diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,26572 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9956467969329705,
+  "eval_steps": 500,
+  "global_step": 3789,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0007918048201118424,
+      "grad_norm": 3.249665515279578,
+      "learning_rate": 2e-07,
+      "loss": 0.7945,
+      "step": 1
+    },
+    {
+      "epoch": 0.0015836096402236848,
+      "grad_norm": 2.8142684051052407,
+      "learning_rate": 4e-07,
+      "loss": 0.7499,
+      "step": 2
+    },
+    {
+      "epoch": 0.0023754144603355272,
+      "grad_norm": 3.1595653950203344,
+      "learning_rate": 6e-07,
+      "loss": 0.7657,
+      "step": 3
+    },
+    {
+      "epoch": 0.0031672192804473695,
+      "grad_norm": 3.077093768185348,
+      "learning_rate": 8e-07,
+      "loss": 0.7546,
+      "step": 4
+    },
+    {
+      "epoch": 0.003959024100559212,
+      "grad_norm": 2.991003227075916,
+      "learning_rate": 1e-06,
+      "loss": 0.7488,
+      "step": 5
+    },
+    {
+      "epoch": 0.0047508289206710545,
+      "grad_norm": 2.664419096676507,
+      "learning_rate": 1.2e-06,
+      "loss": 0.735,
+      "step": 6
+    },
+    {
+      "epoch": 0.005542633740782897,
+      "grad_norm": 2.447554072933163,
+      "learning_rate": 1.4e-06,
+      "loss": 0.7432,
+      "step": 7
+    },
+    {
+      "epoch": 0.006334438560894739,
+      "grad_norm": 1.8641292483217404,
+      "learning_rate": 1.6e-06,
+      "loss": 0.6963,
+      "step": 8
+    },
+    {
+      "epoch": 0.007126243381006582,
+      "grad_norm": 1.9602828612613827,
+      "learning_rate": 1.8e-06,
+      "loss": 0.7213,
+      "step": 9
+    },
+    {
+      "epoch": 0.007918048201118424,
+      "grad_norm": 1.839376447147316,
+      "learning_rate": 2e-06,
+      "loss": 0.7203,
+      "step": 10
+    },
+    {
+      "epoch": 0.008709853021230268,
+      "grad_norm": 1.0894689354185774,
+      "learning_rate": 1.999999654445987e-06,
+      "loss": 0.6868,
+      "step": 11
+    },
+    {
+      "epoch": 0.009501657841342109,
+      "grad_norm": 1.2516859974601064,
+      "learning_rate": 1.9999986177841868e-06,
+      "loss": 0.6474,
+      "step": 12
+    },
+    {
+      "epoch": 0.010293462661453952,
+      "grad_norm": 1.2990716624476766,
+      "learning_rate": 1.999996890015316e-06,
+      "loss": 0.6559,
+      "step": 13
+    },
+    {
+      "epoch": 0.011085267481565794,
+      "grad_norm": 1.271434335142783,
+      "learning_rate": 1.9999944711405684e-06,
+      "loss": 0.6491,
+      "step": 14
+    },
+    {
+      "epoch": 0.011877072301677637,
+      "grad_norm": 2.0436825661525506,
+      "learning_rate": 1.9999913611616162e-06,
+      "loss": 0.6452,
+      "step": 15
+    },
+    {
+      "epoch": 0.012668877121789478,
+      "grad_norm": 2.788044827948402,
+      "learning_rate": 1.999987560080608e-06,
+      "loss": 0.6572,
+      "step": 16
+    },
+    {
+      "epoch": 0.013460681941901321,
+      "grad_norm": 1.817419771671645,
+      "learning_rate": 1.9999830679001716e-06,
+      "loss": 0.6691,
+      "step": 17
+    },
+    {
+      "epoch": 0.014252486762013164,
+      "grad_norm": 1.06301349202853,
+      "learning_rate": 1.999977884623411e-06,
+      "loss": 0.6521,
+      "step": 18
+    },
+    {
+      "epoch": 0.015044291582125006,
+      "grad_norm": 0.7884086864832666,
+      "learning_rate": 1.999972010253908e-06,
+      "loss": 0.6435,
+      "step": 19
+    },
+    {
+      "epoch": 0.01583609640223685,
+      "grad_norm": 0.5927933324880149,
+      "learning_rate": 1.999965444795724e-06,
+      "loss": 0.6378,
+      "step": 20
+    },
+    {
+      "epoch": 0.016627901222348692,
+      "grad_norm": 0.515330764428853,
+      "learning_rate": 1.9999581882533944e-06,
+      "loss": 0.6451,
+      "step": 21
+    },
+    {
+      "epoch": 0.017419706042460535,
+      "grad_norm": 0.46993043798293804,
+      "learning_rate": 1.9999502406319355e-06,
+      "loss": 0.6276,
+      "step": 22
+    },
+    {
+      "epoch": 0.018211510862572375,
+      "grad_norm": 0.4663536484014579,
+      "learning_rate": 1.99994160193684e-06,
+      "loss": 0.644,
+      "step": 23
+    },
+    {
+      "epoch": 0.019003315682684218,
+      "grad_norm": 0.49060116408582094,
+      "learning_rate": 1.999932272174078e-06,
+      "loss": 0.655,
+      "step": 24
+    },
+    {
+      "epoch": 0.01979512050279606,
+      "grad_norm": 0.5480942676252971,
+      "learning_rate": 1.999922251350097e-06,
+      "loss": 0.6515,
+      "step": 25
+    },
+    {
+      "epoch": 0.020586925322907904,
+      "grad_norm": 0.5771916166813391,
+      "learning_rate": 1.999911539471823e-06,
+      "loss": 0.6415,
+      "step": 26
+    },
+    {
+      "epoch": 0.021378730143019744,
+      "grad_norm": 0.5839231264275982,
+      "learning_rate": 1.9999001365466587e-06,
+      "loss": 0.635,
+      "step": 27
+    },
+    {
+      "epoch": 0.022170534963131587,
+      "grad_norm": 0.5301968144588673,
+      "learning_rate": 1.999888042582485e-06,
+      "loss": 0.6375,
+      "step": 28
+    },
+    {
+      "epoch": 0.02296233978324343,
+      "grad_norm": 0.4734334925176316,
+      "learning_rate": 1.9998752575876592e-06,
+      "loss": 0.6184,
+      "step": 29
+    },
+    {
+      "epoch": 0.023754144603355273,
+      "grad_norm": 0.4249379146562315,
+      "learning_rate": 1.9998617815710186e-06,
+      "loss": 0.6452,
+      "step": 30
+    },
+    {
+      "epoch": 0.024545949423467116,
+      "grad_norm": 0.3662617306941407,
+      "learning_rate": 1.999847614541876e-06,
+      "loss": 0.6193,
+      "step": 31
+    },
+    {
+      "epoch": 0.025337754243578956,
+      "grad_norm": 0.33731127892211377,
+      "learning_rate": 1.999832756510022e-06,
+      "loss": 0.6286,
+      "step": 32
+    },
+    {
+      "epoch": 0.0261295590636908,
+      "grad_norm": 0.3663161838243929,
+      "learning_rate": 1.9998172074857257e-06,
+      "loss": 0.6547,
+      "step": 33
+    },
+    {
+      "epoch": 0.026921363883802642,
+      "grad_norm": 0.381291078224402,
+      "learning_rate": 1.999800967479732e-06,
+      "loss": 0.6297,
+      "step": 34
+    },
+    {
+      "epoch": 0.027713168703914486,
+      "grad_norm": 0.3839788122616414,
+      "learning_rate": 1.999784036503266e-06,
+      "loss": 0.6203,
+      "step": 35
+    },
+    {
+      "epoch": 0.02850497352402633,
+      "grad_norm": 0.38298696253878306,
+      "learning_rate": 1.999766414568028e-06,
+      "loss": 0.6167,
+      "step": 36
+    },
+    {
+      "epoch": 0.02929677834413817,
+      "grad_norm": 0.3607911449224398,
+      "learning_rate": 1.999748101686197e-06,
+      "loss": 0.6288,
+      "step": 37
+    },
+    {
+      "epoch": 0.03008858316425001,
+      "grad_norm": 0.3384587080481325,
+      "learning_rate": 1.999729097870429e-06,
+      "loss": 0.6373,
+      "step": 38
+    },
+    {
+      "epoch": 0.030880387984361855,
+      "grad_norm": 0.3437018408721899,
+      "learning_rate": 1.999709403133857e-06,
+      "loss": 0.6158,
+      "step": 39
+    },
+    {
+      "epoch": 0.0316721928044737,
+      "grad_norm": 0.35323126492777207,
+      "learning_rate": 1.9996890174900932e-06,
+      "loss": 0.616,
+      "step": 40
+    },
+    {
+      "epoch": 0.03246399762458554,
+      "grad_norm": 0.31554623480143845,
+      "learning_rate": 1.9996679409532266e-06,
+      "loss": 0.6282,
+      "step": 41
+    },
+    {
+      "epoch": 0.033255802444697384,
+      "grad_norm": 0.32553225221446214,
+      "learning_rate": 1.999646173537822e-06,
+      "loss": 0.6247,
+      "step": 42
+    },
+    {
+      "epoch": 0.034047607264809224,
+      "grad_norm": 0.2931471917183078,
+      "learning_rate": 1.9996237152589236e-06,
+      "loss": 0.598,
+      "step": 43
+    },
+    {
+      "epoch": 0.03483941208492107,
+      "grad_norm": 0.29857347585860283,
+      "learning_rate": 1.999600566132053e-06,
+      "loss": 0.6107,
+      "step": 44
+    },
+    {
+      "epoch": 0.03563121690503291,
+      "grad_norm": 0.2766259754867876,
+      "learning_rate": 1.9995767261732084e-06,
+      "loss": 0.6096,
+      "step": 45
+    },
+    {
+      "epoch": 0.03642302172514475,
+      "grad_norm": 0.2925641957500468,
+      "learning_rate": 1.9995521953988655e-06,
+      "loss": 0.6124,
+      "step": 46
+    },
+    {
+      "epoch": 0.037214826545256596,
+      "grad_norm": 0.27289093170244716,
+      "learning_rate": 1.9995269738259782e-06,
+      "loss": 0.6245,
+      "step": 47
+    },
+    {
+      "epoch": 0.038006631365368436,
+      "grad_norm": 0.2758276253364378,
+      "learning_rate": 1.9995010614719774e-06,
+      "loss": 0.6197,
+      "step": 48
+    },
+    {
+      "epoch": 0.03879843618548028,
+      "grad_norm": 0.264869674303575,
+      "learning_rate": 1.9994744583547704e-06,
+      "loss": 0.622,
+      "step": 49
+    },
+    {
+      "epoch": 0.03959024100559212,
+      "grad_norm": 0.26959516153393015,
+      "learning_rate": 1.999447164492744e-06,
+      "loss": 0.6089,
+      "step": 50
+    },
+    {
+      "epoch": 0.04038204582570396,
+      "grad_norm": 0.2585070187575266,
+      "learning_rate": 1.99941917990476e-06,
+      "loss": 0.6176,
+      "step": 51
+    },
+    {
+      "epoch": 0.04117385064581581,
+      "grad_norm": 0.2535853429181986,
+      "learning_rate": 1.9993905046101603e-06,
+      "loss": 0.5881,
+      "step": 52
+    },
+    {
+      "epoch": 0.04196565546592765,
+      "grad_norm": 0.25964496610236193,
+      "learning_rate": 1.999361138628761e-06,
+      "loss": 0.629,
+      "step": 53
+    },
+    {
+      "epoch": 0.04275746028603949,
+      "grad_norm": 0.3438764551057683,
+      "learning_rate": 1.9993310819808587e-06,
+      "loss": 0.6178,
+      "step": 54
+    },
+    {
+      "epoch": 0.043549265106151334,
+      "grad_norm": 0.24882642114908518,
+      "learning_rate": 1.9993003346872247e-06,
+      "loss": 0.6375,
+      "step": 55
+    },
+    {
+      "epoch": 0.044341069926263174,
+      "grad_norm": 0.2651871648532402,
+      "learning_rate": 1.9992688967691088e-06,
+      "loss": 0.624,
+      "step": 56
+    },
+    {
+      "epoch": 0.04513287474637502,
+      "grad_norm": 0.25284940502692427,
+      "learning_rate": 1.999236768248239e-06,
+      "loss": 0.6125,
+      "step": 57
+    },
+    {
+      "epoch": 0.04592467956648686,
+      "grad_norm": 0.26229841543170196,
+      "learning_rate": 1.999203949146818e-06,
+      "loss": 0.618,
+      "step": 58
+    },
+    {
+      "epoch": 0.0467164843865987,
+      "grad_norm": 0.24920327663759706,
+      "learning_rate": 1.9991704394875287e-06,
+      "loss": 0.6224,
+      "step": 59
+    },
+    {
+      "epoch": 0.04750828920671055,
+      "grad_norm": 0.24646823708084495,
+      "learning_rate": 1.999136239293529e-06,
+      "loss": 0.6083,
+      "step": 60
+    },
+    {
+      "epoch": 0.048300094026822386,
+      "grad_norm": 0.2560335398854992,
+      "learning_rate": 1.9991013485884557e-06,
+      "loss": 0.6122,
+      "step": 61
+    },
+    {
+      "epoch": 0.04909189884693423,
+      "grad_norm": 0.24690995915149744,
+      "learning_rate": 1.9990657673964212e-06,
+      "loss": 0.606,
+      "step": 62
+    },
+    {
+      "epoch": 0.04988370366704607,
+      "grad_norm": 0.2544634101312482,
+      "learning_rate": 1.9990294957420167e-06,
+      "loss": 0.6217,
+      "step": 63
+    },
+    {
+      "epoch": 0.05067550848715791,
+      "grad_norm": 0.25189634044211,
+      "learning_rate": 1.9989925336503093e-06,
+      "loss": 0.6084,
+      "step": 64
+    },
+    {
+      "epoch": 0.05146731330726976,
+      "grad_norm": 0.24802884462346636,
+      "learning_rate": 1.998954881146844e-06,
+      "loss": 0.6237,
+      "step": 65
+    },
+    {
+      "epoch": 0.0522591181273816,
+      "grad_norm": 0.2490473534466852,
+      "learning_rate": 1.998916538257643e-06,
+      "loss": 0.6092,
+      "step": 66
+    },
+    {
+      "epoch": 0.053050922947493445,
+      "grad_norm": 0.24744554617386624,
+      "learning_rate": 1.9988775050092048e-06,
+      "loss": 0.6209,
+      "step": 67
+    },
+    {
+      "epoch": 0.053842727767605285,
+      "grad_norm": 0.24988780685989925,
+      "learning_rate": 1.9988377814285063e-06,
+      "loss": 0.6279,
+      "step": 68
+    },
+    {
+      "epoch": 0.054634532587717125,
+      "grad_norm": 0.24054326057001302,
+      "learning_rate": 1.998797367543e-06,
+      "loss": 0.6181,
+      "step": 69
+    },
+    {
+      "epoch": 0.05542633740782897,
+      "grad_norm": 0.24339247767164976,
+      "learning_rate": 1.998756263380617e-06,
+      "loss": 0.6164,
+      "step": 70
+    },
+    {
+      "epoch": 0.05621814222794081,
+      "grad_norm": 0.25210836407538834,
+      "learning_rate": 1.998714468969764e-06,
+      "loss": 0.6249,
+      "step": 71
+    },
+    {
+      "epoch": 0.05700994704805266,
+      "grad_norm": 0.25071208596853567,
+      "learning_rate": 1.9986719843393265e-06,
+      "loss": 0.6338,
+      "step": 72
+    },
+    {
+      "epoch": 0.0578017518681645,
+      "grad_norm": 0.23776914423319057,
+      "learning_rate": 1.998628809518665e-06,
+      "loss": 0.6033,
+      "step": 73
+    },
+    {
+      "epoch": 0.05859355668827634,
+      "grad_norm": 0.2453910024467796,
+      "learning_rate": 1.998584944537618e-06,
+      "loss": 0.6073,
+      "step": 74
+    },
+    {
+      "epoch": 0.05938536150838818,
+      "grad_norm": 0.24394938336098929,
+      "learning_rate": 1.9985403894265017e-06,
+      "loss": 0.6223,
+      "step": 75
+    },
+    {
+      "epoch": 0.06017716632850002,
+      "grad_norm": 0.2505817471928499,
+      "learning_rate": 1.998495144216108e-06,
+      "loss": 0.6102,
+      "step": 76
+    },
+    {
+      "epoch": 0.06096897114861187,
+      "grad_norm": 0.2437926189291882,
+      "learning_rate": 1.9984492089377057e-06,
+      "loss": 0.6023,
+      "step": 77
+    },
+    {
+      "epoch": 0.06176077596872371,
+      "grad_norm": 0.24885762929340935,
+      "learning_rate": 1.998402583623042e-06,
+      "loss": 0.6075,
+      "step": 78
+    },
+    {
+      "epoch": 0.06255258078883555,
+      "grad_norm": 0.24551220308165528,
+      "learning_rate": 1.9983552683043395e-06,
+      "loss": 0.5987,
+      "step": 79
+    },
+    {
+      "epoch": 0.0633443856089474,
+      "grad_norm": 0.268359520348399,
+      "learning_rate": 1.998307263014298e-06,
+      "loss": 0.6105,
+      "step": 80
+    },
+    {
+      "epoch": 0.06413619042905924,
+      "grad_norm": 0.24925244015298842,
+      "learning_rate": 1.9982585677860954e-06,
+      "loss": 0.6138,
+      "step": 81
+    },
+    {
+      "epoch": 0.06492799524917107,
+      "grad_norm": 0.24162772926660567,
+      "learning_rate": 1.998209182653384e-06,
+      "loss": 0.5903,
+      "step": 82
+    },
+    {
+      "epoch": 0.06571980006928292,
+      "grad_norm": 0.24781516071224408,
+      "learning_rate": 1.998159107650295e-06,
+      "loss": 0.6212,
+      "step": 83
+    },
+    {
+      "epoch": 0.06651160488939477,
+      "grad_norm": 0.24192761327310594,
+      "learning_rate": 1.9981083428114355e-06,
+      "loss": 0.6137,
+      "step": 84
+    },
+    {
+      "epoch": 0.0673034097095066,
+      "grad_norm": 0.24717391550896387,
+      "learning_rate": 1.99805688817189e-06,
+      "loss": 0.6203,
+      "step": 85
+    },
+    {
+      "epoch": 0.06809521452961845,
+      "grad_norm": 0.24088249501234663,
+      "learning_rate": 1.998004743767218e-06,
+      "loss": 0.6208,
+      "step": 86
+    },
+    {
+      "epoch": 0.0688870193497303,
+      "grad_norm": 0.23821640997792096,
+      "learning_rate": 1.997951909633458e-06,
+      "loss": 0.6302,
+      "step": 87
+    },
+    {
+      "epoch": 0.06967882416984214,
+      "grad_norm": 0.24287262149822414,
+      "learning_rate": 1.9978983858071236e-06,
+      "loss": 0.6116,
+      "step": 88
+    },
+    {
+      "epoch": 0.07047062898995397,
+      "grad_norm": 0.24798287495108298,
+      "learning_rate": 1.9978441723252055e-06,
+      "loss": 0.6125,
+      "step": 89
+    },
+    {
+      "epoch": 0.07126243381006582,
+      "grad_norm": 0.2392397494962519,
+      "learning_rate": 1.9977892692251715e-06,
+      "loss": 0.591,
+      "step": 90
+    },
+    {
+      "epoch": 0.07205423863017767,
+      "grad_norm": 0.24837547105981,
+      "learning_rate": 1.997733676544965e-06,
+      "loss": 0.6278,
+      "step": 91
+    },
+    {
+      "epoch": 0.0728460434502895,
+      "grad_norm": 0.2409248642894013,
+      "learning_rate": 1.997677394323007e-06,
+      "loss": 0.6144,
+      "step": 92
+    },
+    {
+      "epoch": 0.07363784827040135,
+      "grad_norm": 0.24434129583738856,
+      "learning_rate": 1.997620422598195e-06,
+      "loss": 0.6261,
+      "step": 93
+    },
+    {
+      "epoch": 0.07442965309051319,
+      "grad_norm": 0.23785260817816306,
+      "learning_rate": 1.9975627614099018e-06,
+      "loss": 0.6137,
+      "step": 94
+    },
+    {
+      "epoch": 0.07522145791062503,
+      "grad_norm": 0.2582438980418932,
+      "learning_rate": 1.997504410797977e-06,
+      "loss": 0.6155,
+      "step": 95
+    },
+    {
+      "epoch": 0.07601326273073687,
+      "grad_norm": 0.23782100531057535,
+      "learning_rate": 1.997445370802749e-06,
+      "loss": 0.5945,
+      "step": 96
+    },
+    {
+      "epoch": 0.07680506755084872,
+      "grad_norm": 0.24354341473147686,
+      "learning_rate": 1.997385641465019e-06,
+      "loss": 0.605,
+      "step": 97
+    },
+    {
+      "epoch": 0.07759687237096057,
+      "grad_norm": 0.24085340933783522,
+      "learning_rate": 1.9973252228260678e-06,
+      "loss": 0.6023,
+      "step": 98
+    },
+    {
+      "epoch": 0.0783886771910724,
+      "grad_norm": 0.24648161069179864,
+      "learning_rate": 1.9972641149276506e-06,
+      "loss": 0.616,
+      "step": 99
+    },
+    {
+      "epoch": 0.07918048201118424,
+      "grad_norm": 0.2544677163439913,
+      "learning_rate": 1.997202317811999e-06,
+      "loss": 0.6067,
+      "step": 100
+    },
+    {
+      "epoch": 0.07997228683129609,
+      "grad_norm": 0.24490827872109347,
+      "learning_rate": 1.9971398315218223e-06,
+      "loss": 0.6237,
+      "step": 101
+    },
+    {
+      "epoch": 0.08076409165140792,
+      "grad_norm": 0.25703203869900015,
+      "learning_rate": 1.997076656100305e-06,
+      "loss": 0.6185,
+      "step": 102
+    },
+    {
+      "epoch": 0.08155589647151977,
+      "grad_norm": 0.23887272337607254,
+      "learning_rate": 1.9970127915911083e-06,
+      "loss": 0.6287,
+      "step": 103
+    },
+    {
+      "epoch": 0.08234770129163162,
+      "grad_norm": 0.2347393500247831,
+      "learning_rate": 1.996948238038369e-06,
+      "loss": 0.6212,
+      "step": 104
+    },
+    {
+      "epoch": 0.08313950611174345,
+      "grad_norm": 0.23400474153598683,
+      "learning_rate": 1.9968829954867015e-06,
+      "loss": 0.6063,
+      "step": 105
+    },
+    {
+      "epoch": 0.0839313109318553,
+      "grad_norm": 0.25444492337175667,
+      "learning_rate": 1.9968170639811943e-06,
+      "loss": 0.6015,
+      "step": 106
+    },
+    {
+      "epoch": 0.08472311575196714,
+      "grad_norm": 0.24767706489511598,
+      "learning_rate": 1.9967504435674142e-06,
+      "loss": 0.6151,
+      "step": 107
+    },
+    {
+      "epoch": 0.08551492057207898,
+      "grad_norm": 0.24857739706161022,
+      "learning_rate": 1.9966831342914025e-06,
+      "loss": 0.6068,
+      "step": 108
+    },
+    {
+      "epoch": 0.08630672539219082,
+      "grad_norm": 0.24013957818221426,
+      "learning_rate": 1.9966151361996774e-06,
+      "loss": 0.6221,
+      "step": 109
+    },
+    {
+      "epoch": 0.08709853021230267,
+      "grad_norm": 0.2552787663397809,
+      "learning_rate": 1.9965464493392326e-06,
+      "loss": 0.6185,
+      "step": 110
+    },
+    {
+      "epoch": 0.08789033503241452,
+      "grad_norm": 0.24393863343993738,
+      "learning_rate": 1.996477073757539e-06,
+      "loss": 0.6106,
+      "step": 111
+    },
+    {
+      "epoch": 0.08868213985252635,
+      "grad_norm": 0.24327566997291633,
+      "learning_rate": 1.9964070095025412e-06,
+      "loss": 0.6158,
+      "step": 112
+    },
+    {
+      "epoch": 0.0894739446726382,
+      "grad_norm": 0.2460212889025927,
+      "learning_rate": 1.9963362566226625e-06,
+      "loss": 0.5918,
+      "step": 113
+    },
+    {
+      "epoch": 0.09026574949275004,
+      "grad_norm": 0.24668386912004803,
+      "learning_rate": 1.9962648151668e-06,
+      "loss": 0.6068,
+      "step": 114
+    },
+    {
+      "epoch": 0.09105755431286187,
+      "grad_norm": 0.24064611635390737,
+      "learning_rate": 1.9961926851843284e-06,
+      "loss": 0.6109,
+      "step": 115
+    },
+    {
+      "epoch": 0.09184935913297372,
+      "grad_norm": 0.2484601429962744,
+      "learning_rate": 1.9961198667250963e-06,
+      "loss": 0.605,
+      "step": 116
+    },
+    {
+      "epoch": 0.09264116395308557,
+      "grad_norm": 0.24394465737688995,
+      "learning_rate": 1.9960463598394293e-06,
+      "loss": 0.6286,
+      "step": 117
+    },
+    {
+      "epoch": 0.0934329687731974,
+      "grad_norm": 0.24679080532649691,
+      "learning_rate": 1.995972164578129e-06,
+      "loss": 0.5859,
+      "step": 118
+    },
+    {
+      "epoch": 0.09422477359330925,
+      "grad_norm": 0.25371887380609054,
+      "learning_rate": 1.995897280992472e-06,
+      "loss": 0.6024,
+      "step": 119
+    },
+    {
+      "epoch": 0.0950165784134211,
+      "grad_norm": 0.24806492536949867,
+      "learning_rate": 1.995821709134211e-06,
+      "loss": 0.6174,
+      "step": 120
+    },
+    {
+      "epoch": 0.09580838323353294,
+      "grad_norm": 0.2565291335234416,
+      "learning_rate": 1.9957454490555745e-06,
+      "loss": 0.6117,
+      "step": 121
+    },
+    {
+      "epoch": 0.09660018805364477,
+      "grad_norm": 0.2510354896290779,
+      "learning_rate": 1.995668500809266e-06,
+      "loss": 0.6083,
+      "step": 122
+    },
+    {
+      "epoch": 0.09739199287375662,
+      "grad_norm": 0.25122352500446965,
+      "learning_rate": 1.995590864448466e-06,
+      "loss": 0.6171,
+      "step": 123
+    },
+    {
+      "epoch": 0.09818379769386847,
+      "grad_norm": 0.2454319191209234,
+      "learning_rate": 1.995512540026829e-06,
+      "loss": 0.6081,
+      "step": 124
+    },
+    {
+      "epoch": 0.0989756025139803,
+      "grad_norm": 0.2520995178709679,
+      "learning_rate": 1.995433527598485e-06,
+      "loss": 0.6001,
+      "step": 125
+    },
+    {
+      "epoch": 0.09976740733409215,
+      "grad_norm": 0.24895886123859298,
+      "learning_rate": 1.995353827218041e-06,
+      "loss": 0.6229,
+      "step": 126
+    },
+    {
+      "epoch": 0.10055921215420399,
+      "grad_norm": 0.2594511545591342,
+      "learning_rate": 1.9952734389405786e-06,
+      "loss": 0.6069,
+      "step": 127
+    },
+    {
+      "epoch": 0.10135101697431582,
+      "grad_norm": 0.24667835691941334,
+      "learning_rate": 1.995192362821654e-06,
+      "loss": 0.6275,
+      "step": 128
+    },
+    {
+      "epoch": 0.10214282179442767,
+      "grad_norm": 0.24318685595006123,
+      "learning_rate": 1.9951105989173006e-06,
+      "loss": 0.6118,
+      "step": 129
+    },
+    {
+      "epoch": 0.10293462661453952,
+      "grad_norm": 0.24353911392852584,
+      "learning_rate": 1.9950281472840255e-06,
+      "loss": 0.6099,
+      "step": 130
+    },
+    {
+      "epoch": 0.10372643143465136,
+      "grad_norm": 0.24328214159825529,
+      "learning_rate": 1.9949450079788114e-06,
+      "loss": 0.595,
+      "step": 131
+    },
+    {
+      "epoch": 0.1045182362547632,
+      "grad_norm": 0.233892898891033,
+      "learning_rate": 1.9948611810591172e-06,
+      "loss": 0.6148,
+      "step": 132
+    },
+    {
+      "epoch": 0.10531004107487504,
+      "grad_norm": 0.23766448029852974,
+      "learning_rate": 1.9947766665828758e-06,
+      "loss": 0.5807,
+      "step": 133
+    },
+    {
+      "epoch": 0.10610184589498689,
+      "grad_norm": 0.2379454146592536,
+      "learning_rate": 1.9946914646084963e-06,
+      "loss": 0.6267,
+      "step": 134
+    },
+    {
+      "epoch": 0.10689365071509872,
+      "grad_norm": 0.24945201019879407,
+      "learning_rate": 1.9946055751948624e-06,
+      "loss": 0.6259,
+      "step": 135
+    },
+    {
+      "epoch": 0.10768545553521057,
+      "grad_norm": 0.24891292482229235,
+      "learning_rate": 1.9945189984013325e-06,
+      "loss": 0.6128,
+      "step": 136
+    },
+    {
+      "epoch": 0.10847726035532242,
+      "grad_norm": 0.2525902603363248,
+      "learning_rate": 1.994431734287741e-06,
+      "loss": 0.5994,
+      "step": 137
+    },
+    {
+      "epoch": 0.10926906517543425,
+      "grad_norm": 0.25443424584665786,
+      "learning_rate": 1.994343782914396e-06,
+      "loss": 0.5965,
+      "step": 138
+    },
+    {
+      "epoch": 0.1100608699955461,
+      "grad_norm": 0.24778523468262087,
+      "learning_rate": 1.994255144342083e-06,
+      "loss": 0.6066,
+      "step": 139
+    },
+    {
+      "epoch": 0.11085267481565794,
+      "grad_norm": 0.24549555159772826,
+      "learning_rate": 1.9941658186320596e-06,
+      "loss": 0.6109,
+      "step": 140
+    },
+    {
+      "epoch": 0.11164447963576979,
+      "grad_norm": 0.2459614675472423,
+      "learning_rate": 1.9940758058460596e-06,
+      "loss": 0.6246,
+      "step": 141
+    },
+    {
+      "epoch": 0.11243628445588162,
+      "grad_norm": 0.25382226746913006,
+      "learning_rate": 1.993985106046292e-06,
+      "loss": 0.6281,
+      "step": 142
+    },
+    {
+      "epoch": 0.11322808927599347,
+      "grad_norm": 0.23892895275893525,
+      "learning_rate": 1.9938937192954394e-06,
+      "loss": 0.6151,
+      "step": 143
+    },
+    {
+      "epoch": 0.11401989409610531,
+      "grad_norm": 0.24606184796512073,
+      "learning_rate": 1.993801645656661e-06,
+      "loss": 0.6215,
+      "step": 144
+    },
+    {
+      "epoch": 0.11481169891621715,
+      "grad_norm": 0.2421058351251952,
+      "learning_rate": 1.993708885193589e-06,
+      "loss": 0.6148,
+      "step": 145
+    },
+    {
+      "epoch": 0.115603503736329,
+      "grad_norm": 0.24631536565411302,
+      "learning_rate": 1.9936154379703307e-06,
+      "loss": 0.6033,
+      "step": 146
+    },
+    {
+      "epoch": 0.11639530855644084,
+      "grad_norm": 0.24147837800339966,
+      "learning_rate": 1.9935213040514685e-06,
+      "loss": 0.6093,
+      "step": 147
+    },
+    {
+      "epoch": 0.11718711337655267,
+      "grad_norm": 0.24339817339043665,
+      "learning_rate": 1.9934264835020596e-06,
+      "loss": 0.6259,
+      "step": 148
+    },
+    {
+      "epoch": 0.11797891819666452,
+      "grad_norm": 0.23842551166815323,
+      "learning_rate": 1.9933309763876342e-06,
+      "loss": 0.6171,
+      "step": 149
+    },
+    {
+      "epoch": 0.11877072301677637,
+      "grad_norm": 0.24104605193929682,
+      "learning_rate": 1.9932347827741987e-06,
+      "loss": 0.6078,
+      "step": 150
+    },
+    {
+      "epoch": 0.11956252783688821,
+      "grad_norm": 0.24688252589023799,
+      "learning_rate": 1.9931379027282333e-06,
+      "loss": 0.6215,
+      "step": 151
+    },
+    {
+      "epoch": 0.12035433265700005,
+      "grad_norm": 0.25022709961910333,
+      "learning_rate": 1.9930403363166925e-06,
+      "loss": 0.6097,
+      "step": 152
+    },
+    {
+      "epoch": 0.12114613747711189,
+      "grad_norm": 0.2525497601786903,
+      "learning_rate": 1.992942083607005e-06,
+      "loss": 0.5999,
+      "step": 153
+    },
+    {
+      "epoch": 0.12193794229722374,
+      "grad_norm": 0.24582397817659582,
+      "learning_rate": 1.992843144667074e-06,
+      "loss": 0.5911,
+      "step": 154
+    },
+    {
+      "epoch": 0.12272974711733557,
+      "grad_norm": 0.24517082411379587,
+      "learning_rate": 1.9927435195652776e-06,
+      "loss": 0.6055,
+      "step": 155
+    },
+    {
+      "epoch": 0.12352155193744742,
+      "grad_norm": 0.24686986627179539,
+      "learning_rate": 1.9926432083704666e-06,
+      "loss": 0.6133,
+      "step": 156
+    },
+    {
+      "epoch": 0.12431335675755927,
+      "grad_norm": 0.24124754454063155,
+      "learning_rate": 1.9925422111519678e-06,
+      "loss": 0.6078,
+      "step": 157
+    },
+    {
+      "epoch": 0.1251051615776711,
+      "grad_norm": 0.25064281296316676,
+      "learning_rate": 1.992440527979581e-06,
+      "loss": 0.6087,
+      "step": 158
+    },
+    {
+      "epoch": 0.12589696639778294,
+      "grad_norm": 0.2422033815243161,
+      "learning_rate": 1.9923381589235794e-06,
+      "loss": 0.5948,
+      "step": 159
+    },
+    {
+      "epoch": 0.1266887712178948,
+      "grad_norm": 0.25651784949814377,
+      "learning_rate": 1.992235104054712e-06,
+      "loss": 0.6184,
+      "step": 160
+    },
+    {
+      "epoch": 0.12748057603800664,
+      "grad_norm": 0.24005721640899194,
+      "learning_rate": 1.9921313634442005e-06,
+      "loss": 0.6178,
+      "step": 161
+    },
+    {
+      "epoch": 0.12827238085811848,
+      "grad_norm": 0.24352522262928494,
+      "learning_rate": 1.9920269371637407e-06,
+      "loss": 0.6085,
+      "step": 162
+    },
+    {
+      "epoch": 0.1290641856782303,
+      "grad_norm": 0.24623314726006892,
+      "learning_rate": 1.991921825285503e-06,
+      "loss": 0.6081,
+      "step": 163
+    },
+    {
+      "epoch": 0.12985599049834215,
+      "grad_norm": 0.2553394676667384,
+      "learning_rate": 1.99181602788213e-06,
+      "loss": 0.6123,
+      "step": 164
+    },
+    {
+      "epoch": 0.130647795318454,
+      "grad_norm": 0.25484888153943747,
+      "learning_rate": 1.9917095450267405e-06,
+      "loss": 0.6053,
+      "step": 165
+    },
+    {
+      "epoch": 0.13143960013856584,
+      "grad_norm": 0.2451398623016501,
+      "learning_rate": 1.991602376792925e-06,
+      "loss": 0.5935,
+      "step": 166
+    },
+    {
+      "epoch": 0.1322314049586777,
+      "grad_norm": 0.23374157148454733,
+      "learning_rate": 1.9914945232547486e-06,
+      "loss": 0.5949,
+      "step": 167
+    },
+    {
+      "epoch": 0.13302320977878954,
+      "grad_norm": 0.2423950980430432,
+      "learning_rate": 1.9913859844867493e-06,
+      "loss": 0.6051,
+      "step": 168
+    },
+    {
+      "epoch": 0.13381501459890138,
+      "grad_norm": 0.2400913194454132,
+      "learning_rate": 1.9912767605639387e-06,
+      "loss": 0.6243,
+      "step": 169
+    },
+    {
+      "epoch": 0.1346068194190132,
+      "grad_norm": 0.24404491950360782,
+      "learning_rate": 1.9911668515618033e-06,
+      "loss": 0.5875,
+      "step": 170
+    },
+    {
+      "epoch": 0.13539862423912505,
+      "grad_norm": 0.24678318176788785,
+      "learning_rate": 1.991056257556302e-06,
+      "loss": 0.6096,
+      "step": 171
+    },
+    {
+      "epoch": 0.1361904290592369,
+      "grad_norm": 0.25735257644134474,
+      "learning_rate": 1.9909449786238666e-06,
+      "loss": 0.6201,
+      "step": 172
+    },
+    {
+      "epoch": 0.13698223387934874,
+      "grad_norm": 0.2519557865307485,
+      "learning_rate": 1.990833014841403e-06,
+      "loss": 0.6067,
+      "step": 173
+    },
+    {
+      "epoch": 0.1377740386994606,
+      "grad_norm": 0.23885770861285516,
+      "learning_rate": 1.9907203662862906e-06,
+      "loss": 0.6019,
+      "step": 174
+    },
+    {
+      "epoch": 0.13856584351957243,
+      "grad_norm": 0.24925712376363354,
+      "learning_rate": 1.990607033036382e-06,
+      "loss": 0.608,
+      "step": 175
+    },
+    {
+      "epoch": 0.13935764833968428,
+      "grad_norm": 0.24712204252168293,
+      "learning_rate": 1.990493015170002e-06,
+      "loss": 0.6057,
+      "step": 176
+    },
+    {
+      "epoch": 0.1401494531597961,
+      "grad_norm": 0.255062892374803,
+      "learning_rate": 1.990378312765949e-06,
+      "loss": 0.6024,
+      "step": 177
+    },
+    {
+      "epoch": 0.14094125797990795,
+      "grad_norm": 0.2412239226129293,
+      "learning_rate": 1.990262925903496e-06,
+      "loss": 0.592,
+      "step": 178
+    },
+    {
+      "epoch": 0.1417330628000198,
+      "grad_norm": 0.24100115776441883,
+      "learning_rate": 1.9901468546623867e-06,
+      "loss": 0.6161,
+      "step": 179
+    },
+    {
+      "epoch": 0.14252486762013164,
+      "grad_norm": 0.2379116890638869,
+      "learning_rate": 1.9900300991228397e-06,
+      "loss": 0.5902,
+      "step": 180
+    },
+    {
+      "epoch": 0.1433166724402435,
+      "grad_norm": 0.2557311003501454,
+      "learning_rate": 1.9899126593655448e-06,
+      "loss": 0.6023,
+      "step": 181
+    },
+    {
+      "epoch": 0.14410847726035533,
+      "grad_norm": 0.2455583575343663,
+      "learning_rate": 1.9897945354716662e-06,
+      "loss": 0.6094,
+      "step": 182
+    },
+    {
+      "epoch": 0.14490028208046715,
+      "grad_norm": 0.24370628898940228,
+      "learning_rate": 1.9896757275228403e-06,
+      "loss": 0.6078,
+      "step": 183
+    },
+    {
+      "epoch": 0.145692086900579,
+      "grad_norm": 0.23759579081298693,
+      "learning_rate": 1.989556235601176e-06,
+      "loss": 0.6039,
+      "step": 184
+    },
+    {
+      "epoch": 0.14648389172069085,
+      "grad_norm": 0.24391920970902445,
+      "learning_rate": 1.9894360597892543e-06,
+      "loss": 0.6078,
+      "step": 185
+    },
+    {
+      "epoch": 0.1472756965408027,
+      "grad_norm": 0.24845711826445283,
+      "learning_rate": 1.989315200170131e-06,
+      "loss": 0.6328,
+      "step": 186
+    },
+    {
+      "epoch": 0.14806750136091454,
+      "grad_norm": 0.2616752365478856,
+      "learning_rate": 1.989193656827333e-06,
+      "loss": 0.6046,
+      "step": 187
+    },
+    {
+      "epoch": 0.14885930618102639,
+      "grad_norm": 0.2446871338572912,
+      "learning_rate": 1.989071429844859e-06,
+      "loss": 0.6061,
+      "step": 188
+    },
+    {
+      "epoch": 0.14965111100113823,
+      "grad_norm": 0.24408589622192395,
+      "learning_rate": 1.988948519307182e-06,
+      "loss": 0.6045,
+      "step": 189
+    },
+    {
+      "epoch": 0.15044291582125005,
+      "grad_norm": 0.23962545305324873,
+      "learning_rate": 1.9888249252992457e-06,
+      "loss": 0.6145,
+      "step": 190
+    },
+    {
+      "epoch": 0.1512347206413619,
+      "grad_norm": 0.25855035028698975,
+      "learning_rate": 1.9887006479064676e-06,
+      "loss": 0.622,
+      "step": 191
+    },
+    {
+      "epoch": 0.15202652546147374,
+      "grad_norm": 0.25096403621981855,
+      "learning_rate": 1.9885756872147367e-06,
+      "loss": 0.599,
+      "step": 192
+    },
+    {
+      "epoch": 0.1528183302815856,
+      "grad_norm": 0.2456345168057193,
+      "learning_rate": 1.9884500433104133e-06,
+      "loss": 0.5971,
+      "step": 193
+    },
+    {
+      "epoch": 0.15361013510169744,
+      "grad_norm": 0.24697511302923733,
+      "learning_rate": 1.988323716280332e-06,
+      "loss": 0.6172,
+      "step": 194
+    },
+    {
+      "epoch": 0.15440193992180928,
+      "grad_norm": 0.2498250649439413,
+      "learning_rate": 1.9881967062117985e-06,
+      "loss": 0.6202,
+      "step": 195
+    },
+    {
+      "epoch": 0.15519374474192113,
+      "grad_norm": 0.2477427405714654,
+      "learning_rate": 1.9880690131925897e-06,
+      "loss": 0.6028,
+      "step": 196
+    },
+    {
+      "epoch": 0.15598554956203295,
+      "grad_norm": 0.25368523649798064,
+      "learning_rate": 1.987940637310956e-06,
+      "loss": 0.6013,
+      "step": 197
+    },
+    {
+      "epoch": 0.1567773543821448,
+      "grad_norm": 0.36000625045653284,
+      "learning_rate": 1.987811578655618e-06,
+      "loss": 0.6118,
+      "step": 198
+    },
+    {
+      "epoch": 0.15756915920225664,
+      "grad_norm": 0.24669657811244383,
+      "learning_rate": 1.9876818373157706e-06,
+      "loss": 0.6099,
+      "step": 199
+    },
+    {
+      "epoch": 0.1583609640223685,
+      "grad_norm": 0.24717856235092012,
+      "learning_rate": 1.987551413381078e-06,
+      "loss": 0.5989,
+      "step": 200
+    },
+    {
+      "epoch": 0.15915276884248034,
+      "grad_norm": 0.24166194958075368,
+      "learning_rate": 1.9874203069416774e-06,
+      "loss": 0.6095,
+      "step": 201
+    },
+    {
+      "epoch": 0.15994457366259218,
+      "grad_norm": 0.2432222315837849,
+      "learning_rate": 1.987288518088178e-06,
+      "loss": 0.5939,
+      "step": 202
+    },
+    {
+      "epoch": 0.160736378482704,
+      "grad_norm": 0.25616990344712176,
+      "learning_rate": 1.9871560469116597e-06,
+      "loss": 0.6319,
+      "step": 203
+    },
+    {
+      "epoch": 0.16152818330281585,
+      "grad_norm": 0.25124893412477034,
+      "learning_rate": 1.9870228935036743e-06,
+      "loss": 0.6104,
+      "step": 204
+    },
+    {
+      "epoch": 0.1623199881229277,
+      "grad_norm": 0.2533422007446896,
+      "learning_rate": 1.9868890579562456e-06,
+      "loss": 0.6187,
+      "step": 205
+    },
+    {
+      "epoch": 0.16311179294303954,
+      "grad_norm": 0.24291758536714064,
+      "learning_rate": 1.9867545403618678e-06,
+      "loss": 0.6145,
+      "step": 206
+    },
+    {
+      "epoch": 0.1639035977631514,
+      "grad_norm": 0.24321672867435246,
+      "learning_rate": 1.9866193408135083e-06,
+      "loss": 0.6037,
+      "step": 207
+    },
+    {
+      "epoch": 0.16469540258326323,
+      "grad_norm": 0.24239392661935416,
+      "learning_rate": 1.986483459404603e-06,
+      "loss": 0.6176,
+      "step": 208
+    },
+    {
+      "epoch": 0.16548720740337508,
+      "grad_norm": 0.2453605563740051,
+      "learning_rate": 1.9863468962290616e-06,
+      "loss": 0.6035,
+      "step": 209
+    },
+    {
+      "epoch": 0.1662790122234869,
+      "grad_norm": 0.25150165507645755,
+      "learning_rate": 1.986209651381264e-06,
+      "loss": 0.5926,
+      "step": 210
+    },
+    {
+      "epoch": 0.16707081704359875,
+      "grad_norm": 0.25363285848077716,
+      "learning_rate": 1.986071724956061e-06,
+      "loss": 0.6122,
+      "step": 211
+    },
+    {
+      "epoch": 0.1678626218637106,
+      "grad_norm": 0.24764125273102747,
+      "learning_rate": 1.9859331170487748e-06,
+      "loss": 0.605,
+      "step": 212
+    },
+    {
+      "epoch": 0.16865442668382244,
+      "grad_norm": 0.24485071327883146,
+      "learning_rate": 1.9857938277551983e-06,
+      "loss": 0.6112,
+      "step": 213
+    },
+    {
+      "epoch": 0.16944623150393429,
+      "grad_norm": 0.24319134264447712,
+      "learning_rate": 1.9856538571715954e-06,
+      "loss": 0.6087,
+      "step": 214
+    },
+    {
+      "epoch": 0.17023803632404613,
+      "grad_norm": 0.24594381812896973,
+      "learning_rate": 1.985513205394701e-06,
+      "loss": 0.5984,
+      "step": 215
+    },
+    {
+      "epoch": 0.17102984114415795,
+      "grad_norm": 0.24682994996910132,
+      "learning_rate": 1.9853718725217207e-06,
+      "loss": 0.607,
+      "step": 216
+    },
+    {
+      "epoch": 0.1718216459642698,
+      "grad_norm": 0.2523031853877198,
+      "learning_rate": 1.9852298586503305e-06,
+      "loss": 0.6154,
+      "step": 217
+    },
+    {
+      "epoch": 0.17261345078438164,
+      "grad_norm": 0.25940404563218156,
+      "learning_rate": 1.985087163878678e-06,
+      "loss": 0.6201,
+      "step": 218
+    },
+    {
+      "epoch": 0.1734052556044935,
+      "grad_norm": 0.24143005184833266,
+      "learning_rate": 1.9849437883053802e-06,
+      "loss": 0.6021,
+      "step": 219
+    },
+    {
+      "epoch": 0.17419706042460534,
+      "grad_norm": 0.24284774715761134,
+      "learning_rate": 1.984799732029525e-06,
+      "loss": 0.6012,
+      "step": 220
+    },
+    {
+      "epoch": 0.17498886524471718,
+      "grad_norm": 0.24534087775707256,
+      "learning_rate": 1.9846549951506712e-06,
+      "loss": 0.6207,
+      "step": 221
+    },
+    {
+      "epoch": 0.17578067006482903,
+      "grad_norm": 0.24717583062758106,
+      "learning_rate": 1.984509577768847e-06,
+      "loss": 0.6202,
+      "step": 222
+    },
+    {
+      "epoch": 0.17657247488494085,
+      "grad_norm": 0.24145918806305072,
+      "learning_rate": 1.984363479984552e-06,
+      "loss": 0.6084,
+      "step": 223
+    },
+    {
+      "epoch": 0.1773642797050527,
+      "grad_norm": 0.24360708384916321,
+      "learning_rate": 1.984216701898756e-06,
+      "loss": 0.6015,
+      "step": 224
+    },
+    {
+      "epoch": 0.17815608452516454,
+      "grad_norm": 0.25269848624865143,
+      "learning_rate": 1.9840692436128975e-06,
+      "loss": 0.595,
+      "step": 225
+    },
+    {
+      "epoch": 0.1789478893452764,
+      "grad_norm": 0.24271430989679857,
+      "learning_rate": 1.9839211052288873e-06,
+      "loss": 0.5995,
+      "step": 226
+    },
+    {
+      "epoch": 0.17973969416538824,
+      "grad_norm": 0.24759560592499152,
+      "learning_rate": 1.983772286849104e-06,
+      "loss": 0.5909,
+      "step": 227
+    },
+    {
+      "epoch": 0.18053149898550008,
+      "grad_norm": 0.25407305995874085,
+      "learning_rate": 1.9836227885763975e-06,
+      "loss": 0.5981,
+      "step": 228
+    },
+    {
+      "epoch": 0.18132330380561193,
+      "grad_norm": 0.24923103654822074,
+      "learning_rate": 1.9834726105140873e-06,
+      "loss": 0.5969,
+      "step": 229
+    },
+    {
+      "epoch": 0.18211510862572375,
+      "grad_norm": 0.2495011332612613,
+      "learning_rate": 1.9833217527659624e-06,
+      "loss": 0.5936,
+      "step": 230
+    },
+    {
+      "epoch": 0.1829069134458356,
+      "grad_norm": 0.25895094181331907,
+      "learning_rate": 1.9831702154362825e-06,
+      "loss": 0.5915,
+      "step": 231
+    },
+    {
+      "epoch": 0.18369871826594744,
+      "grad_norm": 0.25319893187249265,
+      "learning_rate": 1.9830179986297755e-06,
+      "loss": 0.5982,
+      "step": 232
+    },
+    {
+      "epoch": 0.1844905230860593,
+      "grad_norm": 0.24279819253802182,
+      "learning_rate": 1.98286510245164e-06,
+      "loss": 0.6067,
+      "step": 233
+    },
+    {
+      "epoch": 0.18528232790617113,
+      "grad_norm": 0.2619794566140669,
+      "learning_rate": 1.9827115270075438e-06,
+      "loss": 0.6116,
+      "step": 234
+    },
+    {
+      "epoch": 0.18607413272628298,
+      "grad_norm": 0.24762634464898894,
+      "learning_rate": 1.9825572724036242e-06,
+      "loss": 0.5835,
+      "step": 235
+    },
+    {
+      "epoch": 0.1868659375463948,
+      "grad_norm": 0.23136911235072175,
+      "learning_rate": 1.9824023387464877e-06,
+      "loss": 0.6147,
+      "step": 236
+    },
+    {
+      "epoch": 0.18765774236650665,
+      "grad_norm": 0.2635243334169113,
+      "learning_rate": 1.9822467261432097e-06,
+      "loss": 0.6159,
+      "step": 237
+    },
+    {
+      "epoch": 0.1884495471866185,
+      "grad_norm": 0.25635271967861756,
+      "learning_rate": 1.9820904347013364e-06,
+      "loss": 0.614,
+      "step": 238
+    },
+    {
+      "epoch": 0.18924135200673034,
+      "grad_norm": 0.24759710351741077,
+      "learning_rate": 1.9819334645288813e-06,
+      "loss": 0.604,
+      "step": 239
+    },
+    {
+      "epoch": 0.1900331568268422,
+      "grad_norm": 0.25418645288516356,
+      "learning_rate": 1.9817758157343274e-06,
+      "loss": 0.5995,
+      "step": 240
+    },
+    {
+      "epoch": 0.19082496164695403,
+      "grad_norm": 0.24625861500046178,
+      "learning_rate": 1.981617488426628e-06,
+      "loss": 0.6105,
+      "step": 241
+    },
+    {
+      "epoch": 0.19161676646706588,
+      "grad_norm": 0.2477603992651197,
+      "learning_rate": 1.981458482715204e-06,
+      "loss": 0.5975,
+      "step": 242
+    },
+    {
+      "epoch": 0.1924085712871777,
+      "grad_norm": 0.25403326190627773,
+      "learning_rate": 1.9812987987099454e-06,
+      "loss": 0.607,
+      "step": 243
+    },
+    {
+      "epoch": 0.19320037610728955,
+      "grad_norm": 0.254854496754908,
+      "learning_rate": 1.9811384365212108e-06,
+      "loss": 0.595,
+      "step": 244
+    },
+    {
+      "epoch": 0.1939921809274014,
+      "grad_norm": 0.24897773797625972,
+      "learning_rate": 1.9809773962598285e-06,
+      "loss": 0.6066,
+      "step": 245
+    },
+    {
+      "epoch": 0.19478398574751324,
+      "grad_norm": 0.257037218773845,
+      "learning_rate": 1.980815678037094e-06,
+      "loss": 0.6144,
+      "step": 246
+    },
+    {
+      "epoch": 0.19557579056762509,
+      "grad_norm": 0.24510867554760934,
+      "learning_rate": 1.980653281964773e-06,
+      "loss": 0.6091,
+      "step": 247
+    },
+    {
+      "epoch": 0.19636759538773693,
+      "grad_norm": 0.25553709307934286,
+      "learning_rate": 1.9804902081550976e-06,
+      "loss": 0.6204,
+      "step": 248
+    },
+    {
+      "epoch": 0.19715940020784878,
+      "grad_norm": 0.24653872920986297,
+      "learning_rate": 1.98032645672077e-06,
+      "loss": 0.6101,
+      "step": 249
+    },
+    {
+      "epoch": 0.1979512050279606,
+      "grad_norm": 0.2639555201466776,
+      "learning_rate": 1.9801620277749603e-06,
+      "loss": 0.613,
+      "step": 250
+    },
+    {
+      "epoch": 0.19874300984807244,
+      "grad_norm": 0.24866230382323729,
+      "learning_rate": 1.9799969214313065e-06,
+      "loss": 0.5997,
+      "step": 251
+    },
+    {
+      "epoch": 0.1995348146681843,
+      "grad_norm": 0.24273609462370518,
+      "learning_rate": 1.9798311378039146e-06,
+      "loss": 0.5914,
+      "step": 252
+    },
+    {
+      "epoch": 0.20032661948829614,
+      "grad_norm": 0.2506812812970362,
+      "learning_rate": 1.9796646770073594e-06,
+      "loss": 0.6215,
+      "step": 253
+    },
+    {
+      "epoch": 0.20111842430840798,
+      "grad_norm": 0.24602122825448242,
+      "learning_rate": 1.9794975391566832e-06,
+      "loss": 0.6155,
+      "step": 254
+    },
+    {
+      "epoch": 0.20191022912851983,
+      "grad_norm": 0.24750713225507665,
+      "learning_rate": 1.979329724367396e-06,
+      "loss": 0.6088,
+      "step": 255
+    },
+    {
+      "epoch": 0.20270203394863165,
+      "grad_norm": 0.24092722636200453,
+      "learning_rate": 1.9791612327554768e-06,
+      "loss": 0.6022,
+      "step": 256
+    },
+    {
+      "epoch": 0.2034938387687435,
+      "grad_norm": 0.2444047578034734,
+      "learning_rate": 1.9789920644373706e-06,
+      "loss": 0.611,
+      "step": 257
+    },
+    {
+      "epoch": 0.20428564358885534,
+      "grad_norm": 0.2508034294267427,
+      "learning_rate": 1.978822219529991e-06,
+      "loss": 0.6042,
+      "step": 258
+    },
+    {
+      "epoch": 0.2050774484089672,
+      "grad_norm": 0.2485192806241367,
+      "learning_rate": 1.97865169815072e-06,
+      "loss": 0.593,
+      "step": 259
+    },
+    {
+      "epoch": 0.20586925322907904,
+      "grad_norm": 0.25986543914571764,
+      "learning_rate": 1.9784805004174053e-06,
+      "loss": 0.6111,
+      "step": 260
+    },
+    {
+      "epoch": 0.20666105804919088,
+      "grad_norm": 0.24888504318885482,
+      "learning_rate": 1.978308626448364e-06,
+      "loss": 0.6137,
+      "step": 261
+    },
+    {
+      "epoch": 0.20745286286930273,
+      "grad_norm": 0.2509505465202432,
+      "learning_rate": 1.978136076362379e-06,
+      "loss": 0.5987,
+      "step": 262
+    },
+    {
+      "epoch": 0.20824466768941455,
+      "grad_norm": 0.23819610108431402,
+      "learning_rate": 1.977962850278701e-06,
+      "loss": 0.6001,
+      "step": 263
+    },
+    {
+      "epoch": 0.2090364725095264,
+      "grad_norm": 0.25056741041097214,
+      "learning_rate": 1.977788948317048e-06,
+      "loss": 0.6033,
+      "step": 264
+    },
+    {
+      "epoch": 0.20982827732963824,
+      "grad_norm": 0.24498731933250023,
+      "learning_rate": 1.977614370597605e-06,
+      "loss": 0.5956,
+      "step": 265
+    },
+    {
+      "epoch": 0.2106200821497501,
+      "grad_norm": 0.24759436752884384,
+      "learning_rate": 1.9774391172410243e-06,
+      "loss": 0.6031,
+      "step": 266
+    },
+    {
+      "epoch": 0.21141188696986193,
+      "grad_norm": 0.24667679841911644,
+      "learning_rate": 1.977263188368425e-06,
+      "loss": 0.5989,
+      "step": 267
+    },
+    {
+      "epoch": 0.21220369178997378,
+      "grad_norm": 0.24341767258742814,
+      "learning_rate": 1.977086584101393e-06,
+      "loss": 0.5969,
+      "step": 268
+    },
+    {
+      "epoch": 0.21299549661008563,
+      "grad_norm": 0.26466260911509654,
+      "learning_rate": 1.97690930456198e-06,
+      "loss": 0.6138,
+      "step": 269
+    },
+    {
+      "epoch": 0.21378730143019745,
+      "grad_norm": 0.2405167898745803,
+      "learning_rate": 1.9767313498727063e-06,
+      "loss": 0.6064,
+      "step": 270
+    },
+    {
+      "epoch": 0.2145791062503093,
+      "grad_norm": 0.24275673692261862,
+      "learning_rate": 1.9765527201565574e-06,
+      "loss": 0.5975,
+      "step": 271
+    },
+    {
+      "epoch": 0.21537091107042114,
+      "grad_norm": 1.1475853790832273,
+      "learning_rate": 1.976373415536986e-06,
+      "loss": 0.5973,
+      "step": 272
+    },
+    {
+      "epoch": 0.21616271589053299,
+      "grad_norm": 0.25371017212183866,
+      "learning_rate": 1.9761934361379107e-06,
+      "loss": 0.6184,
+      "step": 273
+    },
+    {
+      "epoch": 0.21695452071064483,
+      "grad_norm": 0.24533245392799563,
+      "learning_rate": 1.9760127820837163e-06,
+      "loss": 0.6082,
+      "step": 274
+    },
+    {
+      "epoch": 0.21774632553075668,
+      "grad_norm": 0.24420365359668406,
+      "learning_rate": 1.9758314534992555e-06,
+      "loss": 0.5992,
+      "step": 275
+    },
+    {
+      "epoch": 0.2185381303508685,
+      "grad_norm": 0.24787248749030424,
+      "learning_rate": 1.9756494505098446e-06,
+      "loss": 0.6163,
+      "step": 276
+    },
+    {
+      "epoch": 0.21932993517098034,
+      "grad_norm": 0.25578859269353377,
+      "learning_rate": 1.975466773241268e-06,
+      "loss": 0.6118,
+      "step": 277
+    },
+    {
+      "epoch": 0.2201217399910922,
+      "grad_norm": 0.2407539832006363,
+      "learning_rate": 1.9752834218197757e-06,
+      "loss": 0.6093,
+      "step": 278
+    },
+    {
+      "epoch": 0.22091354481120404,
+      "grad_norm": 0.24025331731692606,
+      "learning_rate": 1.975099396372083e-06,
+      "loss": 0.5972,
+      "step": 279
+    },
+    {
+      "epoch": 0.22170534963131588,
+      "grad_norm": 0.24577431048359374,
+      "learning_rate": 1.9749146970253705e-06,
+      "loss": 0.603,
+      "step": 280
+    },
+    {
+      "epoch": 0.22249715445142773,
+      "grad_norm": 0.25542977696523095,
+      "learning_rate": 1.974729323907287e-06,
+      "loss": 0.6088,
+      "step": 281
+    },
+    {
+      "epoch": 0.22328895927153958,
+      "grad_norm": 0.2406509357436267,
+      "learning_rate": 1.9745432771459447e-06,
+      "loss": 0.6077,
+      "step": 282
+    },
+    {
+      "epoch": 0.2240807640916514,
+      "grad_norm": 0.24706767269031316,
+      "learning_rate": 1.974356556869922e-06,
+      "loss": 0.5984,
+      "step": 283
+    },
+    {
+      "epoch": 0.22487256891176324,
+      "grad_norm": 0.24309976779774528,
+      "learning_rate": 1.9741691632082623e-06,
+      "loss": 0.5834,
+      "step": 284
+    },
+    {
+      "epoch": 0.2256643737318751,
+      "grad_norm": 0.24748833527613592,
+      "learning_rate": 1.9739810962904752e-06,
+      "loss": 0.5814,
+      "step": 285
+    },
+    {
+      "epoch": 0.22645617855198694,
+      "grad_norm": 0.2460942176494084,
+      "learning_rate": 1.9737923562465356e-06,
+      "loss": 0.6015,
+      "step": 286
+    },
+    {
+      "epoch": 0.22724798337209878,
+      "grad_norm": 0.25104869936231566,
+      "learning_rate": 1.973602943206883e-06,
+      "loss": 0.6084,
+      "step": 287
+    },
+    {
+      "epoch": 0.22803978819221063,
+      "grad_norm": 0.2476909073234987,
+      "learning_rate": 1.9734128573024217e-06,
+      "loss": 0.5882,
+      "step": 288
+    },
+    {
+      "epoch": 0.22883159301232248,
+      "grad_norm": 0.24021216899169395,
+      "learning_rate": 1.9732220986645227e-06,
+      "loss": 0.6065,
+      "step": 289
+    },
+    {
+      "epoch": 0.2296233978324343,
+      "grad_norm": 0.26552042901412143,
+      "learning_rate": 1.97303066742502e-06,
+      "loss": 0.6075,
+      "step": 290
+    },
+    {
+      "epoch": 0.23041520265254614,
+      "grad_norm": 0.2391744425572385,
+      "learning_rate": 1.9728385637162136e-06,
+      "loss": 0.5956,
+      "step": 291
+    },
+    {
+      "epoch": 0.231207007472658,
+      "grad_norm": 0.240755445252526,
+      "learning_rate": 1.972645787670868e-06,
+      "loss": 0.5951,
+      "step": 292
+    },
+    {
+      "epoch": 0.23199881229276983,
+      "grad_norm": 0.25114356374370583,
+      "learning_rate": 1.972452339422212e-06,
+      "loss": 0.5855,
+      "step": 293
+    },
+    {
+      "epoch": 0.23279061711288168,
+      "grad_norm": 0.2517140758585886,
+      "learning_rate": 1.9722582191039393e-06,
+      "loss": 0.6042,
+      "step": 294
+    },
+    {
+      "epoch": 0.23358242193299353,
+      "grad_norm": 0.24718597294358718,
+      "learning_rate": 1.972063426850208e-06,
+      "loss": 0.5769,
+      "step": 295
+    },
+    {
+      "epoch": 0.23437422675310535,
+      "grad_norm": 0.2482292753203727,
+      "learning_rate": 1.971867962795641e-06,
+      "loss": 0.5808,
+      "step": 296
+    },
+    {
+      "epoch": 0.2351660315732172,
+      "grad_norm": 0.24794499541491236,
+      "learning_rate": 1.971671827075324e-06,
+      "loss": 0.5968,
+      "step": 297
+    },
+    {
+      "epoch": 0.23595783639332904,
+      "grad_norm": 0.23903391096586601,
+      "learning_rate": 1.9714750198248096e-06,
+      "loss": 0.6063,
+      "step": 298
+    },
+    {
+      "epoch": 0.2367496412134409,
+      "grad_norm": 0.24515278358151316,
+      "learning_rate": 1.9712775411801116e-06,
+      "loss": 0.6026,
+      "step": 299
+    },
+    {
+      "epoch": 0.23754144603355273,
+      "grad_norm": 0.24847240615755817,
+      "learning_rate": 1.9710793912777094e-06,
+      "loss": 0.5978,
+      "step": 300
+    },
+    {
+      "epoch": 0.23833325085366458,
+      "grad_norm": 0.25146845692678066,
+      "learning_rate": 1.9708805702545458e-06,
+      "loss": 0.606,
+      "step": 301
+    },
+    {
+      "epoch": 0.23912505567377643,
+      "grad_norm": 0.2528429551784224,
+      "learning_rate": 1.970681078248028e-06,
+      "loss": 0.5836,
+      "step": 302
+    },
+    {
+      "epoch": 0.23991686049388825,
+      "grad_norm": 0.25244432858861543,
+      "learning_rate": 1.970480915396027e-06,
+      "loss": 0.6076,
+      "step": 303
+    },
+    {
+      "epoch": 0.2407086653140001,
+      "grad_norm": 0.2651349383906822,
+      "learning_rate": 1.9702800818368754e-06,
+      "loss": 0.5888,
+      "step": 304
+    },
+    {
+      "epoch": 0.24150047013411194,
+      "grad_norm": 0.23884501343673606,
+      "learning_rate": 1.9700785777093724e-06,
+      "loss": 0.6105,
+      "step": 305
+    },
+    {
+      "epoch": 0.24229227495422379,
+      "grad_norm": 0.2491501508439908,
+      "learning_rate": 1.969876403152778e-06,
+      "loss": 0.595,
+      "step": 306
+    },
+    {
+      "epoch": 0.24308407977433563,
+      "grad_norm": 0.25783710037749885,
+      "learning_rate": 1.9696735583068173e-06,
+      "loss": 0.6198,
+      "step": 307
+    },
+    {
+      "epoch": 0.24387588459444748,
+      "grad_norm": 0.24815004860558676,
+      "learning_rate": 1.9694700433116783e-06,
+      "loss": 0.5937,
+      "step": 308
+    },
+    {
+      "epoch": 0.2446676894145593,
+      "grad_norm": 0.24426378069410107,
+      "learning_rate": 1.969265858308011e-06,
+      "loss": 0.5973,
+      "step": 309
+    },
+    {
+      "epoch": 0.24545949423467114,
+      "grad_norm": 0.24771046129560645,
+      "learning_rate": 1.96906100343693e-06,
+      "loss": 0.5897,
+      "step": 310
+    },
+    {
+      "epoch": 0.246251299054783,
+      "grad_norm": 0.25749973668903764,
+      "learning_rate": 1.968855478840012e-06,
+      "loss": 0.6004,
+      "step": 311
+    },
+    {
+      "epoch": 0.24704310387489484,
+      "grad_norm": 0.248025320372043,
+      "learning_rate": 1.9686492846592963e-06,
+      "loss": 0.5783,
+      "step": 312
+    },
+    {
+      "epoch": 0.24783490869500668,
+      "grad_norm": 0.24984650039117945,
+      "learning_rate": 1.9684424210372857e-06,
+      "loss": 0.6122,
+      "step": 313
+    },
+    {
+      "epoch": 0.24862671351511853,
+      "grad_norm": 0.24415386192705926,
+      "learning_rate": 1.968234888116945e-06,
+      "loss": 0.5981,
+      "step": 314
+    },
+    {
+      "epoch": 0.24941851833523038,
+      "grad_norm": 0.2509073672891484,
+      "learning_rate": 1.9680266860417025e-06,
+      "loss": 0.5998,
+      "step": 315
+    },
+    {
+      "epoch": 0.2502103231553422,
+      "grad_norm": 0.2552178325232198,
+      "learning_rate": 1.9678178149554477e-06,
+      "loss": 0.5946,
+      "step": 316
+    },
+    {
+      "epoch": 0.25100212797545407,
+      "grad_norm": 0.24539374694206276,
+      "learning_rate": 1.9676082750025335e-06,
+      "loss": 0.6,
+      "step": 317
+    },
+    {
+      "epoch": 0.2517939327955659,
+      "grad_norm": 0.25067363359825495,
+      "learning_rate": 1.967398066327774e-06,
+      "loss": 0.5971,
+      "step": 318
+    },
+    {
+      "epoch": 0.2525857376156777,
+      "grad_norm": 0.2506825238389697,
+      "learning_rate": 1.9671871890764473e-06,
+      "loss": 0.6043,
+      "step": 319
+    },
+    {
+      "epoch": 0.2533775424357896,
+      "grad_norm": 0.2541507375488699,
+      "learning_rate": 1.966975643394291e-06,
+      "loss": 0.599,
+      "step": 320
+    },
+    {
+      "epoch": 0.2541693472559014,
+      "grad_norm": 0.23826752511006694,
+      "learning_rate": 1.966763429427507e-06,
+      "loss": 0.5895,
+      "step": 321
+    },
+    {
+      "epoch": 0.2549611520760133,
+      "grad_norm": 0.2567212307793121,
+      "learning_rate": 1.9665505473227575e-06,
+      "loss": 0.59,
+      "step": 322
+    },
+    {
+      "epoch": 0.2557529568961251,
+      "grad_norm": 0.25250948582083205,
+      "learning_rate": 1.966336997227167e-06,
+      "loss": 0.6027,
+      "step": 323
+    },
+    {
+      "epoch": 0.25654476171623697,
+      "grad_norm": 0.2644211046230142,
+      "learning_rate": 1.9661227792883222e-06,
+      "loss": 0.6136,
+      "step": 324
+    },
+    {
+      "epoch": 0.2573365665363488,
+      "grad_norm": 0.2525750708474708,
+      "learning_rate": 1.9659078936542705e-06,
+      "loss": 0.6057,
+      "step": 325
+    },
+    {
+      "epoch": 0.2581283713564606,
+      "grad_norm": 0.24039651813252683,
+      "learning_rate": 1.9656923404735213e-06,
+      "loss": 0.5997,
+      "step": 326
+    },
+    {
+      "epoch": 0.2589201761765725,
+      "grad_norm": 0.2556325514662131,
+      "learning_rate": 1.965476119895045e-06,
+      "loss": 0.6052,
+      "step": 327
+    },
+    {
+      "epoch": 0.2597119809966843,
+      "grad_norm": 0.2570466737402276,
+      "learning_rate": 1.965259232068273e-06,
+      "loss": 0.5967,
+      "step": 328
+    },
+    {
+      "epoch": 0.2605037858167962,
+      "grad_norm": 0.2548504469875246,
+      "learning_rate": 1.965041677143099e-06,
+      "loss": 0.5956,
+      "step": 329
+    },
+    {
+      "epoch": 0.261295590636908,
+      "grad_norm": 0.24551894747783568,
+      "learning_rate": 1.9648234552698764e-06,
+      "loss": 0.5855,
+      "step": 330
+    },
+    {
+      "epoch": 0.26208739545701987,
+      "grad_norm": 0.2511820658089693,
+      "learning_rate": 1.9646045665994203e-06,
+      "loss": 0.6027,
+      "step": 331
+    },
+    {
+      "epoch": 0.2628792002771317,
+      "grad_norm": 0.25625179410182997,
+      "learning_rate": 1.964385011283006e-06,
+      "loss": 0.5822,
+      "step": 332
+    },
+    {
+      "epoch": 0.2636710050972435,
+      "grad_norm": 0.2492724738434624,
+      "learning_rate": 1.9641647894723706e-06,
+      "loss": 0.5962,
+      "step": 333
+    },
+    {
+      "epoch": 0.2644628099173554,
+      "grad_norm": 0.24639712006197856,
+      "learning_rate": 1.9639439013197107e-06,
+      "loss": 0.5964,
+      "step": 334
+    },
+    {
+      "epoch": 0.2652546147374672,
+      "grad_norm": 0.24284570346349335,
+      "learning_rate": 1.963722346977684e-06,
+      "loss": 0.6077,
+      "step": 335
+    },
+    {
+      "epoch": 0.2660464195575791,
+      "grad_norm": 0.23927956699044922,
+      "learning_rate": 1.9635001265994085e-06,
+      "loss": 0.6148,
+      "step": 336
+    },
+    {
+      "epoch": 0.2668382243776909,
+      "grad_norm": 0.2390886419500776,
+      "learning_rate": 1.9632772403384624e-06,
+      "loss": 0.6072,
+      "step": 337
+    },
+    {
+      "epoch": 0.26763002919780277,
+      "grad_norm": 0.24263488626702964,
+      "learning_rate": 1.9630536883488844e-06,
+      "loss": 0.604,
+      "step": 338
+    },
+    {
+      "epoch": 0.2684218340179146,
+      "grad_norm": 0.24884755916427828,
+      "learning_rate": 1.9628294707851733e-06,
+      "loss": 0.5971,
+      "step": 339
+    },
+    {
+      "epoch": 0.2692136388380264,
+      "grad_norm": 0.2566830533858907,
+      "learning_rate": 1.9626045878022866e-06,
+      "loss": 0.6017,
+      "step": 340
+    },
+    {
+      "epoch": 0.2700054436581383,
+      "grad_norm": 0.2410117164423693,
+      "learning_rate": 1.962379039555644e-06,
+      "loss": 0.6015,
+      "step": 341
+    },
+    {
+      "epoch": 0.2707972484782501,
+      "grad_norm": 0.24423703801214214,
+      "learning_rate": 1.9621528262011227e-06,
+      "loss": 0.6044,
+      "step": 342
+    },
+    {
+      "epoch": 0.27158905329836197,
+      "grad_norm": 0.2515836234811855,
+      "learning_rate": 1.9619259478950614e-06,
+      "loss": 0.5946,
+      "step": 343
+    },
+    {
+      "epoch": 0.2723808581184738,
+      "grad_norm": 0.2597038071689357,
+      "learning_rate": 1.961698404794257e-06,
+      "loss": 0.5985,
+      "step": 344
+    },
+    {
+      "epoch": 0.27317266293858566,
+      "grad_norm": 0.24035427328118955,
+      "learning_rate": 1.9614701970559664e-06,
+      "loss": 0.5903,
+      "step": 345
+    },
+    {
+      "epoch": 0.2739644677586975,
+      "grad_norm": 0.24663368446774298,
+      "learning_rate": 1.961241324837906e-06,
+      "loss": 0.6162,
+      "step": 346
+    },
+    {
+      "epoch": 0.2747562725788093,
+      "grad_norm": 0.26954452255428774,
+      "learning_rate": 1.9610117882982512e-06,
+      "loss": 0.5952,
+      "step": 347
+    },
+    {
+      "epoch": 0.2755480773989212,
+      "grad_norm": 0.2439930239105369,
+      "learning_rate": 1.960781587595636e-06,
+      "loss": 0.5944,
+      "step": 348
+    },
+    {
+      "epoch": 0.276339882219033,
+      "grad_norm": 0.24771750596417796,
+      "learning_rate": 1.9605507228891547e-06,
+      "loss": 0.5801,
+      "step": 349
+    },
+    {
+      "epoch": 0.27713168703914487,
+      "grad_norm": 0.26550547373108035,
+      "learning_rate": 1.9603191943383597e-06,
+      "loss": 0.5961,
+      "step": 350
+    },
+    {
+      "epoch": 0.2779234918592567,
+      "grad_norm": 0.24431239688491097,
+      "learning_rate": 1.9600870021032615e-06,
+      "loss": 0.5907,
+      "step": 351
+    },
+    {
+      "epoch": 0.27871529667936856,
+      "grad_norm": 0.2535682794416749,
+      "learning_rate": 1.9598541463443307e-06,
+      "loss": 0.603,
+      "step": 352
+    },
+    {
+      "epoch": 0.2795071014994804,
+      "grad_norm": 0.2448211841047002,
+      "learning_rate": 1.959620627222496e-06,
+      "loss": 0.5928,
+      "step": 353
+    },
+    {
+      "epoch": 0.2802989063195922,
+      "grad_norm": 0.25103757340926774,
+      "learning_rate": 1.9593864448991435e-06,
+      "loss": 0.6032,
+      "step": 354
+    },
+    {
+      "epoch": 0.2810907111397041,
+      "grad_norm": 0.23940689444983226,
+      "learning_rate": 1.959151599536119e-06,
+      "loss": 0.5938,
+      "step": 355
+    },
+    {
+      "epoch": 0.2818825159598159,
+      "grad_norm": 0.2607433431505887,
+      "learning_rate": 1.958916091295726e-06,
+      "loss": 0.6062,
+      "step": 356
+    },
+    {
+      "epoch": 0.28267432077992777,
+      "grad_norm": 0.25475404406877755,
+      "learning_rate": 1.958679920340726e-06,
+      "loss": 0.5947,
+      "step": 357
+    },
+    {
+      "epoch": 0.2834661256000396,
+      "grad_norm": 0.24802973037438725,
+      "learning_rate": 1.958443086834339e-06,
+      "loss": 0.5991,
+      "step": 358
+    },
+    {
+      "epoch": 0.2842579304201514,
+      "grad_norm": 0.2431769776754948,
+      "learning_rate": 1.958205590940242e-06,
+      "loss": 0.6063,
+      "step": 359
+    },
+    {
+      "epoch": 0.2850497352402633,
+      "grad_norm": 0.2639127813318562,
+      "learning_rate": 1.957967432822571e-06,
+      "loss": 0.6086,
+      "step": 360
+    },
+    {
+      "epoch": 0.2858415400603751,
+      "grad_norm": 0.26584545205362775,
+      "learning_rate": 1.9577286126459176e-06,
+      "loss": 0.6097,
+      "step": 361
+    },
+    {
+      "epoch": 0.286633344880487,
+      "grad_norm": 0.24055194183303985,
+      "learning_rate": 1.957489130575334e-06,
+      "loss": 0.5833,
+      "step": 362
+    },
+    {
+      "epoch": 0.2874251497005988,
+      "grad_norm": 0.25429600055401314,
+      "learning_rate": 1.9572489867763274e-06,
+      "loss": 0.6006,
+      "step": 363
+    },
+    {
+      "epoch": 0.28821695452071067,
+      "grad_norm": 0.2537061934212363,
+      "learning_rate": 1.957008181414863e-06,
+      "loss": 0.6085,
+      "step": 364
+    },
+    {
+      "epoch": 0.2890087593408225,
+      "grad_norm": 0.23596507193683017,
+      "learning_rate": 1.956766714657364e-06,
+      "loss": 0.5909,
+      "step": 365
+    },
+    {
+      "epoch": 0.2898005641609343,
+      "grad_norm": 0.25278423998075816,
+      "learning_rate": 1.956524586670709e-06,
+      "loss": 0.5941,
+      "step": 366
+    },
+    {
+      "epoch": 0.2905923689810462,
+      "grad_norm": 0.2594776908735593,
+      "learning_rate": 1.9562817976222356e-06,
+      "loss": 0.5861,
+      "step": 367
+    },
+    {
+      "epoch": 0.291384173801158,
+      "grad_norm": 0.23709390308086759,
+      "learning_rate": 1.956038347679736e-06,
+      "loss": 0.5982,
+      "step": 368
+    },
+    {
+      "epoch": 0.29217597862126987,
+      "grad_norm": 0.24788413500193693,
+      "learning_rate": 1.9557942370114618e-06,
+      "loss": 0.6196,
+      "step": 369
+    },
+    {
+      "epoch": 0.2929677834413817,
+      "grad_norm": 0.2456033048336526,
+      "learning_rate": 1.955549465786119e-06,
+      "loss": 0.5918,
+      "step": 370
+    },
+    {
+      "epoch": 0.29375958826149356,
+      "grad_norm": 0.24947676521440168,
+      "learning_rate": 1.9553040341728716e-06,
+      "loss": 0.5946,
+      "step": 371
+    },
+    {
+      "epoch": 0.2945513930816054,
+      "grad_norm": 0.2698339661266052,
+      "learning_rate": 1.9550579423413384e-06,
+      "loss": 0.5761,
+      "step": 372
+    },
+    {
+      "epoch": 0.2953431979017172,
+      "grad_norm": 0.25230214410700375,
+      "learning_rate": 1.954811190461596e-06,
+      "loss": 0.5821,
+      "step": 373
+    },
+    {
+      "epoch": 0.2961350027218291,
+      "grad_norm": 0.3105801685091088,
+      "learning_rate": 1.954563778704177e-06,
+      "loss": 0.6046,
+      "step": 374
+    },
+    {
+      "epoch": 0.2969268075419409,
+      "grad_norm": 0.23384241659400207,
+      "learning_rate": 1.954315707240069e-06,
+      "loss": 0.5907,
+      "step": 375
+    },
+    {
+      "epoch": 0.29771861236205277,
+      "grad_norm": 0.24705464372439737,
+      "learning_rate": 1.9540669762407163e-06,
+      "loss": 0.5871,
+      "step": 376
+    },
+    {
+      "epoch": 0.2985104171821646,
+      "grad_norm": 0.2595162897846651,
+      "learning_rate": 1.953817585878019e-06,
+      "loss": 0.5971,
+      "step": 377
+    },
+    {
+      "epoch": 0.29930222200227646,
+      "grad_norm": 0.26096526999231834,
+      "learning_rate": 1.9535675363243334e-06,
+      "loss": 0.594,
+      "step": 378
+    },
+    {
+      "epoch": 0.3000940268223883,
+      "grad_norm": 0.2579837657000217,
+      "learning_rate": 1.9533168277524695e-06,
+      "loss": 0.5993,
+      "step": 379
+    },
+    {
+      "epoch": 0.3008858316425001,
+      "grad_norm": 0.260944446862816,
+      "learning_rate": 1.953065460335695e-06,
+      "loss": 0.6004,
+      "step": 380
+    },
+    {
+      "epoch": 0.301677636462612,
+      "grad_norm": 0.2539201800274544,
+      "learning_rate": 1.9528134342477313e-06,
+      "loss": 0.6043,
+      "step": 381
+    },
+    {
+      "epoch": 0.3024694412827238,
+      "grad_norm": 0.2594416904882131,
+      "learning_rate": 1.952560749662756e-06,
+      "loss": 0.593,
+      "step": 382
+    },
+    {
+      "epoch": 0.30326124610283567,
+      "grad_norm": 0.24637556242607472,
+      "learning_rate": 1.9523074067554017e-06,
+      "loss": 0.6056,
+      "step": 383
+    },
+    {
+      "epoch": 0.3040530509229475,
+      "grad_norm": 0.2613736787318001,
+      "learning_rate": 1.9520534057007548e-06,
+      "loss": 0.5885,
+      "step": 384
+    },
+    {
+      "epoch": 0.30484485574305936,
+      "grad_norm": 0.2576006861385841,
+      "learning_rate": 1.9517987466743585e-06,
+      "loss": 0.6058,
+      "step": 385
+    },
+    {
+      "epoch": 0.3056366605631712,
+      "grad_norm": 0.2559209310303427,
+      "learning_rate": 1.9515434298522093e-06,
+      "loss": 0.5992,
+      "step": 386
+    },
+    {
+      "epoch": 0.306428465383283,
+      "grad_norm": 0.26632881562278343,
+      "learning_rate": 1.9512874554107587e-06,
+      "loss": 0.5859,
+      "step": 387
+    },
+    {
+      "epoch": 0.3072202702033949,
+      "grad_norm": 0.25448466650597745,
+      "learning_rate": 1.9510308235269122e-06,
+      "loss": 0.5952,
+      "step": 388
+    },
+    {
+      "epoch": 0.3080120750235067,
+      "grad_norm": 0.24686184135640277,
+      "learning_rate": 1.9507735343780308e-06,
+      "loss": 0.6017,
+      "step": 389
+    },
+    {
+      "epoch": 0.30880387984361857,
+      "grad_norm": 0.25400298532675814,
+      "learning_rate": 1.9505155881419293e-06,
+      "loss": 0.5845,
+      "step": 390
+    },
+    {
+      "epoch": 0.3095956846637304,
+      "grad_norm": 0.24829689603575703,
+      "learning_rate": 1.9502569849968756e-06,
+      "loss": 0.6061,
+      "step": 391
+    },
+    {
+      "epoch": 0.31038748948384226,
+      "grad_norm": 0.2423789155144904,
+      "learning_rate": 1.949997725121593e-06,
+      "loss": 0.5973,
+      "step": 392
+    },
+    {
+      "epoch": 0.3111792943039541,
+      "grad_norm": 0.2502562090896562,
+      "learning_rate": 1.949737808695258e-06,
+      "loss": 0.5886,
+      "step": 393
+    },
+    {
+      "epoch": 0.3119710991240659,
+      "grad_norm": 0.25580619317505604,
+      "learning_rate": 1.949477235897501e-06,
+      "loss": 0.6038,
+      "step": 394
+    },
+    {
+      "epoch": 0.3127629039441778,
+      "grad_norm": 0.25269772488002135,
+      "learning_rate": 1.9492160069084057e-06,
+      "loss": 0.5898,
+      "step": 395
+    },
+    {
+      "epoch": 0.3135547087642896,
+      "grad_norm": 0.2557818555265261,
+      "learning_rate": 1.948954121908509e-06,
+      "loss": 0.5885,
+      "step": 396
+    },
+    {
+      "epoch": 0.31434651358440147,
+      "grad_norm": 0.2450807783588401,
+      "learning_rate": 1.9486915810788032e-06,
+      "loss": 0.6002,
+      "step": 397
+    },
+    {
+      "epoch": 0.3151383184045133,
+      "grad_norm": 0.2510392036495738,
+      "learning_rate": 1.9484283846007313e-06,
+      "loss": 0.5976,
+      "step": 398
+    },
+    {
+      "epoch": 0.3159301232246251,
+      "grad_norm": 0.25561092633708016,
+      "learning_rate": 1.948164532656191e-06,
+      "loss": 0.5776,
+      "step": 399
+    },
+    {
+      "epoch": 0.316721928044737,
+      "grad_norm": 0.2596884998651969,
+      "learning_rate": 1.9479000254275314e-06,
+      "loss": 0.599,
+      "step": 400
+    },
+    {
+      "epoch": 0.3175137328648488,
+      "grad_norm": 0.26498978664991624,
+      "learning_rate": 1.947634863097557e-06,
+      "loss": 0.5967,
+      "step": 401
+    },
+    {
+      "epoch": 0.31830553768496067,
+      "grad_norm": 0.2501572200213723,
+      "learning_rate": 1.947369045849523e-06,
+      "loss": 0.5943,
+      "step": 402
+    },
+    {
+      "epoch": 0.3190973425050725,
+      "grad_norm": 0.2577630865177056,
+      "learning_rate": 1.947102573867138e-06,
+      "loss": 0.5887,
+      "step": 403
+    },
+    {
+      "epoch": 0.31988914732518436,
+      "grad_norm": 0.24174145169513117,
+      "learning_rate": 1.9468354473345624e-06,
+      "loss": 0.6044,
+      "step": 404
+    },
+    {
+      "epoch": 0.3206809521452962,
+      "grad_norm": 0.24591986565078588,
+      "learning_rate": 1.9465676664364104e-06,
+      "loss": 0.6086,
+      "step": 405
+    },
+    {
+      "epoch": 0.321472756965408,
+      "grad_norm": 0.26307380086163157,
+      "learning_rate": 1.9462992313577462e-06,
+      "loss": 0.6025,
+      "step": 406
+    },
+    {
+      "epoch": 0.3222645617855199,
+      "grad_norm": 0.24700348114115595,
+      "learning_rate": 1.946030142284089e-06,
+      "loss": 0.5896,
+      "step": 407
+    },
+    {
+      "epoch": 0.3230563666056317,
+      "grad_norm": 0.2558304440489774,
+      "learning_rate": 1.9457603994014073e-06,
+      "loss": 0.5995,
+      "step": 408
+    },
+    {
+      "epoch": 0.32384817142574357,
+      "grad_norm": 0.26976961195565224,
+      "learning_rate": 1.9454900028961234e-06,
+      "loss": 0.5806,
+      "step": 409
+    },
+    {
+      "epoch": 0.3246399762458554,
+      "grad_norm": 0.24682486178015295,
+      "learning_rate": 1.9452189529551094e-06,
+      "loss": 0.6029,
+      "step": 410
+    },
+    {
+      "epoch": 0.32543178106596726,
+      "grad_norm": 0.2511422633097611,
+      "learning_rate": 1.944947249765691e-06,
+      "loss": 0.5858,
+      "step": 411
+    },
+    {
+      "epoch": 0.3262235858860791,
+      "grad_norm": 0.26370240792835703,
+      "learning_rate": 1.944674893515644e-06,
+      "loss": 0.5996,
+      "step": 412
+    },
+    {
+      "epoch": 0.3270153907061909,
+      "grad_norm": 0.2523583273695204,
+      "learning_rate": 1.9444018843931965e-06,
+      "loss": 0.6015,
+      "step": 413
+    },
+    {
+      "epoch": 0.3278071955263028,
+      "grad_norm": 0.24129610898224038,
+      "learning_rate": 1.9441282225870267e-06,
+      "loss": 0.5917,
+      "step": 414
+    },
+    {
+      "epoch": 0.3285990003464146,
+      "grad_norm": 0.2616328657607042,
+      "learning_rate": 1.9438539082862648e-06,
+      "loss": 0.5934,
+      "step": 415
+    },
+    {
+      "epoch": 0.32939080516652647,
+      "grad_norm": 0.24557528223674166,
+      "learning_rate": 1.9435789416804915e-06,
+      "loss": 0.5987,
+      "step": 416
+    },
+    {
+      "epoch": 0.3301826099866383,
+      "grad_norm": 0.2514189246130001,
+      "learning_rate": 1.9433033229597384e-06,
+      "loss": 0.5961,
+      "step": 417
+    },
+    {
+      "epoch": 0.33097441480675016,
+      "grad_norm": 0.2466652126527317,
+      "learning_rate": 1.9430270523144882e-06,
+      "loss": 0.5804,
+      "step": 418
+    },
+    {
+      "epoch": 0.331766219626862,
+      "grad_norm": 0.24752095595178564,
+      "learning_rate": 1.942750129935673e-06,
+      "loss": 0.5935,
+      "step": 419
+    },
+    {
+      "epoch": 0.3325580244469738,
+      "grad_norm": 0.24996783596689934,
+      "learning_rate": 1.942472556014677e-06,
+      "loss": 0.5994,
+      "step": 420
+    },
+    {
+      "epoch": 0.3333498292670857,
+      "grad_norm": 0.25223965184297614,
+      "learning_rate": 1.942194330743333e-06,
+      "loss": 0.5898,
+      "step": 421
+    },
+    {
+      "epoch": 0.3341416340871975,
+      "grad_norm": 0.2534843988928022,
+      "learning_rate": 1.9419154543139247e-06,
+      "loss": 0.6157,
+      "step": 422
+    },
+    {
+      "epoch": 0.33493343890730937,
+      "grad_norm": 0.24706484156818997,
+      "learning_rate": 1.9416359269191864e-06,
+      "loss": 0.6089,
+      "step": 423
+    },
+    {
+      "epoch": 0.3357252437274212,
+      "grad_norm": 0.24757094331330678,
+      "learning_rate": 1.9413557487523014e-06,
+      "loss": 0.6049,
+      "step": 424
+    },
+    {
+      "epoch": 0.33651704854753306,
+      "grad_norm": 0.24142843028914535,
+      "learning_rate": 1.9410749200069033e-06,
+      "loss": 0.578,
+      "step": 425
+    },
+    {
+      "epoch": 0.3373088533676449,
+      "grad_norm": 0.2510069841485761,
+      "learning_rate": 1.940793440877075e-06,
+      "loss": 0.5879,
+      "step": 426
+    },
+    {
+      "epoch": 0.3381006581877567,
+      "grad_norm": 0.2609053814825094,
+      "learning_rate": 1.940511311557349e-06,
+      "loss": 0.5848,
+      "step": 427
+    },
+    {
+      "epoch": 0.33889246300786857,
+      "grad_norm": 0.2518737245867738,
+      "learning_rate": 1.940228532242707e-06,
+      "loss": 0.5962,
+      "step": 428
+    },
+    {
+      "epoch": 0.3396842678279804,
+      "grad_norm": 0.25221504281541435,
+      "learning_rate": 1.93994510312858e-06,
+      "loss": 0.6024,
+      "step": 429
+    },
+    {
+      "epoch": 0.34047607264809227,
+      "grad_norm": 0.24951406008176466,
+      "learning_rate": 1.9396610244108486e-06,
+      "loss": 0.5985,
+      "step": 430
+    },
+    {
+      "epoch": 0.3412678774682041,
+      "grad_norm": 0.2453685216063324,
+      "learning_rate": 1.939376296285841e-06,
+      "loss": 0.6025,
+      "step": 431
+    },
+    {
+      "epoch": 0.3420596822883159,
+      "grad_norm": 0.24921034068058728,
+      "learning_rate": 1.939090918950336e-06,
+      "loss": 0.5982,
+      "step": 432
+    },
+    {
+      "epoch": 0.3428514871084278,
+      "grad_norm": 0.24095478985143048,
+      "learning_rate": 1.9388048926015593e-06,
+      "loss": 0.6006,
+      "step": 433
+    },
+    {
+      "epoch": 0.3436432919285396,
+      "grad_norm": 0.23977530214075207,
+      "learning_rate": 1.9385182174371872e-06,
+      "loss": 0.5993,
+      "step": 434
+    },
+    {
+      "epoch": 0.34443509674865147,
+      "grad_norm": 0.241451042853671,
+      "learning_rate": 1.938230893655342e-06,
+      "loss": 0.5947,
+      "step": 435
+    },
+    {
+      "epoch": 0.3452269015687633,
+      "grad_norm": 0.2514189307194158,
+      "learning_rate": 1.937942921454596e-06,
+      "loss": 0.5853,
+      "step": 436
+    },
+    {
+      "epoch": 0.34601870638887516,
+      "grad_norm": 0.24048387845244765,
+      "learning_rate": 1.9376543010339692e-06,
+      "loss": 0.5988,
+      "step": 437
+    },
+    {
+      "epoch": 0.346810511208987,
+      "grad_norm": 0.2536295455265749,
+      "learning_rate": 1.93736503259293e-06,
+      "loss": 0.587,
+      "step": 438
+    },
+    {
+      "epoch": 0.3476023160290988,
+      "grad_norm": 0.261232912394264,
+      "learning_rate": 1.9370751163313927e-06,
+      "loss": 0.5965,
+      "step": 439
+    },
+    {
+      "epoch": 0.3483941208492107,
+      "grad_norm": 0.24959356062955415,
+      "learning_rate": 1.9367845524497217e-06,
+      "loss": 0.6025,
+      "step": 440
+    },
+    {
+      "epoch": 0.3491859256693225,
+      "grad_norm": 0.24606995041096932,
+      "learning_rate": 1.9364933411487283e-06,
+      "loss": 0.5878,
+      "step": 441
+    },
+    {
+      "epoch": 0.34997773048943437,
+      "grad_norm": 0.2525242803889403,
+      "learning_rate": 1.93620148262967e-06,
+      "loss": 0.6033,
+      "step": 442
+    },
+    {
+      "epoch": 0.3507695353095462,
+      "grad_norm": 0.24699322431888046,
+      "learning_rate": 1.9359089770942534e-06,
+      "loss": 0.6022,
+      "step": 443
+    },
+    {
+      "epoch": 0.35156134012965806,
+      "grad_norm": 0.25703417564848346,
+      "learning_rate": 1.935615824744631e-06,
+      "loss": 0.6028,
+      "step": 444
+    },
+    {
+      "epoch": 0.3523531449497699,
+      "grad_norm": 0.25967336842246613,
+      "learning_rate": 1.935322025783403e-06,
+      "loss": 0.6214,
+      "step": 445
+    },
+    {
+      "epoch": 0.3531449497698817,
+      "grad_norm": 0.2540070899335221,
+      "learning_rate": 1.9350275804136166e-06,
+      "loss": 0.5946,
+      "step": 446
+    },
+    {
+      "epoch": 0.3539367545899936,
+      "grad_norm": 0.24440179466401316,
+      "learning_rate": 1.934732488838764e-06,
+      "loss": 0.5949,
+      "step": 447
+    },
+    {
+      "epoch": 0.3547285594101054,
+      "grad_norm": 0.23931278271932108,
+      "learning_rate": 1.934436751262787e-06,
+      "loss": 0.5845,
+      "step": 448
+    },
+    {
+      "epoch": 0.35552036423021727,
+      "grad_norm": 0.24706305509590024,
+      "learning_rate": 1.934140367890071e-06,
+      "loss": 0.5828,
+      "step": 449
+    },
+    {
+      "epoch": 0.3563121690503291,
+      "grad_norm": 0.24703464436340708,
+      "learning_rate": 1.9338433389254495e-06,
+      "loss": 0.5947,
+      "step": 450
+    },
+    {
+      "epoch": 0.35710397387044096,
+      "grad_norm": 0.26567239272076876,
+      "learning_rate": 1.9335456645742014e-06,
+      "loss": 0.5995,
+      "step": 451
+    },
+    {
+      "epoch": 0.3578957786905528,
+      "grad_norm": 0.2623514720491113,
+      "learning_rate": 1.9332473450420523e-06,
+      "loss": 0.5831,
+      "step": 452
+    },
+    {
+      "epoch": 0.3586875835106646,
+      "grad_norm": 0.2518150603558824,
+      "learning_rate": 1.9329483805351723e-06,
+      "loss": 0.5921,
+      "step": 453
+    },
+    {
+      "epoch": 0.3594793883307765,
+      "grad_norm": 0.24844155181430966,
+      "learning_rate": 1.932648771260179e-06,
+      "loss": 0.5992,
+      "step": 454
+    },
+    {
+      "epoch": 0.3602711931508883,
+      "grad_norm": 0.24764470623088883,
+      "learning_rate": 1.932348517424134e-06,
+      "loss": 0.6017,
+      "step": 455
+    },
+    {
+      "epoch": 0.36106299797100017,
+      "grad_norm": 0.2444623913737678,
+      "learning_rate": 1.932047619234546e-06,
+      "loss": 0.6114,
+      "step": 456
+    },
+    {
+      "epoch": 0.361854802791112,
+      "grad_norm": 0.24878794738315252,
+      "learning_rate": 1.9317460768993676e-06,
+      "loss": 0.5979,
+      "step": 457
+    },
+    {
+      "epoch": 0.36264660761122386,
+      "grad_norm": 0.25626001974283374,
+      "learning_rate": 1.9314438906269975e-06,
+      "loss": 0.607,
+      "step": 458
+    },
+    {
+      "epoch": 0.3634384124313357,
+      "grad_norm": 0.24385129423397395,
+      "learning_rate": 1.931141060626279e-06,
+      "loss": 0.5852,
+      "step": 459
+    },
+    {
+      "epoch": 0.3642302172514475,
+      "grad_norm": 0.2561568562068568,
+      "learning_rate": 1.9308375871065003e-06,
+      "loss": 0.6122,
+      "step": 460
+    },
+    {
+      "epoch": 0.36502202207155937,
+      "grad_norm": 0.2511846260355575,
+      "learning_rate": 1.9305334702773936e-06,
+      "loss": 0.5911,
+      "step": 461
+    },
+    {
+      "epoch": 0.3658138268916712,
+      "grad_norm": 0.23904173222463654,
+      "learning_rate": 1.9302287103491376e-06,
+      "loss": 0.58,
+      "step": 462
+    },
+    {
+      "epoch": 0.36660563171178306,
+      "grad_norm": 0.24310034339402387,
+      "learning_rate": 1.9299233075323537e-06,
+      "loss": 0.574,
+      "step": 463
+    },
+    {
+      "epoch": 0.3673974365318949,
+      "grad_norm": 0.23495726837588596,
+      "learning_rate": 1.9296172620381087e-06,
+      "loss": 0.5905,
+      "step": 464
+    },
+    {
+      "epoch": 0.36818924135200676,
+      "grad_norm": 0.2546785897112177,
+      "learning_rate": 1.9293105740779126e-06,
+      "loss": 0.6015,
+      "step": 465
+    },
+    {
+      "epoch": 0.3689810461721186,
+      "grad_norm": 0.252056722545849,
+      "learning_rate": 1.92900324386372e-06,
+      "loss": 0.5817,
+      "step": 466
+    },
+    {
+      "epoch": 0.3697728509922304,
+      "grad_norm": 0.24993882029205636,
+      "learning_rate": 1.9286952716079297e-06,
+      "loss": 0.5917,
+      "step": 467
+    },
+    {
+      "epoch": 0.37056465581234227,
+      "grad_norm": 0.2563838810351198,
+      "learning_rate": 1.928386657523383e-06,
+      "loss": 0.6038,
+      "step": 468
+    },
+    {
+      "epoch": 0.3713564606324541,
+      "grad_norm": 0.244787640148442,
+      "learning_rate": 1.928077401823367e-06,
+      "loss": 0.5781,
+      "step": 469
+    },
+    {
+      "epoch": 0.37214826545256596,
+      "grad_norm": 0.25997369540927734,
+      "learning_rate": 1.9277675047216092e-06,
+      "loss": 0.602,
+      "step": 470
+    },
+    {
+      "epoch": 0.3729400702726778,
+      "grad_norm": 0.2577866099198746,
+      "learning_rate": 1.927456966432283e-06,
+      "loss": 0.5906,
+      "step": 471
+    },
+    {
+      "epoch": 0.3737318750927896,
+      "grad_norm": 0.25343861301131776,
+      "learning_rate": 1.9271457871700032e-06,
+      "loss": 0.5928,
+      "step": 472
+    },
+    {
+      "epoch": 0.3745236799129015,
+      "grad_norm": 0.24326992190799326,
+      "learning_rate": 1.9268339671498287e-06,
+      "loss": 0.5983,
+      "step": 473
+    },
+    {
+      "epoch": 0.3753154847330133,
+      "grad_norm": 0.2565132395383672,
+      "learning_rate": 1.926521506587261e-06,
+      "loss": 0.6021,
+      "step": 474
+    },
+    {
+      "epoch": 0.37610728955312517,
+      "grad_norm": 0.2592413096288339,
+      "learning_rate": 1.9262084056982437e-06,
+      "loss": 0.5897,
+      "step": 475
+    },
+    {
+      "epoch": 0.376899094373237,
+      "grad_norm": 0.25826628366629745,
+      "learning_rate": 1.9258946646991636e-06,
+      "loss": 0.596,
+      "step": 476
+    },
+    {
+      "epoch": 0.37769089919334886,
+      "grad_norm": 0.2479127952507506,
+      "learning_rate": 1.9255802838068497e-06,
+      "loss": 0.6046,
+      "step": 477
+    },
+    {
+      "epoch": 0.3784827040134607,
+      "grad_norm": 0.24028609433471643,
+      "learning_rate": 1.9252652632385727e-06,
+      "loss": 0.6079,
+      "step": 478
+    },
+    {
+      "epoch": 0.3792745088335725,
+      "grad_norm": 0.24444875484814269,
+      "learning_rate": 1.9249496032120463e-06,
+      "loss": 0.584,
+      "step": 479
+    },
+    {
+      "epoch": 0.3800663136536844,
+      "grad_norm": 0.2435843573159472,
+      "learning_rate": 1.924633303945425e-06,
+      "loss": 0.5912,
+      "step": 480
+    },
+    {
+      "epoch": 0.3808581184737962,
+      "grad_norm": 0.2580854272202583,
+      "learning_rate": 1.9243163656573072e-06,
+      "loss": 0.6155,
+      "step": 481
+    },
+    {
+      "epoch": 0.38164992329390807,
+      "grad_norm": 0.24377352220871984,
+      "learning_rate": 1.92399878856673e-06,
+      "loss": 0.5815,
+      "step": 482
+    },
+    {
+      "epoch": 0.3824417281140199,
+      "grad_norm": 0.2611149627044241,
+      "learning_rate": 1.923680572893174e-06,
+      "loss": 0.5874,
+      "step": 483
+    },
+    {
+      "epoch": 0.38323353293413176,
+      "grad_norm": 0.26532404200982734,
+      "learning_rate": 1.9233617188565607e-06,
+      "loss": 0.5857,
+      "step": 484
+    },
+    {
+      "epoch": 0.3840253377542436,
+      "grad_norm": 0.26893467080462036,
+      "learning_rate": 1.9230422266772525e-06,
+      "loss": 0.5952,
+      "step": 485
+    },
+    {
+      "epoch": 0.3848171425743554,
+      "grad_norm": 0.2571456802966826,
+      "learning_rate": 1.9227220965760535e-06,
+      "loss": 0.6135,
+      "step": 486
+    },
+    {
+      "epoch": 0.38560894739446727,
+      "grad_norm": 0.25166009156873304,
+      "learning_rate": 1.9224013287742078e-06,
+      "loss": 0.6047,
+      "step": 487
+    },
+    {
+      "epoch": 0.3864007522145791,
+      "grad_norm": 0.2702394849108666,
+      "learning_rate": 1.9220799234934e-06,
+      "loss": 0.5908,
+      "step": 488
+    },
+    {
+      "epoch": 0.38719255703469097,
+      "grad_norm": 0.25136505111479945,
+      "learning_rate": 1.921757880955757e-06,
+      "loss": 0.6039,
+      "step": 489
+    },
+    {
+      "epoch": 0.3879843618548028,
+      "grad_norm": 0.2534880486433305,
+      "learning_rate": 1.9214352013838445e-06,
+      "loss": 0.5984,
+      "step": 490
+    },
+    {
+      "epoch": 0.38877616667491466,
+      "grad_norm": 0.25108927918306556,
+      "learning_rate": 1.9211118850006687e-06,
+      "loss": 0.5988,
+      "step": 491
+    },
+    {
+      "epoch": 0.3895679714950265,
+      "grad_norm": 0.25633840231196486,
+      "learning_rate": 1.9207879320296764e-06,
+      "loss": 0.5882,
+      "step": 492
+    },
+    {
+      "epoch": 0.3903597763151383,
+      "grad_norm": 0.25275777210177885,
+      "learning_rate": 1.920463342694754e-06,
+      "loss": 0.5824,
+      "step": 493
+    },
+    {
+      "epoch": 0.39115158113525017,
+      "grad_norm": 0.2577018302858132,
+      "learning_rate": 1.9201381172202283e-06,
+      "loss": 0.6049,
+      "step": 494
+    },
+    {
+      "epoch": 0.391943385955362,
+      "grad_norm": 0.25717885246394784,
+      "learning_rate": 1.9198122558308643e-06,
+      "loss": 0.5977,
+      "step": 495
+    },
+    {
+      "epoch": 0.39273519077547386,
+      "grad_norm": 0.24231408485397066,
+      "learning_rate": 1.919485758751868e-06,
+      "loss": 0.5867,
+      "step": 496
+    },
+    {
+      "epoch": 0.3935269955955857,
+      "grad_norm": 0.2503292332824136,
+      "learning_rate": 1.919158626208884e-06,
+      "loss": 0.603,
+      "step": 497
+    },
+    {
+      "epoch": 0.39431880041569756,
+      "grad_norm": 0.2583634281080707,
+      "learning_rate": 1.9188308584279965e-06,
+      "loss": 0.6029,
+      "step": 498
+    },
+    {
+      "epoch": 0.3951106052358094,
+      "grad_norm": 0.25626058933065626,
+      "learning_rate": 1.9185024556357283e-06,
+      "loss": 0.5797,
+      "step": 499
+    },
+    {
+      "epoch": 0.3959024100559212,
+      "grad_norm": 0.2552322552629375,
+      "learning_rate": 1.9181734180590408e-06,
+      "loss": 0.5891,
+      "step": 500
+    },
+    {
+      "epoch": 0.39669421487603307,
+      "grad_norm": 0.2571482255500502,
+      "learning_rate": 1.917843745925335e-06,
+      "loss": 0.5907,
+      "step": 501
+    },
+    {
+      "epoch": 0.3974860196961449,
+      "grad_norm": 0.2558347636656439,
+      "learning_rate": 1.91751343946245e-06,
+      "loss": 0.606,
+      "step": 502
+    },
+    {
+      "epoch": 0.39827782451625676,
+      "grad_norm": 0.24970008982422848,
+      "learning_rate": 1.917182498898663e-06,
+      "loss": 0.5805,
+      "step": 503
+    },
+    {
+      "epoch": 0.3990696293363686,
+      "grad_norm": 0.24874262897937013,
+      "learning_rate": 1.9168509244626897e-06,
+      "loss": 0.6,
+      "step": 504
+    },
+    {
+      "epoch": 0.39986143415648046,
+      "grad_norm": 0.24498750008115716,
+      "learning_rate": 1.916518716383684e-06,
+      "loss": 0.5853,
+      "step": 505
+    },
+    {
+      "epoch": 0.4006532389765923,
+      "grad_norm": 0.26467126876440167,
+      "learning_rate": 1.916185874891237e-06,
+      "loss": 0.6043,
+      "step": 506
+    },
+    {
+      "epoch": 0.4014450437967041,
+      "grad_norm": 0.25168731029496005,
+      "learning_rate": 1.9158524002153787e-06,
+      "loss": 0.5976,
+      "step": 507
+    },
+    {
+      "epoch": 0.40223684861681597,
+      "grad_norm": 0.24247677785798802,
+      "learning_rate": 1.9155182925865762e-06,
+      "loss": 0.5986,
+      "step": 508
+    },
+    {
+      "epoch": 0.4030286534369278,
+      "grad_norm": 0.24947876029329646,
+      "learning_rate": 1.915183552235734e-06,
+      "loss": 0.5836,
+      "step": 509
+    },
+    {
+      "epoch": 0.40382045825703966,
+      "grad_norm": 0.24712297571012012,
+      "learning_rate": 1.9148481793941933e-06,
+      "loss": 0.5876,
+      "step": 510
+    },
+    {
+      "epoch": 0.4046122630771515,
+      "grad_norm": 0.25468129328604866,
+      "learning_rate": 1.914512174293733e-06,
+      "loss": 0.5981,
+      "step": 511
+    },
+    {
+      "epoch": 0.4054040678972633,
+      "grad_norm": 0.24966577324530015,
+      "learning_rate": 1.9141755371665698e-06,
+      "loss": 0.5908,
+      "step": 512
+    },
+    {
+      "epoch": 0.4061958727173752,
+      "grad_norm": 0.23920241476206486,
+      "learning_rate": 1.9138382682453555e-06,
+      "loss": 0.6034,
+      "step": 513
+    },
+    {
+      "epoch": 0.406987677537487,
+      "grad_norm": 0.24572990955319945,
+      "learning_rate": 1.9135003677631794e-06,
+      "loss": 0.6045,
+      "step": 514
+    },
+    {
+      "epoch": 0.40777948235759887,
+      "grad_norm": 0.25333970267239314,
+      "learning_rate": 1.9131618359535676e-06,
+      "loss": 0.5994,
+      "step": 515
+    },
+    {
+      "epoch": 0.4085712871777107,
+      "grad_norm": 0.2452416194903933,
+      "learning_rate": 1.9128226730504816e-06,
+      "loss": 0.6049,
+      "step": 516
+    },
+    {
+      "epoch": 0.40936309199782256,
+      "grad_norm": 0.2554514592884786,
+      "learning_rate": 1.9124828792883204e-06,
+      "loss": 0.6018,
+      "step": 517
+    },
+    {
+      "epoch": 0.4101548968179344,
+      "grad_norm": 0.25668467435797254,
+      "learning_rate": 1.9121424549019176e-06,
+      "loss": 0.5767,
+      "step": 518
+    },
+    {
+      "epoch": 0.4109467016380462,
+      "grad_norm": 0.2440800558700137,
+      "learning_rate": 1.911801400126544e-06,
+      "loss": 0.5874,
+      "step": 519
+    },
+    {
+      "epoch": 0.41173850645815807,
+      "grad_norm": 0.24443693562509336,
+      "learning_rate": 1.9114597151979037e-06,
+      "loss": 0.6054,
+      "step": 520
+    },
+    {
+      "epoch": 0.4125303112782699,
+      "grad_norm": 0.2533333192166624,
+      "learning_rate": 1.9111174003521394e-06,
+      "loss": 0.6006,
+      "step": 521
+    },
+    {
+      "epoch": 0.41332211609838176,
+      "grad_norm": 0.24246229964500904,
+      "learning_rate": 1.910774455825827e-06,
+      "loss": 0.6008,
+      "step": 522
+    },
+    {
+      "epoch": 0.4141139209184936,
+      "grad_norm": 0.23914994890194086,
+      "learning_rate": 1.910430881855978e-06,
+      "loss": 0.5796,
+      "step": 523
+    },
+    {
+      "epoch": 0.41490572573860546,
+      "grad_norm": 0.2521848372112751,
+      "learning_rate": 1.9100866786800403e-06,
+      "loss": 0.5807,
+      "step": 524
+    },
+    {
+      "epoch": 0.4156975305587173,
+      "grad_norm": 0.2490663970974661,
+      "learning_rate": 1.909741846535894e-06,
+      "loss": 0.586,
+      "step": 525
+    },
+    {
+      "epoch": 0.4164893353788291,
+      "grad_norm": 0.2510211645568367,
+      "learning_rate": 1.909396385661856e-06,
+      "loss": 0.5848,
+      "step": 526
+    },
+    {
+      "epoch": 0.41728114019894097,
+      "grad_norm": 0.2520941601934756,
+      "learning_rate": 1.9090502962966773e-06,
+      "loss": 0.5878,
+      "step": 527
+    },
+    {
+      "epoch": 0.4180729450190528,
+      "grad_norm": 0.2549101828097431,
+      "learning_rate": 1.908703578679543e-06,
+      "loss": 0.5949,
+      "step": 528
+    },
+    {
+      "epoch": 0.41886474983916466,
+      "grad_norm": 0.24544990353717744,
+      "learning_rate": 1.908356233050072e-06,
+      "loss": 0.6007,
+      "step": 529
+    },
+    {
+      "epoch": 0.4196565546592765,
+      "grad_norm": 0.2445800594456821,
+      "learning_rate": 1.908008259648318e-06,
+      "loss": 0.5903,
+      "step": 530
+    },
+    {
+      "epoch": 0.42044835947938836,
+      "grad_norm": 0.25257051310738954,
+      "learning_rate": 1.907659658714768e-06,
+      "loss": 0.5976,
+      "step": 531
+    },
+    {
+      "epoch": 0.4212401642995002,
+      "grad_norm": 0.25430432707918715,
+      "learning_rate": 1.9073104304903426e-06,
+      "loss": 0.5932,
+      "step": 532
+    },
+    {
+      "epoch": 0.422031969119612,
+      "grad_norm": 0.2676141397255204,
+      "learning_rate": 1.9069605752163974e-06,
+      "loss": 0.5919,
+      "step": 533
+    },
+    {
+      "epoch": 0.42282377393972387,
+      "grad_norm": 0.24913566877200974,
+      "learning_rate": 1.906610093134719e-06,
+      "loss": 0.5922,
+      "step": 534
+    },
+    {
+      "epoch": 0.4236155787598357,
+      "grad_norm": 0.25142050866297777,
+      "learning_rate": 1.906258984487529e-06,
+      "loss": 0.5981,
+      "step": 535
+    },
+    {
+      "epoch": 0.42440738357994756,
+      "grad_norm": 0.2626434372822137,
+      "learning_rate": 1.9059072495174808e-06,
+      "loss": 0.6167,
+      "step": 536
+    },
+    {
+      "epoch": 0.4251991884000594,
+      "grad_norm": 0.26204574666443564,
+      "learning_rate": 1.9055548884676622e-06,
+      "loss": 0.5898,
+      "step": 537
+    },
+    {
+      "epoch": 0.42599099322017125,
+      "grad_norm": 0.24349345230704145,
+      "learning_rate": 1.905201901581592e-06,
+      "loss": 0.5989,
+      "step": 538
+    },
+    {
+      "epoch": 0.4267827980402831,
+      "grad_norm": 0.25895765874169857,
+      "learning_rate": 1.9048482891032225e-06,
+      "loss": 0.5865,
+      "step": 539
+    },
+    {
+      "epoch": 0.4275746028603949,
+      "grad_norm": 0.2543447511075663,
+      "learning_rate": 1.9044940512769383e-06,
+      "loss": 0.6044,
+      "step": 540
+    },
+    {
+      "epoch": 0.42836640768050677,
+      "grad_norm": 0.25660293201048157,
+      "learning_rate": 1.9041391883475556e-06,
+      "loss": 0.5971,
+      "step": 541
+    },
+    {
+      "epoch": 0.4291582125006186,
+      "grad_norm": 0.25695030395594404,
+      "learning_rate": 1.9037837005603231e-06,
+      "loss": 0.5771,
+      "step": 542
+    },
+    {
+      "epoch": 0.42995001732073046,
+      "grad_norm": 0.25993304117454247,
+      "learning_rate": 1.9034275881609218e-06,
+      "loss": 0.6093,
+      "step": 543
+    },
+    {
+      "epoch": 0.4307418221408423,
+      "grad_norm": 0.24704428549680987,
+      "learning_rate": 1.9030708513954635e-06,
+      "loss": 0.6085,
+      "step": 544
+    },
+    {
+      "epoch": 0.4315336269609541,
+      "grad_norm": 0.2542518306972636,
+      "learning_rate": 1.9027134905104911e-06,
+      "loss": 0.6047,
+      "step": 545
+    },
+    {
+      "epoch": 0.43232543178106597,
+      "grad_norm": 0.25406990466760127,
+      "learning_rate": 1.9023555057529808e-06,
+      "loss": 0.6003,
+      "step": 546
+    },
+    {
+      "epoch": 0.4331172366011778,
+      "grad_norm": 0.25453026870759954,
+      "learning_rate": 1.9019968973703383e-06,
+      "loss": 0.5934,
+      "step": 547
+    },
+    {
+      "epoch": 0.43390904142128967,
+      "grad_norm": 0.2685469556823121,
+      "learning_rate": 1.9016376656104002e-06,
+      "loss": 0.5842,
+      "step": 548
+    },
+    {
+      "epoch": 0.4347008462414015,
+      "grad_norm": 0.2524811149335431,
+      "learning_rate": 1.9012778107214352e-06,
+      "loss": 0.5886,
+      "step": 549
+    },
+    {
+      "epoch": 0.43549265106151336,
+      "grad_norm": 0.2622934631117297,
+      "learning_rate": 1.9009173329521415e-06,
+      "loss": 0.606,
+      "step": 550
+    },
+    {
+      "epoch": 0.4362844558816252,
+      "grad_norm": 0.2379389352813738,
+      "learning_rate": 1.9005562325516481e-06,
+      "loss": 0.5915,
+      "step": 551
+    },
+    {
+      "epoch": 0.437076260701737,
+      "grad_norm": 0.25381718346731963,
+      "learning_rate": 1.900194509769515e-06,
+      "loss": 0.5928,
+      "step": 552
+    },
+    {
+      "epoch": 0.43786806552184887,
+      "grad_norm": 0.24768690289303336,
+      "learning_rate": 1.8998321648557307e-06,
+      "loss": 0.5966,
+      "step": 553
+    },
+    {
+      "epoch": 0.4386598703419607,
+      "grad_norm": 0.24692537647588123,
+      "learning_rate": 1.8994691980607156e-06,
+      "loss": 0.57,
+      "step": 554
+    },
+    {
+      "epoch": 0.43945167516207256,
+      "grad_norm": 0.2425032291152309,
+      "learning_rate": 1.8991056096353186e-06,
+      "loss": 0.5738,
+      "step": 555
+    },
+    {
+      "epoch": 0.4402434799821844,
+      "grad_norm": 0.24564209012352464,
+      "learning_rate": 1.8987413998308185e-06,
+      "loss": 0.596,
+      "step": 556
+    },
+    {
+      "epoch": 0.44103528480229626,
+      "grad_norm": 0.24818474606241464,
+      "learning_rate": 1.8983765688989235e-06,
+      "loss": 0.5855,
+      "step": 557
+    },
+    {
+      "epoch": 0.4418270896224081,
+      "grad_norm": 0.2544777868752887,
+      "learning_rate": 1.8980111170917716e-06,
+      "loss": 0.5802,
+      "step": 558
+    },
+    {
+      "epoch": 0.4426188944425199,
+      "grad_norm": 0.2518961660630504,
+      "learning_rate": 1.8976450446619293e-06,
+      "loss": 0.5973,
+      "step": 559
+    },
+    {
+      "epoch": 0.44341069926263177,
+      "grad_norm": 0.24486041155981994,
+      "learning_rate": 1.8972783518623921e-06,
+      "loss": 0.6097,
+      "step": 560
+    },
+    {
+      "epoch": 0.4442025040827436,
+      "grad_norm": 0.2477100340771778,
+      "learning_rate": 1.8969110389465845e-06,
+      "loss": 0.5931,
+      "step": 561
+    },
+    {
+      "epoch": 0.44499430890285546,
+      "grad_norm": 0.24337367141849547,
+      "learning_rate": 1.8965431061683593e-06,
+      "loss": 0.5801,
+      "step": 562
+    },
+    {
+      "epoch": 0.4457861137229673,
+      "grad_norm": 0.24453218088999523,
+      "learning_rate": 1.8961745537819978e-06,
+      "loss": 0.5827,
+      "step": 563
+    },
+    {
+      "epoch": 0.44657791854307916,
+      "grad_norm": 0.24861572385826658,
+      "learning_rate": 1.8958053820422093e-06,
+      "loss": 0.5917,
+      "step": 564
+    },
+    {
+      "epoch": 0.447369723363191,
+      "grad_norm": 0.24765696666423848,
+      "learning_rate": 1.8954355912041317e-06,
+      "loss": 0.5738,
+      "step": 565
+    },
+    {
+      "epoch": 0.4481615281833028,
+      "grad_norm": 0.2465393453299667,
+      "learning_rate": 1.8950651815233303e-06,
+      "loss": 0.5938,
+      "step": 566
+    },
+    {
+      "epoch": 0.44895333300341467,
+      "grad_norm": 0.2491034515317017,
+      "learning_rate": 1.8946941532557981e-06,
+      "loss": 0.5891,
+      "step": 567
+    },
+    {
+      "epoch": 0.4497451378235265,
+      "grad_norm": 0.24458817821467171,
+      "learning_rate": 1.894322506657956e-06,
+      "loss": 0.6008,
+      "step": 568
+    },
+    {
+      "epoch": 0.45053694264363836,
+      "grad_norm": 0.24138603479841458,
+      "learning_rate": 1.8939502419866517e-06,
+      "loss": 0.5957,
+      "step": 569
+    },
+    {
+      "epoch": 0.4513287474637502,
+      "grad_norm": 0.2632618023896708,
+      "learning_rate": 1.8935773594991602e-06,
+      "loss": 0.5903,
+      "step": 570
+    },
+    {
+      "epoch": 0.45212055228386205,
+      "grad_norm": 0.2624748591331253,
+      "learning_rate": 1.8932038594531839e-06,
+      "loss": 0.5968,
+      "step": 571
+    },
+    {
+      "epoch": 0.4529123571039739,
+      "grad_norm": 0.26217399555794163,
+      "learning_rate": 1.8928297421068513e-06,
+      "loss": 0.5795,
+      "step": 572
+    },
+    {
+      "epoch": 0.4537041619240857,
+      "grad_norm": 0.25490627556170353,
+      "learning_rate": 1.8924550077187183e-06,
+      "loss": 0.5939,
+      "step": 573
+    },
+    {
+      "epoch": 0.45449596674419757,
+      "grad_norm": 0.2507838695486079,
+      "learning_rate": 1.8920796565477666e-06,
+      "loss": 0.6096,
+      "step": 574
+    },
+    {
+      "epoch": 0.4552877715643094,
+      "grad_norm": 0.24142344523476766,
+      "learning_rate": 1.8917036888534042e-06,
+      "loss": 0.5996,
+      "step": 575
+    },
+    {
+      "epoch": 0.45607957638442126,
+      "grad_norm": 0.25846384854319515,
+      "learning_rate": 1.891327104895466e-06,
+      "loss": 0.5852,
+      "step": 576
+    },
+    {
+      "epoch": 0.4568713812045331,
+      "grad_norm": 0.24743738676208468,
+      "learning_rate": 1.8909499049342116e-06,
+      "loss": 0.5844,
+      "step": 577
+    },
+    {
+      "epoch": 0.45766318602464495,
+      "grad_norm": 0.2513197485371807,
+      "learning_rate": 1.8905720892303271e-06,
+      "loss": 0.5948,
+      "step": 578
+    },
+    {
+      "epoch": 0.45845499084475677,
+      "grad_norm": 0.24926525473931793,
+      "learning_rate": 1.8901936580449245e-06,
+      "loss": 0.5819,
+      "step": 579
+    },
+    {
+      "epoch": 0.4592467956648686,
+      "grad_norm": 0.2552388048276722,
+      "learning_rate": 1.8898146116395398e-06,
+      "loss": 0.5974,
+      "step": 580
+    },
+    {
+      "epoch": 0.46003860048498046,
+      "grad_norm": 0.25263820247307917,
+      "learning_rate": 1.8894349502761355e-06,
+      "loss": 0.5762,
+      "step": 581
+    },
+    {
+      "epoch": 0.4608304053050923,
+      "grad_norm": 0.2553733533701803,
+      "learning_rate": 1.8890546742170983e-06,
+      "loss": 0.6036,
+      "step": 582
+    },
+    {
+      "epoch": 0.46162221012520416,
+      "grad_norm": 0.2547596600415412,
+      "learning_rate": 1.8886737837252405e-06,
+      "loss": 0.5796,
+      "step": 583
+    },
+    {
+      "epoch": 0.462414014945316,
+      "grad_norm": 0.2513121780285386,
+      "learning_rate": 1.8882922790637981e-06,
+      "loss": 0.6148,
+      "step": 584
+    },
+    {
+      "epoch": 0.4632058197654278,
+      "grad_norm": 0.26739457310607195,
+      "learning_rate": 1.8879101604964322e-06,
+      "loss": 0.5939,
+      "step": 585
+    },
+    {
+      "epoch": 0.46399762458553967,
+      "grad_norm": 0.25951718112872185,
+      "learning_rate": 1.8875274282872281e-06,
+      "loss": 0.6,
+      "step": 586
+    },
+    {
+      "epoch": 0.4647894294056515,
+      "grad_norm": 0.25631245577556405,
+      "learning_rate": 1.887144082700695e-06,
+      "loss": 0.5859,
+      "step": 587
+    },
+    {
+      "epoch": 0.46558123422576336,
+      "grad_norm": 0.2564914517999274,
+      "learning_rate": 1.8867601240017663e-06,
+      "loss": 0.5853,
+      "step": 588
+    },
+    {
+      "epoch": 0.4663730390458752,
+      "grad_norm": 0.24819504652120772,
+      "learning_rate": 1.8863755524557987e-06,
+      "loss": 0.5941,
+      "step": 589
+    },
+    {
+      "epoch": 0.46716484386598706,
+      "grad_norm": 0.2511821237330362,
+      "learning_rate": 1.885990368328573e-06,
+      "loss": 0.6003,
+      "step": 590
+    },
+    {
+      "epoch": 0.4679566486860989,
+      "grad_norm": 0.2470050771408181,
+      "learning_rate": 1.8856045718862922e-06,
+      "loss": 0.6006,
+      "step": 591
+    },
+    {
+      "epoch": 0.4687484535062107,
+      "grad_norm": 0.2432408248373072,
+      "learning_rate": 1.8852181633955846e-06,
+      "loss": 0.5868,
+      "step": 592
+    },
+    {
+      "epoch": 0.46954025832632257,
+      "grad_norm": 0.2525747260022962,
+      "learning_rate": 1.8848311431234992e-06,
+      "loss": 0.5801,
+      "step": 593
+    },
+    {
+      "epoch": 0.4703320631464344,
+      "grad_norm": 0.261657075936512,
+      "learning_rate": 1.8844435113375093e-06,
+      "loss": 0.5759,
+      "step": 594
+    },
+    {
+      "epoch": 0.47112386796654626,
+      "grad_norm": 0.2642731371394818,
+      "learning_rate": 1.8840552683055102e-06,
+      "loss": 0.5749,
+      "step": 595
+    },
+    {
+      "epoch": 0.4719156727866581,
+      "grad_norm": 0.2670554515440983,
+      "learning_rate": 1.8836664142958197e-06,
+      "loss": 0.5734,
+      "step": 596
+    },
+    {
+      "epoch": 0.47270747760676995,
+      "grad_norm": 0.2550154162106024,
+      "learning_rate": 1.8832769495771782e-06,
+      "loss": 0.5713,
+      "step": 597
+    },
+    {
+      "epoch": 0.4734992824268818,
+      "grad_norm": 0.24938960319128795,
+      "learning_rate": 1.882886874418747e-06,
+      "loss": 0.6046,
+      "step": 598
+    },
+    {
+      "epoch": 0.4742910872469936,
+      "grad_norm": 0.2473340062743241,
+      "learning_rate": 1.8824961890901116e-06,
+      "loss": 0.5819,
+      "step": 599
+    },
+    {
+      "epoch": 0.47508289206710547,
+      "grad_norm": 0.2503894664133491,
+      "learning_rate": 1.8821048938612769e-06,
+      "loss": 0.5971,
+      "step": 600
+    },
+    {
+      "epoch": 0.4758746968872173,
+      "grad_norm": 0.2593563375135493,
+      "learning_rate": 1.8817129890026702e-06,
+      "loss": 0.6008,
+      "step": 601
+    },
+    {
+      "epoch": 0.47666650170732916,
+      "grad_norm": 0.2518731330101114,
+      "learning_rate": 1.8813204747851403e-06,
+      "loss": 0.5929,
+      "step": 602
+    },
+    {
+      "epoch": 0.477458306527441,
+      "grad_norm": 0.2534468089092246,
+      "learning_rate": 1.8809273514799569e-06,
+      "loss": 0.572,
+      "step": 603
+    },
+    {
+      "epoch": 0.47825011134755285,
+      "grad_norm": 0.2550315788078505,
+      "learning_rate": 1.8805336193588103e-06,
+      "loss": 0.6024,
+      "step": 604
+    },
+    {
+      "epoch": 0.47904191616766467,
+      "grad_norm": 0.25608653703733064,
+      "learning_rate": 1.8801392786938125e-06,
+      "loss": 0.5975,
+      "step": 605
+    },
+    {
+      "epoch": 0.4798337209877765,
+      "grad_norm": 0.26515830248195815,
+      "learning_rate": 1.8797443297574952e-06,
+      "loss": 0.5978,
+      "step": 606
+    },
+    {
+      "epoch": 0.48062552580788837,
+      "grad_norm": 0.2598031081750377,
+      "learning_rate": 1.8793487728228104e-06,
+      "loss": 0.5876,
+      "step": 607
+    },
+    {
+      "epoch": 0.4814173306280002,
+      "grad_norm": 0.2579622703822964,
+      "learning_rate": 1.8789526081631313e-06,
+      "loss": 0.5927,
+      "step": 608
+    },
+    {
+      "epoch": 0.48220913544811206,
+      "grad_norm": 0.24929178968775514,
+      "learning_rate": 1.8785558360522505e-06,
+      "loss": 0.5795,
+      "step": 609
+    },
+    {
+      "epoch": 0.4830009402682239,
+      "grad_norm": 0.24609387295872273,
+      "learning_rate": 1.87815845676438e-06,
+      "loss": 0.5887,
+      "step": 610
+    },
+    {
+      "epoch": 0.48379274508833575,
+      "grad_norm": 0.2599644480370626,
+      "learning_rate": 1.877760470574152e-06,
+      "loss": 0.5918,
+      "step": 611
+    },
+    {
+      "epoch": 0.48458454990844757,
+      "grad_norm": 0.25322423338767264,
+      "learning_rate": 1.8773618777566176e-06,
+      "loss": 0.6005,
+      "step": 612
+    },
+    {
+      "epoch": 0.4853763547285594,
+      "grad_norm": 0.24765658323912063,
+      "learning_rate": 1.8769626785872483e-06,
+      "loss": 0.5802,
+      "step": 613
+    },
+    {
+      "epoch": 0.48616815954867126,
+      "grad_norm": 0.25864308269449704,
+      "learning_rate": 1.876562873341933e-06,
+      "loss": 0.5964,
+      "step": 614
+    },
+    {
+      "epoch": 0.4869599643687831,
+      "grad_norm": 0.2406859239191986,
+      "learning_rate": 1.8761624622969807e-06,
+      "loss": 0.5824,
+      "step": 615
+    },
+    {
+      "epoch": 0.48775176918889496,
+      "grad_norm": 0.25436975707855747,
+      "learning_rate": 1.8757614457291187e-06,
+      "loss": 0.5832,
+      "step": 616
+    },
+    {
+      "epoch": 0.4885435740090068,
+      "grad_norm": 0.26589957568490175,
+      "learning_rate": 1.8753598239154926e-06,
+      "loss": 0.5902,
+      "step": 617
+    },
+    {
+      "epoch": 0.4893353788291186,
+      "grad_norm": 0.2823144704981466,
+      "learning_rate": 1.8749575971336672e-06,
+      "loss": 0.6006,
+      "step": 618
+    },
+    {
+      "epoch": 0.49012718364923047,
+      "grad_norm": 0.266502255384463,
+      "learning_rate": 1.874554765661623e-06,
+      "loss": 0.5925,
+      "step": 619
+    },
+    {
+      "epoch": 0.4909189884693423,
+      "grad_norm": 0.25868494586041696,
+      "learning_rate": 1.8741513297777619e-06,
+      "loss": 0.5848,
+      "step": 620
+    },
+    {
+      "epoch": 0.49171079328945416,
+      "grad_norm": 0.24420772635871477,
+      "learning_rate": 1.8737472897609008e-06,
+      "loss": 0.5858,
+      "step": 621
+    },
+    {
+      "epoch": 0.492502598109566,
+      "grad_norm": 0.25686435545350483,
+      "learning_rate": 1.8733426458902748e-06,
+      "loss": 0.5829,
+      "step": 622
+    },
+    {
+      "epoch": 0.49329440292967786,
+      "grad_norm": 0.2555585620025713,
+      "learning_rate": 1.872937398445537e-06,
+      "loss": 0.5967,
+      "step": 623
+    },
+    {
+      "epoch": 0.4940862077497897,
+      "grad_norm": 0.24698406765043235,
+      "learning_rate": 1.872531547706757e-06,
+      "loss": 0.5834,
+      "step": 624
+    },
+    {
+      "epoch": 0.4948780125699015,
+      "grad_norm": 0.2588839733704809,
+      "learning_rate": 1.8721250939544212e-06,
+      "loss": 0.6028,
+      "step": 625
+    },
+    {
+      "epoch": 0.49566981739001337,
+      "grad_norm": 0.2577950792786317,
+      "learning_rate": 1.8717180374694336e-06,
+      "loss": 0.5899,
+      "step": 626
+    },
+    {
+      "epoch": 0.4964616222101252,
+      "grad_norm": 0.2437095818796854,
+      "learning_rate": 1.871310378533114e-06,
+      "loss": 0.5823,
+      "step": 627
+    },
+    {
+      "epoch": 0.49725342703023706,
+      "grad_norm": 0.24544352768150463,
+      "learning_rate": 1.8709021174271982e-06,
+      "loss": 0.5886,
+      "step": 628
+    },
+    {
+      "epoch": 0.4980452318503489,
+      "grad_norm": 0.2607396835887682,
+      "learning_rate": 1.8704932544338395e-06,
+      "loss": 0.5958,
+      "step": 629
+    },
+    {
+      "epoch": 0.49883703667046075,
+      "grad_norm": 0.24314494711697102,
+      "learning_rate": 1.8700837898356063e-06,
+      "loss": 0.5957,
+      "step": 630
+    },
+    {
+      "epoch": 0.4996288414905726,
+      "grad_norm": 0.2548555939972503,
+      "learning_rate": 1.8696737239154823e-06,
+      "loss": 0.5828,
+      "step": 631
+    },
+    {
+      "epoch": 0.5004206463106844,
+      "grad_norm": 0.2507961409164316,
+      "learning_rate": 1.8692630569568682e-06,
+      "loss": 0.5878,
+      "step": 632
+    },
+    {
+      "epoch": 0.5012124511307963,
+      "grad_norm": 0.24630360548273397,
+      "learning_rate": 1.8688517892435782e-06,
+      "loss": 0.5811,
+      "step": 633
+    },
+    {
+      "epoch": 0.5020042559509081,
+      "grad_norm": 0.2583300839160151,
+      "learning_rate": 1.8684399210598438e-06,
+      "loss": 0.595,
+      "step": 634
+    },
+    {
+      "epoch": 0.5027960607710199,
+      "grad_norm": 0.24514303969134071,
+      "learning_rate": 1.8680274526903094e-06,
+      "loss": 0.5882,
+      "step": 635
+    },
+    {
+      "epoch": 0.5035878655911318,
+      "grad_norm": 0.25767312521601865,
+      "learning_rate": 1.867614384420036e-06,
+      "loss": 0.5838,
+      "step": 636
+    },
+    {
+      "epoch": 0.5043796704112437,
+      "grad_norm": 0.2637330975882953,
+      "learning_rate": 1.867200716534498e-06,
+      "loss": 0.605,
+      "step": 637
+    },
+    {
+      "epoch": 0.5051714752313554,
+      "grad_norm": 0.25797054438517375,
+      "learning_rate": 1.8667864493195848e-06,
+      "loss": 0.5871,
+      "step": 638
+    },
+    {
+      "epoch": 0.5059632800514673,
+      "grad_norm": 0.2544771462188513,
+      "learning_rate": 1.8663715830615996e-06,
+      "loss": 0.5956,
+      "step": 639
+    },
+    {
+      "epoch": 0.5067550848715792,
+      "grad_norm": 0.26136970915324076,
+      "learning_rate": 1.8659561180472598e-06,
+      "loss": 0.5904,
+      "step": 640
+    },
+    {
+      "epoch": 0.507546889691691,
+      "grad_norm": 0.2644633648378998,
+      "learning_rate": 1.8655400545636968e-06,
+      "loss": 0.5931,
+      "step": 641
+    },
+    {
+      "epoch": 0.5083386945118028,
+      "grad_norm": 0.25217107553398144,
+      "learning_rate": 1.8651233928984551e-06,
+      "loss": 0.5818,
+      "step": 642
+    },
+    {
+      "epoch": 0.5091304993319147,
+      "grad_norm": 0.2612370653410413,
+      "learning_rate": 1.8647061333394932e-06,
+      "loss": 0.5878,
+      "step": 643
+    },
+    {
+      "epoch": 0.5099223041520266,
+      "grad_norm": 0.25112176662694013,
+      "learning_rate": 1.8642882761751826e-06,
+      "loss": 0.5792,
+      "step": 644
+    },
+    {
+      "epoch": 0.5107141089721383,
+      "grad_norm": 0.2553203153129554,
+      "learning_rate": 1.8638698216943074e-06,
+      "loss": 0.5855,
+      "step": 645
+    },
+    {
+      "epoch": 0.5115059137922502,
+      "grad_norm": 0.257605510384718,
+      "learning_rate": 1.8634507701860651e-06,
+      "loss": 0.5838,
+      "step": 646
+    },
+    {
+      "epoch": 0.5122977186123621,
+      "grad_norm": 0.24739133493766602,
+      "learning_rate": 1.8630311219400655e-06,
+      "loss": 0.5918,
+      "step": 647
+    },
+    {
+      "epoch": 0.5130895234324739,
+      "grad_norm": 0.25614233537193654,
+      "learning_rate": 1.8626108772463307e-06,
+      "loss": 0.5998,
+      "step": 648
+    },
+    {
+      "epoch": 0.5138813282525857,
+      "grad_norm": 0.2593901016358386,
+      "learning_rate": 1.8621900363952957e-06,
+      "loss": 0.5771,
+      "step": 649
+    },
+    {
+      "epoch": 0.5146731330726976,
+      "grad_norm": 0.2580921998074504,
+      "learning_rate": 1.8617685996778066e-06,
+      "loss": 0.5942,
+      "step": 650
+    },
+    {
+      "epoch": 0.5154649378928094,
+      "grad_norm": 0.24518934638594403,
+      "learning_rate": 1.8613465673851214e-06,
+      "loss": 0.5781,
+      "step": 651
+    },
+    {
+      "epoch": 0.5162567427129212,
+      "grad_norm": 0.2684980696453002,
+      "learning_rate": 1.8609239398089105e-06,
+      "loss": 0.5944,
+      "step": 652
+    },
+    {
+      "epoch": 0.5170485475330331,
+      "grad_norm": 0.2467354031644924,
+      "learning_rate": 1.8605007172412553e-06,
+      "loss": 0.5981,
+      "step": 653
+    },
+    {
+      "epoch": 0.517840352353145,
+      "grad_norm": 0.2551470866172334,
+      "learning_rate": 1.8600768999746477e-06,
+      "loss": 0.5777,
+      "step": 654
+    },
+    {
+      "epoch": 0.5186321571732568,
+      "grad_norm": 0.2547867933652486,
+      "learning_rate": 1.8596524883019918e-06,
+      "loss": 0.6042,
+      "step": 655
+    },
+    {
+      "epoch": 0.5194239619933686,
+      "grad_norm": 0.258259134300438,
+      "learning_rate": 1.8592274825166015e-06,
+      "loss": 0.5829,
+      "step": 656
+    },
+    {
+      "epoch": 0.5202157668134805,
+      "grad_norm": 0.24675317898231072,
+      "learning_rate": 1.8588018829122021e-06,
+      "loss": 0.5841,
+      "step": 657
+    },
+    {
+      "epoch": 0.5210075716335923,
+      "grad_norm": 0.26642682255162203,
+      "learning_rate": 1.8583756897829285e-06,
+      "loss": 0.5868,
+      "step": 658
+    },
+    {
+      "epoch": 0.5217993764537041,
+      "grad_norm": 0.2680952507866509,
+      "learning_rate": 1.8579489034233264e-06,
+      "loss": 0.5781,
+      "step": 659
+    },
+    {
+      "epoch": 0.522591181273816,
+      "grad_norm": 0.2692078025041211,
+      "learning_rate": 1.8575215241283515e-06,
+      "loss": 0.6057,
+      "step": 660
+    },
+    {
+      "epoch": 0.5233829860939279,
+      "grad_norm": 0.25931733304226906,
+      "learning_rate": 1.8570935521933686e-06,
+      "loss": 0.5826,
+      "step": 661
+    },
+    {
+      "epoch": 0.5241747909140397,
+      "grad_norm": 0.2541083319258514,
+      "learning_rate": 1.856664987914153e-06,
+      "loss": 0.5786,
+      "step": 662
+    },
+    {
+      "epoch": 0.5249665957341515,
+      "grad_norm": 0.2661944097834986,
+      "learning_rate": 1.8562358315868886e-06,
+      "loss": 0.5887,
+      "step": 663
+    },
+    {
+      "epoch": 0.5257584005542634,
+      "grad_norm": 0.276640962139248,
+      "learning_rate": 1.8558060835081688e-06,
+      "loss": 0.5866,
+      "step": 664
+    },
+    {
+      "epoch": 0.5265502053743752,
+      "grad_norm": 0.2592799780279227,
+      "learning_rate": 1.8553757439749962e-06,
+      "loss": 0.5855,
+      "step": 665
+    },
+    {
+      "epoch": 0.527342010194487,
+      "grad_norm": 0.24091040206785624,
+      "learning_rate": 1.8549448132847814e-06,
+      "loss": 0.5862,
+      "step": 666
+    },
+    {
+      "epoch": 0.5281338150145989,
+      "grad_norm": 0.24197516449912457,
+      "learning_rate": 1.8545132917353445e-06,
+      "loss": 0.6064,
+      "step": 667
+    },
+    {
+      "epoch": 0.5289256198347108,
+      "grad_norm": 0.253739866990384,
+      "learning_rate": 1.8540811796249134e-06,
+      "loss": 0.5791,
+      "step": 668
+    },
+    {
+      "epoch": 0.5297174246548226,
+      "grad_norm": 0.24467570017073437,
+      "learning_rate": 1.8536484772521243e-06,
+      "loss": 0.5718,
+      "step": 669
+    },
+    {
+      "epoch": 0.5305092294749344,
+      "grad_norm": 0.24056093726754427,
+      "learning_rate": 1.8532151849160212e-06,
+      "loss": 0.5768,
+      "step": 670
+    },
+    {
+      "epoch": 0.5313010342950463,
+      "grad_norm": 0.24983386225648294,
+      "learning_rate": 1.852781302916056e-06,
+      "loss": 0.5943,
+      "step": 671
+    },
+    {
+      "epoch": 0.5320928391151581,
+      "grad_norm": 0.2594596210177295,
+      "learning_rate": 1.8523468315520878e-06,
+      "loss": 0.5929,
+      "step": 672
+    },
+    {
+      "epoch": 0.5328846439352699,
+      "grad_norm": 0.2477292597466106,
+      "learning_rate": 1.8519117711243836e-06,
+      "loss": 0.5986,
+      "step": 673
+    },
+    {
+      "epoch": 0.5336764487553818,
+      "grad_norm": 0.23765886344255474,
+      "learning_rate": 1.8514761219336167e-06,
+      "loss": 0.5707,
+      "step": 674
+    },
+    {
+      "epoch": 0.5344682535754937,
+      "grad_norm": 0.24901878692343485,
+      "learning_rate": 1.851039884280868e-06,
+      "loss": 0.5996,
+      "step": 675
+    },
+    {
+      "epoch": 0.5352600583956055,
+      "grad_norm": 0.24421886070902452,
+      "learning_rate": 1.8506030584676253e-06,
+      "loss": 0.5848,
+      "step": 676
+    },
+    {
+      "epoch": 0.5360518632157173,
+      "grad_norm": 0.24280405428062857,
+      "learning_rate": 1.8501656447957817e-06,
+      "loss": 0.5844,
+      "step": 677
+    },
+    {
+      "epoch": 0.5368436680358292,
+      "grad_norm": 0.25268297516887805,
+      "learning_rate": 1.8497276435676375e-06,
+      "loss": 0.5879,
+      "step": 678
+    },
+    {
+      "epoch": 0.537635472855941,
+      "grad_norm": 0.25323598508216055,
+      "learning_rate": 1.8492890550858991e-06,
+      "loss": 0.5749,
+      "step": 679
+    },
+    {
+      "epoch": 0.5384272776760528,
+      "grad_norm": 0.2469965783028501,
+      "learning_rate": 1.8488498796536782e-06,
+      "loss": 0.6041,
+      "step": 680
+    },
+    {
+      "epoch": 0.5392190824961647,
+      "grad_norm": 0.2456833497636686,
+      "learning_rate": 1.8484101175744926e-06,
+      "loss": 0.5829,
+      "step": 681
+    },
+    {
+      "epoch": 0.5400108873162766,
+      "grad_norm": 0.23916202458714156,
+      "learning_rate": 1.8479697691522657e-06,
+      "loss": 0.5711,
+      "step": 682
+    },
+    {
+      "epoch": 0.5408026921363884,
+      "grad_norm": 0.25289719726361376,
+      "learning_rate": 1.8475288346913253e-06,
+      "loss": 0.5968,
+      "step": 683
+    },
+    {
+      "epoch": 0.5415944969565002,
+      "grad_norm": 0.26543836193385023,
+      "learning_rate": 1.847087314496405e-06,
+      "loss": 0.5929,
+      "step": 684
+    },
+    {
+      "epoch": 0.5423863017766121,
+      "grad_norm": 0.2746509392353278,
+      "learning_rate": 1.8466452088726431e-06,
+      "loss": 0.5916,
+      "step": 685
+    },
+    {
+      "epoch": 0.5431781065967239,
+      "grad_norm": 0.2486144683867524,
+      "learning_rate": 1.8462025181255822e-06,
+      "loss": 0.5935,
+      "step": 686
+    },
+    {
+      "epoch": 0.5439699114168357,
+      "grad_norm": 0.2572507467885107,
+      "learning_rate": 1.8457592425611694e-06,
+      "loss": 0.5856,
+      "step": 687
+    },
+    {
+      "epoch": 0.5447617162369476,
+      "grad_norm": 0.25691373702239123,
+      "learning_rate": 1.845315382485756e-06,
+      "loss": 0.5806,
+      "step": 688
+    },
+    {
+      "epoch": 0.5455535210570595,
+      "grad_norm": 0.2697588241292833,
+      "learning_rate": 1.8448709382060975e-06,
+      "loss": 0.5822,
+      "step": 689
+    },
+    {
+      "epoch": 0.5463453258771713,
+      "grad_norm": 0.25767372746150385,
+      "learning_rate": 1.8444259100293523e-06,
+      "loss": 0.5921,
+      "step": 690
+    },
+    {
+      "epoch": 0.5471371306972831,
+      "grad_norm": 0.24439659052624355,
+      "learning_rate": 1.8439802982630837e-06,
+      "loss": 0.6003,
+      "step": 691
+    },
+    {
+      "epoch": 0.547928935517395,
+      "grad_norm": 0.24572923347720313,
+      "learning_rate": 1.8435341032152566e-06,
+      "loss": 0.5978,
+      "step": 692
+    },
+    {
+      "epoch": 0.5487207403375068,
+      "grad_norm": 0.26320713917576716,
+      "learning_rate": 1.8430873251942414e-06,
+      "loss": 0.5826,
+      "step": 693
+    },
+    {
+      "epoch": 0.5495125451576186,
+      "grad_norm": 0.26980082421070706,
+      "learning_rate": 1.842639964508809e-06,
+      "loss": 0.5841,
+      "step": 694
+    },
+    {
+      "epoch": 0.5503043499777305,
+      "grad_norm": 0.24912616528732986,
+      "learning_rate": 1.8421920214681336e-06,
+      "loss": 0.5676,
+      "step": 695
+    },
+    {
+      "epoch": 0.5510961547978424,
+      "grad_norm": 0.25120583657924267,
+      "learning_rate": 1.841743496381793e-06,
+      "loss": 0.5812,
+      "step": 696
+    },
+    {
+      "epoch": 0.5518879596179542,
+      "grad_norm": 0.249207112222779,
+      "learning_rate": 1.8412943895597665e-06,
+      "loss": 0.585,
+      "step": 697
+    },
+    {
+      "epoch": 0.552679764438066,
+      "grad_norm": 0.2678058120982978,
+      "learning_rate": 1.8408447013124352e-06,
+      "loss": 0.5886,
+      "step": 698
+    },
+    {
+      "epoch": 0.5534715692581779,
+      "grad_norm": 0.2552786391349403,
+      "learning_rate": 1.8403944319505817e-06,
+      "loss": 0.5767,
+      "step": 699
+    },
+    {
+      "epoch": 0.5542633740782897,
+      "grad_norm": 0.2465657211337802,
+      "learning_rate": 1.8399435817853916e-06,
+      "loss": 0.5787,
+      "step": 700
+    },
+    {
+      "epoch": 0.5550551788984015,
+      "grad_norm": 0.2499739623405065,
+      "learning_rate": 1.839492151128451e-06,
+      "loss": 0.5882,
+      "step": 701
+    },
+    {
+      "epoch": 0.5558469837185134,
+      "grad_norm": 0.2599297446070384,
+      "learning_rate": 1.8390401402917467e-06,
+      "loss": 0.5834,
+      "step": 702
+    },
+    {
+      "epoch": 0.5566387885386253,
+      "grad_norm": 0.25255278534208114,
+      "learning_rate": 1.8385875495876678e-06,
+      "loss": 0.5988,
+      "step": 703
+    },
+    {
+      "epoch": 0.5574305933587371,
+      "grad_norm": 0.26171628996005464,
+      "learning_rate": 1.8381343793290025e-06,
+      "loss": 0.5885,
+      "step": 704
+    },
+    {
+      "epoch": 0.5582223981788489,
+      "grad_norm": 0.251767699867137,
+      "learning_rate": 1.8376806298289407e-06,
+      "loss": 0.5851,
+      "step": 705
+    },
+    {
+      "epoch": 0.5590142029989608,
+      "grad_norm": 0.22989634696864383,
+      "learning_rate": 1.8372263014010726e-06,
+      "loss": 0.5972,
+      "step": 706
+    },
+    {
+      "epoch": 0.5598060078190726,
+      "grad_norm": 0.2516276753539971,
+      "learning_rate": 1.8367713943593882e-06,
+      "loss": 0.5768,
+      "step": 707
+    },
+    {
+      "epoch": 0.5605978126391844,
+      "grad_norm": 0.2507232449803271,
+      "learning_rate": 1.836315909018277e-06,
+      "loss": 0.6022,
+      "step": 708
+    },
+    {
+      "epoch": 0.5613896174592963,
+      "grad_norm": 0.24432084272529186,
+      "learning_rate": 1.8358598456925288e-06,
+      "loss": 0.5714,
+      "step": 709
+    },
+    {
+      "epoch": 0.5621814222794081,
+      "grad_norm": 0.23384148212599462,
+      "learning_rate": 1.8354032046973328e-06,
+      "loss": 0.5861,
+      "step": 710
+    },
+    {
+      "epoch": 0.5629732270995199,
+      "grad_norm": 0.250622420279018,
+      "learning_rate": 1.8349459863482772e-06,
+      "loss": 0.5801,
+      "step": 711
+    },
+    {
+      "epoch": 0.5637650319196318,
+      "grad_norm": 0.24902384594374952,
+      "learning_rate": 1.8344881909613493e-06,
+      "loss": 0.5961,
+      "step": 712
+    },
+    {
+      "epoch": 0.5645568367397437,
+      "grad_norm": 0.24569769352466586,
+      "learning_rate": 1.8340298188529347e-06,
+      "loss": 0.5862,
+      "step": 713
+    },
+    {
+      "epoch": 0.5653486415598555,
+      "grad_norm": 0.24281376574785227,
+      "learning_rate": 1.8335708703398185e-06,
+      "loss": 0.5748,
+      "step": 714
+    },
+    {
+      "epoch": 0.5661404463799673,
+      "grad_norm": 0.23713171840827416,
+      "learning_rate": 1.8331113457391837e-06,
+      "loss": 0.5933,
+      "step": 715
+    },
+    {
+      "epoch": 0.5669322512000792,
+      "grad_norm": 0.24520199763855147,
+      "learning_rate": 1.832651245368611e-06,
+      "loss": 0.5958,
+      "step": 716
+    },
+    {
+      "epoch": 0.567724056020191,
+      "grad_norm": 0.23584122838719365,
+      "learning_rate": 1.8321905695460802e-06,
+      "loss": 0.5837,
+      "step": 717
+    },
+    {
+      "epoch": 0.5685158608403028,
+      "grad_norm": 0.2548975334264314,
+      "learning_rate": 1.8317293185899676e-06,
+      "loss": 0.5888,
+      "step": 718
+    },
+    {
+      "epoch": 0.5693076656604147,
+      "grad_norm": 0.24645152919598493,
+      "learning_rate": 1.8312674928190472e-06,
+      "loss": 0.5916,
+      "step": 719
+    },
+    {
+      "epoch": 0.5700994704805266,
+      "grad_norm": 0.2508435342752548,
+      "learning_rate": 1.8308050925524912e-06,
+      "loss": 0.6117,
+      "step": 720
+    },
+    {
+      "epoch": 0.5708912753006384,
+      "grad_norm": 0.2541834980979583,
+      "learning_rate": 1.8303421181098676e-06,
+      "loss": 0.5857,
+      "step": 721
+    },
+    {
+      "epoch": 0.5716830801207502,
+      "grad_norm": 0.24482232197513573,
+      "learning_rate": 1.8298785698111415e-06,
+      "loss": 0.6008,
+      "step": 722
+    },
+    {
+      "epoch": 0.5724748849408621,
+      "grad_norm": 0.257618849929514,
+      "learning_rate": 1.8294144479766756e-06,
+      "loss": 0.5841,
+      "step": 723
+    },
+    {
+      "epoch": 0.573266689760974,
+      "grad_norm": 0.23694378769102648,
+      "learning_rate": 1.8289497529272277e-06,
+      "loss": 0.5856,
+      "step": 724
+    },
+    {
+      "epoch": 0.5740584945810857,
+      "grad_norm": 0.2559523294309274,
+      "learning_rate": 1.8284844849839526e-06,
+      "loss": 0.5939,
+      "step": 725
+    },
+    {
+      "epoch": 0.5748502994011976,
+      "grad_norm": 0.2430474922964609,
+      "learning_rate": 1.8280186444684005e-06,
+      "loss": 0.5735,
+      "step": 726
+    },
+    {
+      "epoch": 0.5756421042213095,
+      "grad_norm": 0.2421389562029136,
+      "learning_rate": 1.8275522317025174e-06,
+      "loss": 0.5863,
+      "step": 727
+    },
+    {
+      "epoch": 0.5764339090414213,
+      "grad_norm": 0.24719382586531582,
+      "learning_rate": 1.8270852470086452e-06,
+      "loss": 0.5793,
+      "step": 728
+    },
+    {
+      "epoch": 0.5772257138615331,
+      "grad_norm": 0.2613922322447402,
+      "learning_rate": 1.8266176907095206e-06,
+      "loss": 0.583,
+      "step": 729
+    },
+    {
+      "epoch": 0.578017518681645,
+      "grad_norm": 0.24985652742114342,
+      "learning_rate": 1.8261495631282756e-06,
+      "loss": 0.5867,
+      "step": 730
+    },
+    {
+      "epoch": 0.5788093235017568,
+      "grad_norm": 0.2509190981231187,
+      "learning_rate": 1.8256808645884365e-06,
+      "loss": 0.5842,
+      "step": 731
+    },
+    {
+      "epoch": 0.5796011283218686,
+      "grad_norm": 0.25873515370083255,
+      "learning_rate": 1.8252115954139255e-06,
+      "loss": 0.5866,
+      "step": 732
+    },
+    {
+      "epoch": 0.5803929331419805,
+      "grad_norm": 0.24999759706019414,
+      "learning_rate": 1.8247417559290575e-06,
+      "loss": 0.6023,
+      "step": 733
+    },
+    {
+      "epoch": 0.5811847379620924,
+      "grad_norm": 0.24960925570319883,
+      "learning_rate": 1.8242713464585426e-06,
+      "loss": 0.5838,
+      "step": 734
+    },
+    {
+      "epoch": 0.5819765427822042,
+      "grad_norm": 0.24836711610313497,
+      "learning_rate": 1.8238003673274847e-06,
+      "loss": 0.5827,
+      "step": 735
+    },
+    {
+      "epoch": 0.582768347602316,
+      "grad_norm": 0.24643437314791616,
+      "learning_rate": 1.8233288188613814e-06,
+      "loss": 0.5913,
+      "step": 736
+    },
+    {
+      "epoch": 0.5835601524224279,
+      "grad_norm": 0.26687565546039715,
+      "learning_rate": 1.8228567013861228e-06,
+      "loss": 0.5839,
+      "step": 737
+    },
+    {
+      "epoch": 0.5843519572425397,
+      "grad_norm": 0.24424129443944192,
+      "learning_rate": 1.8223840152279942e-06,
+      "loss": 0.5771,
+      "step": 738
+    },
+    {
+      "epoch": 0.5851437620626515,
+      "grad_norm": 0.2471018097029849,
+      "learning_rate": 1.821910760713672e-06,
+      "loss": 0.5777,
+      "step": 739
+    },
+    {
+      "epoch": 0.5859355668827634,
+      "grad_norm": 0.2547865860289802,
+      "learning_rate": 1.821436938170226e-06,
+      "loss": 0.6005,
+      "step": 740
+    },
+    {
+      "epoch": 0.5867273717028753,
+      "grad_norm": 0.25810692590791984,
+      "learning_rate": 1.82096254792512e-06,
+      "loss": 0.5918,
+      "step": 741
+    },
+    {
+      "epoch": 0.5875191765229871,
+      "grad_norm": 0.2534896323977627,
+      "learning_rate": 1.8204875903062078e-06,
+      "loss": 0.5781,
+      "step": 742
+    },
+    {
+      "epoch": 0.5883109813430989,
+      "grad_norm": 0.250310436243135,
+      "learning_rate": 1.8200120656417365e-06,
+      "loss": 0.5936,
+      "step": 743
+    },
+    {
+      "epoch": 0.5891027861632108,
+      "grad_norm": 0.25846177780365504,
+      "learning_rate": 1.8195359742603455e-06,
+      "loss": 0.5808,
+      "step": 744
+    },
+    {
+      "epoch": 0.5898945909833226,
+      "grad_norm": 0.2434882436517003,
+      "learning_rate": 1.8190593164910651e-06,
+      "loss": 0.5859,
+      "step": 745
+    },
+    {
+      "epoch": 0.5906863958034344,
+      "grad_norm": 0.27177810679363984,
+      "learning_rate": 1.8185820926633173e-06,
+      "loss": 0.6042,
+      "step": 746
+    },
+    {
+      "epoch": 0.5914782006235463,
+      "grad_norm": 0.2586591937858233,
+      "learning_rate": 1.8181043031069155e-06,
+      "loss": 0.5857,
+      "step": 747
+    },
+    {
+      "epoch": 0.5922700054436582,
+      "grad_norm": 0.2573533284689728,
+      "learning_rate": 1.8176259481520635e-06,
+      "loss": 0.5956,
+      "step": 748
+    },
+    {
+      "epoch": 0.59306181026377,
+      "grad_norm": 0.24718324290916885,
+      "learning_rate": 1.8171470281293564e-06,
+      "loss": 0.584,
+      "step": 749
+    },
+    {
+      "epoch": 0.5938536150838818,
+      "grad_norm": 0.2696495436808201,
+      "learning_rate": 1.81666754336978e-06,
+      "loss": 0.6009,
+      "step": 750
+    },
+    {
+      "epoch": 0.5946454199039937,
+      "grad_norm": 0.24839540781894592,
+      "learning_rate": 1.8161874942047097e-06,
+      "loss": 0.583,
+      "step": 751
+    },
+    {
+      "epoch": 0.5954372247241055,
+      "grad_norm": 0.251998162865897,
+      "learning_rate": 1.8157068809659117e-06,
+      "loss": 0.6036,
+      "step": 752
+    },
+    {
+      "epoch": 0.5962290295442173,
+      "grad_norm": 0.27137743828456923,
+      "learning_rate": 1.8152257039855412e-06,
+      "loss": 0.5889,
+      "step": 753
+    },
+    {
+      "epoch": 0.5970208343643292,
+      "grad_norm": 0.26559405382573104,
+      "learning_rate": 1.8147439635961436e-06,
+      "loss": 0.5994,
+      "step": 754
+    },
+    {
+      "epoch": 0.597812639184441,
+      "grad_norm": 0.2582769092884584,
+      "learning_rate": 1.814261660130654e-06,
+      "loss": 0.5912,
+      "step": 755
+    },
+    {
+      "epoch": 0.5986044440045529,
+      "grad_norm": 0.28215847993428234,
+      "learning_rate": 1.8137787939223953e-06,
+      "loss": 0.5902,
+      "step": 756
+    },
+    {
+      "epoch": 0.5993962488246647,
+      "grad_norm": 0.25873844774249477,
+      "learning_rate": 1.8132953653050812e-06,
+      "loss": 0.5822,
+      "step": 757
+    },
+    {
+      "epoch": 0.6001880536447766,
+      "grad_norm": 0.25218363810499717,
+      "learning_rate": 1.8128113746128127e-06,
+      "loss": 0.6108,
+      "step": 758
+    },
+    {
+      "epoch": 0.6009798584648884,
+      "grad_norm": 0.2643151580574007,
+      "learning_rate": 1.8123268221800793e-06,
+      "loss": 0.5815,
+      "step": 759
+    },
+    {
+      "epoch": 0.6017716632850002,
+      "grad_norm": 0.251824401021597,
+      "learning_rate": 1.8118417083417596e-06,
+      "loss": 0.5758,
+      "step": 760
+    },
+    {
+      "epoch": 0.6025634681051121,
+      "grad_norm": 0.24613995055877097,
+      "learning_rate": 1.8113560334331193e-06,
+      "loss": 0.5776,
+      "step": 761
+    },
+    {
+      "epoch": 0.603355272925224,
+      "grad_norm": 0.2601752874991997,
+      "learning_rate": 1.8108697977898126e-06,
+      "loss": 0.5848,
+      "step": 762
+    },
+    {
+      "epoch": 0.6041470777453358,
+      "grad_norm": 0.27174380073397775,
+      "learning_rate": 1.8103830017478805e-06,
+      "loss": 0.5731,
+      "step": 763
+    },
+    {
+      "epoch": 0.6049388825654476,
+      "grad_norm": 0.2430994748190506,
+      "learning_rate": 1.8098956456437518e-06,
+      "loss": 0.593,
+      "step": 764
+    },
+    {
+      "epoch": 0.6057306873855595,
+      "grad_norm": 0.2647416968126511,
+      "learning_rate": 1.8094077298142422e-06,
+      "loss": 0.5857,
+      "step": 765
+    },
+    {
+      "epoch": 0.6065224922056713,
+      "grad_norm": 0.27182892276562726,
+      "learning_rate": 1.8089192545965542e-06,
+      "loss": 0.5986,
+      "step": 766
+    },
+    {
+      "epoch": 0.6073142970257831,
+      "grad_norm": 0.25334048747254595,
+      "learning_rate": 1.808430220328277e-06,
+      "loss": 0.5791,
+      "step": 767
+    },
+    {
+      "epoch": 0.608106101845895,
+      "grad_norm": 0.259321595603846,
+      "learning_rate": 1.807940627347386e-06,
+      "loss": 0.5987,
+      "step": 768
+    },
+    {
+      "epoch": 0.6088979066660068,
+      "grad_norm": 0.2524989237650266,
+      "learning_rate": 1.807450475992243e-06,
+      "loss": 0.5911,
+      "step": 769
+    },
+    {
+      "epoch": 0.6096897114861187,
+      "grad_norm": 0.2535110040211958,
+      "learning_rate": 1.8069597666015956e-06,
+      "loss": 0.5786,
+      "step": 770
+    },
+    {
+      "epoch": 0.6104815163062305,
+      "grad_norm": 0.24658292893806905,
+      "learning_rate": 1.806468499514577e-06,
+      "loss": 0.5727,
+      "step": 771
+    },
+    {
+      "epoch": 0.6112733211263424,
+      "grad_norm": 0.2501300035527694,
+      "learning_rate": 1.8059766750707054e-06,
+      "loss": 0.5856,
+      "step": 772
+    },
+    {
+      "epoch": 0.6120651259464542,
+      "grad_norm": 0.24278539486327075,
+      "learning_rate": 1.8054842936098851e-06,
+      "loss": 0.5946,
+      "step": 773
+    },
+    {
+      "epoch": 0.612856930766566,
+      "grad_norm": 0.24808345795590186,
+      "learning_rate": 1.8049913554724046e-06,
+      "loss": 0.5828,
+      "step": 774
+    },
+    {
+      "epoch": 0.6136487355866779,
+      "grad_norm": 0.24297341717574392,
+      "learning_rate": 1.8044978609989374e-06,
+      "loss": 0.5817,
+      "step": 775
+    },
+    {
+      "epoch": 0.6144405404067897,
+      "grad_norm": 0.2573455405039158,
+      "learning_rate": 1.8040038105305415e-06,
+      "loss": 0.5883,
+      "step": 776
+    },
+    {
+      "epoch": 0.6152323452269016,
+      "grad_norm": 0.25158511627175334,
+      "learning_rate": 1.8035092044086594e-06,
+      "loss": 0.5672,
+      "step": 777
+    },
+    {
+      "epoch": 0.6160241500470134,
+      "grad_norm": 0.2522281238300653,
+      "learning_rate": 1.8030140429751173e-06,
+      "loss": 0.5715,
+      "step": 778
+    },
+    {
+      "epoch": 0.6168159548671253,
+      "grad_norm": 0.24727560215888592,
+      "learning_rate": 1.8025183265721251e-06,
+      "loss": 0.585,
+      "step": 779
+    },
+    {
+      "epoch": 0.6176077596872371,
+      "grad_norm": 0.2541287471761876,
+      "learning_rate": 1.8020220555422763e-06,
+      "loss": 0.5886,
+      "step": 780
+    },
+    {
+      "epoch": 0.6183995645073489,
+      "grad_norm": 0.2533308021305354,
+      "learning_rate": 1.801525230228548e-06,
+      "loss": 0.6008,
+      "step": 781
+    },
+    {
+      "epoch": 0.6191913693274608,
+      "grad_norm": 0.23744897496086267,
+      "learning_rate": 1.8010278509743e-06,
+      "loss": 0.5711,
+      "step": 782
+    },
+    {
+      "epoch": 0.6199831741475726,
+      "grad_norm": 0.25297067693147807,
+      "learning_rate": 1.8005299181232751e-06,
+      "loss": 0.5805,
+      "step": 783
+    },
+    {
+      "epoch": 0.6207749789676845,
+      "grad_norm": 0.24932173741694674,
+      "learning_rate": 1.800031432019599e-06,
+      "loss": 0.5863,
+      "step": 784
+    },
+    {
+      "epoch": 0.6215667837877963,
+      "grad_norm": 0.24535893078651158,
+      "learning_rate": 1.7995323930077791e-06,
+      "loss": 0.5872,
+      "step": 785
+    },
+    {
+      "epoch": 0.6223585886079082,
+      "grad_norm": 0.2383218746738785,
+      "learning_rate": 1.7990328014327054e-06,
+      "loss": 0.5742,
+      "step": 786
+    },
+    {
+      "epoch": 0.62315039342802,
+      "grad_norm": 0.25634064286313396,
+      "learning_rate": 1.7985326576396496e-06,
+      "loss": 0.6043,
+      "step": 787
+    },
+    {
+      "epoch": 0.6239421982481318,
+      "grad_norm": 0.24731036437872886,
+      "learning_rate": 1.798031961974265e-06,
+      "loss": 0.5789,
+      "step": 788
+    },
+    {
+      "epoch": 0.6247340030682437,
+      "grad_norm": 0.25243435536386216,
+      "learning_rate": 1.7975307147825864e-06,
+      "loss": 0.5845,
+      "step": 789
+    },
+    {
+      "epoch": 0.6255258078883555,
+      "grad_norm": 0.2511772367483909,
+      "learning_rate": 1.7970289164110301e-06,
+      "loss": 0.5776,
+      "step": 790
+    },
+    {
+      "epoch": 0.6263176127084673,
+      "grad_norm": 0.2532844706828268,
+      "learning_rate": 1.7965265672063927e-06,
+      "loss": 0.5824,
+      "step": 791
+    },
+    {
+      "epoch": 0.6271094175285792,
+      "grad_norm": 0.2658218019833223,
+      "learning_rate": 1.7960236675158518e-06,
+      "loss": 0.5956,
+      "step": 792
+    },
+    {
+      "epoch": 0.6279012223486911,
+      "grad_norm": 0.25412952245730175,
+      "learning_rate": 1.7955202176869655e-06,
+      "loss": 0.593,
+      "step": 793
+    },
+    {
+      "epoch": 0.6286930271688029,
+      "grad_norm": 0.2561659207216323,
+      "learning_rate": 1.7950162180676717e-06,
+      "loss": 0.5893,
+      "step": 794
+    },
+    {
+      "epoch": 0.6294848319889147,
+      "grad_norm": 0.2517474971714552,
+      "learning_rate": 1.794511669006289e-06,
+      "loss": 0.6108,
+      "step": 795
+    },
+    {
+      "epoch": 0.6302766368090266,
+      "grad_norm": 0.2528102682703397,
+      "learning_rate": 1.7940065708515148e-06,
+      "loss": 0.5992,
+      "step": 796
+    },
+    {
+      "epoch": 0.6310684416291384,
+      "grad_norm": 0.25992360575752016,
+      "learning_rate": 1.7935009239524273e-06,
+      "loss": 0.5795,
+      "step": 797
+    },
+    {
+      "epoch": 0.6318602464492502,
+      "grad_norm": 0.2554063427667969,
+      "learning_rate": 1.7929947286584823e-06,
+      "loss": 0.5985,
+      "step": 798
+    },
+    {
+      "epoch": 0.6326520512693621,
+      "grad_norm": 0.24422735834302362,
+      "learning_rate": 1.792487985319516e-06,
+      "loss": 0.5832,
+      "step": 799
+    },
+    {
+      "epoch": 0.633443856089474,
+      "grad_norm": 0.24622861998843143,
+      "learning_rate": 1.7919806942857426e-06,
+      "loss": 0.5742,
+      "step": 800
+    },
+    {
+      "epoch": 0.6342356609095858,
+      "grad_norm": 0.24029933869830092,
+      "learning_rate": 1.7914728559077546e-06,
+      "loss": 0.5787,
+      "step": 801
+    },
+    {
+      "epoch": 0.6350274657296976,
+      "grad_norm": 0.2429316083100275,
+      "learning_rate": 1.790964470536524e-06,
+      "loss": 0.592,
+      "step": 802
+    },
+    {
+      "epoch": 0.6358192705498095,
+      "grad_norm": 0.24302877995526823,
+      "learning_rate": 1.7904555385233994e-06,
+      "loss": 0.5866,
+      "step": 803
+    },
+    {
+      "epoch": 0.6366110753699213,
+      "grad_norm": 0.24746514901937777,
+      "learning_rate": 1.7899460602201077e-06,
+      "loss": 0.5839,
+      "step": 804
+    },
+    {
+      "epoch": 0.6374028801900331,
+      "grad_norm": 0.2420302965684456,
+      "learning_rate": 1.789436035978754e-06,
+      "loss": 0.5971,
+      "step": 805
+    },
+    {
+      "epoch": 0.638194685010145,
+      "grad_norm": 0.24385045963601545,
+      "learning_rate": 1.7889254661518196e-06,
+      "loss": 0.5815,
+      "step": 806
+    },
+    {
+      "epoch": 0.6389864898302569,
+      "grad_norm": 0.23828991606762473,
+      "learning_rate": 1.7884143510921637e-06,
+      "loss": 0.5953,
+      "step": 807
+    },
+    {
+      "epoch": 0.6397782946503687,
+      "grad_norm": 0.24420936560825504,
+      "learning_rate": 1.7879026911530222e-06,
+      "loss": 0.5926,
+      "step": 808
+    },
+    {
+      "epoch": 0.6405700994704805,
+      "grad_norm": 0.2714421774156976,
+      "learning_rate": 1.787390486688007e-06,
+      "loss": 0.5844,
+      "step": 809
+    },
+    {
+      "epoch": 0.6413619042905924,
+      "grad_norm": 0.24412097745674205,
+      "learning_rate": 1.786877738051107e-06,
+      "loss": 0.5844,
+      "step": 810
+    },
+    {
+      "epoch": 0.6421537091107042,
+      "grad_norm": 0.25422701784506546,
+      "learning_rate": 1.7863644455966865e-06,
+      "loss": 0.5693,
+      "step": 811
+    },
+    {
+      "epoch": 0.642945513930816,
+      "grad_norm": 0.2453540573413632,
+      "learning_rate": 1.7858506096794864e-06,
+      "loss": 0.5951,
+      "step": 812
+    },
+    {
+      "epoch": 0.6437373187509279,
+      "grad_norm": 0.2636005996907456,
+      "learning_rate": 1.7853362306546229e-06,
+      "loss": 0.5916,
+      "step": 813
+    },
+    {
+      "epoch": 0.6445291235710398,
+      "grad_norm": 0.24997894283566946,
+      "learning_rate": 1.7848213088775874e-06,
+      "loss": 0.5794,
+      "step": 814
+    },
+    {
+      "epoch": 0.6453209283911516,
+      "grad_norm": 0.24609967370733982,
+      "learning_rate": 1.784305844704246e-06,
+      "loss": 0.5842,
+      "step": 815
+    },
+    {
+      "epoch": 0.6461127332112634,
+      "grad_norm": 0.25314990024112644,
+      "learning_rate": 1.7837898384908407e-06,
+      "loss": 0.5754,
+      "step": 816
+    },
+    {
+      "epoch": 0.6469045380313753,
+      "grad_norm": 0.2528191893842514,
+      "learning_rate": 1.7832732905939876e-06,
+      "loss": 0.5924,
+      "step": 817
+    },
+    {
+      "epoch": 0.6476963428514871,
+      "grad_norm": 0.24435530651812093,
+      "learning_rate": 1.7827562013706766e-06,
+      "loss": 0.5826,
+      "step": 818
+    },
+    {
+      "epoch": 0.6484881476715989,
+      "grad_norm": 0.2572416831231824,
+      "learning_rate": 1.7822385711782726e-06,
+      "loss": 0.608,
+      "step": 819
+    },
+    {
+      "epoch": 0.6492799524917108,
+      "grad_norm": 0.25617732603538695,
+      "learning_rate": 1.7817204003745134e-06,
+      "loss": 0.5969,
+      "step": 820
+    },
+    {
+      "epoch": 0.6500717573118227,
+      "grad_norm": 0.2521166470503421,
+      "learning_rate": 1.7812016893175119e-06,
+      "loss": 0.5871,
+      "step": 821
+    },
+    {
+      "epoch": 0.6508635621319345,
+      "grad_norm": 0.25056053985682086,
+      "learning_rate": 1.780682438365753e-06,
+      "loss": 0.5848,
+      "step": 822
+    },
+    {
+      "epoch": 0.6516553669520463,
+      "grad_norm": 0.25287431062010535,
+      "learning_rate": 1.780162647878095e-06,
+      "loss": 0.5828,
+      "step": 823
+    },
+    {
+      "epoch": 0.6524471717721582,
+      "grad_norm": 0.25007192663814964,
+      "learning_rate": 1.7796423182137694e-06,
+      "loss": 0.5854,
+      "step": 824
+    },
+    {
+      "epoch": 0.65323897659227,
+      "grad_norm": 0.25043387039495113,
+      "learning_rate": 1.7791214497323805e-06,
+      "loss": 0.5937,
+      "step": 825
+    },
+    {
+      "epoch": 0.6540307814123818,
+      "grad_norm": 0.25615998922421296,
+      "learning_rate": 1.7786000427939047e-06,
+      "loss": 0.5741,
+      "step": 826
+    },
+    {
+      "epoch": 0.6548225862324937,
+      "grad_norm": 0.24930777273834764,
+      "learning_rate": 1.77807809775869e-06,
+      "loss": 0.5908,
+      "step": 827
+    },
+    {
+      "epoch": 0.6556143910526055,
+      "grad_norm": 0.2585417804755247,
+      "learning_rate": 1.777555614987457e-06,
+      "loss": 0.5883,
+      "step": 828
+    },
+    {
+      "epoch": 0.6564061958727174,
+      "grad_norm": 0.24450521436137,
+      "learning_rate": 1.777032594841298e-06,
+      "loss": 0.5678,
+      "step": 829
+    },
+    {
+      "epoch": 0.6571980006928292,
+      "grad_norm": 0.24862190261035272,
+      "learning_rate": 1.7765090376816766e-06,
+      "loss": 0.592,
+      "step": 830
+    },
+    {
+      "epoch": 0.6579898055129411,
+      "grad_norm": 0.25432285465044857,
+      "learning_rate": 1.7759849438704264e-06,
+      "loss": 0.5703,
+      "step": 831
+    },
+    {
+      "epoch": 0.6587816103330529,
+      "grad_norm": 0.25194906787878907,
+      "learning_rate": 1.775460313769754e-06,
+      "loss": 0.5886,
+      "step": 832
+    },
+    {
+      "epoch": 0.6595734151531647,
+      "grad_norm": 0.24467044323457524,
+      "learning_rate": 1.7749351477422345e-06,
+      "loss": 0.594,
+      "step": 833
+    },
+    {
+      "epoch": 0.6603652199732766,
+      "grad_norm": 0.2357389926591831,
+      "learning_rate": 1.774409446150815e-06,
+      "loss": 0.5786,
+      "step": 834
+    },
+    {
+      "epoch": 0.6611570247933884,
+      "grad_norm": 0.25621856938464344,
+      "learning_rate": 1.773883209358812e-06,
+      "loss": 0.5668,
+      "step": 835
+    },
+    {
+      "epoch": 0.6619488296135003,
+      "grad_norm": 0.2538284329877011,
+      "learning_rate": 1.7733564377299115e-06,
+      "loss": 0.5804,
+      "step": 836
+    },
+    {
+      "epoch": 0.6627406344336121,
+      "grad_norm": 0.248575447549464,
+      "learning_rate": 1.7728291316281703e-06,
+      "loss": 0.5987,
+      "step": 837
+    },
+    {
+      "epoch": 0.663532439253724,
+      "grad_norm": 0.2559812446732376,
+      "learning_rate": 1.7723012914180134e-06,
+      "loss": 0.5894,
+      "step": 838
+    },
+    {
+      "epoch": 0.6643242440738358,
+      "grad_norm": 0.26293105509088166,
+      "learning_rate": 1.7717729174642355e-06,
+      "loss": 0.5879,
+      "step": 839
+    },
+    {
+      "epoch": 0.6651160488939476,
+      "grad_norm": 0.23970650462828844,
+      "learning_rate": 1.7712440101320004e-06,
+      "loss": 0.5985,
+      "step": 840
+    },
+    {
+      "epoch": 0.6659078537140595,
+      "grad_norm": 0.25677479955425603,
+      "learning_rate": 1.7707145697868392e-06,
+      "loss": 0.6022,
+      "step": 841
+    },
+    {
+      "epoch": 0.6666996585341713,
+      "grad_norm": 0.24918667579670953,
+      "learning_rate": 1.7701845967946537e-06,
+      "loss": 0.5706,
+      "step": 842
+    },
+    {
+      "epoch": 0.6674914633542832,
+      "grad_norm": 0.2565401820917414,
+      "learning_rate": 1.769654091521712e-06,
+      "loss": 0.5734,
+      "step": 843
+    },
+    {
+      "epoch": 0.668283268174395,
+      "grad_norm": 0.26433306600275414,
+      "learning_rate": 1.7691230543346501e-06,
+      "loss": 0.5932,
+      "step": 844
+    },
+    {
+      "epoch": 0.6690750729945069,
+      "grad_norm": 0.2569470821796716,
+      "learning_rate": 1.7685914856004725e-06,
+      "loss": 0.5822,
+      "step": 845
+    },
+    {
+      "epoch": 0.6698668778146187,
+      "grad_norm": 0.2544063865454498,
+      "learning_rate": 1.7680593856865503e-06,
+      "loss": 0.5874,
+      "step": 846
+    },
+    {
+      "epoch": 0.6706586826347305,
+      "grad_norm": 0.2542509968843097,
+      "learning_rate": 1.7675267549606225e-06,
+      "loss": 0.584,
+      "step": 847
+    },
+    {
+      "epoch": 0.6714504874548424,
+      "grad_norm": 0.24318698066947034,
+      "learning_rate": 1.7669935937907941e-06,
+      "loss": 0.5906,
+      "step": 848
+    },
+    {
+      "epoch": 0.6722422922749542,
+      "grad_norm": 0.24661612792719936,
+      "learning_rate": 1.766459902545537e-06,
+      "loss": 0.5813,
+      "step": 849
+    },
+    {
+      "epoch": 0.6730340970950661,
+      "grad_norm": 0.2492753457810347,
+      "learning_rate": 1.76592568159369e-06,
+      "loss": 0.5801,
+      "step": 850
+    },
+    {
+      "epoch": 0.6738259019151779,
+      "grad_norm": 0.24830446834246703,
+      "learning_rate": 1.765390931304457e-06,
+      "loss": 0.5915,
+      "step": 851
+    },
+    {
+      "epoch": 0.6746177067352898,
+      "grad_norm": 0.25431438924953426,
+      "learning_rate": 1.7648556520474083e-06,
+      "loss": 0.5836,
+      "step": 852
+    },
+    {
+      "epoch": 0.6754095115554016,
+      "grad_norm": 0.2588824070511301,
+      "learning_rate": 1.7643198441924797e-06,
+      "loss": 0.5879,
+      "step": 853
+    },
+    {
+      "epoch": 0.6762013163755134,
+      "grad_norm": 0.25038875476912814,
+      "learning_rate": 1.7637835081099729e-06,
+      "loss": 0.5725,
+      "step": 854
+    },
+    {
+      "epoch": 0.6769931211956253,
+      "grad_norm": 0.2514821846393738,
+      "learning_rate": 1.763246644170553e-06,
+      "loss": 0.5868,
+      "step": 855
+    },
+    {
+      "epoch": 0.6777849260157371,
+      "grad_norm": 0.2851916766730875,
+      "learning_rate": 1.7627092527452515e-06,
+      "loss": 0.5976,
+      "step": 856
+    },
+    {
+      "epoch": 0.678576730835849,
+      "grad_norm": 0.24179081935232716,
+      "learning_rate": 1.762171334205464e-06,
+      "loss": 0.5793,
+      "step": 857
+    },
+    {
+      "epoch": 0.6793685356559608,
+      "grad_norm": 0.25292715648491026,
+      "learning_rate": 1.7616328889229505e-06,
+      "loss": 0.5827,
+      "step": 858
+    },
+    {
+      "epoch": 0.6801603404760727,
+      "grad_norm": 0.23964732502085143,
+      "learning_rate": 1.7610939172698344e-06,
+      "loss": 0.5846,
+      "step": 859
+    },
+    {
+      "epoch": 0.6809521452961845,
+      "grad_norm": 0.24641216304573305,
+      "learning_rate": 1.7605544196186037e-06,
+      "loss": 0.6043,
+      "step": 860
+    },
+    {
+      "epoch": 0.6817439501162963,
+      "grad_norm": 0.2591671468302859,
+      "learning_rate": 1.7600143963421094e-06,
+      "loss": 0.5837,
+      "step": 861
+    },
+    {
+      "epoch": 0.6825357549364082,
+      "grad_norm": 0.24314631653055893,
+      "learning_rate": 1.7594738478135658e-06,
+      "loss": 0.5926,
+      "step": 862
+    },
+    {
+      "epoch": 0.68332755975652,
+      "grad_norm": 0.24738598633453043,
+      "learning_rate": 1.7589327744065507e-06,
+      "loss": 0.5869,
+      "step": 863
+    },
+    {
+      "epoch": 0.6841193645766318,
+      "grad_norm": 0.25688559624231405,
+      "learning_rate": 1.7583911764950039e-06,
+      "loss": 0.5937,
+      "step": 864
+    },
+    {
+      "epoch": 0.6849111693967437,
+      "grad_norm": 0.24885371925651226,
+      "learning_rate": 1.7578490544532283e-06,
+      "loss": 0.5856,
+      "step": 865
+    },
+    {
+      "epoch": 0.6857029742168556,
+      "grad_norm": 0.24690074827520492,
+      "learning_rate": 1.7573064086558884e-06,
+      "loss": 0.574,
+      "step": 866
+    },
+    {
+      "epoch": 0.6864947790369674,
+      "grad_norm": 0.24760162187909207,
+      "learning_rate": 1.7567632394780118e-06,
+      "loss": 0.5665,
+      "step": 867
+    },
+    {
+      "epoch": 0.6872865838570792,
+      "grad_norm": 0.24577712965038068,
+      "learning_rate": 1.7562195472949863e-06,
+      "loss": 0.584,
+      "step": 868
+    },
+    {
+      "epoch": 0.6880783886771911,
+      "grad_norm": 0.24707196829506817,
+      "learning_rate": 1.7556753324825624e-06,
+      "loss": 0.5728,
+      "step": 869
+    },
+    {
+      "epoch": 0.6888701934973029,
+      "grad_norm": 0.2427499693077296,
+      "learning_rate": 1.7551305954168512e-06,
+      "loss": 0.5907,
+      "step": 870
+    },
+    {
+      "epoch": 0.6896619983174147,
+      "grad_norm": 0.2557832991364167,
+      "learning_rate": 1.7545853364743248e-06,
+      "loss": 0.5817,
+      "step": 871
+    },
+    {
+      "epoch": 0.6904538031375266,
+      "grad_norm": 0.2480115469674397,
+      "learning_rate": 1.7540395560318162e-06,
+      "loss": 0.5878,
+      "step": 872
+    },
+    {
+      "epoch": 0.6912456079576385,
+      "grad_norm": 0.24717915980270053,
+      "learning_rate": 1.7534932544665183e-06,
+      "loss": 0.5734,
+      "step": 873
+    },
+    {
+      "epoch": 0.6920374127777503,
+      "grad_norm": 0.2524836513942343,
+      "learning_rate": 1.7529464321559848e-06,
+      "loss": 0.5948,
+      "step": 874
+    },
+    {
+      "epoch": 0.6928292175978621,
+      "grad_norm": 0.245275353092912,
+      "learning_rate": 1.752399089478129e-06,
+      "loss": 0.5744,
+      "step": 875
+    },
+    {
+      "epoch": 0.693621022417974,
+      "grad_norm": 0.24062703187758544,
+      "learning_rate": 1.7518512268112238e-06,
+      "loss": 0.5784,
+      "step": 876
+    },
+    {
+      "epoch": 0.6944128272380858,
+      "grad_norm": 0.24530600631442345,
+      "learning_rate": 1.7513028445339014e-06,
+      "loss": 0.5782,
+      "step": 877
+    },
+    {
+      "epoch": 0.6952046320581976,
+      "grad_norm": 0.24669451967407843,
+      "learning_rate": 1.750753943025153e-06,
+      "loss": 0.575,
+      "step": 878
+    },
+    {
+      "epoch": 0.6959964368783095,
+      "grad_norm": 0.2447861480839371,
+      "learning_rate": 1.750204522664329e-06,
+      "loss": 0.5886,
+      "step": 879
+    },
+    {
+      "epoch": 0.6967882416984214,
+      "grad_norm": 0.24324802163539397,
+      "learning_rate": 1.7496545838311386e-06,
+      "loss": 0.5978,
+      "step": 880
+    },
+    {
+      "epoch": 0.6975800465185332,
+      "grad_norm": 0.25376150676455245,
+      "learning_rate": 1.7491041269056485e-06,
+      "loss": 0.5718,
+      "step": 881
+    },
+    {
+      "epoch": 0.698371851338645,
+      "grad_norm": 0.24267518899970847,
+      "learning_rate": 1.7485531522682839e-06,
+      "loss": 0.5711,
+      "step": 882
+    },
+    {
+      "epoch": 0.6991636561587569,
+      "grad_norm": 0.24977132670485755,
+      "learning_rate": 1.748001660299828e-06,
+      "loss": 0.5941,
+      "step": 883
+    },
+    {
+      "epoch": 0.6999554609788687,
+      "grad_norm": 0.2459853658010188,
+      "learning_rate": 1.7474496513814213e-06,
+      "loss": 0.5925,
+      "step": 884
+    },
+    {
+      "epoch": 0.7007472657989805,
+      "grad_norm": 0.26409388113089094,
+      "learning_rate": 1.7468971258945615e-06,
+      "loss": 0.5922,
+      "step": 885
+    },
+    {
+      "epoch": 0.7015390706190924,
+      "grad_norm": 0.24702804605528103,
+      "learning_rate": 1.7463440842211035e-06,
+      "loss": 0.573,
+      "step": 886
+    },
+    {
+      "epoch": 0.7023308754392042,
+      "grad_norm": 0.2541245452020192,
+      "learning_rate": 1.7457905267432585e-06,
+      "loss": 0.5839,
+      "step": 887
+    },
+    {
+      "epoch": 0.7031226802593161,
+      "grad_norm": 0.25110485274127725,
+      "learning_rate": 1.7452364538435949e-06,
+      "loss": 0.5743,
+      "step": 888
+    },
+    {
+      "epoch": 0.7039144850794279,
+      "grad_norm": 0.25186565532483407,
+      "learning_rate": 1.7446818659050369e-06,
+      "loss": 0.5948,
+      "step": 889
+    },
+    {
+      "epoch": 0.7047062898995398,
+      "grad_norm": 0.2429131508592695,
+      "learning_rate": 1.7441267633108642e-06,
+      "loss": 0.5891,
+      "step": 890
+    },
+    {
+      "epoch": 0.7054980947196516,
+      "grad_norm": 0.23545565600107377,
+      "learning_rate": 1.7435711464447133e-06,
+      "loss": 0.5697,
+      "step": 891
+    },
+    {
+      "epoch": 0.7062898995397634,
+      "grad_norm": 0.24937194476011745,
+      "learning_rate": 1.7430150156905752e-06,
+      "loss": 0.5797,
+      "step": 892
+    },
+    {
+      "epoch": 0.7070817043598753,
+      "grad_norm": 0.244376238101983,
+      "learning_rate": 1.7424583714327965e-06,
+      "loss": 0.5884,
+      "step": 893
+    },
+    {
+      "epoch": 0.7078735091799871,
+      "grad_norm": 0.23916929044981503,
+      "learning_rate": 1.7419012140560782e-06,
+      "loss": 0.5729,
+      "step": 894
+    },
+    {
+      "epoch": 0.708665314000099,
+      "grad_norm": 0.25363642242342044,
+      "learning_rate": 1.7413435439454762e-06,
+      "loss": 0.5714,
+      "step": 895
+    },
+    {
+      "epoch": 0.7094571188202108,
+      "grad_norm": 0.24211294200683903,
+      "learning_rate": 1.7407853614864013e-06,
+      "loss": 0.5658,
+      "step": 896
+    },
+    {
+      "epoch": 0.7102489236403227,
+      "grad_norm": 0.24860427991387388,
+      "learning_rate": 1.7402266670646175e-06,
+      "loss": 0.5775,
+      "step": 897
+    },
+    {
+      "epoch": 0.7110407284604345,
+      "grad_norm": 0.2561766068193548,
+      "learning_rate": 1.7396674610662431e-06,
+      "loss": 0.5866,
+      "step": 898
+    },
+    {
+      "epoch": 0.7118325332805463,
+      "grad_norm": 0.25062849494456546,
+      "learning_rate": 1.7391077438777498e-06,
+      "loss": 0.5808,
+      "step": 899
+    },
+    {
+      "epoch": 0.7126243381006582,
+      "grad_norm": 0.25441600351502675,
+      "learning_rate": 1.7385475158859624e-06,
+      "loss": 0.6,
+      "step": 900
+    },
+    {
+      "epoch": 0.71341614292077,
+      "grad_norm": 0.2464011228232308,
+      "learning_rate": 1.7379867774780594e-06,
+      "loss": 0.6006,
+      "step": 901
+    },
+    {
+      "epoch": 0.7142079477408819,
+      "grad_norm": 0.2544724679474216,
+      "learning_rate": 1.7374255290415714e-06,
+      "loss": 0.5975,
+      "step": 902
+    },
+    {
+      "epoch": 0.7149997525609937,
+      "grad_norm": 0.2519462028169189,
+      "learning_rate": 1.7368637709643817e-06,
+      "loss": 0.5717,
+      "step": 903
+    },
+    {
+      "epoch": 0.7157915573811056,
+      "grad_norm": 0.24456334735486743,
+      "learning_rate": 1.7363015036347259e-06,
+      "loss": 0.5829,
+      "step": 904
+    },
+    {
+      "epoch": 0.7165833622012174,
+      "grad_norm": 0.249011649931778,
+      "learning_rate": 1.7357387274411912e-06,
+      "loss": 0.5876,
+      "step": 905
+    },
+    {
+      "epoch": 0.7173751670213292,
+      "grad_norm": 0.24076372543115893,
+      "learning_rate": 1.7351754427727174e-06,
+      "loss": 0.5847,
+      "step": 906
+    },
+    {
+      "epoch": 0.7181669718414411,
+      "grad_norm": 0.25032599232475183,
+      "learning_rate": 1.734611650018594e-06,
+      "loss": 0.5949,
+      "step": 907
+    },
+    {
+      "epoch": 0.718958776661553,
+      "grad_norm": 0.2485987378016031,
+      "learning_rate": 1.7340473495684637e-06,
+      "loss": 0.5679,
+      "step": 908
+    },
+    {
+      "epoch": 0.7197505814816648,
+      "grad_norm": 0.24943278117786208,
+      "learning_rate": 1.7334825418123187e-06,
+      "loss": 0.5966,
+      "step": 909
+    },
+    {
+      "epoch": 0.7205423863017766,
+      "grad_norm": 0.24813196439993077,
+      "learning_rate": 1.7329172271405019e-06,
+      "loss": 0.5852,
+      "step": 910
+    },
+    {
+      "epoch": 0.7213341911218885,
+      "grad_norm": 0.2535279623656571,
+      "learning_rate": 1.7323514059437075e-06,
+      "loss": 0.5879,
+      "step": 911
+    },
+    {
+      "epoch": 0.7221259959420003,
+      "grad_norm": 0.2520772188455546,
+      "learning_rate": 1.7317850786129782e-06,
+      "loss": 0.5838,
+      "step": 912
+    },
+    {
+      "epoch": 0.7229178007621121,
+      "grad_norm": 0.2525009987208763,
+      "learning_rate": 1.731218245539708e-06,
+      "loss": 0.5857,
+      "step": 913
+    },
+    {
+      "epoch": 0.723709605582224,
+      "grad_norm": 0.2433018153161353,
+      "learning_rate": 1.7306509071156394e-06,
+      "loss": 0.5914,
+      "step": 914
+    },
+    {
+      "epoch": 0.7245014104023358,
+      "grad_norm": 0.24811807449526793,
+      "learning_rate": 1.7300830637328648e-06,
+      "loss": 0.5832,
+      "step": 915
+    },
+    {
+      "epoch": 0.7252932152224477,
+      "grad_norm": 0.2469823253728138,
+      "learning_rate": 1.7295147157838253e-06,
+      "loss": 0.5834,
+      "step": 916
+    },
+    {
+      "epoch": 0.7260850200425595,
+      "grad_norm": 0.24970511504740275,
+      "learning_rate": 1.7289458636613105e-06,
+      "loss": 0.5753,
+      "step": 917
+    },
+    {
+      "epoch": 0.7268768248626714,
+      "grad_norm": 0.2430854348807761,
+      "learning_rate": 1.7283765077584592e-06,
+      "loss": 0.5761,
+      "step": 918
+    },
+    {
+      "epoch": 0.7276686296827832,
+      "grad_norm": 0.24594292469061335,
+      "learning_rate": 1.7278066484687571e-06,
+      "loss": 0.5891,
+      "step": 919
+    },
+    {
+      "epoch": 0.728460434502895,
+      "grad_norm": 0.23539605756545798,
+      "learning_rate": 1.7272362861860388e-06,
+      "loss": 0.5778,
+      "step": 920
+    },
+    {
+      "epoch": 0.7292522393230069,
+      "grad_norm": 0.25176104671877025,
+      "learning_rate": 1.7266654213044865e-06,
+      "loss": 0.5723,
+      "step": 921
+    },
+    {
+      "epoch": 0.7300440441431187,
+      "grad_norm": 0.2450854097972758,
+      "learning_rate": 1.7260940542186293e-06,
+      "loss": 0.5822,
+      "step": 922
+    },
+    {
+      "epoch": 0.7308358489632306,
+      "grad_norm": 0.25608007063845983,
+      "learning_rate": 1.7255221853233438e-06,
+      "loss": 0.5839,
+      "step": 923
+    },
+    {
+      "epoch": 0.7316276537833424,
+      "grad_norm": 0.24542022770248775,
+      "learning_rate": 1.7249498150138528e-06,
+      "loss": 0.5733,
+      "step": 924
+    },
+    {
+      "epoch": 0.7324194586034543,
+      "grad_norm": 0.24927466094881492,
+      "learning_rate": 1.7243769436857262e-06,
+      "loss": 0.5847,
+      "step": 925
+    },
+    {
+      "epoch": 0.7332112634235661,
+      "grad_norm": 0.26812137197973235,
+      "learning_rate": 1.7238035717348801e-06,
+      "loss": 0.582,
+      "step": 926
+    },
+    {
+      "epoch": 0.7340030682436779,
+      "grad_norm": 0.2463608580688781,
+      "learning_rate": 1.7232296995575765e-06,
+      "loss": 0.5867,
+      "step": 927
+    },
+    {
+      "epoch": 0.7347948730637898,
+      "grad_norm": 0.2688322608735481,
+      "learning_rate": 1.7226553275504229e-06,
+      "loss": 0.5806,
+      "step": 928
+    },
+    {
+      "epoch": 0.7355866778839016,
+      "grad_norm": 0.26945234351830955,
+      "learning_rate": 1.722080456110372e-06,
+      "loss": 0.5846,
+      "step": 929
+    },
+    {
+      "epoch": 0.7363784827040135,
+      "grad_norm": 0.25408035335687396,
+      "learning_rate": 1.721505085634723e-06,
+      "loss": 0.5755,
+      "step": 930
+    },
+    {
+      "epoch": 0.7371702875241253,
+      "grad_norm": 0.2465434921741765,
+      "learning_rate": 1.7209292165211182e-06,
+      "loss": 0.5798,
+      "step": 931
+    },
+    {
+      "epoch": 0.7379620923442372,
+      "grad_norm": 0.2540837663212656,
+      "learning_rate": 1.720352849167546e-06,
+      "loss": 0.5802,
+      "step": 932
+    },
+    {
+      "epoch": 0.738753897164349,
+      "grad_norm": 0.25238195730729085,
+      "learning_rate": 1.7197759839723375e-06,
+      "loss": 0.5718,
+      "step": 933
+    },
+    {
+      "epoch": 0.7395457019844608,
+      "grad_norm": 0.25437203628781707,
+      "learning_rate": 1.7191986213341703e-06,
+      "loss": 0.5668,
+      "step": 934
+    },
+    {
+      "epoch": 0.7403375068045727,
+      "grad_norm": 0.25335680248875186,
+      "learning_rate": 1.7186207616520635e-06,
+      "loss": 0.5697,
+      "step": 935
+    },
+    {
+      "epoch": 0.7411293116246845,
+      "grad_norm": 0.2470089828067232,
+      "learning_rate": 1.7180424053253808e-06,
+      "loss": 0.5788,
+      "step": 936
+    },
+    {
+      "epoch": 0.7419211164447963,
+      "grad_norm": 0.25981652910641084,
+      "learning_rate": 1.7174635527538284e-06,
+      "loss": 0.5666,
+      "step": 937
+    },
+    {
+      "epoch": 0.7427129212649082,
+      "grad_norm": 0.25609013882045345,
+      "learning_rate": 1.7168842043374566e-06,
+      "loss": 0.5809,
+      "step": 938
+    },
+    {
+      "epoch": 0.74350472608502,
+      "grad_norm": 0.25837506251027265,
+      "learning_rate": 1.7163043604766573e-06,
+      "loss": 0.5711,
+      "step": 939
+    },
+    {
+      "epoch": 0.7442965309051319,
+      "grad_norm": 0.2555723939908134,
+      "learning_rate": 1.715724021572166e-06,
+      "loss": 0.5863,
+      "step": 940
+    },
+    {
+      "epoch": 0.7450883357252437,
+      "grad_norm": 0.24707881931050837,
+      "learning_rate": 1.7151431880250585e-06,
+      "loss": 0.5863,
+      "step": 941
+    },
+    {
+      "epoch": 0.7458801405453556,
+      "grad_norm": 0.24351053097539035,
+      "learning_rate": 1.714561860236754e-06,
+      "loss": 0.599,
+      "step": 942
+    },
+    {
+      "epoch": 0.7466719453654674,
+      "grad_norm": 0.24475544145118233,
+      "learning_rate": 1.7139800386090131e-06,
+      "loss": 0.5885,
+      "step": 943
+    },
+    {
+      "epoch": 0.7474637501855792,
+      "grad_norm": 0.25716157533418266,
+      "learning_rate": 1.7133977235439371e-06,
+      "loss": 0.5594,
+      "step": 944
+    },
+    {
+      "epoch": 0.7482555550056911,
+      "grad_norm": 0.25286004003751,
+      "learning_rate": 1.7128149154439684e-06,
+      "loss": 0.5801,
+      "step": 945
+    },
+    {
+      "epoch": 0.749047359825803,
+      "grad_norm": 0.24534973996948886,
+      "learning_rate": 1.712231614711891e-06,
+      "loss": 0.5963,
+      "step": 946
+    },
+    {
+      "epoch": 0.7498391646459148,
+      "grad_norm": 0.2478474592213585,
+      "learning_rate": 1.711647821750828e-06,
+      "loss": 0.582,
+      "step": 947
+    },
+    {
+      "epoch": 0.7506309694660266,
+      "grad_norm": 0.2516958656147911,
+      "learning_rate": 1.711063536964244e-06,
+      "loss": 0.5758,
+      "step": 948
+    },
+    {
+      "epoch": 0.7514227742861385,
+      "grad_norm": 0.25361160405447486,
+      "learning_rate": 1.7104787607559424e-06,
+      "loss": 0.5843,
+      "step": 949
+    },
+    {
+      "epoch": 0.7522145791062503,
+      "grad_norm": 0.2536184915636103,
+      "learning_rate": 1.709893493530067e-06,
+      "loss": 0.5957,
+      "step": 950
+    },
+    {
+      "epoch": 0.7530063839263621,
+      "grad_norm": 0.2532291449126456,
+      "learning_rate": 1.7093077356911008e-06,
+      "loss": 0.5847,
+      "step": 951
+    },
+    {
+      "epoch": 0.753798188746474,
+      "grad_norm": 0.24017158006580785,
+      "learning_rate": 1.7087214876438653e-06,
+      "loss": 0.594,
+      "step": 952
+    },
+    {
+      "epoch": 0.7545899935665858,
+      "grad_norm": 0.24723979397219245,
+      "learning_rate": 1.7081347497935216e-06,
+      "loss": 0.5778,
+      "step": 953
+    },
+    {
+      "epoch": 0.7553817983866977,
+      "grad_norm": 0.2564595111532207,
+      "learning_rate": 1.7075475225455688e-06,
+      "loss": 0.5888,
+      "step": 954
+    },
+    {
+      "epoch": 0.7561736032068095,
+      "grad_norm": 0.24437381393671384,
+      "learning_rate": 1.7069598063058448e-06,
+      "loss": 0.5761,
+      "step": 955
+    },
+    {
+      "epoch": 0.7569654080269214,
+      "grad_norm": 0.2521756274505913,
+      "learning_rate": 1.7063716014805242e-06,
+      "loss": 0.5735,
+      "step": 956
+    },
+    {
+      "epoch": 0.7577572128470332,
+      "grad_norm": 0.24695700919885905,
+      "learning_rate": 1.7057829084761207e-06,
+      "loss": 0.6006,
+      "step": 957
+    },
+    {
+      "epoch": 0.758549017667145,
+      "grad_norm": 0.24688421725514956,
+      "learning_rate": 1.7051937276994845e-06,
+      "loss": 0.5892,
+      "step": 958
+    },
+    {
+      "epoch": 0.7593408224872569,
+      "grad_norm": 0.24928610762379938,
+      "learning_rate": 1.7046040595578033e-06,
+      "loss": 0.5894,
+      "step": 959
+    },
+    {
+      "epoch": 0.7601326273073687,
+      "grad_norm": 0.2521753453756982,
+      "learning_rate": 1.7040139044586016e-06,
+      "loss": 0.5855,
+      "step": 960
+    },
+    {
+      "epoch": 0.7609244321274806,
+      "grad_norm": 0.24392734940364796,
+      "learning_rate": 1.7034232628097395e-06,
+      "loss": 0.5789,
+      "step": 961
+    },
+    {
+      "epoch": 0.7617162369475924,
+      "grad_norm": 0.25157346435761563,
+      "learning_rate": 1.7028321350194155e-06,
+      "loss": 0.5783,
+      "step": 962
+    },
+    {
+      "epoch": 0.7625080417677043,
+      "grad_norm": 0.24407210040846353,
+      "learning_rate": 1.7022405214961617e-06,
+      "loss": 0.5908,
+      "step": 963
+    },
+    {
+      "epoch": 0.7632998465878161,
+      "grad_norm": 0.24231258628359728,
+      "learning_rate": 1.7016484226488474e-06,
+      "loss": 0.5752,
+      "step": 964
+    },
+    {
+      "epoch": 0.7640916514079279,
+      "grad_norm": 0.24266590901680207,
+      "learning_rate": 1.7010558388866766e-06,
+      "loss": 0.589,
+      "step": 965
+    },
+    {
+      "epoch": 0.7648834562280398,
+      "grad_norm": 0.24311944620382953,
+      "learning_rate": 1.7004627706191895e-06,
+      "loss": 0.6077,
+      "step": 966
+    },
+    {
+      "epoch": 0.7656752610481516,
+      "grad_norm": 0.24549077766115834,
+      "learning_rate": 1.6998692182562592e-06,
+      "loss": 0.5778,
+      "step": 967
+    },
+    {
+      "epoch": 0.7664670658682635,
+      "grad_norm": 0.24241932841306488,
+      "learning_rate": 1.6992751822080952e-06,
+      "loss": 0.5909,
+      "step": 968
+    },
+    {
+      "epoch": 0.7672588706883753,
+      "grad_norm": 0.25902837727029465,
+      "learning_rate": 1.6986806628852404e-06,
+      "loss": 0.5709,
+      "step": 969
+    },
+    {
+      "epoch": 0.7680506755084872,
+      "grad_norm": 0.2489981742220158,
+      "learning_rate": 1.6980856606985717e-06,
+      "loss": 0.6042,
+      "step": 970
+    },
+    {
+      "epoch": 0.768842480328599,
+      "grad_norm": 0.25941636084634623,
+      "learning_rate": 1.6974901760593006e-06,
+      "loss": 0.5758,
+      "step": 971
+    },
+    {
+      "epoch": 0.7696342851487108,
+      "grad_norm": 0.25092704109760794,
+      "learning_rate": 1.6968942093789704e-06,
+      "loss": 0.5987,
+      "step": 972
+    },
+    {
+      "epoch": 0.7704260899688227,
+      "grad_norm": 0.2604196556366055,
+      "learning_rate": 1.6962977610694588e-06,
+      "loss": 0.5834,
+      "step": 973
+    },
+    {
+      "epoch": 0.7712178947889345,
+      "grad_norm": 0.2583034640293634,
+      "learning_rate": 1.6957008315429761e-06,
+      "loss": 0.5732,
+      "step": 974
+    },
+    {
+      "epoch": 0.7720096996090464,
+      "grad_norm": 0.24564918682645784,
+      "learning_rate": 1.6951034212120652e-06,
+      "loss": 0.5776,
+      "step": 975
+    },
+    {
+      "epoch": 0.7728015044291582,
+      "grad_norm": 0.24530129884926857,
+      "learning_rate": 1.694505530489601e-06,
+      "loss": 0.581,
+      "step": 976
+    },
+    {
+      "epoch": 0.7735933092492701,
+      "grad_norm": 0.25839773288616386,
+      "learning_rate": 1.6939071597887903e-06,
+      "loss": 0.5649,
+      "step": 977
+    },
+    {
+      "epoch": 0.7743851140693819,
+      "grad_norm": 0.2633398810124407,
+      "learning_rate": 1.6933083095231726e-06,
+      "loss": 0.5903,
+      "step": 978
+    },
+    {
+      "epoch": 0.7751769188894937,
+      "grad_norm": 0.25183824076871514,
+      "learning_rate": 1.6927089801066176e-06,
+      "loss": 0.5742,
+      "step": 979
+    },
+    {
+      "epoch": 0.7759687237096056,
+      "grad_norm": 0.24269781478025124,
+      "learning_rate": 1.6921091719533267e-06,
+      "loss": 0.5772,
+      "step": 980
+    },
+    {
+      "epoch": 0.7767605285297174,
+      "grad_norm": 0.2418600370580417,
+      "learning_rate": 1.6915088854778324e-06,
+      "loss": 0.5715,
+      "step": 981
+    },
+    {
+      "epoch": 0.7775523333498293,
+      "grad_norm": 0.2572191934436083,
+      "learning_rate": 1.690908121094997e-06,
+      "loss": 0.575,
+      "step": 982
+    },
+    {
+      "epoch": 0.7783441381699411,
+      "grad_norm": 0.25724074356800636,
+      "learning_rate": 1.690306879220014e-06,
+      "loss": 0.5847,
+      "step": 983
+    },
+    {
+      "epoch": 0.779135942990053,
+      "grad_norm": 0.2459839841353052,
+      "learning_rate": 1.6897051602684067e-06,
+      "loss": 0.5895,
+      "step": 984
+    },
+    {
+      "epoch": 0.7799277478101648,
+      "grad_norm": 0.2523185194394151,
+      "learning_rate": 1.689102964656027e-06,
+      "loss": 0.5763,
+      "step": 985
+    },
+    {
+      "epoch": 0.7807195526302766,
+      "grad_norm": 0.2514690273057696,
+      "learning_rate": 1.688500292799058e-06,
+      "loss": 0.5891,
+      "step": 986
+    },
+    {
+      "epoch": 0.7815113574503885,
+      "grad_norm": 0.2554984712579457,
+      "learning_rate": 1.6878971451140112e-06,
+      "loss": 0.5871,
+      "step": 987
+    },
+    {
+      "epoch": 0.7823031622705003,
+      "grad_norm": 0.24460510413890155,
+      "learning_rate": 1.687293522017726e-06,
+      "loss": 0.5719,
+      "step": 988
+    },
+    {
+      "epoch": 0.7830949670906122,
+      "grad_norm": 0.25006028531547964,
+      "learning_rate": 1.6866894239273716e-06,
+      "loss": 0.5879,
+      "step": 989
+    },
+    {
+      "epoch": 0.783886771910724,
+      "grad_norm": 0.2720793329328961,
+      "learning_rate": 1.6860848512604449e-06,
+      "loss": 0.5805,
+      "step": 990
+    },
+    {
+      "epoch": 0.7846785767308359,
+      "grad_norm": 0.26031706750706746,
+      "learning_rate": 1.6854798044347708e-06,
+      "loss": 0.6049,
+      "step": 991
+    },
+    {
+      "epoch": 0.7854703815509477,
+      "grad_norm": 0.25280789896335876,
+      "learning_rate": 1.6848742838685026e-06,
+      "loss": 0.5813,
+      "step": 992
+    },
+    {
+      "epoch": 0.7862621863710595,
+      "grad_norm": 0.2686687521746721,
+      "learning_rate": 1.6842682899801198e-06,
+      "loss": 0.5984,
+      "step": 993
+    },
+    {
+      "epoch": 0.7870539911911714,
+      "grad_norm": 0.2696539128983892,
+      "learning_rate": 1.6836618231884299e-06,
+      "loss": 0.5746,
+      "step": 994
+    },
+    {
+      "epoch": 0.7878457960112832,
+      "grad_norm": 0.24873630460097296,
+      "learning_rate": 1.6830548839125672e-06,
+      "loss": 0.5792,
+      "step": 995
+    },
+    {
+      "epoch": 0.7886376008313951,
+      "grad_norm": 0.25588691360488625,
+      "learning_rate": 1.6824474725719916e-06,
+      "loss": 0.5781,
+      "step": 996
+    },
+    {
+      "epoch": 0.7894294056515069,
+      "grad_norm": 0.2593738561169873,
+      "learning_rate": 1.6818395895864908e-06,
+      "loss": 0.592,
+      "step": 997
+    },
+    {
+      "epoch": 0.7902212104716188,
+      "grad_norm": 0.2503435439903613,
+      "learning_rate": 1.681231235376177e-06,
+      "loss": 0.5796,
+      "step": 998
+    },
+    {
+      "epoch": 0.7910130152917306,
+      "grad_norm": 0.24341615985355178,
+      "learning_rate": 1.6806224103614887e-06,
+      "loss": 0.598,
+      "step": 999
+    },
+    {
+      "epoch": 0.7918048201118424,
+      "grad_norm": 0.23918323278794373,
+      "learning_rate": 1.6800131149631907e-06,
+      "loss": 0.5857,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7925966249319543,
+      "grad_norm": 0.27123704462934584,
+      "learning_rate": 1.6794033496023704e-06,
+      "loss": 0.5856,
+      "step": 1001
+    },
+    {
+      "epoch": 0.7933884297520661,
+      "grad_norm": 0.24976229528373015,
+      "learning_rate": 1.6787931147004426e-06,
+      "loss": 0.5697,
+      "step": 1002
+    },
+    {
+      "epoch": 0.794180234572178,
+      "grad_norm": 0.25081951371265254,
+      "learning_rate": 1.6781824106791453e-06,
+      "loss": 0.6036,
+      "step": 1003
+    },
+    {
+      "epoch": 0.7949720393922898,
+      "grad_norm": 0.2543707728394622,
+      "learning_rate": 1.6775712379605409e-06,
+      "loss": 0.5757,
+      "step": 1004
+    },
+    {
+      "epoch": 0.7957638442124016,
+      "grad_norm": 0.2517746465053035,
+      "learning_rate": 1.6769595969670158e-06,
+      "loss": 0.5631,
+      "step": 1005
+    },
+    {
+      "epoch": 0.7965556490325135,
+      "grad_norm": 0.2505079887028291,
+      "learning_rate": 1.67634748812128e-06,
+      "loss": 0.5649,
+      "step": 1006
+    },
+    {
+      "epoch": 0.7973474538526253,
+      "grad_norm": 0.25100526219511066,
+      "learning_rate": 1.6757349118463666e-06,
+      "loss": 0.586,
+      "step": 1007
+    },
+    {
+      "epoch": 0.7981392586727372,
+      "grad_norm": 0.25008384189162813,
+      "learning_rate": 1.6751218685656325e-06,
+      "loss": 0.587,
+      "step": 1008
+    },
+    {
+      "epoch": 0.798931063492849,
+      "grad_norm": 0.2571108017946599,
+      "learning_rate": 1.6745083587027562e-06,
+      "loss": 0.5875,
+      "step": 1009
+    },
+    {
+      "epoch": 0.7997228683129609,
+      "grad_norm": 0.24575800730786365,
+      "learning_rate": 1.6738943826817397e-06,
+      "loss": 0.5772,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8005146731330727,
+      "grad_norm": 0.25158102693738865,
+      "learning_rate": 1.6732799409269069e-06,
+      "loss": 0.5866,
+      "step": 1011
+    },
+    {
+      "epoch": 0.8013064779531845,
+      "grad_norm": 0.2583811964469298,
+      "learning_rate": 1.672665033862903e-06,
+      "loss": 0.5719,
+      "step": 1012
+    },
+    {
+      "epoch": 0.8020982827732964,
+      "grad_norm": 0.24370943351611393,
+      "learning_rate": 1.6720496619146955e-06,
+      "loss": 0.5778,
+      "step": 1013
+    },
+    {
+      "epoch": 0.8028900875934082,
+      "grad_norm": 0.25906184718775943,
+      "learning_rate": 1.6714338255075728e-06,
+      "loss": 0.5775,
+      "step": 1014
+    },
+    {
+      "epoch": 0.8036818924135201,
+      "grad_norm": 0.2511585348986055,
+      "learning_rate": 1.6708175250671444e-06,
+      "loss": 0.5716,
+      "step": 1015
+    },
+    {
+      "epoch": 0.8044736972336319,
+      "grad_norm": 0.24707065260669275,
+      "learning_rate": 1.6702007610193403e-06,
+      "loss": 0.5888,
+      "step": 1016
+    },
+    {
+      "epoch": 0.8052655020537437,
+      "grad_norm": 0.24987917740310078,
+      "learning_rate": 1.6695835337904117e-06,
+      "loss": 0.5717,
+      "step": 1017
+    },
+    {
+      "epoch": 0.8060573068738556,
+      "grad_norm": 0.25196440898953126,
+      "learning_rate": 1.6689658438069284e-06,
+      "loss": 0.5666,
+      "step": 1018
+    },
+    {
+      "epoch": 0.8068491116939674,
+      "grad_norm": 0.26111092994260254,
+      "learning_rate": 1.6683476914957812e-06,
+      "loss": 0.5692,
+      "step": 1019
+    },
+    {
+      "epoch": 0.8076409165140793,
+      "grad_norm": 0.24716955694239662,
+      "learning_rate": 1.6677290772841807e-06,
+      "loss": 0.5736,
+      "step": 1020
+    },
+    {
+      "epoch": 0.8084327213341911,
+      "grad_norm": 0.2519859731877666,
+      "learning_rate": 1.6671100015996556e-06,
+      "loss": 0.5752,
+      "step": 1021
+    },
+    {
+      "epoch": 0.809224526154303,
+      "grad_norm": 0.2403464065951902,
+      "learning_rate": 1.666490464870054e-06,
+      "loss": 0.5955,
+      "step": 1022
+    },
+    {
+      "epoch": 0.8100163309744148,
+      "grad_norm": 0.2567687570720557,
+      "learning_rate": 1.6658704675235434e-06,
+      "loss": 0.6026,
+      "step": 1023
+    },
+    {
+      "epoch": 0.8108081357945266,
+      "grad_norm": 0.23933427768870075,
+      "learning_rate": 1.665250009988608e-06,
+      "loss": 0.5652,
+      "step": 1024
+    },
+    {
+      "epoch": 0.8115999406146385,
+      "grad_norm": 0.2442575208595484,
+      "learning_rate": 1.6646290926940512e-06,
+      "loss": 0.5802,
+      "step": 1025
+    },
+    {
+      "epoch": 0.8123917454347503,
+      "grad_norm": 0.25096321452317977,
+      "learning_rate": 1.6640077160689945e-06,
+      "loss": 0.5776,
+      "step": 1026
+    },
+    {
+      "epoch": 0.8131835502548622,
+      "grad_norm": 0.2533874645488365,
+      "learning_rate": 1.6633858805428758e-06,
+      "loss": 0.5835,
+      "step": 1027
+    },
+    {
+      "epoch": 0.813975355074974,
+      "grad_norm": 0.2543786497379953,
+      "learning_rate": 1.6627635865454507e-06,
+      "loss": 0.5869,
+      "step": 1028
+    },
+    {
+      "epoch": 0.8147671598950859,
+      "grad_norm": 0.2546574175602153,
+      "learning_rate": 1.662140834506792e-06,
+      "loss": 0.5873,
+      "step": 1029
+    },
+    {
+      "epoch": 0.8155589647151977,
+      "grad_norm": 0.24039403366732648,
+      "learning_rate": 1.661517624857288e-06,
+      "loss": 0.5739,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8163507695353095,
+      "grad_norm": 0.24657766248695567,
+      "learning_rate": 1.660893958027644e-06,
+      "loss": 0.5831,
+      "step": 1031
+    },
+    {
+      "epoch": 0.8171425743554214,
+      "grad_norm": 0.24389759783969212,
+      "learning_rate": 1.6602698344488816e-06,
+      "loss": 0.5872,
+      "step": 1032
+    },
+    {
+      "epoch": 0.8179343791755332,
+      "grad_norm": 0.2554998501325067,
+      "learning_rate": 1.6596452545523368e-06,
+      "loss": 0.568,
+      "step": 1033
+    },
+    {
+      "epoch": 0.8187261839956451,
+      "grad_norm": 0.24487900079848257,
+      "learning_rate": 1.6590202187696626e-06,
+      "loss": 0.5778,
+      "step": 1034
+    },
+    {
+      "epoch": 0.8195179888157569,
+      "grad_norm": 0.2409374318162725,
+      "learning_rate": 1.6583947275328262e-06,
+      "loss": 0.5726,
+      "step": 1035
+    },
+    {
+      "epoch": 0.8203097936358688,
+      "grad_norm": 0.2555336201677589,
+      "learning_rate": 1.6577687812741092e-06,
+      "loss": 0.5763,
+      "step": 1036
+    },
+    {
+      "epoch": 0.8211015984559806,
+      "grad_norm": 0.24788955965702197,
+      "learning_rate": 1.657142380426108e-06,
+      "loss": 0.5738,
+      "step": 1037
+    },
+    {
+      "epoch": 0.8218934032760924,
+      "grad_norm": 0.2434532019250245,
+      "learning_rate": 1.6565155254217335e-06,
+      "loss": 0.5849,
+      "step": 1038
+    },
+    {
+      "epoch": 0.8226852080962043,
+      "grad_norm": 0.25339375614835435,
+      "learning_rate": 1.6558882166942105e-06,
+      "loss": 0.6043,
+      "step": 1039
+    },
+    {
+      "epoch": 0.8234770129163161,
+      "grad_norm": 0.25543803777050145,
+      "learning_rate": 1.6552604546770766e-06,
+      "loss": 0.5861,
+      "step": 1040
+    },
+    {
+      "epoch": 0.824268817736428,
+      "grad_norm": 0.2651661036586476,
+      "learning_rate": 1.6546322398041832e-06,
+      "loss": 0.5689,
+      "step": 1041
+    },
+    {
+      "epoch": 0.8250606225565398,
+      "grad_norm": 0.24732454750675104,
+      "learning_rate": 1.654003572509695e-06,
+      "loss": 0.5632,
+      "step": 1042
+    },
+    {
+      "epoch": 0.8258524273766517,
+      "grad_norm": 0.2542908482879605,
+      "learning_rate": 1.653374453228089e-06,
+      "loss": 0.5957,
+      "step": 1043
+    },
+    {
+      "epoch": 0.8266442321967635,
+      "grad_norm": 0.25243605180228523,
+      "learning_rate": 1.6527448823941543e-06,
+      "loss": 0.5893,
+      "step": 1044
+    },
+    {
+      "epoch": 0.8274360370168753,
+      "grad_norm": 0.26166682310011685,
+      "learning_rate": 1.6521148604429925e-06,
+      "loss": 0.5986,
+      "step": 1045
+    },
+    {
+      "epoch": 0.8282278418369872,
+      "grad_norm": 0.24835269137254626,
+      "learning_rate": 1.6514843878100163e-06,
+      "loss": 0.5743,
+      "step": 1046
+    },
+    {
+      "epoch": 0.829019646657099,
+      "grad_norm": 0.25029496480012176,
+      "learning_rate": 1.6508534649309516e-06,
+      "loss": 0.5858,
+      "step": 1047
+    },
+    {
+      "epoch": 0.8298114514772109,
+      "grad_norm": 0.2484934813624596,
+      "learning_rate": 1.650222092241833e-06,
+      "loss": 0.5779,
+      "step": 1048
+    },
+    {
+      "epoch": 0.8306032562973227,
+      "grad_norm": 0.24747045389662214,
+      "learning_rate": 1.649590270179008e-06,
+      "loss": 0.5825,
+      "step": 1049
+    },
+    {
+      "epoch": 0.8313950611174346,
+      "grad_norm": 0.25678312654227714,
+      "learning_rate": 1.6489579991791337e-06,
+      "loss": 0.5779,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8321868659375464,
+      "grad_norm": 0.25001373774852653,
+      "learning_rate": 1.6483252796791778e-06,
+      "loss": 0.5884,
+      "step": 1051
+    },
+    {
+      "epoch": 0.8329786707576582,
+      "grad_norm": 0.24649088580005798,
+      "learning_rate": 1.6476921121164176e-06,
+      "loss": 0.5987,
+      "step": 1052
+    },
+    {
+      "epoch": 0.8337704755777701,
+      "grad_norm": 0.2537535387527769,
+      "learning_rate": 1.6470584969284405e-06,
+      "loss": 0.5851,
+      "step": 1053
+    },
+    {
+      "epoch": 0.8345622803978819,
+      "grad_norm": 0.24145915608683344,
+      "learning_rate": 1.6464244345531423e-06,
+      "loss": 0.5722,
+      "step": 1054
+    },
+    {
+      "epoch": 0.8353540852179938,
+      "grad_norm": 0.24705155210352187,
+      "learning_rate": 1.64578992542873e-06,
+      "loss": 0.5831,
+      "step": 1055
+    },
+    {
+      "epoch": 0.8361458900381056,
+      "grad_norm": 0.24564210050744636,
+      "learning_rate": 1.6451549699937166e-06,
+      "loss": 0.5719,
+      "step": 1056
+    },
+    {
+      "epoch": 0.8369376948582175,
+      "grad_norm": 0.23904533252059798,
+      "learning_rate": 1.6445195686869259e-06,
+      "loss": 0.5941,
+      "step": 1057
+    },
+    {
+      "epoch": 0.8377294996783293,
+      "grad_norm": 0.2518695754783361,
+      "learning_rate": 1.6438837219474876e-06,
+      "loss": 0.5857,
+      "step": 1058
+    },
+    {
+      "epoch": 0.8385213044984411,
+      "grad_norm": 0.24304954610057472,
+      "learning_rate": 1.643247430214842e-06,
+      "loss": 0.5688,
+      "step": 1059
+    },
+    {
+      "epoch": 0.839313109318553,
+      "grad_norm": 0.2502788325538013,
+      "learning_rate": 1.6426106939287345e-06,
+      "loss": 0.5763,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8401049141386648,
+      "grad_norm": 0.23595313937091714,
+      "learning_rate": 1.6419735135292188e-06,
+      "loss": 0.5737,
+      "step": 1061
+    },
+    {
+      "epoch": 0.8408967189587767,
+      "grad_norm": 0.23719473568301383,
+      "learning_rate": 1.6413358894566553e-06,
+      "loss": 0.5767,
+      "step": 1062
+    },
+    {
+      "epoch": 0.8416885237788885,
+      "grad_norm": 0.24272651721099175,
+      "learning_rate": 1.6406978221517116e-06,
+      "loss": 0.5761,
+      "step": 1063
+    },
+    {
+      "epoch": 0.8424803285990003,
+      "grad_norm": 0.2642905302556802,
+      "learning_rate": 1.6400593120553606e-06,
+      "loss": 0.5588,
+      "step": 1064
+    },
+    {
+      "epoch": 0.8432721334191122,
+      "grad_norm": 0.2488237526188394,
+      "learning_rate": 1.639420359608882e-06,
+      "loss": 0.566,
+      "step": 1065
+    },
+    {
+      "epoch": 0.844063938239224,
+      "grad_norm": 0.25131458781350147,
+      "learning_rate": 1.6387809652538608e-06,
+      "loss": 0.5748,
+      "step": 1066
+    },
+    {
+      "epoch": 0.8448557430593359,
+      "grad_norm": 0.24645204360337702,
+      "learning_rate": 1.6381411294321876e-06,
+      "loss": 0.5829,
+      "step": 1067
+    },
+    {
+      "epoch": 0.8456475478794477,
+      "grad_norm": 0.2517558246624942,
+      "learning_rate": 1.6375008525860584e-06,
+      "loss": 0.5697,
+      "step": 1068
+    },
+    {
+      "epoch": 0.8464393526995596,
+      "grad_norm": 0.24583594853286805,
+      "learning_rate": 1.6368601351579732e-06,
+      "loss": 0.5699,
+      "step": 1069
+    },
+    {
+      "epoch": 0.8472311575196714,
+      "grad_norm": 0.2501707400525852,
+      "learning_rate": 1.6362189775907373e-06,
+      "loss": 0.5828,
+      "step": 1070
+    },
+    {
+      "epoch": 0.8480229623397832,
+      "grad_norm": 0.255049508563797,
+      "learning_rate": 1.6355773803274597e-06,
+      "loss": 0.5893,
+      "step": 1071
+    },
+    {
+      "epoch": 0.8488147671598951,
+      "grad_norm": 0.2539879381739073,
+      "learning_rate": 1.6349353438115535e-06,
+      "loss": 0.5725,
+      "step": 1072
+    },
+    {
+      "epoch": 0.8496065719800069,
+      "grad_norm": 0.24145487424539905,
+      "learning_rate": 1.6342928684867351e-06,
+      "loss": 0.5838,
+      "step": 1073
+    },
+    {
+      "epoch": 0.8503983768001188,
+      "grad_norm": 0.2520150705783432,
+      "learning_rate": 1.6336499547970246e-06,
+      "loss": 0.5687,
+      "step": 1074
+    },
+    {
+      "epoch": 0.8511901816202306,
+      "grad_norm": 0.2475644312970042,
+      "learning_rate": 1.6330066031867445e-06,
+      "loss": 0.5794,
+      "step": 1075
+    },
+    {
+      "epoch": 0.8519819864403425,
+      "grad_norm": 0.24259026078041718,
+      "learning_rate": 1.6323628141005206e-06,
+      "loss": 0.5689,
+      "step": 1076
+    },
+    {
+      "epoch": 0.8527737912604543,
+      "grad_norm": 0.2539525583876839,
+      "learning_rate": 1.6317185879832805e-06,
+      "loss": 0.5739,
+      "step": 1077
+    },
+    {
+      "epoch": 0.8535655960805661,
+      "grad_norm": 0.24574991580450073,
+      "learning_rate": 1.631073925280254e-06,
+      "loss": 0.58,
+      "step": 1078
+    },
+    {
+      "epoch": 0.854357400900678,
+      "grad_norm": 0.26035074540088715,
+      "learning_rate": 1.6304288264369727e-06,
+      "loss": 0.5709,
+      "step": 1079
+    },
+    {
+      "epoch": 0.8551492057207898,
+      "grad_norm": 0.25118729860231354,
+      "learning_rate": 1.6297832918992698e-06,
+      "loss": 0.577,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8559410105409017,
+      "grad_norm": 0.24088051605213703,
+      "learning_rate": 1.6291373221132793e-06,
+      "loss": 0.5702,
+      "step": 1081
+    },
+    {
+      "epoch": 0.8567328153610135,
+      "grad_norm": 0.2686187728277244,
+      "learning_rate": 1.628490917525436e-06,
+      "loss": 0.5916,
+      "step": 1082
+    },
+    {
+      "epoch": 0.8575246201811254,
+      "grad_norm": 0.23663725614238204,
+      "learning_rate": 1.6278440785824752e-06,
+      "loss": 0.572,
+      "step": 1083
+    },
+    {
+      "epoch": 0.8583164250012372,
+      "grad_norm": 0.25765616886797094,
+      "learning_rate": 1.6271968057314329e-06,
+      "loss": 0.5909,
+      "step": 1084
+    },
+    {
+      "epoch": 0.859108229821349,
+      "grad_norm": 0.24534663703048665,
+      "learning_rate": 1.6265490994196442e-06,
+      "loss": 0.5707,
+      "step": 1085
+    },
+    {
+      "epoch": 0.8599000346414609,
+      "grad_norm": 0.265161136643667,
+      "learning_rate": 1.6259009600947443e-06,
+      "loss": 0.572,
+      "step": 1086
+    },
+    {
+      "epoch": 0.8606918394615727,
+      "grad_norm": 0.24368310689432676,
+      "learning_rate": 1.625252388204667e-06,
+      "loss": 0.5872,
+      "step": 1087
+    },
+    {
+      "epoch": 0.8614836442816846,
+      "grad_norm": 0.2459058837532321,
+      "learning_rate": 1.6246033841976465e-06,
+      "loss": 0.5725,
+      "step": 1088
+    },
+    {
+      "epoch": 0.8622754491017964,
+      "grad_norm": 0.25441194174474324,
+      "learning_rate": 1.6239539485222139e-06,
+      "loss": 0.5888,
+      "step": 1089
+    },
+    {
+      "epoch": 0.8630672539219082,
+      "grad_norm": 0.24136222032791688,
+      "learning_rate": 1.6233040816271996e-06,
+      "loss": 0.5779,
+      "step": 1090
+    },
+    {
+      "epoch": 0.8638590587420201,
+      "grad_norm": 0.24333904193511258,
+      "learning_rate": 1.6226537839617321e-06,
+      "loss": 0.5834,
+      "step": 1091
+    },
+    {
+      "epoch": 0.8646508635621319,
+      "grad_norm": 0.24651480245565502,
+      "learning_rate": 1.6220030559752367e-06,
+      "loss": 0.5941,
+      "step": 1092
+    },
+    {
+      "epoch": 0.8654426683822438,
+      "grad_norm": 0.24766500609673905,
+      "learning_rate": 1.6213518981174374e-06,
+      "loss": 0.5679,
+      "step": 1093
+    },
+    {
+      "epoch": 0.8662344732023556,
+      "grad_norm": 0.25470747691542456,
+      "learning_rate": 1.6207003108383543e-06,
+      "loss": 0.5996,
+      "step": 1094
+    },
+    {
+      "epoch": 0.8670262780224675,
+      "grad_norm": 0.2460007287413227,
+      "learning_rate": 1.6200482945883046e-06,
+      "loss": 0.5808,
+      "step": 1095
+    },
+    {
+      "epoch": 0.8678180828425793,
+      "grad_norm": 0.25337461555203933,
+      "learning_rate": 1.6193958498179022e-06,
+      "loss": 0.5672,
+      "step": 1096
+    },
+    {
+      "epoch": 0.8686098876626911,
+      "grad_norm": 0.25552804611255225,
+      "learning_rate": 1.6187429769780568e-06,
+      "loss": 0.5723,
+      "step": 1097
+    },
+    {
+      "epoch": 0.869401692482803,
+      "grad_norm": 0.2420545397711388,
+      "learning_rate": 1.6180896765199737e-06,
+      "loss": 0.5778,
+      "step": 1098
+    },
+    {
+      "epoch": 0.8701934973029148,
+      "grad_norm": 0.25071874604846495,
+      "learning_rate": 1.6174359488951545e-06,
+      "loss": 0.5847,
+      "step": 1099
+    },
+    {
+      "epoch": 0.8709853021230267,
+      "grad_norm": 0.26163100705727776,
+      "learning_rate": 1.6167817945553956e-06,
+      "loss": 0.5891,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8717771069431385,
+      "grad_norm": 0.2553614991918954,
+      "learning_rate": 1.6161272139527882e-06,
+      "loss": 0.5691,
+      "step": 1101
+    },
+    {
+      "epoch": 0.8725689117632504,
+      "grad_norm": 0.24966352488537705,
+      "learning_rate": 1.6154722075397183e-06,
+      "loss": 0.5848,
+      "step": 1102
+    },
+    {
+      "epoch": 0.8733607165833622,
+      "grad_norm": 0.24514512458122928,
+      "learning_rate": 1.6148167757688657e-06,
+      "loss": 0.577,
+      "step": 1103
+    },
+    {
+      "epoch": 0.874152521403474,
+      "grad_norm": 0.2637944713446023,
+      "learning_rate": 1.6141609190932051e-06,
+      "loss": 0.5941,
+      "step": 1104
+    },
+    {
+      "epoch": 0.8749443262235859,
+      "grad_norm": 0.2456325755441037,
+      "learning_rate": 1.6135046379660037e-06,
+      "loss": 0.588,
+      "step": 1105
+    },
+    {
+      "epoch": 0.8757361310436977,
+      "grad_norm": 0.23929723192140198,
+      "learning_rate": 1.6128479328408236e-06,
+      "loss": 0.5803,
+      "step": 1106
+    },
+    {
+      "epoch": 0.8765279358638096,
+      "grad_norm": 0.25769065276824593,
+      "learning_rate": 1.6121908041715181e-06,
+      "loss": 0.5828,
+      "step": 1107
+    },
+    {
+      "epoch": 0.8773197406839214,
+      "grad_norm": 0.26935441628478257,
+      "learning_rate": 1.6115332524122345e-06,
+      "loss": 0.5796,
+      "step": 1108
+    },
+    {
+      "epoch": 0.8781115455040333,
+      "grad_norm": 0.24643537762947543,
+      "learning_rate": 1.610875278017412e-06,
+      "loss": 0.5792,
+      "step": 1109
+    },
+    {
+      "epoch": 0.8789033503241451,
+      "grad_norm": 0.25353326502965123,
+      "learning_rate": 1.610216881441782e-06,
+      "loss": 0.5905,
+      "step": 1110
+    },
+    {
+      "epoch": 0.8796951551442569,
+      "grad_norm": 0.24281644334466226,
+      "learning_rate": 1.6095580631403676e-06,
+      "loss": 0.5937,
+      "step": 1111
+    },
+    {
+      "epoch": 0.8804869599643688,
+      "grad_norm": 0.25565001965436085,
+      "learning_rate": 1.6088988235684834e-06,
+      "loss": 0.5859,
+      "step": 1112
+    },
+    {
+      "epoch": 0.8812787647844806,
+      "grad_norm": 0.2541655484308428,
+      "learning_rate": 1.6082391631817354e-06,
+      "loss": 0.5829,
+      "step": 1113
+    },
+    {
+      "epoch": 0.8820705696045925,
+      "grad_norm": 0.24937428478641344,
+      "learning_rate": 1.60757908243602e-06,
+      "loss": 0.5668,
+      "step": 1114
+    },
+    {
+      "epoch": 0.8828623744247043,
+      "grad_norm": 0.24928470722270202,
+      "learning_rate": 1.6069185817875242e-06,
+      "loss": 0.5815,
+      "step": 1115
+    },
+    {
+      "epoch": 0.8836541792448162,
+      "grad_norm": 0.2560545020452616,
+      "learning_rate": 1.6062576616927256e-06,
+      "loss": 0.5887,
+      "step": 1116
+    },
+    {
+      "epoch": 0.884445984064928,
+      "grad_norm": 0.27127526912359223,
+      "learning_rate": 1.6055963226083909e-06,
+      "loss": 0.5969,
+      "step": 1117
+    },
+    {
+      "epoch": 0.8852377888850398,
+      "grad_norm": 0.2514824688319713,
+      "learning_rate": 1.6049345649915775e-06,
+      "loss": 0.5842,
+      "step": 1118
+    },
+    {
+      "epoch": 0.8860295937051517,
+      "grad_norm": 0.24375986127866897,
+      "learning_rate": 1.6042723892996309e-06,
+      "loss": 0.5777,
+      "step": 1119
+    },
+    {
+      "epoch": 0.8868213985252635,
+      "grad_norm": 0.2546590226484071,
+      "learning_rate": 1.6036097959901862e-06,
+      "loss": 0.5874,
+      "step": 1120
+    },
+    {
+      "epoch": 0.8876132033453754,
+      "grad_norm": 0.25490643966330057,
+      "learning_rate": 1.602946785521167e-06,
+      "loss": 0.5829,
+      "step": 1121
+    },
+    {
+      "epoch": 0.8884050081654872,
+      "grad_norm": 0.2490222536319817,
+      "learning_rate": 1.6022833583507848e-06,
+      "loss": 0.5798,
+      "step": 1122
+    },
+    {
+      "epoch": 0.889196812985599,
+      "grad_norm": 0.25536422958159516,
+      "learning_rate": 1.60161951493754e-06,
+      "loss": 0.5954,
+      "step": 1123
+    },
+    {
+      "epoch": 0.8899886178057109,
+      "grad_norm": 0.24968603426729527,
+      "learning_rate": 1.6009552557402198e-06,
+      "loss": 0.5752,
+      "step": 1124
+    },
+    {
+      "epoch": 0.8907804226258227,
+      "grad_norm": 0.24393303439491235,
+      "learning_rate": 1.6002905812178988e-06,
+      "loss": 0.5743,
+      "step": 1125
+    },
+    {
+      "epoch": 0.8915722274459346,
+      "grad_norm": 0.24528004931946223,
+      "learning_rate": 1.59962549182994e-06,
+      "loss": 0.5691,
+      "step": 1126
+    },
+    {
+      "epoch": 0.8923640322660464,
+      "grad_norm": 0.24646740204639261,
+      "learning_rate": 1.5989599880359906e-06,
+      "loss": 0.5688,
+      "step": 1127
+    },
+    {
+      "epoch": 0.8931558370861583,
+      "grad_norm": 0.2529708580577012,
+      "learning_rate": 1.5982940702959864e-06,
+      "loss": 0.5744,
+      "step": 1128
+    },
+    {
+      "epoch": 0.8939476419062701,
+      "grad_norm": 0.24649907291895468,
+      "learning_rate": 1.597627739070148e-06,
+      "loss": 0.5754,
+      "step": 1129
+    },
+    {
+      "epoch": 0.894739446726382,
+      "grad_norm": 0.25002828092540585,
+      "learning_rate": 1.596960994818983e-06,
+      "loss": 0.5862,
+      "step": 1130
+    },
+    {
+      "epoch": 0.8955312515464938,
+      "grad_norm": 0.24390990267195928,
+      "learning_rate": 1.5962938380032833e-06,
+      "loss": 0.5766,
+      "step": 1131
+    },
+    {
+      "epoch": 0.8963230563666056,
+      "grad_norm": 0.2736736092042765,
+      "learning_rate": 1.595626269084126e-06,
+      "loss": 0.598,
+      "step": 1132
+    },
+    {
+      "epoch": 0.8971148611867175,
+      "grad_norm": 0.25173453453576516,
+      "learning_rate": 1.5949582885228741e-06,
+      "loss": 0.5731,
+      "step": 1133
+    },
+    {
+      "epoch": 0.8979066660068293,
+      "grad_norm": 0.24310431339486674,
+      "learning_rate": 1.5942898967811735e-06,
+      "loss": 0.5721,
+      "step": 1134
+    },
+    {
+      "epoch": 0.8986984708269412,
+      "grad_norm": 0.2508954998227375,
+      "learning_rate": 1.5936210943209556e-06,
+      "loss": 0.5712,
+      "step": 1135
+    },
+    {
+      "epoch": 0.899490275647053,
+      "grad_norm": 0.25916566992709084,
+      "learning_rate": 1.592951881604435e-06,
+      "loss": 0.5665,
+      "step": 1136
+    },
+    {
+      "epoch": 0.9002820804671648,
+      "grad_norm": 0.2609148285652845,
+      "learning_rate": 1.5922822590941101e-06,
+      "loss": 0.5934,
+      "step": 1137
+    },
+    {
+      "epoch": 0.9010738852872767,
+      "grad_norm": 0.24396936865607036,
+      "learning_rate": 1.5916122272527622e-06,
+      "loss": 0.5819,
+      "step": 1138
+    },
+    {
+      "epoch": 0.9018656901073885,
+      "grad_norm": 0.24518618613401058,
+      "learning_rate": 1.590941786543456e-06,
+      "loss": 0.5727,
+      "step": 1139
+    },
+    {
+      "epoch": 0.9026574949275004,
+      "grad_norm": 0.270738638719993,
+      "learning_rate": 1.5902709374295378e-06,
+      "loss": 0.5752,
+      "step": 1140
+    },
+    {
+      "epoch": 0.9034492997476122,
+      "grad_norm": 0.2538745924611469,
+      "learning_rate": 1.5895996803746377e-06,
+      "loss": 0.5749,
+      "step": 1141
+    },
+    {
+      "epoch": 0.9042411045677241,
+      "grad_norm": 0.24372233219327333,
+      "learning_rate": 1.588928015842666e-06,
+      "loss": 0.568,
+      "step": 1142
+    },
+    {
+      "epoch": 0.9050329093878359,
+      "grad_norm": 0.25103700311305904,
+      "learning_rate": 1.5882559442978161e-06,
+      "loss": 0.5814,
+      "step": 1143
+    },
+    {
+      "epoch": 0.9058247142079477,
+      "grad_norm": 0.25214324010928046,
+      "learning_rate": 1.587583466204562e-06,
+      "loss": 0.5752,
+      "step": 1144
+    },
+    {
+      "epoch": 0.9066165190280596,
+      "grad_norm": 0.26722987577286583,
+      "learning_rate": 1.586910582027658e-06,
+      "loss": 0.5807,
+      "step": 1145
+    },
+    {
+      "epoch": 0.9074083238481714,
+      "grad_norm": 0.2428861218982861,
+      "learning_rate": 1.5862372922321403e-06,
+      "loss": 0.5805,
+      "step": 1146
+    },
+    {
+      "epoch": 0.9082001286682833,
+      "grad_norm": 0.2373379401322153,
+      "learning_rate": 1.585563597283325e-06,
+      "loss": 0.5759,
+      "step": 1147
+    },
+    {
+      "epoch": 0.9089919334883951,
+      "grad_norm": 0.23509882651000052,
+      "learning_rate": 1.5848894976468079e-06,
+      "loss": 0.5962,
+      "step": 1148
+    },
+    {
+      "epoch": 0.909783738308507,
+      "grad_norm": 0.2534606776590353,
+      "learning_rate": 1.5842149937884645e-06,
+      "loss": 0.5717,
+      "step": 1149
+    },
+    {
+      "epoch": 0.9105755431286188,
+      "grad_norm": 0.24492279368136252,
+      "learning_rate": 1.5835400861744504e-06,
+      "loss": 0.5799,
+      "step": 1150
+    },
+    {
+      "epoch": 0.9113673479487306,
+      "grad_norm": 0.23969145535922307,
+      "learning_rate": 1.582864775271199e-06,
+      "loss": 0.5955,
+      "step": 1151
+    },
+    {
+      "epoch": 0.9121591527688425,
+      "grad_norm": 0.24042106482444472,
+      "learning_rate": 1.5821890615454235e-06,
+      "loss": 0.5906,
+      "step": 1152
+    },
+    {
+      "epoch": 0.9129509575889543,
+      "grad_norm": 0.25129518800116574,
+      "learning_rate": 1.5815129454641144e-06,
+      "loss": 0.5743,
+      "step": 1153
+    },
+    {
+      "epoch": 0.9137427624090662,
+      "grad_norm": 0.2545510658300164,
+      "learning_rate": 1.580836427494542e-06,
+      "loss": 0.5925,
+      "step": 1154
+    },
+    {
+      "epoch": 0.914534567229178,
+      "grad_norm": 0.24624218829669162,
+      "learning_rate": 1.5801595081042524e-06,
+      "loss": 0.5945,
+      "step": 1155
+    },
+    {
+      "epoch": 0.9153263720492899,
+      "grad_norm": 0.24503633144908887,
+      "learning_rate": 1.579482187761071e-06,
+      "loss": 0.5595,
+      "step": 1156
+    },
+    {
+      "epoch": 0.9161181768694017,
+      "grad_norm": 0.23811483557631802,
+      "learning_rate": 1.5788044669330982e-06,
+      "loss": 0.5615,
+      "step": 1157
+    },
+    {
+      "epoch": 0.9169099816895135,
+      "grad_norm": 0.23977190697285644,
+      "learning_rate": 1.5781263460887134e-06,
+      "loss": 0.5685,
+      "step": 1158
+    },
+    {
+      "epoch": 0.9177017865096254,
+      "grad_norm": 0.24954718549245242,
+      "learning_rate": 1.57744782569657e-06,
+      "loss": 0.5949,
+      "step": 1159
+    },
+    {
+      "epoch": 0.9184935913297372,
+      "grad_norm": 0.2568842831723131,
+      "learning_rate": 1.5767689062256002e-06,
+      "loss": 0.5929,
+      "step": 1160
+    },
+    {
+      "epoch": 0.919285396149849,
+      "grad_norm": 0.2544427897096277,
+      "learning_rate": 1.57608958814501e-06,
+      "loss": 0.5801,
+      "step": 1161
+    },
+    {
+      "epoch": 0.9200772009699609,
+      "grad_norm": 0.24426385349699706,
+      "learning_rate": 1.575409871924282e-06,
+      "loss": 0.5755,
+      "step": 1162
+    },
+    {
+      "epoch": 0.9208690057900727,
+      "grad_norm": 0.2365754597960453,
+      "learning_rate": 1.574729758033173e-06,
+      "loss": 0.5632,
+      "step": 1163
+    },
+    {
+      "epoch": 0.9216608106101846,
+      "grad_norm": 0.2505054501781746,
+      "learning_rate": 1.5740492469417158e-06,
+      "loss": 0.5689,
+      "step": 1164
+    },
+    {
+      "epoch": 0.9224526154302964,
+      "grad_norm": 0.2479162053654987,
+      "learning_rate": 1.5733683391202167e-06,
+      "loss": 0.573,
+      "step": 1165
+    },
+    {
+      "epoch": 0.9232444202504083,
+      "grad_norm": 0.24524728490335904,
+      "learning_rate": 1.572687035039256e-06,
+      "loss": 0.5929,
+      "step": 1166
+    },
+    {
+      "epoch": 0.9240362250705201,
+      "grad_norm": 0.25587349357699457,
+      "learning_rate": 1.5720053351696896e-06,
+      "loss": 0.5815,
+      "step": 1167
+    },
+    {
+      "epoch": 0.924828029890632,
+      "grad_norm": 0.24760385439967084,
+      "learning_rate": 1.5713232399826453e-06,
+      "loss": 0.5889,
+      "step": 1168
+    },
+    {
+      "epoch": 0.9256198347107438,
+      "grad_norm": 0.24903741397827486,
+      "learning_rate": 1.5706407499495239e-06,
+      "loss": 0.567,
+      "step": 1169
+    },
+    {
+      "epoch": 0.9264116395308556,
+      "grad_norm": 0.26127318105688313,
+      "learning_rate": 1.5699578655420007e-06,
+      "loss": 0.5804,
+      "step": 1170
+    },
+    {
+      "epoch": 0.9272034443509675,
+      "grad_norm": 0.252238856846934,
+      "learning_rate": 1.5692745872320216e-06,
+      "loss": 0.5815,
+      "step": 1171
+    },
+    {
+      "epoch": 0.9279952491710793,
+      "grad_norm": 0.24192869754410945,
+      "learning_rate": 1.5685909154918067e-06,
+      "loss": 0.566,
+      "step": 1172
+    },
+    {
+      "epoch": 0.9287870539911912,
+      "grad_norm": 0.24442472614445832,
+      "learning_rate": 1.5679068507938466e-06,
+      "loss": 0.5688,
+      "step": 1173
+    },
+    {
+      "epoch": 0.929578858811303,
+      "grad_norm": 0.2521094423336885,
+      "learning_rate": 1.567222393610904e-06,
+      "loss": 0.5765,
+      "step": 1174
+    },
+    {
+      "epoch": 0.9303706636314149,
+      "grad_norm": 0.24684251870455634,
+      "learning_rate": 1.5665375444160126e-06,
+      "loss": 0.5873,
+      "step": 1175
+    },
+    {
+      "epoch": 0.9311624684515267,
+      "grad_norm": 0.24192921881193505,
+      "learning_rate": 1.565852303682477e-06,
+      "loss": 0.5773,
+      "step": 1176
+    },
+    {
+      "epoch": 0.9319542732716385,
+      "grad_norm": 0.24349858877928449,
+      "learning_rate": 1.565166671883873e-06,
+      "loss": 0.584,
+      "step": 1177
+    },
+    {
+      "epoch": 0.9327460780917504,
+      "grad_norm": 0.24894847440005802,
+      "learning_rate": 1.5644806494940462e-06,
+      "loss": 0.5741,
+      "step": 1178
+    },
+    {
+      "epoch": 0.9335378829118622,
+      "grad_norm": 0.23837509286108713,
+      "learning_rate": 1.5637942369871117e-06,
+      "loss": 0.5917,
+      "step": 1179
+    },
+    {
+      "epoch": 0.9343296877319741,
+      "grad_norm": 0.2568809865381604,
+      "learning_rate": 1.5631074348374553e-06,
+      "loss": 0.5794,
+      "step": 1180
+    },
+    {
+      "epoch": 0.9351214925520859,
+      "grad_norm": 0.24918522428776926,
+      "learning_rate": 1.562420243519731e-06,
+      "loss": 0.5838,
+      "step": 1181
+    },
+    {
+      "epoch": 0.9359132973721977,
+      "grad_norm": 0.24540553438114424,
+      "learning_rate": 1.5617326635088625e-06,
+      "loss": 0.5789,
+      "step": 1182
+    },
+    {
+      "epoch": 0.9367051021923096,
+      "grad_norm": 0.24774553423368362,
+      "learning_rate": 1.5610446952800422e-06,
+      "loss": 0.5848,
+      "step": 1183
+    },
+    {
+      "epoch": 0.9374969070124214,
+      "grad_norm": 0.2660905672231104,
+      "learning_rate": 1.5603563393087296e-06,
+      "loss": 0.5681,
+      "step": 1184
+    },
+    {
+      "epoch": 0.9382887118325333,
+      "grad_norm": 0.2427522788591016,
+      "learning_rate": 1.5596675960706534e-06,
+      "loss": 0.5858,
+      "step": 1185
+    },
+    {
+      "epoch": 0.9390805166526451,
+      "grad_norm": 0.24652400083416762,
+      "learning_rate": 1.5589784660418102e-06,
+      "loss": 0.5695,
+      "step": 1186
+    },
+    {
+      "epoch": 0.939872321472757,
+      "grad_norm": 0.24013995413355602,
+      "learning_rate": 1.5582889496984621e-06,
+      "loss": 0.5857,
+      "step": 1187
+    },
+    {
+      "epoch": 0.9406641262928688,
+      "grad_norm": 0.24921304508425615,
+      "learning_rate": 1.5575990475171405e-06,
+      "loss": 0.5793,
+      "step": 1188
+    },
+    {
+      "epoch": 0.9414559311129806,
+      "grad_norm": 0.24070654325224555,
+      "learning_rate": 1.556908759974642e-06,
+      "loss": 0.5845,
+      "step": 1189
+    },
+    {
+      "epoch": 0.9422477359330925,
+      "grad_norm": 0.25443445419963834,
+      "learning_rate": 1.5562180875480296e-06,
+      "loss": 0.5685,
+      "step": 1190
+    },
+    {
+      "epoch": 0.9430395407532043,
+      "grad_norm": 0.24770278707419008,
+      "learning_rate": 1.5555270307146326e-06,
+      "loss": 0.589,
+      "step": 1191
+    },
+    {
+      "epoch": 0.9438313455733162,
+      "grad_norm": 0.2525597333456365,
+      "learning_rate": 1.554835589952046e-06,
+      "loss": 0.5891,
+      "step": 1192
+    },
+    {
+      "epoch": 0.944623150393428,
+      "grad_norm": 0.24177694953763318,
+      "learning_rate": 1.5541437657381303e-06,
+      "loss": 0.5869,
+      "step": 1193
+    },
+    {
+      "epoch": 0.9454149552135399,
+      "grad_norm": 0.24453407397790725,
+      "learning_rate": 1.5534515585510105e-06,
+      "loss": 0.5855,
+      "step": 1194
+    },
+    {
+      "epoch": 0.9462067600336517,
+      "grad_norm": 0.24980528629373336,
+      "learning_rate": 1.5527589688690763e-06,
+      "loss": 0.5943,
+      "step": 1195
+    },
+    {
+      "epoch": 0.9469985648537635,
+      "grad_norm": 0.23728415171069342,
+      "learning_rate": 1.5520659971709829e-06,
+      "loss": 0.5894,
+      "step": 1196
+    },
+    {
+      "epoch": 0.9477903696738754,
+      "grad_norm": 0.24119160485242816,
+      "learning_rate": 1.5513726439356475e-06,
+      "loss": 0.5889,
+      "step": 1197
+    },
+    {
+      "epoch": 0.9485821744939872,
+      "grad_norm": 0.25828470554024263,
+      "learning_rate": 1.5506789096422526e-06,
+      "loss": 0.5784,
+      "step": 1198
+    },
+    {
+      "epoch": 0.9493739793140991,
+      "grad_norm": 0.2387583307346421,
+      "learning_rate": 1.5499847947702436e-06,
+      "loss": 0.5715,
+      "step": 1199
+    },
+    {
+      "epoch": 0.9501657841342109,
+      "grad_norm": 0.2577285083898989,
+      "learning_rate": 1.549290299799329e-06,
+      "loss": 0.5792,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9509575889543228,
+      "grad_norm": 0.24901630093178415,
+      "learning_rate": 1.5485954252094792e-06,
+      "loss": 0.5697,
+      "step": 1201
+    },
+    {
+      "epoch": 0.9517493937744346,
+      "grad_norm": 0.25010870398297697,
+      "learning_rate": 1.5479001714809283e-06,
+      "loss": 0.5774,
+      "step": 1202
+    },
+    {
+      "epoch": 0.9525411985945464,
+      "grad_norm": 0.24680871093997186,
+      "learning_rate": 1.5472045390941713e-06,
+      "loss": 0.5932,
+      "step": 1203
+    },
+    {
+      "epoch": 0.9533330034146583,
+      "grad_norm": 0.24422855655402023,
+      "learning_rate": 1.5465085285299658e-06,
+      "loss": 0.5849,
+      "step": 1204
+    },
+    {
+      "epoch": 0.9541248082347701,
+      "grad_norm": 0.24598397229101523,
+      "learning_rate": 1.5458121402693297e-06,
+      "loss": 0.5598,
+      "step": 1205
+    },
+    {
+      "epoch": 0.954916613054882,
+      "grad_norm": 0.2599607802640301,
+      "learning_rate": 1.5451153747935427e-06,
+      "loss": 0.5818,
+      "step": 1206
+    },
+    {
+      "epoch": 0.9557084178749938,
+      "grad_norm": 0.27695634670719327,
+      "learning_rate": 1.5444182325841453e-06,
+      "loss": 0.5892,
+      "step": 1207
+    },
+    {
+      "epoch": 0.9565002226951057,
+      "grad_norm": 0.25211054985453246,
+      "learning_rate": 1.5437207141229378e-06,
+      "loss": 0.5919,
+      "step": 1208
+    },
+    {
+      "epoch": 0.9572920275152175,
+      "grad_norm": 0.2451060531835357,
+      "learning_rate": 1.543022819891981e-06,
+      "loss": 0.5928,
+      "step": 1209
+    },
+    {
+      "epoch": 0.9580838323353293,
+      "grad_norm": 0.24547033201409746,
+      "learning_rate": 1.5423245503735948e-06,
+      "loss": 0.5825,
+      "step": 1210
+    },
+    {
+      "epoch": 0.9588756371554412,
+      "grad_norm": 0.24596145085571985,
+      "learning_rate": 1.5416259060503594e-06,
+      "loss": 0.5941,
+      "step": 1211
+    },
+    {
+      "epoch": 0.959667441975553,
+      "grad_norm": 0.2597094087310838,
+      "learning_rate": 1.5409268874051134e-06,
+      "loss": 0.574,
+      "step": 1212
+    },
+    {
+      "epoch": 0.9604592467956649,
+      "grad_norm": 0.2514603318556578,
+      "learning_rate": 1.5402274949209538e-06,
+      "loss": 0.5627,
+      "step": 1213
+    },
+    {
+      "epoch": 0.9612510516157767,
+      "grad_norm": 0.2545090445589627,
+      "learning_rate": 1.5395277290812365e-06,
+      "loss": 0.5861,
+      "step": 1214
+    },
+    {
+      "epoch": 0.9620428564358886,
+      "grad_norm": 0.23975577653391356,
+      "learning_rate": 1.5388275903695758e-06,
+      "loss": 0.5882,
+      "step": 1215
+    },
+    {
+      "epoch": 0.9628346612560004,
+      "grad_norm": 0.24911852486492334,
+      "learning_rate": 1.5381270792698425e-06,
+      "loss": 0.5762,
+      "step": 1216
+    },
+    {
+      "epoch": 0.9636264660761122,
+      "grad_norm": 0.24117914928164164,
+      "learning_rate": 1.537426196266166e-06,
+      "loss": 0.5884,
+      "step": 1217
+    },
+    {
+      "epoch": 0.9644182708962241,
+      "grad_norm": 0.24333257333117847,
+      "learning_rate": 1.5367249418429316e-06,
+      "loss": 0.5795,
+      "step": 1218
+    },
+    {
+      "epoch": 0.9652100757163359,
+      "grad_norm": 0.25614483117781256,
+      "learning_rate": 1.536023316484782e-06,
+      "loss": 0.5468,
+      "step": 1219
+    },
+    {
+      "epoch": 0.9660018805364478,
+      "grad_norm": 0.24673039533023608,
+      "learning_rate": 1.5353213206766165e-06,
+      "loss": 0.5846,
+      "step": 1220
+    },
+    {
+      "epoch": 0.9667936853565596,
+      "grad_norm": 0.2583575821473352,
+      "learning_rate": 1.53461895490359e-06,
+      "loss": 0.5806,
+      "step": 1221
+    },
+    {
+      "epoch": 0.9675854901766715,
+      "grad_norm": 0.24876753094990142,
+      "learning_rate": 1.5339162196511126e-06,
+      "loss": 0.5815,
+      "step": 1222
+    },
+    {
+      "epoch": 0.9683772949967833,
+      "grad_norm": 0.2498128758337844,
+      "learning_rate": 1.5332131154048509e-06,
+      "loss": 0.5705,
+      "step": 1223
+    },
+    {
+      "epoch": 0.9691690998168951,
+      "grad_norm": 0.2459633172393119,
+      "learning_rate": 1.5325096426507252e-06,
+      "loss": 0.5677,
+      "step": 1224
+    },
+    {
+      "epoch": 0.969960904637007,
+      "grad_norm": 0.2509699664179796,
+      "learning_rate": 1.531805801874912e-06,
+      "loss": 0.5712,
+      "step": 1225
+    },
+    {
+      "epoch": 0.9707527094571188,
+      "grad_norm": 0.2586042819186474,
+      "learning_rate": 1.5311015935638408e-06,
+      "loss": 0.5677,
+      "step": 1226
+    },
+    {
+      "epoch": 0.9715445142772307,
+      "grad_norm": 0.24904022429580894,
+      "learning_rate": 1.5303970182041953e-06,
+      "loss": 0.5728,
+      "step": 1227
+    },
+    {
+      "epoch": 0.9723363190973425,
+      "grad_norm": 0.24197372346486745,
+      "learning_rate": 1.5296920762829139e-06,
+      "loss": 0.5804,
+      "step": 1228
+    },
+    {
+      "epoch": 0.9731281239174544,
+      "grad_norm": 0.24753782636573635,
+      "learning_rate": 1.5289867682871871e-06,
+      "loss": 0.5758,
+      "step": 1229
+    },
+    {
+      "epoch": 0.9739199287375662,
+      "grad_norm": 0.25113388858593033,
+      "learning_rate": 1.5282810947044594e-06,
+      "loss": 0.5649,
+      "step": 1230
+    },
+    {
+      "epoch": 0.974711733557678,
+      "grad_norm": 0.23987306987017853,
+      "learning_rate": 1.527575056022427e-06,
+      "loss": 0.5809,
+      "step": 1231
+    },
+    {
+      "epoch": 0.9755035383777899,
+      "grad_norm": 0.2525822493786336,
+      "learning_rate": 1.526868652729039e-06,
+      "loss": 0.5865,
+      "step": 1232
+    },
+    {
+      "epoch": 0.9762953431979017,
+      "grad_norm": 0.2486476903505399,
+      "learning_rate": 1.5261618853124967e-06,
+      "loss": 0.5996,
+      "step": 1233
+    },
+    {
+      "epoch": 0.9770871480180136,
+      "grad_norm": 0.25031091187174087,
+      "learning_rate": 1.525454754261252e-06,
+      "loss": 0.5946,
+      "step": 1234
+    },
+    {
+      "epoch": 0.9778789528381254,
+      "grad_norm": 0.24463685058744283,
+      "learning_rate": 1.52474726006401e-06,
+      "loss": 0.5817,
+      "step": 1235
+    },
+    {
+      "epoch": 0.9786707576582372,
+      "grad_norm": 0.24512456350133258,
+      "learning_rate": 1.5240394032097248e-06,
+      "loss": 0.5878,
+      "step": 1236
+    },
+    {
+      "epoch": 0.9794625624783491,
+      "grad_norm": 0.243488137783844,
+      "learning_rate": 1.5233311841876018e-06,
+      "loss": 0.5839,
+      "step": 1237
+    },
+    {
+      "epoch": 0.9802543672984609,
+      "grad_norm": 0.24976775565439793,
+      "learning_rate": 1.5226226034870973e-06,
+      "loss": 0.569,
+      "step": 1238
+    },
+    {
+      "epoch": 0.9810461721185728,
+      "grad_norm": 0.24828869737831172,
+      "learning_rate": 1.5219136615979168e-06,
+      "loss": 0.5788,
+      "step": 1239
+    },
+    {
+      "epoch": 0.9818379769386846,
+      "grad_norm": 0.248520412702361,
+      "learning_rate": 1.5212043590100163e-06,
+      "loss": 0.6001,
+      "step": 1240
+    },
+    {
+      "epoch": 0.9826297817587965,
+      "grad_norm": 0.25457793621661734,
+      "learning_rate": 1.5204946962136002e-06,
+      "loss": 0.5705,
+      "step": 1241
+    },
+    {
+      "epoch": 0.9834215865789083,
+      "grad_norm": 0.24422652650754226,
+      "learning_rate": 1.5197846736991218e-06,
+      "loss": 0.5827,
+      "step": 1242
+    },
+    {
+      "epoch": 0.9842133913990201,
+      "grad_norm": 0.2538600213012752,
+      "learning_rate": 1.5190742919572834e-06,
+      "loss": 0.5856,
+      "step": 1243
+    },
+    {
+      "epoch": 0.985005196219132,
+      "grad_norm": 0.2438152381132691,
+      "learning_rate": 1.5183635514790363e-06,
+      "loss": 0.5815,
+      "step": 1244
+    },
+    {
+      "epoch": 0.9857970010392438,
+      "grad_norm": 0.24446696920789407,
+      "learning_rate": 1.517652452755578e-06,
+      "loss": 0.5613,
+      "step": 1245
+    },
+    {
+      "epoch": 0.9865888058593557,
+      "grad_norm": 0.25778425733766247,
+      "learning_rate": 1.5169409962783552e-06,
+      "loss": 0.5796,
+      "step": 1246
+    },
+    {
+      "epoch": 0.9873806106794675,
+      "grad_norm": 0.25735735869613113,
+      "learning_rate": 1.5162291825390608e-06,
+      "loss": 0.5792,
+      "step": 1247
+    },
+    {
+      "epoch": 0.9881724154995793,
+      "grad_norm": 0.25348670485786373,
+      "learning_rate": 1.5155170120296347e-06,
+      "loss": 0.5825,
+      "step": 1248
+    },
+    {
+      "epoch": 0.9889642203196912,
+      "grad_norm": 0.24604259316927563,
+      "learning_rate": 1.5148044852422647e-06,
+      "loss": 0.5826,
+      "step": 1249
+    },
+    {
+      "epoch": 0.989756025139803,
+      "grad_norm": 0.23488252274168747,
+      "learning_rate": 1.5140916026693829e-06,
+      "loss": 0.5632,
+      "step": 1250
+    },
+    {
+      "epoch": 0.9905478299599149,
+      "grad_norm": 0.24395727151980548,
+      "learning_rate": 1.5133783648036685e-06,
+      "loss": 0.5911,
+      "step": 1251
+    },
+    {
+      "epoch": 0.9913396347800267,
+      "grad_norm": 0.2773273461281478,
+      "learning_rate": 1.5126647721380455e-06,
+      "loss": 0.5774,
+      "step": 1252
+    },
+    {
+      "epoch": 0.9921314396001386,
+      "grad_norm": 0.24715168847541846,
+      "learning_rate": 1.5119508251656842e-06,
+      "loss": 0.5644,
+      "step": 1253
+    },
+    {
+      "epoch": 0.9929232444202504,
+      "grad_norm": 0.2445962547720453,
+      "learning_rate": 1.5112365243799986e-06,
+      "loss": 0.5777,
+      "step": 1254
+    },
+    {
+      "epoch": 0.9937150492403622,
+      "grad_norm": 0.2428786228336218,
+      "learning_rate": 1.5105218702746478e-06,
+      "loss": 0.5826,
+      "step": 1255
+    },
+    {
+      "epoch": 0.9945068540604741,
+      "grad_norm": 0.2515089428556661,
+      "learning_rate": 1.5098068633435351e-06,
+      "loss": 0.568,
+      "step": 1256
+    },
+    {
+      "epoch": 0.9952986588805859,
+      "grad_norm": 0.25692707472958654,
+      "learning_rate": 1.5090915040808073e-06,
+      "loss": 0.5793,
+      "step": 1257
+    },
+    {
+      "epoch": 0.9960904637006978,
+      "grad_norm": 0.25172227371334727,
+      "learning_rate": 1.5083757929808551e-06,
+      "loss": 0.5879,
+      "step": 1258
+    },
+    {
+      "epoch": 0.9968822685208096,
+      "grad_norm": 0.24433804978221943,
+      "learning_rate": 1.5076597305383125e-06,
+      "loss": 0.5704,
+      "step": 1259
+    },
+    {
+      "epoch": 0.9976740733409215,
+      "grad_norm": 0.2538319779590854,
+      "learning_rate": 1.5069433172480555e-06,
+      "loss": 0.5904,
+      "step": 1260
+    },
+    {
+      "epoch": 0.9984658781610333,
+      "grad_norm": 0.2416778852725012,
+      "learning_rate": 1.5062265536052029e-06,
+      "loss": 0.5644,
+      "step": 1261
+    },
+    {
+      "epoch": 0.9992576829811451,
+      "grad_norm": 0.2460468272195175,
+      "learning_rate": 1.5055094401051163e-06,
+      "loss": 0.5752,
+      "step": 1262
+    },
+    {
+      "epoch": 1.000049487801257,
+      "grad_norm": 0.2443164680146125,
+      "learning_rate": 1.5047919772433985e-06,
+      "loss": 0.5692,
+      "step": 1263
+    },
+    {
+      "epoch": 1.0008412926213688,
+      "grad_norm": 0.2493012691970272,
+      "learning_rate": 1.5040741655158942e-06,
+      "loss": 0.5739,
+      "step": 1264
+    },
+    {
+      "epoch": 1.0016330974414807,
+      "grad_norm": 0.25776351156930466,
+      "learning_rate": 1.5033560054186878e-06,
+      "loss": 0.5812,
+      "step": 1265
+    },
+    {
+      "epoch": 1.0016330974414807,
+      "eval_loss": 0.6005386114120483,
+      "eval_runtime": 2126.3335,
+      "eval_samples_per_second": 12.825,
+      "eval_steps_per_second": 1.603,
+      "step": 1265
+    },
+    {
+      "epoch": 1.000346294647274,
+      "grad_norm": 0.24794191536714286,
+      "learning_rate": 1.5026374974481063e-06,
+      "loss": 0.5743,
+      "step": 1266
+    },
+    {
+      "epoch": 1.001137825269615,
+      "grad_norm": 0.2535145399680773,
+      "learning_rate": 1.5019186421007165e-06,
+      "loss": 0.579,
+      "step": 1267
+    },
+    {
+      "epoch": 1.001929355891956,
+      "grad_norm": 0.24982740232306555,
+      "learning_rate": 1.5011994398733245e-06,
+      "loss": 0.5713,
+      "step": 1268
+    },
+    {
+      "epoch": 1.002720886514297,
+      "grad_norm": 0.2395927596073369,
+      "learning_rate": 1.5004798912629767e-06,
+      "loss": 0.5617,
+      "step": 1269
+    },
+    {
+      "epoch": 1.003512417136638,
+      "grad_norm": 0.26152042185423313,
+      "learning_rate": 1.4997599967669597e-06,
+      "loss": 0.5728,
+      "step": 1270
+    },
+    {
+      "epoch": 1.004303947758979,
+      "grad_norm": 0.2710493799893223,
+      "learning_rate": 1.4990397568827974e-06,
+      "loss": 0.5673,
+      "step": 1271
+    },
+    {
+      "epoch": 1.00509547838132,
+      "grad_norm": 0.2597138637459402,
+      "learning_rate": 1.4983191721082544e-06,
+      "loss": 0.5675,
+      "step": 1272
+    },
+    {
+      "epoch": 1.005887009003661,
+      "grad_norm": 0.2546589756506511,
+      "learning_rate": 1.497598242941332e-06,
+      "loss": 0.5655,
+      "step": 1273
+    },
+    {
+      "epoch": 1.0066785396260018,
+      "grad_norm": 0.2510955685288097,
+      "learning_rate": 1.4968769698802697e-06,
+      "loss": 0.5755,
+      "step": 1274
+    },
+    {
+      "epoch": 1.0074700702483428,
+      "grad_norm": 0.24678269709753914,
+      "learning_rate": 1.4961553534235458e-06,
+      "loss": 0.5598,
+      "step": 1275
+    },
+    {
+      "epoch": 1.0082616008706837,
+      "grad_norm": 0.25559076819263843,
+      "learning_rate": 1.495433394069875e-06,
+      "loss": 0.5668,
+      "step": 1276
+    },
+    {
+      "epoch": 1.0090531314930247,
+      "grad_norm": 0.26026624192884135,
+      "learning_rate": 1.4947110923182096e-06,
+      "loss": 0.5542,
+      "step": 1277
+    },
+    {
+      "epoch": 1.0098446621153656,
+      "grad_norm": 0.24564726543394408,
+      "learning_rate": 1.4939884486677375e-06,
+      "loss": 0.5702,
+      "step": 1278
+    },
+    {
+      "epoch": 1.0106361927377066,
+      "grad_norm": 0.25567765139893456,
+      "learning_rate": 1.4932654636178836e-06,
+      "loss": 0.5673,
+      "step": 1279
+    },
+    {
+      "epoch": 1.0114277233600475,
+      "grad_norm": 0.25527953402860915,
+      "learning_rate": 1.4925421376683092e-06,
+      "loss": 0.5682,
+      "step": 1280
+    },
+    {
+      "epoch": 1.0122192539823884,
+      "grad_norm": 0.2498700299454256,
+      "learning_rate": 1.49181847131891e-06,
+      "loss": 0.5644,
+      "step": 1281
+    },
+    {
+      "epoch": 1.0130107846047294,
+      "grad_norm": 0.256882496818525,
+      "learning_rate": 1.4910944650698183e-06,
+      "loss": 0.5819,
+      "step": 1282
+    },
+    {
+      "epoch": 1.0138023152270703,
+      "grad_norm": 0.25510467285097205,
+      "learning_rate": 1.4903701194214002e-06,
+      "loss": 0.5472,
+      "step": 1283
+    },
+    {
+      "epoch": 1.0145938458494113,
+      "grad_norm": 0.26764921372549993,
+      "learning_rate": 1.4896454348742565e-06,
+      "loss": 0.5611,
+      "step": 1284
+    },
+    {
+      "epoch": 1.0153853764717522,
+      "grad_norm": 0.25076367991337567,
+      "learning_rate": 1.4889204119292234e-06,
+      "loss": 0.5452,
+      "step": 1285
+    },
+    {
+      "epoch": 1.0161769070940931,
+      "grad_norm": 0.24422920897401315,
+      "learning_rate": 1.488195051087369e-06,
+      "loss": 0.5489,
+      "step": 1286
+    },
+    {
+      "epoch": 1.016968437716434,
+      "grad_norm": 0.24688556667418038,
+      "learning_rate": 1.487469352849997e-06,
+      "loss": 0.5602,
+      "step": 1287
+    },
+    {
+      "epoch": 1.017759968338775,
+      "grad_norm": 0.2452056807184004,
+      "learning_rate": 1.486743317718643e-06,
+      "loss": 0.5603,
+      "step": 1288
+    },
+    {
+      "epoch": 1.018551498961116,
+      "grad_norm": 0.2591709342340121,
+      "learning_rate": 1.486016946195075e-06,
+      "loss": 0.5654,
+      "step": 1289
+    },
+    {
+      "epoch": 1.019343029583457,
+      "grad_norm": 0.25914934971045006,
+      "learning_rate": 1.4852902387812949e-06,
+      "loss": 0.5519,
+      "step": 1290
+    },
+    {
+      "epoch": 1.020134560205798,
+      "grad_norm": 0.25066275706999996,
+      "learning_rate": 1.484563195979536e-06,
+      "loss": 0.5712,
+      "step": 1291
+    },
+    {
+      "epoch": 1.020926090828139,
+      "grad_norm": 0.23897396839102916,
+      "learning_rate": 1.4838358182922632e-06,
+      "loss": 0.5553,
+      "step": 1292
+    },
+    {
+      "epoch": 1.02171762145048,
+      "grad_norm": 0.24223046692686187,
+      "learning_rate": 1.4831081062221734e-06,
+      "loss": 0.5585,
+      "step": 1293
+    },
+    {
+      "epoch": 1.022509152072821,
+      "grad_norm": 0.25774822941778874,
+      "learning_rate": 1.4823800602721935e-06,
+      "loss": 0.5618,
+      "step": 1294
+    },
+    {
+      "epoch": 1.0233006826951618,
+      "grad_norm": 0.24912409745324013,
+      "learning_rate": 1.4816516809454825e-06,
+      "loss": 0.5708,
+      "step": 1295
+    },
+    {
+      "epoch": 1.0240922133175028,
+      "grad_norm": 0.25126066328858004,
+      "learning_rate": 1.4809229687454292e-06,
+      "loss": 0.5619,
+      "step": 1296
+    },
+    {
+      "epoch": 1.0248837439398437,
+      "grad_norm": 0.25175714301525737,
+      "learning_rate": 1.4801939241756522e-06,
+      "loss": 0.5439,
+      "step": 1297
+    },
+    {
+      "epoch": 1.0256752745621847,
+      "grad_norm": 0.2534938847816727,
+      "learning_rate": 1.4794645477400002e-06,
+      "loss": 0.5757,
+      "step": 1298
+    },
+    {
+      "epoch": 1.0264668051845256,
+      "grad_norm": 0.2654739919991637,
+      "learning_rate": 1.4787348399425511e-06,
+      "loss": 0.55,
+      "step": 1299
+    },
+    {
+      "epoch": 1.0272583358068665,
+      "grad_norm": 0.24991454038211489,
+      "learning_rate": 1.4780048012876118e-06,
+      "loss": 0.5585,
+      "step": 1300
+    },
+    {
+      "epoch": 1.0280498664292075,
+      "grad_norm": 0.24786821821932004,
+      "learning_rate": 1.4772744322797185e-06,
+      "loss": 0.5632,
+      "step": 1301
+    },
+    {
+      "epoch": 1.0288413970515484,
+      "grad_norm": 0.2536266778374565,
+      "learning_rate": 1.4765437334236337e-06,
+      "loss": 0.5616,
+      "step": 1302
+    },
+    {
+      "epoch": 1.0296329276738894,
+      "grad_norm": 0.269318489855077,
+      "learning_rate": 1.4758127052243502e-06,
+      "loss": 0.5687,
+      "step": 1303
+    },
+    {
+      "epoch": 1.0304244582962303,
+      "grad_norm": 0.25582220993180177,
+      "learning_rate": 1.475081348187087e-06,
+      "loss": 0.5773,
+      "step": 1304
+    },
+    {
+      "epoch": 1.0312159889185712,
+      "grad_norm": 0.24828004800561954,
+      "learning_rate": 1.4743496628172915e-06,
+      "loss": 0.5585,
+      "step": 1305
+    },
+    {
+      "epoch": 1.0320075195409122,
+      "grad_norm": 0.24955982603745422,
+      "learning_rate": 1.4736176496206368e-06,
+      "loss": 0.5446,
+      "step": 1306
+    },
+    {
+      "epoch": 1.0327990501632531,
+      "grad_norm": 0.2467057872121939,
+      "learning_rate": 1.472885309103023e-06,
+      "loss": 0.5578,
+      "step": 1307
+    },
+    {
+      "epoch": 1.033590580785594,
+      "grad_norm": 0.24803505195333375,
+      "learning_rate": 1.472152641770577e-06,
+      "loss": 0.561,
+      "step": 1308
+    },
+    {
+      "epoch": 1.034382111407935,
+      "grad_norm": 0.244150143950206,
+      "learning_rate": 1.4714196481296508e-06,
+      "loss": 0.5523,
+      "step": 1309
+    },
+    {
+      "epoch": 1.035173642030276,
+      "grad_norm": 0.24392729840689864,
+      "learning_rate": 1.470686328686822e-06,
+      "loss": 0.5553,
+      "step": 1310
+    },
+    {
+      "epoch": 1.035965172652617,
+      "grad_norm": 0.2398461843564468,
+      "learning_rate": 1.4699526839488936e-06,
+      "loss": 0.5692,
+      "step": 1311
+    },
+    {
+      "epoch": 1.0367567032749578,
+      "grad_norm": 0.24964130220548783,
+      "learning_rate": 1.4692187144228939e-06,
+      "loss": 0.581,
+      "step": 1312
+    },
+    {
+      "epoch": 1.037548233897299,
+      "grad_norm": 0.25270711459088857,
+      "learning_rate": 1.4684844206160744e-06,
+      "loss": 0.5731,
+      "step": 1313
+    },
+    {
+      "epoch": 1.03833976451964,
+      "grad_norm": 0.25311734915864864,
+      "learning_rate": 1.467749803035912e-06,
+      "loss": 0.5559,
+      "step": 1314
+    },
+    {
+      "epoch": 1.0391312951419809,
+      "grad_norm": 0.24895949119922856,
+      "learning_rate": 1.4670148621901064e-06,
+      "loss": 0.5646,
+      "step": 1315
+    },
+    {
+      "epoch": 1.0399228257643218,
+      "grad_norm": 0.24692807710604875,
+      "learning_rate": 1.4662795985865813e-06,
+      "loss": 0.5633,
+      "step": 1316
+    },
+    {
+      "epoch": 1.0407143563866628,
+      "grad_norm": 0.24666659458126228,
+      "learning_rate": 1.4655440127334835e-06,
+      "loss": 0.5561,
+      "step": 1317
+    },
+    {
+      "epoch": 1.0415058870090037,
+      "grad_norm": 0.24781497169803066,
+      "learning_rate": 1.4648081051391817e-06,
+      "loss": 0.5509,
+      "step": 1318
+    },
+    {
+      "epoch": 1.0422974176313446,
+      "grad_norm": 0.24987595540755106,
+      "learning_rate": 1.4640718763122682e-06,
+      "loss": 0.56,
+      "step": 1319
+    },
+    {
+      "epoch": 1.0430889482536856,
+      "grad_norm": 0.24412348946717424,
+      "learning_rate": 1.4633353267615562e-06,
+      "loss": 0.5465,
+      "step": 1320
+    },
+    {
+      "epoch": 1.0438804788760265,
+      "grad_norm": 0.25007361354164936,
+      "learning_rate": 1.4625984569960812e-06,
+      "loss": 0.5588,
+      "step": 1321
+    },
+    {
+      "epoch": 1.0446720094983675,
+      "grad_norm": 0.2603620223234983,
+      "learning_rate": 1.4618612675250996e-06,
+      "loss": 0.5546,
+      "step": 1322
+    },
+    {
+      "epoch": 1.0454635401207084,
+      "grad_norm": 0.24236235522821056,
+      "learning_rate": 1.4611237588580893e-06,
+      "loss": 0.5666,
+      "step": 1323
+    },
+    {
+      "epoch": 1.0462550707430494,
+      "grad_norm": 0.24539893131607105,
+      "learning_rate": 1.4603859315047483e-06,
+      "loss": 0.5574,
+      "step": 1324
+    },
+    {
+      "epoch": 1.0470466013653903,
+      "grad_norm": 0.25026973097816974,
+      "learning_rate": 1.4596477859749953e-06,
+      "loss": 0.5619,
+      "step": 1325
+    },
+    {
+      "epoch": 1.0478381319877312,
+      "grad_norm": 0.2512554046209749,
+      "learning_rate": 1.4589093227789678e-06,
+      "loss": 0.5648,
+      "step": 1326
+    },
+    {
+      "epoch": 1.0486296626100722,
+      "grad_norm": 0.2502456623416568,
+      "learning_rate": 1.4581705424270244e-06,
+      "loss": 0.5556,
+      "step": 1327
+    },
+    {
+      "epoch": 1.0494211932324131,
+      "grad_norm": 0.25955130771015766,
+      "learning_rate": 1.4574314454297416e-06,
+      "loss": 0.5632,
+      "step": 1328
+    },
+    {
+      "epoch": 1.050212723854754,
+      "grad_norm": 0.24964711378746987,
+      "learning_rate": 1.4566920322979157e-06,
+      "loss": 0.5525,
+      "step": 1329
+    },
+    {
+      "epoch": 1.051004254477095,
+      "grad_norm": 0.2465198285458817,
+      "learning_rate": 1.4559523035425612e-06,
+      "loss": 0.5706,
+      "step": 1330
+    },
+    {
+      "epoch": 1.051795785099436,
+      "grad_norm": 0.25385504922573,
+      "learning_rate": 1.4552122596749097e-06,
+      "loss": 0.5634,
+      "step": 1331
+    },
+    {
+      "epoch": 1.0525873157217769,
+      "grad_norm": 0.2458409041004127,
+      "learning_rate": 1.4544719012064122e-06,
+      "loss": 0.5642,
+      "step": 1332
+    },
+    {
+      "epoch": 1.0533788463441178,
+      "grad_norm": 0.2560080115195914,
+      "learning_rate": 1.4537312286487361e-06,
+      "loss": 0.5701,
+      "step": 1333
+    },
+    {
+      "epoch": 1.054170376966459,
+      "grad_norm": 0.2478834879649551,
+      "learning_rate": 1.4529902425137666e-06,
+      "loss": 0.5704,
+      "step": 1334
+    },
+    {
+      "epoch": 1.0549619075888,
+      "grad_norm": 0.2517473335379542,
+      "learning_rate": 1.4522489433136049e-06,
+      "loss": 0.5647,
+      "step": 1335
+    },
+    {
+      "epoch": 1.0557534382111409,
+      "grad_norm": 0.26244292716307693,
+      "learning_rate": 1.4515073315605681e-06,
+      "loss": 0.5702,
+      "step": 1336
+    },
+    {
+      "epoch": 1.0565449688334818,
+      "grad_norm": 0.24955021623416915,
+      "learning_rate": 1.4507654077671911e-06,
+      "loss": 0.576,
+      "step": 1337
+    },
+    {
+      "epoch": 1.0573364994558228,
+      "grad_norm": 0.2434535878102341,
+      "learning_rate": 1.4500231724462232e-06,
+      "loss": 0.5675,
+      "step": 1338
+    },
+    {
+      "epoch": 1.0581280300781637,
+      "grad_norm": 0.24865477438057995,
+      "learning_rate": 1.4492806261106285e-06,
+      "loss": 0.5631,
+      "step": 1339
+    },
+    {
+      "epoch": 1.0589195607005046,
+      "grad_norm": 0.2516834885516957,
+      "learning_rate": 1.4485377692735875e-06,
+      "loss": 0.5415,
+      "step": 1340
+    },
+    {
+      "epoch": 1.0597110913228456,
+      "grad_norm": 0.24427606981869535,
+      "learning_rate": 1.447794602448494e-06,
+      "loss": 0.5651,
+      "step": 1341
+    },
+    {
+      "epoch": 1.0605026219451865,
+      "grad_norm": 0.24943473935695282,
+      "learning_rate": 1.4470511261489567e-06,
+      "loss": 0.5621,
+      "step": 1342
+    },
+    {
+      "epoch": 1.0612941525675275,
+      "grad_norm": 0.2507908265549793,
+      "learning_rate": 1.4463073408887982e-06,
+      "loss": 0.5531,
+      "step": 1343
+    },
+    {
+      "epoch": 1.0620856831898684,
+      "grad_norm": 0.23597709877327022,
+      "learning_rate": 1.4455632471820544e-06,
+      "loss": 0.5548,
+      "step": 1344
+    },
+    {
+      "epoch": 1.0628772138122093,
+      "grad_norm": 0.24294964099822539,
+      "learning_rate": 1.4448188455429744e-06,
+      "loss": 0.5611,
+      "step": 1345
+    },
+    {
+      "epoch": 1.0636687444345503,
+      "grad_norm": 0.24044756054247945,
+      "learning_rate": 1.4440741364860204e-06,
+      "loss": 0.5705,
+      "step": 1346
+    },
+    {
+      "epoch": 1.0644602750568912,
+      "grad_norm": 0.2488697572231574,
+      "learning_rate": 1.443329120525866e-06,
+      "loss": 0.5679,
+      "step": 1347
+    },
+    {
+      "epoch": 1.0652518056792322,
+      "grad_norm": 0.2475552616581592,
+      "learning_rate": 1.4425837981773988e-06,
+      "loss": 0.5673,
+      "step": 1348
+    },
+    {
+      "epoch": 1.066043336301573,
+      "grad_norm": 0.24975796330795333,
+      "learning_rate": 1.4418381699557161e-06,
+      "loss": 0.5775,
+      "step": 1349
+    },
+    {
+      "epoch": 1.066834866923914,
+      "grad_norm": 0.24276956911990535,
+      "learning_rate": 1.441092236376128e-06,
+      "loss": 0.5477,
+      "step": 1350
+    },
+    {
+      "epoch": 1.067626397546255,
+      "grad_norm": 0.2554956875493257,
+      "learning_rate": 1.4403459979541552e-06,
+      "loss": 0.5527,
+      "step": 1351
+    },
+    {
+      "epoch": 1.068417928168596,
+      "grad_norm": 0.25159070985489457,
+      "learning_rate": 1.4395994552055291e-06,
+      "loss": 0.5715,
+      "step": 1352
+    },
+    {
+      "epoch": 1.0692094587909369,
+      "grad_norm": 0.2494368117656817,
+      "learning_rate": 1.4388526086461912e-06,
+      "loss": 0.5755,
+      "step": 1353
+    },
+    {
+      "epoch": 1.0700009894132778,
+      "grad_norm": 0.26034209301072064,
+      "learning_rate": 1.4381054587922934e-06,
+      "loss": 0.5603,
+      "step": 1354
+    },
+    {
+      "epoch": 1.070792520035619,
+      "grad_norm": 0.26023920582587345,
+      "learning_rate": 1.4373580061601967e-06,
+      "loss": 0.5651,
+      "step": 1355
+    },
+    {
+      "epoch": 1.07158405065796,
+      "grad_norm": 0.24420554855163062,
+      "learning_rate": 1.4366102512664717e-06,
+      "loss": 0.5681,
+      "step": 1356
+    },
+    {
+      "epoch": 1.0723755812803009,
+      "grad_norm": 0.2415399016964875,
+      "learning_rate": 1.4358621946278979e-06,
+      "loss": 0.5558,
+      "step": 1357
+    },
+    {
+      "epoch": 1.0731671119026418,
+      "grad_norm": 0.25368243576055927,
+      "learning_rate": 1.435113836761463e-06,
+      "loss": 0.5599,
+      "step": 1358
+    },
+    {
+      "epoch": 1.0739586425249827,
+      "grad_norm": 0.2544067704434178,
+      "learning_rate": 1.4343651781843638e-06,
+      "loss": 0.5682,
+      "step": 1359
+    },
+    {
+      "epoch": 1.0747501731473237,
+      "grad_norm": 0.25308737067053566,
+      "learning_rate": 1.4336162194140032e-06,
+      "loss": 0.5765,
+      "step": 1360
+    },
+    {
+      "epoch": 1.0755417037696646,
+      "grad_norm": 0.2506401510851039,
+      "learning_rate": 1.4328669609679936e-06,
+      "loss": 0.559,
+      "step": 1361
+    },
+    {
+      "epoch": 1.0763332343920056,
+      "grad_norm": 0.25875349828322386,
+      "learning_rate": 1.4321174033641527e-06,
+      "loss": 0.5655,
+      "step": 1362
+    },
+    {
+      "epoch": 1.0771247650143465,
+      "grad_norm": 0.25059238334641404,
+      "learning_rate": 1.4313675471205064e-06,
+      "loss": 0.5656,
+      "step": 1363
+    },
+    {
+      "epoch": 1.0779162956366874,
+      "grad_norm": 0.25768255084006086,
+      "learning_rate": 1.4306173927552863e-06,
+      "loss": 0.5601,
+      "step": 1364
+    },
+    {
+      "epoch": 1.0787078262590284,
+      "grad_norm": 0.23875002879243465,
+      "learning_rate": 1.4298669407869294e-06,
+      "loss": 0.5557,
+      "step": 1365
+    },
+    {
+      "epoch": 1.0794993568813693,
+      "grad_norm": 0.2445795213472089,
+      "learning_rate": 1.4291161917340802e-06,
+      "loss": 0.5722,
+      "step": 1366
+    },
+    {
+      "epoch": 1.0802908875037103,
+      "grad_norm": 0.2510424953327719,
+      "learning_rate": 1.4283651461155865e-06,
+      "loss": 0.5607,
+      "step": 1367
+    },
+    {
+      "epoch": 1.0810824181260512,
+      "grad_norm": 0.25830090042117565,
+      "learning_rate": 1.427613804450502e-06,
+      "loss": 0.573,
+      "step": 1368
+    },
+    {
+      "epoch": 1.0818739487483922,
+      "grad_norm": 0.24261289539474715,
+      "learning_rate": 1.426862167258086e-06,
+      "loss": 0.5583,
+      "step": 1369
+    },
+    {
+      "epoch": 1.082665479370733,
+      "grad_norm": 0.24311075923898381,
+      "learning_rate": 1.4261102350577996e-06,
+      "loss": 0.5458,
+      "step": 1370
+    },
+    {
+      "epoch": 1.083457009993074,
+      "grad_norm": 0.25599732005442116,
+      "learning_rate": 1.4253580083693103e-06,
+      "loss": 0.5567,
+      "step": 1371
+    },
+    {
+      "epoch": 1.084248540615415,
+      "grad_norm": 0.26359055736961196,
+      "learning_rate": 1.424605487712487e-06,
+      "loss": 0.5521,
+      "step": 1372
+    },
+    {
+      "epoch": 1.085040071237756,
+      "grad_norm": 0.24429222575076712,
+      "learning_rate": 1.4238526736074037e-06,
+      "loss": 0.5547,
+      "step": 1373
+    },
+    {
+      "epoch": 1.0858316018600969,
+      "grad_norm": 0.2592387090737342,
+      "learning_rate": 1.4230995665743354e-06,
+      "loss": 0.5639,
+      "step": 1374
+    },
+    {
+      "epoch": 1.0866231324824378,
+      "grad_norm": 0.2576967328448918,
+      "learning_rate": 1.4223461671337611e-06,
+      "loss": 0.5633,
+      "step": 1375
+    },
+    {
+      "epoch": 1.087414663104779,
+      "grad_norm": 0.2539279510513475,
+      "learning_rate": 1.4215924758063612e-06,
+      "loss": 0.5693,
+      "step": 1376
+    },
+    {
+      "epoch": 1.08820619372712,
+      "grad_norm": 0.250940301496672,
+      "learning_rate": 1.4208384931130173e-06,
+      "loss": 0.5561,
+      "step": 1377
+    },
+    {
+      "epoch": 1.0889977243494608,
+      "grad_norm": 0.24976103373673933,
+      "learning_rate": 1.420084219574813e-06,
+      "loss": 0.5705,
+      "step": 1378
+    },
+    {
+      "epoch": 1.0897892549718018,
+      "grad_norm": 0.24263277731436725,
+      "learning_rate": 1.4193296557130332e-06,
+      "loss": 0.566,
+      "step": 1379
+    },
+    {
+      "epoch": 1.0905807855941427,
+      "grad_norm": 0.2633927958881685,
+      "learning_rate": 1.4185748020491628e-06,
+      "loss": 0.5788,
+      "step": 1380
+    },
+    {
+      "epoch": 1.0913723162164837,
+      "grad_norm": 0.24815938962315232,
+      "learning_rate": 1.4178196591048873e-06,
+      "loss": 0.5557,
+      "step": 1381
+    },
+    {
+      "epoch": 1.0921638468388246,
+      "grad_norm": 0.25115346718353765,
+      "learning_rate": 1.4170642274020921e-06,
+      "loss": 0.5644,
+      "step": 1382
+    },
+    {
+      "epoch": 1.0929553774611656,
+      "grad_norm": 0.25936916512413793,
+      "learning_rate": 1.4163085074628615e-06,
+      "loss": 0.5705,
+      "step": 1383
+    },
+    {
+      "epoch": 1.0937469080835065,
+      "grad_norm": 0.2540161130865951,
+      "learning_rate": 1.4155524998094804e-06,
+      "loss": 0.5703,
+      "step": 1384
+    },
+    {
+      "epoch": 1.0945384387058474,
+      "grad_norm": 0.2473168264796292,
+      "learning_rate": 1.4147962049644315e-06,
+      "loss": 0.5663,
+      "step": 1385
+    },
+    {
+      "epoch": 1.0953299693281884,
+      "grad_norm": 0.2478505866906445,
+      "learning_rate": 1.4140396234503964e-06,
+      "loss": 0.5641,
+      "step": 1386
+    },
+    {
+      "epoch": 1.0961214999505293,
+      "grad_norm": 0.24769702910603478,
+      "learning_rate": 1.4132827557902541e-06,
+      "loss": 0.5505,
+      "step": 1387
+    },
+    {
+      "epoch": 1.0969130305728703,
+      "grad_norm": 0.26318448683729706,
+      "learning_rate": 1.4125256025070824e-06,
+      "loss": 0.5709,
+      "step": 1388
+    },
+    {
+      "epoch": 1.0977045611952112,
+      "grad_norm": 0.24246624551299162,
+      "learning_rate": 1.4117681641241559e-06,
+      "loss": 0.5518,
+      "step": 1389
+    },
+    {
+      "epoch": 1.0984960918175521,
+      "grad_norm": 0.26244592210781836,
+      "learning_rate": 1.4110104411649463e-06,
+      "loss": 0.5692,
+      "step": 1390
+    },
+    {
+      "epoch": 1.099287622439893,
+      "grad_norm": 0.25937353050620887,
+      "learning_rate": 1.4102524341531223e-06,
+      "loss": 0.5543,
+      "step": 1391
+    },
+    {
+      "epoch": 1.100079153062234,
+      "grad_norm": 0.2606143038056361,
+      "learning_rate": 1.4094941436125482e-06,
+      "loss": 0.5683,
+      "step": 1392
+    },
+    {
+      "epoch": 1.100870683684575,
+      "grad_norm": 0.2589377474035992,
+      "learning_rate": 1.408735570067285e-06,
+      "loss": 0.5755,
+      "step": 1393
+    },
+    {
+      "epoch": 1.101662214306916,
+      "grad_norm": 0.2560676432428256,
+      "learning_rate": 1.4079767140415886e-06,
+      "loss": 0.5607,
+      "step": 1394
+    },
+    {
+      "epoch": 1.1024537449292569,
+      "grad_norm": 0.2544295215042865,
+      "learning_rate": 1.4072175760599107e-06,
+      "loss": 0.5658,
+      "step": 1395
+    },
+    {
+      "epoch": 1.1032452755515978,
+      "grad_norm": 0.25692156591295223,
+      "learning_rate": 1.406458156646898e-06,
+      "loss": 0.5491,
+      "step": 1396
+    },
+    {
+      "epoch": 1.104036806173939,
+      "grad_norm": 0.2615088469762251,
+      "learning_rate": 1.4056984563273906e-06,
+      "loss": 0.5663,
+      "step": 1397
+    },
+    {
+      "epoch": 1.10482833679628,
+      "grad_norm": 0.40982086382454475,
+      "learning_rate": 1.4049384756264243e-06,
+      "loss": 0.5676,
+      "step": 1398
+    },
+    {
+      "epoch": 1.1056198674186208,
+      "grad_norm": 0.2557534167110459,
+      "learning_rate": 1.4041782150692273e-06,
+      "loss": 0.5677,
+      "step": 1399
+    },
+    {
+      "epoch": 1.1064113980409618,
+      "grad_norm": 0.24867310110850344,
+      "learning_rate": 1.4034176751812219e-06,
+      "loss": 0.5698,
+      "step": 1400
+    },
+    {
+      "epoch": 1.1072029286633027,
+      "grad_norm": 0.24449685446842004,
+      "learning_rate": 1.4026568564880235e-06,
+      "loss": 0.5603,
+      "step": 1401
+    },
+    {
+      "epoch": 1.1079944592856437,
+      "grad_norm": 0.2629831409137662,
+      "learning_rate": 1.4018957595154395e-06,
+      "loss": 0.551,
+      "step": 1402
+    },
+    {
+      "epoch": 1.1087859899079846,
+      "grad_norm": 0.26123133768605006,
+      "learning_rate": 1.401134384789471e-06,
+      "loss": 0.5852,
+      "step": 1403
+    },
+    {
+      "epoch": 1.1095775205303255,
+      "grad_norm": 0.2440411786605679,
+      "learning_rate": 1.4003727328363092e-06,
+      "loss": 0.5623,
+      "step": 1404
+    },
+    {
+      "epoch": 1.1103690511526665,
+      "grad_norm": 0.23987090708226047,
+      "learning_rate": 1.3996108041823386e-06,
+      "loss": 0.5579,
+      "step": 1405
+    },
+    {
+      "epoch": 1.1111605817750074,
+      "grad_norm": 0.24206684985462035,
+      "learning_rate": 1.3988485993541342e-06,
+      "loss": 0.557,
+      "step": 1406
+    },
+    {
+      "epoch": 1.1119521123973484,
+      "grad_norm": 0.25278949872560796,
+      "learning_rate": 1.3980861188784613e-06,
+      "loss": 0.5633,
+      "step": 1407
+    },
+    {
+      "epoch": 1.1127436430196893,
+      "grad_norm": 0.2542737514370265,
+      "learning_rate": 1.3973233632822766e-06,
+      "loss": 0.5628,
+      "step": 1408
+    },
+    {
+      "epoch": 1.1135351736420303,
+      "grad_norm": 0.24384929625513102,
+      "learning_rate": 1.3965603330927267e-06,
+      "loss": 0.5773,
+      "step": 1409
+    },
+    {
+      "epoch": 1.1143267042643712,
+      "grad_norm": 0.24890884561690166,
+      "learning_rate": 1.3957970288371481e-06,
+      "loss": 0.5596,
+      "step": 1410
+    },
+    {
+      "epoch": 1.1151182348867121,
+      "grad_norm": 0.24306152813754728,
+      "learning_rate": 1.3950334510430662e-06,
+      "loss": 0.5689,
+      "step": 1411
+    },
+    {
+      "epoch": 1.115909765509053,
+      "grad_norm": 0.24893814741240725,
+      "learning_rate": 1.3942696002381954e-06,
+      "loss": 0.5643,
+      "step": 1412
+    },
+    {
+      "epoch": 1.116701296131394,
+      "grad_norm": 0.25140564614699834,
+      "learning_rate": 1.3935054769504392e-06,
+      "loss": 0.5675,
+      "step": 1413
+    },
+    {
+      "epoch": 1.117492826753735,
+      "grad_norm": 0.24362217835135425,
+      "learning_rate": 1.3927410817078907e-06,
+      "loss": 0.5595,
+      "step": 1414
+    },
+    {
+      "epoch": 1.118284357376076,
+      "grad_norm": 0.24781969989964167,
+      "learning_rate": 1.3919764150388276e-06,
+      "loss": 0.5696,
+      "step": 1415
+    },
+    {
+      "epoch": 1.1190758879984168,
+      "grad_norm": 0.24760728428101522,
+      "learning_rate": 1.3912114774717183e-06,
+      "loss": 0.5499,
+      "step": 1416
+    },
+    {
+      "epoch": 1.1198674186207578,
+      "grad_norm": 0.23910844633467185,
+      "learning_rate": 1.3904462695352171e-06,
+      "loss": 0.5443,
+      "step": 1417
+    },
+    {
+      "epoch": 1.120658949243099,
+      "grad_norm": 0.24489015265636474,
+      "learning_rate": 1.389680791758165e-06,
+      "loss": 0.5572,
+      "step": 1418
+    },
+    {
+      "epoch": 1.1214504798654399,
+      "grad_norm": 0.2460657292272078,
+      "learning_rate": 1.3889150446695907e-06,
+      "loss": 0.576,
+      "step": 1419
+    },
+    {
+      "epoch": 1.1222420104877808,
+      "grad_norm": 0.25033130711864643,
+      "learning_rate": 1.3881490287987075e-06,
+      "loss": 0.5544,
+      "step": 1420
+    },
+    {
+      "epoch": 1.1230335411101218,
+      "grad_norm": 0.24827949995915674,
+      "learning_rate": 1.387382744674915e-06,
+      "loss": 0.572,
+      "step": 1421
+    },
+    {
+      "epoch": 1.1238250717324627,
+      "grad_norm": 0.2378346073817711,
+      "learning_rate": 1.3866161928277986e-06,
+      "loss": 0.5682,
+      "step": 1422
+    },
+    {
+      "epoch": 1.1246166023548037,
+      "grad_norm": 0.2571384344357915,
+      "learning_rate": 1.3858493737871284e-06,
+      "loss": 0.5617,
+      "step": 1423
+    },
+    {
+      "epoch": 1.1254081329771446,
+      "grad_norm": 0.2384405621085864,
+      "learning_rate": 1.3850822880828593e-06,
+      "loss": 0.5564,
+      "step": 1424
+    },
+    {
+      "epoch": 1.1261996635994855,
+      "grad_norm": 0.24597836673469964,
+      "learning_rate": 1.38431493624513e-06,
+      "loss": 0.5633,
+      "step": 1425
+    },
+    {
+      "epoch": 1.1269911942218265,
+      "grad_norm": 0.24829013323231672,
+      "learning_rate": 1.3835473188042636e-06,
+      "loss": 0.5738,
+      "step": 1426
+    },
+    {
+      "epoch": 1.1277827248441674,
+      "grad_norm": 0.24780693733556894,
+      "learning_rate": 1.3827794362907672e-06,
+      "loss": 0.5496,
+      "step": 1427
+    },
+    {
+      "epoch": 1.1285742554665084,
+      "grad_norm": 0.2496677391898154,
+      "learning_rate": 1.3820112892353301e-06,
+      "loss": 0.5473,
+      "step": 1428
+    },
+    {
+      "epoch": 1.1293657860888493,
+      "grad_norm": 0.24347934169416524,
+      "learning_rate": 1.381242878168825e-06,
+      "loss": 0.5591,
+      "step": 1429
+    },
+    {
+      "epoch": 1.1301573167111902,
+      "grad_norm": 0.2949242019925856,
+      "learning_rate": 1.3804742036223071e-06,
+      "loss": 0.563,
+      "step": 1430
+    },
+    {
+      "epoch": 1.1309488473335312,
+      "grad_norm": 0.25587647799333957,
+      "learning_rate": 1.3797052661270132e-06,
+      "loss": 0.5595,
+      "step": 1431
+    },
+    {
+      "epoch": 1.1317403779558721,
+      "grad_norm": 0.2492426640162163,
+      "learning_rate": 1.3789360662143622e-06,
+      "loss": 0.5708,
+      "step": 1432
+    },
+    {
+      "epoch": 1.132531908578213,
+      "grad_norm": 0.24691132782837508,
+      "learning_rate": 1.378166604415955e-06,
+      "loss": 0.5669,
+      "step": 1433
+    },
+    {
+      "epoch": 1.133323439200554,
+      "grad_norm": 0.25270399928726467,
+      "learning_rate": 1.377396881263572e-06,
+      "loss": 0.5624,
+      "step": 1434
+    },
+    {
+      "epoch": 1.134114969822895,
+      "grad_norm": 0.24548804339917815,
+      "learning_rate": 1.3766268972891757e-06,
+      "loss": 0.5639,
+      "step": 1435
+    },
+    {
+      "epoch": 1.1349065004452359,
+      "grad_norm": 0.23617628723323206,
+      "learning_rate": 1.3758566530249074e-06,
+      "loss": 0.5604,
+      "step": 1436
+    },
+    {
+      "epoch": 1.1356980310675768,
+      "grad_norm": 0.25460525737316764,
+      "learning_rate": 1.3750861490030898e-06,
+      "loss": 0.5637,
+      "step": 1437
+    },
+    {
+      "epoch": 1.1364895616899178,
+      "grad_norm": 0.25667284385399564,
+      "learning_rate": 1.3743153857562244e-06,
+      "loss": 0.5674,
+      "step": 1438
+    },
+    {
+      "epoch": 1.137281092312259,
+      "grad_norm": 0.2435879950474742,
+      "learning_rate": 1.3735443638169914e-06,
+      "loss": 0.5698,
+      "step": 1439
+    },
+    {
+      "epoch": 1.1380726229345999,
+      "grad_norm": 0.2469201860892531,
+      "learning_rate": 1.3727730837182508e-06,
+      "loss": 0.562,
+      "step": 1440
+    },
+    {
+      "epoch": 1.1388641535569408,
+      "grad_norm": 0.2504993856121079,
+      "learning_rate": 1.3720015459930396e-06,
+      "loss": 0.575,
+      "step": 1441
+    },
+    {
+      "epoch": 1.1396556841792818,
+      "grad_norm": 0.23936229592820304,
+      "learning_rate": 1.3712297511745749e-06,
+      "loss": 0.5773,
+      "step": 1442
+    },
+    {
+      "epoch": 1.1404472148016227,
+      "grad_norm": 0.2519788767852563,
+      "learning_rate": 1.3704576997962495e-06,
+      "loss": 0.5725,
+      "step": 1443
+    },
+    {
+      "epoch": 1.1412387454239636,
+      "grad_norm": 0.2380806327883765,
+      "learning_rate": 1.3696853923916343e-06,
+      "loss": 0.558,
+      "step": 1444
+    },
+    {
+      "epoch": 1.1420302760463046,
+      "grad_norm": 0.25879478539242484,
+      "learning_rate": 1.3689128294944777e-06,
+      "loss": 0.5612,
+      "step": 1445
+    },
+    {
+      "epoch": 1.1428218066686455,
+      "grad_norm": 0.2525430153085939,
+      "learning_rate": 1.3681400116387033e-06,
+      "loss": 0.5626,
+      "step": 1446
+    },
+    {
+      "epoch": 1.1436133372909865,
+      "grad_norm": 0.2450210620431473,
+      "learning_rate": 1.3673669393584123e-06,
+      "loss": 0.5649,
+      "step": 1447
+    },
+    {
+      "epoch": 1.1444048679133274,
+      "grad_norm": 0.24000612889609255,
+      "learning_rate": 1.3665936131878814e-06,
+      "loss": 0.5488,
+      "step": 1448
+    },
+    {
+      "epoch": 1.1451963985356683,
+      "grad_norm": 0.25832369735084254,
+      "learning_rate": 1.3658200336615618e-06,
+      "loss": 0.5698,
+      "step": 1449
+    },
+    {
+      "epoch": 1.1459879291580093,
+      "grad_norm": 0.24163981156196715,
+      "learning_rate": 1.365046201314081e-06,
+      "loss": 0.5596,
+      "step": 1450
+    },
+    {
+      "epoch": 1.1467794597803502,
+      "grad_norm": 0.24601436851831515,
+      "learning_rate": 1.3642721166802406e-06,
+      "loss": 0.5575,
+      "step": 1451
+    },
+    {
+      "epoch": 1.1475709904026912,
+      "grad_norm": 0.25582215178071455,
+      "learning_rate": 1.3634977802950166e-06,
+      "loss": 0.5672,
+      "step": 1452
+    },
+    {
+      "epoch": 1.1483625210250321,
+      "grad_norm": 0.25421730792632513,
+      "learning_rate": 1.3627231926935596e-06,
+      "loss": 0.5867,
+      "step": 1453
+    },
+    {
+      "epoch": 1.149154051647373,
+      "grad_norm": 0.24605176555963193,
+      "learning_rate": 1.3619483544111925e-06,
+      "loss": 0.5779,
+      "step": 1454
+    },
+    {
+      "epoch": 1.149945582269714,
+      "grad_norm": 0.2339753161272484,
+      "learning_rate": 1.3611732659834128e-06,
+      "loss": 0.5494,
+      "step": 1455
+    },
+    {
+      "epoch": 1.150737112892055,
+      "grad_norm": 0.24505913123384715,
+      "learning_rate": 1.3603979279458903e-06,
+      "loss": 0.5484,
+      "step": 1456
+    },
+    {
+      "epoch": 1.1515286435143959,
+      "grad_norm": 0.24447597123717954,
+      "learning_rate": 1.3596223408344674e-06,
+      "loss": 0.5601,
+      "step": 1457
+    },
+    {
+      "epoch": 1.1523201741367368,
+      "grad_norm": 0.25432198557858027,
+      "learning_rate": 1.3588465051851583e-06,
+      "loss": 0.5635,
+      "step": 1458
+    },
+    {
+      "epoch": 1.1531117047590778,
+      "grad_norm": 0.23893493265840152,
+      "learning_rate": 1.3580704215341493e-06,
+      "loss": 0.5589,
+      "step": 1459
+    },
+    {
+      "epoch": 1.153903235381419,
+      "grad_norm": 0.24689412322769985,
+      "learning_rate": 1.3572940904177978e-06,
+      "loss": 0.5642,
+      "step": 1460
+    },
+    {
+      "epoch": 1.1546947660037599,
+      "grad_norm": 0.24518858861845222,
+      "learning_rate": 1.3565175123726334e-06,
+      "loss": 0.5741,
+      "step": 1461
+    },
+    {
+      "epoch": 1.1554862966261008,
+      "grad_norm": 0.27557275432437467,
+      "learning_rate": 1.3557406879353544e-06,
+      "loss": 0.5558,
+      "step": 1462
+    },
+    {
+      "epoch": 1.1562778272484417,
+      "grad_norm": 0.24806836507237567,
+      "learning_rate": 1.3549636176428309e-06,
+      "loss": 0.5459,
+      "step": 1463
+    },
+    {
+      "epoch": 1.1570693578707827,
+      "grad_norm": 0.24573547223625097,
+      "learning_rate": 1.3541863020321025e-06,
+      "loss": 0.5619,
+      "step": 1464
+    },
+    {
+      "epoch": 1.1578608884931236,
+      "grad_norm": 0.23943167718446362,
+      "learning_rate": 1.3534087416403777e-06,
+      "loss": 0.5597,
+      "step": 1465
+    },
+    {
+      "epoch": 1.1586524191154646,
+      "grad_norm": 0.24002190583674485,
+      "learning_rate": 1.3526309370050357e-06,
+      "loss": 0.5616,
+      "step": 1466
+    },
+    {
+      "epoch": 1.1594439497378055,
+      "grad_norm": 0.2516513879614623,
+      "learning_rate": 1.3518528886636223e-06,
+      "loss": 0.5689,
+      "step": 1467
+    },
+    {
+      "epoch": 1.1602354803601465,
+      "grad_norm": 0.24065964510116664,
+      "learning_rate": 1.3510745971538537e-06,
+      "loss": 0.5502,
+      "step": 1468
+    },
+    {
+      "epoch": 1.1610270109824874,
+      "grad_norm": 0.25407611161783095,
+      "learning_rate": 1.3502960630136135e-06,
+      "loss": 0.5848,
+      "step": 1469
+    },
+    {
+      "epoch": 1.1618185416048283,
+      "grad_norm": 0.2522333241246855,
+      "learning_rate": 1.3495172867809525e-06,
+      "loss": 0.5619,
+      "step": 1470
+    },
+    {
+      "epoch": 1.1626100722271693,
+      "grad_norm": 0.24130571124051212,
+      "learning_rate": 1.3487382689940895e-06,
+      "loss": 0.5702,
+      "step": 1471
+    },
+    {
+      "epoch": 1.1634016028495102,
+      "grad_norm": 0.24551806796107772,
+      "learning_rate": 1.3479590101914102e-06,
+      "loss": 0.56,
+      "step": 1472
+    },
+    {
+      "epoch": 1.1641931334718512,
+      "grad_norm": 0.24466831966052552,
+      "learning_rate": 1.3471795109114657e-06,
+      "loss": 0.56,
+      "step": 1473
+    },
+    {
+      "epoch": 1.164984664094192,
+      "grad_norm": 0.24683947507642787,
+      "learning_rate": 1.346399771692975e-06,
+      "loss": 0.5764,
+      "step": 1474
+    },
+    {
+      "epoch": 1.165776194716533,
+      "grad_norm": 0.2562690119188769,
+      "learning_rate": 1.3456197930748218e-06,
+      "loss": 0.5612,
+      "step": 1475
+    },
+    {
+      "epoch": 1.166567725338874,
+      "grad_norm": 0.25396326381412243,
+      "learning_rate": 1.3448395755960557e-06,
+      "loss": 0.5618,
+      "step": 1476
+    },
+    {
+      "epoch": 1.167359255961215,
+      "grad_norm": 0.24145350778239516,
+      "learning_rate": 1.3440591197958914e-06,
+      "loss": 0.5624,
+      "step": 1477
+    },
+    {
+      "epoch": 1.1681507865835559,
+      "grad_norm": 0.2570848035221008,
+      "learning_rate": 1.343278426213708e-06,
+      "loss": 0.566,
+      "step": 1478
+    },
+    {
+      "epoch": 1.1689423172058968,
+      "grad_norm": 0.25821666285717104,
+      "learning_rate": 1.3424974953890486e-06,
+      "loss": 0.5481,
+      "step": 1479
+    },
+    {
+      "epoch": 1.1697338478282377,
+      "grad_norm": 0.2540322071803146,
+      "learning_rate": 1.3417163278616219e-06,
+      "loss": 0.5614,
+      "step": 1480
+    },
+    {
+      "epoch": 1.170525378450579,
+      "grad_norm": 0.24994903320454964,
+      "learning_rate": 1.340934924171298e-06,
+      "loss": 0.5683,
+      "step": 1481
+    },
+    {
+      "epoch": 1.1713169090729199,
+      "grad_norm": 0.24481055255582856,
+      "learning_rate": 1.3401532848581118e-06,
+      "loss": 0.5522,
+      "step": 1482
+    },
+    {
+      "epoch": 1.1721084396952608,
+      "grad_norm": 0.24942373276188304,
+      "learning_rate": 1.3393714104622605e-06,
+      "loss": 0.5594,
+      "step": 1483
+    },
+    {
+      "epoch": 1.1728999703176017,
+      "grad_norm": 0.26081266748278387,
+      "learning_rate": 1.3385893015241034e-06,
+      "loss": 0.5656,
+      "step": 1484
+    },
+    {
+      "epoch": 1.1736915009399427,
+      "grad_norm": 0.26034434454258165,
+      "learning_rate": 1.3378069585841628e-06,
+      "loss": 0.5638,
+      "step": 1485
+    },
+    {
+      "epoch": 1.1744830315622836,
+      "grad_norm": 0.2513850183241021,
+      "learning_rate": 1.3370243821831215e-06,
+      "loss": 0.5566,
+      "step": 1486
+    },
+    {
+      "epoch": 1.1752745621846246,
+      "grad_norm": 0.24043759547874047,
+      "learning_rate": 1.336241572861825e-06,
+      "loss": 0.5731,
+      "step": 1487
+    },
+    {
+      "epoch": 1.1760660928069655,
+      "grad_norm": 0.2580029426084782,
+      "learning_rate": 1.3354585311612787e-06,
+      "loss": 0.5627,
+      "step": 1488
+    },
+    {
+      "epoch": 1.1768576234293064,
+      "grad_norm": 0.27269605348413717,
+      "learning_rate": 1.3346752576226495e-06,
+      "loss": 0.5471,
+      "step": 1489
+    },
+    {
+      "epoch": 1.1776491540516474,
+      "grad_norm": 0.25760244919508485,
+      "learning_rate": 1.3338917527872633e-06,
+      "loss": 0.5577,
+      "step": 1490
+    },
+    {
+      "epoch": 1.1784406846739883,
+      "grad_norm": 0.24204885480871255,
+      "learning_rate": 1.3331080171966069e-06,
+      "loss": 0.5507,
+      "step": 1491
+    },
+    {
+      "epoch": 1.1792322152963293,
+      "grad_norm": 0.24490566217327128,
+      "learning_rate": 1.3323240513923266e-06,
+      "loss": 0.559,
+      "step": 1492
+    },
+    {
+      "epoch": 1.1800237459186702,
+      "grad_norm": 0.24394288222382954,
+      "learning_rate": 1.3315398559162268e-06,
+      "loss": 0.5681,
+      "step": 1493
+    },
+    {
+      "epoch": 1.1808152765410111,
+      "grad_norm": 0.2487697419681786,
+      "learning_rate": 1.330755431310272e-06,
+      "loss": 0.5627,
+      "step": 1494
+    },
+    {
+      "epoch": 1.181606807163352,
+      "grad_norm": 0.2577056813523912,
+      "learning_rate": 1.3299707781165838e-06,
+      "loss": 0.5682,
+      "step": 1495
+    },
+    {
+      "epoch": 1.182398337785693,
+      "grad_norm": 0.252776565485377,
+      "learning_rate": 1.3291858968774425e-06,
+      "loss": 0.5504,
+      "step": 1496
+    },
+    {
+      "epoch": 1.183189868408034,
+      "grad_norm": 0.23213210854724767,
+      "learning_rate": 1.3284007881352857e-06,
+      "loss": 0.5657,
+      "step": 1497
+    },
+    {
+      "epoch": 1.183981399030375,
+      "grad_norm": 0.24680774056848112,
+      "learning_rate": 1.3276154524327088e-06,
+      "loss": 0.5723,
+      "step": 1498
+    },
+    {
+      "epoch": 1.1847729296527159,
+      "grad_norm": 0.25150523401632685,
+      "learning_rate": 1.3268298903124632e-06,
+      "loss": 0.5589,
+      "step": 1499
+    },
+    {
+      "epoch": 1.1855644602750568,
+      "grad_norm": 0.251403052409526,
+      "learning_rate": 1.3260441023174573e-06,
+      "loss": 0.5676,
+      "step": 1500
+    },
+    {
+      "epoch": 1.1863559908973977,
+      "grad_norm": 0.25303547261663323,
+      "learning_rate": 1.3252580889907552e-06,
+      "loss": 0.571,
+      "step": 1501
+    },
+    {
+      "epoch": 1.187147521519739,
+      "grad_norm": 0.24525609893856698,
+      "learning_rate": 1.3244718508755775e-06,
+      "loss": 0.5723,
+      "step": 1502
+    },
+    {
+      "epoch": 1.1879390521420798,
+      "grad_norm": 0.24875562634094203,
+      "learning_rate": 1.3236853885152993e-06,
+      "loss": 0.5567,
+      "step": 1503
+    },
+    {
+      "epoch": 1.1887305827644208,
+      "grad_norm": 0.2463317871919473,
+      "learning_rate": 1.3228987024534515e-06,
+      "loss": 0.5715,
+      "step": 1504
+    },
+    {
+      "epoch": 1.1895221133867617,
+      "grad_norm": 0.24545871745197517,
+      "learning_rate": 1.3221117932337187e-06,
+      "loss": 0.5717,
+      "step": 1505
+    },
+    {
+      "epoch": 1.1903136440091027,
+      "grad_norm": 0.2473448767957027,
+      "learning_rate": 1.3213246613999403e-06,
+      "loss": 0.5675,
+      "step": 1506
+    },
+    {
+      "epoch": 1.1911051746314436,
+      "grad_norm": 0.24629434812573725,
+      "learning_rate": 1.3205373074961095e-06,
+      "loss": 0.559,
+      "step": 1507
+    },
+    {
+      "epoch": 1.1918967052537845,
+      "grad_norm": 0.24434319715655584,
+      "learning_rate": 1.319749732066373e-06,
+      "loss": 0.5592,
+      "step": 1508
+    },
+    {
+      "epoch": 1.1926882358761255,
+      "grad_norm": 0.24964508059499005,
+      "learning_rate": 1.3189619356550301e-06,
+      "loss": 0.5745,
+      "step": 1509
+    },
+    {
+      "epoch": 1.1934797664984664,
+      "grad_norm": 0.25254233963159106,
+      "learning_rate": 1.3181739188065337e-06,
+      "loss": 0.5565,
+      "step": 1510
+    },
+    {
+      "epoch": 1.1942712971208074,
+      "grad_norm": 0.24142885431486732,
+      "learning_rate": 1.3173856820654884e-06,
+      "loss": 0.5544,
+      "step": 1511
+    },
+    {
+      "epoch": 1.1950628277431483,
+      "grad_norm": 0.24365351975783034,
+      "learning_rate": 1.3165972259766504e-06,
+      "loss": 0.5678,
+      "step": 1512
+    },
+    {
+      "epoch": 1.1958543583654893,
+      "grad_norm": 0.24247554455887,
+      "learning_rate": 1.3158085510849292e-06,
+      "loss": 0.5498,
+      "step": 1513
+    },
+    {
+      "epoch": 1.1966458889878302,
+      "grad_norm": 0.24241098462339952,
+      "learning_rate": 1.3150196579353833e-06,
+      "loss": 0.5648,
+      "step": 1514
+    },
+    {
+      "epoch": 1.1974374196101711,
+      "grad_norm": 0.24249053802488338,
+      "learning_rate": 1.3142305470732235e-06,
+      "loss": 0.5608,
+      "step": 1515
+    },
+    {
+      "epoch": 1.198228950232512,
+      "grad_norm": 0.24206188838756354,
+      "learning_rate": 1.3134412190438108e-06,
+      "loss": 0.5813,
+      "step": 1516
+    },
+    {
+      "epoch": 1.199020480854853,
+      "grad_norm": 0.2514494077696141,
+      "learning_rate": 1.3126516743926557e-06,
+      "loss": 0.555,
+      "step": 1517
+    },
+    {
+      "epoch": 1.199812011477194,
+      "grad_norm": 0.24250342755119866,
+      "learning_rate": 1.3118619136654194e-06,
+      "loss": 0.5824,
+      "step": 1518
+    },
+    {
+      "epoch": 1.200603542099535,
+      "grad_norm": 0.23787699251894545,
+      "learning_rate": 1.3110719374079114e-06,
+      "loss": 0.5562,
+      "step": 1519
+    },
+    {
+      "epoch": 1.2013950727218758,
+      "grad_norm": 0.2479687422231374,
+      "learning_rate": 1.3102817461660907e-06,
+      "loss": 0.56,
+      "step": 1520
+    },
+    {
+      "epoch": 1.2021866033442168,
+      "grad_norm": 0.23767241740347347,
+      "learning_rate": 1.309491340486065e-06,
+      "loss": 0.5574,
+      "step": 1521
+    },
+    {
+      "epoch": 1.2029781339665577,
+      "grad_norm": 0.2446681705477864,
+      "learning_rate": 1.30870072091409e-06,
+      "loss": 0.5622,
+      "step": 1522
+    },
+    {
+      "epoch": 1.203769664588899,
+      "grad_norm": 0.24475978456393724,
+      "learning_rate": 1.3079098879965692e-06,
+      "loss": 0.5478,
+      "step": 1523
+    },
+    {
+      "epoch": 1.2045611952112396,
+      "grad_norm": 0.24473406227114267,
+      "learning_rate": 1.3071188422800533e-06,
+      "loss": 0.5373,
+      "step": 1524
+    },
+    {
+      "epoch": 1.2053527258335808,
+      "grad_norm": 0.2429482384203616,
+      "learning_rate": 1.3063275843112408e-06,
+      "loss": 0.5667,
+      "step": 1525
+    },
+    {
+      "epoch": 1.2061442564559217,
+      "grad_norm": 0.2569384178130532,
+      "learning_rate": 1.305536114636976e-06,
+      "loss": 0.5667,
+      "step": 1526
+    },
+    {
+      "epoch": 1.2069357870782627,
+      "grad_norm": 0.25473603378901205,
+      "learning_rate": 1.3047444338042503e-06,
+      "loss": 0.5591,
+      "step": 1527
+    },
+    {
+      "epoch": 1.2077273177006036,
+      "grad_norm": 0.25697854502479367,
+      "learning_rate": 1.3039525423602005e-06,
+      "loss": 0.5775,
+      "step": 1528
+    },
+    {
+      "epoch": 1.2085188483229445,
+      "grad_norm": 0.25238019491301633,
+      "learning_rate": 1.3031604408521091e-06,
+      "loss": 0.5844,
+      "step": 1529
+    },
+    {
+      "epoch": 1.2093103789452855,
+      "grad_norm": 0.25229905848431244,
+      "learning_rate": 1.302368129827404e-06,
+      "loss": 0.5854,
+      "step": 1530
+    },
+    {
+      "epoch": 1.2101019095676264,
+      "grad_norm": 0.2664347693257625,
+      "learning_rate": 1.301575609833657e-06,
+      "loss": 0.5562,
+      "step": 1531
+    },
+    {
+      "epoch": 1.2108934401899674,
+      "grad_norm": 0.2496922180924536,
+      "learning_rate": 1.3007828814185865e-06,
+      "loss": 0.5684,
+      "step": 1532
+    },
+    {
+      "epoch": 1.2116849708123083,
+      "grad_norm": 0.2570095455485703,
+      "learning_rate": 1.299989945130052e-06,
+      "loss": 0.5587,
+      "step": 1533
+    },
+    {
+      "epoch": 1.2124765014346492,
+      "grad_norm": 0.2504829012924918,
+      "learning_rate": 1.299196801516059e-06,
+      "loss": 0.5739,
+      "step": 1534
+    },
+    {
+      "epoch": 1.2132680320569902,
+      "grad_norm": 0.25609771518114627,
+      "learning_rate": 1.2984034511247549e-06,
+      "loss": 0.5503,
+      "step": 1535
+    },
+    {
+      "epoch": 1.2140595626793311,
+      "grad_norm": 0.2588638522992403,
+      "learning_rate": 1.297609894504431e-06,
+      "loss": 0.5642,
+      "step": 1536
+    },
+    {
+      "epoch": 1.214851093301672,
+      "grad_norm": 0.24021387096343427,
+      "learning_rate": 1.2968161322035203e-06,
+      "loss": 0.5589,
+      "step": 1537
+    },
+    {
+      "epoch": 1.215642623924013,
+      "grad_norm": 0.2466955191888222,
+      "learning_rate": 1.2960221647705983e-06,
+      "loss": 0.5628,
+      "step": 1538
+    },
+    {
+      "epoch": 1.216434154546354,
+      "grad_norm": 0.24947350639256644,
+      "learning_rate": 1.2952279927543824e-06,
+      "loss": 0.5776,
+      "step": 1539
+    },
+    {
+      "epoch": 1.217225685168695,
+      "grad_norm": 0.2560276558387007,
+      "learning_rate": 1.294433616703731e-06,
+      "loss": 0.575,
+      "step": 1540
+    },
+    {
+      "epoch": 1.2180172157910358,
+      "grad_norm": 0.2460633195241213,
+      "learning_rate": 1.2936390371676444e-06,
+      "loss": 0.5613,
+      "step": 1541
+    },
+    {
+      "epoch": 1.2188087464133768,
+      "grad_norm": 0.24500623054045792,
+      "learning_rate": 1.2928442546952624e-06,
+      "loss": 0.5732,
+      "step": 1542
+    },
+    {
+      "epoch": 1.2196002770357177,
+      "grad_norm": 0.2519740921886011,
+      "learning_rate": 1.2920492698358654e-06,
+      "loss": 0.5606,
+      "step": 1543
+    },
+    {
+      "epoch": 1.2203918076580589,
+      "grad_norm": 0.25154306115101094,
+      "learning_rate": 1.291254083138874e-06,
+      "loss": 0.5673,
+      "step": 1544
+    },
+    {
+      "epoch": 1.2211833382803996,
+      "grad_norm": 0.23946440421747092,
+      "learning_rate": 1.2904586951538482e-06,
+      "loss": 0.5735,
+      "step": 1545
+    },
+    {
+      "epoch": 1.2219748689027408,
+      "grad_norm": 0.2444312352710892,
+      "learning_rate": 1.2896631064304875e-06,
+      "loss": 0.5624,
+      "step": 1546
+    },
+    {
+      "epoch": 1.2227663995250817,
+      "grad_norm": 0.2417617325038712,
+      "learning_rate": 1.2888673175186285e-06,
+      "loss": 0.5539,
+      "step": 1547
+    },
+    {
+      "epoch": 1.2235579301474226,
+      "grad_norm": 0.2433204443864132,
+      "learning_rate": 1.2880713289682482e-06,
+      "loss": 0.5792,
+      "step": 1548
+    },
+    {
+      "epoch": 1.2243494607697636,
+      "grad_norm": 0.2412703302687474,
+      "learning_rate": 1.2872751413294603e-06,
+      "loss": 0.5412,
+      "step": 1549
+    },
+    {
+      "epoch": 1.2251409913921045,
+      "grad_norm": 0.244643351278545,
+      "learning_rate": 1.2864787551525165e-06,
+      "loss": 0.5694,
+      "step": 1550
+    },
+    {
+      "epoch": 1.2259325220144455,
+      "grad_norm": 0.23973020727942237,
+      "learning_rate": 1.285682170987806e-06,
+      "loss": 0.5735,
+      "step": 1551
+    },
+    {
+      "epoch": 1.2267240526367864,
+      "grad_norm": 0.24234936017315478,
+      "learning_rate": 1.2848853893858538e-06,
+      "loss": 0.5652,
+      "step": 1552
+    },
+    {
+      "epoch": 1.2275155832591274,
+      "grad_norm": 0.2586282557349905,
+      "learning_rate": 1.284088410897323e-06,
+      "loss": 0.5605,
+      "step": 1553
+    },
+    {
+      "epoch": 1.2283071138814683,
+      "grad_norm": 0.24505684808942604,
+      "learning_rate": 1.283291236073011e-06,
+      "loss": 0.5547,
+      "step": 1554
+    },
+    {
+      "epoch": 1.2290986445038092,
+      "grad_norm": 0.251500866370209,
+      "learning_rate": 1.2824938654638523e-06,
+      "loss": 0.5574,
+      "step": 1555
+    },
+    {
+      "epoch": 1.2298901751261502,
+      "grad_norm": 0.24257617276229285,
+      "learning_rate": 1.2816962996209158e-06,
+      "loss": 0.5623,
+      "step": 1556
+    },
+    {
+      "epoch": 1.2306817057484911,
+      "grad_norm": 0.24439097721753178,
+      "learning_rate": 1.2808985390954058e-06,
+      "loss": 0.5528,
+      "step": 1557
+    },
+    {
+      "epoch": 1.231473236370832,
+      "grad_norm": 0.24002520337491023,
+      "learning_rate": 1.2801005844386608e-06,
+      "loss": 0.5457,
+      "step": 1558
+    },
+    {
+      "epoch": 1.232264766993173,
+      "grad_norm": 0.2489675951780969,
+      "learning_rate": 1.2793024362021539e-06,
+      "loss": 0.5625,
+      "step": 1559
+    },
+    {
+      "epoch": 1.233056297615514,
+      "grad_norm": 0.24263360465866113,
+      "learning_rate": 1.2785040949374913e-06,
+      "loss": 0.574,
+      "step": 1560
+    },
+    {
+      "epoch": 1.2338478282378549,
+      "grad_norm": 0.23885755915714316,
+      "learning_rate": 1.2777055611964143e-06,
+      "loss": 0.552,
+      "step": 1561
+    },
+    {
+      "epoch": 1.2346393588601958,
+      "grad_norm": 0.24079241769982115,
+      "learning_rate": 1.2769068355307944e-06,
+      "loss": 0.5677,
+      "step": 1562
+    },
+    {
+      "epoch": 1.2354308894825368,
+      "grad_norm": 0.2609744072790651,
+      "learning_rate": 1.2761079184926383e-06,
+      "loss": 0.5477,
+      "step": 1563
+    },
+    {
+      "epoch": 1.2362224201048777,
+      "grad_norm": 0.24786241690689384,
+      "learning_rate": 1.2753088106340834e-06,
+      "loss": 0.5567,
+      "step": 1564
+    },
+    {
+      "epoch": 1.2370139507272189,
+      "grad_norm": 0.24911141713974588,
+      "learning_rate": 1.2745095125074e-06,
+      "loss": 0.5405,
+      "step": 1565
+    },
+    {
+      "epoch": 1.2378054813495596,
+      "grad_norm": 0.24570528505748546,
+      "learning_rate": 1.2737100246649894e-06,
+      "loss": 0.5622,
+      "step": 1566
+    },
+    {
+      "epoch": 1.2385970119719008,
+      "grad_norm": 0.23695499893006872,
+      "learning_rate": 1.2729103476593839e-06,
+      "loss": 0.5591,
+      "step": 1567
+    },
+    {
+      "epoch": 1.2393885425942417,
+      "grad_norm": 0.25451569776294236,
+      "learning_rate": 1.272110482043247e-06,
+      "loss": 0.5478,
+      "step": 1568
+    },
+    {
+      "epoch": 1.2401800732165826,
+      "grad_norm": 0.26976294471550605,
+      "learning_rate": 1.2713104283693717e-06,
+      "loss": 0.5625,
+      "step": 1569
+    },
+    {
+      "epoch": 1.2409716038389236,
+      "grad_norm": 0.25424281949570265,
+      "learning_rate": 1.2705101871906819e-06,
+      "loss": 0.5721,
+      "step": 1570
+    },
+    {
+      "epoch": 1.2417631344612645,
+      "grad_norm": 0.3145157934672575,
+      "learning_rate": 1.2697097590602309e-06,
+      "loss": 0.5653,
+      "step": 1571
+    },
+    {
+      "epoch": 1.2425546650836055,
+      "grad_norm": 0.26214677001673914,
+      "learning_rate": 1.2689091445312002e-06,
+      "loss": 0.5602,
+      "step": 1572
+    },
+    {
+      "epoch": 1.2433461957059464,
+      "grad_norm": 0.248967012073164,
+      "learning_rate": 1.2681083441569016e-06,
+      "loss": 0.5699,
+      "step": 1573
+    },
+    {
+      "epoch": 1.2441377263282873,
+      "grad_norm": 0.24808362195885564,
+      "learning_rate": 1.2673073584907748e-06,
+      "loss": 0.5566,
+      "step": 1574
+    },
+    {
+      "epoch": 1.2449292569506283,
+      "grad_norm": 0.24101514028608262,
+      "learning_rate": 1.266506188086387e-06,
+      "loss": 0.5426,
+      "step": 1575
+    },
+    {
+      "epoch": 1.2457207875729692,
+      "grad_norm": 0.2483643234417837,
+      "learning_rate": 1.265704833497434e-06,
+      "loss": 0.5553,
+      "step": 1576
+    },
+    {
+      "epoch": 1.2465123181953102,
+      "grad_norm": 0.25406125539863356,
+      "learning_rate": 1.2649032952777374e-06,
+      "loss": 0.5452,
+      "step": 1577
+    },
+    {
+      "epoch": 1.247303848817651,
+      "grad_norm": 0.25646760949952596,
+      "learning_rate": 1.2641015739812474e-06,
+      "loss": 0.5733,
+      "step": 1578
+    },
+    {
+      "epoch": 1.248095379439992,
+      "grad_norm": 0.2501605676457125,
+      "learning_rate": 1.2632996701620405e-06,
+      "loss": 0.565,
+      "step": 1579
+    },
+    {
+      "epoch": 1.248886910062333,
+      "grad_norm": 0.25303082598696736,
+      "learning_rate": 1.262497584374318e-06,
+      "loss": 0.5587,
+      "step": 1580
+    },
+    {
+      "epoch": 1.249678440684674,
+      "grad_norm": 0.24583501946750871,
+      "learning_rate": 1.2616953171724078e-06,
+      "loss": 0.5726,
+      "step": 1581
+    },
+    {
+      "epoch": 1.2504699713070149,
+      "grad_norm": 0.2597718060699548,
+      "learning_rate": 1.260892869110764e-06,
+      "loss": 0.563,
+      "step": 1582
+    },
+    {
+      "epoch": 1.2512615019293558,
+      "grad_norm": 0.23711734212394223,
+      "learning_rate": 1.2600902407439643e-06,
+      "loss": 0.5581,
+      "step": 1583
+    },
+    {
+      "epoch": 1.252053032551697,
+      "grad_norm": 0.2379141319323817,
+      "learning_rate": 1.259287432626712e-06,
+      "loss": 0.5453,
+      "step": 1584
+    },
+    {
+      "epoch": 1.2528445631740377,
+      "grad_norm": 0.24675536545058904,
+      "learning_rate": 1.2584844453138334e-06,
+      "loss": 0.5534,
+      "step": 1585
+    },
+    {
+      "epoch": 1.2536360937963789,
+      "grad_norm": 0.2416337478880738,
+      "learning_rate": 1.2576812793602804e-06,
+      "loss": 0.5531,
+      "step": 1586
+    },
+    {
+      "epoch": 1.2544276244187196,
+      "grad_norm": 0.25820029418212614,
+      "learning_rate": 1.2568779353211272e-06,
+      "loss": 0.5551,
+      "step": 1587
+    },
+    {
+      "epoch": 1.2552191550410607,
+      "grad_norm": 0.2493044952822321,
+      "learning_rate": 1.256074413751571e-06,
+      "loss": 0.5646,
+      "step": 1588
+    },
+    {
+      "epoch": 1.2560106856634017,
+      "grad_norm": 0.2494315770295232,
+      "learning_rate": 1.2552707152069325e-06,
+      "loss": 0.5644,
+      "step": 1589
+    },
+    {
+      "epoch": 1.2568022162857426,
+      "grad_norm": 0.2430525866343491,
+      "learning_rate": 1.2544668402426538e-06,
+      "loss": 0.5618,
+      "step": 1590
+    },
+    {
+      "epoch": 1.2575937469080836,
+      "grad_norm": 0.24358367177137008,
+      "learning_rate": 1.253662789414299e-06,
+      "loss": 0.5691,
+      "step": 1591
+    },
+    {
+      "epoch": 1.2583852775304245,
+      "grad_norm": 0.2535198189922417,
+      "learning_rate": 1.252858563277555e-06,
+      "loss": 0.5737,
+      "step": 1592
+    },
+    {
+      "epoch": 1.2591768081527654,
+      "grad_norm": 0.24582225593517482,
+      "learning_rate": 1.2520541623882287e-06,
+      "loss": 0.5499,
+      "step": 1593
+    },
+    {
+      "epoch": 1.2599683387751064,
+      "grad_norm": 0.24472109624030428,
+      "learning_rate": 1.2512495873022474e-06,
+      "loss": 0.5543,
+      "step": 1594
+    },
+    {
+      "epoch": 1.2607598693974473,
+      "grad_norm": 0.25137341214156705,
+      "learning_rate": 1.2504448385756598e-06,
+      "loss": 0.5464,
+      "step": 1595
+    },
+    {
+      "epoch": 1.2615514000197883,
+      "grad_norm": 0.2627554199798165,
+      "learning_rate": 1.2496399167646342e-06,
+      "loss": 0.5566,
+      "step": 1596
+    },
+    {
+      "epoch": 1.2623429306421292,
+      "grad_norm": 0.24264954666822985,
+      "learning_rate": 1.2488348224254588e-06,
+      "loss": 0.5537,
+      "step": 1597
+    },
+    {
+      "epoch": 1.2631344612644702,
+      "grad_norm": 0.23472343110165828,
+      "learning_rate": 1.2480295561145405e-06,
+      "loss": 0.5536,
+      "step": 1598
+    },
+    {
+      "epoch": 1.263925991886811,
+      "grad_norm": 0.23799331683738528,
+      "learning_rate": 1.247224118388405e-06,
+      "loss": 0.5651,
+      "step": 1599
+    },
+    {
+      "epoch": 1.264717522509152,
+      "grad_norm": 0.24011674675311853,
+      "learning_rate": 1.246418509803697e-06,
+      "loss": 0.5637,
+      "step": 1600
+    },
+    {
+      "epoch": 1.265509053131493,
+      "grad_norm": 0.2417891185532008,
+      "learning_rate": 1.2456127309171793e-06,
+      "loss": 0.5638,
+      "step": 1601
+    },
+    {
+      "epoch": 1.266300583753834,
+      "grad_norm": 0.24520499447330632,
+      "learning_rate": 1.244806782285732e-06,
+      "loss": 0.5686,
+      "step": 1602
+    },
+    {
+      "epoch": 1.2670921143761749,
+      "grad_norm": 0.24736570319096188,
+      "learning_rate": 1.2440006644663528e-06,
+      "loss": 0.5573,
+      "step": 1603
+    },
+    {
+      "epoch": 1.2678836449985158,
+      "grad_norm": 0.2394906420948201,
+      "learning_rate": 1.2431943780161556e-06,
+      "loss": 0.5683,
+      "step": 1604
+    },
+    {
+      "epoch": 1.268675175620857,
+      "grad_norm": 0.24619462045414997,
+      "learning_rate": 1.2423879234923723e-06,
+      "loss": 0.5663,
+      "step": 1605
+    },
+    {
+      "epoch": 1.2694667062431977,
+      "grad_norm": 0.24828261671946197,
+      "learning_rate": 1.2415813014523496e-06,
+      "loss": 0.566,
+      "step": 1606
+    },
+    {
+      "epoch": 1.2702582368655388,
+      "grad_norm": 0.23611278428927193,
+      "learning_rate": 1.2407745124535505e-06,
+      "loss": 0.5645,
+      "step": 1607
+    },
+    {
+      "epoch": 1.2710497674878796,
+      "grad_norm": 0.24202287638992181,
+      "learning_rate": 1.2399675570535532e-06,
+      "loss": 0.58,
+      "step": 1608
+    },
+    {
+      "epoch": 1.2718412981102207,
+      "grad_norm": 0.23897979132953442,
+      "learning_rate": 1.2391604358100514e-06,
+      "loss": 0.5557,
+      "step": 1609
+    },
+    {
+      "epoch": 1.2726328287325617,
+      "grad_norm": 3.69549056371507,
+      "learning_rate": 1.2383531492808529e-06,
+      "loss": 0.5745,
+      "step": 1610
+    },
+    {
+      "epoch": 1.2734243593549026,
+      "grad_norm": 0.24544340196952252,
+      "learning_rate": 1.23754569802388e-06,
+      "loss": 0.5661,
+      "step": 1611
+    },
+    {
+      "epoch": 1.2742158899772436,
+      "grad_norm": 0.25838718505166175,
+      "learning_rate": 1.2367380825971686e-06,
+      "loss": 0.5431,
+      "step": 1612
+    },
+    {
+      "epoch": 1.2750074205995845,
+      "grad_norm": 0.25681360600743736,
+      "learning_rate": 1.2359303035588683e-06,
+      "loss": 0.5499,
+      "step": 1613
+    },
+    {
+      "epoch": 1.2757989512219254,
+      "grad_norm": 0.2360565094592681,
+      "learning_rate": 1.2351223614672417e-06,
+      "loss": 0.5618,
+      "step": 1614
+    },
+    {
+      "epoch": 1.2765904818442664,
+      "grad_norm": 0.23704001453136386,
+      "learning_rate": 1.2343142568806638e-06,
+      "loss": 0.557,
+      "step": 1615
+    },
+    {
+      "epoch": 1.2773820124666073,
+      "grad_norm": 0.2550329749989943,
+      "learning_rate": 1.2335059903576227e-06,
+      "loss": 0.5657,
+      "step": 1616
+    },
+    {
+      "epoch": 1.2781735430889483,
+      "grad_norm": 0.24638931620607757,
+      "learning_rate": 1.2326975624567173e-06,
+      "loss": 0.5601,
+      "step": 1617
+    },
+    {
+      "epoch": 1.2789650737112892,
+      "grad_norm": 0.2451404275754642,
+      "learning_rate": 1.231888973736659e-06,
+      "loss": 0.5636,
+      "step": 1618
+    },
+    {
+      "epoch": 1.2797566043336301,
+      "grad_norm": 0.2490864823150462,
+      "learning_rate": 1.2310802247562696e-06,
+      "loss": 0.5668,
+      "step": 1619
+    },
+    {
+      "epoch": 1.280548134955971,
+      "grad_norm": 0.2494801103850093,
+      "learning_rate": 1.2302713160744822e-06,
+      "loss": 0.5563,
+      "step": 1620
+    },
+    {
+      "epoch": 1.281339665578312,
+      "grad_norm": 0.2455574765759398,
+      "learning_rate": 1.2294622482503403e-06,
+      "loss": 0.5471,
+      "step": 1621
+    },
+    {
+      "epoch": 1.282131196200653,
+      "grad_norm": 0.24880822483734186,
+      "learning_rate": 1.2286530218429972e-06,
+      "loss": 0.5616,
+      "step": 1622
+    },
+    {
+      "epoch": 1.282922726822994,
+      "grad_norm": 0.2392897415582679,
+      "learning_rate": 1.2278436374117151e-06,
+      "loss": 0.5693,
+      "step": 1623
+    },
+    {
+      "epoch": 1.2837142574453348,
+      "grad_norm": 0.24940610953675144,
+      "learning_rate": 1.2270340955158668e-06,
+      "loss": 0.5542,
+      "step": 1624
+    },
+    {
+      "epoch": 1.2845057880676758,
+      "grad_norm": 0.2500966139602465,
+      "learning_rate": 1.2262243967149326e-06,
+      "loss": 0.5415,
+      "step": 1625
+    },
+    {
+      "epoch": 1.285297318690017,
+      "grad_norm": 0.24248185785508394,
+      "learning_rate": 1.2254145415685025e-06,
+      "loss": 0.5603,
+      "step": 1626
+    },
+    {
+      "epoch": 1.2860888493123577,
+      "grad_norm": 0.24258695020659132,
+      "learning_rate": 1.2246045306362736e-06,
+      "loss": 0.5651,
+      "step": 1627
+    },
+    {
+      "epoch": 1.2868803799346988,
+      "grad_norm": 0.254653360005497,
+      "learning_rate": 1.2237943644780507e-06,
+      "loss": 0.5442,
+      "step": 1628
+    },
+    {
+      "epoch": 1.2876719105570396,
+      "grad_norm": 0.2493302726077809,
+      "learning_rate": 1.2229840436537464e-06,
+      "loss": 0.5724,
+      "step": 1629
+    },
+    {
+      "epoch": 1.2884634411793807,
+      "grad_norm": 0.2402514109423032,
+      "learning_rate": 1.2221735687233797e-06,
+      "loss": 0.5564,
+      "step": 1630
+    },
+    {
+      "epoch": 1.2892549718017217,
+      "grad_norm": 0.2471478455884797,
+      "learning_rate": 1.2213629402470766e-06,
+      "loss": 0.5627,
+      "step": 1631
+    },
+    {
+      "epoch": 1.2900465024240626,
+      "grad_norm": 0.23997860829436274,
+      "learning_rate": 1.2205521587850692e-06,
+      "loss": 0.5465,
+      "step": 1632
+    },
+    {
+      "epoch": 1.2908380330464035,
+      "grad_norm": 0.24985884693129512,
+      "learning_rate": 1.2197412248976942e-06,
+      "loss": 0.5575,
+      "step": 1633
+    },
+    {
+      "epoch": 1.2916295636687445,
+      "grad_norm": 0.24793115694862064,
+      "learning_rate": 1.2189301391453954e-06,
+      "loss": 0.5561,
+      "step": 1634
+    },
+    {
+      "epoch": 1.2924210942910854,
+      "grad_norm": 0.24439507970308486,
+      "learning_rate": 1.21811890208872e-06,
+      "loss": 0.5728,
+      "step": 1635
+    },
+    {
+      "epoch": 1.2932126249134264,
+      "grad_norm": 0.24945868420510264,
+      "learning_rate": 1.217307514288321e-06,
+      "loss": 0.5568,
+      "step": 1636
+    },
+    {
+      "epoch": 1.2940041555357673,
+      "grad_norm": 0.24873126082475502,
+      "learning_rate": 1.2164959763049547e-06,
+      "loss": 0.5465,
+      "step": 1637
+    },
+    {
+      "epoch": 1.2947956861581082,
+      "grad_norm": 0.2456054378861574,
+      "learning_rate": 1.2156842886994815e-06,
+      "loss": 0.5637,
+      "step": 1638
+    },
+    {
+      "epoch": 1.2955872167804492,
+      "grad_norm": 0.240049154102141,
+      "learning_rate": 1.2148724520328655e-06,
+      "loss": 0.554,
+      "step": 1639
+    },
+    {
+      "epoch": 1.2963787474027901,
+      "grad_norm": 0.25353406818399277,
+      "learning_rate": 1.214060466866173e-06,
+      "loss": 0.5584,
+      "step": 1640
+    },
+    {
+      "epoch": 1.297170278025131,
+      "grad_norm": 0.2365662911935809,
+      "learning_rate": 1.2132483337605744e-06,
+      "loss": 0.5671,
+      "step": 1641
+    },
+    {
+      "epoch": 1.297961808647472,
+      "grad_norm": 0.2463877003258589,
+      "learning_rate": 1.2124360532773402e-06,
+      "loss": 0.5588,
+      "step": 1642
+    },
+    {
+      "epoch": 1.298753339269813,
+      "grad_norm": 0.2459247729337839,
+      "learning_rate": 1.211623625977845e-06,
+      "loss": 0.5651,
+      "step": 1643
+    },
+    {
+      "epoch": 1.299544869892154,
+      "grad_norm": 0.2460048320809122,
+      "learning_rate": 1.210811052423563e-06,
+      "loss": 0.5464,
+      "step": 1644
+    },
+    {
+      "epoch": 1.3003364005144948,
+      "grad_norm": 0.25144426047242724,
+      "learning_rate": 1.2099983331760713e-06,
+      "loss": 0.5591,
+      "step": 1645
+    },
+    {
+      "epoch": 1.3011279311368358,
+      "grad_norm": 0.24129146619153063,
+      "learning_rate": 1.2091854687970455e-06,
+      "loss": 0.5486,
+      "step": 1646
+    },
+    {
+      "epoch": 1.301919461759177,
+      "grad_norm": 0.2452087727819343,
+      "learning_rate": 1.2083724598482636e-06,
+      "loss": 0.5606,
+      "step": 1647
+    },
+    {
+      "epoch": 1.3027109923815177,
+      "grad_norm": 0.2558418227604141,
+      "learning_rate": 1.2075593068916022e-06,
+      "loss": 0.567,
+      "step": 1648
+    },
+    {
+      "epoch": 1.3035025230038588,
+      "grad_norm": 0.24655924383451752,
+      "learning_rate": 1.2067460104890377e-06,
+      "loss": 0.5533,
+      "step": 1649
+    },
+    {
+      "epoch": 1.3042940536261995,
+      "grad_norm": 0.2444611861271756,
+      "learning_rate": 1.2059325712026468e-06,
+      "loss": 0.5631,
+      "step": 1650
+    },
+    {
+      "epoch": 1.3050855842485407,
+      "grad_norm": 0.24142955295091997,
+      "learning_rate": 1.2051189895946024e-06,
+      "loss": 0.5577,
+      "step": 1651
+    },
+    {
+      "epoch": 1.3058771148708817,
+      "grad_norm": 0.25144848411228726,
+      "learning_rate": 1.2043052662271781e-06,
+      "loss": 0.5451,
+      "step": 1652
+    },
+    {
+      "epoch": 1.3066686454932226,
+      "grad_norm": 0.24858016163654087,
+      "learning_rate": 1.2034914016627447e-06,
+      "loss": 0.5624,
+      "step": 1653
+    },
+    {
+      "epoch": 1.3074601761155635,
+      "grad_norm": 0.25024588430180106,
+      "learning_rate": 1.2026773964637703e-06,
+      "loss": 0.5632,
+      "step": 1654
+    },
+    {
+      "epoch": 1.3082517067379045,
+      "grad_norm": 0.23830548334228782,
+      "learning_rate": 1.2018632511928208e-06,
+      "loss": 0.548,
+      "step": 1655
+    },
+    {
+      "epoch": 1.3090432373602454,
+      "grad_norm": 0.2538608293049789,
+      "learning_rate": 1.201048966412558e-06,
+      "loss": 0.5675,
+      "step": 1656
+    },
+    {
+      "epoch": 1.3098347679825864,
+      "grad_norm": 0.2516917880013272,
+      "learning_rate": 1.200234542685741e-06,
+      "loss": 0.58,
+      "step": 1657
+    },
+    {
+      "epoch": 1.3106262986049273,
+      "grad_norm": 0.24721531887082182,
+      "learning_rate": 1.1994199805752243e-06,
+      "loss": 0.5444,
+      "step": 1658
+    },
+    {
+      "epoch": 1.3114178292272682,
+      "grad_norm": 0.2520269105815416,
+      "learning_rate": 1.1986052806439589e-06,
+      "loss": 0.5606,
+      "step": 1659
+    },
+    {
+      "epoch": 1.3122093598496092,
+      "grad_norm": 0.2350653945276768,
+      "learning_rate": 1.1977904434549898e-06,
+      "loss": 0.5516,
+      "step": 1660
+    },
+    {
+      "epoch": 1.3130008904719501,
+      "grad_norm": 0.24388750167583753,
+      "learning_rate": 1.1969754695714575e-06,
+      "loss": 0.5591,
+      "step": 1661
+    },
+    {
+      "epoch": 1.313792421094291,
+      "grad_norm": 0.252058942375097,
+      "learning_rate": 1.1961603595565974e-06,
+      "loss": 0.5503,
+      "step": 1662
+    },
+    {
+      "epoch": 1.314583951716632,
+      "grad_norm": 0.24874839822478895,
+      "learning_rate": 1.1953451139737388e-06,
+      "loss": 0.5686,
+      "step": 1663
+    },
+    {
+      "epoch": 1.315375482338973,
+      "grad_norm": 0.23075194020607312,
+      "learning_rate": 1.194529733386304e-06,
+      "loss": 0.579,
+      "step": 1664
+    },
+    {
+      "epoch": 1.3161670129613139,
+      "grad_norm": 0.2552051894720154,
+      "learning_rate": 1.193714218357809e-06,
+      "loss": 0.5486,
+      "step": 1665
+    },
+    {
+      "epoch": 1.3169585435836548,
+      "grad_norm": 0.25481829889713686,
+      "learning_rate": 1.1928985694518627e-06,
+      "loss": 0.5487,
+      "step": 1666
+    },
+    {
+      "epoch": 1.3177500742059958,
+      "grad_norm": 0.24478122947030165,
+      "learning_rate": 1.192082787232167e-06,
+      "loss": 0.5586,
+      "step": 1667
+    },
+    {
+      "epoch": 1.318541604828337,
+      "grad_norm": 0.24904610342558647,
+      "learning_rate": 1.1912668722625157e-06,
+      "loss": 0.5588,
+      "step": 1668
+    },
+    {
+      "epoch": 1.3193331354506777,
+      "grad_norm": 0.2529525708987727,
+      "learning_rate": 1.1904508251067938e-06,
+      "loss": 0.5873,
+      "step": 1669
+    },
+    {
+      "epoch": 1.3201246660730188,
+      "grad_norm": 0.2453123415283199,
+      "learning_rate": 1.1896346463289778e-06,
+      "loss": 0.5562,
+      "step": 1670
+    },
+    {
+      "epoch": 1.3209161966953595,
+      "grad_norm": 0.2416389082493275,
+      "learning_rate": 1.188818336493136e-06,
+      "loss": 0.5439,
+      "step": 1671
+    },
+    {
+      "epoch": 1.3217077273177007,
+      "grad_norm": 0.23635897181983642,
+      "learning_rate": 1.1880018961634261e-06,
+      "loss": 0.5428,
+      "step": 1672
+    },
+    {
+      "epoch": 1.3224992579400414,
+      "grad_norm": 0.25313133701478413,
+      "learning_rate": 1.1871853259040972e-06,
+      "loss": 0.55,
+      "step": 1673
+    },
+    {
+      "epoch": 1.3232907885623826,
+      "grad_norm": 0.2415427081507558,
+      "learning_rate": 1.186368626279487e-06,
+      "loss": 0.5496,
+      "step": 1674
+    },
+    {
+      "epoch": 1.3240823191847235,
+      "grad_norm": 0.24239180033556626,
+      "learning_rate": 1.1855517978540234e-06,
+      "loss": 0.5612,
+      "step": 1675
+    },
+    {
+      "epoch": 1.3248738498070645,
+      "grad_norm": 0.24195732545796123,
+      "learning_rate": 1.1847348411922235e-06,
+      "loss": 0.5473,
+      "step": 1676
+    },
+    {
+      "epoch": 1.3256653804294054,
+      "grad_norm": 0.26194573198725696,
+      "learning_rate": 1.1839177568586918e-06,
+      "loss": 0.5614,
+      "step": 1677
+    },
+    {
+      "epoch": 1.3264569110517463,
+      "grad_norm": 0.24902746243757803,
+      "learning_rate": 1.1831005454181223e-06,
+      "loss": 0.5561,
+      "step": 1678
+    },
+    {
+      "epoch": 1.3272484416740873,
+      "grad_norm": 0.24489567020624506,
+      "learning_rate": 1.1822832074352962e-06,
+      "loss": 0.5507,
+      "step": 1679
+    },
+    {
+      "epoch": 1.3280399722964282,
+      "grad_norm": 0.25648305921788134,
+      "learning_rate": 1.1814657434750825e-06,
+      "loss": 0.5772,
+      "step": 1680
+    },
+    {
+      "epoch": 1.3288315029187692,
+      "grad_norm": 0.24089216814733655,
+      "learning_rate": 1.1806481541024371e-06,
+      "loss": 0.5544,
+      "step": 1681
+    },
+    {
+      "epoch": 1.32962303354111,
+      "grad_norm": 0.24541187842548162,
+      "learning_rate": 1.1798304398824024e-06,
+      "loss": 0.5487,
+      "step": 1682
+    },
+    {
+      "epoch": 1.330414564163451,
+      "grad_norm": 0.24598070389513868,
+      "learning_rate": 1.1790126013801074e-06,
+      "loss": 0.5645,
+      "step": 1683
+    },
+    {
+      "epoch": 1.331206094785792,
+      "grad_norm": 0.24629092938401864,
+      "learning_rate": 1.1781946391607672e-06,
+      "loss": 0.5656,
+      "step": 1684
+    },
+    {
+      "epoch": 1.331997625408133,
+      "grad_norm": 0.24923038130366343,
+      "learning_rate": 1.177376553789681e-06,
+      "loss": 0.5606,
+      "step": 1685
+    },
+    {
+      "epoch": 1.3327891560304739,
+      "grad_norm": 0.24311655903151758,
+      "learning_rate": 1.1765583458322354e-06,
+      "loss": 0.5614,
+      "step": 1686
+    },
+    {
+      "epoch": 1.3335806866528148,
+      "grad_norm": 0.24387615776609642,
+      "learning_rate": 1.1757400158538997e-06,
+      "loss": 0.5603,
+      "step": 1687
+    },
+    {
+      "epoch": 1.3343722172751558,
+      "grad_norm": 0.25462270378559737,
+      "learning_rate": 1.1749215644202286e-06,
+      "loss": 0.5749,
+      "step": 1688
+    },
+    {
+      "epoch": 1.335163747897497,
+      "grad_norm": 0.24532158932525416,
+      "learning_rate": 1.1741029920968605e-06,
+      "loss": 0.5583,
+      "step": 1689
+    },
+    {
+      "epoch": 1.3359552785198376,
+      "grad_norm": 0.24808814626487177,
+      "learning_rate": 1.1732842994495172e-06,
+      "loss": 0.5477,
+      "step": 1690
+    },
+    {
+      "epoch": 1.3367468091421788,
+      "grad_norm": 0.26179824952528297,
+      "learning_rate": 1.1724654870440036e-06,
+      "loss": 0.5657,
+      "step": 1691
+    },
+    {
+      "epoch": 1.3375383397645195,
+      "grad_norm": 0.23685666923300686,
+      "learning_rate": 1.171646555446208e-06,
+      "loss": 0.549,
+      "step": 1692
+    },
+    {
+      "epoch": 1.3383298703868607,
+      "grad_norm": 0.2514157285787257,
+      "learning_rate": 1.1708275052221e-06,
+      "loss": 0.558,
+      "step": 1693
+    },
+    {
+      "epoch": 1.3391214010092014,
+      "grad_norm": 0.2435334830438849,
+      "learning_rate": 1.1700083369377326e-06,
+      "loss": 0.56,
+      "step": 1694
+    },
+    {
+      "epoch": 1.3399129316315426,
+      "grad_norm": 0.24647470455772713,
+      "learning_rate": 1.1691890511592383e-06,
+      "loss": 0.5529,
+      "step": 1695
+    },
+    {
+      "epoch": 1.3407044622538835,
+      "grad_norm": 0.257918680442798,
+      "learning_rate": 1.1683696484528335e-06,
+      "loss": 0.5645,
+      "step": 1696
+    },
+    {
+      "epoch": 1.3414959928762245,
+      "grad_norm": 0.24001912836512246,
+      "learning_rate": 1.167550129384813e-06,
+      "loss": 0.569,
+      "step": 1697
+    },
+    {
+      "epoch": 1.3422875234985654,
+      "grad_norm": 0.25045209939003,
+      "learning_rate": 1.1667304945215534e-06,
+      "loss": 0.5629,
+      "step": 1698
+    },
+    {
+      "epoch": 1.3430790541209063,
+      "grad_norm": 0.24377854341294275,
+      "learning_rate": 1.1659107444295106e-06,
+      "loss": 0.5714,
+      "step": 1699
+    },
+    {
+      "epoch": 1.3438705847432473,
+      "grad_norm": 0.25114587450270015,
+      "learning_rate": 1.1650908796752213e-06,
+      "loss": 0.5643,
+      "step": 1700
+    },
+    {
+      "epoch": 1.3446621153655882,
+      "grad_norm": 0.23726228094488144,
+      "learning_rate": 1.1642709008252995e-06,
+      "loss": 0.5572,
+      "step": 1701
+    },
+    {
+      "epoch": 1.3454536459879292,
+      "grad_norm": 0.2570557148889931,
+      "learning_rate": 1.1634508084464402e-06,
+      "loss": 0.5536,
+      "step": 1702
+    },
+    {
+      "epoch": 1.34624517661027,
+      "grad_norm": 0.24462330596739756,
+      "learning_rate": 1.162630603105415e-06,
+      "loss": 0.5702,
+      "step": 1703
+    },
+    {
+      "epoch": 1.347036707232611,
+      "grad_norm": 0.24375145390909486,
+      "learning_rate": 1.1618102853690747e-06,
+      "loss": 0.5644,
+      "step": 1704
+    },
+    {
+      "epoch": 1.347828237854952,
+      "grad_norm": 0.24861484106318527,
+      "learning_rate": 1.1609898558043477e-06,
+      "loss": 0.5558,
+      "step": 1705
+    },
+    {
+      "epoch": 1.348619768477293,
+      "grad_norm": 0.24589755830104046,
+      "learning_rate": 1.1601693149782392e-06,
+      "loss": 0.5548,
+      "step": 1706
+    },
+    {
+      "epoch": 1.3494112990996339,
+      "grad_norm": 0.2405777155845157,
+      "learning_rate": 1.1593486634578316e-06,
+      "loss": 0.5659,
+      "step": 1707
+    },
+    {
+      "epoch": 1.3502028297219748,
+      "grad_norm": 0.2525715780086026,
+      "learning_rate": 1.1585279018102837e-06,
+      "loss": 0.5526,
+      "step": 1708
+    },
+    {
+      "epoch": 1.3509943603443157,
+      "grad_norm": 0.2482565319576305,
+      "learning_rate": 1.1577070306028302e-06,
+      "loss": 0.5443,
+      "step": 1709
+    },
+    {
+      "epoch": 1.351785890966657,
+      "grad_norm": 0.2447384221840171,
+      "learning_rate": 1.1568860504027827e-06,
+      "loss": 0.5533,
+      "step": 1710
+    },
+    {
+      "epoch": 1.3525774215889976,
+      "grad_norm": 0.24922245484056646,
+      "learning_rate": 1.1560649617775262e-06,
+      "loss": 0.5671,
+      "step": 1711
+    },
+    {
+      "epoch": 1.3533689522113388,
+      "grad_norm": 0.24148496427549768,
+      "learning_rate": 1.1552437652945223e-06,
+      "loss": 0.5564,
+      "step": 1712
+    },
+    {
+      "epoch": 1.3541604828336795,
+      "grad_norm": 0.2406317622445464,
+      "learning_rate": 1.154422461521306e-06,
+      "loss": 0.5392,
+      "step": 1713
+    },
+    {
+      "epoch": 1.3549520134560207,
+      "grad_norm": 0.244910794346466,
+      "learning_rate": 1.1536010510254872e-06,
+      "loss": 0.5642,
+      "step": 1714
+    },
+    {
+      "epoch": 1.3557435440783614,
+      "grad_norm": 0.24531774867589026,
+      "learning_rate": 1.1527795343747493e-06,
+      "loss": 0.5609,
+      "step": 1715
+    },
+    {
+      "epoch": 1.3565350747007026,
+      "grad_norm": 0.24205552864277713,
+      "learning_rate": 1.151957912136849e-06,
+      "loss": 0.5609,
+      "step": 1716
+    },
+    {
+      "epoch": 1.3573266053230435,
+      "grad_norm": 0.245561835662967,
+      "learning_rate": 1.151136184879616e-06,
+      "loss": 0.5704,
+      "step": 1717
+    },
+    {
+      "epoch": 1.3581181359453844,
+      "grad_norm": 0.23983199229942967,
+      "learning_rate": 1.1503143531709527e-06,
+      "loss": 0.5676,
+      "step": 1718
+    },
+    {
+      "epoch": 1.3589096665677254,
+      "grad_norm": 0.24191950280705846,
+      "learning_rate": 1.1494924175788332e-06,
+      "loss": 0.5805,
+      "step": 1719
+    },
+    {
+      "epoch": 1.3597011971900663,
+      "grad_norm": 0.24827783815317256,
+      "learning_rate": 1.1486703786713043e-06,
+      "loss": 0.5605,
+      "step": 1720
+    },
+    {
+      "epoch": 1.3604927278124073,
+      "grad_norm": 0.24507837042657588,
+      "learning_rate": 1.1478482370164837e-06,
+      "loss": 0.5518,
+      "step": 1721
+    },
+    {
+      "epoch": 1.3612842584347482,
+      "grad_norm": 0.24947413634372814,
+      "learning_rate": 1.1470259931825595e-06,
+      "loss": 0.5432,
+      "step": 1722
+    },
+    {
+      "epoch": 1.3620757890570891,
+      "grad_norm": 0.2426407520934941,
+      "learning_rate": 1.1462036477377917e-06,
+      "loss": 0.558,
+      "step": 1723
+    },
+    {
+      "epoch": 1.36286731967943,
+      "grad_norm": 0.2412205789291043,
+      "learning_rate": 1.1453812012505091e-06,
+      "loss": 0.566,
+      "step": 1724
+    },
+    {
+      "epoch": 1.363658850301771,
+      "grad_norm": 0.24108633129742646,
+      "learning_rate": 1.144558654289112e-06,
+      "loss": 0.5559,
+      "step": 1725
+    },
+    {
+      "epoch": 1.364450380924112,
+      "grad_norm": 0.23714632161758006,
+      "learning_rate": 1.1437360074220683e-06,
+      "loss": 0.5488,
+      "step": 1726
+    },
+    {
+      "epoch": 1.365241911546453,
+      "grad_norm": 0.24382604336059036,
+      "learning_rate": 1.1429132612179164e-06,
+      "loss": 0.5615,
+      "step": 1727
+    },
+    {
+      "epoch": 1.3660334421687939,
+      "grad_norm": 0.23894135962205929,
+      "learning_rate": 1.1420904162452626e-06,
+      "loss": 0.5655,
+      "step": 1728
+    },
+    {
+      "epoch": 1.3668249727911348,
+      "grad_norm": 0.24385031573903024,
+      "learning_rate": 1.1412674730727818e-06,
+      "loss": 0.5459,
+      "step": 1729
+    },
+    {
+      "epoch": 1.3676165034134757,
+      "grad_norm": 0.24695290459092584,
+      "learning_rate": 1.1404444322692167e-06,
+      "loss": 0.5648,
+      "step": 1730
+    },
+    {
+      "epoch": 1.3684080340358167,
+      "grad_norm": 0.24493741494489846,
+      "learning_rate": 1.139621294403377e-06,
+      "loss": 0.5676,
+      "step": 1731
+    },
+    {
+      "epoch": 1.3691995646581576,
+      "grad_norm": 0.24886579934473138,
+      "learning_rate": 1.1387980600441403e-06,
+      "loss": 0.5616,
+      "step": 1732
+    },
+    {
+      "epoch": 1.3699910952804988,
+      "grad_norm": 0.2475984162720158,
+      "learning_rate": 1.1379747297604502e-06,
+      "loss": 0.5557,
+      "step": 1733
+    },
+    {
+      "epoch": 1.3707826259028395,
+      "grad_norm": 0.24074846356777077,
+      "learning_rate": 1.1371513041213168e-06,
+      "loss": 0.5468,
+      "step": 1734
+    },
+    {
+      "epoch": 1.3715741565251807,
+      "grad_norm": 0.2481390305891257,
+      "learning_rate": 1.1363277836958167e-06,
+      "loss": 0.5719,
+      "step": 1735
+    },
+    {
+      "epoch": 1.3723656871475214,
+      "grad_norm": 0.24410534282618412,
+      "learning_rate": 1.1355041690530911e-06,
+      "loss": 0.5489,
+      "step": 1736
+    },
+    {
+      "epoch": 1.3731572177698625,
+      "grad_norm": 0.251570922054409,
+      "learning_rate": 1.1346804607623463e-06,
+      "loss": 0.5585,
+      "step": 1737
+    },
+    {
+      "epoch": 1.3739487483922035,
+      "grad_norm": 0.24773556027804694,
+      "learning_rate": 1.1338566593928544e-06,
+      "loss": 0.5665,
+      "step": 1738
+    },
+    {
+      "epoch": 1.3747402790145444,
+      "grad_norm": 0.2523851567742354,
+      "learning_rate": 1.1330327655139512e-06,
+      "loss": 0.5585,
+      "step": 1739
+    },
+    {
+      "epoch": 1.3755318096368854,
+      "grad_norm": 0.24722026729991606,
+      "learning_rate": 1.1322087796950358e-06,
+      "loss": 0.5665,
+      "step": 1740
+    },
+    {
+      "epoch": 1.3763233402592263,
+      "grad_norm": 0.24359214618999236,
+      "learning_rate": 1.1313847025055716e-06,
+      "loss": 0.539,
+      "step": 1741
+    },
+    {
+      "epoch": 1.3771148708815673,
+      "grad_norm": 0.24823250980538344,
+      "learning_rate": 1.1305605345150848e-06,
+      "loss": 0.5679,
+      "step": 1742
+    },
+    {
+      "epoch": 1.3779064015039082,
+      "grad_norm": 0.24950308412089048,
+      "learning_rate": 1.1297362762931647e-06,
+      "loss": 0.5513,
+      "step": 1743
+    },
+    {
+      "epoch": 1.3786979321262491,
+      "grad_norm": 0.24984903403329928,
+      "learning_rate": 1.1289119284094633e-06,
+      "loss": 0.5708,
+      "step": 1744
+    },
+    {
+      "epoch": 1.37948946274859,
+      "grad_norm": 0.2540393327158127,
+      "learning_rate": 1.128087491433693e-06,
+      "loss": 0.5502,
+      "step": 1745
+    },
+    {
+      "epoch": 1.380280993370931,
+      "grad_norm": 0.2397092671429131,
+      "learning_rate": 1.1272629659356296e-06,
+      "loss": 0.5514,
+      "step": 1746
+    },
+    {
+      "epoch": 1.381072523993272,
+      "grad_norm": 0.2545424160596748,
+      "learning_rate": 1.126438352485109e-06,
+      "loss": 0.5594,
+      "step": 1747
+    },
+    {
+      "epoch": 1.381864054615613,
+      "grad_norm": 0.25389933261603476,
+      "learning_rate": 1.1256136516520283e-06,
+      "loss": 0.553,
+      "step": 1748
+    },
+    {
+      "epoch": 1.3826555852379538,
+      "grad_norm": 0.25595186731390124,
+      "learning_rate": 1.124788864006345e-06,
+      "loss": 0.5545,
+      "step": 1749
+    },
+    {
+      "epoch": 1.3834471158602948,
+      "grad_norm": 0.2538215322222843,
+      "learning_rate": 1.1239639901180757e-06,
+      "loss": 0.552,
+      "step": 1750
+    },
+    {
+      "epoch": 1.3842386464826357,
+      "grad_norm": 0.2434561664825095,
+      "learning_rate": 1.1231390305572981e-06,
+      "loss": 0.5571,
+      "step": 1751
+    },
+    {
+      "epoch": 1.3850301771049767,
+      "grad_norm": 0.2515826753187997,
+      "learning_rate": 1.1223139858941481e-06,
+      "loss": 0.5621,
+      "step": 1752
+    },
+    {
+      "epoch": 1.3858217077273176,
+      "grad_norm": 0.24304199261022497,
+      "learning_rate": 1.121488856698821e-06,
+      "loss": 0.5636,
+      "step": 1753
+    },
+    {
+      "epoch": 1.3866132383496588,
+      "grad_norm": 0.249221543401245,
+      "learning_rate": 1.1206636435415696e-06,
+      "loss": 0.5621,
+      "step": 1754
+    },
+    {
+      "epoch": 1.3874047689719995,
+      "grad_norm": 0.25381314826168094,
+      "learning_rate": 1.1198383469927061e-06,
+      "loss": 0.5834,
+      "step": 1755
+    },
+    {
+      "epoch": 1.3881962995943407,
+      "grad_norm": 0.2461090617556699,
+      "learning_rate": 1.119012967622599e-06,
+      "loss": 0.5696,
+      "step": 1756
+    },
+    {
+      "epoch": 1.3889878302166814,
+      "grad_norm": 0.2522213466470216,
+      "learning_rate": 1.118187506001675e-06,
+      "loss": 0.5501,
+      "step": 1757
+    },
+    {
+      "epoch": 1.3897793608390225,
+      "grad_norm": 0.24781293816561925,
+      "learning_rate": 1.1173619627004168e-06,
+      "loss": 0.5707,
+      "step": 1758
+    },
+    {
+      "epoch": 1.3905708914613635,
+      "grad_norm": 0.24020257431457925,
+      "learning_rate": 1.1165363382893642e-06,
+      "loss": 0.5676,
+      "step": 1759
+    },
+    {
+      "epoch": 1.3913624220837044,
+      "grad_norm": 0.2501915603787139,
+      "learning_rate": 1.1157106333391131e-06,
+      "loss": 0.5584,
+      "step": 1760
+    },
+    {
+      "epoch": 1.3921539527060454,
+      "grad_norm": 0.2473703246046045,
+      "learning_rate": 1.1148848484203143e-06,
+      "loss": 0.5645,
+      "step": 1761
+    },
+    {
+      "epoch": 1.3929454833283863,
+      "grad_norm": 0.23890245427011528,
+      "learning_rate": 1.1140589841036749e-06,
+      "loss": 0.5673,
+      "step": 1762
+    },
+    {
+      "epoch": 1.3937370139507272,
+      "grad_norm": 0.23776869694051625,
+      "learning_rate": 1.1132330409599562e-06,
+      "loss": 0.574,
+      "step": 1763
+    },
+    {
+      "epoch": 1.3945285445730682,
+      "grad_norm": 0.23519164494572908,
+      "learning_rate": 1.1124070195599738e-06,
+      "loss": 0.5553,
+      "step": 1764
+    },
+    {
+      "epoch": 1.3953200751954091,
+      "grad_norm": 0.23840089959621127,
+      "learning_rate": 1.1115809204745982e-06,
+      "loss": 0.5665,
+      "step": 1765
+    },
+    {
+      "epoch": 1.39611160581775,
+      "grad_norm": 0.25152689581922855,
+      "learning_rate": 1.1107547442747529e-06,
+      "loss": 0.5474,
+      "step": 1766
+    },
+    {
+      "epoch": 1.396903136440091,
+      "grad_norm": 0.2470266862599932,
+      "learning_rate": 1.1099284915314152e-06,
+      "loss": 0.5752,
+      "step": 1767
+    },
+    {
+      "epoch": 1.397694667062432,
+      "grad_norm": 0.2412530565253777,
+      "learning_rate": 1.1091021628156143e-06,
+      "loss": 0.5498,
+      "step": 1768
+    },
+    {
+      "epoch": 1.398486197684773,
+      "grad_norm": 0.23721111685798252,
+      "learning_rate": 1.108275758698433e-06,
+      "loss": 0.5464,
+      "step": 1769
+    },
+    {
+      "epoch": 1.3992777283071138,
+      "grad_norm": 0.24351562526834195,
+      "learning_rate": 1.1074492797510062e-06,
+      "loss": 0.5564,
+      "step": 1770
+    },
+    {
+      "epoch": 1.4000692589294548,
+      "grad_norm": 0.25588642956275404,
+      "learning_rate": 1.1066227265445196e-06,
+      "loss": 0.5474,
+      "step": 1771
+    },
+    {
+      "epoch": 1.4008607895517957,
+      "grad_norm": 0.24325585961125706,
+      "learning_rate": 1.1057960996502109e-06,
+      "loss": 0.5559,
+      "step": 1772
+    },
+    {
+      "epoch": 1.4016523201741367,
+      "grad_norm": 0.24380364989827122,
+      "learning_rate": 1.1049693996393687e-06,
+      "loss": 0.5648,
+      "step": 1773
+    },
+    {
+      "epoch": 1.4024438507964776,
+      "grad_norm": 0.23597633461039252,
+      "learning_rate": 1.1041426270833318e-06,
+      "loss": 0.5673,
+      "step": 1774
+    },
+    {
+      "epoch": 1.4032353814188188,
+      "grad_norm": 0.24278524172378807,
+      "learning_rate": 1.1033157825534896e-06,
+      "loss": 0.551,
+      "step": 1775
+    },
+    {
+      "epoch": 1.4040269120411595,
+      "grad_norm": 0.25289497514364,
+      "learning_rate": 1.1024888666212808e-06,
+      "loss": 0.5515,
+      "step": 1776
+    },
+    {
+      "epoch": 1.4048184426635006,
+      "grad_norm": 0.2532208606043583,
+      "learning_rate": 1.1016618798581937e-06,
+      "loss": 0.5692,
+      "step": 1777
+    },
+    {
+      "epoch": 1.4056099732858414,
+      "grad_norm": 0.25692567593300203,
+      "learning_rate": 1.1008348228357657e-06,
+      "loss": 0.5424,
+      "step": 1778
+    },
+    {
+      "epoch": 1.4064015039081825,
+      "grad_norm": 0.23890317619691112,
+      "learning_rate": 1.1000076961255822e-06,
+      "loss": 0.559,
+      "step": 1779
+    },
+    {
+      "epoch": 1.4071930345305235,
+      "grad_norm": 0.2485982325211479,
+      "learning_rate": 1.099180500299277e-06,
+      "loss": 0.559,
+      "step": 1780
+    },
+    {
+      "epoch": 1.4079845651528644,
+      "grad_norm": 0.25210770277351785,
+      "learning_rate": 1.0983532359285324e-06,
+      "loss": 0.5713,
+      "step": 1781
+    },
+    {
+      "epoch": 1.4087760957752054,
+      "grad_norm": 0.2553458928622144,
+      "learning_rate": 1.097525903585077e-06,
+      "loss": 0.5756,
+      "step": 1782
+    },
+    {
+      "epoch": 1.4095676263975463,
+      "grad_norm": 0.2511017029175708,
+      "learning_rate": 1.0966985038406868e-06,
+      "loss": 0.5518,
+      "step": 1783
+    },
+    {
+      "epoch": 1.4103591570198872,
+      "grad_norm": 0.2475404708066992,
+      "learning_rate": 1.0958710372671845e-06,
+      "loss": 0.5637,
+      "step": 1784
+    },
+    {
+      "epoch": 1.4111506876422282,
+      "grad_norm": 0.23797302343362914,
+      "learning_rate": 1.0950435044364388e-06,
+      "loss": 0.5518,
+      "step": 1785
+    },
+    {
+      "epoch": 1.4119422182645691,
+      "grad_norm": 0.24656859756288255,
+      "learning_rate": 1.0942159059203644e-06,
+      "loss": 0.5431,
+      "step": 1786
+    },
+    {
+      "epoch": 1.41273374888691,
+      "grad_norm": 0.24107739978828868,
+      "learning_rate": 1.0933882422909216e-06,
+      "loss": 0.5567,
+      "step": 1787
+    },
+    {
+      "epoch": 1.413525279509251,
+      "grad_norm": 0.2356718550731877,
+      "learning_rate": 1.0925605141201144e-06,
+      "loss": 0.556,
+      "step": 1788
+    },
+    {
+      "epoch": 1.414316810131592,
+      "grad_norm": 0.24175815505353185,
+      "learning_rate": 1.0917327219799935e-06,
+      "loss": 0.5672,
+      "step": 1789
+    },
+    {
+      "epoch": 1.4151083407539329,
+      "grad_norm": 0.259273210045425,
+      "learning_rate": 1.0909048664426519e-06,
+      "loss": 0.5474,
+      "step": 1790
+    },
+    {
+      "epoch": 1.4158998713762738,
+      "grad_norm": 0.2452236914724908,
+      "learning_rate": 1.0900769480802278e-06,
+      "loss": 0.5657,
+      "step": 1791
+    },
+    {
+      "epoch": 1.4166914019986148,
+      "grad_norm": 0.24380905191208496,
+      "learning_rate": 1.0892489674649016e-06,
+      "loss": 0.5725,
+      "step": 1792
+    },
+    {
+      "epoch": 1.4174829326209557,
+      "grad_norm": 0.2509559233325704,
+      "learning_rate": 1.0884209251688978e-06,
+      "loss": 0.5376,
+      "step": 1793
+    },
+    {
+      "epoch": 1.4182744632432966,
+      "grad_norm": 0.24322450352928576,
+      "learning_rate": 1.0875928217644829e-06,
+      "loss": 0.5681,
+      "step": 1794
+    },
+    {
+      "epoch": 1.4190659938656376,
+      "grad_norm": 0.2504916663222412,
+      "learning_rate": 1.086764657823966e-06,
+      "loss": 0.5512,
+      "step": 1795
+    },
+    {
+      "epoch": 1.4198575244879788,
+      "grad_norm": 0.267150023425375,
+      "learning_rate": 1.0859364339196978e-06,
+      "loss": 0.5625,
+      "step": 1796
+    },
+    {
+      "epoch": 1.4206490551103195,
+      "grad_norm": 0.24717884320440905,
+      "learning_rate": 1.0851081506240703e-06,
+      "loss": 0.5569,
+      "step": 1797
+    },
+    {
+      "epoch": 1.4214405857326606,
+      "grad_norm": 0.2446443226487652,
+      "learning_rate": 1.0842798085095165e-06,
+      "loss": 0.5647,
+      "step": 1798
+    },
+    {
+      "epoch": 1.4222321163550014,
+      "grad_norm": 0.25692581921323493,
+      "learning_rate": 1.0834514081485111e-06,
+      "loss": 0.5653,
+      "step": 1799
+    },
+    {
+      "epoch": 1.4230236469773425,
+      "grad_norm": 0.2439742451677067,
+      "learning_rate": 1.0826229501135675e-06,
+      "loss": 0.5434,
+      "step": 1800
+    },
+    {
+      "epoch": 1.4238151775996835,
+      "grad_norm": 0.25058440196747644,
+      "learning_rate": 1.0817944349772398e-06,
+      "loss": 0.5554,
+      "step": 1801
+    },
+    {
+      "epoch": 1.4246067082220244,
+      "grad_norm": 0.25052414894956365,
+      "learning_rate": 1.0809658633121222e-06,
+      "loss": 0.5813,
+      "step": 1802
+    },
+    {
+      "epoch": 1.4253982388443653,
+      "grad_norm": 0.2551531754057182,
+      "learning_rate": 1.0801372356908461e-06,
+      "loss": 0.5689,
+      "step": 1803
+    },
+    {
+      "epoch": 1.4261897694667063,
+      "grad_norm": 0.24395819104907066,
+      "learning_rate": 1.0793085526860832e-06,
+      "loss": 0.55,
+      "step": 1804
+    },
+    {
+      "epoch": 1.4269813000890472,
+      "grad_norm": 0.253654833757268,
+      "learning_rate": 1.0784798148705432e-06,
+      "loss": 0.5646,
+      "step": 1805
+    },
+    {
+      "epoch": 1.4277728307113882,
+      "grad_norm": 0.2472594700686716,
+      "learning_rate": 1.0776510228169733e-06,
+      "loss": 0.5604,
+      "step": 1806
+    },
+    {
+      "epoch": 1.428564361333729,
+      "grad_norm": 0.2570127725543254,
+      "learning_rate": 1.0768221770981583e-06,
+      "loss": 0.5665,
+      "step": 1807
+    },
+    {
+      "epoch": 1.42935589195607,
+      "grad_norm": 0.25219949596877267,
+      "learning_rate": 1.0759932782869198e-06,
+      "loss": 0.5668,
+      "step": 1808
+    },
+    {
+      "epoch": 1.430147422578411,
+      "grad_norm": 0.24888935758981642,
+      "learning_rate": 1.075164326956117e-06,
+      "loss": 0.565,
+      "step": 1809
+    },
+    {
+      "epoch": 1.430938953200752,
+      "grad_norm": 0.2508456731183149,
+      "learning_rate": 1.0743353236786449e-06,
+      "loss": 0.5429,
+      "step": 1810
+    },
+    {
+      "epoch": 1.4317304838230929,
+      "grad_norm": 0.25010152903566374,
+      "learning_rate": 1.0735062690274337e-06,
+      "loss": 0.5517,
+      "step": 1811
+    },
+    {
+      "epoch": 1.4325220144454338,
+      "grad_norm": 0.2623661368348324,
+      "learning_rate": 1.0726771635754502e-06,
+      "loss": 0.5554,
+      "step": 1812
+    },
+    {
+      "epoch": 1.4333135450677748,
+      "grad_norm": 0.26566701270731186,
+      "learning_rate": 1.0718480078956952e-06,
+      "loss": 0.5676,
+      "step": 1813
+    },
+    {
+      "epoch": 1.4341050756901157,
+      "grad_norm": 0.2582041514202188,
+      "learning_rate": 1.0710188025612055e-06,
+      "loss": 0.5643,
+      "step": 1814
+    },
+    {
+      "epoch": 1.4348966063124566,
+      "grad_norm": 0.24095795169398565,
+      "learning_rate": 1.0701895481450516e-06,
+      "loss": 0.577,
+      "step": 1815
+    },
+    {
+      "epoch": 1.4356881369347976,
+      "grad_norm": 0.24552980144547748,
+      "learning_rate": 1.0693602452203374e-06,
+      "loss": 0.5434,
+      "step": 1816
+    },
+    {
+      "epoch": 1.4364796675571387,
+      "grad_norm": 0.24802734115308014,
+      "learning_rate": 1.0685308943602013e-06,
+      "loss": 0.5495,
+      "step": 1817
+    },
+    {
+      "epoch": 1.4372711981794795,
+      "grad_norm": 0.24063232282451458,
+      "learning_rate": 1.0677014961378136e-06,
+      "loss": 0.5643,
+      "step": 1818
+    },
+    {
+      "epoch": 1.4380627288018206,
+      "grad_norm": 0.26674826901817056,
+      "learning_rate": 1.0668720511263786e-06,
+      "loss": 0.5563,
+      "step": 1819
+    },
+    {
+      "epoch": 1.4388542594241613,
+      "grad_norm": 0.2519268868939359,
+      "learning_rate": 1.0660425598991333e-06,
+      "loss": 0.5708,
+      "step": 1820
+    },
+    {
+      "epoch": 1.4396457900465025,
+      "grad_norm": 0.25537139429719885,
+      "learning_rate": 1.0652130230293437e-06,
+      "loss": 0.5528,
+      "step": 1821
+    },
+    {
+      "epoch": 1.4404373206688434,
+      "grad_norm": 0.24245967377929403,
+      "learning_rate": 1.0643834410903108e-06,
+      "loss": 0.5683,
+      "step": 1822
+    },
+    {
+      "epoch": 1.4412288512911844,
+      "grad_norm": 0.24700653920398574,
+      "learning_rate": 1.063553814655365e-06,
+      "loss": 0.5645,
+      "step": 1823
+    },
+    {
+      "epoch": 1.4420203819135253,
+      "grad_norm": 0.25569340348431774,
+      "learning_rate": 1.062724144297868e-06,
+      "loss": 0.5684,
+      "step": 1824
+    },
+    {
+      "epoch": 1.4428119125358663,
+      "grad_norm": 0.2442426148762711,
+      "learning_rate": 1.0618944305912117e-06,
+      "loss": 0.5611,
+      "step": 1825
+    },
+    {
+      "epoch": 1.4436034431582072,
+      "grad_norm": 0.24668468930856702,
+      "learning_rate": 1.0610646741088171e-06,
+      "loss": 0.5542,
+      "step": 1826
+    },
+    {
+      "epoch": 1.4443949737805482,
+      "grad_norm": 0.24180925399348951,
+      "learning_rate": 1.060234875424136e-06,
+      "loss": 0.5496,
+      "step": 1827
+    },
+    {
+      "epoch": 1.445186504402889,
+      "grad_norm": 0.24548292791451223,
+      "learning_rate": 1.0594050351106495e-06,
+      "loss": 0.5429,
+      "step": 1828
+    },
+    {
+      "epoch": 1.44597803502523,
+      "grad_norm": 0.2457166387439888,
+      "learning_rate": 1.0585751537418665e-06,
+      "loss": 0.5571,
+      "step": 1829
+    },
+    {
+      "epoch": 1.446769565647571,
+      "grad_norm": 0.2517923689801228,
+      "learning_rate": 1.0577452318913244e-06,
+      "loss": 0.5599,
+      "step": 1830
+    },
+    {
+      "epoch": 1.447561096269912,
+      "grad_norm": 0.2434257429077564,
+      "learning_rate": 1.0569152701325891e-06,
+      "loss": 0.5687,
+      "step": 1831
+    },
+    {
+      "epoch": 1.4483526268922529,
+      "grad_norm": 0.2536063863905166,
+      "learning_rate": 1.0560852690392538e-06,
+      "loss": 0.569,
+      "step": 1832
+    },
+    {
+      "epoch": 1.4491441575145938,
+      "grad_norm": 0.24997586186843135,
+      "learning_rate": 1.055255229184939e-06,
+      "loss": 0.5678,
+      "step": 1833
+    },
+    {
+      "epoch": 1.4499356881369347,
+      "grad_norm": 0.2394202668192573,
+      "learning_rate": 1.054425151143292e-06,
+      "loss": 0.5684,
+      "step": 1834
+    },
+    {
+      "epoch": 1.4507272187592757,
+      "grad_norm": 0.24622681685601036,
+      "learning_rate": 1.053595035487986e-06,
+      "loss": 0.56,
+      "step": 1835
+    },
+    {
+      "epoch": 1.4515187493816166,
+      "grad_norm": 0.24766022547943686,
+      "learning_rate": 1.052764882792721e-06,
+      "loss": 0.5507,
+      "step": 1836
+    },
+    {
+      "epoch": 1.4523102800039576,
+      "grad_norm": 0.24766891264630292,
+      "learning_rate": 1.0519346936312217e-06,
+      "loss": 0.5534,
+      "step": 1837
+    },
+    {
+      "epoch": 1.4531018106262987,
+      "grad_norm": 0.24371750012654173,
+      "learning_rate": 1.051104468577239e-06,
+      "loss": 0.5603,
+      "step": 1838
+    },
+    {
+      "epoch": 1.4538933412486394,
+      "grad_norm": 0.24773989442809743,
+      "learning_rate": 1.0502742082045478e-06,
+      "loss": 0.5482,
+      "step": 1839
+    },
+    {
+      "epoch": 1.4546848718709806,
+      "grad_norm": 0.24593443044170213,
+      "learning_rate": 1.049443913086948e-06,
+      "loss": 0.5613,
+      "step": 1840
+    },
+    {
+      "epoch": 1.4554764024933213,
+      "grad_norm": 0.2520294939428411,
+      "learning_rate": 1.048613583798263e-06,
+      "loss": 0.55,
+      "step": 1841
+    },
+    {
+      "epoch": 1.4562679331156625,
+      "grad_norm": 0.2568090128922785,
+      "learning_rate": 1.0477832209123399e-06,
+      "loss": 0.562,
+      "step": 1842
+    },
+    {
+      "epoch": 1.4570594637380034,
+      "grad_norm": 0.24861420329563133,
+      "learning_rate": 1.0469528250030495e-06,
+      "loss": 0.5554,
+      "step": 1843
+    },
+    {
+      "epoch": 1.4578509943603444,
+      "grad_norm": 0.2397636208883377,
+      "learning_rate": 1.046122396644285e-06,
+      "loss": 0.5527,
+      "step": 1844
+    },
+    {
+      "epoch": 1.4586425249826853,
+      "grad_norm": 0.25211289237402956,
+      "learning_rate": 1.0452919364099616e-06,
+      "loss": 0.5578,
+      "step": 1845
+    },
+    {
+      "epoch": 1.4594340556050263,
+      "grad_norm": 0.2622014777508393,
+      "learning_rate": 1.0444614448740177e-06,
+      "loss": 0.5712,
+      "step": 1846
+    },
+    {
+      "epoch": 1.4602255862273672,
+      "grad_norm": 0.24645617140580847,
+      "learning_rate": 1.0436309226104124e-06,
+      "loss": 0.5594,
+      "step": 1847
+    },
+    {
+      "epoch": 1.4610171168497081,
+      "grad_norm": 0.2459939996889011,
+      "learning_rate": 1.0428003701931263e-06,
+      "loss": 0.5311,
+      "step": 1848
+    },
+    {
+      "epoch": 1.461808647472049,
+      "grad_norm": 0.24216005432750698,
+      "learning_rate": 1.0419697881961606e-06,
+      "loss": 0.5449,
+      "step": 1849
+    },
+    {
+      "epoch": 1.46260017809439,
+      "grad_norm": 0.24265662602239907,
+      "learning_rate": 1.0411391771935377e-06,
+      "loss": 0.5483,
+      "step": 1850
+    },
+    {
+      "epoch": 1.463391708716731,
+      "grad_norm": 0.24575939677213582,
+      "learning_rate": 1.0403085377592991e-06,
+      "loss": 0.5509,
+      "step": 1851
+    },
+    {
+      "epoch": 1.464183239339072,
+      "grad_norm": 0.23687738850752296,
+      "learning_rate": 1.0394778704675066e-06,
+      "loss": 0.5555,
+      "step": 1852
+    },
+    {
+      "epoch": 1.4649747699614128,
+      "grad_norm": 0.2499892511365357,
+      "learning_rate": 1.038647175892241e-06,
+      "loss": 0.5551,
+      "step": 1853
+    },
+    {
+      "epoch": 1.4657663005837538,
+      "grad_norm": 0.24192706787427268,
+      "learning_rate": 1.0378164546076022e-06,
+      "loss": 0.5658,
+      "step": 1854
+    },
+    {
+      "epoch": 1.4665578312060947,
+      "grad_norm": 0.23184326927057436,
+      "learning_rate": 1.0369857071877077e-06,
+      "loss": 0.5574,
+      "step": 1855
+    },
+    {
+      "epoch": 1.4673493618284357,
+      "grad_norm": 0.2687981320629577,
+      "learning_rate": 1.0361549342066942e-06,
+      "loss": 0.5525,
+      "step": 1856
+    },
+    {
+      "epoch": 1.4681408924507766,
+      "grad_norm": 0.24080649133852142,
+      "learning_rate": 1.0353241362387156e-06,
+      "loss": 0.5472,
+      "step": 1857
+    },
+    {
+      "epoch": 1.4689324230731176,
+      "grad_norm": 0.24991388863098019,
+      "learning_rate": 1.034493313857943e-06,
+      "loss": 0.5484,
+      "step": 1858
+    },
+    {
+      "epoch": 1.4697239536954587,
+      "grad_norm": 0.2645689623760348,
+      "learning_rate": 1.0336624676385644e-06,
+      "loss": 0.5531,
+      "step": 1859
+    },
+    {
+      "epoch": 1.4705154843177994,
+      "grad_norm": 0.26137992808052546,
+      "learning_rate": 1.032831598154784e-06,
+      "loss": 0.5653,
+      "step": 1860
+    },
+    {
+      "epoch": 1.4713070149401406,
+      "grad_norm": 0.25109005763019854,
+      "learning_rate": 1.0320007059808226e-06,
+      "loss": 0.5462,
+      "step": 1861
+    },
+    {
+      "epoch": 1.4720985455624813,
+      "grad_norm": 0.2438954035418305,
+      "learning_rate": 1.0311697916909165e-06,
+      "loss": 0.5605,
+      "step": 1862
+    },
+    {
+      "epoch": 1.4728900761848225,
+      "grad_norm": 0.24324977710114248,
+      "learning_rate": 1.0303388558593176e-06,
+      "loss": 0.5494,
+      "step": 1863
+    },
+    {
+      "epoch": 1.4736816068071634,
+      "grad_norm": 0.24616857673174009,
+      "learning_rate": 1.0295078990602914e-06,
+      "loss": 0.555,
+      "step": 1864
+    },
+    {
+      "epoch": 1.4744731374295044,
+      "grad_norm": 0.24526355098388605,
+      "learning_rate": 1.0286769218681194e-06,
+      "loss": 0.5538,
+      "step": 1865
+    },
+    {
+      "epoch": 1.4752646680518453,
+      "grad_norm": 0.2522719853942459,
+      "learning_rate": 1.0278459248570968e-06,
+      "loss": 0.5632,
+      "step": 1866
+    },
+    {
+      "epoch": 1.4760561986741862,
+      "grad_norm": 0.24668814131099567,
+      "learning_rate": 1.027014908601532e-06,
+      "loss": 0.557,
+      "step": 1867
+    },
+    {
+      "epoch": 1.4768477292965272,
+      "grad_norm": 0.24958807685657602,
+      "learning_rate": 1.0261838736757469e-06,
+      "loss": 0.5638,
+      "step": 1868
+    },
+    {
+      "epoch": 1.4776392599188681,
+      "grad_norm": 0.2416674442464471,
+      "learning_rate": 1.0253528206540764e-06,
+      "loss": 0.5408,
+      "step": 1869
+    },
+    {
+      "epoch": 1.478430790541209,
+      "grad_norm": 0.24183631113589926,
+      "learning_rate": 1.0245217501108686e-06,
+      "loss": 0.5465,
+      "step": 1870
+    },
+    {
+      "epoch": 1.47922232116355,
+      "grad_norm": 0.24684608161971405,
+      "learning_rate": 1.0236906626204818e-06,
+      "loss": 0.5679,
+      "step": 1871
+    },
+    {
+      "epoch": 1.480013851785891,
+      "grad_norm": 0.2577969021987222,
+      "learning_rate": 1.0228595587572886e-06,
+      "loss": 0.5726,
+      "step": 1872
+    },
+    {
+      "epoch": 1.480805382408232,
+      "grad_norm": 0.24719018121256667,
+      "learning_rate": 1.0220284390956703e-06,
+      "loss": 0.5541,
+      "step": 1873
+    },
+    {
+      "epoch": 1.4815969130305728,
+      "grad_norm": 0.24327660265334178,
+      "learning_rate": 1.0211973042100212e-06,
+      "loss": 0.5593,
+      "step": 1874
+    },
+    {
+      "epoch": 1.4823884436529138,
+      "grad_norm": 0.24716476661970774,
+      "learning_rate": 1.020366154674745e-06,
+      "loss": 0.5582,
+      "step": 1875
+    },
+    {
+      "epoch": 1.4831799742752547,
+      "grad_norm": 0.24889099861675523,
+      "learning_rate": 1.019534991064256e-06,
+      "loss": 0.5617,
+      "step": 1876
+    },
+    {
+      "epoch": 1.4839715048975957,
+      "grad_norm": 0.24104354323552155,
+      "learning_rate": 1.0187038139529775e-06,
+      "loss": 0.54,
+      "step": 1877
+    },
+    {
+      "epoch": 1.4847630355199366,
+      "grad_norm": 0.25631507617820665,
+      "learning_rate": 1.0178726239153435e-06,
+      "loss": 0.5551,
+      "step": 1878
+    },
+    {
+      "epoch": 1.4855545661422775,
+      "grad_norm": 0.24265480781006588,
+      "learning_rate": 1.0170414215257955e-06,
+      "loss": 0.5538,
+      "step": 1879
+    },
+    {
+      "epoch": 1.4863460967646187,
+      "grad_norm": 0.2571920866849079,
+      "learning_rate": 1.016210207358784e-06,
+      "loss": 0.5477,
+      "step": 1880
+    },
+    {
+      "epoch": 1.4871376273869594,
+      "grad_norm": 0.25687643816628813,
+      "learning_rate": 1.0153789819887688e-06,
+      "loss": 0.5546,
+      "step": 1881
+    },
+    {
+      "epoch": 1.4879291580093006,
+      "grad_norm": 0.24634232950761223,
+      "learning_rate": 1.014547745990215e-06,
+      "loss": 0.5716,
+      "step": 1882
+    },
+    {
+      "epoch": 1.4887206886316413,
+      "grad_norm": 0.24512103236922167,
+      "learning_rate": 1.0137164999375977e-06,
+      "loss": 0.5591,
+      "step": 1883
+    },
+    {
+      "epoch": 1.4895122192539825,
+      "grad_norm": 0.24374321098015242,
+      "learning_rate": 1.0128852444053969e-06,
+      "loss": 0.5535,
+      "step": 1884
+    },
+    {
+      "epoch": 1.4903037498763234,
+      "grad_norm": 0.2524403358422412,
+      "learning_rate": 1.0120539799681003e-06,
+      "loss": 0.557,
+      "step": 1885
+    },
+    {
+      "epoch": 1.4910952804986644,
+      "grad_norm": 0.2517792905086232,
+      "learning_rate": 1.0112227072002016e-06,
+      "loss": 0.5492,
+      "step": 1886
+    },
+    {
+      "epoch": 1.4918868111210053,
+      "grad_norm": 0.24995874361292328,
+      "learning_rate": 1.0103914266761997e-06,
+      "loss": 0.5465,
+      "step": 1887
+    },
+    {
+      "epoch": 1.4926783417433462,
+      "grad_norm": 0.2415788919243182,
+      "learning_rate": 1.0095601389705995e-06,
+      "loss": 0.5537,
+      "step": 1888
+    },
+    {
+      "epoch": 1.4934698723656872,
+      "grad_norm": 0.252420574246598,
+      "learning_rate": 1.0087288446579104e-06,
+      "loss": 0.5525,
+      "step": 1889
+    },
+    {
+      "epoch": 1.4942614029880281,
+      "grad_norm": 0.26350997377549623,
+      "learning_rate": 1.0078975443126467e-06,
+      "loss": 0.5566,
+      "step": 1890
+    },
+    {
+      "epoch": 1.495052933610369,
+      "grad_norm": 0.25324136736305286,
+      "learning_rate": 1.0070662385093268e-06,
+      "loss": 0.5621,
+      "step": 1891
+    },
+    {
+      "epoch": 1.49584446423271,
+      "grad_norm": 0.253270804896081,
+      "learning_rate": 1.006234927822473e-06,
+      "loss": 0.5667,
+      "step": 1892
+    },
+    {
+      "epoch": 1.496635994855051,
+      "grad_norm": 0.24124282896723453,
+      "learning_rate": 1.0054036128266103e-06,
+      "loss": 0.5607,
+      "step": 1893
+    },
+    {
+      "epoch": 1.4974275254773919,
+      "grad_norm": 0.2448640921733728,
+      "learning_rate": 1.0045722940962674e-06,
+      "loss": 0.5442,
+      "step": 1894
+    },
+    {
+      "epoch": 1.4982190560997328,
+      "grad_norm": 0.2640041793280316,
+      "learning_rate": 1.0037409722059753e-06,
+      "loss": 0.5509,
+      "step": 1895
+    },
+    {
+      "epoch": 1.4990105867220738,
+      "grad_norm": 0.25178257836527557,
+      "learning_rate": 1.0029096477302675e-06,
+      "loss": 0.5663,
+      "step": 1896
+    },
+    {
+      "epoch": 1.4998021173444147,
+      "grad_norm": 0.2522885663597851,
+      "learning_rate": 1.0020783212436786e-06,
+      "loss": 0.5513,
+      "step": 1897
+    },
+    {
+      "epoch": 1.5005936479667557,
+      "grad_norm": 0.23903412879321914,
+      "learning_rate": 1.001246993320745e-06,
+      "loss": 0.5472,
+      "step": 1898
+    },
+    {
+      "epoch": 1.5013851785890968,
+      "grad_norm": 0.24148865550325693,
+      "learning_rate": 1.0004156645360046e-06,
+      "loss": 0.5478,
+      "step": 1899
+    },
+    {
+      "epoch": 1.5021767092114375,
+      "grad_norm": 0.24617269278643503,
+      "learning_rate": 9.995843354639953e-07,
+      "loss": 0.5697,
+      "step": 1900
+    },
+    {
+      "epoch": 1.5029682398337787,
+      "grad_norm": 0.2512853954967072,
+      "learning_rate": 9.98753006679255e-07,
+      "loss": 0.5468,
+      "step": 1901
+    },
+    {
+      "epoch": 1.5037597704561194,
+      "grad_norm": 0.24242798075534466,
+      "learning_rate": 9.979216787563215e-07,
+      "loss": 0.5474,
+      "step": 1902
+    },
+    {
+      "epoch": 1.5045513010784606,
+      "grad_norm": 0.2500089048504842,
+      "learning_rate": 9.970903522697327e-07,
+      "loss": 0.5641,
+      "step": 1903
+    },
+    {
+      "epoch": 1.5053428317008013,
+      "grad_norm": 0.2506717772095021,
+      "learning_rate": 9.962590277940248e-07,
+      "loss": 0.5689,
+      "step": 1904
+    },
+    {
+      "epoch": 1.5061343623231425,
+      "grad_norm": 0.23509731460848576,
+      "learning_rate": 9.954277059037328e-07,
+      "loss": 0.5665,
+      "step": 1905
+    },
+    {
+      "epoch": 1.5069258929454832,
+      "grad_norm": 0.23577031562634193,
+      "learning_rate": 9.9459638717339e-07,
+      "loss": 0.5592,
+      "step": 1906
+    },
+    {
+      "epoch": 1.5077174235678243,
+      "grad_norm": 0.23301075318780953,
+      "learning_rate": 9.937650721775271e-07,
+      "loss": 0.5453,
+      "step": 1907
+    },
+    {
+      "epoch": 1.5085089541901653,
+      "grad_norm": 0.23510510224196848,
+      "learning_rate": 9.92933761490673e-07,
+      "loss": 0.5653,
+      "step": 1908
+    },
+    {
+      "epoch": 1.5093004848125062,
+      "grad_norm": 0.2508185976569224,
+      "learning_rate": 9.921024556873534e-07,
+      "loss": 0.5653,
+      "step": 1909
+    },
+    {
+      "epoch": 1.5100920154348472,
+      "grad_norm": 0.24532833686758868,
+      "learning_rate": 9.9127115534209e-07,
+      "loss": 0.5659,
+      "step": 1910
+    },
+    {
+      "epoch": 1.510883546057188,
+      "grad_norm": 0.24982293913960318,
+      "learning_rate": 9.904398610294009e-07,
+      "loss": 0.5475,
+      "step": 1911
+    },
+    {
+      "epoch": 1.511675076679529,
+      "grad_norm": 0.2461122246563086,
+      "learning_rate": 9.896085733238005e-07,
+      "loss": 0.5663,
+      "step": 1912
+    },
+    {
+      "epoch": 1.51246660730187,
+      "grad_norm": 0.24342830334949983,
+      "learning_rate": 9.887772927997985e-07,
+      "loss": 0.5636,
+      "step": 1913
+    },
+    {
+      "epoch": 1.513258137924211,
+      "grad_norm": 0.26595073547637565,
+      "learning_rate": 9.879460200318996e-07,
+      "loss": 0.5505,
+      "step": 1914
+    },
+    {
+      "epoch": 1.5140496685465519,
+      "grad_norm": 0.2547639076075029,
+      "learning_rate": 9.87114755594603e-07,
+      "loss": 0.5672,
+      "step": 1915
+    },
+    {
+      "epoch": 1.5148411991688928,
+      "grad_norm": 0.24435738914065724,
+      "learning_rate": 9.862835000624027e-07,
+      "loss": 0.5614,
+      "step": 1916
+    },
+    {
+      "epoch": 1.5156327297912338,
+      "grad_norm": 0.24217728482583128,
+      "learning_rate": 9.854522540097849e-07,
+      "loss": 0.5654,
+      "step": 1917
+    },
+    {
+      "epoch": 1.5164242604135747,
+      "grad_norm": 0.2438592351187668,
+      "learning_rate": 9.846210180112313e-07,
+      "loss": 0.554,
+      "step": 1918
+    },
+    {
+      "epoch": 1.5172157910359156,
+      "grad_norm": 0.2543337565157083,
+      "learning_rate": 9.837897926412158e-07,
+      "loss": 0.5678,
+      "step": 1919
+    },
+    {
+      "epoch": 1.5180073216582568,
+      "grad_norm": 0.2475247777369915,
+      "learning_rate": 9.829585784742044e-07,
+      "loss": 0.5557,
+      "step": 1920
+    },
+    {
+      "epoch": 1.5187988522805975,
+      "grad_norm": 0.24541528673676888,
+      "learning_rate": 9.821273760846566e-07,
+      "loss": 0.5692,
+      "step": 1921
+    },
+    {
+      "epoch": 1.5195903829029387,
+      "grad_norm": 0.2440537926695465,
+      "learning_rate": 9.812961860470224e-07,
+      "loss": 0.5506,
+      "step": 1922
+    },
+    {
+      "epoch": 1.5203819135252794,
+      "grad_norm": 0.242479518117656,
+      "learning_rate": 9.80465008935744e-07,
+      "loss": 0.5486,
+      "step": 1923
+    },
+    {
+      "epoch": 1.5211734441476206,
+      "grad_norm": 0.2391267667552566,
+      "learning_rate": 9.79633845325255e-07,
+      "loss": 0.5679,
+      "step": 1924
+    },
+    {
+      "epoch": 1.5219649747699613,
+      "grad_norm": 0.25073091567616873,
+      "learning_rate": 9.788026957899787e-07,
+      "loss": 0.5559,
+      "step": 1925
+    },
+    {
+      "epoch": 1.5227565053923025,
+      "grad_norm": 0.251673253025521,
+      "learning_rate": 9.779715609043298e-07,
+      "loss": 0.564,
+      "step": 1926
+    },
+    {
+      "epoch": 1.5235480360146432,
+      "grad_norm": 0.23796244303345004,
+      "learning_rate": 9.771404412427118e-07,
+      "loss": 0.5499,
+      "step": 1927
+    },
+    {
+      "epoch": 1.5243395666369843,
+      "grad_norm": 0.2451054292089041,
+      "learning_rate": 9.763093373795182e-07,
+      "loss": 0.5538,
+      "step": 1928
+    },
+    {
+      "epoch": 1.5251310972593253,
+      "grad_norm": 0.2624505164949271,
+      "learning_rate": 9.754782498891315e-07,
+      "loss": 0.5539,
+      "step": 1929
+    },
+    {
+      "epoch": 1.5259226278816662,
+      "grad_norm": 0.23533235614968642,
+      "learning_rate": 9.746471793459233e-07,
+      "loss": 0.575,
+      "step": 1930
+    },
+    {
+      "epoch": 1.5267141585040072,
+      "grad_norm": 0.2475762834224649,
+      "learning_rate": 9.738161263242533e-07,
+      "loss": 0.5417,
+      "step": 1931
+    },
+    {
+      "epoch": 1.527505689126348,
+      "grad_norm": 0.24674239111653284,
+      "learning_rate": 9.72985091398468e-07,
+      "loss": 0.552,
+      "step": 1932
+    },
+    {
+      "epoch": 1.528297219748689,
+      "grad_norm": 0.25891616435498827,
+      "learning_rate": 9.721540751429033e-07,
+      "loss": 0.5555,
+      "step": 1933
+    },
+    {
+      "epoch": 1.52908875037103,
+      "grad_norm": 0.2416627505402866,
+      "learning_rate": 9.713230781318805e-07,
+      "loss": 0.5576,
+      "step": 1934
+    },
+    {
+      "epoch": 1.529880280993371,
+      "grad_norm": 0.2382625660179014,
+      "learning_rate": 9.704921009397085e-07,
+      "loss": 0.5602,
+      "step": 1935
+    },
+    {
+      "epoch": 1.5306718116157119,
+      "grad_norm": 0.25361790791674116,
+      "learning_rate": 9.696611441406823e-07,
+      "loss": 0.5692,
+      "step": 1936
+    },
+    {
+      "epoch": 1.5314633422380528,
+      "grad_norm": 0.2590692844289554,
+      "learning_rate": 9.688302083090834e-07,
+      "loss": 0.5623,
+      "step": 1937
+    },
+    {
+      "epoch": 1.5322548728603937,
+      "grad_norm": 0.24290857121052734,
+      "learning_rate": 9.679992940191775e-07,
+      "loss": 0.5589,
+      "step": 1938
+    },
+    {
+      "epoch": 1.5330464034827347,
+      "grad_norm": 0.2590396521661701,
+      "learning_rate": 9.67168401845216e-07,
+      "loss": 0.5614,
+      "step": 1939
+    },
+    {
+      "epoch": 1.5338379341050756,
+      "grad_norm": 0.24220575644838827,
+      "learning_rate": 9.663375323614355e-07,
+      "loss": 0.5563,
+      "step": 1940
+    },
+    {
+      "epoch": 1.5346294647274168,
+      "grad_norm": 0.2431427963659207,
+      "learning_rate": 9.655066861420568e-07,
+      "loss": 0.5686,
+      "step": 1941
+    },
+    {
+      "epoch": 1.5354209953497575,
+      "grad_norm": 0.2445220192707097,
+      "learning_rate": 9.646758637612845e-07,
+      "loss": 0.5563,
+      "step": 1942
+    },
+    {
+      "epoch": 1.5362125259720987,
+      "grad_norm": 0.24382547746742983,
+      "learning_rate": 9.638450657933057e-07,
+      "loss": 0.5733,
+      "step": 1943
+    },
+    {
+      "epoch": 1.5370040565944394,
+      "grad_norm": 0.2499805349979736,
+      "learning_rate": 9.630142928122922e-07,
+      "loss": 0.5623,
+      "step": 1944
+    },
+    {
+      "epoch": 1.5377955872167806,
+      "grad_norm": 0.25483671995828583,
+      "learning_rate": 9.62183545392398e-07,
+      "loss": 0.5698,
+      "step": 1945
+    },
+    {
+      "epoch": 1.5385871178391213,
+      "grad_norm": 0.24186776789464984,
+      "learning_rate": 9.613528241077589e-07,
+      "loss": 0.5491,
+      "step": 1946
+    },
+    {
+      "epoch": 1.5393786484614624,
+      "grad_norm": 0.24206995994822683,
+      "learning_rate": 9.605221295324933e-07,
+      "loss": 0.5571,
+      "step": 1947
+    },
+    {
+      "epoch": 1.5401701790838032,
+      "grad_norm": 0.24485203216287849,
+      "learning_rate": 9.59691462240701e-07,
+      "loss": 0.5635,
+      "step": 1948
+    },
+    {
+      "epoch": 1.5409617097061443,
+      "grad_norm": 0.2496955131917218,
+      "learning_rate": 9.588608228064624e-07,
+      "loss": 0.5516,
+      "step": 1949
+    },
+    {
+      "epoch": 1.541753240328485,
+      "grad_norm": 0.24139459751072465,
+      "learning_rate": 9.580302118038393e-07,
+      "loss": 0.5441,
+      "step": 1950
+    },
+    {
+      "epoch": 1.5425447709508262,
+      "grad_norm": 0.24458136696925753,
+      "learning_rate": 9.571996298068738e-07,
+      "loss": 0.5548,
+      "step": 1951
+    },
+    {
+      "epoch": 1.5433363015731671,
+      "grad_norm": 0.2397884266348811,
+      "learning_rate": 9.563690773895877e-07,
+      "loss": 0.5526,
+      "step": 1952
+    },
+    {
+      "epoch": 1.544127832195508,
+      "grad_norm": 0.24533557300048986,
+      "learning_rate": 9.555385551259826e-07,
+      "loss": 0.5611,
+      "step": 1953
+    },
+    {
+      "epoch": 1.544919362817849,
+      "grad_norm": 0.24802243370126517,
+      "learning_rate": 9.547080635900384e-07,
+      "loss": 0.5549,
+      "step": 1954
+    },
+    {
+      "epoch": 1.54571089344019,
+      "grad_norm": 0.24638276392236952,
+      "learning_rate": 9.538776033557151e-07,
+      "loss": 0.5422,
+      "step": 1955
+    },
+    {
+      "epoch": 1.546502424062531,
+      "grad_norm": 0.23797566068655734,
+      "learning_rate": 9.530471749969505e-07,
+      "loss": 0.5658,
+      "step": 1956
+    },
+    {
+      "epoch": 1.5472939546848719,
+      "grad_norm": 0.2498023988712014,
+      "learning_rate": 9.522167790876599e-07,
+      "loss": 0.5593,
+      "step": 1957
+    },
+    {
+      "epoch": 1.5480854853072128,
+      "grad_norm": 0.2347849795627273,
+      "learning_rate": 9.513864162017372e-07,
+      "loss": 0.5608,
+      "step": 1958
+    },
+    {
+      "epoch": 1.5488770159295537,
+      "grad_norm": 0.24766480533354307,
+      "learning_rate": 9.50556086913052e-07,
+      "loss": 0.5539,
+      "step": 1959
+    },
+    {
+      "epoch": 1.5496685465518947,
+      "grad_norm": 0.25857272748452553,
+      "learning_rate": 9.49725791795452e-07,
+      "loss": 0.542,
+      "step": 1960
+    },
+    {
+      "epoch": 1.5504600771742356,
+      "grad_norm": 0.23965192361455062,
+      "learning_rate": 9.48895531422761e-07,
+      "loss": 0.558,
+      "step": 1961
+    },
+    {
+      "epoch": 1.5512516077965768,
+      "grad_norm": 0.2392492882182199,
+      "learning_rate": 9.480653063687781e-07,
+      "loss": 0.5469,
+      "step": 1962
+    },
+    {
+      "epoch": 1.5520431384189175,
+      "grad_norm": 0.24234911620503033,
+      "learning_rate": 9.472351172072794e-07,
+      "loss": 0.5508,
+      "step": 1963
+    },
+    {
+      "epoch": 1.5528346690412587,
+      "grad_norm": 0.2489087947909424,
+      "learning_rate": 9.464049645120141e-07,
+      "loss": 0.5683,
+      "step": 1964
+    },
+    {
+      "epoch": 1.5536261996635994,
+      "grad_norm": 0.24187863610564125,
+      "learning_rate": 9.45574848856708e-07,
+      "loss": 0.5583,
+      "step": 1965
+    },
+    {
+      "epoch": 1.5544177302859405,
+      "grad_norm": 0.25331174039850096,
+      "learning_rate": 9.447447708150609e-07,
+      "loss": 0.5668,
+      "step": 1966
+    },
+    {
+      "epoch": 1.5552092609082813,
+      "grad_norm": 0.2504989290198329,
+      "learning_rate": 9.439147309607459e-07,
+      "loss": 0.5684,
+      "step": 1967
+    },
+    {
+      "epoch": 1.5560007915306224,
+      "grad_norm": 0.24419936683861262,
+      "learning_rate": 9.430847298674111e-07,
+      "loss": 0.5584,
+      "step": 1968
+    },
+    {
+      "epoch": 1.5567923221529631,
+      "grad_norm": 0.2585052556632517,
+      "learning_rate": 9.422547681086757e-07,
+      "loss": 0.5518,
+      "step": 1969
+    },
+    {
+      "epoch": 1.5575838527753043,
+      "grad_norm": 0.24257383662044224,
+      "learning_rate": 9.414248462581334e-07,
+      "loss": 0.5767,
+      "step": 1970
+    },
+    {
+      "epoch": 1.558375383397645,
+      "grad_norm": 0.24407245554214702,
+      "learning_rate": 9.405949648893504e-07,
+      "loss": 0.5561,
+      "step": 1971
+    },
+    {
+      "epoch": 1.5591669140199862,
+      "grad_norm": 0.24918179240562197,
+      "learning_rate": 9.397651245758636e-07,
+      "loss": 0.5554,
+      "step": 1972
+    },
+    {
+      "epoch": 1.5599584446423271,
+      "grad_norm": 0.24634678811139119,
+      "learning_rate": 9.389353258911832e-07,
+      "loss": 0.5564,
+      "step": 1973
+    },
+    {
+      "epoch": 1.560749975264668,
+      "grad_norm": 0.23634915706130913,
+      "learning_rate": 9.381055694087887e-07,
+      "loss": 0.5716,
+      "step": 1974
+    },
+    {
+      "epoch": 1.561541505887009,
+      "grad_norm": 0.240530519493639,
+      "learning_rate": 9.372758557021319e-07,
+      "loss": 0.5362,
+      "step": 1975
+    },
+    {
+      "epoch": 1.56233303650935,
+      "grad_norm": 0.24612319684632025,
+      "learning_rate": 9.364461853446349e-07,
+      "loss": 0.5675,
+      "step": 1976
+    },
+    {
+      "epoch": 1.563124567131691,
+      "grad_norm": 0.24726122032951803,
+      "learning_rate": 9.35616558909689e-07,
+      "loss": 0.5646,
+      "step": 1977
+    },
+    {
+      "epoch": 1.5639160977540318,
+      "grad_norm": 0.2431712736071995,
+      "learning_rate": 9.347869769706561e-07,
+      "loss": 0.5471,
+      "step": 1978
+    },
+    {
+      "epoch": 1.5647076283763728,
+      "grad_norm": 0.24200958775563516,
+      "learning_rate": 9.339574401008672e-07,
+      "loss": 0.5732,
+      "step": 1979
+    },
+    {
+      "epoch": 1.5654991589987137,
+      "grad_norm": 0.24121456560440047,
+      "learning_rate": 9.331279488736213e-07,
+      "loss": 0.5536,
+      "step": 1980
+    },
+    {
+      "epoch": 1.5662906896210547,
+      "grad_norm": 0.24310098306754682,
+      "learning_rate": 9.322985038621863e-07,
+      "loss": 0.5526,
+      "step": 1981
+    },
+    {
+      "epoch": 1.5670822202433956,
+      "grad_norm": 0.23658339428483585,
+      "learning_rate": 9.314691056397989e-07,
+      "loss": 0.5496,
+      "step": 1982
+    },
+    {
+      "epoch": 1.5678737508657368,
+      "grad_norm": 0.2435279041919117,
+      "learning_rate": 9.306397547796624e-07,
+      "loss": 0.5628,
+      "step": 1983
+    },
+    {
+      "epoch": 1.5686652814880775,
+      "grad_norm": 0.23426171320870792,
+      "learning_rate": 9.298104518549486e-07,
+      "loss": 0.561,
+      "step": 1984
+    },
+    {
+      "epoch": 1.5694568121104187,
+      "grad_norm": 0.24702648257739543,
+      "learning_rate": 9.289811974387945e-07,
+      "loss": 0.5456,
+      "step": 1985
+    },
+    {
+      "epoch": 1.5702483427327594,
+      "grad_norm": 0.24863404130824132,
+      "learning_rate": 9.281519921043048e-07,
+      "loss": 0.5587,
+      "step": 1986
+    },
+    {
+      "epoch": 1.5710398733551005,
+      "grad_norm": 0.2332512958343926,
+      "learning_rate": 9.273228364245499e-07,
+      "loss": 0.5547,
+      "step": 1987
+    },
+    {
+      "epoch": 1.5718314039774413,
+      "grad_norm": 0.24031475053794954,
+      "learning_rate": 9.264937309725662e-07,
+      "loss": 0.5749,
+      "step": 1988
+    },
+    {
+      "epoch": 1.5726229345997824,
+      "grad_norm": 0.24036406010731085,
+      "learning_rate": 9.256646763213553e-07,
+      "loss": 0.5568,
+      "step": 1989
+    },
+    {
+      "epoch": 1.5734144652221231,
+      "grad_norm": 0.25120888713194434,
+      "learning_rate": 9.24835673043883e-07,
+      "loss": 0.5475,
+      "step": 1990
+    },
+    {
+      "epoch": 1.5742059958444643,
+      "grad_norm": 0.24591089376286998,
+      "learning_rate": 9.240067217130801e-07,
+      "loss": 0.5597,
+      "step": 1991
+    },
+    {
+      "epoch": 1.574997526466805,
+      "grad_norm": 0.23881311139117428,
+      "learning_rate": 9.231778229018416e-07,
+      "loss": 0.5462,
+      "step": 1992
+    },
+    {
+      "epoch": 1.5757890570891462,
+      "grad_norm": 0.24112698391348775,
+      "learning_rate": 9.223489771830265e-07,
+      "loss": 0.5473,
+      "step": 1993
+    },
+    {
+      "epoch": 1.5765805877114871,
+      "grad_norm": 0.2429405631746065,
+      "learning_rate": 9.215201851294569e-07,
+      "loss": 0.5654,
+      "step": 1994
+    },
+    {
+      "epoch": 1.577372118333828,
+      "grad_norm": 0.2646333136373766,
+      "learning_rate": 9.206914473139169e-07,
+      "loss": 0.5557,
+      "step": 1995
+    },
+    {
+      "epoch": 1.578163648956169,
+      "grad_norm": 0.2486794216251206,
+      "learning_rate": 9.198627643091541e-07,
+      "loss": 0.5505,
+      "step": 1996
+    },
+    {
+      "epoch": 1.57895517957851,
+      "grad_norm": 0.23557275143197978,
+      "learning_rate": 9.19034136687878e-07,
+      "loss": 0.5642,
+      "step": 1997
+    },
+    {
+      "epoch": 1.579746710200851,
+      "grad_norm": 0.24290837948404265,
+      "learning_rate": 9.1820556502276e-07,
+      "loss": 0.5527,
+      "step": 1998
+    },
+    {
+      "epoch": 1.5805382408231918,
+      "grad_norm": 0.2445531298782679,
+      "learning_rate": 9.173770498864324e-07,
+      "loss": 0.5406,
+      "step": 1999
+    },
+    {
+      "epoch": 1.5813297714455328,
+      "grad_norm": 0.24521079534272908,
+      "learning_rate": 9.165485918514891e-07,
+      "loss": 0.5719,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5821213020678737,
+      "grad_norm": 0.24382747432485855,
+      "learning_rate": 9.157201914904834e-07,
+      "loss": 0.5643,
+      "step": 2001
+    },
+    {
+      "epoch": 1.5829128326902147,
+      "grad_norm": 0.24316398481862123,
+      "learning_rate": 9.148918493759297e-07,
+      "loss": 0.5561,
+      "step": 2002
+    },
+    {
+      "epoch": 1.5837043633125556,
+      "grad_norm": 0.23884960368469524,
+      "learning_rate": 9.140635660803024e-07,
+      "loss": 0.5526,
+      "step": 2003
+    },
+    {
+      "epoch": 1.5844958939348968,
+      "grad_norm": 0.23978880256137527,
+      "learning_rate": 9.132353421760339e-07,
+      "loss": 0.5375,
+      "step": 2004
+    },
+    {
+      "epoch": 1.5852874245572375,
+      "grad_norm": 0.2412768962551659,
+      "learning_rate": 9.124071782355171e-07,
+      "loss": 0.5518,
+      "step": 2005
+    },
+    {
+      "epoch": 1.5860789551795786,
+      "grad_norm": 0.2439193302045052,
+      "learning_rate": 9.115790748311022e-07,
+      "loss": 0.5583,
+      "step": 2006
+    },
+    {
+      "epoch": 1.5868704858019194,
+      "grad_norm": 0.24777383189207242,
+      "learning_rate": 9.107510325350983e-07,
+      "loss": 0.5474,
+      "step": 2007
+    },
+    {
+      "epoch": 1.5876620164242605,
+      "grad_norm": 0.25061556066450913,
+      "learning_rate": 9.099230519197724e-07,
+      "loss": 0.568,
+      "step": 2008
+    },
+    {
+      "epoch": 1.5884535470466012,
+      "grad_norm": 0.23769449043634142,
+      "learning_rate": 9.09095133557348e-07,
+      "loss": 0.5518,
+      "step": 2009
+    },
+    {
+      "epoch": 1.5892450776689424,
+      "grad_norm": 0.2437545233145305,
+      "learning_rate": 9.082672780200068e-07,
+      "loss": 0.5682,
+      "step": 2010
+    },
+    {
+      "epoch": 1.5900366082912831,
+      "grad_norm": 0.24597414418264885,
+      "learning_rate": 9.074394858798856e-07,
+      "loss": 0.558,
+      "step": 2011
+    },
+    {
+      "epoch": 1.5908281389136243,
+      "grad_norm": 0.2505269187142412,
+      "learning_rate": 9.066117577090786e-07,
+      "loss": 0.5542,
+      "step": 2012
+    },
+    {
+      "epoch": 1.591619669535965,
+      "grad_norm": 0.2438859735554627,
+      "learning_rate": 9.057840940796356e-07,
+      "loss": 0.5527,
+      "step": 2013
+    },
+    {
+      "epoch": 1.5924112001583062,
+      "grad_norm": 0.24398672777416325,
+      "learning_rate": 9.049564955635612e-07,
+      "loss": 0.5516,
+      "step": 2014
+    },
+    {
+      "epoch": 1.5932027307806471,
+      "grad_norm": 0.23802712142850696,
+      "learning_rate": 9.041289627328158e-07,
+      "loss": 0.5522,
+      "step": 2015
+    },
+    {
+      "epoch": 1.593994261402988,
+      "grad_norm": 0.23947980168767818,
+      "learning_rate": 9.033014961593134e-07,
+      "loss": 0.5511,
+      "step": 2016
+    },
+    {
+      "epoch": 1.594785792025329,
+      "grad_norm": 0.24309875949637275,
+      "learning_rate": 9.024740964149231e-07,
+      "loss": 0.5485,
+      "step": 2017
+    },
+    {
+      "epoch": 1.59557732264767,
+      "grad_norm": 0.24157386327966096,
+      "learning_rate": 9.016467640714677e-07,
+      "loss": 0.5507,
+      "step": 2018
+    },
+    {
+      "epoch": 1.5963688532700109,
+      "grad_norm": 0.2445120784282942,
+      "learning_rate": 9.008194997007227e-07,
+      "loss": 0.5445,
+      "step": 2019
+    },
+    {
+      "epoch": 1.5971603838923518,
+      "grad_norm": 0.23954075205120254,
+      "learning_rate": 8.999923038744177e-07,
+      "loss": 0.55,
+      "step": 2020
+    },
+    {
+      "epoch": 1.5979519145146928,
+      "grad_norm": 0.23531685414910658,
+      "learning_rate": 8.991651771642345e-07,
+      "loss": 0.5547,
+      "step": 2021
+    },
+    {
+      "epoch": 1.5987434451370337,
+      "grad_norm": 0.2518174676964054,
+      "learning_rate": 8.983381201418062e-07,
+      "loss": 0.5564,
+      "step": 2022
+    },
+    {
+      "epoch": 1.5995349757593746,
+      "grad_norm": 0.24207774498953183,
+      "learning_rate": 8.975111333787192e-07,
+      "loss": 0.553,
+      "step": 2023
+    },
+    {
+      "epoch": 1.6003265063817156,
+      "grad_norm": 0.24858406061862517,
+      "learning_rate": 8.966842174465103e-07,
+      "loss": 0.5536,
+      "step": 2024
+    },
+    {
+      "epoch": 1.6011180370040567,
+      "grad_norm": 0.23619562290454044,
+      "learning_rate": 8.95857372916668e-07,
+      "loss": 0.546,
+      "step": 2025
+    },
+    {
+      "epoch": 1.6019095676263975,
+      "grad_norm": 0.2400375915791126,
+      "learning_rate": 8.950306003606314e-07,
+      "loss": 0.5724,
+      "step": 2026
+    },
+    {
+      "epoch": 1.6027010982487386,
+      "grad_norm": 0.24226127560950453,
+      "learning_rate": 8.942039003497892e-07,
+      "loss": 0.5531,
+      "step": 2027
+    },
+    {
+      "epoch": 1.6034926288710794,
+      "grad_norm": 0.24417083063861214,
+      "learning_rate": 8.933772734554806e-07,
+      "loss": 0.5589,
+      "step": 2028
+    },
+    {
+      "epoch": 1.6042841594934205,
+      "grad_norm": 0.24056487563446027,
+      "learning_rate": 8.925507202489938e-07,
+      "loss": 0.5691,
+      "step": 2029
+    },
+    {
+      "epoch": 1.6050756901157612,
+      "grad_norm": 0.24375248747069628,
+      "learning_rate": 8.917242413015666e-07,
+      "loss": 0.536,
+      "step": 2030
+    },
+    {
+      "epoch": 1.6058672207381024,
+      "grad_norm": 0.24527152468161173,
+      "learning_rate": 8.908978371843858e-07,
+      "loss": 0.5534,
+      "step": 2031
+    },
+    {
+      "epoch": 1.6066587513604431,
+      "grad_norm": 0.2387293578057444,
+      "learning_rate": 8.90071508468585e-07,
+      "loss": 0.5635,
+      "step": 2032
+    },
+    {
+      "epoch": 1.6074502819827843,
+      "grad_norm": 0.2402138297428413,
+      "learning_rate": 8.892452557252471e-07,
+      "loss": 0.5698,
+      "step": 2033
+    },
+    {
+      "epoch": 1.608241812605125,
+      "grad_norm": 0.2519986495145601,
+      "learning_rate": 8.884190795254017e-07,
+      "loss": 0.5664,
+      "step": 2034
+    },
+    {
+      "epoch": 1.6090333432274662,
+      "grad_norm": 0.24916973598540385,
+      "learning_rate": 8.875929804400259e-07,
+      "loss": 0.5667,
+      "step": 2035
+    },
+    {
+      "epoch": 1.609824873849807,
+      "grad_norm": 0.23952696403139964,
+      "learning_rate": 8.86766959040044e-07,
+      "loss": 0.5477,
+      "step": 2036
+    },
+    {
+      "epoch": 1.610616404472148,
+      "grad_norm": 0.25479237829538365,
+      "learning_rate": 8.859410158963254e-07,
+      "loss": 0.5617,
+      "step": 2037
+    },
+    {
+      "epoch": 1.611407935094489,
+      "grad_norm": 0.24759472080214803,
+      "learning_rate": 8.851151515796859e-07,
+      "loss": 0.5547,
+      "step": 2038
+    },
+    {
+      "epoch": 1.61219946571683,
+      "grad_norm": 0.2472420808791739,
+      "learning_rate": 8.842893666608871e-07,
+      "loss": 0.547,
+      "step": 2039
+    },
+    {
+      "epoch": 1.6129909963391709,
+      "grad_norm": 0.25737531625494264,
+      "learning_rate": 8.834636617106356e-07,
+      "loss": 0.5485,
+      "step": 2040
+    },
+    {
+      "epoch": 1.6137825269615118,
+      "grad_norm": 0.2427318858871427,
+      "learning_rate": 8.826380372995832e-07,
+      "loss": 0.5449,
+      "step": 2041
+    },
+    {
+      "epoch": 1.6145740575838528,
+      "grad_norm": 0.24872616426240185,
+      "learning_rate": 8.818124939983254e-07,
+      "loss": 0.5502,
+      "step": 2042
+    },
+    {
+      "epoch": 1.6153655882061937,
+      "grad_norm": 0.23712940097713245,
+      "learning_rate": 8.80987032377401e-07,
+      "loss": 0.5593,
+      "step": 2043
+    },
+    {
+      "epoch": 1.6161571188285346,
+      "grad_norm": 0.24970016687100757,
+      "learning_rate": 8.801616530072939e-07,
+      "loss": 0.5492,
+      "step": 2044
+    },
+    {
+      "epoch": 1.6169486494508756,
+      "grad_norm": 0.2408996082081383,
+      "learning_rate": 8.7933635645843e-07,
+      "loss": 0.5664,
+      "step": 2045
+    },
+    {
+      "epoch": 1.6177401800732167,
+      "grad_norm": 0.24399569773186586,
+      "learning_rate": 8.785111433011789e-07,
+      "loss": 0.5461,
+      "step": 2046
+    },
+    {
+      "epoch": 1.6185317106955575,
+      "grad_norm": 0.24474799760899071,
+      "learning_rate": 8.77686014105852e-07,
+      "loss": 0.5484,
+      "step": 2047
+    },
+    {
+      "epoch": 1.6193232413178986,
+      "grad_norm": 0.2546632175346895,
+      "learning_rate": 8.76860969442702e-07,
+      "loss": 0.5576,
+      "step": 2048
+    },
+    {
+      "epoch": 1.6201147719402393,
+      "grad_norm": 0.2409988974971737,
+      "learning_rate": 8.760360098819243e-07,
+      "loss": 0.5665,
+      "step": 2049
+    },
+    {
+      "epoch": 1.6209063025625805,
+      "grad_norm": 0.2513223866448745,
+      "learning_rate": 8.75211135993655e-07,
+      "loss": 0.5668,
+      "step": 2050
+    },
+    {
+      "epoch": 1.6216978331849212,
+      "grad_norm": 0.24429719225143925,
+      "learning_rate": 8.743863483479716e-07,
+      "loss": 0.5684,
+      "step": 2051
+    },
+    {
+      "epoch": 1.6224893638072624,
+      "grad_norm": 0.23998508846294653,
+      "learning_rate": 8.73561647514891e-07,
+      "loss": 0.5568,
+      "step": 2052
+    },
+    {
+      "epoch": 1.623280894429603,
+      "grad_norm": 0.24080160339323878,
+      "learning_rate": 8.727370340643704e-07,
+      "loss": 0.5601,
+      "step": 2053
+    },
+    {
+      "epoch": 1.6240724250519443,
+      "grad_norm": 0.24644207357074463,
+      "learning_rate": 8.719125085663069e-07,
+      "loss": 0.5485,
+      "step": 2054
+    },
+    {
+      "epoch": 1.624863955674285,
+      "grad_norm": 0.25001783187191484,
+      "learning_rate": 8.710880715905369e-07,
+      "loss": 0.561,
+      "step": 2055
+    },
+    {
+      "epoch": 1.6256554862966262,
+      "grad_norm": 0.25594656785194075,
+      "learning_rate": 8.702637237068351e-07,
+      "loss": 0.5545,
+      "step": 2056
+    },
+    {
+      "epoch": 1.626447016918967,
+      "grad_norm": 0.24342006382441958,
+      "learning_rate": 8.694394654849155e-07,
+      "loss": 0.5522,
+      "step": 2057
+    },
+    {
+      "epoch": 1.627238547541308,
+      "grad_norm": 0.23963624568506198,
+      "learning_rate": 8.686152974944287e-07,
+      "loss": 0.558,
+      "step": 2058
+    },
+    {
+      "epoch": 1.628030078163649,
+      "grad_norm": 0.24318954257524006,
+      "learning_rate": 8.677912203049645e-07,
+      "loss": 0.557,
+      "step": 2059
+    },
+    {
+      "epoch": 1.62882160878599,
+      "grad_norm": 0.24052317362646064,
+      "learning_rate": 8.669672344860491e-07,
+      "loss": 0.556,
+      "step": 2060
+    },
+    {
+      "epoch": 1.6296131394083309,
+      "grad_norm": 0.25711502671895936,
+      "learning_rate": 8.661433406071453e-07,
+      "loss": 0.5693,
+      "step": 2061
+    },
+    {
+      "epoch": 1.6304046700306718,
+      "grad_norm": 0.24102907986820435,
+      "learning_rate": 8.653195392376538e-07,
+      "loss": 0.558,
+      "step": 2062
+    },
+    {
+      "epoch": 1.6311962006530127,
+      "grad_norm": 0.2373550834520336,
+      "learning_rate": 8.644958309469091e-07,
+      "loss": 0.5395,
+      "step": 2063
+    },
+    {
+      "epoch": 1.6319877312753537,
+      "grad_norm": 0.2570678590573862,
+      "learning_rate": 8.636722163041833e-07,
+      "loss": 0.5391,
+      "step": 2064
+    },
+    {
+      "epoch": 1.6327792618976946,
+      "grad_norm": 0.2445396534415899,
+      "learning_rate": 8.628486958786831e-07,
+      "loss": 0.5457,
+      "step": 2065
+    },
+    {
+      "epoch": 1.6335707925200356,
+      "grad_norm": 0.24447082953743518,
+      "learning_rate": 8.620252702395499e-07,
+      "loss": 0.5519,
+      "step": 2066
+    },
+    {
+      "epoch": 1.6343623231423767,
+      "grad_norm": 0.2455708603720528,
+      "learning_rate": 8.612019399558597e-07,
+      "loss": 0.5618,
+      "step": 2067
+    },
+    {
+      "epoch": 1.6351538537647174,
+      "grad_norm": 0.2428458574211925,
+      "learning_rate": 8.603787055966232e-07,
+      "loss": 0.5711,
+      "step": 2068
+    },
+    {
+      "epoch": 1.6359453843870586,
+      "grad_norm": 0.2428632953712524,
+      "learning_rate": 8.595555677307835e-07,
+      "loss": 0.5469,
+      "step": 2069
+    },
+    {
+      "epoch": 1.6367369150093993,
+      "grad_norm": 0.25187082947017325,
+      "learning_rate": 8.587325269272183e-07,
+      "loss": 0.5558,
+      "step": 2070
+    },
+    {
+      "epoch": 1.6375284456317405,
+      "grad_norm": 0.23378164006817775,
+      "learning_rate": 8.579095837547373e-07,
+      "loss": 0.5487,
+      "step": 2071
+    },
+    {
+      "epoch": 1.6383199762540812,
+      "grad_norm": 0.23723769921773163,
+      "learning_rate": 8.570867387820835e-07,
+      "loss": 0.5571,
+      "step": 2072
+    },
+    {
+      "epoch": 1.6391115068764224,
+      "grad_norm": 0.23914851721873825,
+      "learning_rate": 8.562639925779318e-07,
+      "loss": 0.5451,
+      "step": 2073
+    },
+    {
+      "epoch": 1.639903037498763,
+      "grad_norm": 0.24889143215976497,
+      "learning_rate": 8.554413457108881e-07,
+      "loss": 0.5422,
+      "step": 2074
+    },
+    {
+      "epoch": 1.6406945681211043,
+      "grad_norm": 0.23847255202101136,
+      "learning_rate": 8.546187987494908e-07,
+      "loss": 0.5659,
+      "step": 2075
+    },
+    {
+      "epoch": 1.641486098743445,
+      "grad_norm": 0.24034831626255304,
+      "learning_rate": 8.537963522622084e-07,
+      "loss": 0.5581,
+      "step": 2076
+    },
+    {
+      "epoch": 1.6422776293657861,
+      "grad_norm": 0.24280446433904804,
+      "learning_rate": 8.529740068174402e-07,
+      "loss": 0.5598,
+      "step": 2077
+    },
+    {
+      "epoch": 1.643069159988127,
+      "grad_norm": 0.24181242354981677,
+      "learning_rate": 8.521517629835165e-07,
+      "loss": 0.5477,
+      "step": 2078
+    },
+    {
+      "epoch": 1.643860690610468,
+      "grad_norm": 0.24745841762119297,
+      "learning_rate": 8.513296213286956e-07,
+      "loss": 0.5708,
+      "step": 2079
+    },
+    {
+      "epoch": 1.644652221232809,
+      "grad_norm": 0.23458264483259214,
+      "learning_rate": 8.505075824211668e-07,
+      "loss": 0.5372,
+      "step": 2080
+    },
+    {
+      "epoch": 1.64544375185515,
+      "grad_norm": 0.2474157511212525,
+      "learning_rate": 8.496856468290474e-07,
+      "loss": 0.5667,
+      "step": 2081
+    },
+    {
+      "epoch": 1.6462352824774908,
+      "grad_norm": 0.23564463504954694,
+      "learning_rate": 8.488638151203839e-07,
+      "loss": 0.5609,
+      "step": 2082
+    },
+    {
+      "epoch": 1.6470268130998318,
+      "grad_norm": 0.2503948137425057,
+      "learning_rate": 8.480420878631511e-07,
+      "loss": 0.5577,
+      "step": 2083
+    },
+    {
+      "epoch": 1.6478183437221727,
+      "grad_norm": 0.24463697719998737,
+      "learning_rate": 8.472204656252506e-07,
+      "loss": 0.564,
+      "step": 2084
+    },
+    {
+      "epoch": 1.6486098743445137,
+      "grad_norm": 0.24132887295057562,
+      "learning_rate": 8.463989489745129e-07,
+      "loss": 0.5471,
+      "step": 2085
+    },
+    {
+      "epoch": 1.6494014049668546,
+      "grad_norm": 0.23539199733933527,
+      "learning_rate": 8.455775384786939e-07,
+      "loss": 0.5628,
+      "step": 2086
+    },
+    {
+      "epoch": 1.6501929355891956,
+      "grad_norm": 0.24810760237934773,
+      "learning_rate": 8.447562347054776e-07,
+      "loss": 0.5551,
+      "step": 2087
+    },
+    {
+      "epoch": 1.6509844662115367,
+      "grad_norm": 0.2428508266769543,
+      "learning_rate": 8.439350382224737e-07,
+      "loss": 0.5633,
+      "step": 2088
+    },
+    {
+      "epoch": 1.6517759968338774,
+      "grad_norm": 0.24540948486221204,
+      "learning_rate": 8.431139495972176e-07,
+      "loss": 0.561,
+      "step": 2089
+    },
+    {
+      "epoch": 1.6525675274562186,
+      "grad_norm": 0.25065727786285197,
+      "learning_rate": 8.422929693971698e-07,
+      "loss": 0.5455,
+      "step": 2090
+    },
+    {
+      "epoch": 1.6533590580785593,
+      "grad_norm": 0.23792356960861794,
+      "learning_rate": 8.414720981897164e-07,
+      "loss": 0.5583,
+      "step": 2091
+    },
+    {
+      "epoch": 1.6541505887009005,
+      "grad_norm": 0.24225723128309418,
+      "learning_rate": 8.406513365421683e-07,
+      "loss": 0.5663,
+      "step": 2092
+    },
+    {
+      "epoch": 1.6549421193232412,
+      "grad_norm": 0.24501700400121662,
+      "learning_rate": 8.398306850217608e-07,
+      "loss": 0.5467,
+      "step": 2093
+    },
+    {
+      "epoch": 1.6557336499455824,
+      "grad_norm": 0.24501439601092997,
+      "learning_rate": 8.390101441956525e-07,
+      "loss": 0.5626,
+      "step": 2094
+    },
+    {
+      "epoch": 1.656525180567923,
+      "grad_norm": 0.24856531462758444,
+      "learning_rate": 8.381897146309253e-07,
+      "loss": 0.5629,
+      "step": 2095
+    },
+    {
+      "epoch": 1.6573167111902642,
+      "grad_norm": 0.23883102164757536,
+      "learning_rate": 8.373693968945849e-07,
+      "loss": 0.5398,
+      "step": 2096
+    },
+    {
+      "epoch": 1.658108241812605,
+      "grad_norm": 0.2445122932551097,
+      "learning_rate": 8.365491915535597e-07,
+      "loss": 0.547,
+      "step": 2097
+    },
+    {
+      "epoch": 1.6588997724349461,
+      "grad_norm": 0.25961728718405314,
+      "learning_rate": 8.357290991747002e-07,
+      "loss": 0.5549,
+      "step": 2098
+    },
+    {
+      "epoch": 1.659691303057287,
+      "grad_norm": 0.2375815462137297,
+      "learning_rate": 8.34909120324779e-07,
+      "loss": 0.5574,
+      "step": 2099
+    },
+    {
+      "epoch": 1.660482833679628,
+      "grad_norm": 0.24519708664453363,
+      "learning_rate": 8.340892555704893e-07,
+      "loss": 0.5568,
+      "step": 2100
+    },
+    {
+      "epoch": 1.661274364301969,
+      "grad_norm": 0.2438913296131192,
+      "learning_rate": 8.332695054784466e-07,
+      "loss": 0.5609,
+      "step": 2101
+    },
+    {
+      "epoch": 1.66206589492431,
+      "grad_norm": 0.23898030997530736,
+      "learning_rate": 8.324498706151868e-07,
+      "loss": 0.5559,
+      "step": 2102
+    },
+    {
+      "epoch": 1.6628574255466508,
+      "grad_norm": 0.24166461498282207,
+      "learning_rate": 8.316303515471665e-07,
+      "loss": 0.5595,
+      "step": 2103
+    },
+    {
+      "epoch": 1.6636489561689918,
+      "grad_norm": 0.2446555279103593,
+      "learning_rate": 8.308109488407618e-07,
+      "loss": 0.5632,
+      "step": 2104
+    },
+    {
+      "epoch": 1.6644404867913327,
+      "grad_norm": 0.2567484699955933,
+      "learning_rate": 8.299916630622678e-07,
+      "loss": 0.5565,
+      "step": 2105
+    },
+    {
+      "epoch": 1.6652320174136737,
+      "grad_norm": 0.23880975061449405,
+      "learning_rate": 8.291724947778998e-07,
+      "loss": 0.5487,
+      "step": 2106
+    },
+    {
+      "epoch": 1.6660235480360146,
+      "grad_norm": 0.25089707885771734,
+      "learning_rate": 8.283534445537921e-07,
+      "loss": 0.5632,
+      "step": 2107
+    },
+    {
+      "epoch": 1.6668150786583555,
+      "grad_norm": 0.24072108293658043,
+      "learning_rate": 8.275345129559962e-07,
+      "loss": 0.5523,
+      "step": 2108
+    },
+    {
+      "epoch": 1.6676066092806967,
+      "grad_norm": 0.2475226641466154,
+      "learning_rate": 8.267157005504825e-07,
+      "loss": 0.5569,
+      "step": 2109
+    },
+    {
+      "epoch": 1.6683981399030374,
+      "grad_norm": 0.24007485038291715,
+      "learning_rate": 8.258970079031395e-07,
+      "loss": 0.5552,
+      "step": 2110
+    },
+    {
+      "epoch": 1.6691896705253786,
+      "grad_norm": 0.23882560906148753,
+      "learning_rate": 8.250784355797714e-07,
+      "loss": 0.5673,
+      "step": 2111
+    },
+    {
+      "epoch": 1.6699812011477193,
+      "grad_norm": 0.2517824349186474,
+      "learning_rate": 8.242599841461005e-07,
+      "loss": 0.5574,
+      "step": 2112
+    },
+    {
+      "epoch": 1.6707727317700605,
+      "grad_norm": 0.24295826491829725,
+      "learning_rate": 8.234416541677647e-07,
+      "loss": 0.5523,
+      "step": 2113
+    },
+    {
+      "epoch": 1.6715642623924012,
+      "grad_norm": 0.24388404869935035,
+      "learning_rate": 8.226234462103188e-07,
+      "loss": 0.5777,
+      "step": 2114
+    },
+    {
+      "epoch": 1.6723557930147424,
+      "grad_norm": 0.23786869322677054,
+      "learning_rate": 8.218053608392332e-07,
+      "loss": 0.5562,
+      "step": 2115
+    },
+    {
+      "epoch": 1.673147323637083,
+      "grad_norm": 0.2531283298486136,
+      "learning_rate": 8.209873986198926e-07,
+      "loss": 0.5528,
+      "step": 2116
+    },
+    {
+      "epoch": 1.6739388542594242,
+      "grad_norm": 0.23718460284078302,
+      "learning_rate": 8.201695601175977e-07,
+      "loss": 0.5672,
+      "step": 2117
+    },
+    {
+      "epoch": 1.674730384881765,
+      "grad_norm": 0.23993927480694083,
+      "learning_rate": 8.19351845897563e-07,
+      "loss": 0.5699,
+      "step": 2118
+    },
+    {
+      "epoch": 1.6755219155041061,
+      "grad_norm": 0.23794714997443736,
+      "learning_rate": 8.185342565249174e-07,
+      "loss": 0.5582,
+      "step": 2119
+    },
+    {
+      "epoch": 1.676313446126447,
+      "grad_norm": 0.23950708340571722,
+      "learning_rate": 8.177167925647039e-07,
+      "loss": 0.556,
+      "step": 2120
+    },
+    {
+      "epoch": 1.677104976748788,
+      "grad_norm": 0.24068667787476933,
+      "learning_rate": 8.168994545818779e-07,
+      "loss": 0.5464,
+      "step": 2121
+    },
+    {
+      "epoch": 1.677896507371129,
+      "grad_norm": 0.2507718932955019,
+      "learning_rate": 8.160822431413084e-07,
+      "loss": 0.5501,
+      "step": 2122
+    },
+    {
+      "epoch": 1.6786880379934699,
+      "grad_norm": 0.24007692263093897,
+      "learning_rate": 8.152651588077765e-07,
+      "loss": 0.5517,
+      "step": 2123
+    },
+    {
+      "epoch": 1.6794795686158108,
+      "grad_norm": 0.2515251510367636,
+      "learning_rate": 8.144482021459762e-07,
+      "loss": 0.5518,
+      "step": 2124
+    },
+    {
+      "epoch": 1.6802710992381518,
+      "grad_norm": 0.25013459716253933,
+      "learning_rate": 8.13631373720513e-07,
+      "loss": 0.5598,
+      "step": 2125
+    },
+    {
+      "epoch": 1.6810626298604927,
+      "grad_norm": 0.23832331109270669,
+      "learning_rate": 8.128146740959028e-07,
+      "loss": 0.5549,
+      "step": 2126
+    },
+    {
+      "epoch": 1.6818541604828336,
+      "grad_norm": 0.24691594384946502,
+      "learning_rate": 8.119981038365738e-07,
+      "loss": 0.5635,
+      "step": 2127
+    },
+    {
+      "epoch": 1.6826456911051746,
+      "grad_norm": 0.24515478763570772,
+      "learning_rate": 8.11181663506864e-07,
+      "loss": 0.5448,
+      "step": 2128
+    },
+    {
+      "epoch": 1.6834372217275155,
+      "grad_norm": 0.24830744464257612,
+      "learning_rate": 8.103653536710219e-07,
+      "loss": 0.5619,
+      "step": 2129
+    },
+    {
+      "epoch": 1.6842287523498567,
+      "grad_norm": 0.24240100747542648,
+      "learning_rate": 8.095491748932063e-07,
+      "loss": 0.5568,
+      "step": 2130
+    },
+    {
+      "epoch": 1.6850202829721974,
+      "grad_norm": 0.23911385549167222,
+      "learning_rate": 8.087331277374844e-07,
+      "loss": 0.5642,
+      "step": 2131
+    },
+    {
+      "epoch": 1.6858118135945386,
+      "grad_norm": 0.23455278963650725,
+      "learning_rate": 8.079172127678329e-07,
+      "loss": 0.5471,
+      "step": 2132
+    },
+    {
+      "epoch": 1.6866033442168793,
+      "grad_norm": 0.2380396825467588,
+      "learning_rate": 8.071014305481373e-07,
+      "loss": 0.5496,
+      "step": 2133
+    },
+    {
+      "epoch": 1.6873948748392205,
+      "grad_norm": 0.24561620775271423,
+      "learning_rate": 8.06285781642191e-07,
+      "loss": 0.5628,
+      "step": 2134
+    },
+    {
+      "epoch": 1.6881864054615612,
+      "grad_norm": 0.23719363419038508,
+      "learning_rate": 8.054702666136962e-07,
+      "loss": 0.5424,
+      "step": 2135
+    },
+    {
+      "epoch": 1.6889779360839023,
+      "grad_norm": 0.24429327343347726,
+      "learning_rate": 8.046548860262613e-07,
+      "loss": 0.5617,
+      "step": 2136
+    },
+    {
+      "epoch": 1.689769466706243,
+      "grad_norm": 0.23427956646463197,
+      "learning_rate": 8.038396404434024e-07,
+      "loss": 0.5662,
+      "step": 2137
+    },
+    {
+      "epoch": 1.6905609973285842,
+      "grad_norm": 0.24412993380524262,
+      "learning_rate": 8.030245304285424e-07,
+      "loss": 0.5381,
+      "step": 2138
+    },
+    {
+      "epoch": 1.691352527950925,
+      "grad_norm": 0.2442892324940378,
+      "learning_rate": 8.022095565450102e-07,
+      "loss": 0.5581,
+      "step": 2139
+    },
+    {
+      "epoch": 1.692144058573266,
+      "grad_norm": 0.24154065834023128,
+      "learning_rate": 8.013947193560412e-07,
+      "loss": 0.5585,
+      "step": 2140
+    },
+    {
+      "epoch": 1.692935589195607,
+      "grad_norm": 0.23448738096386273,
+      "learning_rate": 8.005800194247759e-07,
+      "loss": 0.5538,
+      "step": 2141
+    },
+    {
+      "epoch": 1.693727119817948,
+      "grad_norm": 0.24132536336211838,
+      "learning_rate": 7.997654573142592e-07,
+      "loss": 0.556,
+      "step": 2142
+    },
+    {
+      "epoch": 1.694518650440289,
+      "grad_norm": 0.24420785160970682,
+      "learning_rate": 7.989510335874419e-07,
+      "loss": 0.5581,
+      "step": 2143
+    },
+    {
+      "epoch": 1.6953101810626299,
+      "grad_norm": 0.24196611171201077,
+      "learning_rate": 7.98136748807179e-07,
+      "loss": 0.5595,
+      "step": 2144
+    },
+    {
+      "epoch": 1.6961017116849708,
+      "grad_norm": 0.23855984750560175,
+      "learning_rate": 7.973226035362295e-07,
+      "loss": 0.5598,
+      "step": 2145
+    },
+    {
+      "epoch": 1.6968932423073118,
+      "grad_norm": 0.24608009031288522,
+      "learning_rate": 7.965085983372555e-07,
+      "loss": 0.5561,
+      "step": 2146
+    },
+    {
+      "epoch": 1.6976847729296527,
+      "grad_norm": 0.24465724540214806,
+      "learning_rate": 7.956947337728221e-07,
+      "loss": 0.5571,
+      "step": 2147
+    },
+    {
+      "epoch": 1.6984763035519936,
+      "grad_norm": 0.2392334027180945,
+      "learning_rate": 7.948810104053976e-07,
+      "loss": 0.5595,
+      "step": 2148
+    },
+    {
+      "epoch": 1.6992678341743346,
+      "grad_norm": 0.24978335197145127,
+      "learning_rate": 7.940674287973535e-07,
+      "loss": 0.567,
+      "step": 2149
+    },
+    {
+      "epoch": 1.7000593647966755,
+      "grad_norm": 0.23399604183084974,
+      "learning_rate": 7.932539895109621e-07,
+      "loss": 0.541,
+      "step": 2150
+    },
+    {
+      "epoch": 1.7008508954190167,
+      "grad_norm": 0.24858612960019044,
+      "learning_rate": 7.924406931083978e-07,
+      "loss": 0.5496,
+      "step": 2151
+    },
+    {
+      "epoch": 1.7016424260413574,
+      "grad_norm": 0.25648218351749696,
+      "learning_rate": 7.916275401517368e-07,
+      "loss": 0.5654,
+      "step": 2152
+    },
+    {
+      "epoch": 1.7024339566636986,
+      "grad_norm": 0.23677924634192826,
+      "learning_rate": 7.908145312029546e-07,
+      "loss": 0.5546,
+      "step": 2153
+    },
+    {
+      "epoch": 1.7032254872860393,
+      "grad_norm": 0.24112980505116388,
+      "learning_rate": 7.900016668239289e-07,
+      "loss": 0.5431,
+      "step": 2154
+    },
+    {
+      "epoch": 1.7040170179083804,
+      "grad_norm": 0.2409520761415017,
+      "learning_rate": 7.891889475764369e-07,
+      "loss": 0.5485,
+      "step": 2155
+    },
+    {
+      "epoch": 1.7048085485307212,
+      "grad_norm": 0.24313335964792243,
+      "learning_rate": 7.88376374022155e-07,
+      "loss": 0.5582,
+      "step": 2156
+    },
+    {
+      "epoch": 1.7056000791530623,
+      "grad_norm": 0.24219718663063755,
+      "learning_rate": 7.875639467226599e-07,
+      "loss": 0.5618,
+      "step": 2157
+    },
+    {
+      "epoch": 1.706391609775403,
+      "grad_norm": 0.23917136316741616,
+      "learning_rate": 7.867516662394258e-07,
+      "loss": 0.5447,
+      "step": 2158
+    },
+    {
+      "epoch": 1.7071831403977442,
+      "grad_norm": 0.2446518091624776,
+      "learning_rate": 7.85939533133827e-07,
+      "loss": 0.5489,
+      "step": 2159
+    },
+    {
+      "epoch": 1.707974671020085,
+      "grad_norm": 0.2462233916332169,
+      "learning_rate": 7.851275479671346e-07,
+      "loss": 0.5643,
+      "step": 2160
+    },
+    {
+      "epoch": 1.708766201642426,
+      "grad_norm": 0.25948166380731974,
+      "learning_rate": 7.843157113005184e-07,
+      "loss": 0.555,
+      "step": 2161
+    },
+    {
+      "epoch": 1.709557732264767,
+      "grad_norm": 0.23802898169647893,
+      "learning_rate": 7.835040236950455e-07,
+      "loss": 0.5491,
+      "step": 2162
+    },
+    {
+      "epoch": 1.710349262887108,
+      "grad_norm": 0.2449931462774046,
+      "learning_rate": 7.82692485711679e-07,
+      "loss": 0.5539,
+      "step": 2163
+    },
+    {
+      "epoch": 1.711140793509449,
+      "grad_norm": 0.2370948135147026,
+      "learning_rate": 7.8188109791128e-07,
+      "loss": 0.5529,
+      "step": 2164
+    },
+    {
+      "epoch": 1.7119323241317899,
+      "grad_norm": 0.24817303915339273,
+      "learning_rate": 7.810698608546046e-07,
+      "loss": 0.5572,
+      "step": 2165
+    },
+    {
+      "epoch": 1.7127238547541308,
+      "grad_norm": 0.23626202654201317,
+      "learning_rate": 7.802587751023055e-07,
+      "loss": 0.5571,
+      "step": 2166
+    },
+    {
+      "epoch": 1.7135153853764717,
+      "grad_norm": 0.23842992920460468,
+      "learning_rate": 7.794478412149311e-07,
+      "loss": 0.5556,
+      "step": 2167
+    },
+    {
+      "epoch": 1.7143069159988127,
+      "grad_norm": 0.24572744490954124,
+      "learning_rate": 7.786370597529233e-07,
+      "loss": 0.5509,
+      "step": 2168
+    },
+    {
+      "epoch": 1.7150984466211536,
+      "grad_norm": 0.23783734579789367,
+      "learning_rate": 7.778264312766203e-07,
+      "loss": 0.5642,
+      "step": 2169
+    },
+    {
+      "epoch": 1.7158899772434946,
+      "grad_norm": 0.24430720648740292,
+      "learning_rate": 7.770159563462537e-07,
+      "loss": 0.5657,
+      "step": 2170
+    },
+    {
+      "epoch": 1.7166815078658355,
+      "grad_norm": 0.23201775490562962,
+      "learning_rate": 7.762056355219493e-07,
+      "loss": 0.5535,
+      "step": 2171
+    },
+    {
+      "epoch": 1.7174730384881767,
+      "grad_norm": 0.23706798034542195,
+      "learning_rate": 7.753954693637262e-07,
+      "loss": 0.5505,
+      "step": 2172
+    },
+    {
+      "epoch": 1.7182645691105174,
+      "grad_norm": 0.23849100010630137,
+      "learning_rate": 7.745854584314975e-07,
+      "loss": 0.5614,
+      "step": 2173
+    },
+    {
+      "epoch": 1.7190560997328586,
+      "grad_norm": 0.24251874043226324,
+      "learning_rate": 7.737756032850674e-07,
+      "loss": 0.562,
+      "step": 2174
+    },
+    {
+      "epoch": 1.7198476303551993,
+      "grad_norm": 0.23995004484046845,
+      "learning_rate": 7.729659044841333e-07,
+      "loss": 0.5527,
+      "step": 2175
+    },
+    {
+      "epoch": 1.7206391609775404,
+      "grad_norm": 0.24590505624862533,
+      "learning_rate": 7.721563625882848e-07,
+      "loss": 0.5582,
+      "step": 2176
+    },
+    {
+      "epoch": 1.7214306915998812,
+      "grad_norm": 0.24601574831899525,
+      "learning_rate": 7.713469781570029e-07,
+      "loss": 0.5554,
+      "step": 2177
+    },
+    {
+      "epoch": 1.7222222222222223,
+      "grad_norm": 0.24380639161955578,
+      "learning_rate": 7.705377517496596e-07,
+      "loss": 0.554,
+      "step": 2178
+    },
+    {
+      "epoch": 1.723013752844563,
+      "grad_norm": 0.2407684830091183,
+      "learning_rate": 7.697286839255178e-07,
+      "loss": 0.5533,
+      "step": 2179
+    },
+    {
+      "epoch": 1.7238052834669042,
+      "grad_norm": 0.23927455389202604,
+      "learning_rate": 7.689197752437305e-07,
+      "loss": 0.5408,
+      "step": 2180
+    },
+    {
+      "epoch": 1.724596814089245,
+      "grad_norm": 0.23757612892397442,
+      "learning_rate": 7.68111026263341e-07,
+      "loss": 0.5464,
+      "step": 2181
+    },
+    {
+      "epoch": 1.725388344711586,
+      "grad_norm": 0.24157743935413045,
+      "learning_rate": 7.673024375432827e-07,
+      "loss": 0.5578,
+      "step": 2182
+    },
+    {
+      "epoch": 1.726179875333927,
+      "grad_norm": 0.2340409671932611,
+      "learning_rate": 7.664940096423776e-07,
+      "loss": 0.5564,
+      "step": 2183
+    },
+    {
+      "epoch": 1.726971405956268,
+      "grad_norm": 0.23546781514033982,
+      "learning_rate": 7.656857431193363e-07,
+      "loss": 0.5582,
+      "step": 2184
+    },
+    {
+      "epoch": 1.727762936578609,
+      "grad_norm": 0.24013399422650472,
+      "learning_rate": 7.648776385327585e-07,
+      "loss": 0.5644,
+      "step": 2185
+    },
+    {
+      "epoch": 1.7285544672009499,
+      "grad_norm": 0.24190172048969363,
+      "learning_rate": 7.640696964411315e-07,
+      "loss": 0.5537,
+      "step": 2186
+    },
+    {
+      "epoch": 1.7293459978232908,
+      "grad_norm": 0.24171309768875512,
+      "learning_rate": 7.632619174028313e-07,
+      "loss": 0.5354,
+      "step": 2187
+    },
+    {
+      "epoch": 1.7301375284456317,
+      "grad_norm": 0.24035832391791973,
+      "learning_rate": 7.624543019761201e-07,
+      "loss": 0.5545,
+      "step": 2188
+    },
+    {
+      "epoch": 1.7309290590679727,
+      "grad_norm": 0.2406686377007751,
+      "learning_rate": 7.616468507191471e-07,
+      "loss": 0.5572,
+      "step": 2189
+    },
+    {
+      "epoch": 1.7317205896903136,
+      "grad_norm": 0.2427060129926691,
+      "learning_rate": 7.608395641899486e-07,
+      "loss": 0.559,
+      "step": 2190
+    },
+    {
+      "epoch": 1.7325121203126546,
+      "grad_norm": 0.23709023805563265,
+      "learning_rate": 7.600324429464466e-07,
+      "loss": 0.555,
+      "step": 2191
+    },
+    {
+      "epoch": 1.7333036509349955,
+      "grad_norm": 0.24132054975928716,
+      "learning_rate": 7.592254875464496e-07,
+      "loss": 0.5638,
+      "step": 2192
+    },
+    {
+      "epoch": 1.7340951815573367,
+      "grad_norm": 0.24230958726738466,
+      "learning_rate": 7.584186985476504e-07,
+      "loss": 0.5615,
+      "step": 2193
+    },
+    {
+      "epoch": 1.7348867121796774,
+      "grad_norm": 0.23703804551595645,
+      "learning_rate": 7.576120765076279e-07,
+      "loss": 0.5619,
+      "step": 2194
+    },
+    {
+      "epoch": 1.7356782428020185,
+      "grad_norm": 0.24522677182745897,
+      "learning_rate": 7.568056219838444e-07,
+      "loss": 0.5613,
+      "step": 2195
+    },
+    {
+      "epoch": 1.7364697734243593,
+      "grad_norm": 0.24551047082297947,
+      "learning_rate": 7.559993355336473e-07,
+      "loss": 0.5777,
+      "step": 2196
+    },
+    {
+      "epoch": 1.7372613040467004,
+      "grad_norm": 0.24554824554869178,
+      "learning_rate": 7.55193217714268e-07,
+      "loss": 0.557,
+      "step": 2197
+    },
+    {
+      "epoch": 1.7380528346690411,
+      "grad_norm": 0.24293205701393464,
+      "learning_rate": 7.543872690828206e-07,
+      "loss": 0.5303,
+      "step": 2198
+    },
+    {
+      "epoch": 1.7388443652913823,
+      "grad_norm": 0.24302354125452177,
+      "learning_rate": 7.535814901963031e-07,
+      "loss": 0.5454,
+      "step": 2199
+    },
+    {
+      "epoch": 1.739635895913723,
+      "grad_norm": 0.2431861330386916,
+      "learning_rate": 7.527758816115953e-07,
+      "loss": 0.5453,
+      "step": 2200
+    },
+    {
+      "epoch": 1.7404274265360642,
+      "grad_norm": 0.23923352331044176,
+      "learning_rate": 7.519704438854598e-07,
+      "loss": 0.5636,
+      "step": 2201
+    },
+    {
+      "epoch": 1.741218957158405,
+      "grad_norm": 0.23972811842788172,
+      "learning_rate": 7.511651775745413e-07,
+      "loss": 0.5326,
+      "step": 2202
+    },
+    {
+      "epoch": 1.742010487780746,
+      "grad_norm": 0.24985436390755797,
+      "learning_rate": 7.503600832353656e-07,
+      "loss": 0.5644,
+      "step": 2203
+    },
+    {
+      "epoch": 1.742802018403087,
+      "grad_norm": 0.2424396509739881,
+      "learning_rate": 7.495551614243404e-07,
+      "loss": 0.5453,
+      "step": 2204
+    },
+    {
+      "epoch": 1.743593549025428,
+      "grad_norm": 0.24083600712817166,
+      "learning_rate": 7.487504126977528e-07,
+      "loss": 0.5458,
+      "step": 2205
+    },
+    {
+      "epoch": 1.744385079647769,
+      "grad_norm": 0.23736625232490574,
+      "learning_rate": 7.479458376117715e-07,
+      "loss": 0.5559,
+      "step": 2206
+    },
+    {
+      "epoch": 1.7451766102701098,
+      "grad_norm": 0.2496008231762276,
+      "learning_rate": 7.471414367224449e-07,
+      "loss": 0.5501,
+      "step": 2207
+    },
+    {
+      "epoch": 1.7459681408924508,
+      "grad_norm": 0.2416875135646901,
+      "learning_rate": 7.463372105857008e-07,
+      "loss": 0.5551,
+      "step": 2208
+    },
+    {
+      "epoch": 1.7467596715147917,
+      "grad_norm": 0.23910205162063805,
+      "learning_rate": 7.455331597573467e-07,
+      "loss": 0.5546,
+      "step": 2209
+    },
+    {
+      "epoch": 1.7475512021371327,
+      "grad_norm": 0.23642334161783982,
+      "learning_rate": 7.447292847930677e-07,
+      "loss": 0.5406,
+      "step": 2210
+    },
+    {
+      "epoch": 1.7483427327594736,
+      "grad_norm": 0.2402782176963215,
+      "learning_rate": 7.439255862484288e-07,
+      "loss": 0.5556,
+      "step": 2211
+    },
+    {
+      "epoch": 1.7491342633818145,
+      "grad_norm": 0.24319025847672862,
+      "learning_rate": 7.431220646788728e-07,
+      "loss": 0.5604,
+      "step": 2212
+    },
+    {
+      "epoch": 1.7499257940041555,
+      "grad_norm": 0.24183606139928246,
+      "learning_rate": 7.423187206397194e-07,
+      "loss": 0.561,
+      "step": 2213
+    },
+    {
+      "epoch": 1.7507173246264967,
+      "grad_norm": 0.23647144330481903,
+      "learning_rate": 7.415155546861667e-07,
+      "loss": 0.5508,
+      "step": 2214
+    },
+    {
+      "epoch": 1.7515088552488374,
+      "grad_norm": 0.24047185581119493,
+      "learning_rate": 7.407125673732884e-07,
+      "loss": 0.5695,
+      "step": 2215
+    },
+    {
+      "epoch": 1.7523003858711785,
+      "grad_norm": 0.24779995107088315,
+      "learning_rate": 7.399097592560359e-07,
+      "loss": 0.5609,
+      "step": 2216
+    },
+    {
+      "epoch": 1.7530919164935193,
+      "grad_norm": 0.23624790745492177,
+      "learning_rate": 7.39107130889236e-07,
+      "loss": 0.5554,
+      "step": 2217
+    },
+    {
+      "epoch": 1.7538834471158604,
+      "grad_norm": 0.24288340429453822,
+      "learning_rate": 7.38304682827592e-07,
+      "loss": 0.5552,
+      "step": 2218
+    },
+    {
+      "epoch": 1.7546749777382011,
+      "grad_norm": 0.23878211179399783,
+      "learning_rate": 7.37502415625682e-07,
+      "loss": 0.5525,
+      "step": 2219
+    },
+    {
+      "epoch": 1.7554665083605423,
+      "grad_norm": 0.24104183274827198,
+      "learning_rate": 7.367003298379597e-07,
+      "loss": 0.549,
+      "step": 2220
+    },
+    {
+      "epoch": 1.756258038982883,
+      "grad_norm": 0.23501424265923035,
+      "learning_rate": 7.358984260187526e-07,
+      "loss": 0.5657,
+      "step": 2221
+    },
+    {
+      "epoch": 1.7570495696052242,
+      "grad_norm": 0.2372833396939188,
+      "learning_rate": 7.350967047222627e-07,
+      "loss": 0.5619,
+      "step": 2222
+    },
+    {
+      "epoch": 1.757841100227565,
+      "grad_norm": 0.23553083697812366,
+      "learning_rate": 7.342951665025663e-07,
+      "loss": 0.5366,
+      "step": 2223
+    },
+    {
+      "epoch": 1.758632630849906,
+      "grad_norm": 0.23548253077937908,
+      "learning_rate": 7.334938119136127e-07,
+      "loss": 0.5563,
+      "step": 2224
+    },
+    {
+      "epoch": 1.759424161472247,
+      "grad_norm": 0.23773384962064845,
+      "learning_rate": 7.326926415092252e-07,
+      "loss": 0.55,
+      "step": 2225
+    },
+    {
+      "epoch": 1.760215692094588,
+      "grad_norm": 0.24050851822944924,
+      "learning_rate": 7.318916558430984e-07,
+      "loss": 0.569,
+      "step": 2226
+    },
+    {
+      "epoch": 1.7610072227169289,
+      "grad_norm": 0.23953819099930582,
+      "learning_rate": 7.310908554687998e-07,
+      "loss": 0.5495,
+      "step": 2227
+    },
+    {
+      "epoch": 1.7617987533392698,
+      "grad_norm": 0.24258422612890876,
+      "learning_rate": 7.302902409397693e-07,
+      "loss": 0.5422,
+      "step": 2228
+    },
+    {
+      "epoch": 1.7625902839616108,
+      "grad_norm": 0.23696866660215898,
+      "learning_rate": 7.294898128093179e-07,
+      "loss": 0.5662,
+      "step": 2229
+    },
+    {
+      "epoch": 1.7633818145839517,
+      "grad_norm": 0.2361512263061421,
+      "learning_rate": 7.286895716306284e-07,
+      "loss": 0.5499,
+      "step": 2230
+    },
+    {
+      "epoch": 1.7641733452062927,
+      "grad_norm": 0.24095374873912595,
+      "learning_rate": 7.278895179567533e-07,
+      "loss": 0.5498,
+      "step": 2231
+    },
+    {
+      "epoch": 1.7649648758286336,
+      "grad_norm": 0.23322150218966986,
+      "learning_rate": 7.27089652340616e-07,
+      "loss": 0.5526,
+      "step": 2232
+    },
+    {
+      "epoch": 1.7657564064509745,
+      "grad_norm": 0.24547347021758917,
+      "learning_rate": 7.262899753350104e-07,
+      "loss": 0.5477,
+      "step": 2233
+    },
+    {
+      "epoch": 1.7665479370733155,
+      "grad_norm": 0.2428990539423586,
+      "learning_rate": 7.254904874925999e-07,
+      "loss": 0.5612,
+      "step": 2234
+    },
+    {
+      "epoch": 1.7673394676956566,
+      "grad_norm": 0.24596385189491438,
+      "learning_rate": 7.246911893659167e-07,
+      "loss": 0.5489,
+      "step": 2235
+    },
+    {
+      "epoch": 1.7681309983179974,
+      "grad_norm": 0.24544528922836092,
+      "learning_rate": 7.238920815073621e-07,
+      "loss": 0.5585,
+      "step": 2236
+    },
+    {
+      "epoch": 1.7689225289403385,
+      "grad_norm": 0.2380119745734608,
+      "learning_rate": 7.230931644692057e-07,
+      "loss": 0.5519,
+      "step": 2237
+    },
+    {
+      "epoch": 1.7697140595626792,
+      "grad_norm": 0.23862840909204722,
+      "learning_rate": 7.222944388035859e-07,
+      "loss": 0.57,
+      "step": 2238
+    },
+    {
+      "epoch": 1.7705055901850204,
+      "grad_norm": 0.23283778262407775,
+      "learning_rate": 7.214959050625085e-07,
+      "loss": 0.5456,
+      "step": 2239
+    },
+    {
+      "epoch": 1.7712971208073611,
+      "grad_norm": 0.2374660754544533,
+      "learning_rate": 7.20697563797846e-07,
+      "loss": 0.5622,
+      "step": 2240
+    },
+    {
+      "epoch": 1.7720886514297023,
+      "grad_norm": 0.23978966065378568,
+      "learning_rate": 7.198994155613395e-07,
+      "loss": 0.5546,
+      "step": 2241
+    },
+    {
+      "epoch": 1.772880182052043,
+      "grad_norm": 0.2514758149552116,
+      "learning_rate": 7.191014609045945e-07,
+      "loss": 0.5608,
+      "step": 2242
+    },
+    {
+      "epoch": 1.7736717126743842,
+      "grad_norm": 0.239362411396253,
+      "learning_rate": 7.183037003790842e-07,
+      "loss": 0.5596,
+      "step": 2243
+    },
+    {
+      "epoch": 1.774463243296725,
+      "grad_norm": 0.23869763089069795,
+      "learning_rate": 7.175061345361479e-07,
+      "loss": 0.5588,
+      "step": 2244
+    },
+    {
+      "epoch": 1.775254773919066,
+      "grad_norm": 0.22948716001022224,
+      "learning_rate": 7.167087639269889e-07,
+      "loss": 0.5463,
+      "step": 2245
+    },
+    {
+      "epoch": 1.776046304541407,
+      "grad_norm": 0.23597151449855,
+      "learning_rate": 7.159115891026773e-07,
+      "loss": 0.5551,
+      "step": 2246
+    },
+    {
+      "epoch": 1.776837835163748,
+      "grad_norm": 0.23445464954183326,
+      "learning_rate": 7.151146106141461e-07,
+      "loss": 0.5483,
+      "step": 2247
+    },
+    {
+      "epoch": 1.7776293657860889,
+      "grad_norm": 0.24079164518017426,
+      "learning_rate": 7.143178290121941e-07,
+      "loss": 0.5798,
+      "step": 2248
+    },
+    {
+      "epoch": 1.7784208964084298,
+      "grad_norm": 0.24548032825300617,
+      "learning_rate": 7.135212448474836e-07,
+      "loss": 0.5673,
+      "step": 2249
+    },
+    {
+      "epoch": 1.7792124270307708,
+      "grad_norm": 0.24462915858719766,
+      "learning_rate": 7.127248586705397e-07,
+      "loss": 0.564,
+      "step": 2250
+    },
+    {
+      "epoch": 1.7800039576531117,
+      "grad_norm": 0.25313748158719196,
+      "learning_rate": 7.11928671031752e-07,
+      "loss": 0.5662,
+      "step": 2251
+    },
+    {
+      "epoch": 1.7807954882754526,
+      "grad_norm": 0.2509929007772327,
+      "learning_rate": 7.111326824813716e-07,
+      "loss": 0.5525,
+      "step": 2252
+    },
+    {
+      "epoch": 1.7815870188977936,
+      "grad_norm": 0.2392518481507003,
+      "learning_rate": 7.103368935695127e-07,
+      "loss": 0.5615,
+      "step": 2253
+    },
+    {
+      "epoch": 1.7823785495201345,
+      "grad_norm": 0.23491375049248236,
+      "learning_rate": 7.095413048461516e-07,
+      "loss": 0.5434,
+      "step": 2254
+    },
+    {
+      "epoch": 1.7831700801424755,
+      "grad_norm": 0.24024078798241236,
+      "learning_rate": 7.087459168611259e-07,
+      "loss": 0.5448,
+      "step": 2255
+    },
+    {
+      "epoch": 1.7839616107648166,
+      "grad_norm": 0.2439628732463763,
+      "learning_rate": 7.079507301641348e-07,
+      "loss": 0.5556,
+      "step": 2256
+    },
+    {
+      "epoch": 1.7847531413871573,
+      "grad_norm": 0.23730279363087756,
+      "learning_rate": 7.071557453047378e-07,
+      "loss": 0.5615,
+      "step": 2257
+    },
+    {
+      "epoch": 1.7855446720094985,
+      "grad_norm": 0.23959308056939885,
+      "learning_rate": 7.063609628323555e-07,
+      "loss": 0.5593,
+      "step": 2258
+    },
+    {
+      "epoch": 1.7863362026318392,
+      "grad_norm": 0.23250071223661176,
+      "learning_rate": 7.055663832962688e-07,
+      "loss": 0.5662,
+      "step": 2259
+    },
+    {
+      "epoch": 1.7871277332541804,
+      "grad_norm": 0.2376050572540389,
+      "learning_rate": 7.047720072456175e-07,
+      "loss": 0.5508,
+      "step": 2260
+    },
+    {
+      "epoch": 1.7879192638765211,
+      "grad_norm": 0.23641651237842087,
+      "learning_rate": 7.039778352294015e-07,
+      "loss": 0.5733,
+      "step": 2261
+    },
+    {
+      "epoch": 1.7887107944988623,
+      "grad_norm": 0.23982577133592545,
+      "learning_rate": 7.031838677964798e-07,
+      "loss": 0.5654,
+      "step": 2262
+    },
+    {
+      "epoch": 1.789502325121203,
+      "grad_norm": 0.2521637985444799,
+      "learning_rate": 7.023901054955691e-07,
+      "loss": 0.5689,
+      "step": 2263
+    },
+    {
+      "epoch": 1.7902938557435442,
+      "grad_norm": 0.24211730729014924,
+      "learning_rate": 7.015965488752451e-07,
+      "loss": 0.558,
+      "step": 2264
+    },
+    {
+      "epoch": 1.7910853863658849,
+      "grad_norm": 0.2496910755279375,
+      "learning_rate": 7.00803198483941e-07,
+      "loss": 0.5686,
+      "step": 2265
+    },
+    {
+      "epoch": 1.791876916988226,
+      "grad_norm": 0.23018708943321897,
+      "learning_rate": 7.000100548699476e-07,
+      "loss": 0.547,
+      "step": 2266
+    },
+    {
+      "epoch": 1.792668447610567,
+      "grad_norm": 0.2311351296522197,
+      "learning_rate": 6.992171185814137e-07,
+      "loss": 0.5619,
+      "step": 2267
+    },
+    {
+      "epoch": 1.793459978232908,
+      "grad_norm": 0.24818603262501357,
+      "learning_rate": 6.984243901663429e-07,
+      "loss": 0.5645,
+      "step": 2268
+    },
+    {
+      "epoch": 1.7942515088552489,
+      "grad_norm": 0.23417529603719672,
+      "learning_rate": 6.976318701725963e-07,
+      "loss": 0.5504,
+      "step": 2269
+    },
+    {
+      "epoch": 1.7950430394775898,
+      "grad_norm": 0.2513033097750515,
+      "learning_rate": 6.96839559147891e-07,
+      "loss": 0.5433,
+      "step": 2270
+    },
+    {
+      "epoch": 1.7958345700999307,
+      "grad_norm": 0.24566331048313653,
+      "learning_rate": 6.960474576397994e-07,
+      "loss": 0.5519,
+      "step": 2271
+    },
+    {
+      "epoch": 1.7966261007222717,
+      "grad_norm": 0.23738942204593105,
+      "learning_rate": 6.9525556619575e-07,
+      "loss": 0.5337,
+      "step": 2272
+    },
+    {
+      "epoch": 1.7974176313446126,
+      "grad_norm": 0.2366066672752667,
+      "learning_rate": 6.944638853630243e-07,
+      "loss": 0.558,
+      "step": 2273
+    },
+    {
+      "epoch": 1.7982091619669536,
+      "grad_norm": 0.2510111639222807,
+      "learning_rate": 6.936724156887595e-07,
+      "loss": 0.5508,
+      "step": 2274
+    },
+    {
+      "epoch": 1.7990006925892945,
+      "grad_norm": 0.23754972873567884,
+      "learning_rate": 6.928811577199467e-07,
+      "loss": 0.5647,
+      "step": 2275
+    },
+    {
+      "epoch": 1.7997922232116355,
+      "grad_norm": 0.238934512310531,
+      "learning_rate": 6.920901120034308e-07,
+      "loss": 0.5667,
+      "step": 2276
+    },
+    {
+      "epoch": 1.8005837538339764,
+      "grad_norm": 0.24434541535275991,
+      "learning_rate": 6.912992790859101e-07,
+      "loss": 0.5359,
+      "step": 2277
+    },
+    {
+      "epoch": 1.8013752844563173,
+      "grad_norm": 0.23673331978972978,
+      "learning_rate": 6.905086595139351e-07,
+      "loss": 0.5524,
+      "step": 2278
+    },
+    {
+      "epoch": 1.8021668150786585,
+      "grad_norm": 0.23799036271887128,
+      "learning_rate": 6.897182538339093e-07,
+      "loss": 0.5653,
+      "step": 2279
+    },
+    {
+      "epoch": 1.8029583457009992,
+      "grad_norm": 0.2430485402993927,
+      "learning_rate": 6.889280625920886e-07,
+      "loss": 0.5488,
+      "step": 2280
+    },
+    {
+      "epoch": 1.8037498763233404,
+      "grad_norm": 0.24479649312428692,
+      "learning_rate": 6.881380863345805e-07,
+      "loss": 0.5624,
+      "step": 2281
+    },
+    {
+      "epoch": 1.804541406945681,
+      "grad_norm": 0.24286408865509948,
+      "learning_rate": 6.87348325607344e-07,
+      "loss": 0.5424,
+      "step": 2282
+    },
+    {
+      "epoch": 1.8053329375680223,
+      "grad_norm": 0.2409483758178677,
+      "learning_rate": 6.865587809561893e-07,
+      "loss": 0.556,
+      "step": 2283
+    },
+    {
+      "epoch": 1.806124468190363,
+      "grad_norm": 0.24298280116705154,
+      "learning_rate": 6.857694529267763e-07,
+      "loss": 0.548,
+      "step": 2284
+    },
+    {
+      "epoch": 1.8069159988127041,
+      "grad_norm": 0.23905238009781168,
+      "learning_rate": 6.849803420646165e-07,
+      "loss": 0.5525,
+      "step": 2285
+    },
+    {
+      "epoch": 1.8077075294350449,
+      "grad_norm": 0.23604023844260322,
+      "learning_rate": 6.841914489150708e-07,
+      "loss": 0.5501,
+      "step": 2286
+    },
+    {
+      "epoch": 1.808499060057386,
+      "grad_norm": 0.23906787494105905,
+      "learning_rate": 6.834027740233492e-07,
+      "loss": 0.5565,
+      "step": 2287
+    },
+    {
+      "epoch": 1.809290590679727,
+      "grad_norm": 0.24029615943963029,
+      "learning_rate": 6.826143179345118e-07,
+      "loss": 0.5571,
+      "step": 2288
+    },
+    {
+      "epoch": 1.810082121302068,
+      "grad_norm": 0.24357915428979005,
+      "learning_rate": 6.818260811934664e-07,
+      "loss": 0.557,
+      "step": 2289
+    },
+    {
+      "epoch": 1.8108736519244089,
+      "grad_norm": 0.2498477135697297,
+      "learning_rate": 6.810380643449698e-07,
+      "loss": 0.5488,
+      "step": 2290
+    },
+    {
+      "epoch": 1.8116651825467498,
+      "grad_norm": 0.2459033357611074,
+      "learning_rate": 6.80250267933627e-07,
+      "loss": 0.5605,
+      "step": 2291
+    },
+    {
+      "epoch": 1.8124567131690907,
+      "grad_norm": 0.23632460590930843,
+      "learning_rate": 6.794626925038905e-07,
+      "loss": 0.5761,
+      "step": 2292
+    },
+    {
+      "epoch": 1.8132482437914317,
+      "grad_norm": 0.24900934919267886,
+      "learning_rate": 6.786753386000599e-07,
+      "loss": 0.5698,
+      "step": 2293
+    },
+    {
+      "epoch": 1.8140397744137726,
+      "grad_norm": 0.24210728450429736,
+      "learning_rate": 6.778882067662815e-07,
+      "loss": 0.5535,
+      "step": 2294
+    },
+    {
+      "epoch": 1.8148313050361136,
+      "grad_norm": 0.24209896764555133,
+      "learning_rate": 6.771012975465484e-07,
+      "loss": 0.557,
+      "step": 2295
+    },
+    {
+      "epoch": 1.8156228356584545,
+      "grad_norm": 0.23664483800454386,
+      "learning_rate": 6.763146114847006e-07,
+      "loss": 0.5587,
+      "step": 2296
+    },
+    {
+      "epoch": 1.8164143662807954,
+      "grad_norm": 0.23539531716987616,
+      "learning_rate": 6.755281491244224e-07,
+      "loss": 0.5552,
+      "step": 2297
+    },
+    {
+      "epoch": 1.8172058969031364,
+      "grad_norm": 0.23970394453645905,
+      "learning_rate": 6.747419110092449e-07,
+      "loss": 0.5496,
+      "step": 2298
+    },
+    {
+      "epoch": 1.8179974275254773,
+      "grad_norm": 0.23951359745776754,
+      "learning_rate": 6.73955897682543e-07,
+      "loss": 0.5459,
+      "step": 2299
+    },
+    {
+      "epoch": 1.8187889581478185,
+      "grad_norm": 0.23611766505624682,
+      "learning_rate": 6.731701096875368e-07,
+      "loss": 0.5638,
+      "step": 2300
+    },
+    {
+      "epoch": 1.8195804887701592,
+      "grad_norm": 0.24096794686089035,
+      "learning_rate": 6.723845475672914e-07,
+      "loss": 0.5517,
+      "step": 2301
+    },
+    {
+      "epoch": 1.8203720193925004,
+      "grad_norm": 0.2387576102864395,
+      "learning_rate": 6.715992118647142e-07,
+      "loss": 0.5554,
+      "step": 2302
+    },
+    {
+      "epoch": 1.821163550014841,
+      "grad_norm": 0.23919209056115712,
+      "learning_rate": 6.708141031225574e-07,
+      "loss": 0.5702,
+      "step": 2303
+    },
+    {
+      "epoch": 1.8219550806371823,
+      "grad_norm": 0.23916623288866273,
+      "learning_rate": 6.700292218834164e-07,
+      "loss": 0.5597,
+      "step": 2304
+    },
+    {
+      "epoch": 1.822746611259523,
+      "grad_norm": 0.23514975978391578,
+      "learning_rate": 6.692445686897281e-07,
+      "loss": 0.5449,
+      "step": 2305
+    },
+    {
+      "epoch": 1.8235381418818641,
+      "grad_norm": 0.235941676409706,
+      "learning_rate": 6.684601440837731e-07,
+      "loss": 0.5506,
+      "step": 2306
+    },
+    {
+      "epoch": 1.8243296725042049,
+      "grad_norm": 0.24204959008901106,
+      "learning_rate": 6.676759486076735e-07,
+      "loss": 0.5673,
+      "step": 2307
+    },
+    {
+      "epoch": 1.825121203126546,
+      "grad_norm": 0.2409770722540558,
+      "learning_rate": 6.668919828033929e-07,
+      "loss": 0.5469,
+      "step": 2308
+    },
+    {
+      "epoch": 1.825912733748887,
+      "grad_norm": 0.23392382638544185,
+      "learning_rate": 6.661082472127368e-07,
+      "loss": 0.5621,
+      "step": 2309
+    },
+    {
+      "epoch": 1.826704264371228,
+      "grad_norm": 0.23542651782289323,
+      "learning_rate": 6.653247423773507e-07,
+      "loss": 0.5548,
+      "step": 2310
+    },
+    {
+      "epoch": 1.8274957949935688,
+      "grad_norm": 0.23736117326812825,
+      "learning_rate": 6.645414688387212e-07,
+      "loss": 0.5618,
+      "step": 2311
+    },
+    {
+      "epoch": 1.8282873256159098,
+      "grad_norm": 0.24155604449826412,
+      "learning_rate": 6.637584271381749e-07,
+      "loss": 0.5447,
+      "step": 2312
+    },
+    {
+      "epoch": 1.8290788562382507,
+      "grad_norm": 0.23430895391792395,
+      "learning_rate": 6.629756178168783e-07,
+      "loss": 0.5708,
+      "step": 2313
+    },
+    {
+      "epoch": 1.8298703868605917,
+      "grad_norm": 0.24636770889127307,
+      "learning_rate": 6.621930414158374e-07,
+      "loss": 0.5645,
+      "step": 2314
+    },
+    {
+      "epoch": 1.8306619174829326,
+      "grad_norm": 0.23744470174473464,
+      "learning_rate": 6.614106984758965e-07,
+      "loss": 0.5563,
+      "step": 2315
+    },
+    {
+      "epoch": 1.8314534481052736,
+      "grad_norm": 0.2356690086717818,
+      "learning_rate": 6.606285895377397e-07,
+      "loss": 0.564,
+      "step": 2316
+    },
+    {
+      "epoch": 1.8322449787276145,
+      "grad_norm": 0.23488291816006648,
+      "learning_rate": 6.59846715141888e-07,
+      "loss": 0.5422,
+      "step": 2317
+    },
+    {
+      "epoch": 1.8330365093499554,
+      "grad_norm": 0.24926387103850173,
+      "learning_rate": 6.590650758287018e-07,
+      "loss": 0.5474,
+      "step": 2318
+    },
+    {
+      "epoch": 1.8338280399722964,
+      "grad_norm": 0.2341450028715121,
+      "learning_rate": 6.582836721383784e-07,
+      "loss": 0.5407,
+      "step": 2319
+    },
+    {
+      "epoch": 1.8346195705946373,
+      "grad_norm": 0.24123193482149877,
+      "learning_rate": 6.575025046109515e-07,
+      "loss": 0.5595,
+      "step": 2320
+    },
+    {
+      "epoch": 1.8354111012169785,
+      "grad_norm": 0.2389911607109484,
+      "learning_rate": 6.567215737862923e-07,
+      "loss": 0.5613,
+      "step": 2321
+    },
+    {
+      "epoch": 1.8362026318393192,
+      "grad_norm": 0.23176945932886384,
+      "learning_rate": 6.559408802041087e-07,
+      "loss": 0.55,
+      "step": 2322
+    },
+    {
+      "epoch": 1.8369941624616604,
+      "grad_norm": 0.24386516463106725,
+      "learning_rate": 6.551604244039441e-07,
+      "loss": 0.5684,
+      "step": 2323
+    },
+    {
+      "epoch": 1.837785693084001,
+      "grad_norm": 0.24491123056474223,
+      "learning_rate": 6.54380206925178e-07,
+      "loss": 0.5697,
+      "step": 2324
+    },
+    {
+      "epoch": 1.8385772237063422,
+      "grad_norm": 0.23959194903187608,
+      "learning_rate": 6.536002283070253e-07,
+      "loss": 0.5514,
+      "step": 2325
+    },
+    {
+      "epoch": 1.839368754328683,
+      "grad_norm": 0.23343505688377264,
+      "learning_rate": 6.528204890885342e-07,
+      "loss": 0.5588,
+      "step": 2326
+    },
+    {
+      "epoch": 1.8401602849510241,
+      "grad_norm": 0.24011311991343826,
+      "learning_rate": 6.520409898085899e-07,
+      "loss": 0.5658,
+      "step": 2327
+    },
+    {
+      "epoch": 1.8409518155733648,
+      "grad_norm": 0.23997235651902463,
+      "learning_rate": 6.512617310059101e-07,
+      "loss": 0.5457,
+      "step": 2328
+    },
+    {
+      "epoch": 1.841743346195706,
+      "grad_norm": 0.23664962533257064,
+      "learning_rate": 6.504827132190471e-07,
+      "loss": 0.5754,
+      "step": 2329
+    },
+    {
+      "epoch": 1.842534876818047,
+      "grad_norm": 0.2408846776707785,
+      "learning_rate": 6.497039369863866e-07,
+      "loss": 0.5515,
+      "step": 2330
+    },
+    {
+      "epoch": 1.843326407440388,
+      "grad_norm": 0.24200784743574702,
+      "learning_rate": 6.489254028461462e-07,
+      "loss": 0.5587,
+      "step": 2331
+    },
+    {
+      "epoch": 1.8441179380627288,
+      "grad_norm": 0.23869103910857442,
+      "learning_rate": 6.481471113363776e-07,
+      "loss": 0.5485,
+      "step": 2332
+    },
+    {
+      "epoch": 1.8449094686850698,
+      "grad_norm": 0.23916108197027652,
+      "learning_rate": 6.473690629949644e-07,
+      "loss": 0.5581,
+      "step": 2333
+    },
+    {
+      "epoch": 1.8457009993074107,
+      "grad_norm": 0.24265705116236377,
+      "learning_rate": 6.465912583596221e-07,
+      "loss": 0.5509,
+      "step": 2334
+    },
+    {
+      "epoch": 1.8464925299297517,
+      "grad_norm": 0.23998194252671326,
+      "learning_rate": 6.458136979678978e-07,
+      "loss": 0.554,
+      "step": 2335
+    },
+    {
+      "epoch": 1.8472840605520926,
+      "grad_norm": 0.24258965204353683,
+      "learning_rate": 6.450363823571691e-07,
+      "loss": 0.5508,
+      "step": 2336
+    },
+    {
+      "epoch": 1.8480755911744335,
+      "grad_norm": 0.24387263782873647,
+      "learning_rate": 6.442593120646456e-07,
+      "loss": 0.5532,
+      "step": 2337
+    },
+    {
+      "epoch": 1.8488671217967745,
+      "grad_norm": 0.24070305923056098,
+      "learning_rate": 6.434824876273668e-07,
+      "loss": 0.5654,
+      "step": 2338
+    },
+    {
+      "epoch": 1.8496586524191154,
+      "grad_norm": 0.23472031609301547,
+      "learning_rate": 6.427059095822019e-07,
+      "loss": 0.5594,
+      "step": 2339
+    },
+    {
+      "epoch": 1.8504501830414564,
+      "grad_norm": 0.23574981654708868,
+      "learning_rate": 6.419295784658511e-07,
+      "loss": 0.5433,
+      "step": 2340
+    },
+    {
+      "epoch": 1.8512417136637973,
+      "grad_norm": 0.2455516349980044,
+      "learning_rate": 6.411534948148421e-07,
+      "loss": 0.5616,
+      "step": 2341
+    },
+    {
+      "epoch": 1.8520332442861385,
+      "grad_norm": 0.23866347416769718,
+      "learning_rate": 6.403776591655328e-07,
+      "loss": 0.5549,
+      "step": 2342
+    },
+    {
+      "epoch": 1.8528247749084792,
+      "grad_norm": 0.23097699749081696,
+      "learning_rate": 6.396020720541097e-07,
+      "loss": 0.5491,
+      "step": 2343
+    },
+    {
+      "epoch": 1.8536163055308204,
+      "grad_norm": 0.23200056818734432,
+      "learning_rate": 6.38826734016587e-07,
+      "loss": 0.5645,
+      "step": 2344
+    },
+    {
+      "epoch": 1.854407836153161,
+      "grad_norm": 0.2417124508988092,
+      "learning_rate": 6.380516455888072e-07,
+      "loss": 0.5527,
+      "step": 2345
+    },
+    {
+      "epoch": 1.8551993667755022,
+      "grad_norm": 0.24046045058609652,
+      "learning_rate": 6.372768073064406e-07,
+      "loss": 0.5409,
+      "step": 2346
+    },
+    {
+      "epoch": 1.855990897397843,
+      "grad_norm": 0.23470534804207058,
+      "learning_rate": 6.365022197049832e-07,
+      "loss": 0.5492,
+      "step": 2347
+    },
+    {
+      "epoch": 1.8567824280201841,
+      "grad_norm": 0.2390242999424857,
+      "learning_rate": 6.357278833197594e-07,
+      "loss": 0.5525,
+      "step": 2348
+    },
+    {
+      "epoch": 1.8575739586425248,
+      "grad_norm": 0.2388123870630192,
+      "learning_rate": 6.349537986859189e-07,
+      "loss": 0.5519,
+      "step": 2349
+    },
+    {
+      "epoch": 1.858365489264866,
+      "grad_norm": 0.2354214384007597,
+      "learning_rate": 6.341799663384379e-07,
+      "loss": 0.5506,
+      "step": 2350
+    },
+    {
+      "epoch": 1.859157019887207,
+      "grad_norm": 0.24107434793571558,
+      "learning_rate": 6.334063868121188e-07,
+      "loss": 0.5609,
+      "step": 2351
+    },
+    {
+      "epoch": 1.8599485505095479,
+      "grad_norm": 0.24345399329835843,
+      "learning_rate": 6.326330606415875e-07,
+      "loss": 0.5561,
+      "step": 2352
+    },
+    {
+      "epoch": 1.8607400811318888,
+      "grad_norm": 0.23938860381305677,
+      "learning_rate": 6.318599883612968e-07,
+      "loss": 0.5533,
+      "step": 2353
+    },
+    {
+      "epoch": 1.8615316117542298,
+      "grad_norm": 0.24017682844728316,
+      "learning_rate": 6.310871705055224e-07,
+      "loss": 0.5547,
+      "step": 2354
+    },
+    {
+      "epoch": 1.8623231423765707,
+      "grad_norm": 0.2432670533361399,
+      "learning_rate": 6.303146076083654e-07,
+      "loss": 0.5762,
+      "step": 2355
+    },
+    {
+      "epoch": 1.8631146729989116,
+      "grad_norm": 0.2363381041473353,
+      "learning_rate": 6.295423002037507e-07,
+      "loss": 0.5681,
+      "step": 2356
+    },
+    {
+      "epoch": 1.8639062036212526,
+      "grad_norm": 0.24283339079057742,
+      "learning_rate": 6.287702488254252e-07,
+      "loss": 0.5604,
+      "step": 2357
+    },
+    {
+      "epoch": 1.8646977342435935,
+      "grad_norm": 0.2325040714927763,
+      "learning_rate": 6.279984540069602e-07,
+      "loss": 0.5595,
+      "step": 2358
+    },
+    {
+      "epoch": 1.8654892648659345,
+      "grad_norm": 0.24415758673149396,
+      "learning_rate": 6.272269162817494e-07,
+      "loss": 0.5716,
+      "step": 2359
+    },
+    {
+      "epoch": 1.8662807954882754,
+      "grad_norm": 0.2428025755069738,
+      "learning_rate": 6.264556361830084e-07,
+      "loss": 0.5538,
+      "step": 2360
+    },
+    {
+      "epoch": 1.8670723261106164,
+      "grad_norm": 0.23313158379518548,
+      "learning_rate": 6.256846142437757e-07,
+      "loss": 0.5596,
+      "step": 2361
+    },
+    {
+      "epoch": 1.8678638567329573,
+      "grad_norm": 0.23042869370617095,
+      "learning_rate": 6.249138509969101e-07,
+      "loss": 0.5574,
+      "step": 2362
+    },
+    {
+      "epoch": 1.8686553873552985,
+      "grad_norm": 0.24431359866881122,
+      "learning_rate": 6.241433469750926e-07,
+      "loss": 0.5639,
+      "step": 2363
+    },
+    {
+      "epoch": 1.8694469179776392,
+      "grad_norm": 0.23373404423656063,
+      "learning_rate": 6.233731027108244e-07,
+      "loss": 0.5535,
+      "step": 2364
+    },
+    {
+      "epoch": 1.8702384485999803,
+      "grad_norm": 0.25129592335429896,
+      "learning_rate": 6.226031187364278e-07,
+      "loss": 0.5632,
+      "step": 2365
+    },
+    {
+      "epoch": 1.871029979222321,
+      "grad_norm": 0.2305640766257186,
+      "learning_rate": 6.218333955840451e-07,
+      "loss": 0.553,
+      "step": 2366
+    },
+    {
+      "epoch": 1.8718215098446622,
+      "grad_norm": 0.23897738138262625,
+      "learning_rate": 6.210639337856376e-07,
+      "loss": 0.5408,
+      "step": 2367
+    },
+    {
+      "epoch": 1.872613040467003,
+      "grad_norm": 0.2380331124139679,
+      "learning_rate": 6.20294733872987e-07,
+      "loss": 0.5518,
+      "step": 2368
+    },
+    {
+      "epoch": 1.873404571089344,
+      "grad_norm": 0.2406179368363775,
+      "learning_rate": 6.19525796377693e-07,
+      "loss": 0.5719,
+      "step": 2369
+    },
+    {
+      "epoch": 1.8741961017116848,
+      "grad_norm": 0.24157444867831274,
+      "learning_rate": 6.187571218311748e-07,
+      "loss": 0.5535,
+      "step": 2370
+    },
+    {
+      "epoch": 1.874987632334026,
+      "grad_norm": 0.2409576246223222,
+      "learning_rate": 6.179887107646698e-07,
+      "loss": 0.5587,
+      "step": 2371
+    },
+    {
+      "epoch": 1.8757791629563667,
+      "grad_norm": 0.24402319965739602,
+      "learning_rate": 6.172205637092329e-07,
+      "loss": 0.5508,
+      "step": 2372
+    },
+    {
+      "epoch": 1.8765706935787079,
+      "grad_norm": 0.2365119379994907,
+      "learning_rate": 6.164526811957363e-07,
+      "loss": 0.5598,
+      "step": 2373
+    },
+    {
+      "epoch": 1.8773622242010488,
+      "grad_norm": 0.2526132954101458,
+      "learning_rate": 6.156850637548701e-07,
+      "loss": 0.5718,
+      "step": 2374
+    },
+    {
+      "epoch": 1.8781537548233898,
+      "grad_norm": 0.24510438664346818,
+      "learning_rate": 6.149177119171407e-07,
+      "loss": 0.5497,
+      "step": 2375
+    },
+    {
+      "epoch": 1.8789452854457307,
+      "grad_norm": 0.24376795264398032,
+      "learning_rate": 6.141506262128715e-07,
+      "loss": 0.554,
+      "step": 2376
+    },
+    {
+      "epoch": 1.8797368160680716,
+      "grad_norm": 0.23899904359061272,
+      "learning_rate": 6.133838071722017e-07,
+      "loss": 0.5432,
+      "step": 2377
+    },
+    {
+      "epoch": 1.8805283466904126,
+      "grad_norm": 0.24760673460820454,
+      "learning_rate": 6.126172553250852e-07,
+      "loss": 0.5603,
+      "step": 2378
+    },
+    {
+      "epoch": 1.8813198773127535,
+      "grad_norm": 0.24439397727668621,
+      "learning_rate": 6.118509712012927e-07,
+      "loss": 0.5667,
+      "step": 2379
+    },
+    {
+      "epoch": 1.8821114079350945,
+      "grad_norm": 0.23570472393437591,
+      "learning_rate": 6.110849553304091e-07,
+      "loss": 0.5487,
+      "step": 2380
+    },
+    {
+      "epoch": 1.8829029385574354,
+      "grad_norm": 0.2396300432209492,
+      "learning_rate": 6.103192082418347e-07,
+      "loss": 0.5571,
+      "step": 2381
+    },
+    {
+      "epoch": 1.8836944691797763,
+      "grad_norm": 0.2359323187904901,
+      "learning_rate": 6.095537304647832e-07,
+      "loss": 0.5584,
+      "step": 2382
+    },
+    {
+      "epoch": 1.8844859998021173,
+      "grad_norm": 0.24259348781786255,
+      "learning_rate": 6.087885225282818e-07,
+      "loss": 0.5483,
+      "step": 2383
+    },
+    {
+      "epoch": 1.8852775304244584,
+      "grad_norm": 0.23938494203387234,
+      "learning_rate": 6.080235849611726e-07,
+      "loss": 0.5554,
+      "step": 2384
+    },
+    {
+      "epoch": 1.8860690610467992,
+      "grad_norm": 0.23551924646792072,
+      "learning_rate": 6.072589182921095e-07,
+      "loss": 0.5498,
+      "step": 2385
+    },
+    {
+      "epoch": 1.8868605916691403,
+      "grad_norm": 0.2370825388647084,
+      "learning_rate": 6.064945230495604e-07,
+      "loss": 0.5759,
+      "step": 2386
+    },
+    {
+      "epoch": 1.887652122291481,
+      "grad_norm": 0.236241511559059,
+      "learning_rate": 6.057303997618049e-07,
+      "loss": 0.5632,
+      "step": 2387
+    },
+    {
+      "epoch": 1.8884436529138222,
+      "grad_norm": 0.23687697173096994,
+      "learning_rate": 6.049665489569342e-07,
+      "loss": 0.5522,
+      "step": 2388
+    },
+    {
+      "epoch": 1.889235183536163,
+      "grad_norm": 0.2561281312211803,
+      "learning_rate": 6.04202971162852e-07,
+      "loss": 0.5401,
+      "step": 2389
+    },
+    {
+      "epoch": 1.890026714158504,
+      "grad_norm": 0.23887935233284158,
+      "learning_rate": 6.034396669072732e-07,
+      "loss": 0.553,
+      "step": 2390
+    },
+    {
+      "epoch": 1.8908182447808448,
+      "grad_norm": 0.24033422577314426,
+      "learning_rate": 6.026766367177232e-07,
+      "loss": 0.5634,
+      "step": 2391
+    },
+    {
+      "epoch": 1.891609775403186,
+      "grad_norm": 0.23369547767029367,
+      "learning_rate": 6.019138811215385e-07,
+      "loss": 0.537,
+      "step": 2392
+    },
+    {
+      "epoch": 1.8924013060255267,
+      "grad_norm": 0.23490532270272227,
+      "learning_rate": 6.01151400645866e-07,
+      "loss": 0.5571,
+      "step": 2393
+    },
+    {
+      "epoch": 1.8931928366478679,
+      "grad_norm": 0.2312910104944147,
+      "learning_rate": 6.003891958176612e-07,
+      "loss": 0.5607,
+      "step": 2394
+    },
+    {
+      "epoch": 1.8939843672702088,
+      "grad_norm": 0.23983094007084702,
+      "learning_rate": 5.996272671636909e-07,
+      "loss": 0.5445,
+      "step": 2395
+    },
+    {
+      "epoch": 1.8947758978925497,
+      "grad_norm": 0.24533562974078352,
+      "learning_rate": 5.988656152105292e-07,
+      "loss": 0.5649,
+      "step": 2396
+    },
+    {
+      "epoch": 1.8955674285148907,
+      "grad_norm": 0.24138758840329555,
+      "learning_rate": 5.981042404845602e-07,
+      "loss": 0.547,
+      "step": 2397
+    },
+    {
+      "epoch": 1.8963589591372316,
+      "grad_norm": 0.25398257402202423,
+      "learning_rate": 5.973431435119768e-07,
+      "loss": 0.5688,
+      "step": 2398
+    },
+    {
+      "epoch": 1.8971504897595726,
+      "grad_norm": 0.23325624017989163,
+      "learning_rate": 5.965823248187782e-07,
+      "loss": 0.5582,
+      "step": 2399
+    },
+    {
+      "epoch": 1.8979420203819135,
+      "grad_norm": 0.2405526192243757,
+      "learning_rate": 5.95821784930773e-07,
+      "loss": 0.5285,
+      "step": 2400
+    },
+    {
+      "epoch": 1.8987335510042544,
+      "grad_norm": 0.2433360587435491,
+      "learning_rate": 5.950615243735757e-07,
+      "loss": 0.557,
+      "step": 2401
+    },
+    {
+      "epoch": 1.8995250816265954,
+      "grad_norm": 0.26265562627560823,
+      "learning_rate": 5.943015436726091e-07,
+      "loss": 0.5465,
+      "step": 2402
+    },
+    {
+      "epoch": 1.9003166122489363,
+      "grad_norm": 0.23714179154726767,
+      "learning_rate": 5.935418433531022e-07,
+      "loss": 0.5653,
+      "step": 2403
+    },
+    {
+      "epoch": 1.9011081428712773,
+      "grad_norm": 0.2379637823598576,
+      "learning_rate": 5.927824239400892e-07,
+      "loss": 0.5566,
+      "step": 2404
+    },
+    {
+      "epoch": 1.9018996734936184,
+      "grad_norm": 0.24168197952379378,
+      "learning_rate": 5.920232859584115e-07,
+      "loss": 0.559,
+      "step": 2405
+    },
+    {
+      "epoch": 1.9026912041159592,
+      "grad_norm": 0.2364198883706187,
+      "learning_rate": 5.912644299327151e-07,
+      "loss": 0.5751,
+      "step": 2406
+    },
+    {
+      "epoch": 1.9034827347383003,
+      "grad_norm": 0.2461630194769159,
+      "learning_rate": 5.905058563874517e-07,
+      "loss": 0.544,
+      "step": 2407
+    },
+    {
+      "epoch": 1.904274265360641,
+      "grad_norm": 0.23906810091061048,
+      "learning_rate": 5.897475658468778e-07,
+      "loss": 0.5563,
+      "step": 2408
+    },
+    {
+      "epoch": 1.9050657959829822,
+      "grad_norm": 0.23987543128055613,
+      "learning_rate": 5.889895588350535e-07,
+      "loss": 0.5491,
+      "step": 2409
+    },
+    {
+      "epoch": 1.905857326605323,
+      "grad_norm": 0.2355722118955971,
+      "learning_rate": 5.88231835875844e-07,
+      "loss": 0.5519,
+      "step": 2410
+    },
+    {
+      "epoch": 1.906648857227664,
+      "grad_norm": 0.2358550161853384,
+      "learning_rate": 5.874743974929175e-07,
+      "loss": 0.5517,
+      "step": 2411
+    },
+    {
+      "epoch": 1.9074403878500048,
+      "grad_norm": 0.24080033229834438,
+      "learning_rate": 5.867172442097457e-07,
+      "loss": 0.5496,
+      "step": 2412
+    },
+    {
+      "epoch": 1.908231918472346,
+      "grad_norm": 0.22818664797067068,
+      "learning_rate": 5.859603765496036e-07,
+      "loss": 0.544,
+      "step": 2413
+    },
+    {
+      "epoch": 1.9090234490946867,
+      "grad_norm": 0.24635159524151135,
+      "learning_rate": 5.852037950355684e-07,
+      "loss": 0.5536,
+      "step": 2414
+    },
+    {
+      "epoch": 1.9098149797170278,
+      "grad_norm": 0.24616316436901883,
+      "learning_rate": 5.844475001905196e-07,
+      "loss": 0.5595,
+      "step": 2415
+    },
+    {
+      "epoch": 1.9106065103393688,
+      "grad_norm": 0.2368376079471897,
+      "learning_rate": 5.836914925371384e-07,
+      "loss": 0.5354,
+      "step": 2416
+    },
+    {
+      "epoch": 1.9113980409617097,
+      "grad_norm": 0.237265488635935,
+      "learning_rate": 5.829357725979083e-07,
+      "loss": 0.5644,
+      "step": 2417
+    },
+    {
+      "epoch": 1.9121895715840507,
+      "grad_norm": 0.2367113836381169,
+      "learning_rate": 5.821803408951125e-07,
+      "loss": 0.5449,
+      "step": 2418
+    },
+    {
+      "epoch": 1.9129811022063916,
+      "grad_norm": 0.23783758483957596,
+      "learning_rate": 5.814251979508373e-07,
+      "loss": 0.548,
+      "step": 2419
+    },
+    {
+      "epoch": 1.9137726328287326,
+      "grad_norm": 0.23645640597685252,
+      "learning_rate": 5.806703442869667e-07,
+      "loss": 0.5488,
+      "step": 2420
+    },
+    {
+      "epoch": 1.9145641634510735,
+      "grad_norm": 0.24353731461583766,
+      "learning_rate": 5.799157804251868e-07,
+      "loss": 0.5484,
+      "step": 2421
+    },
+    {
+      "epoch": 1.9153556940734144,
+      "grad_norm": 0.235479765701421,
+      "learning_rate": 5.791615068869828e-07,
+      "loss": 0.5651,
+      "step": 2422
+    },
+    {
+      "epoch": 1.9161472246957554,
+      "grad_norm": 0.24182147611689045,
+      "learning_rate": 5.784075241936387e-07,
+      "loss": 0.5543,
+      "step": 2423
+    },
+    {
+      "epoch": 1.9169387553180963,
+      "grad_norm": 0.234808676925544,
+      "learning_rate": 5.77653832866239e-07,
+      "loss": 0.5561,
+      "step": 2424
+    },
+    {
+      "epoch": 1.9177302859404373,
+      "grad_norm": 0.23768948880263918,
+      "learning_rate": 5.769004334256645e-07,
+      "loss": 0.5645,
+      "step": 2425
+    },
+    {
+      "epoch": 1.9185218165627784,
+      "grad_norm": 0.22840772589416633,
+      "learning_rate": 5.761473263925964e-07,
+      "loss": 0.5409,
+      "step": 2426
+    },
+    {
+      "epoch": 1.9193133471851191,
+      "grad_norm": 0.23969343364944234,
+      "learning_rate": 5.75394512287513e-07,
+      "loss": 0.5564,
+      "step": 2427
+    },
+    {
+      "epoch": 1.9201048778074603,
+      "grad_norm": 0.24547323203686924,
+      "learning_rate": 5.746419916306898e-07,
+      "loss": 0.5451,
+      "step": 2428
+    },
+    {
+      "epoch": 1.920896408429801,
+      "grad_norm": 0.23758892838939458,
+      "learning_rate": 5.738897649422005e-07,
+      "loss": 0.5494,
+      "step": 2429
+    },
+    {
+      "epoch": 1.9216879390521422,
+      "grad_norm": 0.250850511325333,
+      "learning_rate": 5.731378327419141e-07,
+      "loss": 0.5557,
+      "step": 2430
+    },
+    {
+      "epoch": 1.922479469674483,
+      "grad_norm": 0.2474183927050043,
+      "learning_rate": 5.723861955494977e-07,
+      "loss": 0.5575,
+      "step": 2431
+    },
+    {
+      "epoch": 1.923271000296824,
+      "grad_norm": 0.24081355385277323,
+      "learning_rate": 5.716348538844136e-07,
+      "loss": 0.5499,
+      "step": 2432
+    },
+    {
+      "epoch": 1.9240625309191648,
+      "grad_norm": 0.23944059010205088,
+      "learning_rate": 5.708838082659198e-07,
+      "loss": 0.5524,
+      "step": 2433
+    },
+    {
+      "epoch": 1.924854061541506,
+      "grad_norm": 0.24110743457635514,
+      "learning_rate": 5.701330592130704e-07,
+      "loss": 0.5526,
+      "step": 2434
+    },
+    {
+      "epoch": 1.9256455921638467,
+      "grad_norm": 0.24173994727975137,
+      "learning_rate": 5.693826072447139e-07,
+      "loss": 0.557,
+      "step": 2435
+    },
+    {
+      "epoch": 1.9264371227861878,
+      "grad_norm": 0.2464393142787827,
+      "learning_rate": 5.686324528794937e-07,
+      "loss": 0.5385,
+      "step": 2436
+    },
+    {
+      "epoch": 1.9272286534085288,
+      "grad_norm": 0.24029954615659102,
+      "learning_rate": 5.678825966358474e-07,
+      "loss": 0.5513,
+      "step": 2437
+    },
+    {
+      "epoch": 1.9280201840308697,
+      "grad_norm": 0.2348592985551055,
+      "learning_rate": 5.671330390320063e-07,
+      "loss": 0.5387,
+      "step": 2438
+    },
+    {
+      "epoch": 1.9288117146532107,
+      "grad_norm": 0.2427186720103737,
+      "learning_rate": 5.663837805859966e-07,
+      "loss": 0.5583,
+      "step": 2439
+    },
+    {
+      "epoch": 1.9296032452755516,
+      "grad_norm": 0.2377050876736877,
+      "learning_rate": 5.656348218156363e-07,
+      "loss": 0.5505,
+      "step": 2440
+    },
+    {
+      "epoch": 1.9303947758978925,
+      "grad_norm": 0.24523301092284647,
+      "learning_rate": 5.648861632385368e-07,
+      "loss": 0.5317,
+      "step": 2441
+    },
+    {
+      "epoch": 1.9311863065202335,
+      "grad_norm": 0.23440624414493083,
+      "learning_rate": 5.641378053721022e-07,
+      "loss": 0.5537,
+      "step": 2442
+    },
+    {
+      "epoch": 1.9319778371425744,
+      "grad_norm": 0.24180475598605752,
+      "learning_rate": 5.633897487335279e-07,
+      "loss": 0.5429,
+      "step": 2443
+    },
+    {
+      "epoch": 1.9327693677649154,
+      "grad_norm": 0.23849571115961757,
+      "learning_rate": 5.626419938398032e-07,
+      "loss": 0.5468,
+      "step": 2444
+    },
+    {
+      "epoch": 1.9335608983872563,
+      "grad_norm": 0.2460470566941156,
+      "learning_rate": 5.618945412077065e-07,
+      "loss": 0.5614,
+      "step": 2445
+    },
+    {
+      "epoch": 1.9343524290095973,
+      "grad_norm": 0.23681777891851846,
+      "learning_rate": 5.611473913538087e-07,
+      "loss": 0.5575,
+      "step": 2446
+    },
+    {
+      "epoch": 1.9351439596319384,
+      "grad_norm": 0.23023517497313897,
+      "learning_rate": 5.604005447944711e-07,
+      "loss": 0.5458,
+      "step": 2447
+    },
+    {
+      "epoch": 1.9359354902542791,
+      "grad_norm": 0.23989326935200456,
+      "learning_rate": 5.596540020458445e-07,
+      "loss": 0.5606,
+      "step": 2448
+    },
+    {
+      "epoch": 1.9367270208766203,
+      "grad_norm": 0.24559666397466526,
+      "learning_rate": 5.589077636238719e-07,
+      "loss": 0.5495,
+      "step": 2449
+    },
+    {
+      "epoch": 1.937518551498961,
+      "grad_norm": 0.2315148502891609,
+      "learning_rate": 5.58161830044284e-07,
+      "loss": 0.5523,
+      "step": 2450
+    },
+    {
+      "epoch": 1.9383100821213022,
+      "grad_norm": 0.24074998530427816,
+      "learning_rate": 5.574162018226015e-07,
+      "loss": 0.5455,
+      "step": 2451
+    },
+    {
+      "epoch": 1.939101612743643,
+      "grad_norm": 0.23730949762800163,
+      "learning_rate": 5.566708794741341e-07,
+      "loss": 0.5483,
+      "step": 2452
+    },
+    {
+      "epoch": 1.939893143365984,
+      "grad_norm": 0.24244224941050907,
+      "learning_rate": 5.559258635139796e-07,
+      "loss": 0.5654,
+      "step": 2453
+    },
+    {
+      "epoch": 1.9406846739883248,
+      "grad_norm": 0.24326649471398717,
+      "learning_rate": 5.551811544570255e-07,
+      "loss": 0.558,
+      "step": 2454
+    },
+    {
+      "epoch": 1.941476204610666,
+      "grad_norm": 0.25468287478151763,
+      "learning_rate": 5.544367528179456e-07,
+      "loss": 0.5591,
+      "step": 2455
+    },
+    {
+      "epoch": 1.9422677352330067,
+      "grad_norm": 0.23107902824037338,
+      "learning_rate": 5.536926591112018e-07,
+      "loss": 0.5448,
+      "step": 2456
+    },
+    {
+      "epoch": 1.9430592658553478,
+      "grad_norm": 0.24051650444495715,
+      "learning_rate": 5.529488738510434e-07,
+      "loss": 0.5637,
+      "step": 2457
+    },
+    {
+      "epoch": 1.9438507964776888,
+      "grad_norm": 0.23882938825699424,
+      "learning_rate": 5.52205397551506e-07,
+      "loss": 0.5531,
+      "step": 2458
+    },
+    {
+      "epoch": 1.9446423271000297,
+      "grad_norm": 0.23670699492810005,
+      "learning_rate": 5.514622307264125e-07,
+      "loss": 0.5665,
+      "step": 2459
+    },
+    {
+      "epoch": 1.9454338577223707,
+      "grad_norm": 0.23841907768128848,
+      "learning_rate": 5.507193738893714e-07,
+      "loss": 0.5495,
+      "step": 2460
+    },
+    {
+      "epoch": 1.9462253883447116,
+      "grad_norm": 0.23961571643571905,
+      "learning_rate": 5.49976827553777e-07,
+      "loss": 0.5494,
+      "step": 2461
+    },
+    {
+      "epoch": 1.9470169189670525,
+      "grad_norm": 0.24335249171704146,
+      "learning_rate": 5.492345922328089e-07,
+      "loss": 0.5661,
+      "step": 2462
+    },
+    {
+      "epoch": 1.9478084495893935,
+      "grad_norm": 0.2314967581467103,
+      "learning_rate": 5.484926684394316e-07,
+      "loss": 0.5583,
+      "step": 2463
+    },
+    {
+      "epoch": 1.9485999802117344,
+      "grad_norm": 0.239064802795224,
+      "learning_rate": 5.477510566863953e-07,
+      "loss": 0.541,
+      "step": 2464
+    },
+    {
+      "epoch": 1.9493915108340754,
+      "grad_norm": 0.2284958524975405,
+      "learning_rate": 5.470097574862334e-07,
+      "loss": 0.5603,
+      "step": 2465
+    },
+    {
+      "epoch": 1.9501830414564163,
+      "grad_norm": 0.23631318786547262,
+      "learning_rate": 5.462687713512637e-07,
+      "loss": 0.5551,
+      "step": 2466
+    },
+    {
+      "epoch": 1.9509745720787572,
+      "grad_norm": 0.2367226366344625,
+      "learning_rate": 5.455280987935879e-07,
+      "loss": 0.5578,
+      "step": 2467
+    },
+    {
+      "epoch": 1.9517661027010984,
+      "grad_norm": 0.2383950714885435,
+      "learning_rate": 5.447877403250901e-07,
+      "loss": 0.567,
+      "step": 2468
+    },
+    {
+      "epoch": 1.9525576333234391,
+      "grad_norm": 0.2350701691920177,
+      "learning_rate": 5.44047696457439e-07,
+      "loss": 0.54,
+      "step": 2469
+    },
+    {
+      "epoch": 1.9533491639457803,
+      "grad_norm": 0.23771857910109087,
+      "learning_rate": 5.433079677020842e-07,
+      "loss": 0.554,
+      "step": 2470
+    },
+    {
+      "epoch": 1.954140694568121,
+      "grad_norm": 0.23727350879604792,
+      "learning_rate": 5.425685545702585e-07,
+      "loss": 0.5467,
+      "step": 2471
+    },
+    {
+      "epoch": 1.9549322251904622,
+      "grad_norm": 0.22989087330747757,
+      "learning_rate": 5.418294575729758e-07,
+      "loss": 0.5455,
+      "step": 2472
+    },
+    {
+      "epoch": 1.9557237558128029,
+      "grad_norm": 0.23442124535998407,
+      "learning_rate": 5.410906772210325e-07,
+      "loss": 0.5519,
+      "step": 2473
+    },
+    {
+      "epoch": 1.956515286435144,
+      "grad_norm": 0.23784311728729576,
+      "learning_rate": 5.403522140250049e-07,
+      "loss": 0.5597,
+      "step": 2474
+    },
+    {
+      "epoch": 1.9573068170574848,
+      "grad_norm": 0.2397409728072015,
+      "learning_rate": 5.396140684952516e-07,
+      "loss": 0.5431,
+      "step": 2475
+    },
+    {
+      "epoch": 1.958098347679826,
+      "grad_norm": 0.2398035543797548,
+      "learning_rate": 5.388762411419104e-07,
+      "loss": 0.5614,
+      "step": 2476
+    },
+    {
+      "epoch": 1.9588898783021667,
+      "grad_norm": 0.23994136890298104,
+      "learning_rate": 5.381387324749005e-07,
+      "loss": 0.5554,
+      "step": 2477
+    },
+    {
+      "epoch": 1.9596814089245078,
+      "grad_norm": 0.24169417187721992,
+      "learning_rate": 5.374015430039191e-07,
+      "loss": 0.5536,
+      "step": 2478
+    },
+    {
+      "epoch": 1.9604729395468488,
+      "grad_norm": 0.23709731857020674,
+      "learning_rate": 5.366646732384437e-07,
+      "loss": 0.557,
+      "step": 2479
+    },
+    {
+      "epoch": 1.9612644701691897,
+      "grad_norm": 0.2402655953454577,
+      "learning_rate": 5.35928123687732e-07,
+      "loss": 0.551,
+      "step": 2480
+    },
+    {
+      "epoch": 1.9620560007915306,
+      "grad_norm": 0.24224574444439703,
+      "learning_rate": 5.351918948608181e-07,
+      "loss": 0.5646,
+      "step": 2481
+    },
+    {
+      "epoch": 1.9628475314138716,
+      "grad_norm": 0.2457113684861732,
+      "learning_rate": 5.344559872665168e-07,
+      "loss": 0.542,
+      "step": 2482
+    },
+    {
+      "epoch": 1.9636390620362125,
+      "grad_norm": 0.24704185219831643,
+      "learning_rate": 5.337204014134189e-07,
+      "loss": 0.5589,
+      "step": 2483
+    },
+    {
+      "epoch": 1.9644305926585535,
+      "grad_norm": 0.24402006383958874,
+      "learning_rate": 5.329851378098935e-07,
+      "loss": 0.557,
+      "step": 2484
+    },
+    {
+      "epoch": 1.9652221232808944,
+      "grad_norm": 0.23717589771406625,
+      "learning_rate": 5.322501969640882e-07,
+      "loss": 0.5657,
+      "step": 2485
+    },
+    {
+      "epoch": 1.9660136539032353,
+      "grad_norm": 0.24098898733719273,
+      "learning_rate": 5.315155793839252e-07,
+      "loss": 0.5568,
+      "step": 2486
+    },
+    {
+      "epoch": 1.9668051845255763,
+      "grad_norm": 0.23162980850798595,
+      "learning_rate": 5.307812855771064e-07,
+      "loss": 0.5632,
+      "step": 2487
+    },
+    {
+      "epoch": 1.9675967151479172,
+      "grad_norm": 0.2451998753327035,
+      "learning_rate": 5.300473160511065e-07,
+      "loss": 0.5519,
+      "step": 2488
+    },
+    {
+      "epoch": 1.9683882457702584,
+      "grad_norm": 0.24608389236797332,
+      "learning_rate": 5.29313671313178e-07,
+      "loss": 0.5385,
+      "step": 2489
+    },
+    {
+      "epoch": 1.9691797763925991,
+      "grad_norm": 0.23457900464866493,
+      "learning_rate": 5.285803518703493e-07,
+      "loss": 0.5576,
+      "step": 2490
+    },
+    {
+      "epoch": 1.9699713070149403,
+      "grad_norm": 0.2371699065569416,
+      "learning_rate": 5.27847358229423e-07,
+      "loss": 0.557,
+      "step": 2491
+    },
+    {
+      "epoch": 1.970762837637281,
+      "grad_norm": 0.24009406599812902,
+      "learning_rate": 5.27114690896977e-07,
+      "loss": 0.5525,
+      "step": 2492
+    },
+    {
+      "epoch": 1.9715543682596222,
+      "grad_norm": 0.23655243716817018,
+      "learning_rate": 5.263823503793634e-07,
+      "loss": 0.5609,
+      "step": 2493
+    },
+    {
+      "epoch": 1.9723458988819629,
+      "grad_norm": 0.24445936204627025,
+      "learning_rate": 5.256503371827085e-07,
+      "loss": 0.5435,
+      "step": 2494
+    },
+    {
+      "epoch": 1.973137429504304,
+      "grad_norm": 0.24478990006565868,
+      "learning_rate": 5.249186518129128e-07,
+      "loss": 0.5612,
+      "step": 2495
+    },
+    {
+      "epoch": 1.9739289601266448,
+      "grad_norm": 0.23624292044273565,
+      "learning_rate": 5.241872947756501e-07,
+      "loss": 0.5525,
+      "step": 2496
+    },
+    {
+      "epoch": 1.974720490748986,
+      "grad_norm": 0.23910063314875665,
+      "learning_rate": 5.234562665763662e-07,
+      "loss": 0.5534,
+      "step": 2497
+    },
+    {
+      "epoch": 1.9755120213713266,
+      "grad_norm": 0.23513898688014817,
+      "learning_rate": 5.227255677202821e-07,
+      "loss": 0.5469,
+      "step": 2498
+    },
+    {
+      "epoch": 1.9763035519936678,
+      "grad_norm": 0.24291059633629505,
+      "learning_rate": 5.219951987123878e-07,
+      "loss": 0.5585,
+      "step": 2499
+    },
+    {
+      "epoch": 1.9770950826160087,
+      "grad_norm": 0.2387987313855466,
+      "learning_rate": 5.212651600574487e-07,
+      "loss": 0.5384,
+      "step": 2500
+    },
+    {
+      "epoch": 1.9778866132383497,
+      "grad_norm": 0.2450052462307983,
+      "learning_rate": 5.205354522599998e-07,
+      "loss": 0.5673,
+      "step": 2501
+    },
+    {
+      "epoch": 1.9786781438606906,
+      "grad_norm": 0.23726334743327526,
+      "learning_rate": 5.198060758243475e-07,
+      "loss": 0.5511,
+      "step": 2502
+    },
+    {
+      "epoch": 1.9794696744830316,
+      "grad_norm": 0.241254532507191,
+      "learning_rate": 5.190770312545711e-07,
+      "loss": 0.5596,
+      "step": 2503
+    },
+    {
+      "epoch": 1.9802612051053725,
+      "grad_norm": 0.2536493764522025,
+      "learning_rate": 5.183483190545175e-07,
+      "loss": 0.5662,
+      "step": 2504
+    },
+    {
+      "epoch": 1.9810527357277135,
+      "grad_norm": 0.2359397363385922,
+      "learning_rate": 5.176199397278066e-07,
+      "loss": 0.5412,
+      "step": 2505
+    },
+    {
+      "epoch": 1.9818442663500544,
+      "grad_norm": 0.2353383907282585,
+      "learning_rate": 5.16891893777827e-07,
+      "loss": 0.5562,
+      "step": 2506
+    },
+    {
+      "epoch": 1.9826357969723953,
+      "grad_norm": 0.2420766802841231,
+      "learning_rate": 5.161641817077366e-07,
+      "loss": 0.5502,
+      "step": 2507
+    },
+    {
+      "epoch": 1.9834273275947363,
+      "grad_norm": 0.3428472738074613,
+      "learning_rate": 5.154368040204642e-07,
+      "loss": 0.5485,
+      "step": 2508
+    },
+    {
+      "epoch": 1.9842188582170772,
+      "grad_norm": 0.2383569482950343,
+      "learning_rate": 5.14709761218705e-07,
+      "loss": 0.567,
+      "step": 2509
+    },
+    {
+      "epoch": 1.9850103888394184,
+      "grad_norm": 0.2412249723452846,
+      "learning_rate": 5.13983053804925e-07,
+      "loss": 0.5726,
+      "step": 2510
+    },
+    {
+      "epoch": 1.985801919461759,
+      "grad_norm": 0.24101319591698722,
+      "learning_rate": 5.132566822813574e-07,
+      "loss": 0.5551,
+      "step": 2511
+    },
+    {
+      "epoch": 1.9865934500841003,
+      "grad_norm": 0.23809317461331223,
+      "learning_rate": 5.125306471500027e-07,
+      "loss": 0.5561,
+      "step": 2512
+    },
+    {
+      "epoch": 1.987384980706441,
+      "grad_norm": 0.24225892795271026,
+      "learning_rate": 5.118049489126311e-07,
+      "loss": 0.5441,
+      "step": 2513
+    },
+    {
+      "epoch": 1.9881765113287821,
+      "grad_norm": 0.2320756177212838,
+      "learning_rate": 5.110795880707766e-07,
+      "loss": 0.5407,
+      "step": 2514
+    },
+    {
+      "epoch": 1.9889680419511229,
+      "grad_norm": 0.23590815149266026,
+      "learning_rate": 5.103545651257434e-07,
+      "loss": 0.5581,
+      "step": 2515
+    },
+    {
+      "epoch": 1.989759572573464,
+      "grad_norm": 0.2310993563834982,
+      "learning_rate": 5.096298805786001e-07,
+      "loss": 0.5519,
+      "step": 2516
+    },
+    {
+      "epoch": 1.9905511031958047,
+      "grad_norm": 0.2465016184253646,
+      "learning_rate": 5.089055349301816e-07,
+      "loss": 0.5543,
+      "step": 2517
+    },
+    {
+      "epoch": 1.991342633818146,
+      "grad_norm": 0.24490968259524826,
+      "learning_rate": 5.081815286810902e-07,
+      "loss": 0.5473,
+      "step": 2518
+    },
+    {
+      "epoch": 1.9921341644404866,
+      "grad_norm": 0.2394550568566711,
+      "learning_rate": 5.074578623316907e-07,
+      "loss": 0.5717,
+      "step": 2519
+    },
+    {
+      "epoch": 1.9929256950628278,
+      "grad_norm": 0.2351647758535822,
+      "learning_rate": 5.067345363821162e-07,
+      "loss": 0.554,
+      "step": 2520
+    },
+    {
+      "epoch": 1.9937172256851687,
+      "grad_norm": 0.24227970138606264,
+      "learning_rate": 5.060115513322627e-07,
+      "loss": 0.5676,
+      "step": 2521
+    },
+    {
+      "epoch": 1.9945087563075097,
+      "grad_norm": 0.24532665453467145,
+      "learning_rate": 5.052889076817903e-07,
+      "loss": 0.5405,
+      "step": 2522
+    },
+    {
+      "epoch": 1.9953002869298506,
+      "grad_norm": 0.23808811316520417,
+      "learning_rate": 5.045666059301246e-07,
+      "loss": 0.5599,
+      "step": 2523
+    },
+    {
+      "epoch": 1.9960918175521916,
+      "grad_norm": 0.24540727646746247,
+      "learning_rate": 5.038446465764541e-07,
+      "loss": 0.5624,
+      "step": 2524
+    },
+    {
+      "epoch": 1.9968833481745325,
+      "grad_norm": 0.24379096659303265,
+      "learning_rate": 5.031230301197303e-07,
+      "loss": 0.5636,
+      "step": 2525
+    },
+    {
+      "epoch": 1.9976748787968734,
+      "grad_norm": 0.23696774982991786,
+      "learning_rate": 5.024017570586684e-07,
+      "loss": 0.5613,
+      "step": 2526
+    },
+    {
+      "epoch": 1.9984664094192144,
+      "grad_norm": 0.2438080404125615,
+      "learning_rate": 5.016808278917455e-07,
+      "loss": 0.5457,
+      "step": 2527
+    },
+    {
+      "epoch": 1.9992579400415553,
+      "grad_norm": 0.24294862123670521,
+      "learning_rate": 5.009602431172022e-07,
+      "loss": 0.5512,
+      "step": 2528
+    },
+    {
+      "epoch": 2.000049470663896,
+      "grad_norm": 0.23540041623911523,
+      "learning_rate": 5.002400032330403e-07,
+      "loss": 0.5471,
+      "step": 2529
+    },
+    {
+      "epoch": 2.000841001286237,
+      "grad_norm": 0.24136383377777998,
+      "learning_rate": 4.995201087370232e-07,
+      "loss": 0.5639,
+      "step": 2530
+    },
+    {
+      "epoch": 2.0016325319085784,
+      "grad_norm": 0.24016731049277348,
+      "learning_rate": 4.988005601266758e-07,
+      "loss": 0.5538,
+      "step": 2531
+    },
+    {
+      "epoch": 2.0016325319085784,
+      "eval_loss": 0.5885822772979736,
+      "eval_runtime": 2118.5562,
+      "eval_samples_per_second": 12.872,
+      "eval_steps_per_second": 1.609,
+      "step": 2531
+    },
+    {
+      "epoch": 2.0007420232500617,
+      "grad_norm": 0.24773872144799403,
+      "learning_rate": 4.980813578992835e-07,
+      "loss": 0.5381,
+      "step": 2532
+    },
+    {
+      "epoch": 2.0015335147167943,
+      "grad_norm": 0.2385073683372732,
+      "learning_rate": 4.973625025518934e-07,
+      "loss": 0.5313,
+      "step": 2533
+    },
+    {
+      "epoch": 2.002325006183527,
+      "grad_norm": 0.24022890170412228,
+      "learning_rate": 4.966439945813121e-07,
+      "loss": 0.5483,
+      "step": 2534
+    },
+    {
+      "epoch": 2.0031164976502596,
+      "grad_norm": 0.24169024205808562,
+      "learning_rate": 4.959258344841061e-07,
+      "loss": 0.5524,
+      "step": 2535
+    },
+    {
+      "epoch": 2.003907989116992,
+      "grad_norm": 0.23275002001052666,
+      "learning_rate": 4.952080227566014e-07,
+      "loss": 0.5295,
+      "step": 2536
+    },
+    {
+      "epoch": 2.004699480583725,
+      "grad_norm": 0.23000403348109522,
+      "learning_rate": 4.944905598948834e-07,
+      "loss": 0.5476,
+      "step": 2537
+    },
+    {
+      "epoch": 2.0054909720504575,
+      "grad_norm": 0.24300537805341524,
+      "learning_rate": 4.93773446394797e-07,
+      "loss": 0.5485,
+      "step": 2538
+    },
+    {
+      "epoch": 2.00628246351719,
+      "grad_norm": 0.24273370758876583,
+      "learning_rate": 4.930566827519447e-07,
+      "loss": 0.539,
+      "step": 2539
+    },
+    {
+      "epoch": 2.0070739549839227,
+      "grad_norm": 0.24281565983217607,
+      "learning_rate": 4.923402694616877e-07,
+      "loss": 0.5393,
+      "step": 2540
+    },
+    {
+      "epoch": 2.0078654464506553,
+      "grad_norm": 0.24177680184458014,
+      "learning_rate": 4.91624207019145e-07,
+      "loss": 0.5435,
+      "step": 2541
+    },
+    {
+      "epoch": 2.008656937917388,
+      "grad_norm": 0.24674405255012505,
+      "learning_rate": 4.909084959191925e-07,
+      "loss": 0.5552,
+      "step": 2542
+    },
+    {
+      "epoch": 2.0094484293841206,
+      "grad_norm": 0.23832192168061545,
+      "learning_rate": 4.901931366564649e-07,
+      "loss": 0.541,
+      "step": 2543
+    },
+    {
+      "epoch": 2.0102399208508532,
+      "grad_norm": 0.23311165779172743,
+      "learning_rate": 4.894781297253522e-07,
+      "loss": 0.5371,
+      "step": 2544
+    },
+    {
+      "epoch": 2.011031412317586,
+      "grad_norm": 0.23750875001267496,
+      "learning_rate": 4.887634756200015e-07,
+      "loss": 0.549,
+      "step": 2545
+    },
+    {
+      "epoch": 2.0118229037843185,
+      "grad_norm": 0.23927372447424175,
+      "learning_rate": 4.880491748343159e-07,
+      "loss": 0.5311,
+      "step": 2546
+    },
+    {
+      "epoch": 2.012614395251051,
+      "grad_norm": 0.23826738272169978,
+      "learning_rate": 4.873352278619542e-07,
+      "loss": 0.5356,
+      "step": 2547
+    },
+    {
+      "epoch": 2.0134058867177838,
+      "grad_norm": 0.2430163007596369,
+      "learning_rate": 4.866216351963316e-07,
+      "loss": 0.5495,
+      "step": 2548
+    },
+    {
+      "epoch": 2.0141973781845164,
+      "grad_norm": 0.23496655183307583,
+      "learning_rate": 4.859083973306171e-07,
+      "loss": 0.5374,
+      "step": 2549
+    },
+    {
+      "epoch": 2.014988869651249,
+      "grad_norm": 0.23169821982515804,
+      "learning_rate": 4.851955147577354e-07,
+      "loss": 0.5462,
+      "step": 2550
+    },
+    {
+      "epoch": 2.0157803611179816,
+      "grad_norm": 0.23525321337936167,
+      "learning_rate": 4.844829879703653e-07,
+      "loss": 0.5204,
+      "step": 2551
+    },
+    {
+      "epoch": 2.0165718525847143,
+      "grad_norm": 0.2412893577983348,
+      "learning_rate": 4.837708174609393e-07,
+      "loss": 0.5543,
+      "step": 2552
+    },
+    {
+      "epoch": 2.017363344051447,
+      "grad_norm": 0.24020913441771027,
+      "learning_rate": 4.830590037216449e-07,
+      "loss": 0.5492,
+      "step": 2553
+    },
+    {
+      "epoch": 2.0181548355181795,
+      "grad_norm": 0.2388973935767555,
+      "learning_rate": 4.823475472444221e-07,
+      "loss": 0.5522,
+      "step": 2554
+    },
+    {
+      "epoch": 2.018946326984912,
+      "grad_norm": 0.23177782154784332,
+      "learning_rate": 4.81636448520964e-07,
+      "loss": 0.5365,
+      "step": 2555
+    },
+    {
+      "epoch": 2.019737818451645,
+      "grad_norm": 0.23893233132437353,
+      "learning_rate": 4.809257080427166e-07,
+      "loss": 0.5405,
+      "step": 2556
+    },
+    {
+      "epoch": 2.0205293099183774,
+      "grad_norm": 0.24141766708765985,
+      "learning_rate": 4.802153263008782e-07,
+      "loss": 0.5479,
+      "step": 2557
+    },
+    {
+      "epoch": 2.02132080138511,
+      "grad_norm": 0.2333453756743298,
+      "learning_rate": 4.795053037863999e-07,
+      "loss": 0.5502,
+      "step": 2558
+    },
+    {
+      "epoch": 2.0221122928518427,
+      "grad_norm": 0.2383671275575243,
+      "learning_rate": 4.787956409899836e-07,
+      "loss": 0.5439,
+      "step": 2559
+    },
+    {
+      "epoch": 2.0229037843185753,
+      "grad_norm": 0.2361967368521129,
+      "learning_rate": 4.78086338402083e-07,
+      "loss": 0.5508,
+      "step": 2560
+    },
+    {
+      "epoch": 2.023695275785308,
+      "grad_norm": 0.23710023443369116,
+      "learning_rate": 4.773773965129028e-07,
+      "loss": 0.5514,
+      "step": 2561
+    },
+    {
+      "epoch": 2.0244867672520406,
+      "grad_norm": 0.24349644792421318,
+      "learning_rate": 4.76668815812398e-07,
+      "loss": 0.5489,
+      "step": 2562
+    },
+    {
+      "epoch": 2.025278258718773,
+      "grad_norm": 0.23121129785638533,
+      "learning_rate": 4.7596059679027536e-07,
+      "loss": 0.5465,
+      "step": 2563
+    },
+    {
+      "epoch": 2.026069750185506,
+      "grad_norm": 0.23738825335000968,
+      "learning_rate": 4.7525273993599015e-07,
+      "loss": 0.5532,
+      "step": 2564
+    },
+    {
+      "epoch": 2.0268612416522385,
+      "grad_norm": 0.2372106152865567,
+      "learning_rate": 4.7454524573874744e-07,
+      "loss": 0.544,
+      "step": 2565
+    },
+    {
+      "epoch": 2.027652733118971,
+      "grad_norm": 0.2367507110443533,
+      "learning_rate": 4.7383811468750356e-07,
+      "loss": 0.5398,
+      "step": 2566
+    },
+    {
+      "epoch": 2.0284442245857037,
+      "grad_norm": 0.23602386237918596,
+      "learning_rate": 4.7313134727096083e-07,
+      "loss": 0.5348,
+      "step": 2567
+    },
+    {
+      "epoch": 2.0292357160524364,
+      "grad_norm": 0.2420129410819989,
+      "learning_rate": 4.7242494397757303e-07,
+      "loss": 0.5344,
+      "step": 2568
+    },
+    {
+      "epoch": 2.030027207519169,
+      "grad_norm": 0.2380690253218786,
+      "learning_rate": 4.717189052955407e-07,
+      "loss": 0.5395,
+      "step": 2569
+    },
+    {
+      "epoch": 2.0308186989859016,
+      "grad_norm": 0.2409138666986156,
+      "learning_rate": 4.710132317128125e-07,
+      "loss": 0.5538,
+      "step": 2570
+    },
+    {
+      "epoch": 2.0316101904526342,
+      "grad_norm": 0.24516111762394113,
+      "learning_rate": 4.7030792371708625e-07,
+      "loss": 0.5539,
+      "step": 2571
+    },
+    {
+      "epoch": 2.032401681919367,
+      "grad_norm": 0.24333049370617602,
+      "learning_rate": 4.696029817958045e-07,
+      "loss": 0.5523,
+      "step": 2572
+    },
+    {
+      "epoch": 2.0331931733860995,
+      "grad_norm": 0.23243381498396162,
+      "learning_rate": 4.688984064361593e-07,
+      "loss": 0.5478,
+      "step": 2573
+    },
+    {
+      "epoch": 2.033984664852832,
+      "grad_norm": 0.23501751026576856,
+      "learning_rate": 4.6819419812508807e-07,
+      "loss": 0.5401,
+      "step": 2574
+    },
+    {
+      "epoch": 2.0347761563195648,
+      "grad_norm": 0.2370957909666972,
+      "learning_rate": 4.6749035734927434e-07,
+      "loss": 0.5365,
+      "step": 2575
+    },
+    {
+      "epoch": 2.0355676477862974,
+      "grad_norm": 0.2350543528277413,
+      "learning_rate": 4.6678688459514937e-07,
+      "loss": 0.5292,
+      "step": 2576
+    },
+    {
+      "epoch": 2.03635913925303,
+      "grad_norm": 0.23770123833645423,
+      "learning_rate": 4.6608378034888763e-07,
+      "loss": 0.5568,
+      "step": 2577
+    },
+    {
+      "epoch": 2.0371506307197627,
+      "grad_norm": 0.24308606133493194,
+      "learning_rate": 4.6538104509641007e-07,
+      "loss": 0.5452,
+      "step": 2578
+    },
+    {
+      "epoch": 2.0379421221864953,
+      "grad_norm": 0.2363203380658553,
+      "learning_rate": 4.6467867932338344e-07,
+      "loss": 0.5296,
+      "step": 2579
+    },
+    {
+      "epoch": 2.038733613653228,
+      "grad_norm": 0.24286814987173197,
+      "learning_rate": 4.639766835152177e-07,
+      "loss": 0.5401,
+      "step": 2580
+    },
+    {
+      "epoch": 2.0395251051199605,
+      "grad_norm": 0.23125599966751756,
+      "learning_rate": 4.632750581570687e-07,
+      "loss": 0.5508,
+      "step": 2581
+    },
+    {
+      "epoch": 2.040316596586693,
+      "grad_norm": 0.25038102575664556,
+      "learning_rate": 4.625738037338345e-07,
+      "loss": 0.5452,
+      "step": 2582
+    },
+    {
+      "epoch": 2.041108088053426,
+      "grad_norm": 0.2356423019517063,
+      "learning_rate": 4.618729207301575e-07,
+      "loss": 0.5423,
+      "step": 2583
+    },
+    {
+      "epoch": 2.0418995795201584,
+      "grad_norm": 0.23772180389882316,
+      "learning_rate": 4.611724096304244e-07,
+      "loss": 0.5395,
+      "step": 2584
+    },
+    {
+      "epoch": 2.042691070986891,
+      "grad_norm": 0.23939242487627102,
+      "learning_rate": 4.6047227091876305e-07,
+      "loss": 0.5283,
+      "step": 2585
+    },
+    {
+      "epoch": 2.0434825624536237,
+      "grad_norm": 0.24759837515967656,
+      "learning_rate": 4.597725050790461e-07,
+      "loss": 0.5594,
+      "step": 2586
+    },
+    {
+      "epoch": 2.0442740539203563,
+      "grad_norm": 0.238897498076543,
+      "learning_rate": 4.59073112594887e-07,
+      "loss": 0.544,
+      "step": 2587
+    },
+    {
+      "epoch": 2.045065545387089,
+      "grad_norm": 0.2385577619041336,
+      "learning_rate": 4.583740939496404e-07,
+      "loss": 0.5599,
+      "step": 2588
+    },
+    {
+      "epoch": 2.0458570368538216,
+      "grad_norm": 0.23923905909510285,
+      "learning_rate": 4.576754496264051e-07,
+      "loss": 0.548,
+      "step": 2589
+    },
+    {
+      "epoch": 2.046648528320554,
+      "grad_norm": 0.23462535030007492,
+      "learning_rate": 4.5697718010801877e-07,
+      "loss": 0.5519,
+      "step": 2590
+    },
+    {
+      "epoch": 2.047440019787287,
+      "grad_norm": 0.24857974456978588,
+      "learning_rate": 4.5627928587706197e-07,
+      "loss": 0.5494,
+      "step": 2591
+    },
+    {
+      "epoch": 2.0482315112540195,
+      "grad_norm": 0.23635724494456387,
+      "learning_rate": 4.5558176741585497e-07,
+      "loss": 0.55,
+      "step": 2592
+    },
+    {
+      "epoch": 2.049023002720752,
+      "grad_norm": 0.23616867778197356,
+      "learning_rate": 4.548846252064572e-07,
+      "loss": 0.551,
+      "step": 2593
+    },
+    {
+      "epoch": 2.0498144941874847,
+      "grad_norm": 0.2383207284099584,
+      "learning_rate": 4.541878597306704e-07,
+      "loss": 0.5335,
+      "step": 2594
+    },
+    {
+      "epoch": 2.0506059856542174,
+      "grad_norm": 0.23933865356005052,
+      "learning_rate": 4.534914714700345e-07,
+      "loss": 0.5545,
+      "step": 2595
+    },
+    {
+      "epoch": 2.05139747712095,
+      "grad_norm": 0.2475085228852165,
+      "learning_rate": 4.5279546090582855e-07,
+      "loss": 0.5459,
+      "step": 2596
+    },
+    {
+      "epoch": 2.0521889685876826,
+      "grad_norm": 0.23124372369872748,
+      "learning_rate": 4.5209982851907205e-07,
+      "loss": 0.5432,
+      "step": 2597
+    },
+    {
+      "epoch": 2.052980460054415,
+      "grad_norm": 0.23836302808378285,
+      "learning_rate": 4.5140457479052076e-07,
+      "loss": 0.5439,
+      "step": 2598
+    },
+    {
+      "epoch": 2.0537719515211474,
+      "grad_norm": 0.24070375492131732,
+      "learning_rate": 4.507097002006712e-07,
+      "loss": 0.5298,
+      "step": 2599
+    },
+    {
+      "epoch": 2.05456344298788,
+      "grad_norm": 0.24882750453944374,
+      "learning_rate": 4.5001520522975655e-07,
+      "loss": 0.5477,
+      "step": 2600
+    },
+    {
+      "epoch": 2.0553549344546127,
+      "grad_norm": 0.23998037775805112,
+      "learning_rate": 4.493210903577472e-07,
+      "loss": 0.5445,
+      "step": 2601
+    },
+    {
+      "epoch": 2.0561464259213453,
+      "grad_norm": 0.22944328523082502,
+      "learning_rate": 4.4862735606435286e-07,
+      "loss": 0.5391,
+      "step": 2602
+    },
+    {
+      "epoch": 2.056937917388078,
+      "grad_norm": 0.2375348653351975,
+      "learning_rate": 4.479340028290174e-07,
+      "loss": 0.5415,
+      "step": 2603
+    },
+    {
+      "epoch": 2.0577294088548106,
+      "grad_norm": 0.24575707157167756,
+      "learning_rate": 4.472410311309236e-07,
+      "loss": 0.5375,
+      "step": 2604
+    },
+    {
+      "epoch": 2.0585209003215432,
+      "grad_norm": 0.23996252446730126,
+      "learning_rate": 4.465484414489897e-07,
+      "loss": 0.5558,
+      "step": 2605
+    },
+    {
+      "epoch": 2.059312391788276,
+      "grad_norm": 0.22860285437539668,
+      "learning_rate": 4.458562342618696e-07,
+      "loss": 0.5521,
+      "step": 2606
+    },
+    {
+      "epoch": 2.0601038832550085,
+      "grad_norm": 0.23868020440907983,
+      "learning_rate": 4.4516441004795393e-07,
+      "loss": 0.5372,
+      "step": 2607
+    },
+    {
+      "epoch": 2.060895374721741,
+      "grad_norm": 0.24299037579234517,
+      "learning_rate": 4.444729692853675e-07,
+      "loss": 0.5387,
+      "step": 2608
+    },
+    {
+      "epoch": 2.0616868661884737,
+      "grad_norm": 0.23582683768296428,
+      "learning_rate": 4.437819124519706e-07,
+      "loss": 0.5398,
+      "step": 2609
+    },
+    {
+      "epoch": 2.0624783576552064,
+      "grad_norm": 0.2388253141147197,
+      "learning_rate": 4.4309124002535824e-07,
+      "loss": 0.554,
+      "step": 2610
+    },
+    {
+      "epoch": 2.063269849121939,
+      "grad_norm": 0.23809082107721263,
+      "learning_rate": 4.4240095248285923e-07,
+      "loss": 0.5452,
+      "step": 2611
+    },
+    {
+      "epoch": 2.0640613405886716,
+      "grad_norm": 0.24457771276029272,
+      "learning_rate": 4.4171105030153764e-07,
+      "loss": 0.5442,
+      "step": 2612
+    },
+    {
+      "epoch": 2.0648528320554043,
+      "grad_norm": 0.2327143037804538,
+      "learning_rate": 4.4102153395819e-07,
+      "loss": 0.5448,
+      "step": 2613
+    },
+    {
+      "epoch": 2.065644323522137,
+      "grad_norm": 0.23749102179802173,
+      "learning_rate": 4.4033240392934656e-07,
+      "loss": 0.5432,
+      "step": 2614
+    },
+    {
+      "epoch": 2.0664358149888695,
+      "grad_norm": 0.23936681376959112,
+      "learning_rate": 4.3964366069127055e-07,
+      "loss": 0.5578,
+      "step": 2615
+    },
+    {
+      "epoch": 2.067227306455602,
+      "grad_norm": 0.24895254802281508,
+      "learning_rate": 4.389553047199578e-07,
+      "loss": 0.5441,
+      "step": 2616
+    },
+    {
+      "epoch": 2.068018797922335,
+      "grad_norm": 0.24822475647937434,
+      "learning_rate": 4.382673364911372e-07,
+      "loss": 0.5389,
+      "step": 2617
+    },
+    {
+      "epoch": 2.0688102893890674,
+      "grad_norm": 0.23421098129477877,
+      "learning_rate": 4.375797564802689e-07,
+      "loss": 0.5378,
+      "step": 2618
+    },
+    {
+      "epoch": 2.0696017808558,
+      "grad_norm": 0.23370889073902257,
+      "learning_rate": 4.368925651625448e-07,
+      "loss": 0.5446,
+      "step": 2619
+    },
+    {
+      "epoch": 2.0703932723225327,
+      "grad_norm": 0.2420375741406236,
+      "learning_rate": 4.3620576301288834e-07,
+      "loss": 0.5523,
+      "step": 2620
+    },
+    {
+      "epoch": 2.0711847637892653,
+      "grad_norm": 0.23700570041043656,
+      "learning_rate": 4.355193505059537e-07,
+      "loss": 0.5402,
+      "step": 2621
+    },
+    {
+      "epoch": 2.071976255255998,
+      "grad_norm": 0.2401536630396451,
+      "learning_rate": 4.348333281161268e-07,
+      "loss": 0.5491,
+      "step": 2622
+    },
+    {
+      "epoch": 2.0727677467227306,
+      "grad_norm": 0.24256794326623432,
+      "learning_rate": 4.341476963175229e-07,
+      "loss": 0.5427,
+      "step": 2623
+    },
+    {
+      "epoch": 2.073559238189463,
+      "grad_norm": 0.24137579698066972,
+      "learning_rate": 4.3346245558398753e-07,
+      "loss": 0.5473,
+      "step": 2624
+    },
+    {
+      "epoch": 2.074350729656196,
+      "grad_norm": 0.2414256494195632,
+      "learning_rate": 4.327776063890962e-07,
+      "loss": 0.5502,
+      "step": 2625
+    },
+    {
+      "epoch": 2.0751422211229285,
+      "grad_norm": 0.2350478866333628,
+      "learning_rate": 4.3209314920615314e-07,
+      "loss": 0.5479,
+      "step": 2626
+    },
+    {
+      "epoch": 2.075933712589661,
+      "grad_norm": 0.24060422874115028,
+      "learning_rate": 4.314090845081931e-07,
+      "loss": 0.5416,
+      "step": 2627
+    },
+    {
+      "epoch": 2.0767252040563937,
+      "grad_norm": 0.24005254014681046,
+      "learning_rate": 4.307254127679782e-07,
+      "loss": 0.5334,
+      "step": 2628
+    },
+    {
+      "epoch": 2.0775166955231263,
+      "grad_norm": 0.23726947495812292,
+      "learning_rate": 4.300421344579996e-07,
+      "loss": 0.5365,
+      "step": 2629
+    },
+    {
+      "epoch": 2.078308186989859,
+      "grad_norm": 0.23920009310993925,
+      "learning_rate": 4.2935925005047624e-07,
+      "loss": 0.535,
+      "step": 2630
+    },
+    {
+      "epoch": 2.0790996784565916,
+      "grad_norm": 0.23120415369048664,
+      "learning_rate": 4.2867676001735486e-07,
+      "loss": 0.537,
+      "step": 2631
+    },
+    {
+      "epoch": 2.0798911699233242,
+      "grad_norm": 0.23700420227631172,
+      "learning_rate": 4.2799466483031033e-07,
+      "loss": 0.5442,
+      "step": 2632
+    },
+    {
+      "epoch": 2.080682661390057,
+      "grad_norm": 0.23965011423463708,
+      "learning_rate": 4.273129649607439e-07,
+      "loss": 0.5435,
+      "step": 2633
+    },
+    {
+      "epoch": 2.0814741528567895,
+      "grad_norm": 0.2346525230558761,
+      "learning_rate": 4.266316608797836e-07,
+      "loss": 0.5268,
+      "step": 2634
+    },
+    {
+      "epoch": 2.082265644323522,
+      "grad_norm": 0.23502062739655638,
+      "learning_rate": 4.2595075305828445e-07,
+      "loss": 0.5304,
+      "step": 2635
+    },
+    {
+      "epoch": 2.0830571357902548,
+      "grad_norm": 0.23455472138946823,
+      "learning_rate": 4.252702419668267e-07,
+      "loss": 0.5307,
+      "step": 2636
+    },
+    {
+      "epoch": 2.0838486272569874,
+      "grad_norm": 0.23408216408686333,
+      "learning_rate": 4.245901280757179e-07,
+      "loss": 0.5501,
+      "step": 2637
+    },
+    {
+      "epoch": 2.08464011872372,
+      "grad_norm": 0.23716650040576878,
+      "learning_rate": 4.2391041185498987e-07,
+      "loss": 0.5493,
+      "step": 2638
+    },
+    {
+      "epoch": 2.0854316101904526,
+      "grad_norm": 0.2376582721488692,
+      "learning_rate": 4.232310937743998e-07,
+      "loss": 0.5548,
+      "step": 2639
+    },
+    {
+      "epoch": 2.0862231016571853,
+      "grad_norm": 0.23617707987290446,
+      "learning_rate": 4.2255217430343005e-07,
+      "loss": 0.5238,
+      "step": 2640
+    },
+    {
+      "epoch": 2.087014593123918,
+      "grad_norm": 0.24201311469735495,
+      "learning_rate": 4.218736539112868e-07,
+      "loss": 0.5417,
+      "step": 2641
+    },
+    {
+      "epoch": 2.0878060845906505,
+      "grad_norm": 0.2377382925311404,
+      "learning_rate": 4.211955330669017e-07,
+      "loss": 0.5458,
+      "step": 2642
+    },
+    {
+      "epoch": 2.088597576057383,
+      "grad_norm": 0.23487792135662103,
+      "learning_rate": 4.205178122389291e-07,
+      "loss": 0.5457,
+      "step": 2643
+    },
+    {
+      "epoch": 2.089389067524116,
+      "grad_norm": 0.2407008574725615,
+      "learning_rate": 4.198404918957474e-07,
+      "loss": 0.5407,
+      "step": 2644
+    },
+    {
+      "epoch": 2.0901805589908484,
+      "grad_norm": 0.24245992583135814,
+      "learning_rate": 4.191635725054582e-07,
+      "loss": 0.5448,
+      "step": 2645
+    },
+    {
+      "epoch": 2.090972050457581,
+      "grad_norm": 0.23423440438327897,
+      "learning_rate": 4.184870545358854e-07,
+      "loss": 0.5296,
+      "step": 2646
+    },
+    {
+      "epoch": 2.0917635419243137,
+      "grad_norm": 0.23723200022314908,
+      "learning_rate": 4.178109384545767e-07,
+      "loss": 0.5424,
+      "step": 2647
+    },
+    {
+      "epoch": 2.0925550333910463,
+      "grad_norm": 0.24589976472930544,
+      "learning_rate": 4.1713522472880115e-07,
+      "loss": 0.544,
+      "step": 2648
+    },
+    {
+      "epoch": 2.093346524857779,
+      "grad_norm": 0.23780461189688573,
+      "learning_rate": 4.1645991382554945e-07,
+      "loss": 0.528,
+      "step": 2649
+    },
+    {
+      "epoch": 2.0941380163245116,
+      "grad_norm": 0.23329018650218644,
+      "learning_rate": 4.1578500621153556e-07,
+      "loss": 0.5597,
+      "step": 2650
+    },
+    {
+      "epoch": 2.094929507791244,
+      "grad_norm": 0.24396786237580703,
+      "learning_rate": 4.15110502353192e-07,
+      "loss": 0.5427,
+      "step": 2651
+    },
+    {
+      "epoch": 2.095720999257977,
+      "grad_norm": 0.2390954572392998,
+      "learning_rate": 4.144364027166749e-07,
+      "loss": 0.5389,
+      "step": 2652
+    },
+    {
+      "epoch": 2.0965124907247095,
+      "grad_norm": 0.2309555686291755,
+      "learning_rate": 4.137627077678596e-07,
+      "loss": 0.5372,
+      "step": 2653
+    },
+    {
+      "epoch": 2.097303982191442,
+      "grad_norm": 0.24066560664586562,
+      "learning_rate": 4.1308941797234174e-07,
+      "loss": 0.5585,
+      "step": 2654
+    },
+    {
+      "epoch": 2.0980954736581747,
+      "grad_norm": 0.23979703252462348,
+      "learning_rate": 4.124165337954384e-07,
+      "loss": 0.543,
+      "step": 2655
+    },
+    {
+      "epoch": 2.0988869651249074,
+      "grad_norm": 0.2462660341072815,
+      "learning_rate": 4.1174405570218375e-07,
+      "loss": 0.5466,
+      "step": 2656
+    },
+    {
+      "epoch": 2.09967845659164,
+      "grad_norm": 0.23253145245822232,
+      "learning_rate": 4.110719841573339e-07,
+      "loss": 0.5253,
+      "step": 2657
+    },
+    {
+      "epoch": 2.1004699480583726,
+      "grad_norm": 0.2399868131554026,
+      "learning_rate": 4.104003196253625e-07,
+      "loss": 0.5537,
+      "step": 2658
+    },
+    {
+      "epoch": 2.1012614395251052,
+      "grad_norm": 0.2323392761111308,
+      "learning_rate": 4.097290625704619e-07,
+      "loss": 0.5472,
+      "step": 2659
+    },
+    {
+      "epoch": 2.102052930991838,
+      "grad_norm": 0.24553950723811055,
+      "learning_rate": 4.0905821345654437e-07,
+      "loss": 0.5445,
+      "step": 2660
+    },
+    {
+      "epoch": 2.1028444224585705,
+      "grad_norm": 0.2428412102787307,
+      "learning_rate": 4.083877727472378e-07,
+      "loss": 0.5346,
+      "step": 2661
+    },
+    {
+      "epoch": 2.103635913925303,
+      "grad_norm": 0.235044009071209,
+      "learning_rate": 4.0771774090588994e-07,
+      "loss": 0.5267,
+      "step": 2662
+    },
+    {
+      "epoch": 2.1044274053920358,
+      "grad_norm": 0.23369441567291457,
+      "learning_rate": 4.0704811839556506e-07,
+      "loss": 0.5317,
+      "step": 2663
+    },
+    {
+      "epoch": 2.1052188968587684,
+      "grad_norm": 0.23642228760111267,
+      "learning_rate": 4.063789056790442e-07,
+      "loss": 0.5437,
+      "step": 2664
+    },
+    {
+      "epoch": 2.106010388325501,
+      "grad_norm": 0.24511624494914874,
+      "learning_rate": 4.0571010321882683e-07,
+      "loss": 0.5569,
+      "step": 2665
+    },
+    {
+      "epoch": 2.1068018797922337,
+      "grad_norm": 0.23807134925894802,
+      "learning_rate": 4.050417114771261e-07,
+      "loss": 0.5445,
+      "step": 2666
+    },
+    {
+      "epoch": 2.1075933712589663,
+      "grad_norm": 0.2450748062872641,
+      "learning_rate": 4.04373730915874e-07,
+      "loss": 0.5403,
+      "step": 2667
+    },
+    {
+      "epoch": 2.108384862725699,
+      "grad_norm": 0.22752898002700206,
+      "learning_rate": 4.037061619967169e-07,
+      "loss": 0.5344,
+      "step": 2668
+    },
+    {
+      "epoch": 2.1091763541924315,
+      "grad_norm": 0.235979027692674,
+      "learning_rate": 4.0303900518101685e-07,
+      "loss": 0.5514,
+      "step": 2669
+    },
+    {
+      "epoch": 2.109967845659164,
+      "grad_norm": 0.2358535907753104,
+      "learning_rate": 4.023722609298521e-07,
+      "loss": 0.5616,
+      "step": 2670
+    },
+    {
+      "epoch": 2.110759337125897,
+      "grad_norm": 0.23800463564472196,
+      "learning_rate": 4.0170592970401373e-07,
+      "loss": 0.5383,
+      "step": 2671
+    },
+    {
+      "epoch": 2.1115508285926294,
+      "grad_norm": 0.24083426114622639,
+      "learning_rate": 4.010400119640095e-07,
+      "loss": 0.5347,
+      "step": 2672
+    },
+    {
+      "epoch": 2.112342320059362,
+      "grad_norm": 0.24370066558921152,
+      "learning_rate": 4.0037450817006037e-07,
+      "loss": 0.5503,
+      "step": 2673
+    },
+    {
+      "epoch": 2.1131338115260947,
+      "grad_norm": 0.23224467494246734,
+      "learning_rate": 3.9970941878210076e-07,
+      "loss": 0.5505,
+      "step": 2674
+    },
+    {
+      "epoch": 2.1139253029928273,
+      "grad_norm": 0.23644803338116782,
+      "learning_rate": 3.990447442597801e-07,
+      "loss": 0.5529,
+      "step": 2675
+    },
+    {
+      "epoch": 2.11471679445956,
+      "grad_norm": 0.2391182606163028,
+      "learning_rate": 3.9838048506246e-07,
+      "loss": 0.5438,
+      "step": 2676
+    },
+    {
+      "epoch": 2.115508285926292,
+      "grad_norm": 0.235304197389743,
+      "learning_rate": 3.9771664164921516e-07,
+      "loss": 0.5535,
+      "step": 2677
+    },
+    {
+      "epoch": 2.1162997773930248,
+      "grad_norm": 0.23629465426666046,
+      "learning_rate": 3.970532144788332e-07,
+      "loss": 0.5467,
+      "step": 2678
+    },
+    {
+      "epoch": 2.1170912688597574,
+      "grad_norm": 0.23914376324070952,
+      "learning_rate": 3.963902040098137e-07,
+      "loss": 0.5376,
+      "step": 2679
+    },
+    {
+      "epoch": 2.11788276032649,
+      "grad_norm": 0.2416501276838314,
+      "learning_rate": 3.957276107003691e-07,
+      "loss": 0.5537,
+      "step": 2680
+    },
+    {
+      "epoch": 2.1186742517932227,
+      "grad_norm": 0.23513994997497853,
+      "learning_rate": 3.950654350084225e-07,
+      "loss": 0.5361,
+      "step": 2681
+    },
+    {
+      "epoch": 2.1194657432599553,
+      "grad_norm": 0.23945735733077111,
+      "learning_rate": 3.9440367739160894e-07,
+      "loss": 0.5435,
+      "step": 2682
+    },
+    {
+      "epoch": 2.120257234726688,
+      "grad_norm": 0.24057992870637823,
+      "learning_rate": 3.9374233830727453e-07,
+      "loss": 0.5642,
+      "step": 2683
+    },
+    {
+      "epoch": 2.1210487261934206,
+      "grad_norm": 0.23542654349540829,
+      "learning_rate": 3.9308141821247555e-07,
+      "loss": 0.5329,
+      "step": 2684
+    },
+    {
+      "epoch": 2.121840217660153,
+      "grad_norm": 0.23207027638096936,
+      "learning_rate": 3.924209175639799e-07,
+      "loss": 0.5372,
+      "step": 2685
+    },
+    {
+      "epoch": 2.122631709126886,
+      "grad_norm": 0.241740908022761,
+      "learning_rate": 3.9176083681826487e-07,
+      "loss": 0.5293,
+      "step": 2686
+    },
+    {
+      "epoch": 2.1234232005936184,
+      "grad_norm": 0.23303805496442506,
+      "learning_rate": 3.911011764315165e-07,
+      "loss": 0.5365,
+      "step": 2687
+    },
+    {
+      "epoch": 2.124214692060351,
+      "grad_norm": 0.23518516070727688,
+      "learning_rate": 3.9044193685963254e-07,
+      "loss": 0.5359,
+      "step": 2688
+    },
+    {
+      "epoch": 2.1250061835270837,
+      "grad_norm": 0.2312937285854721,
+      "learning_rate": 3.8978311855821787e-07,
+      "loss": 0.5273,
+      "step": 2689
+    },
+    {
+      "epoch": 2.1257976749938163,
+      "grad_norm": 0.240795395078905,
+      "learning_rate": 3.891247219825879e-07,
+      "loss": 0.545,
+      "step": 2690
+    },
+    {
+      "epoch": 2.126589166460549,
+      "grad_norm": 0.22992709791566163,
+      "learning_rate": 3.884667475877659e-07,
+      "loss": 0.5427,
+      "step": 2691
+    },
+    {
+      "epoch": 2.1273806579272816,
+      "grad_norm": 0.23547723324550332,
+      "learning_rate": 3.8780919582848194e-07,
+      "loss": 0.5518,
+      "step": 2692
+    },
+    {
+      "epoch": 2.1281721493940142,
+      "grad_norm": 0.2376023112368025,
+      "learning_rate": 3.871520671591766e-07,
+      "loss": 0.544,
+      "step": 2693
+    },
+    {
+      "epoch": 2.128963640860747,
+      "grad_norm": 0.2485137223948424,
+      "learning_rate": 3.864953620339959e-07,
+      "loss": 0.5314,
+      "step": 2694
+    },
+    {
+      "epoch": 2.1297551323274795,
+      "grad_norm": 0.23854155144796496,
+      "learning_rate": 3.8583908090679495e-07,
+      "loss": 0.553,
+      "step": 2695
+    },
+    {
+      "epoch": 2.130546623794212,
+      "grad_norm": 0.24172009255298904,
+      "learning_rate": 3.8518322423113423e-07,
+      "loss": 0.5381,
+      "step": 2696
+    },
+    {
+      "epoch": 2.1313381152609447,
+      "grad_norm": 0.24317528384140413,
+      "learning_rate": 3.84527792460282e-07,
+      "loss": 0.5581,
+      "step": 2697
+    },
+    {
+      "epoch": 2.1321296067276774,
+      "grad_norm": 0.23843831512216013,
+      "learning_rate": 3.83872786047212e-07,
+      "loss": 0.5414,
+      "step": 2698
+    },
+    {
+      "epoch": 2.13292109819441,
+      "grad_norm": 0.2375907806760504,
+      "learning_rate": 3.8321820544460427e-07,
+      "loss": 0.5536,
+      "step": 2699
+    },
+    {
+      "epoch": 2.1337125896611426,
+      "grad_norm": 0.23910046786798353,
+      "learning_rate": 3.8256405110484535e-07,
+      "loss": 0.5492,
+      "step": 2700
+    },
+    {
+      "epoch": 2.1345040811278753,
+      "grad_norm": 0.2314723668263214,
+      "learning_rate": 3.819103234800263e-07,
+      "loss": 0.5225,
+      "step": 2701
+    },
+    {
+      "epoch": 2.135295572594608,
+      "grad_norm": 0.23372676483647006,
+      "learning_rate": 3.812570230219433e-07,
+      "loss": 0.5447,
+      "step": 2702
+    },
+    {
+      "epoch": 2.1360870640613405,
+      "grad_norm": 0.24784857348868877,
+      "learning_rate": 3.806041501820979e-07,
+      "loss": 0.5445,
+      "step": 2703
+    },
+    {
+      "epoch": 2.136878555528073,
+      "grad_norm": 0.23839361925335847,
+      "learning_rate": 3.799517054116954e-07,
+      "loss": 0.5341,
+      "step": 2704
+    },
+    {
+      "epoch": 2.137670046994806,
+      "grad_norm": 0.2376321030641214,
+      "learning_rate": 3.7929968916164554e-07,
+      "loss": 0.5401,
+      "step": 2705
+    },
+    {
+      "epoch": 2.1384615384615384,
+      "grad_norm": 0.23541406069684748,
+      "learning_rate": 3.7864810188256255e-07,
+      "loss": 0.542,
+      "step": 2706
+    },
+    {
+      "epoch": 2.139253029928271,
+      "grad_norm": 0.23830045324424498,
+      "learning_rate": 3.779969440247632e-07,
+      "loss": 0.5509,
+      "step": 2707
+    },
+    {
+      "epoch": 2.1400445213950037,
+      "grad_norm": 0.23907942240071353,
+      "learning_rate": 3.773462160382681e-07,
+      "loss": 0.5471,
+      "step": 2708
+    },
+    {
+      "epoch": 2.1408360128617363,
+      "grad_norm": 0.2396199043443259,
+      "learning_rate": 3.766959183728006e-07,
+      "loss": 0.5434,
+      "step": 2709
+    },
+    {
+      "epoch": 2.141627504328469,
+      "grad_norm": 0.24243347753188646,
+      "learning_rate": 3.76046051477786e-07,
+      "loss": 0.5396,
+      "step": 2710
+    },
+    {
+      "epoch": 2.1424189957952016,
+      "grad_norm": 0.23490660986564252,
+      "learning_rate": 3.753966158023535e-07,
+      "loss": 0.5494,
+      "step": 2711
+    },
+    {
+      "epoch": 2.143210487261934,
+      "grad_norm": 0.23643994727318168,
+      "learning_rate": 3.7474761179533287e-07,
+      "loss": 0.5364,
+      "step": 2712
+    },
+    {
+      "epoch": 2.144001978728667,
+      "grad_norm": 0.23813687206078374,
+      "learning_rate": 3.7409903990525593e-07,
+      "loss": 0.5655,
+      "step": 2713
+    },
+    {
+      "epoch": 2.1447934701953995,
+      "grad_norm": 0.23658533828618422,
+      "learning_rate": 3.73450900580356e-07,
+      "loss": 0.5499,
+      "step": 2714
+    },
+    {
+      "epoch": 2.145584961662132,
+      "grad_norm": 0.2378604628576131,
+      "learning_rate": 3.72803194268567e-07,
+      "loss": 0.5421,
+      "step": 2715
+    },
+    {
+      "epoch": 2.1463764531288647,
+      "grad_norm": 0.23632934850343384,
+      "learning_rate": 3.721559214175246e-07,
+      "loss": 0.5422,
+      "step": 2716
+    },
+    {
+      "epoch": 2.1471679445955973,
+      "grad_norm": 0.23413369926924474,
+      "learning_rate": 3.715090824745637e-07,
+      "loss": 0.5408,
+      "step": 2717
+    },
+    {
+      "epoch": 2.14795943606233,
+      "grad_norm": 0.23640766332836996,
+      "learning_rate": 3.708626778867208e-07,
+      "loss": 0.5502,
+      "step": 2718
+    },
+    {
+      "epoch": 2.1487509275290626,
+      "grad_norm": 0.23533575904719933,
+      "learning_rate": 3.7021670810073023e-07,
+      "loss": 0.5363,
+      "step": 2719
+    },
+    {
+      "epoch": 2.1495424189957952,
+      "grad_norm": 0.24218923848920135,
+      "learning_rate": 3.6957117356302715e-07,
+      "loss": 0.5403,
+      "step": 2720
+    },
+    {
+      "epoch": 2.150333910462528,
+      "grad_norm": 0.236820793190202,
+      "learning_rate": 3.6892607471974603e-07,
+      "loss": 0.551,
+      "step": 2721
+    },
+    {
+      "epoch": 2.1511254019292605,
+      "grad_norm": 0.2410691061573119,
+      "learning_rate": 3.682814120167197e-07,
+      "loss": 0.5397,
+      "step": 2722
+    },
+    {
+      "epoch": 2.151916893395993,
+      "grad_norm": 0.23352531474515983,
+      "learning_rate": 3.6763718589947967e-07,
+      "loss": 0.5391,
+      "step": 2723
+    },
+    {
+      "epoch": 2.1527083848627258,
+      "grad_norm": 0.2365230560264958,
+      "learning_rate": 3.6699339681325567e-07,
+      "loss": 0.5392,
+      "step": 2724
+    },
+    {
+      "epoch": 2.1534998763294584,
+      "grad_norm": 0.2360532844037301,
+      "learning_rate": 3.663500452029755e-07,
+      "loss": 0.554,
+      "step": 2725
+    },
+    {
+      "epoch": 2.154291367796191,
+      "grad_norm": 0.2420408481557582,
+      "learning_rate": 3.6570713151326494e-07,
+      "loss": 0.5411,
+      "step": 2726
+    },
+    {
+      "epoch": 2.1550828592629236,
+      "grad_norm": 0.23247463401551066,
+      "learning_rate": 3.6506465618844664e-07,
+      "loss": 0.5322,
+      "step": 2727
+    },
+    {
+      "epoch": 2.1558743507296563,
+      "grad_norm": 0.23236867983238696,
+      "learning_rate": 3.644226196725404e-07,
+      "loss": 0.5312,
+      "step": 2728
+    },
+    {
+      "epoch": 2.156665842196389,
+      "grad_norm": 0.24358870002051564,
+      "learning_rate": 3.637810224092629e-07,
+      "loss": 0.5464,
+      "step": 2729
+    },
+    {
+      "epoch": 2.1574573336631215,
+      "grad_norm": 0.2364477616915208,
+      "learning_rate": 3.6313986484202666e-07,
+      "loss": 0.5411,
+      "step": 2730
+    },
+    {
+      "epoch": 2.158248825129854,
+      "grad_norm": 0.23768339433671795,
+      "learning_rate": 3.624991474139416e-07,
+      "loss": 0.5406,
+      "step": 2731
+    },
+    {
+      "epoch": 2.159040316596587,
+      "grad_norm": 0.24196534113116164,
+      "learning_rate": 3.618588705678124e-07,
+      "loss": 0.5455,
+      "step": 2732
+    },
+    {
+      "epoch": 2.1598318080633194,
+      "grad_norm": 0.2352209662500066,
+      "learning_rate": 3.6121903474613935e-07,
+      "loss": 0.528,
+      "step": 2733
+    },
+    {
+      "epoch": 2.160623299530052,
+      "grad_norm": 0.23640826060048364,
+      "learning_rate": 3.6057964039111834e-07,
+      "loss": 0.5527,
+      "step": 2734
+    },
+    {
+      "epoch": 2.1614147909967847,
+      "grad_norm": 0.23550542411389946,
+      "learning_rate": 3.599406879446394e-07,
+      "loss": 0.5534,
+      "step": 2735
+    },
+    {
+      "epoch": 2.1622062824635173,
+      "grad_norm": 0.23195874916767886,
+      "learning_rate": 3.593021778482884e-07,
+      "loss": 0.5356,
+      "step": 2736
+    },
+    {
+      "epoch": 2.16299777393025,
+      "grad_norm": 0.23604071060154344,
+      "learning_rate": 3.586641105433447e-07,
+      "loss": 0.5375,
+      "step": 2737
+    },
+    {
+      "epoch": 2.1637892653969826,
+      "grad_norm": 0.23808981850245756,
+      "learning_rate": 3.5802648647078107e-07,
+      "loss": 0.5546,
+      "step": 2738
+    },
+    {
+      "epoch": 2.164580756863715,
+      "grad_norm": 0.23328256503214606,
+      "learning_rate": 3.573893060712657e-07,
+      "loss": 0.5396,
+      "step": 2739
+    },
+    {
+      "epoch": 2.165372248330448,
+      "grad_norm": 0.23538990673898114,
+      "learning_rate": 3.567525697851579e-07,
+      "loss": 0.5349,
+      "step": 2740
+    },
+    {
+      "epoch": 2.1661637397971805,
+      "grad_norm": 0.24033583361609867,
+      "learning_rate": 3.561162780525122e-07,
+      "loss": 0.5375,
+      "step": 2741
+    },
+    {
+      "epoch": 2.166955231263913,
+      "grad_norm": 0.23752082108251496,
+      "learning_rate": 3.5548043131307437e-07,
+      "loss": 0.546,
+      "step": 2742
+    },
+    {
+      "epoch": 2.1677467227306457,
+      "grad_norm": 0.2420787082419534,
+      "learning_rate": 3.548450300062832e-07,
+      "loss": 0.5408,
+      "step": 2743
+    },
+    {
+      "epoch": 2.1685382141973784,
+      "grad_norm": 0.239040103059141,
+      "learning_rate": 3.542100745712704e-07,
+      "loss": 0.5351,
+      "step": 2744
+    },
+    {
+      "epoch": 2.169329705664111,
+      "grad_norm": 0.23789310410737252,
+      "learning_rate": 3.535755654468574e-07,
+      "loss": 0.5364,
+      "step": 2745
+    },
+    {
+      "epoch": 2.1701211971308436,
+      "grad_norm": 0.2509356328833936,
+      "learning_rate": 3.5294150307155965e-07,
+      "loss": 0.5467,
+      "step": 2746
+    },
+    {
+      "epoch": 2.1709126885975762,
+      "grad_norm": 0.23855493455386934,
+      "learning_rate": 3.5230788788358244e-07,
+      "loss": 0.5541,
+      "step": 2747
+    },
+    {
+      "epoch": 2.171704180064309,
+      "grad_norm": 0.2320380484076634,
+      "learning_rate": 3.5167472032082193e-07,
+      "loss": 0.5419,
+      "step": 2748
+    },
+    {
+      "epoch": 2.1724956715310415,
+      "grad_norm": 0.24097966011385014,
+      "learning_rate": 3.510420008208663e-07,
+      "loss": 0.5472,
+      "step": 2749
+    },
+    {
+      "epoch": 2.173287162997774,
+      "grad_norm": 0.24582658971565477,
+      "learning_rate": 3.504097298209918e-07,
+      "loss": 0.5444,
+      "step": 2750
+    },
+    {
+      "epoch": 2.1740786544645063,
+      "grad_norm": 0.2410878215392785,
+      "learning_rate": 3.4977790775816694e-07,
+      "loss": 0.5412,
+      "step": 2751
+    },
+    {
+      "epoch": 2.174870145931239,
+      "grad_norm": 0.23521436046476052,
+      "learning_rate": 3.491465350690487e-07,
+      "loss": 0.5375,
+      "step": 2752
+    },
+    {
+      "epoch": 2.1756616373979716,
+      "grad_norm": 0.23996796346983834,
+      "learning_rate": 3.4851561218998337e-07,
+      "loss": 0.5446,
+      "step": 2753
+    },
+    {
+      "epoch": 2.176453128864704,
+      "grad_norm": 0.23994080060671666,
+      "learning_rate": 3.4788513955700794e-07,
+      "loss": 0.5511,
+      "step": 2754
+    },
+    {
+      "epoch": 2.177244620331437,
+      "grad_norm": 0.24092313468444498,
+      "learning_rate": 3.472551176058458e-07,
+      "loss": 0.5527,
+      "step": 2755
+    },
+    {
+      "epoch": 2.1780361117981695,
+      "grad_norm": 0.24460037463040793,
+      "learning_rate": 3.4662554677191115e-07,
+      "loss": 0.5436,
+      "step": 2756
+    },
+    {
+      "epoch": 2.178827603264902,
+      "grad_norm": 0.2352021106165701,
+      "learning_rate": 3.45996427490305e-07,
+      "loss": 0.5514,
+      "step": 2757
+    },
+    {
+      "epoch": 2.1796190947316347,
+      "grad_norm": 0.23613075232978054,
+      "learning_rate": 3.453677601958166e-07,
+      "loss": 0.5373,
+      "step": 2758
+    },
+    {
+      "epoch": 2.1804105861983674,
+      "grad_norm": 0.23833584609528025,
+      "learning_rate": 3.4473954532292336e-07,
+      "loss": 0.5386,
+      "step": 2759
+    },
+    {
+      "epoch": 2.1812020776651,
+      "grad_norm": 0.24048857569727802,
+      "learning_rate": 3.441117833057896e-07,
+      "loss": 0.5381,
+      "step": 2760
+    },
+    {
+      "epoch": 2.1819935691318326,
+      "grad_norm": 0.2314595284658549,
+      "learning_rate": 3.4348447457826646e-07,
+      "loss": 0.5464,
+      "step": 2761
+    },
+    {
+      "epoch": 2.1827850605985653,
+      "grad_norm": 0.23498305258167132,
+      "learning_rate": 3.428576195738921e-07,
+      "loss": 0.5318,
+      "step": 2762
+    },
+    {
+      "epoch": 2.183576552065298,
+      "grad_norm": 0.23942660920743883,
+      "learning_rate": 3.4223121872589076e-07,
+      "loss": 0.5429,
+      "step": 2763
+    },
+    {
+      "epoch": 2.1843680435320305,
+      "grad_norm": 0.23629933727881583,
+      "learning_rate": 3.416052724671737e-07,
+      "loss": 0.5389,
+      "step": 2764
+    },
+    {
+      "epoch": 2.185159534998763,
+      "grad_norm": 0.2484889529273199,
+      "learning_rate": 3.409797812303371e-07,
+      "loss": 0.5379,
+      "step": 2765
+    },
+    {
+      "epoch": 2.1859510264654958,
+      "grad_norm": 0.2342655173341307,
+      "learning_rate": 3.4035474544766306e-07,
+      "loss": 0.5437,
+      "step": 2766
+    },
+    {
+      "epoch": 2.1867425179322284,
+      "grad_norm": 0.24166378020389223,
+      "learning_rate": 3.3973016555111865e-07,
+      "loss": 0.5357,
+      "step": 2767
+    },
+    {
+      "epoch": 2.187534009398961,
+      "grad_norm": 0.2332169509558839,
+      "learning_rate": 3.391060419723558e-07,
+      "loss": 0.5469,
+      "step": 2768
+    },
+    {
+      "epoch": 2.1883255008656937,
+      "grad_norm": 0.24274548354086564,
+      "learning_rate": 3.3848237514271196e-07,
+      "loss": 0.5408,
+      "step": 2769
+    },
+    {
+      "epoch": 2.1891169923324263,
+      "grad_norm": 0.23325643021174935,
+      "learning_rate": 3.3785916549320804e-07,
+      "loss": 0.5429,
+      "step": 2770
+    },
+    {
+      "epoch": 2.189908483799159,
+      "grad_norm": 0.23626208169532442,
+      "learning_rate": 3.372364134545491e-07,
+      "loss": 0.5428,
+      "step": 2771
+    },
+    {
+      "epoch": 2.1906999752658916,
+      "grad_norm": 0.23943777925930157,
+      "learning_rate": 3.3661411945712415e-07,
+      "loss": 0.5461,
+      "step": 2772
+    },
+    {
+      "epoch": 2.191491466732624,
+      "grad_norm": 0.23883505497909377,
+      "learning_rate": 3.359922839310053e-07,
+      "loss": 0.5426,
+      "step": 2773
+    },
+    {
+      "epoch": 2.192282958199357,
+      "grad_norm": 0.23593992088353088,
+      "learning_rate": 3.3537090730594864e-07,
+      "loss": 0.5539,
+      "step": 2774
+    },
+    {
+      "epoch": 2.1930744496660894,
+      "grad_norm": 0.24296391543886683,
+      "learning_rate": 3.3474999001139214e-07,
+      "loss": 0.5502,
+      "step": 2775
+    },
+    {
+      "epoch": 2.193865941132822,
+      "grad_norm": 0.23568483856630298,
+      "learning_rate": 3.341295324764569e-07,
+      "loss": 0.5388,
+      "step": 2776
+    },
+    {
+      "epoch": 2.1946574325995547,
+      "grad_norm": 0.23613563619882116,
+      "learning_rate": 3.3350953512994606e-07,
+      "loss": 0.5534,
+      "step": 2777
+    },
+    {
+      "epoch": 2.1954489240662873,
+      "grad_norm": 0.24378112398366505,
+      "learning_rate": 3.3288999840034425e-07,
+      "loss": 0.548,
+      "step": 2778
+    },
+    {
+      "epoch": 2.19624041553302,
+      "grad_norm": 0.23251859526158788,
+      "learning_rate": 3.3227092271581924e-07,
+      "loss": 0.5424,
+      "step": 2779
+    },
+    {
+      "epoch": 2.1970319069997526,
+      "grad_norm": 0.23370396151069003,
+      "learning_rate": 3.3165230850421864e-07,
+      "loss": 0.5496,
+      "step": 2780
+    },
+    {
+      "epoch": 2.1978233984664852,
+      "grad_norm": 0.24493607278999252,
+      "learning_rate": 3.3103415619307175e-07,
+      "loss": 0.5425,
+      "step": 2781
+    },
+    {
+      "epoch": 2.198614889933218,
+      "grad_norm": 0.2441568731723912,
+      "learning_rate": 3.304164662095886e-07,
+      "loss": 0.5433,
+      "step": 2782
+    },
+    {
+      "epoch": 2.1994063813999505,
+      "grad_norm": 0.23975135718916102,
+      "learning_rate": 3.297992389806594e-07,
+      "loss": 0.5467,
+      "step": 2783
+    },
+    {
+      "epoch": 2.200197872866683,
+      "grad_norm": 0.23788564669911225,
+      "learning_rate": 3.2918247493285556e-07,
+      "loss": 0.551,
+      "step": 2784
+    },
+    {
+      "epoch": 2.2009893643334157,
+      "grad_norm": 0.23649951902652033,
+      "learning_rate": 3.285661744924272e-07,
+      "loss": 0.5573,
+      "step": 2785
+    },
+    {
+      "epoch": 2.2017808558001484,
+      "grad_norm": 0.2345774311166057,
+      "learning_rate": 3.2795033808530445e-07,
+      "loss": 0.5507,
+      "step": 2786
+    },
+    {
+      "epoch": 2.202572347266881,
+      "grad_norm": 0.23692663247155593,
+      "learning_rate": 3.273349661370971e-07,
+      "loss": 0.5511,
+      "step": 2787
+    },
+    {
+      "epoch": 2.2033638387336136,
+      "grad_norm": 0.23441563157854112,
+      "learning_rate": 3.2672005907309297e-07,
+      "loss": 0.5446,
+      "step": 2788
+    },
+    {
+      "epoch": 2.2041553302003463,
+      "grad_norm": 0.23310473967990403,
+      "learning_rate": 3.261056173182601e-07,
+      "loss": 0.5402,
+      "step": 2789
+    },
+    {
+      "epoch": 2.204946821667079,
+      "grad_norm": 0.24161903549455568,
+      "learning_rate": 3.254916412972437e-07,
+      "loss": 0.5405,
+      "step": 2790
+    },
+    {
+      "epoch": 2.2057383131338115,
+      "grad_norm": 0.24051136701477105,
+      "learning_rate": 3.248781314343677e-07,
+      "loss": 0.5501,
+      "step": 2791
+    },
+    {
+      "epoch": 2.206529804600544,
+      "grad_norm": 0.2348785076805987,
+      "learning_rate": 3.242650881536335e-07,
+      "loss": 0.5424,
+      "step": 2792
+    },
+    {
+      "epoch": 2.207321296067277,
+      "grad_norm": 0.23919692370677437,
+      "learning_rate": 3.2365251187872e-07,
+      "loss": 0.5327,
+      "step": 2793
+    },
+    {
+      "epoch": 2.2081127875340094,
+      "grad_norm": 0.23552891269201812,
+      "learning_rate": 3.230404030329842e-07,
+      "loss": 0.545,
+      "step": 2794
+    },
+    {
+      "epoch": 2.208904279000742,
+      "grad_norm": 0.24049913539703127,
+      "learning_rate": 3.224287620394591e-07,
+      "loss": 0.5412,
+      "step": 2795
+    },
+    {
+      "epoch": 2.2096957704674747,
+      "grad_norm": 0.2350548822292516,
+      "learning_rate": 3.218175893208548e-07,
+      "loss": 0.5505,
+      "step": 2796
+    },
+    {
+      "epoch": 2.2104872619342073,
+      "grad_norm": 0.2340448502858718,
+      "learning_rate": 3.212068852995574e-07,
+      "loss": 0.5516,
+      "step": 2797
+    },
+    {
+      "epoch": 2.21127875340094,
+      "grad_norm": 0.23892522416827608,
+      "learning_rate": 3.205966503976294e-07,
+      "loss": 0.5501,
+      "step": 2798
+    },
+    {
+      "epoch": 2.2120702448676726,
+      "grad_norm": 0.23321367696906653,
+      "learning_rate": 3.199868850368095e-07,
+      "loss": 0.5481,
+      "step": 2799
+    },
+    {
+      "epoch": 2.212861736334405,
+      "grad_norm": 0.24740200464824078,
+      "learning_rate": 3.1937758963851105e-07,
+      "loss": 0.5539,
+      "step": 2800
+    },
+    {
+      "epoch": 2.213653227801138,
+      "grad_norm": 0.2316305960967838,
+      "learning_rate": 3.187687646238227e-07,
+      "loss": 0.5377,
+      "step": 2801
+    },
+    {
+      "epoch": 2.2144447192678705,
+      "grad_norm": 0.24081111597547666,
+      "learning_rate": 3.181604104135094e-07,
+      "loss": 0.5358,
+      "step": 2802
+    },
+    {
+      "epoch": 2.215236210734603,
+      "grad_norm": 0.23645305344607873,
+      "learning_rate": 3.1755252742800817e-07,
+      "loss": 0.5387,
+      "step": 2803
+    },
+    {
+      "epoch": 2.2160277022013357,
+      "grad_norm": 0.23275317040884852,
+      "learning_rate": 3.169451160874329e-07,
+      "loss": 0.5323,
+      "step": 2804
+    },
+    {
+      "epoch": 2.2168191936680683,
+      "grad_norm": 0.2334082778496196,
+      "learning_rate": 3.1633817681157005e-07,
+      "loss": 0.5423,
+      "step": 2805
+    },
+    {
+      "epoch": 2.217610685134801,
+      "grad_norm": 0.2427650789454405,
+      "learning_rate": 3.1573171001987996e-07,
+      "loss": 0.5258,
+      "step": 2806
+    },
+    {
+      "epoch": 2.2184021766015336,
+      "grad_norm": 0.23771869244918228,
+      "learning_rate": 3.1512571613149764e-07,
+      "loss": 0.5377,
+      "step": 2807
+    },
+    {
+      "epoch": 2.2191936680682662,
+      "grad_norm": 0.2415593130470467,
+      "learning_rate": 3.145201955652289e-07,
+      "loss": 0.534,
+      "step": 2808
+    },
+    {
+      "epoch": 2.219985159534999,
+      "grad_norm": 0.2340140824836716,
+      "learning_rate": 3.139151487395552e-07,
+      "loss": 0.5502,
+      "step": 2809
+    },
+    {
+      "epoch": 2.2207766510017315,
+      "grad_norm": 0.2401592208945918,
+      "learning_rate": 3.1331057607262854e-07,
+      "loss": 0.5316,
+      "step": 2810
+    },
+    {
+      "epoch": 2.221568142468464,
+      "grad_norm": 0.24120828857298485,
+      "learning_rate": 3.127064779822738e-07,
+      "loss": 0.5503,
+      "step": 2811
+    },
+    {
+      "epoch": 2.2223596339351968,
+      "grad_norm": 0.24117083049827034,
+      "learning_rate": 3.1210285488598897e-07,
+      "loss": 0.5647,
+      "step": 2812
+    },
+    {
+      "epoch": 2.2231511254019294,
+      "grad_norm": 0.23528875437046673,
+      "learning_rate": 3.114997072009419e-07,
+      "loss": 0.544,
+      "step": 2813
+    },
+    {
+      "epoch": 2.223942616868662,
+      "grad_norm": 0.23486315716716744,
+      "learning_rate": 3.108970353439727e-07,
+      "loss": 0.5389,
+      "step": 2814
+    },
+    {
+      "epoch": 2.2247341083353946,
+      "grad_norm": 0.23108068584911928,
+      "learning_rate": 3.102948397315934e-07,
+      "loss": 0.5139,
+      "step": 2815
+    },
+    {
+      "epoch": 2.2255255998021273,
+      "grad_norm": 0.23643343155253604,
+      "learning_rate": 3.0969312077998564e-07,
+      "loss": 0.5353,
+      "step": 2816
+    },
+    {
+      "epoch": 2.22631709126886,
+      "grad_norm": 0.24404072862273804,
+      "learning_rate": 3.090918789050031e-07,
+      "loss": 0.5417,
+      "step": 2817
+    },
+    {
+      "epoch": 2.2271085827355925,
+      "grad_norm": 0.234829036271383,
+      "learning_rate": 3.084911145221679e-07,
+      "loss": 0.5307,
+      "step": 2818
+    },
+    {
+      "epoch": 2.227900074202325,
+      "grad_norm": 0.23909490101015032,
+      "learning_rate": 3.0789082804667323e-07,
+      "loss": 0.5587,
+      "step": 2819
+    },
+    {
+      "epoch": 2.228691565669058,
+      "grad_norm": 0.24775122596074298,
+      "learning_rate": 3.0729101989338247e-07,
+      "loss": 0.5468,
+      "step": 2820
+    },
+    {
+      "epoch": 2.2294830571357904,
+      "grad_norm": 0.2342429049625052,
+      "learning_rate": 3.066916904768271e-07,
+      "loss": 0.5328,
+      "step": 2821
+    },
+    {
+      "epoch": 2.230274548602523,
+      "grad_norm": 0.24281846191245626,
+      "learning_rate": 3.060928402112096e-07,
+      "loss": 0.5346,
+      "step": 2822
+    },
+    {
+      "epoch": 2.2310660400692557,
+      "grad_norm": 0.2444610729786211,
+      "learning_rate": 3.0549446951039927e-07,
+      "loss": 0.5505,
+      "step": 2823
+    },
+    {
+      "epoch": 2.2318575315359883,
+      "grad_norm": 0.24526065759415705,
+      "learning_rate": 3.0489657878793473e-07,
+      "loss": 0.5463,
+      "step": 2824
+    },
+    {
+      "epoch": 2.232649023002721,
+      "grad_norm": 0.23808868335388847,
+      "learning_rate": 3.0429916845702387e-07,
+      "loss": 0.5461,
+      "step": 2825
+    },
+    {
+      "epoch": 2.2334405144694536,
+      "grad_norm": 0.24672306196776342,
+      "learning_rate": 3.0370223893054094e-07,
+      "loss": 0.5595,
+      "step": 2826
+    },
+    {
+      "epoch": 2.234232005936186,
+      "grad_norm": 0.2380471364714382,
+      "learning_rate": 3.0310579062102946e-07,
+      "loss": 0.5377,
+      "step": 2827
+    },
+    {
+      "epoch": 2.235023497402919,
+      "grad_norm": 0.23080436472819818,
+      "learning_rate": 3.025098239406998e-07,
+      "loss": 0.5486,
+      "step": 2828
+    },
+    {
+      "epoch": 2.2358149888696515,
+      "grad_norm": 0.23672293323577456,
+      "learning_rate": 3.0191433930142805e-07,
+      "loss": 0.5439,
+      "step": 2829
+    },
+    {
+      "epoch": 2.236606480336384,
+      "grad_norm": 0.2441709080931778,
+      "learning_rate": 3.0131933711475965e-07,
+      "loss": 0.5559,
+      "step": 2830
+    },
+    {
+      "epoch": 2.2373979718031167,
+      "grad_norm": 0.240840113760501,
+      "learning_rate": 3.007248177919048e-07,
+      "loss": 0.5478,
+      "step": 2831
+    },
+    {
+      "epoch": 2.2381894632698494,
+      "grad_norm": 0.24233695817539602,
+      "learning_rate": 3.001307817437406e-07,
+      "loss": 0.5394,
+      "step": 2832
+    },
+    {
+      "epoch": 2.2389809547365815,
+      "grad_norm": 0.24063609595208607,
+      "learning_rate": 2.995372293808108e-07,
+      "loss": 0.5471,
+      "step": 2833
+    },
+    {
+      "epoch": 2.239772446203314,
+      "grad_norm": 0.23924643250796646,
+      "learning_rate": 2.989441611133231e-07,
+      "loss": 0.5333,
+      "step": 2834
+    },
+    {
+      "epoch": 2.240563937670047,
+      "grad_norm": 0.24813417318421102,
+      "learning_rate": 2.9835157735115257e-07,
+      "loss": 0.5533,
+      "step": 2835
+    },
+    {
+      "epoch": 2.2413554291367794,
+      "grad_norm": 0.24649175721522854,
+      "learning_rate": 2.9775947850383844e-07,
+      "loss": 0.5354,
+      "step": 2836
+    },
+    {
+      "epoch": 2.242146920603512,
+      "grad_norm": 0.23923480918210283,
+      "learning_rate": 2.9716786498058446e-07,
+      "loss": 0.5319,
+      "step": 2837
+    },
+    {
+      "epoch": 2.2429384120702447,
+      "grad_norm": 0.23767840866812015,
+      "learning_rate": 2.9657673719026054e-07,
+      "loss": 0.5542,
+      "step": 2838
+    },
+    {
+      "epoch": 2.2437299035369773,
+      "grad_norm": 0.22920949091170925,
+      "learning_rate": 2.9598609554139854e-07,
+      "loss": 0.5314,
+      "step": 2839
+    },
+    {
+      "epoch": 2.24452139500371,
+      "grad_norm": 0.24646364189934578,
+      "learning_rate": 2.9539594044219673e-07,
+      "loss": 0.5409,
+      "step": 2840
+    },
+    {
+      "epoch": 2.2453128864704426,
+      "grad_norm": 0.23608561878992967,
+      "learning_rate": 2.948062723005155e-07,
+      "loss": 0.5242,
+      "step": 2841
+    },
+    {
+      "epoch": 2.246104377937175,
+      "grad_norm": 0.23586637749198103,
+      "learning_rate": 2.9421709152387906e-07,
+      "loss": 0.5418,
+      "step": 2842
+    },
+    {
+      "epoch": 2.246895869403908,
+      "grad_norm": 0.23898935280822234,
+      "learning_rate": 2.9362839851947596e-07,
+      "loss": 0.5384,
+      "step": 2843
+    },
+    {
+      "epoch": 2.2476873608706405,
+      "grad_norm": 0.2413763631419046,
+      "learning_rate": 2.9304019369415525e-07,
+      "loss": 0.5437,
+      "step": 2844
+    },
+    {
+      "epoch": 2.248478852337373,
+      "grad_norm": 0.23564479828315718,
+      "learning_rate": 2.9245247745443105e-07,
+      "loss": 0.5321,
+      "step": 2845
+    },
+    {
+      "epoch": 2.2492703438041057,
+      "grad_norm": 0.23321194155400574,
+      "learning_rate": 2.918652502064785e-07,
+      "loss": 0.539,
+      "step": 2846
+    },
+    {
+      "epoch": 2.2500618352708384,
+      "grad_norm": 0.2254629116591103,
+      "learning_rate": 2.9127851235613453e-07,
+      "loss": 0.5282,
+      "step": 2847
+    },
+    {
+      "epoch": 2.250853326737571,
+      "grad_norm": 0.23478463008165587,
+      "learning_rate": 2.906922643088993e-07,
+      "loss": 0.5645,
+      "step": 2848
+    },
+    {
+      "epoch": 2.2516448182043036,
+      "grad_norm": 0.23452103089076926,
+      "learning_rate": 2.9010650646993295e-07,
+      "loss": 0.5372,
+      "step": 2849
+    },
+    {
+      "epoch": 2.2524363096710363,
+      "grad_norm": 0.22881393447554058,
+      "learning_rate": 2.895212392440577e-07,
+      "loss": 0.5331,
+      "step": 2850
+    },
+    {
+      "epoch": 2.253227801137769,
+      "grad_norm": 0.24671940165475068,
+      "learning_rate": 2.8893646303575625e-07,
+      "loss": 0.5399,
+      "step": 2851
+    },
+    {
+      "epoch": 2.2540192926045015,
+      "grad_norm": 0.244503937172185,
+      "learning_rate": 2.883521782491718e-07,
+      "loss": 0.5384,
+      "step": 2852
+    },
+    {
+      "epoch": 2.254810784071234,
+      "grad_norm": 0.24163619148587315,
+      "learning_rate": 2.87768385288109e-07,
+      "loss": 0.5646,
+      "step": 2853
+    },
+    {
+      "epoch": 2.2556022755379668,
+      "grad_norm": 0.2411564234197795,
+      "learning_rate": 2.871850845560314e-07,
+      "loss": 0.5508,
+      "step": 2854
+    },
+    {
+      "epoch": 2.2563937670046994,
+      "grad_norm": 0.23508813064354636,
+      "learning_rate": 2.8660227645606296e-07,
+      "loss": 0.549,
+      "step": 2855
+    },
+    {
+      "epoch": 2.257185258471432,
+      "grad_norm": 0.23477381984369078,
+      "learning_rate": 2.8601996139098695e-07,
+      "loss": 0.5426,
+      "step": 2856
+    },
+    {
+      "epoch": 2.2579767499381647,
+      "grad_norm": 0.2370525453618061,
+      "learning_rate": 2.854381397632457e-07,
+      "loss": 0.5391,
+      "step": 2857
+    },
+    {
+      "epoch": 2.2587682414048973,
+      "grad_norm": 0.2390607000164692,
+      "learning_rate": 2.848568119749415e-07,
+      "loss": 0.5426,
+      "step": 2858
+    },
+    {
+      "epoch": 2.25955973287163,
+      "grad_norm": 0.23094266808799557,
+      "learning_rate": 2.8427597842783414e-07,
+      "loss": 0.5394,
+      "step": 2859
+    },
+    {
+      "epoch": 2.2603512243383626,
+      "grad_norm": 0.24087702331429744,
+      "learning_rate": 2.836956395233425e-07,
+      "loss": 0.5395,
+      "step": 2860
+    },
+    {
+      "epoch": 2.261142715805095,
+      "grad_norm": 0.23751207190954038,
+      "learning_rate": 2.831157956625434e-07,
+      "loss": 0.5483,
+      "step": 2861
+    },
+    {
+      "epoch": 2.261934207271828,
+      "grad_norm": 0.247604081371722,
+      "learning_rate": 2.825364472461714e-07,
+      "loss": 0.5416,
+      "step": 2862
+    },
+    {
+      "epoch": 2.2627256987385604,
+      "grad_norm": 0.23789480125692128,
+      "learning_rate": 2.819575946746192e-07,
+      "loss": 0.5409,
+      "step": 2863
+    },
+    {
+      "epoch": 2.263517190205293,
+      "grad_norm": 0.2349177918274698,
+      "learning_rate": 2.813792383479364e-07,
+      "loss": 0.547,
+      "step": 2864
+    },
+    {
+      "epoch": 2.2643086816720257,
+      "grad_norm": 0.24211911719136156,
+      "learning_rate": 2.8080137866582967e-07,
+      "loss": 0.5437,
+      "step": 2865
+    },
+    {
+      "epoch": 2.2651001731387583,
+      "grad_norm": 0.2313582053135601,
+      "learning_rate": 2.802240160276623e-07,
+      "loss": 0.5324,
+      "step": 2866
+    },
+    {
+      "epoch": 2.265891664605491,
+      "grad_norm": 0.23156413476438797,
+      "learning_rate": 2.7964715083245406e-07,
+      "loss": 0.5443,
+      "step": 2867
+    },
+    {
+      "epoch": 2.2666831560722236,
+      "grad_norm": 0.23733214690280127,
+      "learning_rate": 2.790707834788818e-07,
+      "loss": 0.5418,
+      "step": 2868
+    },
+    {
+      "epoch": 2.2674746475389562,
+      "grad_norm": 0.23342416904972457,
+      "learning_rate": 2.784949143652772e-07,
+      "loss": 0.5376,
+      "step": 2869
+    },
+    {
+      "epoch": 2.268266139005689,
+      "grad_norm": 0.24012437478052984,
+      "learning_rate": 2.7791954388962804e-07,
+      "loss": 0.5341,
+      "step": 2870
+    },
+    {
+      "epoch": 2.2690576304724215,
+      "grad_norm": 0.2416102948062892,
+      "learning_rate": 2.773446724495774e-07,
+      "loss": 0.5421,
+      "step": 2871
+    },
+    {
+      "epoch": 2.269849121939154,
+      "grad_norm": 0.24175842485162605,
+      "learning_rate": 2.7677030044242345e-07,
+      "loss": 0.5449,
+      "step": 2872
+    },
+    {
+      "epoch": 2.2706406134058867,
+      "grad_norm": 0.23826793457919446,
+      "learning_rate": 2.761964282651197e-07,
+      "loss": 0.5625,
+      "step": 2873
+    },
+    {
+      "epoch": 2.2714321048726194,
+      "grad_norm": 0.24175799585516589,
+      "learning_rate": 2.7562305631427373e-07,
+      "loss": 0.5524,
+      "step": 2874
+    },
+    {
+      "epoch": 2.272223596339352,
+      "grad_norm": 0.23639796460897217,
+      "learning_rate": 2.750501849861473e-07,
+      "loss": 0.542,
+      "step": 2875
+    },
+    {
+      "epoch": 2.2730150878060846,
+      "grad_norm": 0.23684692153515702,
+      "learning_rate": 2.744778146766564e-07,
+      "loss": 0.5319,
+      "step": 2876
+    },
+    {
+      "epoch": 2.2738065792728173,
+      "grad_norm": 0.2353212441198773,
+      "learning_rate": 2.739059457813705e-07,
+      "loss": 0.5374,
+      "step": 2877
+    },
+    {
+      "epoch": 2.27459807073955,
+      "grad_norm": 0.2419176199222997,
+      "learning_rate": 2.733345786955135e-07,
+      "loss": 0.5427,
+      "step": 2878
+    },
+    {
+      "epoch": 2.2753895622062825,
+      "grad_norm": 0.24536013058010053,
+      "learning_rate": 2.727637138139612e-07,
+      "loss": 0.5374,
+      "step": 2879
+    },
+    {
+      "epoch": 2.276181053673015,
+      "grad_norm": 0.23024852786858077,
+      "learning_rate": 2.721933515312431e-07,
+      "loss": 0.5296,
+      "step": 2880
+    },
+    {
+      "epoch": 2.276972545139748,
+      "grad_norm": 0.23470183671325456,
+      "learning_rate": 2.7162349224154114e-07,
+      "loss": 0.5441,
+      "step": 2881
+    },
+    {
+      "epoch": 2.2777640366064804,
+      "grad_norm": 0.2322393667970669,
+      "learning_rate": 2.7105413633868936e-07,
+      "loss": 0.5427,
+      "step": 2882
+    },
+    {
+      "epoch": 2.278555528073213,
+      "grad_norm": 0.24380547084349297,
+      "learning_rate": 2.704852842161747e-07,
+      "loss": 0.5531,
+      "step": 2883
+    },
+    {
+      "epoch": 2.2793470195399457,
+      "grad_norm": 0.23162637501137084,
+      "learning_rate": 2.699169362671352e-07,
+      "loss": 0.5505,
+      "step": 2884
+    },
+    {
+      "epoch": 2.2801385110066783,
+      "grad_norm": 0.23554080571533584,
+      "learning_rate": 2.693490928843607e-07,
+      "loss": 0.5411,
+      "step": 2885
+    },
+    {
+      "epoch": 2.280930002473411,
+      "grad_norm": 0.24389813206815725,
+      "learning_rate": 2.6878175446029217e-07,
+      "loss": 0.5522,
+      "step": 2886
+    },
+    {
+      "epoch": 2.2817214939401436,
+      "grad_norm": 0.23710487701566618,
+      "learning_rate": 2.6821492138702164e-07,
+      "loss": 0.5404,
+      "step": 2887
+    },
+    {
+      "epoch": 2.282512985406876,
+      "grad_norm": 0.23897379699795512,
+      "learning_rate": 2.6764859405629257e-07,
+      "loss": 0.5405,
+      "step": 2888
+    },
+    {
+      "epoch": 2.283304476873609,
+      "grad_norm": 0.23482860235947972,
+      "learning_rate": 2.6708277285949786e-07,
+      "loss": 0.5499,
+      "step": 2889
+    },
+    {
+      "epoch": 2.2840959683403415,
+      "grad_norm": 0.24132942821579026,
+      "learning_rate": 2.6651745818768103e-07,
+      "loss": 0.5501,
+      "step": 2890
+    },
+    {
+      "epoch": 2.284887459807074,
+      "grad_norm": 0.2352002575063765,
+      "learning_rate": 2.6595265043153623e-07,
+      "loss": 0.5451,
+      "step": 2891
+    },
+    {
+      "epoch": 2.2856789512738067,
+      "grad_norm": 0.24083178361958704,
+      "learning_rate": 2.653883499814057e-07,
+      "loss": 0.5477,
+      "step": 2892
+    },
+    {
+      "epoch": 2.2864704427405393,
+      "grad_norm": 0.23212187081542618,
+      "learning_rate": 2.648245572272827e-07,
+      "loss": 0.5447,
+      "step": 2893
+    },
+    {
+      "epoch": 2.287261934207272,
+      "grad_norm": 0.2337076125479155,
+      "learning_rate": 2.642612725588087e-07,
+      "loss": 0.5304,
+      "step": 2894
+    },
+    {
+      "epoch": 2.2880534256740046,
+      "grad_norm": 0.2413573851502974,
+      "learning_rate": 2.6369849636527386e-07,
+      "loss": 0.5313,
+      "step": 2895
+    },
+    {
+      "epoch": 2.2888449171407372,
+      "grad_norm": 0.2413442351417876,
+      "learning_rate": 2.6313622903561836e-07,
+      "loss": 0.5392,
+      "step": 2896
+    },
+    {
+      "epoch": 2.28963640860747,
+      "grad_norm": 0.23452248538934567,
+      "learning_rate": 2.6257447095842854e-07,
+      "loss": 0.5429,
+      "step": 2897
+    },
+    {
+      "epoch": 2.2904279000742025,
+      "grad_norm": 0.23223040055805008,
+      "learning_rate": 2.620132225219406e-07,
+      "loss": 0.5476,
+      "step": 2898
+    },
+    {
+      "epoch": 2.291219391540935,
+      "grad_norm": 0.2329488894717753,
+      "learning_rate": 2.6145248411403765e-07,
+      "loss": 0.5429,
+      "step": 2899
+    },
+    {
+      "epoch": 2.2920108830076678,
+      "grad_norm": 0.23360015005402276,
+      "learning_rate": 2.608922561222502e-07,
+      "loss": 0.5373,
+      "step": 2900
+    },
+    {
+      "epoch": 2.2928023744744004,
+      "grad_norm": 0.2362549233317612,
+      "learning_rate": 2.6033253893375715e-07,
+      "loss": 0.5352,
+      "step": 2901
+    },
+    {
+      "epoch": 2.293593865941133,
+      "grad_norm": 0.24488868241053144,
+      "learning_rate": 2.597733329353824e-07,
+      "loss": 0.5322,
+      "step": 2902
+    },
+    {
+      "epoch": 2.294385357407865,
+      "grad_norm": 0.24004854502643702,
+      "learning_rate": 2.592146385135986e-07,
+      "loss": 0.5383,
+      "step": 2903
+    },
+    {
+      "epoch": 2.295176848874598,
+      "grad_norm": 0.23926218240031688,
+      "learning_rate": 2.586564560545237e-07,
+      "loss": 0.5496,
+      "step": 2904
+    },
+    {
+      "epoch": 2.2959683403413305,
+      "grad_norm": 0.2319858869114076,
+      "learning_rate": 2.5809878594392166e-07,
+      "loss": 0.5455,
+      "step": 2905
+    },
+    {
+      "epoch": 2.296759831808063,
+      "grad_norm": 0.23041785337081805,
+      "learning_rate": 2.5754162856720375e-07,
+      "loss": 0.5387,
+      "step": 2906
+    },
+    {
+      "epoch": 2.2975513232747957,
+      "grad_norm": 0.23762253769348288,
+      "learning_rate": 2.569849843094246e-07,
+      "loss": 0.5543,
+      "step": 2907
+    },
+    {
+      "epoch": 2.2983428147415284,
+      "grad_norm": 0.2337980834477202,
+      "learning_rate": 2.5642885355528663e-07,
+      "loss": 0.5443,
+      "step": 2908
+    },
+    {
+      "epoch": 2.299134306208261,
+      "grad_norm": 0.23676957670959684,
+      "learning_rate": 2.5587323668913575e-07,
+      "loss": 0.536,
+      "step": 2909
+    },
+    {
+      "epoch": 2.2999257976749936,
+      "grad_norm": 0.23441760003192555,
+      "learning_rate": 2.553181340949632e-07,
+      "loss": 0.5497,
+      "step": 2910
+    },
+    {
+      "epoch": 2.3007172891417262,
+      "grad_norm": 0.23508857955133236,
+      "learning_rate": 2.5476354615640504e-07,
+      "loss": 0.5435,
+      "step": 2911
+    },
+    {
+      "epoch": 2.301508780608459,
+      "grad_norm": 0.2422690010031859,
+      "learning_rate": 2.5420947325674145e-07,
+      "loss": 0.544,
+      "step": 2912
+    },
+    {
+      "epoch": 2.3023002720751915,
+      "grad_norm": 0.23926052851832066,
+      "learning_rate": 2.536559157788967e-07,
+      "loss": 0.5512,
+      "step": 2913
+    },
+    {
+      "epoch": 2.303091763541924,
+      "grad_norm": 0.23921418589979995,
+      "learning_rate": 2.531028741054386e-07,
+      "loss": 0.5562,
+      "step": 2914
+    },
+    {
+      "epoch": 2.3038832550086568,
+      "grad_norm": 0.23848422835146474,
+      "learning_rate": 2.525503486185785e-07,
+      "loss": 0.5439,
+      "step": 2915
+    },
+    {
+      "epoch": 2.3046747464753894,
+      "grad_norm": 0.23711473131955782,
+      "learning_rate": 2.519983397001717e-07,
+      "loss": 0.5576,
+      "step": 2916
+    },
+    {
+      "epoch": 2.305466237942122,
+      "grad_norm": 0.23309687717908328,
+      "learning_rate": 2.5144684773171623e-07,
+      "loss": 0.5317,
+      "step": 2917
+    },
+    {
+      "epoch": 2.3062577294088547,
+      "grad_norm": 0.23776376999681284,
+      "learning_rate": 2.5089587309435144e-07,
+      "loss": 0.5415,
+      "step": 2918
+    },
+    {
+      "epoch": 2.3070492208755873,
+      "grad_norm": 0.2379389945672291,
+      "learning_rate": 2.503454161688614e-07,
+      "loss": 0.5323,
+      "step": 2919
+    },
+    {
+      "epoch": 2.30784071234232,
+      "grad_norm": 0.23274745607177677,
+      "learning_rate": 2.497954773356705e-07,
+      "loss": 0.5417,
+      "step": 2920
+    },
+    {
+      "epoch": 2.3086322038090525,
+      "grad_norm": 0.23283432148544148,
+      "learning_rate": 2.492460569748468e-07,
+      "loss": 0.5382,
+      "step": 2921
+    },
+    {
+      "epoch": 2.309423695275785,
+      "grad_norm": 0.23672261901751507,
+      "learning_rate": 2.4869715546609894e-07,
+      "loss": 0.5272,
+      "step": 2922
+    },
+    {
+      "epoch": 2.310215186742518,
+      "grad_norm": 0.24404734310473322,
+      "learning_rate": 2.481487731887761e-07,
+      "loss": 0.5353,
+      "step": 2923
+    },
+    {
+      "epoch": 2.3110066782092504,
+      "grad_norm": 0.23489632952562617,
+      "learning_rate": 2.476009105218709e-07,
+      "loss": 0.5453,
+      "step": 2924
+    },
+    {
+      "epoch": 2.311798169675983,
+      "grad_norm": 0.23290588005746646,
+      "learning_rate": 2.4705356784401486e-07,
+      "loss": 0.5379,
+      "step": 2925
+    },
+    {
+      "epoch": 2.3125896611427157,
+      "grad_norm": 0.23709934150914688,
+      "learning_rate": 2.465067455334815e-07,
+      "loss": 0.5374,
+      "step": 2926
+    },
+    {
+      "epoch": 2.3133811526094483,
+      "grad_norm": 0.24157094934977066,
+      "learning_rate": 2.4596044396818415e-07,
+      "loss": 0.5346,
+      "step": 2927
+    },
+    {
+      "epoch": 2.314172644076181,
+      "grad_norm": 0.23607873462862783,
+      "learning_rate": 2.454146635256752e-07,
+      "loss": 0.527,
+      "step": 2928
+    },
+    {
+      "epoch": 2.3149641355429136,
+      "grad_norm": 0.2327484448089389,
+      "learning_rate": 2.448694045831489e-07,
+      "loss": 0.5506,
+      "step": 2929
+    },
+    {
+      "epoch": 2.315755627009646,
+      "grad_norm": 0.2292591355480856,
+      "learning_rate": 2.443246675174374e-07,
+      "loss": 0.5402,
+      "step": 2930
+    },
+    {
+      "epoch": 2.316547118476379,
+      "grad_norm": 0.23424073187304081,
+      "learning_rate": 2.4378045270501356e-07,
+      "loss": 0.5206,
+      "step": 2931
+    },
+    {
+      "epoch": 2.3173386099431115,
+      "grad_norm": 0.23523815302769308,
+      "learning_rate": 2.432367605219883e-07,
+      "loss": 0.5415,
+      "step": 2932
+    },
+    {
+      "epoch": 2.318130101409844,
+      "grad_norm": 0.24322651728949735,
+      "learning_rate": 2.426935913441115e-07,
+      "loss": 0.5556,
+      "step": 2933
+    },
+    {
+      "epoch": 2.3189215928765767,
+      "grad_norm": 0.2308249704959735,
+      "learning_rate": 2.4215094554677186e-07,
+      "loss": 0.5383,
+      "step": 2934
+    },
+    {
+      "epoch": 2.3197130843433094,
+      "grad_norm": 0.23709984308039514,
+      "learning_rate": 2.4160882350499625e-07,
+      "loss": 0.5517,
+      "step": 2935
+    },
+    {
+      "epoch": 2.320504575810042,
+      "grad_norm": 0.23740547802018616,
+      "learning_rate": 2.4106722559344914e-07,
+      "loss": 0.54,
+      "step": 2936
+    },
+    {
+      "epoch": 2.3212960672767746,
+      "grad_norm": 0.25173178577131733,
+      "learning_rate": 2.4052615218643403e-07,
+      "loss": 0.5498,
+      "step": 2937
+    },
+    {
+      "epoch": 2.3220875587435073,
+      "grad_norm": 0.24070201028750376,
+      "learning_rate": 2.3998560365789055e-07,
+      "loss": 0.5484,
+      "step": 2938
+    },
+    {
+      "epoch": 2.32287905021024,
+      "grad_norm": 0.23765755990070234,
+      "learning_rate": 2.394455803813963e-07,
+      "loss": 0.5491,
+      "step": 2939
+    },
+    {
+      "epoch": 2.3236705416769725,
+      "grad_norm": 0.2389945952429259,
+      "learning_rate": 2.389060827301657e-07,
+      "loss": 0.539,
+      "step": 2940
+    },
+    {
+      "epoch": 2.324462033143705,
+      "grad_norm": 0.23250324232305528,
+      "learning_rate": 2.3836711107704944e-07,
+      "loss": 0.5471,
+      "step": 2941
+    },
+    {
+      "epoch": 2.3252535246104378,
+      "grad_norm": 0.23999911548685604,
+      "learning_rate": 2.3782866579453586e-07,
+      "loss": 0.5543,
+      "step": 2942
+    },
+    {
+      "epoch": 2.3260450160771704,
+      "grad_norm": 0.23675045453576574,
+      "learning_rate": 2.372907472547485e-07,
+      "loss": 0.5576,
+      "step": 2943
+    },
+    {
+      "epoch": 2.326836507543903,
+      "grad_norm": 0.24175296270765648,
+      "learning_rate": 2.367533558294472e-07,
+      "loss": 0.5439,
+      "step": 2944
+    },
+    {
+      "epoch": 2.3276279990106357,
+      "grad_norm": 0.23572156866038801,
+      "learning_rate": 2.3621649189002745e-07,
+      "loss": 0.5458,
+      "step": 2945
+    },
+    {
+      "epoch": 2.3284194904773683,
+      "grad_norm": 0.2400512235425696,
+      "learning_rate": 2.356801558075201e-07,
+      "loss": 0.5466,
+      "step": 2946
+    },
+    {
+      "epoch": 2.329210981944101,
+      "grad_norm": 0.23140180592120257,
+      "learning_rate": 2.3514434795259164e-07,
+      "loss": 0.5476,
+      "step": 2947
+    },
+    {
+      "epoch": 2.3300024734108336,
+      "grad_norm": 0.23740514054863193,
+      "learning_rate": 2.3460906869554308e-07,
+      "loss": 0.5385,
+      "step": 2948
+    },
+    {
+      "epoch": 2.330793964877566,
+      "grad_norm": 0.2353411242490673,
+      "learning_rate": 2.3407431840631008e-07,
+      "loss": 0.5297,
+      "step": 2949
+    },
+    {
+      "epoch": 2.331585456344299,
+      "grad_norm": 0.2361315966652804,
+      "learning_rate": 2.3354009745446303e-07,
+      "loss": 0.5393,
+      "step": 2950
+    },
+    {
+      "epoch": 2.3323769478110314,
+      "grad_norm": 0.23801386911590616,
+      "learning_rate": 2.3300640620920587e-07,
+      "loss": 0.538,
+      "step": 2951
+    },
+    {
+      "epoch": 2.333168439277764,
+      "grad_norm": 0.2378470424093242,
+      "learning_rate": 2.3247324503937748e-07,
+      "loss": 0.5377,
+      "step": 2952
+    },
+    {
+      "epoch": 2.3339599307444967,
+      "grad_norm": 0.23316348768934472,
+      "learning_rate": 2.3194061431344968e-07,
+      "loss": 0.5225,
+      "step": 2953
+    },
+    {
+      "epoch": 2.3347514222112293,
+      "grad_norm": 0.24397046916608647,
+      "learning_rate": 2.314085143995277e-07,
+      "loss": 0.5639,
+      "step": 2954
+    },
+    {
+      "epoch": 2.335542913677962,
+      "grad_norm": 0.23741322935364065,
+      "learning_rate": 2.308769456653501e-07,
+      "loss": 0.5544,
+      "step": 2955
+    },
+    {
+      "epoch": 2.3363344051446946,
+      "grad_norm": 0.23296794692410755,
+      "learning_rate": 2.3034590847828806e-07,
+      "loss": 0.5369,
+      "step": 2956
+    },
+    {
+      "epoch": 2.3371258966114272,
+      "grad_norm": 0.24641419323860192,
+      "learning_rate": 2.298154032053461e-07,
+      "loss": 0.5392,
+      "step": 2957
+    },
+    {
+      "epoch": 2.33791738807816,
+      "grad_norm": 0.2341220817496533,
+      "learning_rate": 2.292854302131606e-07,
+      "loss": 0.5255,
+      "step": 2958
+    },
+    {
+      "epoch": 2.3387088795448925,
+      "grad_norm": 0.2417259848899059,
+      "learning_rate": 2.2875598986799992e-07,
+      "loss": 0.5533,
+      "step": 2959
+    },
+    {
+      "epoch": 2.339500371011625,
+      "grad_norm": 0.2384720082016968,
+      "learning_rate": 2.2822708253576462e-07,
+      "loss": 0.5403,
+      "step": 2960
+    },
+    {
+      "epoch": 2.3402918624783577,
+      "grad_norm": 0.24021821063661916,
+      "learning_rate": 2.2769870858198647e-07,
+      "loss": 0.5351,
+      "step": 2961
+    },
+    {
+      "epoch": 2.3410833539450904,
+      "grad_norm": 0.23341851615304104,
+      "learning_rate": 2.2717086837182964e-07,
+      "loss": 0.5465,
+      "step": 2962
+    },
+    {
+      "epoch": 2.341874845411823,
+      "grad_norm": 0.2282609483014954,
+      "learning_rate": 2.266435622700884e-07,
+      "loss": 0.5386,
+      "step": 2963
+    },
+    {
+      "epoch": 2.3426663368785556,
+      "grad_norm": 0.2381507009889597,
+      "learning_rate": 2.2611679064118816e-07,
+      "loss": 0.5413,
+      "step": 2964
+    },
+    {
+      "epoch": 2.3434578283452883,
+      "grad_norm": 0.23369932358400436,
+      "learning_rate": 2.2559055384918512e-07,
+      "loss": 0.5423,
+      "step": 2965
+    },
+    {
+      "epoch": 2.344249319812021,
+      "grad_norm": 0.2372626435269223,
+      "learning_rate": 2.250648522577654e-07,
+      "loss": 0.5346,
+      "step": 2966
+    },
+    {
+      "epoch": 2.3450408112787535,
+      "grad_norm": 0.23575100203991795,
+      "learning_rate": 2.245396862302461e-07,
+      "loss": 0.553,
+      "step": 2967
+    },
+    {
+      "epoch": 2.345832302745486,
+      "grad_norm": 0.24193031770458479,
+      "learning_rate": 2.240150561295735e-07,
+      "loss": 0.541,
+      "step": 2968
+    },
+    {
+      "epoch": 2.346623794212219,
+      "grad_norm": 0.23871335460532514,
+      "learning_rate": 2.2349096231832366e-07,
+      "loss": 0.5381,
+      "step": 2969
+    },
+    {
+      "epoch": 2.3474152856789514,
+      "grad_norm": 0.23360995485358593,
+      "learning_rate": 2.2296740515870205e-07,
+      "loss": 0.5405,
+      "step": 2970
+    },
+    {
+      "epoch": 2.348206777145684,
+      "grad_norm": 0.24073556478269748,
+      "learning_rate": 2.224443850125428e-07,
+      "loss": 0.5475,
+      "step": 2971
+    },
+    {
+      "epoch": 2.3489982686124167,
+      "grad_norm": 0.2316980279828743,
+      "learning_rate": 2.2192190224131003e-07,
+      "loss": 0.5437,
+      "step": 2972
+    },
+    {
+      "epoch": 2.3497897600791493,
+      "grad_norm": 0.23504108629980844,
+      "learning_rate": 2.213999572060955e-07,
+      "loss": 0.5437,
+      "step": 2973
+    },
+    {
+      "epoch": 2.350581251545882,
+      "grad_norm": 0.23090311838451977,
+      "learning_rate": 2.208785502676195e-07,
+      "loss": 0.5346,
+      "step": 2974
+    },
+    {
+      "epoch": 2.3513727430126146,
+      "grad_norm": 0.23252338648942963,
+      "learning_rate": 2.203576817862306e-07,
+      "loss": 0.5399,
+      "step": 2975
+    },
+    {
+      "epoch": 2.352164234479347,
+      "grad_norm": 0.23479332311143924,
+      "learning_rate": 2.198373521219049e-07,
+      "loss": 0.5358,
+      "step": 2976
+    },
+    {
+      "epoch": 2.35295572594608,
+      "grad_norm": 0.23545280272525002,
+      "learning_rate": 2.1931756163424708e-07,
+      "loss": 0.5535,
+      "step": 2977
+    },
+    {
+      "epoch": 2.3537472174128125,
+      "grad_norm": 0.23562554623559326,
+      "learning_rate": 2.187983106824881e-07,
+      "loss": 0.5399,
+      "step": 2978
+    },
+    {
+      "epoch": 2.354538708879545,
+      "grad_norm": 0.2385058108290646,
+      "learning_rate": 2.1827959962548626e-07,
+      "loss": 0.5387,
+      "step": 2979
+    },
+    {
+      "epoch": 2.3553302003462777,
+      "grad_norm": 0.2444467679434618,
+      "learning_rate": 2.1776142882172766e-07,
+      "loss": 0.544,
+      "step": 2980
+    },
+    {
+      "epoch": 2.3561216918130103,
+      "grad_norm": 0.23492791468464386,
+      "learning_rate": 2.172437986293233e-07,
+      "loss": 0.5404,
+      "step": 2981
+    },
+    {
+      "epoch": 2.356913183279743,
+      "grad_norm": 0.23734604714372437,
+      "learning_rate": 2.1672670940601244e-07,
+      "loss": 0.5659,
+      "step": 2982
+    },
+    {
+      "epoch": 2.3577046747464756,
+      "grad_norm": 0.23559863206262585,
+      "learning_rate": 2.1621016150915916e-07,
+      "loss": 0.5253,
+      "step": 2983
+    },
+    {
+      "epoch": 2.3584961662132082,
+      "grad_norm": 0.23137078846940448,
+      "learning_rate": 2.1569415529575363e-07,
+      "loss": 0.5188,
+      "step": 2984
+    },
+    {
+      "epoch": 2.359287657679941,
+      "grad_norm": 0.23127965442800444,
+      "learning_rate": 2.1517869112241282e-07,
+      "loss": 0.5367,
+      "step": 2985
+    },
+    {
+      "epoch": 2.3600791491466735,
+      "grad_norm": 0.24431856585504763,
+      "learning_rate": 2.14663769345377e-07,
+      "loss": 0.5358,
+      "step": 2986
+    },
+    {
+      "epoch": 2.360870640613406,
+      "grad_norm": 0.23671112881562123,
+      "learning_rate": 2.1414939032051338e-07,
+      "loss": 0.5415,
+      "step": 2987
+    },
+    {
+      "epoch": 2.3616621320801388,
+      "grad_norm": 0.23467375875259117,
+      "learning_rate": 2.1363555440331349e-07,
+      "loss": 0.5583,
+      "step": 2988
+    },
+    {
+      "epoch": 2.3624536235468714,
+      "grad_norm": 0.23551544465173518,
+      "learning_rate": 2.1312226194889294e-07,
+      "loss": 0.5297,
+      "step": 2989
+    },
+    {
+      "epoch": 2.3632451150136036,
+      "grad_norm": 0.24252246719560705,
+      "learning_rate": 2.1260951331199318e-07,
+      "loss": 0.5438,
+      "step": 2990
+    },
+    {
+      "epoch": 2.364036606480336,
+      "grad_norm": 0.23527427658269348,
+      "learning_rate": 2.1209730884697773e-07,
+      "loss": 0.5585,
+      "step": 2991
+    },
+    {
+      "epoch": 2.364828097947069,
+      "grad_norm": 0.2369771155174878,
+      "learning_rate": 2.1158564890783614e-07,
+      "loss": 0.5461,
+      "step": 2992
+    },
+    {
+      "epoch": 2.3656195894138015,
+      "grad_norm": 0.22913161426046072,
+      "learning_rate": 2.1107453384818041e-07,
+      "loss": 0.5471,
+      "step": 2993
+    },
+    {
+      "epoch": 2.366411080880534,
+      "grad_norm": 0.241289509398457,
+      "learning_rate": 2.1056396402124598e-07,
+      "loss": 0.5555,
+      "step": 2994
+    },
+    {
+      "epoch": 2.3672025723472667,
+      "grad_norm": 0.23499416053446745,
+      "learning_rate": 2.100539397798925e-07,
+      "loss": 0.5315,
+      "step": 2995
+    },
+    {
+      "epoch": 2.3679940638139994,
+      "grad_norm": 0.23796863417235548,
+      "learning_rate": 2.0954446147660076e-07,
+      "loss": 0.5353,
+      "step": 2996
+    },
+    {
+      "epoch": 2.368785555280732,
+      "grad_norm": 0.23104230521797295,
+      "learning_rate": 2.0903552946347613e-07,
+      "loss": 0.5213,
+      "step": 2997
+    },
+    {
+      "epoch": 2.3695770467474646,
+      "grad_norm": 0.23531598896105188,
+      "learning_rate": 2.0852714409224537e-07,
+      "loss": 0.5365,
+      "step": 2998
+    },
+    {
+      "epoch": 2.3703685382141972,
+      "grad_norm": 0.23130820172264557,
+      "learning_rate": 2.080193057142574e-07,
+      "loss": 0.5447,
+      "step": 2999
+    },
+    {
+      "epoch": 2.37116002968093,
+      "grad_norm": 0.24572529630980294,
+      "learning_rate": 2.0751201468048386e-07,
+      "loss": 0.554,
+      "step": 3000
+    },
+    {
+      "epoch": 2.3719515211476625,
+      "grad_norm": 0.23708559641086727,
+      "learning_rate": 2.0700527134151757e-07,
+      "loss": 0.5413,
+      "step": 3001
+    },
+    {
+      "epoch": 2.372743012614395,
+      "grad_norm": 0.23449673402384064,
+      "learning_rate": 2.0649907604757267e-07,
+      "loss": 0.542,
+      "step": 3002
+    },
+    {
+      "epoch": 2.3735345040811278,
+      "grad_norm": 0.23661757635590114,
+      "learning_rate": 2.0599342914848507e-07,
+      "loss": 0.5368,
+      "step": 3003
+    },
+    {
+      "epoch": 2.3743259955478604,
+      "grad_norm": 0.23694048392448555,
+      "learning_rate": 2.054883309937109e-07,
+      "loss": 0.5395,
+      "step": 3004
+    },
+    {
+      "epoch": 2.375117487014593,
+      "grad_norm": 0.24350189562956667,
+      "learning_rate": 2.0498378193232824e-07,
+      "loss": 0.5491,
+      "step": 3005
+    },
+    {
+      "epoch": 2.3759089784813257,
+      "grad_norm": 0.24335964799360535,
+      "learning_rate": 2.0447978231303465e-07,
+      "loss": 0.5536,
+      "step": 3006
+    },
+    {
+      "epoch": 2.3767004699480583,
+      "grad_norm": 0.23622368180971987,
+      "learning_rate": 2.0397633248414824e-07,
+      "loss": 0.5394,
+      "step": 3007
+    },
+    {
+      "epoch": 2.377491961414791,
+      "grad_norm": 0.23085108037031152,
+      "learning_rate": 2.034734327936074e-07,
+      "loss": 0.5408,
+      "step": 3008
+    },
+    {
+      "epoch": 2.3782834528815235,
+      "grad_norm": 0.2338776831000615,
+      "learning_rate": 2.0297108358896976e-07,
+      "loss": 0.5464,
+      "step": 3009
+    },
+    {
+      "epoch": 2.379074944348256,
+      "grad_norm": 0.23398572930753478,
+      "learning_rate": 2.0246928521741336e-07,
+      "loss": 0.5311,
+      "step": 3010
+    },
+    {
+      "epoch": 2.379866435814989,
+      "grad_norm": 0.23550836865287855,
+      "learning_rate": 2.0196803802573504e-07,
+      "loss": 0.5302,
+      "step": 3011
+    },
+    {
+      "epoch": 2.3806579272817214,
+      "grad_norm": 0.2401035646778562,
+      "learning_rate": 2.0146734236035056e-07,
+      "loss": 0.5411,
+      "step": 3012
+    },
+    {
+      "epoch": 2.381449418748454,
+      "grad_norm": 0.2428198648386074,
+      "learning_rate": 2.009671985672947e-07,
+      "loss": 0.5578,
+      "step": 3013
+    },
+    {
+      "epoch": 2.3822409102151867,
+      "grad_norm": 0.23172459856296815,
+      "learning_rate": 2.0046760699222077e-07,
+      "loss": 0.5395,
+      "step": 3014
+    },
+    {
+      "epoch": 2.3830324016819193,
+      "grad_norm": 0.23267258196412796,
+      "learning_rate": 1.9996856798040096e-07,
+      "loss": 0.5285,
+      "step": 3015
+    },
+    {
+      "epoch": 2.383823893148652,
+      "grad_norm": 0.23379681011650655,
+      "learning_rate": 1.9947008187672476e-07,
+      "loss": 0.5385,
+      "step": 3016
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.23882978808989294,
+      "learning_rate": 1.9897214902570004e-07,
+      "loss": 0.5454,
+      "step": 3017
+    },
+    {
+      "epoch": 2.385406876082117,
+      "grad_norm": 0.23279583356351305,
+      "learning_rate": 1.9847476977145206e-07,
+      "loss": 0.5553,
+      "step": 3018
+    },
+    {
+      "epoch": 2.38619836754885,
+      "grad_norm": 0.2358515290040839,
+      "learning_rate": 1.9797794445772354e-07,
+      "loss": 0.5465,
+      "step": 3019
+    },
+    {
+      "epoch": 2.3869898590155825,
+      "grad_norm": 0.23214625087536142,
+      "learning_rate": 1.974816734278748e-07,
+      "loss": 0.5341,
+      "step": 3020
+    },
+    {
+      "epoch": 2.387781350482315,
+      "grad_norm": 0.23957482651366258,
+      "learning_rate": 1.9698595702488264e-07,
+      "loss": 0.5382,
+      "step": 3021
+    },
+    {
+      "epoch": 2.3885728419490477,
+      "grad_norm": 0.2375334170362343,
+      "learning_rate": 1.9649079559134052e-07,
+      "loss": 0.5431,
+      "step": 3022
+    },
+    {
+      "epoch": 2.3893643334157804,
+      "grad_norm": 0.23426344141299393,
+      "learning_rate": 1.959961894694585e-07,
+      "loss": 0.5465,
+      "step": 3023
+    },
+    {
+      "epoch": 2.390155824882513,
+      "grad_norm": 0.2427116950044707,
+      "learning_rate": 1.955021390010625e-07,
+      "loss": 0.5422,
+      "step": 3024
+    },
+    {
+      "epoch": 2.3909473163492456,
+      "grad_norm": 0.23268379552551366,
+      "learning_rate": 1.9500864452759546e-07,
+      "loss": 0.5367,
+      "step": 3025
+    },
+    {
+      "epoch": 2.3917388078159783,
+      "grad_norm": 0.2396047627331565,
+      "learning_rate": 1.9451570639011506e-07,
+      "loss": 0.5332,
+      "step": 3026
+    },
+    {
+      "epoch": 2.392530299282711,
+      "grad_norm": 0.23972633958727738,
+      "learning_rate": 1.9402332492929462e-07,
+      "loss": 0.5378,
+      "step": 3027
+    },
+    {
+      "epoch": 2.3933217907494435,
+      "grad_norm": 0.24085801974620788,
+      "learning_rate": 1.9353150048542323e-07,
+      "loss": 0.5467,
+      "step": 3028
+    },
+    {
+      "epoch": 2.394113282216176,
+      "grad_norm": 0.23765836270601126,
+      "learning_rate": 1.9304023339840413e-07,
+      "loss": 0.5373,
+      "step": 3029
+    },
+    {
+      "epoch": 2.3949047736829088,
+      "grad_norm": 0.23759056797531627,
+      "learning_rate": 1.925495240077567e-07,
+      "loss": 0.5429,
+      "step": 3030
+    },
+    {
+      "epoch": 2.3956962651496414,
+      "grad_norm": 0.24133582058090594,
+      "learning_rate": 1.9205937265261386e-07,
+      "loss": 0.5277,
+      "step": 3031
+    },
+    {
+      "epoch": 2.396487756616374,
+      "grad_norm": 0.23810254393805264,
+      "learning_rate": 1.9156977967172305e-07,
+      "loss": 0.5452,
+      "step": 3032
+    },
+    {
+      "epoch": 2.3972792480831067,
+      "grad_norm": 0.23935644607567932,
+      "learning_rate": 1.9108074540344589e-07,
+      "loss": 0.5533,
+      "step": 3033
+    },
+    {
+      "epoch": 2.3980707395498393,
+      "grad_norm": 0.23402378461251894,
+      "learning_rate": 1.9059227018575773e-07,
+      "loss": 0.536,
+      "step": 3034
+    },
+    {
+      "epoch": 2.398862231016572,
+      "grad_norm": 0.234941047429903,
+      "learning_rate": 1.9010435435624817e-07,
+      "loss": 0.5312,
+      "step": 3035
+    },
+    {
+      "epoch": 2.3996537224833046,
+      "grad_norm": 0.23274382332399768,
+      "learning_rate": 1.896169982521194e-07,
+      "loss": 0.543,
+      "step": 3036
+    },
+    {
+      "epoch": 2.400445213950037,
+      "grad_norm": 0.2346041612978699,
+      "learning_rate": 1.8913020221018738e-07,
+      "loss": 0.5319,
+      "step": 3037
+    },
+    {
+      "epoch": 2.40123670541677,
+      "grad_norm": 0.23119817591818537,
+      "learning_rate": 1.8864396656688065e-07,
+      "loss": 0.5347,
+      "step": 3038
+    },
+    {
+      "epoch": 2.4020281968835024,
+      "grad_norm": 0.23559666560723652,
+      "learning_rate": 1.8815829165824027e-07,
+      "loss": 0.557,
+      "step": 3039
+    },
+    {
+      "epoch": 2.402819688350235,
+      "grad_norm": 0.23902309595885962,
+      "learning_rate": 1.8767317781992053e-07,
+      "loss": 0.5455,
+      "step": 3040
+    },
+    {
+      "epoch": 2.4036111798169677,
+      "grad_norm": 0.23874849537561182,
+      "learning_rate": 1.8718862538718727e-07,
+      "loss": 0.561,
+      "step": 3041
+    },
+    {
+      "epoch": 2.4044026712837003,
+      "grad_norm": 0.23487207124785586,
+      "learning_rate": 1.8670463469491848e-07,
+      "loss": 0.5493,
+      "step": 3042
+    },
+    {
+      "epoch": 2.405194162750433,
+      "grad_norm": 0.2348907171177612,
+      "learning_rate": 1.8622120607760462e-07,
+      "loss": 0.5445,
+      "step": 3043
+    },
+    {
+      "epoch": 2.4059856542171656,
+      "grad_norm": 0.23511639513535781,
+      "learning_rate": 1.857383398693463e-07,
+      "loss": 0.5361,
+      "step": 3044
+    },
+    {
+      "epoch": 2.4067771456838982,
+      "grad_norm": 0.23172284220984307,
+      "learning_rate": 1.8525603640385623e-07,
+      "loss": 0.5398,
+      "step": 3045
+    },
+    {
+      "epoch": 2.407568637150631,
+      "grad_norm": 0.24145893268253132,
+      "learning_rate": 1.8477429601445883e-07,
+      "loss": 0.555,
+      "step": 3046
+    },
+    {
+      "epoch": 2.4083601286173635,
+      "grad_norm": 0.23923201446021647,
+      "learning_rate": 1.8429311903408818e-07,
+      "loss": 0.5425,
+      "step": 3047
+    },
+    {
+      "epoch": 2.409151620084096,
+      "grad_norm": 0.23142822151905634,
+      "learning_rate": 1.8381250579529028e-07,
+      "loss": 0.5246,
+      "step": 3048
+    },
+    {
+      "epoch": 2.4099431115508287,
+      "grad_norm": 0.23318458482271598,
+      "learning_rate": 1.8333245663022012e-07,
+      "loss": 0.5334,
+      "step": 3049
+    },
+    {
+      "epoch": 2.4107346030175614,
+      "grad_norm": 0.234765056105339,
+      "learning_rate": 1.828529718706434e-07,
+      "loss": 0.5484,
+      "step": 3050
+    },
+    {
+      "epoch": 2.411526094484294,
+      "grad_norm": 0.24291943608056135,
+      "learning_rate": 1.8237405184793653e-07,
+      "loss": 0.5258,
+      "step": 3051
+    },
+    {
+      "epoch": 2.4123175859510266,
+      "grad_norm": 0.23763549523903982,
+      "learning_rate": 1.8189569689308447e-07,
+      "loss": 0.5503,
+      "step": 3052
+    },
+    {
+      "epoch": 2.4131090774177593,
+      "grad_norm": 0.2345372918505549,
+      "learning_rate": 1.8141790733668284e-07,
+      "loss": 0.5476,
+      "step": 3053
+    },
+    {
+      "epoch": 2.413900568884492,
+      "grad_norm": 0.2328438476755046,
+      "learning_rate": 1.8094068350893498e-07,
+      "loss": 0.5449,
+      "step": 3054
+    },
+    {
+      "epoch": 2.4146920603512245,
+      "grad_norm": 0.2397623560412376,
+      "learning_rate": 1.8046402573965435e-07,
+      "loss": 0.5486,
+      "step": 3055
+    },
+    {
+      "epoch": 2.415483551817957,
+      "grad_norm": 0.2350165835587842,
+      "learning_rate": 1.7998793435826331e-07,
+      "loss": 0.5375,
+      "step": 3056
+    },
+    {
+      "epoch": 2.4162750432846893,
+      "grad_norm": 0.2271690396423099,
+      "learning_rate": 1.79512409693792e-07,
+      "loss": 0.5225,
+      "step": 3057
+    },
+    {
+      "epoch": 2.417066534751422,
+      "grad_norm": 0.2363044776651791,
+      "learning_rate": 1.7903745207488008e-07,
+      "loss": 0.5519,
+      "step": 3058
+    },
+    {
+      "epoch": 2.4178580262181546,
+      "grad_norm": 0.2389641908921406,
+      "learning_rate": 1.7856306182977377e-07,
+      "loss": 0.5484,
+      "step": 3059
+    },
+    {
+      "epoch": 2.4186495176848872,
+      "grad_norm": 0.23733756879821363,
+      "learning_rate": 1.7808923928632802e-07,
+      "loss": 0.5407,
+      "step": 3060
+    },
+    {
+      "epoch": 2.41944100915162,
+      "grad_norm": 0.2367910945018087,
+      "learning_rate": 1.776159847720059e-07,
+      "loss": 0.5529,
+      "step": 3061
+    },
+    {
+      "epoch": 2.4202325006183525,
+      "grad_norm": 0.2404032568053986,
+      "learning_rate": 1.771432986138771e-07,
+      "loss": 0.5474,
+      "step": 3062
+    },
+    {
+      "epoch": 2.421023992085085,
+      "grad_norm": 0.2361832628890172,
+      "learning_rate": 1.7667118113861856e-07,
+      "loss": 0.5515,
+      "step": 3063
+    },
+    {
+      "epoch": 2.4218154835518177,
+      "grad_norm": 0.23336322102034698,
+      "learning_rate": 1.7619963267251535e-07,
+      "loss": 0.5362,
+      "step": 3064
+    },
+    {
+      "epoch": 2.4226069750185504,
+      "grad_norm": 0.23227217709511253,
+      "learning_rate": 1.7572865354145727e-07,
+      "loss": 0.5285,
+      "step": 3065
+    },
+    {
+      "epoch": 2.423398466485283,
+      "grad_norm": 0.23915673343967842,
+      "learning_rate": 1.7525824407094248e-07,
+      "loss": 0.5461,
+      "step": 3066
+    },
+    {
+      "epoch": 2.4241899579520156,
+      "grad_norm": 0.22756092226139843,
+      "learning_rate": 1.747884045860746e-07,
+      "loss": 0.5242,
+      "step": 3067
+    },
+    {
+      "epoch": 2.4249814494187483,
+      "grad_norm": 0.23655072304336305,
+      "learning_rate": 1.7431913541156317e-07,
+      "loss": 0.5375,
+      "step": 3068
+    },
+    {
+      "epoch": 2.425772940885481,
+      "grad_norm": 0.2408872647317191,
+      "learning_rate": 1.7385043687172462e-07,
+      "loss": 0.5299,
+      "step": 3069
+    },
+    {
+      "epoch": 2.4265644323522135,
+      "grad_norm": 0.2350455613550991,
+      "learning_rate": 1.7338230929047935e-07,
+      "loss": 0.5458,
+      "step": 3070
+    },
+    {
+      "epoch": 2.427355923818946,
+      "grad_norm": 0.23106468062471258,
+      "learning_rate": 1.7291475299135483e-07,
+      "loss": 0.5506,
+      "step": 3071
+    },
+    {
+      "epoch": 2.428147415285679,
+      "grad_norm": 0.2347809663190683,
+      "learning_rate": 1.7244776829748265e-07,
+      "loss": 0.5441,
+      "step": 3072
+    },
+    {
+      "epoch": 2.4289389067524114,
+      "grad_norm": 0.2359511824980337,
+      "learning_rate": 1.7198135553159941e-07,
+      "loss": 0.535,
+      "step": 3073
+    },
+    {
+      "epoch": 2.429730398219144,
+      "grad_norm": 0.23348670547692796,
+      "learning_rate": 1.7151551501604754e-07,
+      "loss": 0.5357,
+      "step": 3074
+    },
+    {
+      "epoch": 2.4305218896858767,
+      "grad_norm": 0.23205260864080945,
+      "learning_rate": 1.7105024707277216e-07,
+      "loss": 0.532,
+      "step": 3075
+    },
+    {
+      "epoch": 2.4313133811526093,
+      "grad_norm": 0.23874489046367486,
+      "learning_rate": 1.7058555202332435e-07,
+      "loss": 0.5517,
+      "step": 3076
+    },
+    {
+      "epoch": 2.432104872619342,
+      "grad_norm": 0.23587560295683307,
+      "learning_rate": 1.7012143018885837e-07,
+      "loss": 0.5446,
+      "step": 3077
+    },
+    {
+      "epoch": 2.4328963640860746,
+      "grad_norm": 0.23924043326356567,
+      "learning_rate": 1.696578818901324e-07,
+      "loss": 0.5496,
+      "step": 3078
+    },
+    {
+      "epoch": 2.433687855552807,
+      "grad_norm": 0.22826423171367932,
+      "learning_rate": 1.6919490744750898e-07,
+      "loss": 0.5417,
+      "step": 3079
+    },
+    {
+      "epoch": 2.43447934701954,
+      "grad_norm": 0.23315932588856364,
+      "learning_rate": 1.6873250718095257e-07,
+      "loss": 0.5374,
+      "step": 3080
+    },
+    {
+      "epoch": 2.4352708384862725,
+      "grad_norm": 0.24658934266044152,
+      "learning_rate": 1.6827068141003242e-07,
+      "loss": 0.5349,
+      "step": 3081
+    },
+    {
+      "epoch": 2.436062329953005,
+      "grad_norm": 0.24013862399555566,
+      "learning_rate": 1.678094304539198e-07,
+      "loss": 0.5452,
+      "step": 3082
+    },
+    {
+      "epoch": 2.4368538214197377,
+      "grad_norm": 0.23589336828081162,
+      "learning_rate": 1.6734875463138864e-07,
+      "loss": 0.5296,
+      "step": 3083
+    },
+    {
+      "epoch": 2.4376453128864704,
+      "grad_norm": 0.23847558779221928,
+      "learning_rate": 1.668886542608162e-07,
+      "loss": 0.5397,
+      "step": 3084
+    },
+    {
+      "epoch": 2.438436804353203,
+      "grad_norm": 0.2385144819390216,
+      "learning_rate": 1.664291296601814e-07,
+      "loss": 0.5465,
+      "step": 3085
+    },
+    {
+      "epoch": 2.4392282958199356,
+      "grad_norm": 0.2374882046971228,
+      "learning_rate": 1.659701811470653e-07,
+      "loss": 0.5437,
+      "step": 3086
+    },
+    {
+      "epoch": 2.4400197872866682,
+      "grad_norm": 0.23157400279180596,
+      "learning_rate": 1.655118090386509e-07,
+      "loss": 0.5132,
+      "step": 3087
+    },
+    {
+      "epoch": 2.440811278753401,
+      "grad_norm": 0.24130126154312947,
+      "learning_rate": 1.6505401365172267e-07,
+      "loss": 0.5638,
+      "step": 3088
+    },
+    {
+      "epoch": 2.4416027702201335,
+      "grad_norm": 0.23707093723590836,
+      "learning_rate": 1.6459679530266703e-07,
+      "loss": 0.5619,
+      "step": 3089
+    },
+    {
+      "epoch": 2.442394261686866,
+      "grad_norm": 0.23990481490120463,
+      "learning_rate": 1.6414015430747107e-07,
+      "loss": 0.5401,
+      "step": 3090
+    },
+    {
+      "epoch": 2.4431857531535988,
+      "grad_norm": 0.23655098262787114,
+      "learning_rate": 1.6368409098172297e-07,
+      "loss": 0.5426,
+      "step": 3091
+    },
+    {
+      "epoch": 2.4439772446203314,
+      "grad_norm": 0.23289934639423135,
+      "learning_rate": 1.6322860564061203e-07,
+      "loss": 0.541,
+      "step": 3092
+    },
+    {
+      "epoch": 2.444768736087064,
+      "grad_norm": 0.24182482596804827,
+      "learning_rate": 1.627736985989272e-07,
+      "loss": 0.5579,
+      "step": 3093
+    },
+    {
+      "epoch": 2.4455602275537967,
+      "grad_norm": 0.24072551065716197,
+      "learning_rate": 1.6231937017105923e-07,
+      "loss": 0.5397,
+      "step": 3094
+    },
+    {
+      "epoch": 2.4463517190205293,
+      "grad_norm": 0.24121211759958325,
+      "learning_rate": 1.6186562067099764e-07,
+      "loss": 0.5433,
+      "step": 3095
+    },
+    {
+      "epoch": 2.447143210487262,
+      "grad_norm": 0.23902272198133934,
+      "learning_rate": 1.6141245041233242e-07,
+      "loss": 0.5397,
+      "step": 3096
+    },
+    {
+      "epoch": 2.4479347019539945,
+      "grad_norm": 0.23477298657527262,
+      "learning_rate": 1.6095985970825321e-07,
+      "loss": 0.5447,
+      "step": 3097
+    },
+    {
+      "epoch": 2.448726193420727,
+      "grad_norm": 0.23495547741671985,
+      "learning_rate": 1.605078488715489e-07,
+      "loss": 0.5361,
+      "step": 3098
+    },
+    {
+      "epoch": 2.44951768488746,
+      "grad_norm": 0.23688597208873882,
+      "learning_rate": 1.6005641821460813e-07,
+      "loss": 0.5295,
+      "step": 3099
+    },
+    {
+      "epoch": 2.4503091763541924,
+      "grad_norm": 0.23306646738205733,
+      "learning_rate": 1.5960556804941815e-07,
+      "loss": 0.5443,
+      "step": 3100
+    },
+    {
+      "epoch": 2.451100667820925,
+      "grad_norm": 0.2333315996373385,
+      "learning_rate": 1.5915529868756507e-07,
+      "loss": 0.5453,
+      "step": 3101
+    },
+    {
+      "epoch": 2.4518921592876577,
+      "grad_norm": 0.23531650793015646,
+      "learning_rate": 1.5870561044023355e-07,
+      "loss": 0.5406,
+      "step": 3102
+    },
+    {
+      "epoch": 2.4526836507543903,
+      "grad_norm": 0.23207235034413456,
+      "learning_rate": 1.5825650361820674e-07,
+      "loss": 0.5603,
+      "step": 3103
+    },
+    {
+      "epoch": 2.453475142221123,
+      "grad_norm": 0.23593052853948004,
+      "learning_rate": 1.5780797853186623e-07,
+      "loss": 0.5483,
+      "step": 3104
+    },
+    {
+      "epoch": 2.4542666336878556,
+      "grad_norm": 0.23842629564774298,
+      "learning_rate": 1.5736003549119125e-07,
+      "loss": 0.5443,
+      "step": 3105
+    },
+    {
+      "epoch": 2.455058125154588,
+      "grad_norm": 0.22824167790126995,
+      "learning_rate": 1.5691267480575865e-07,
+      "loss": 0.5334,
+      "step": 3106
+    },
+    {
+      "epoch": 2.455849616621321,
+      "grad_norm": 0.23843678125334108,
+      "learning_rate": 1.564658967847432e-07,
+      "loss": 0.5324,
+      "step": 3107
+    },
+    {
+      "epoch": 2.4566411080880535,
+      "grad_norm": 0.23499125374752075,
+      "learning_rate": 1.5601970173691635e-07,
+      "loss": 0.5448,
+      "step": 3108
+    },
+    {
+      "epoch": 2.457432599554786,
+      "grad_norm": 0.22915750706814803,
+      "learning_rate": 1.5557408997064768e-07,
+      "loss": 0.5362,
+      "step": 3109
+    },
+    {
+      "epoch": 2.4582240910215187,
+      "grad_norm": 0.2353437999296168,
+      "learning_rate": 1.5512906179390272e-07,
+      "loss": 0.5346,
+      "step": 3110
+    },
+    {
+      "epoch": 2.4590155824882514,
+      "grad_norm": 0.23919039171583964,
+      "learning_rate": 1.54684617514244e-07,
+      "loss": 0.5435,
+      "step": 3111
+    },
+    {
+      "epoch": 2.459807073954984,
+      "grad_norm": 0.23043079808369574,
+      "learning_rate": 1.542407574388307e-07,
+      "loss": 0.5459,
+      "step": 3112
+    },
+    {
+      "epoch": 2.4605985654217166,
+      "grad_norm": 0.23705034936994757,
+      "learning_rate": 1.5379748187441766e-07,
+      "loss": 0.5497,
+      "step": 3113
+    },
+    {
+      "epoch": 2.4613900568884493,
+      "grad_norm": 0.23271866752152964,
+      "learning_rate": 1.5335479112735683e-07,
+      "loss": 0.5361,
+      "step": 3114
+    },
+    {
+      "epoch": 2.462181548355182,
+      "grad_norm": 0.24397729770321486,
+      "learning_rate": 1.5291268550359493e-07,
+      "loss": 0.5549,
+      "step": 3115
+    },
+    {
+      "epoch": 2.4629730398219145,
+      "grad_norm": 0.23664546071419895,
+      "learning_rate": 1.5247116530867477e-07,
+      "loss": 0.5461,
+      "step": 3116
+    },
+    {
+      "epoch": 2.463764531288647,
+      "grad_norm": 0.23449734137667236,
+      "learning_rate": 1.5203023084773448e-07,
+      "loss": 0.5532,
+      "step": 3117
+    },
+    {
+      "epoch": 2.4645560227553798,
+      "grad_norm": 0.22847149227790406,
+      "learning_rate": 1.5158988242550718e-07,
+      "loss": 0.5367,
+      "step": 3118
+    },
+    {
+      "epoch": 2.4653475142221124,
+      "grad_norm": 0.23977652202005234,
+      "learning_rate": 1.511501203463218e-07,
+      "loss": 0.5444,
+      "step": 3119
+    },
+    {
+      "epoch": 2.466139005688845,
+      "grad_norm": 0.237980361612846,
+      "learning_rate": 1.5071094491410098e-07,
+      "loss": 0.5435,
+      "step": 3120
+    },
+    {
+      "epoch": 2.4669304971555777,
+      "grad_norm": 0.23745647639017792,
+      "learning_rate": 1.5027235643236257e-07,
+      "loss": 0.5543,
+      "step": 3121
+    },
+    {
+      "epoch": 2.4677219886223103,
+      "grad_norm": 0.23550908797276845,
+      "learning_rate": 1.4983435520421849e-07,
+      "loss": 0.556,
+      "step": 3122
+    },
+    {
+      "epoch": 2.468513480089043,
+      "grad_norm": 0.23840297172092187,
+      "learning_rate": 1.493969415323747e-07,
+      "loss": 0.5395,
+      "step": 3123
+    },
+    {
+      "epoch": 2.4693049715557756,
+      "grad_norm": 0.23904245391064374,
+      "learning_rate": 1.4896011571913169e-07,
+      "loss": 0.5482,
+      "step": 3124
+    },
+    {
+      "epoch": 2.470096463022508,
+      "grad_norm": 0.23456396224995432,
+      "learning_rate": 1.485238780663832e-07,
+      "loss": 0.5415,
+      "step": 3125
+    },
+    {
+      "epoch": 2.470887954489241,
+      "grad_norm": 0.23216838017885,
+      "learning_rate": 1.4808822887561645e-07,
+      "loss": 0.5297,
+      "step": 3126
+    },
+    {
+      "epoch": 2.4716794459559734,
+      "grad_norm": 0.23579546284320718,
+      "learning_rate": 1.476531684479122e-07,
+      "loss": 0.5356,
+      "step": 3127
+    },
+    {
+      "epoch": 2.472470937422706,
+      "grad_norm": 0.23692145416780397,
+      "learning_rate": 1.4721869708394396e-07,
+      "loss": 0.5412,
+      "step": 3128
+    },
+    {
+      "epoch": 2.4732624288894387,
+      "grad_norm": 0.23638279385837965,
+      "learning_rate": 1.4678481508397868e-07,
+      "loss": 0.5545,
+      "step": 3129
+    },
+    {
+      "epoch": 2.4740539203561713,
+      "grad_norm": 0.23120172321373875,
+      "learning_rate": 1.4635152274787564e-07,
+      "loss": 0.5474,
+      "step": 3130
+    },
+    {
+      "epoch": 2.474845411822904,
+      "grad_norm": 0.22828148544503526,
+      "learning_rate": 1.4591882037508629e-07,
+      "loss": 0.541,
+      "step": 3131
+    },
+    {
+      "epoch": 2.4756369032896366,
+      "grad_norm": 0.24079017997885094,
+      "learning_rate": 1.4548670826465548e-07,
+      "loss": 0.5342,
+      "step": 3132
+    },
+    {
+      "epoch": 2.4764283947563692,
+      "grad_norm": 0.23558816283982148,
+      "learning_rate": 1.4505518671521854e-07,
+      "loss": 0.5485,
+      "step": 3133
+    },
+    {
+      "epoch": 2.477219886223102,
+      "grad_norm": 0.24578326210122056,
+      "learning_rate": 1.4462425602500395e-07,
+      "loss": 0.5331,
+      "step": 3134
+    },
+    {
+      "epoch": 2.4780113776898345,
+      "grad_norm": 0.2345514476597932,
+      "learning_rate": 1.441939164918312e-07,
+      "loss": 0.544,
+      "step": 3135
+    },
+    {
+      "epoch": 2.478802869156567,
+      "grad_norm": 0.24189182491688455,
+      "learning_rate": 1.4376416841311122e-07,
+      "loss": 0.5596,
+      "step": 3136
+    },
+    {
+      "epoch": 2.4795943606232997,
+      "grad_norm": 0.23566027035426818,
+      "learning_rate": 1.4333501208584708e-07,
+      "loss": 0.5493,
+      "step": 3137
+    },
+    {
+      "epoch": 2.4803858520900324,
+      "grad_norm": 0.23381446085837193,
+      "learning_rate": 1.429064478066312e-07,
+      "loss": 0.5291,
+      "step": 3138
+    },
+    {
+      "epoch": 2.481177343556765,
+      "grad_norm": 0.22876724413413854,
+      "learning_rate": 1.424784758716485e-07,
+      "loss": 0.541,
+      "step": 3139
+    },
+    {
+      "epoch": 2.4819688350234976,
+      "grad_norm": 0.23719487554609764,
+      "learning_rate": 1.4205109657667346e-07,
+      "loss": 0.5432,
+      "step": 3140
+    },
+    {
+      "epoch": 2.4827603264902303,
+      "grad_norm": 0.23126560447353134,
+      "learning_rate": 1.4162431021707134e-07,
+      "loss": 0.5397,
+      "step": 3141
+    },
+    {
+      "epoch": 2.483551817956963,
+      "grad_norm": 0.23569293373808486,
+      "learning_rate": 1.4119811708779816e-07,
+      "loss": 0.5503,
+      "step": 3142
+    },
+    {
+      "epoch": 2.4843433094236955,
+      "grad_norm": 0.24010869254677691,
+      "learning_rate": 1.4077251748339845e-07,
+      "loss": 0.5485,
+      "step": 3143
+    },
+    {
+      "epoch": 2.4851348008904277,
+      "grad_norm": 0.23668252067028273,
+      "learning_rate": 1.4034751169800828e-07,
+      "loss": 0.548,
+      "step": 3144
+    },
+    {
+      "epoch": 2.4859262923571603,
+      "grad_norm": 0.23061861151519614,
+      "learning_rate": 1.3992310002535235e-07,
+      "loss": 0.5334,
+      "step": 3145
+    },
+    {
+      "epoch": 2.486717783823893,
+      "grad_norm": 0.23424778790966844,
+      "learning_rate": 1.3949928275874468e-07,
+      "loss": 0.5522,
+      "step": 3146
+    },
+    {
+      "epoch": 2.4875092752906256,
+      "grad_norm": 0.23764531062904293,
+      "learning_rate": 1.3907606019108952e-07,
+      "loss": 0.5382,
+      "step": 3147
+    },
+    {
+      "epoch": 2.4883007667573582,
+      "grad_norm": 0.2344565009397954,
+      "learning_rate": 1.386534326148787e-07,
+      "loss": 0.544,
+      "step": 3148
+    },
+    {
+      "epoch": 2.489092258224091,
+      "grad_norm": 0.24085363881369712,
+      "learning_rate": 1.382314003221935e-07,
+      "loss": 0.5498,
+      "step": 3149
+    },
+    {
+      "epoch": 2.4898837496908235,
+      "grad_norm": 0.24149543811050822,
+      "learning_rate": 1.3780996360470432e-07,
+      "loss": 0.5442,
+      "step": 3150
+    },
+    {
+      "epoch": 2.490675241157556,
+      "grad_norm": 0.23687706561240804,
+      "learning_rate": 1.3738912275366898e-07,
+      "loss": 0.5445,
+      "step": 3151
+    },
+    {
+      "epoch": 2.4914667326242887,
+      "grad_norm": 0.24028087206167023,
+      "learning_rate": 1.369688780599344e-07,
+      "loss": 0.553,
+      "step": 3152
+    },
+    {
+      "epoch": 2.4922582240910214,
+      "grad_norm": 0.2362314693025883,
+      "learning_rate": 1.3654922981393503e-07,
+      "loss": 0.5555,
+      "step": 3153
+    },
+    {
+      "epoch": 2.493049715557754,
+      "grad_norm": 0.24010830855718476,
+      "learning_rate": 1.3613017830569252e-07,
+      "loss": 0.5493,
+      "step": 3154
+    },
+    {
+      "epoch": 2.4938412070244866,
+      "grad_norm": 0.23996396449725285,
+      "learning_rate": 1.3571172382481744e-07,
+      "loss": 0.5532,
+      "step": 3155
+    },
+    {
+      "epoch": 2.4946326984912193,
+      "grad_norm": 0.2565057060388274,
+      "learning_rate": 1.3529386666050656e-07,
+      "loss": 0.5539,
+      "step": 3156
+    },
+    {
+      "epoch": 2.495424189957952,
+      "grad_norm": 0.2280272817404365,
+      "learning_rate": 1.348766071015447e-07,
+      "loss": 0.5386,
+      "step": 3157
+    },
+    {
+      "epoch": 2.4962156814246845,
+      "grad_norm": 0.2332522637920383,
+      "learning_rate": 1.3445994543630346e-07,
+      "loss": 0.5246,
+      "step": 3158
+    },
+    {
+      "epoch": 2.497007172891417,
+      "grad_norm": 0.2321020310190722,
+      "learning_rate": 1.340438819527402e-07,
+      "loss": 0.5472,
+      "step": 3159
+    },
+    {
+      "epoch": 2.49779866435815,
+      "grad_norm": 0.23551347986342572,
+      "learning_rate": 1.3362841693840044e-07,
+      "loss": 0.548,
+      "step": 3160
+    },
+    {
+      "epoch": 2.4985901558248824,
+      "grad_norm": 0.23739641225781086,
+      "learning_rate": 1.3321355068041505e-07,
+      "loss": 0.555,
+      "step": 3161
+    },
+    {
+      "epoch": 2.499381647291615,
+      "grad_norm": 0.2383612647706636,
+      "learning_rate": 1.3279928346550185e-07,
+      "loss": 0.5559,
+      "step": 3162
+    },
+    {
+      "epoch": 2.5001731387583477,
+      "grad_norm": 0.23159310295039023,
+      "learning_rate": 1.3238561557996408e-07,
+      "loss": 0.5431,
+      "step": 3163
+    },
+    {
+      "epoch": 2.5009646302250803,
+      "grad_norm": 0.2364148670331119,
+      "learning_rate": 1.3197254730969042e-07,
+      "loss": 0.5365,
+      "step": 3164
+    },
+    {
+      "epoch": 2.501756121691813,
+      "grad_norm": 0.23841468904139962,
+      "learning_rate": 1.315600789401563e-07,
+      "loss": 0.5346,
+      "step": 3165
+    },
+    {
+      "epoch": 2.5025476131585456,
+      "grad_norm": 0.23440476605397462,
+      "learning_rate": 1.311482107564218e-07,
+      "loss": 0.5566,
+      "step": 3166
+    },
+    {
+      "epoch": 2.503339104625278,
+      "grad_norm": 0.2367995177563655,
+      "learning_rate": 1.3073694304313186e-07,
+      "loss": 0.553,
+      "step": 3167
+    },
+    {
+      "epoch": 2.504130596092011,
+      "grad_norm": 0.23088595959065358,
+      "learning_rate": 1.3032627608451774e-07,
+      "loss": 0.5359,
+      "step": 3168
+    },
+    {
+      "epoch": 2.5049220875587435,
+      "grad_norm": 0.23754198283978084,
+      "learning_rate": 1.299162101643938e-07,
+      "loss": 0.5498,
+      "step": 3169
+    },
+    {
+      "epoch": 2.505713579025476,
+      "grad_norm": 0.23484073685539716,
+      "learning_rate": 1.2950674556616047e-07,
+      "loss": 0.5263,
+      "step": 3170
+    },
+    {
+      "epoch": 2.5065050704922087,
+      "grad_norm": 0.23690510100612922,
+      "learning_rate": 1.2909788257280185e-07,
+      "loss": 0.5466,
+      "step": 3171
+    },
+    {
+      "epoch": 2.5072965619589413,
+      "grad_norm": 0.24288340940723652,
+      "learning_rate": 1.2868962146688612e-07,
+      "loss": 0.5522,
+      "step": 3172
+    },
+    {
+      "epoch": 2.508088053425674,
+      "grad_norm": 0.22880200656756094,
+      "learning_rate": 1.2828196253056633e-07,
+      "loss": 0.5294,
+      "step": 3173
+    },
+    {
+      "epoch": 2.5088795448924066,
+      "grad_norm": 0.23376621779238194,
+      "learning_rate": 1.278749060455787e-07,
+      "loss": 0.5447,
+      "step": 3174
+    },
+    {
+      "epoch": 2.5096710363591392,
+      "grad_norm": 0.23597945277757415,
+      "learning_rate": 1.274684522932431e-07,
+      "loss": 0.5561,
+      "step": 3175
+    },
+    {
+      "epoch": 2.510462527825872,
+      "grad_norm": 0.23957405682153782,
+      "learning_rate": 1.2706260155446302e-07,
+      "loss": 0.5466,
+      "step": 3176
+    },
+    {
+      "epoch": 2.5112540192926045,
+      "grad_norm": 0.2595199288251904,
+      "learning_rate": 1.26657354109725e-07,
+      "loss": 0.5421,
+      "step": 3177
+    },
+    {
+      "epoch": 2.512045510759337,
+      "grad_norm": 0.23222753790037834,
+      "learning_rate": 1.2625271023909924e-07,
+      "loss": 0.5526,
+      "step": 3178
+    },
+    {
+      "epoch": 2.5128370022260698,
+      "grad_norm": 0.23166962662010931,
+      "learning_rate": 1.25848670222238e-07,
+      "loss": 0.5343,
+      "step": 3179
+    },
+    {
+      "epoch": 2.5136284936928024,
+      "grad_norm": 0.23028563286774928,
+      "learning_rate": 1.2544523433837672e-07,
+      "loss": 0.5422,
+      "step": 3180
+    },
+    {
+      "epoch": 2.514419985159535,
+      "grad_norm": 0.2348245119892915,
+      "learning_rate": 1.2504240286633317e-07,
+      "loss": 0.5405,
+      "step": 3181
+    },
+    {
+      "epoch": 2.5152114766262677,
+      "grad_norm": 0.2334086471897704,
+      "learning_rate": 1.2464017608450706e-07,
+      "loss": 0.5252,
+      "step": 3182
+    },
+    {
+      "epoch": 2.5160029680930003,
+      "grad_norm": 0.23597440381205284,
+      "learning_rate": 1.2423855427088114e-07,
+      "loss": 0.5424,
+      "step": 3183
+    },
+    {
+      "epoch": 2.516794459559733,
+      "grad_norm": 0.24481159177007725,
+      "learning_rate": 1.238375377030192e-07,
+      "loss": 0.5425,
+      "step": 3184
+    },
+    {
+      "epoch": 2.5175859510264655,
+      "grad_norm": 0.23343778404294818,
+      "learning_rate": 1.2343712665806705e-07,
+      "loss": 0.5462,
+      "step": 3185
+    },
+    {
+      "epoch": 2.518377442493198,
+      "grad_norm": 0.23356755763259637,
+      "learning_rate": 1.2303732141275192e-07,
+      "loss": 0.5608,
+      "step": 3186
+    },
+    {
+      "epoch": 2.519168933959931,
+      "grad_norm": 0.23451727738070482,
+      "learning_rate": 1.2263812224338222e-07,
+      "loss": 0.5444,
+      "step": 3187
+    },
+    {
+      "epoch": 2.5199604254266634,
+      "grad_norm": 0.24347780073046107,
+      "learning_rate": 1.2223952942584804e-07,
+      "loss": 0.5452,
+      "step": 3188
+    },
+    {
+      "epoch": 2.520751916893396,
+      "grad_norm": 0.23632368926850109,
+      "learning_rate": 1.2184154323562003e-07,
+      "loss": 0.5468,
+      "step": 3189
+    },
+    {
+      "epoch": 2.5215434083601287,
+      "grad_norm": 0.2320789472342085,
+      "learning_rate": 1.2144416394774958e-07,
+      "loss": 0.5404,
+      "step": 3190
+    },
+    {
+      "epoch": 2.5223348998268613,
+      "grad_norm": 0.23496457576160137,
+      "learning_rate": 1.2104739183686863e-07,
+      "loss": 0.5344,
+      "step": 3191
+    },
+    {
+      "epoch": 2.523126391293594,
+      "grad_norm": 0.23491603881659645,
+      "learning_rate": 1.2065122717718946e-07,
+      "loss": 0.5444,
+      "step": 3192
+    },
+    {
+      "epoch": 2.5239178827603266,
+      "grad_norm": 0.23610267089430673,
+      "learning_rate": 1.2025567024250504e-07,
+      "loss": 0.545,
+      "step": 3193
+    },
+    {
+      "epoch": 2.524709374227059,
+      "grad_norm": 0.2373998158761749,
+      "learning_rate": 1.1986072130618762e-07,
+      "loss": 0.5372,
+      "step": 3194
+    },
+    {
+      "epoch": 2.525500865693792,
+      "grad_norm": 0.23146349394195206,
+      "learning_rate": 1.1946638064118975e-07,
+      "loss": 0.5409,
+      "step": 3195
+    },
+    {
+      "epoch": 2.5262923571605245,
+      "grad_norm": 0.23361231660690018,
+      "learning_rate": 1.190726485200434e-07,
+      "loss": 0.5412,
+      "step": 3196
+    },
+    {
+      "epoch": 2.527083848627257,
+      "grad_norm": 0.23289859428625342,
+      "learning_rate": 1.1867952521485958e-07,
+      "loss": 0.5352,
+      "step": 3197
+    },
+    {
+      "epoch": 2.5278753400939897,
+      "grad_norm": 0.23007999272443694,
+      "learning_rate": 1.1828701099732973e-07,
+      "loss": 0.547,
+      "step": 3198
+    },
+    {
+      "epoch": 2.5286668315607224,
+      "grad_norm": 0.23535159755948495,
+      "learning_rate": 1.1789510613872311e-07,
+      "loss": 0.5238,
+      "step": 3199
+    },
+    {
+      "epoch": 2.529458323027455,
+      "grad_norm": 0.2341115865553397,
+      "learning_rate": 1.1750381090988837e-07,
+      "loss": 0.5449,
+      "step": 3200
+    },
+    {
+      "epoch": 2.5302498144941876,
+      "grad_norm": 0.2315684136615418,
+      "learning_rate": 1.1711312558125275e-07,
+      "loss": 0.5455,
+      "step": 3201
+    },
+    {
+      "epoch": 2.5310413059609203,
+      "grad_norm": 0.2342899619112105,
+      "learning_rate": 1.1672305042282193e-07,
+      "loss": 0.5417,
+      "step": 3202
+    },
+    {
+      "epoch": 2.531832797427653,
+      "grad_norm": 0.23359039775215643,
+      "learning_rate": 1.1633358570418028e-07,
+      "loss": 0.5529,
+      "step": 3203
+    },
+    {
+      "epoch": 2.5326242888943855,
+      "grad_norm": 0.23345481895639528,
+      "learning_rate": 1.1594473169448982e-07,
+      "loss": 0.5338,
+      "step": 3204
+    },
+    {
+      "epoch": 2.533415780361118,
+      "grad_norm": 0.23947237723853748,
+      "learning_rate": 1.1555648866249068e-07,
+      "loss": 0.553,
+      "step": 3205
+    },
+    {
+      "epoch": 2.5342072718278508,
+      "grad_norm": 0.23313844462433372,
+      "learning_rate": 1.151688568765008e-07,
+      "loss": 0.5368,
+      "step": 3206
+    },
+    {
+      "epoch": 2.534998763294583,
+      "grad_norm": 0.23168108763650186,
+      "learning_rate": 1.1478183660441532e-07,
+      "loss": 0.549,
+      "step": 3207
+    },
+    {
+      "epoch": 2.5357902547613156,
+      "grad_norm": 0.23784493111942692,
+      "learning_rate": 1.143954281137075e-07,
+      "loss": 0.535,
+      "step": 3208
+    },
+    {
+      "epoch": 2.536581746228048,
+      "grad_norm": 0.2374851402734962,
+      "learning_rate": 1.1400963167142719e-07,
+      "loss": 0.5491,
+      "step": 3209
+    },
+    {
+      "epoch": 2.537373237694781,
+      "grad_norm": 0.2317170391103134,
+      "learning_rate": 1.1362444754420131e-07,
+      "loss": 0.5384,
+      "step": 3210
+    },
+    {
+      "epoch": 2.5381647291615135,
+      "grad_norm": 0.23267068568888466,
+      "learning_rate": 1.132398759982337e-07,
+      "loss": 0.5313,
+      "step": 3211
+    },
+    {
+      "epoch": 2.538956220628246,
+      "grad_norm": 0.23703211103344654,
+      "learning_rate": 1.1285591729930477e-07,
+      "loss": 0.546,
+      "step": 3212
+    },
+    {
+      "epoch": 2.5397477120949787,
+      "grad_norm": 0.2321030061925981,
+      "learning_rate": 1.124725717127718e-07,
+      "loss": 0.5463,
+      "step": 3213
+    },
+    {
+      "epoch": 2.5405392035617114,
+      "grad_norm": 0.23448127548658504,
+      "learning_rate": 1.1208983950356776e-07,
+      "loss": 0.5394,
+      "step": 3214
+    },
+    {
+      "epoch": 2.541330695028444,
+      "grad_norm": 0.22892239445385304,
+      "learning_rate": 1.1170772093620184e-07,
+      "loss": 0.5384,
+      "step": 3215
+    },
+    {
+      "epoch": 2.5421221864951766,
+      "grad_norm": 0.23527838040468413,
+      "learning_rate": 1.1132621627475969e-07,
+      "loss": 0.5435,
+      "step": 3216
+    },
+    {
+      "epoch": 2.5429136779619093,
+      "grad_norm": 0.2357248356865842,
+      "learning_rate": 1.1094532578290161e-07,
+      "loss": 0.5417,
+      "step": 3217
+    },
+    {
+      "epoch": 2.543705169428642,
+      "grad_norm": 0.23333939799332037,
+      "learning_rate": 1.1056504972386449e-07,
+      "loss": 0.5267,
+      "step": 3218
+    },
+    {
+      "epoch": 2.5444966608953745,
+      "grad_norm": 0.2355635099505826,
+      "learning_rate": 1.1018538836046021e-07,
+      "loss": 0.5296,
+      "step": 3219
+    },
+    {
+      "epoch": 2.545288152362107,
+      "grad_norm": 0.23679869926609523,
+      "learning_rate": 1.0980634195507543e-07,
+      "loss": 0.5419,
+      "step": 3220
+    },
+    {
+      "epoch": 2.54607964382884,
+      "grad_norm": 0.22770204545009404,
+      "learning_rate": 1.0942791076967272e-07,
+      "loss": 0.5326,
+      "step": 3221
+    },
+    {
+      "epoch": 2.5468711352955724,
+      "grad_norm": 0.24015483577927543,
+      "learning_rate": 1.0905009506578822e-07,
+      "loss": 0.5336,
+      "step": 3222
+    },
+    {
+      "epoch": 2.547662626762305,
+      "grad_norm": 0.24046943015257521,
+      "learning_rate": 1.0867289510453403e-07,
+      "loss": 0.5595,
+      "step": 3223
+    },
+    {
+      "epoch": 2.5484541182290377,
+      "grad_norm": 0.23767828535833865,
+      "learning_rate": 1.0829631114659576e-07,
+      "loss": 0.5229,
+      "step": 3224
+    },
+    {
+      "epoch": 2.5492456096957703,
+      "grad_norm": 0.2404685202276093,
+      "learning_rate": 1.0792034345223333e-07,
+      "loss": 0.5632,
+      "step": 3225
+    },
+    {
+      "epoch": 2.550037101162503,
+      "grad_norm": 0.23068258897467078,
+      "learning_rate": 1.0754499228128189e-07,
+      "loss": 0.532,
+      "step": 3226
+    },
+    {
+      "epoch": 2.5508285926292356,
+      "grad_norm": 0.23602688581386108,
+      "learning_rate": 1.071702578931486e-07,
+      "loss": 0.5358,
+      "step": 3227
+    },
+    {
+      "epoch": 2.551620084095968,
+      "grad_norm": 0.22894322587648852,
+      "learning_rate": 1.0679614054681618e-07,
+      "loss": 0.5329,
+      "step": 3228
+    },
+    {
+      "epoch": 2.552411575562701,
+      "grad_norm": 0.2312081834433616,
+      "learning_rate": 1.0642264050083983e-07,
+      "loss": 0.546,
+      "step": 3229
+    },
+    {
+      "epoch": 2.5532030670294334,
+      "grad_norm": 0.2419104611354076,
+      "learning_rate": 1.0604975801334826e-07,
+      "loss": 0.5473,
+      "step": 3230
+    },
+    {
+      "epoch": 2.553994558496166,
+      "grad_norm": 0.23702332247088082,
+      "learning_rate": 1.0567749334204412e-07,
+      "loss": 0.5442,
+      "step": 3231
+    },
+    {
+      "epoch": 2.5547860499628987,
+      "grad_norm": 0.23270703236793058,
+      "learning_rate": 1.0530584674420173e-07,
+      "loss": 0.5456,
+      "step": 3232
+    },
+    {
+      "epoch": 2.5555775414296313,
+      "grad_norm": 0.22961756966436309,
+      "learning_rate": 1.0493481847666963e-07,
+      "loss": 0.5352,
+      "step": 3233
+    },
+    {
+      "epoch": 2.556369032896364,
+      "grad_norm": 0.28629052688498435,
+      "learning_rate": 1.0456440879586825e-07,
+      "loss": 0.5511,
+      "step": 3234
+    },
+    {
+      "epoch": 2.5571605243630966,
+      "grad_norm": 0.24356184836417788,
+      "learning_rate": 1.041946179577905e-07,
+      "loss": 0.5294,
+      "step": 3235
+    },
+    {
+      "epoch": 2.5579520158298292,
+      "grad_norm": 0.23696380822952862,
+      "learning_rate": 1.0382544621800216e-07,
+      "loss": 0.5577,
+      "step": 3236
+    },
+    {
+      "epoch": 2.558743507296562,
+      "grad_norm": 0.23649415572248325,
+      "learning_rate": 1.0345689383164068e-07,
+      "loss": 0.5571,
+      "step": 3237
+    },
+    {
+      "epoch": 2.5595349987632945,
+      "grad_norm": 0.2425831872092614,
+      "learning_rate": 1.0308896105341547e-07,
+      "loss": 0.5494,
+      "step": 3238
+    },
+    {
+      "epoch": 2.560326490230027,
+      "grad_norm": 0.23098879668762667,
+      "learning_rate": 1.0272164813760787e-07,
+      "loss": 0.5318,
+      "step": 3239
+    },
+    {
+      "epoch": 2.5611179816967597,
+      "grad_norm": 0.23372087151188703,
+      "learning_rate": 1.0235495533807048e-07,
+      "loss": 0.5287,
+      "step": 3240
+    },
+    {
+      "epoch": 2.5619094731634924,
+      "grad_norm": 0.23265948131008293,
+      "learning_rate": 1.0198888290822827e-07,
+      "loss": 0.5396,
+      "step": 3241
+    },
+    {
+      "epoch": 2.562700964630225,
+      "grad_norm": 0.23426685852800547,
+      "learning_rate": 1.0162343110107641e-07,
+      "loss": 0.5274,
+      "step": 3242
+    },
+    {
+      "epoch": 2.5634924560969576,
+      "grad_norm": 0.2271260522432803,
+      "learning_rate": 1.0125860016918163e-07,
+      "loss": 0.5324,
+      "step": 3243
+    },
+    {
+      "epoch": 2.5642839475636903,
+      "grad_norm": 0.23571815367832072,
+      "learning_rate": 1.008943903646815e-07,
+      "loss": 0.5447,
+      "step": 3244
+    },
+    {
+      "epoch": 2.565075439030423,
+      "grad_norm": 0.23668867736100763,
+      "learning_rate": 1.0053080193928432e-07,
+      "loss": 0.5511,
+      "step": 3245
+    },
+    {
+      "epoch": 2.5658669304971555,
+      "grad_norm": 0.2295744892562428,
+      "learning_rate": 1.0016783514426919e-07,
+      "loss": 0.5336,
+      "step": 3246
+    },
+    {
+      "epoch": 2.566658421963888,
+      "grad_norm": 0.23245559143610908,
+      "learning_rate": 9.980549023048512e-08,
+      "loss": 0.5434,
+      "step": 3247
+    },
+    {
+      "epoch": 2.567449913430621,
+      "grad_norm": 0.23430477667816743,
+      "learning_rate": 9.944376744835181e-08,
+      "loss": 0.528,
+      "step": 3248
+    },
+    {
+      "epoch": 2.5682414048973534,
+      "grad_norm": 0.23550847832675992,
+      "learning_rate": 9.90826670478585e-08,
+      "loss": 0.5451,
+      "step": 3249
+    },
+    {
+      "epoch": 2.569032896364086,
+      "grad_norm": 0.23578221932952387,
+      "learning_rate": 9.872218927856468e-08,
+      "loss": 0.5546,
+      "step": 3250
+    },
+    {
+      "epoch": 2.5698243878308187,
+      "grad_norm": 0.2355131785530226,
+      "learning_rate": 9.836233438959962e-08,
+      "loss": 0.5599,
+      "step": 3251
+    },
+    {
+      "epoch": 2.5706158792975513,
+      "grad_norm": 0.23085690150964858,
+      "learning_rate": 9.800310262966183e-08,
+      "loss": 0.5392,
+      "step": 3252
+    },
+    {
+      "epoch": 2.571407370764284,
+      "grad_norm": 0.23610506349929572,
+      "learning_rate": 9.764449424701915e-08,
+      "loss": 0.5524,
+      "step": 3253
+    },
+    {
+      "epoch": 2.5721988622310166,
+      "grad_norm": 0.2320838939245817,
+      "learning_rate": 9.728650948950878e-08,
+      "loss": 0.5533,
+      "step": 3254
+    },
+    {
+      "epoch": 2.572990353697749,
+      "grad_norm": 0.2371740138803932,
+      "learning_rate": 9.69291486045366e-08,
+      "loss": 0.5498,
+      "step": 3255
+    },
+    {
+      "epoch": 2.573781845164482,
+      "grad_norm": 0.2345519697901397,
+      "learning_rate": 9.657241183907805e-08,
+      "loss": 0.5384,
+      "step": 3256
+    },
+    {
+      "epoch": 2.5745733366312145,
+      "grad_norm": 0.22977283745907118,
+      "learning_rate": 9.621629943967669e-08,
+      "loss": 0.536,
+      "step": 3257
+    },
+    {
+      "epoch": 2.575364828097947,
+      "grad_norm": 0.2317062114794596,
+      "learning_rate": 9.586081165244442e-08,
+      "loss": 0.5463,
+      "step": 3258
+    },
+    {
+      "epoch": 2.5761563195646797,
+      "grad_norm": 0.2361978413822048,
+      "learning_rate": 9.550594872306183e-08,
+      "loss": 0.55,
+      "step": 3259
+    },
+    {
+      "epoch": 2.5769478110314123,
+      "grad_norm": 0.22798913560472214,
+      "learning_rate": 9.515171089677731e-08,
+      "loss": 0.5372,
+      "step": 3260
+    },
+    {
+      "epoch": 2.577739302498145,
+      "grad_norm": 0.23256596597325183,
+      "learning_rate": 9.47980984184078e-08,
+      "loss": 0.5409,
+      "step": 3261
+    },
+    {
+      "epoch": 2.5785307939648776,
+      "grad_norm": 0.23189532545734856,
+      "learning_rate": 9.44451115323377e-08,
+      "loss": 0.5446,
+      "step": 3262
+    },
+    {
+      "epoch": 2.5793222854316102,
+      "grad_norm": 0.2336721072308425,
+      "learning_rate": 9.409275048251896e-08,
+      "loss": 0.5427,
+      "step": 3263
+    },
+    {
+      "epoch": 2.580113776898343,
+      "grad_norm": 0.24228150907741597,
+      "learning_rate": 9.374101551247115e-08,
+      "loss": 0.5629,
+      "step": 3264
+    },
+    {
+      "epoch": 2.5809052683650755,
+      "grad_norm": 0.2332936061305752,
+      "learning_rate": 9.338990686528092e-08,
+      "loss": 0.5439,
+      "step": 3265
+    },
+    {
+      "epoch": 2.581696759831808,
+      "grad_norm": 0.2603182451765577,
+      "learning_rate": 9.30394247836026e-08,
+      "loss": 0.5356,
+      "step": 3266
+    },
+    {
+      "epoch": 2.5824882512985408,
+      "grad_norm": 0.23445377307724014,
+      "learning_rate": 9.268956950965712e-08,
+      "loss": 0.5566,
+      "step": 3267
+    },
+    {
+      "epoch": 2.5832797427652734,
+      "grad_norm": 0.23170666610167254,
+      "learning_rate": 9.234034128523227e-08,
+      "loss": 0.5456,
+      "step": 3268
+    },
+    {
+      "epoch": 2.584071234232006,
+      "grad_norm": 0.23973468636497786,
+      "learning_rate": 9.199174035168222e-08,
+      "loss": 0.5471,
+      "step": 3269
+    },
+    {
+      "epoch": 2.5848627256987387,
+      "grad_norm": 0.23884964300759012,
+      "learning_rate": 9.164376694992804e-08,
+      "loss": 0.5288,
+      "step": 3270
+    },
+    {
+      "epoch": 2.5856542171654713,
+      "grad_norm": 0.23616746385245097,
+      "learning_rate": 9.129642132045712e-08,
+      "loss": 0.5297,
+      "step": 3271
+    },
+    {
+      "epoch": 2.586445708632204,
+      "grad_norm": 0.23211053247086316,
+      "learning_rate": 9.094970370332256e-08,
+      "loss": 0.5422,
+      "step": 3272
+    },
+    {
+      "epoch": 2.5872372000989365,
+      "grad_norm": 0.23237141188613064,
+      "learning_rate": 9.060361433814378e-08,
+      "loss": 0.5376,
+      "step": 3273
+    },
+    {
+      "epoch": 2.588028691565669,
+      "grad_norm": 0.23244278897303372,
+      "learning_rate": 9.025815346410604e-08,
+      "loss": 0.5499,
+      "step": 3274
+    },
+    {
+      "epoch": 2.588820183032402,
+      "grad_norm": 0.2405070173127924,
+      "learning_rate": 8.991332131995987e-08,
+      "loss": 0.55,
+      "step": 3275
+    },
+    {
+      "epoch": 2.5896116744991344,
+      "grad_norm": 0.22891260878735048,
+      "learning_rate": 8.956911814402157e-08,
+      "loss": 0.5409,
+      "step": 3276
+    },
+    {
+      "epoch": 2.590403165965867,
+      "grad_norm": 0.23211457918407827,
+      "learning_rate": 8.922554417417294e-08,
+      "loss": 0.5439,
+      "step": 3277
+    },
+    {
+      "epoch": 2.5911946574325997,
+      "grad_norm": 0.23800191370437923,
+      "learning_rate": 8.88825996478606e-08,
+      "loss": 0.5369,
+      "step": 3278
+    },
+    {
+      "epoch": 2.5919861488993323,
+      "grad_norm": 0.2422727156327912,
+      "learning_rate": 8.85402848020963e-08,
+      "loss": 0.5432,
+      "step": 3279
+    },
+    {
+      "epoch": 2.592777640366065,
+      "grad_norm": 0.24076231174928742,
+      "learning_rate": 8.819859987345645e-08,
+      "loss": 0.5611,
+      "step": 3280
+    },
+    {
+      "epoch": 2.5935691318327976,
+      "grad_norm": 0.23758532741580235,
+      "learning_rate": 8.785754509808207e-08,
+      "loss": 0.5401,
+      "step": 3281
+    },
+    {
+      "epoch": 2.59436062329953,
+      "grad_norm": 0.2349556462570344,
+      "learning_rate": 8.751712071167939e-08,
+      "loss": 0.5432,
+      "step": 3282
+    },
+    {
+      "epoch": 2.595152114766263,
+      "grad_norm": 0.22976396515524597,
+      "learning_rate": 8.717732694951796e-08,
+      "loss": 0.5421,
+      "step": 3283
+    },
+    {
+      "epoch": 2.5959436062329955,
+      "grad_norm": 0.23680472103924952,
+      "learning_rate": 8.683816404643252e-08,
+      "loss": 0.5355,
+      "step": 3284
+    },
+    {
+      "epoch": 2.596735097699728,
+      "grad_norm": 0.24204305535113652,
+      "learning_rate": 8.649963223682066e-08,
+      "loss": 0.5404,
+      "step": 3285
+    },
+    {
+      "epoch": 2.5975265891664607,
+      "grad_norm": 0.23402848897488596,
+      "learning_rate": 8.616173175464458e-08,
+      "loss": 0.5391,
+      "step": 3286
+    },
+    {
+      "epoch": 2.5983180806331934,
+      "grad_norm": 0.23380288641913877,
+      "learning_rate": 8.582446283343026e-08,
+      "loss": 0.5356,
+      "step": 3287
+    },
+    {
+      "epoch": 2.599109572099926,
+      "grad_norm": 0.23782795634634146,
+      "learning_rate": 8.548782570626666e-08,
+      "loss": 0.5286,
+      "step": 3288
+    },
+    {
+      "epoch": 2.5999010635666586,
+      "grad_norm": 0.23406536154328317,
+      "learning_rate": 8.515182060580683e-08,
+      "loss": 0.533,
+      "step": 3289
+    },
+    {
+      "epoch": 2.6006925550333913,
+      "grad_norm": 0.23267787452687938,
+      "learning_rate": 8.481644776426633e-08,
+      "loss": 0.5418,
+      "step": 3290
+    },
+    {
+      "epoch": 2.601484046500124,
+      "grad_norm": 0.23271510406162305,
+      "learning_rate": 8.448170741342364e-08,
+      "loss": 0.5467,
+      "step": 3291
+    },
+    {
+      "epoch": 2.6022755379668565,
+      "grad_norm": 0.23772719690013952,
+      "learning_rate": 8.414759978462116e-08,
+      "loss": 0.538,
+      "step": 3292
+    },
+    {
+      "epoch": 2.603067029433589,
+      "grad_norm": 0.2364391397005506,
+      "learning_rate": 8.381412510876307e-08,
+      "loss": 0.5479,
+      "step": 3293
+    },
+    {
+      "epoch": 2.6038585209003218,
+      "grad_norm": 0.23015522710192796,
+      "learning_rate": 8.348128361631634e-08,
+      "loss": 0.5447,
+      "step": 3294
+    },
+    {
+      "epoch": 2.6046500123670544,
+      "grad_norm": 0.2370660271484739,
+      "learning_rate": 8.314907553731054e-08,
+      "loss": 0.5463,
+      "step": 3295
+    },
+    {
+      "epoch": 2.605441503833787,
+      "grad_norm": 0.23698433146290473,
+      "learning_rate": 8.281750110133701e-08,
+      "loss": 0.5367,
+      "step": 3296
+    },
+    {
+      "epoch": 2.6062329953005197,
+      "grad_norm": 0.23910359816647553,
+      "learning_rate": 8.248656053754999e-08,
+      "loss": 0.5407,
+      "step": 3297
+    },
+    {
+      "epoch": 2.6070244867672523,
+      "grad_norm": 0.23728545288367459,
+      "learning_rate": 8.215625407466486e-08,
+      "loss": 0.5453,
+      "step": 3298
+    },
+    {
+      "epoch": 2.607815978233985,
+      "grad_norm": 0.23480932634169566,
+      "learning_rate": 8.182658194095914e-08,
+      "loss": 0.5515,
+      "step": 3299
+    },
+    {
+      "epoch": 2.6086074697007176,
+      "grad_norm": 0.23802606396141185,
+      "learning_rate": 8.14975443642718e-08,
+      "loss": 0.5509,
+      "step": 3300
+    },
+    {
+      "epoch": 2.60939896116745,
+      "grad_norm": 0.22972271494621538,
+      "learning_rate": 8.116914157200339e-08,
+      "loss": 0.5362,
+      "step": 3301
+    },
+    {
+      "epoch": 2.610190452634183,
+      "grad_norm": 0.23486271426552657,
+      "learning_rate": 8.084137379111577e-08,
+      "loss": 0.5218,
+      "step": 3302
+    },
+    {
+      "epoch": 2.610981944100915,
+      "grad_norm": 0.2306382742158135,
+      "learning_rate": 8.051424124813199e-08,
+      "loss": 0.5499,
+      "step": 3303
+    },
+    {
+      "epoch": 2.6117734355676476,
+      "grad_norm": 0.2343152822354507,
+      "learning_rate": 8.018774416913565e-08,
+      "loss": 0.5483,
+      "step": 3304
+    },
+    {
+      "epoch": 2.6125649270343803,
+      "grad_norm": 0.23138535389201664,
+      "learning_rate": 7.986188277977202e-08,
+      "loss": 0.5304,
+      "step": 3305
+    },
+    {
+      "epoch": 2.613356418501113,
+      "grad_norm": 0.23748623310209613,
+      "learning_rate": 7.953665730524584e-08,
+      "loss": 0.5455,
+      "step": 3306
+    },
+    {
+      "epoch": 2.6141479099678455,
+      "grad_norm": 0.23805936946734416,
+      "learning_rate": 7.921206797032354e-08,
+      "loss": 0.5558,
+      "step": 3307
+    },
+    {
+      "epoch": 2.614939401434578,
+      "grad_norm": 0.23203940724739996,
+      "learning_rate": 7.888811499933135e-08,
+      "loss": 0.5291,
+      "step": 3308
+    },
+    {
+      "epoch": 2.6157308929013108,
+      "grad_norm": 0.24196687723997037,
+      "learning_rate": 7.856479861615551e-08,
+      "loss": 0.5568,
+      "step": 3309
+    },
+    {
+      "epoch": 2.6165223843680434,
+      "grad_norm": 0.22756235172460287,
+      "learning_rate": 7.824211904424305e-08,
+      "loss": 0.5415,
+      "step": 3310
+    },
+    {
+      "epoch": 2.617313875834776,
+      "grad_norm": 0.2345438116625913,
+      "learning_rate": 7.792007650659981e-08,
+      "loss": 0.5402,
+      "step": 3311
+    },
+    {
+      "epoch": 2.6181053673015087,
+      "grad_norm": 0.2321566853556492,
+      "learning_rate": 7.759867122579245e-08,
+      "loss": 0.5306,
+      "step": 3312
+    },
+    {
+      "epoch": 2.6188968587682413,
+      "grad_norm": 0.2303537722632866,
+      "learning_rate": 7.727790342394657e-08,
+      "loss": 0.5417,
+      "step": 3313
+    },
+    {
+      "epoch": 2.619688350234974,
+      "grad_norm": 0.23668844423079852,
+      "learning_rate": 7.695777332274722e-08,
+      "loss": 0.5382,
+      "step": 3314
+    },
+    {
+      "epoch": 2.6204798417017066,
+      "grad_norm": 0.2414251156724988,
+      "learning_rate": 7.663828114343939e-08,
+      "loss": 0.5461,
+      "step": 3315
+    },
+    {
+      "epoch": 2.621271333168439,
+      "grad_norm": 0.2288192327340346,
+      "learning_rate": 7.631942710682593e-08,
+      "loss": 0.5508,
+      "step": 3316
+    },
+    {
+      "epoch": 2.622062824635172,
+      "grad_norm": 0.2309913863417435,
+      "learning_rate": 7.60012114332701e-08,
+      "loss": 0.5526,
+      "step": 3317
+    },
+    {
+      "epoch": 2.6228543161019044,
+      "grad_norm": 0.23524931378287434,
+      "learning_rate": 7.568363434269299e-08,
+      "loss": 0.5471,
+      "step": 3318
+    },
+    {
+      "epoch": 2.623645807568637,
+      "grad_norm": 0.23661072595053065,
+      "learning_rate": 7.536669605457446e-08,
+      "loss": 0.5576,
+      "step": 3319
+    },
+    {
+      "epoch": 2.6244372990353697,
+      "grad_norm": 0.24557479693371542,
+      "learning_rate": 7.505039678795377e-08,
+      "loss": 0.5527,
+      "step": 3320
+    },
+    {
+      "epoch": 2.6252287905021023,
+      "grad_norm": 0.23182584962680358,
+      "learning_rate": 7.473473676142716e-08,
+      "loss": 0.5423,
+      "step": 3321
+    },
+    {
+      "epoch": 2.626020281968835,
+      "grad_norm": 0.2320309639633608,
+      "learning_rate": 7.441971619315024e-08,
+      "loss": 0.5417,
+      "step": 3322
+    },
+    {
+      "epoch": 2.6268117734355676,
+      "grad_norm": 0.22920057909618372,
+      "learning_rate": 7.410533530083618e-08,
+      "loss": 0.5313,
+      "step": 3323
+    },
+    {
+      "epoch": 2.6276032649023002,
+      "grad_norm": 0.24020858834656983,
+      "learning_rate": 7.379159430175596e-08,
+      "loss": 0.5501,
+      "step": 3324
+    },
+    {
+      "epoch": 2.628394756369033,
+      "grad_norm": 0.2339868081826666,
+      "learning_rate": 7.34784934127387e-08,
+      "loss": 0.5492,
+      "step": 3325
+    },
+    {
+      "epoch": 2.6291862478357655,
+      "grad_norm": 0.23996926487304107,
+      "learning_rate": 7.316603285017098e-08,
+      "loss": 0.5329,
+      "step": 3326
+    },
+    {
+      "epoch": 2.629977739302498,
+      "grad_norm": 0.23393305474716156,
+      "learning_rate": 7.28542128299967e-08,
+      "loss": 0.5494,
+      "step": 3327
+    },
+    {
+      "epoch": 2.6307692307692307,
+      "grad_norm": 0.22963241060417952,
+      "learning_rate": 7.254303356771707e-08,
+      "loss": 0.5406,
+      "step": 3328
+    },
+    {
+      "epoch": 2.6315607222359634,
+      "grad_norm": 0.23886740051813526,
+      "learning_rate": 7.223249527839059e-08,
+      "loss": 0.5446,
+      "step": 3329
+    },
+    {
+      "epoch": 2.632352213702696,
+      "grad_norm": 0.22744165041862927,
+      "learning_rate": 7.192259817663304e-08,
+      "loss": 0.545,
+      "step": 3330
+    },
+    {
+      "epoch": 2.6331437051694286,
+      "grad_norm": 0.23670022225669177,
+      "learning_rate": 7.161334247661655e-08,
+      "loss": 0.5391,
+      "step": 3331
+    },
+    {
+      "epoch": 2.6339351966361613,
+      "grad_norm": 0.23971268561491108,
+      "learning_rate": 7.130472839207025e-08,
+      "loss": 0.5492,
+      "step": 3332
+    },
+    {
+      "epoch": 2.634726688102894,
+      "grad_norm": 0.2383035981779266,
+      "learning_rate": 7.099675613627998e-08,
+      "loss": 0.5399,
+      "step": 3333
+    },
+    {
+      "epoch": 2.6355181795696265,
+      "grad_norm": 0.23579330856816882,
+      "learning_rate": 7.068942592208737e-08,
+      "loss": 0.5416,
+      "step": 3334
+    },
+    {
+      "epoch": 2.636309671036359,
+      "grad_norm": 0.23575770594660464,
+      "learning_rate": 7.038273796189142e-08,
+      "loss": 0.5385,
+      "step": 3335
+    },
+    {
+      "epoch": 2.637101162503092,
+      "grad_norm": 0.23548350110621555,
+      "learning_rate": 7.007669246764625e-08,
+      "loss": 0.5537,
+      "step": 3336
+    },
+    {
+      "epoch": 2.6378926539698244,
+      "grad_norm": 0.23844529878142853,
+      "learning_rate": 6.977128965086243e-08,
+      "loss": 0.5469,
+      "step": 3337
+    },
+    {
+      "epoch": 2.638684145436557,
+      "grad_norm": 0.22686586098540332,
+      "learning_rate": 6.946652972260636e-08,
+      "loss": 0.54,
+      "step": 3338
+    },
+    {
+      "epoch": 2.6394756369032897,
+      "grad_norm": 0.23694325529283686,
+      "learning_rate": 6.916241289349988e-08,
+      "loss": 0.5447,
+      "step": 3339
+    },
+    {
+      "epoch": 2.6402671283700223,
+      "grad_norm": 0.2348179002723714,
+      "learning_rate": 6.885893937372089e-08,
+      "loss": 0.5433,
+      "step": 3340
+    },
+    {
+      "epoch": 2.641058619836755,
+      "grad_norm": 0.22958201656947394,
+      "learning_rate": 6.855610937300216e-08,
+      "loss": 0.5482,
+      "step": 3341
+    },
+    {
+      "epoch": 2.6418501113034876,
+      "grad_norm": 0.24174400808676,
+      "learning_rate": 6.825392310063205e-08,
+      "loss": 0.5617,
+      "step": 3342
+    },
+    {
+      "epoch": 2.64264160277022,
+      "grad_norm": 0.23418458670941242,
+      "learning_rate": 6.795238076545384e-08,
+      "loss": 0.54,
+      "step": 3343
+    },
+    {
+      "epoch": 2.643433094236953,
+      "grad_norm": 0.22909574160163376,
+      "learning_rate": 6.765148257586573e-08,
+      "loss": 0.5269,
+      "step": 3344
+    },
+    {
+      "epoch": 2.6442245857036855,
+      "grad_norm": 0.24226787531773467,
+      "learning_rate": 6.735122873982113e-08,
+      "loss": 0.5481,
+      "step": 3345
+    },
+    {
+      "epoch": 2.645016077170418,
+      "grad_norm": 0.2379068130862316,
+      "learning_rate": 6.705161946482773e-08,
+      "loss": 0.5585,
+      "step": 3346
+    },
+    {
+      "epoch": 2.6458075686371507,
+      "grad_norm": 0.23731897910040356,
+      "learning_rate": 6.67526549579479e-08,
+      "loss": 0.5412,
+      "step": 3347
+    },
+    {
+      "epoch": 2.6465990601038833,
+      "grad_norm": 0.23372179787258096,
+      "learning_rate": 6.645433542579848e-08,
+      "loss": 0.539,
+      "step": 3348
+    },
+    {
+      "epoch": 2.647390551570616,
+      "grad_norm": 0.23682260207578243,
+      "learning_rate": 6.615666107455031e-08,
+      "loss": 0.5603,
+      "step": 3349
+    },
+    {
+      "epoch": 2.6481820430373486,
+      "grad_norm": 0.23206983260338943,
+      "learning_rate": 6.585963210992885e-08,
+      "loss": 0.538,
+      "step": 3350
+    },
+    {
+      "epoch": 2.6489735345040812,
+      "grad_norm": 0.23993646439239033,
+      "learning_rate": 6.556324873721297e-08,
+      "loss": 0.5514,
+      "step": 3351
+    },
+    {
+      "epoch": 2.649765025970814,
+      "grad_norm": 0.236293576505746,
+      "learning_rate": 6.526751116123574e-08,
+      "loss": 0.5468,
+      "step": 3352
+    },
+    {
+      "epoch": 2.6505565174375465,
+      "grad_norm": 0.23182344719612016,
+      "learning_rate": 6.497241958638367e-08,
+      "loss": 0.5404,
+      "step": 3353
+    },
+    {
+      "epoch": 2.651348008904279,
+      "grad_norm": 0.2301902403497955,
+      "learning_rate": 6.46779742165967e-08,
+      "loss": 0.5458,
+      "step": 3354
+    },
+    {
+      "epoch": 2.6521395003710118,
+      "grad_norm": 0.2343777032733534,
+      "learning_rate": 6.438417525536876e-08,
+      "loss": 0.5409,
+      "step": 3355
+    },
+    {
+      "epoch": 2.6529309918377444,
+      "grad_norm": 0.23665232079532303,
+      "learning_rate": 6.409102290574653e-08,
+      "loss": 0.5459,
+      "step": 3356
+    },
+    {
+      "epoch": 2.653722483304477,
+      "grad_norm": 0.2376784514350585,
+      "learning_rate": 6.379851737032993e-08,
+      "loss": 0.5381,
+      "step": 3357
+    },
+    {
+      "epoch": 2.6545139747712097,
+      "grad_norm": 0.23866739464754536,
+      "learning_rate": 6.350665885127193e-08,
+      "loss": 0.5529,
+      "step": 3358
+    },
+    {
+      "epoch": 2.6553054662379423,
+      "grad_norm": 0.2293476005110921,
+      "learning_rate": 6.321544755027819e-08,
+      "loss": 0.5401,
+      "step": 3359
+    },
+    {
+      "epoch": 2.656096957704675,
+      "grad_norm": 0.23295232015536255,
+      "learning_rate": 6.292488366860726e-08,
+      "loss": 0.54,
+      "step": 3360
+    },
+    {
+      "epoch": 2.656888449171407,
+      "grad_norm": 0.23398681265126528,
+      "learning_rate": 6.26349674070703e-08,
+      "loss": 0.5423,
+      "step": 3361
+    },
+    {
+      "epoch": 2.6576799406381397,
+      "grad_norm": 0.2425264702879415,
+      "learning_rate": 6.234569896603049e-08,
+      "loss": 0.5454,
+      "step": 3362
+    },
+    {
+      "epoch": 2.6584714321048724,
+      "grad_norm": 0.2312996214940595,
+      "learning_rate": 6.205707854540376e-08,
+      "loss": 0.5454,
+      "step": 3363
+    },
+    {
+      "epoch": 2.659262923571605,
+      "grad_norm": 0.2277981269472317,
+      "learning_rate": 6.176910634465781e-08,
+      "loss": 0.5387,
+      "step": 3364
+    },
+    {
+      "epoch": 2.6600544150383376,
+      "grad_norm": 0.23853364569520374,
+      "learning_rate": 6.148178256281289e-08,
+      "loss": 0.5528,
+      "step": 3365
+    },
+    {
+      "epoch": 2.6608459065050702,
+      "grad_norm": 0.22907449802683474,
+      "learning_rate": 6.119510739844046e-08,
+      "loss": 0.5474,
+      "step": 3366
+    },
+    {
+      "epoch": 2.661637397971803,
+      "grad_norm": 0.2299247984715103,
+      "learning_rate": 6.090908104966397e-08,
+      "loss": 0.5411,
+      "step": 3367
+    },
+    {
+      "epoch": 2.6624288894385355,
+      "grad_norm": 0.227551049767968,
+      "learning_rate": 6.062370371415903e-08,
+      "loss": 0.5402,
+      "step": 3368
+    },
+    {
+      "epoch": 2.663220380905268,
+      "grad_norm": 0.23378723243047844,
+      "learning_rate": 6.033897558915158e-08,
+      "loss": 0.5485,
+      "step": 3369
+    },
+    {
+      "epoch": 2.6640118723720008,
+      "grad_norm": 0.2394556610967616,
+      "learning_rate": 6.005489687141996e-08,
+      "loss": 0.5486,
+      "step": 3370
+    },
+    {
+      "epoch": 2.6648033638387334,
+      "grad_norm": 0.23009070758179592,
+      "learning_rate": 5.977146775729302e-08,
+      "loss": 0.5354,
+      "step": 3371
+    },
+    {
+      "epoch": 2.665594855305466,
+      "grad_norm": 0.2363466914378266,
+      "learning_rate": 5.948868844265076e-08,
+      "loss": 0.5447,
+      "step": 3372
+    },
+    {
+      "epoch": 2.6663863467721987,
+      "grad_norm": 0.23746064750217005,
+      "learning_rate": 5.9206559122924826e-08,
+      "loss": 0.5326,
+      "step": 3373
+    },
+    {
+      "epoch": 2.6671778382389313,
+      "grad_norm": 0.2366549271451058,
+      "learning_rate": 5.892507999309637e-08,
+      "loss": 0.5402,
+      "step": 3374
+    },
+    {
+      "epoch": 2.667969329705664,
+      "grad_norm": 0.2394632783060834,
+      "learning_rate": 5.864425124769823e-08,
+      "loss": 0.5428,
+      "step": 3375
+    },
+    {
+      "epoch": 2.6687608211723965,
+      "grad_norm": 0.2412885040753179,
+      "learning_rate": 5.836407308081337e-08,
+      "loss": 0.5401,
+      "step": 3376
+    },
+    {
+      "epoch": 2.669552312639129,
+      "grad_norm": 0.23311997952084815,
+      "learning_rate": 5.808454568607501e-08,
+      "loss": 0.5518,
+      "step": 3377
+    },
+    {
+      "epoch": 2.670343804105862,
+      "grad_norm": 0.23394786556968425,
+      "learning_rate": 5.780566925666719e-08,
+      "loss": 0.527,
+      "step": 3378
+    },
+    {
+      "epoch": 2.6711352955725944,
+      "grad_norm": 0.2316218940832262,
+      "learning_rate": 5.752744398532317e-08,
+      "loss": 0.5427,
+      "step": 3379
+    },
+    {
+      "epoch": 2.671926787039327,
+      "grad_norm": 0.23739609076706067,
+      "learning_rate": 5.724987006432691e-08,
+      "loss": 0.5413,
+      "step": 3380
+    },
+    {
+      "epoch": 2.6727182785060597,
+      "grad_norm": 0.23654663494691486,
+      "learning_rate": 5.697294768551197e-08,
+      "loss": 0.543,
+      "step": 3381
+    },
+    {
+      "epoch": 2.6735097699727923,
+      "grad_norm": 0.24486486134295715,
+      "learning_rate": 5.6696677040261376e-08,
+      "loss": 0.5555,
+      "step": 3382
+    },
+    {
+      "epoch": 2.674301261439525,
+      "grad_norm": 0.22913015899333763,
+      "learning_rate": 5.6421058319508495e-08,
+      "loss": 0.5444,
+      "step": 3383
+    },
+    {
+      "epoch": 2.6750927529062576,
+      "grad_norm": 0.23127123534788113,
+      "learning_rate": 5.61460917137353e-08,
+      "loss": 0.5307,
+      "step": 3384
+    },
+    {
+      "epoch": 2.67588424437299,
+      "grad_norm": 0.23662149668399493,
+      "learning_rate": 5.587177741297322e-08,
+      "loss": 0.5403,
+      "step": 3385
+    },
+    {
+      "epoch": 2.676675735839723,
+      "grad_norm": 0.22941545147855613,
+      "learning_rate": 5.559811560680361e-08,
+      "loss": 0.538,
+      "step": 3386
+    },
+    {
+      "epoch": 2.6774672273064555,
+      "grad_norm": 0.23557226516061405,
+      "learning_rate": 5.5325106484355735e-08,
+      "loss": 0.549,
+      "step": 3387
+    },
+    {
+      "epoch": 2.678258718773188,
+      "grad_norm": 0.23903798783195193,
+      "learning_rate": 5.5052750234308884e-08,
+      "loss": 0.5384,
+      "step": 3388
+    },
+    {
+      "epoch": 2.6790502102399207,
+      "grad_norm": 0.23390759042618223,
+      "learning_rate": 5.478104704489061e-08,
+      "loss": 0.55,
+      "step": 3389
+    },
+    {
+      "epoch": 2.6798417017066534,
+      "grad_norm": 0.24319314506092046,
+      "learning_rate": 5.4509997103876824e-08,
+      "loss": 0.5478,
+      "step": 3390
+    },
+    {
+      "epoch": 2.680633193173386,
+      "grad_norm": 0.2324985845713776,
+      "learning_rate": 5.423960059859256e-08,
+      "loss": 0.5503,
+      "step": 3391
+    },
+    {
+      "epoch": 2.6814246846401186,
+      "grad_norm": 0.23451652089274386,
+      "learning_rate": 5.396985771591078e-08,
+      "loss": 0.5471,
+      "step": 3392
+    },
+    {
+      "epoch": 2.6822161761068513,
+      "grad_norm": 0.24989306663252006,
+      "learning_rate": 5.3700768642253366e-08,
+      "loss": 0.5356,
+      "step": 3393
+    },
+    {
+      "epoch": 2.683007667573584,
+      "grad_norm": 0.24397093985512394,
+      "learning_rate": 5.343233356358989e-08,
+      "loss": 0.5412,
+      "step": 3394
+    },
+    {
+      "epoch": 2.6837991590403165,
+      "grad_norm": 0.229773232873424,
+      "learning_rate": 5.3164552665437404e-08,
+      "loss": 0.5456,
+      "step": 3395
+    },
+    {
+      "epoch": 2.684590650507049,
+      "grad_norm": 0.23722507770809703,
+      "learning_rate": 5.289742613286208e-08,
+      "loss": 0.5481,
+      "step": 3396
+    },
+    {
+      "epoch": 2.6853821419737818,
+      "grad_norm": 0.2373573136152371,
+      "learning_rate": 5.2630954150476914e-08,
+      "loss": 0.5358,
+      "step": 3397
+    },
+    {
+      "epoch": 2.6861736334405144,
+      "grad_norm": 0.23428134873211282,
+      "learning_rate": 5.236513690244282e-08,
+      "loss": 0.553,
+      "step": 3398
+    },
+    {
+      "epoch": 2.686965124907247,
+      "grad_norm": 0.23310966726833876,
+      "learning_rate": 5.209997457246851e-08,
+      "loss": 0.5243,
+      "step": 3399
+    },
+    {
+      "epoch": 2.6877566163739797,
+      "grad_norm": 0.23758896816669747,
+      "learning_rate": 5.183546734380928e-08,
+      "loss": 0.5468,
+      "step": 3400
+    },
+    {
+      "epoch": 2.6885481078407123,
+      "grad_norm": 0.23864284541641398,
+      "learning_rate": 5.15716153992688e-08,
+      "loss": 0.5437,
+      "step": 3401
+    },
+    {
+      "epoch": 2.689339599307445,
+      "grad_norm": 0.2277969446365102,
+      "learning_rate": 5.130841892119686e-08,
+      "loss": 0.5457,
+      "step": 3402
+    },
+    {
+      "epoch": 2.6901310907741776,
+      "grad_norm": 0.22992842908824299,
+      "learning_rate": 5.1045878091490636e-08,
+      "loss": 0.5465,
+      "step": 3403
+    },
+    {
+      "epoch": 2.69092258224091,
+      "grad_norm": 0.22899571568499077,
+      "learning_rate": 5.0783993091594643e-08,
+      "loss": 0.537,
+      "step": 3404
+    },
+    {
+      "epoch": 2.691714073707643,
+      "grad_norm": 0.2426667299516774,
+      "learning_rate": 5.052276410249912e-08,
+      "loss": 0.5419,
+      "step": 3405
+    },
+    {
+      "epoch": 2.6925055651743754,
+      "grad_norm": 0.23553616708092487,
+      "learning_rate": 5.026219130474185e-08,
+      "loss": 0.5412,
+      "step": 3406
+    },
+    {
+      "epoch": 2.693297056641108,
+      "grad_norm": 0.23398828031835345,
+      "learning_rate": 5.0002274878406804e-08,
+      "loss": 0.5396,
+      "step": 3407
+    },
+    {
+      "epoch": 2.6940885481078407,
+      "grad_norm": 0.23580587669446443,
+      "learning_rate": 4.974301500312417e-08,
+      "loss": 0.5334,
+      "step": 3408
+    },
+    {
+      "epoch": 2.6948800395745733,
+      "grad_norm": 0.24035198235528163,
+      "learning_rate": 4.948441185807062e-08,
+      "loss": 0.5468,
+      "step": 3409
+    },
+    {
+      "epoch": 2.695671531041306,
+      "grad_norm": 0.23806444946182978,
+      "learning_rate": 4.9226465621968974e-08,
+      "loss": 0.55,
+      "step": 3410
+    },
+    {
+      "epoch": 2.6964630225080386,
+      "grad_norm": 0.23194104144441952,
+      "learning_rate": 4.8969176473087737e-08,
+      "loss": 0.5496,
+      "step": 3411
+    },
+    {
+      "epoch": 2.6972545139747712,
+      "grad_norm": 0.23867628766648052,
+      "learning_rate": 4.871254458924157e-08,
+      "loss": 0.5381,
+      "step": 3412
+    },
+    {
+      "epoch": 2.698046005441504,
+      "grad_norm": 0.23339696658980788,
+      "learning_rate": 4.8456570147790696e-08,
+      "loss": 0.5436,
+      "step": 3413
+    },
+    {
+      "epoch": 2.6988374969082365,
+      "grad_norm": 0.2320492748563772,
+      "learning_rate": 4.82012533256414e-08,
+      "loss": 0.552,
+      "step": 3414
+    },
+    {
+      "epoch": 2.699628988374969,
+      "grad_norm": 0.2325736734263487,
+      "learning_rate": 4.794659429924508e-08,
+      "loss": 0.5392,
+      "step": 3415
+    },
+    {
+      "epoch": 2.7004204798417017,
+      "grad_norm": 0.22975884339493674,
+      "learning_rate": 4.769259324459851e-08,
+      "loss": 0.5343,
+      "step": 3416
+    },
+    {
+      "epoch": 2.7012119713084344,
+      "grad_norm": 0.2298298087474154,
+      "learning_rate": 4.743925033724405e-08,
+      "loss": 0.5543,
+      "step": 3417
+    },
+    {
+      "epoch": 2.702003462775167,
+      "grad_norm": 0.2344361780445073,
+      "learning_rate": 4.718656575226865e-08,
+      "loss": 0.5499,
+      "step": 3418
+    },
+    {
+      "epoch": 2.7027949542418996,
+      "grad_norm": 0.24206674411947315,
+      "learning_rate": 4.693453966430505e-08,
+      "loss": 0.5465,
+      "step": 3419
+    },
+    {
+      "epoch": 2.7035864457086323,
+      "grad_norm": 0.22954596310049344,
+      "learning_rate": 4.668317224753049e-08,
+      "loss": 0.5249,
+      "step": 3420
+    },
+    {
+      "epoch": 2.704377937175365,
+      "grad_norm": 0.23727187656397647,
+      "learning_rate": 4.643246367566678e-08,
+      "loss": 0.5427,
+      "step": 3421
+    },
+    {
+      "epoch": 2.7051694286420975,
+      "grad_norm": 0.2363763923476531,
+      "learning_rate": 4.618241412198076e-08,
+      "loss": 0.5473,
+      "step": 3422
+    },
+    {
+      "epoch": 2.70596092010883,
+      "grad_norm": 0.2401557462835347,
+      "learning_rate": 4.593302375928343e-08,
+      "loss": 0.5621,
+      "step": 3423
+    },
+    {
+      "epoch": 2.706752411575563,
+      "grad_norm": 0.23243396201237482,
+      "learning_rate": 4.568429275993091e-08,
+      "loss": 0.5389,
+      "step": 3424
+    },
+    {
+      "epoch": 2.7075439030422954,
+      "grad_norm": 0.23562065303751875,
+      "learning_rate": 4.543622129582303e-08,
+      "loss": 0.5527,
+      "step": 3425
+    },
+    {
+      "epoch": 2.708335394509028,
+      "grad_norm": 0.2419214235285186,
+      "learning_rate": 4.518880953840376e-08,
+      "loss": 0.5419,
+      "step": 3426
+    },
+    {
+      "epoch": 2.7091268859757607,
+      "grad_norm": 0.231034387406549,
+      "learning_rate": 4.4942057658661655e-08,
+      "loss": 0.5427,
+      "step": 3427
+    },
+    {
+      "epoch": 2.7099183774424933,
+      "grad_norm": 0.23365120173504692,
+      "learning_rate": 4.469596582712854e-08,
+      "loss": 0.5377,
+      "step": 3428
+    },
+    {
+      "epoch": 2.710709868909226,
+      "grad_norm": 0.23276674721623392,
+      "learning_rate": 4.4450534213880785e-08,
+      "loss": 0.5447,
+      "step": 3429
+    },
+    {
+      "epoch": 2.7115013603759586,
+      "grad_norm": 0.23521469344731916,
+      "learning_rate": 4.4205762988538175e-08,
+      "loss": 0.536,
+      "step": 3430
+    },
+    {
+      "epoch": 2.712292851842691,
+      "grad_norm": 0.23243297975161775,
+      "learning_rate": 4.3961652320263834e-08,
+      "loss": 0.5462,
+      "step": 3431
+    },
+    {
+      "epoch": 2.713084343309424,
+      "grad_norm": 0.23899991736523044,
+      "learning_rate": 4.371820237776469e-08,
+      "loss": 0.5569,
+      "step": 3432
+    },
+    {
+      "epoch": 2.7138758347761565,
+      "grad_norm": 0.2350694293606461,
+      "learning_rate": 4.347541332929095e-08,
+      "loss": 0.5585,
+      "step": 3433
+    },
+    {
+      "epoch": 2.714667326242889,
+      "grad_norm": 0.2398373071592293,
+      "learning_rate": 4.323328534263615e-08,
+      "loss": 0.5427,
+      "step": 3434
+    },
+    {
+      "epoch": 2.7154588177096217,
+      "grad_norm": 0.22978786890343167,
+      "learning_rate": 4.2991818585136786e-08,
+      "loss": 0.537,
+      "step": 3435
+    },
+    {
+      "epoch": 2.7162503091763543,
+      "grad_norm": 0.23521253337923248,
+      "learning_rate": 4.275101322367258e-08,
+      "loss": 0.54,
+      "step": 3436
+    },
+    {
+      "epoch": 2.717041800643087,
+      "grad_norm": 0.23354481297301627,
+      "learning_rate": 4.2510869424665974e-08,
+      "loss": 0.5434,
+      "step": 3437
+    },
+    {
+      "epoch": 2.7178332921098196,
+      "grad_norm": 0.22686230379711145,
+      "learning_rate": 4.227138735408209e-08,
+      "loss": 0.533,
+      "step": 3438
+    },
+    {
+      "epoch": 2.7186247835765522,
+      "grad_norm": 0.2309054443517431,
+      "learning_rate": 4.203256717742931e-08,
+      "loss": 0.533,
+      "step": 3439
+    },
+    {
+      "epoch": 2.719416275043285,
+      "grad_norm": 0.24156887530710502,
+      "learning_rate": 4.1794409059757954e-08,
+      "loss": 0.5467,
+      "step": 3440
+    },
+    {
+      "epoch": 2.7202077665100175,
+      "grad_norm": 0.23660085736384973,
+      "learning_rate": 4.155691316566101e-08,
+      "loss": 0.54,
+      "step": 3441
+    },
+    {
+      "epoch": 2.72099925797675,
+      "grad_norm": 0.24514687244327169,
+      "learning_rate": 4.1320079659273864e-08,
+      "loss": 0.5469,
+      "step": 3442
+    },
+    {
+      "epoch": 2.7217907494434828,
+      "grad_norm": 0.23847074510715047,
+      "learning_rate": 4.1083908704273805e-08,
+      "loss": 0.5561,
+      "step": 3443
+    },
+    {
+      "epoch": 2.7225822409102154,
+      "grad_norm": 0.2358355223625499,
+      "learning_rate": 4.084840046388083e-08,
+      "loss": 0.5403,
+      "step": 3444
+    },
+    {
+      "epoch": 2.723373732376948,
+      "grad_norm": 0.23573668088691632,
+      "learning_rate": 4.06135551008564e-08,
+      "loss": 0.5549,
+      "step": 3445
+    },
+    {
+      "epoch": 2.7241652238436806,
+      "grad_norm": 0.23490689572976287,
+      "learning_rate": 4.037937277750403e-08,
+      "loss": 0.5453,
+      "step": 3446
+    },
+    {
+      "epoch": 2.7249567153104133,
+      "grad_norm": 0.23070750162727222,
+      "learning_rate": 4.014585365566914e-08,
+      "loss": 0.5508,
+      "step": 3447
+    },
+    {
+      "epoch": 2.725748206777146,
+      "grad_norm": 0.23341170525102728,
+      "learning_rate": 3.99129978967383e-08,
+      "loss": 0.5528,
+      "step": 3448
+    },
+    {
+      "epoch": 2.7265396982438785,
+      "grad_norm": 0.23470833326746446,
+      "learning_rate": 3.9680805661640335e-08,
+      "loss": 0.5148,
+      "step": 3449
+    },
+    {
+      "epoch": 2.727331189710611,
+      "grad_norm": 0.2311784537193434,
+      "learning_rate": 3.944927711084511e-08,
+      "loss": 0.5426,
+      "step": 3450
+    },
+    {
+      "epoch": 2.728122681177344,
+      "grad_norm": 0.2356104303311427,
+      "learning_rate": 3.9218412404363854e-08,
+      "loss": 0.5282,
+      "step": 3451
+    },
+    {
+      "epoch": 2.7289141726440764,
+      "grad_norm": 0.23252944127723466,
+      "learning_rate": 3.898821170174904e-08,
+      "loss": 0.537,
+      "step": 3452
+    },
+    {
+      "epoch": 2.729705664110809,
+      "grad_norm": 0.24246467763086027,
+      "learning_rate": 3.875867516209397e-08,
+      "loss": 0.5523,
+      "step": 3453
+    },
+    {
+      "epoch": 2.7304971555775417,
+      "grad_norm": 0.23288654085502317,
+      "learning_rate": 3.852980294403352e-08,
+      "loss": 0.5458,
+      "step": 3454
+    },
+    {
+      "epoch": 2.7312886470442743,
+      "grad_norm": 0.23268975477803808,
+      "learning_rate": 3.830159520574294e-08,
+      "loss": 0.5381,
+      "step": 3455
+    },
+    {
+      "epoch": 2.732080138511007,
+      "grad_norm": 0.23980609129077946,
+      "learning_rate": 3.807405210493841e-08,
+      "loss": 0.5461,
+      "step": 3456
+    },
+    {
+      "epoch": 2.732871629977739,
+      "grad_norm": 0.23630067177042066,
+      "learning_rate": 3.7847173798877033e-08,
+      "loss": 0.5478,
+      "step": 3457
+    },
+    {
+      "epoch": 2.7336631214444718,
+      "grad_norm": 0.23318535717061248,
+      "learning_rate": 3.762096044435592e-08,
+      "loss": 0.5508,
+      "step": 3458
+    },
+    {
+      "epoch": 2.7344546129112044,
+      "grad_norm": 0.23267507197446813,
+      "learning_rate": 3.739541219771314e-08,
+      "loss": 0.5473,
+      "step": 3459
+    },
+    {
+      "epoch": 2.735246104377937,
+      "grad_norm": 0.2399675008164859,
+      "learning_rate": 3.7170529214826774e-08,
+      "loss": 0.548,
+      "step": 3460
+    },
+    {
+      "epoch": 2.7360375958446697,
+      "grad_norm": 0.22669170257218765,
+      "learning_rate": 3.694631165111528e-08,
+      "loss": 0.5336,
+      "step": 3461
+    },
+    {
+      "epoch": 2.7368290873114023,
+      "grad_norm": 0.23240953576710244,
+      "learning_rate": 3.6722759661537374e-08,
+      "loss": 0.5335,
+      "step": 3462
+    },
+    {
+      "epoch": 2.737620578778135,
+      "grad_norm": 0.23444655359985708,
+      "learning_rate": 3.6499873400591353e-08,
+      "loss": 0.5403,
+      "step": 3463
+    },
+    {
+      "epoch": 2.7384120702448675,
+      "grad_norm": 0.23233076560046437,
+      "learning_rate": 3.627765302231589e-08,
+      "loss": 0.5384,
+      "step": 3464
+    },
+    {
+      "epoch": 2.7392035617116,
+      "grad_norm": 0.2347489961324914,
+      "learning_rate": 3.605609868028925e-08,
+      "loss": 0.5277,
+      "step": 3465
+    },
+    {
+      "epoch": 2.739995053178333,
+      "grad_norm": 0.2299054594238578,
+      "learning_rate": 3.583521052762928e-08,
+      "loss": 0.5346,
+      "step": 3466
+    },
+    {
+      "epoch": 2.7407865446450654,
+      "grad_norm": 0.23102598850353004,
+      "learning_rate": 3.5614988716993975e-08,
+      "loss": 0.5383,
+      "step": 3467
+    },
+    {
+      "epoch": 2.741578036111798,
+      "grad_norm": 0.23021780179591364,
+      "learning_rate": 3.5395433400579823e-08,
+      "loss": 0.5416,
+      "step": 3468
+    },
+    {
+      "epoch": 2.7423695275785307,
+      "grad_norm": 0.2333466076786601,
+      "learning_rate": 3.5176544730123546e-08,
+      "loss": 0.5482,
+      "step": 3469
+    },
+    {
+      "epoch": 2.7431610190452633,
+      "grad_norm": 0.2296607821578593,
+      "learning_rate": 3.4958322856900925e-08,
+      "loss": 0.5385,
+      "step": 3470
+    },
+    {
+      "epoch": 2.743952510511996,
+      "grad_norm": 0.22970187576139794,
+      "learning_rate": 3.4740767931726754e-08,
+      "loss": 0.5465,
+      "step": 3471
+    },
+    {
+      "epoch": 2.7447440019787286,
+      "grad_norm": 0.23140483990915722,
+      "learning_rate": 3.4523880104955105e-08,
+      "loss": 0.5371,
+      "step": 3472
+    },
+    {
+      "epoch": 2.745535493445461,
+      "grad_norm": 0.24235975648798602,
+      "learning_rate": 3.430765952647863e-08,
+      "loss": 0.5448,
+      "step": 3473
+    },
+    {
+      "epoch": 2.746326984912194,
+      "grad_norm": 0.23555635868935854,
+      "learning_rate": 3.409210634572934e-08,
+      "loss": 0.5394,
+      "step": 3474
+    },
+    {
+      "epoch": 2.7471184763789265,
+      "grad_norm": 0.23966553106887273,
+      "learning_rate": 3.387722071167765e-08,
+      "loss": 0.5424,
+      "step": 3475
+    },
+    {
+      "epoch": 2.747909967845659,
+      "grad_norm": 0.24839035325363107,
+      "learning_rate": 3.366300277283285e-08,
+      "loss": 0.5536,
+      "step": 3476
+    },
+    {
+      "epoch": 2.7487014593123917,
+      "grad_norm": 0.2286134790665496,
+      "learning_rate": 3.344945267724264e-08,
+      "loss": 0.5379,
+      "step": 3477
+    },
+    {
+      "epoch": 2.7494929507791244,
+      "grad_norm": 0.2333887450627868,
+      "learning_rate": 3.323657057249318e-08,
+      "loss": 0.5345,
+      "step": 3478
+    },
+    {
+      "epoch": 2.750284442245857,
+      "grad_norm": 0.2304774829009927,
+      "learning_rate": 3.302435660570901e-08,
+      "loss": 0.529,
+      "step": 3479
+    },
+    {
+      "epoch": 2.7510759337125896,
+      "grad_norm": 0.2364774765020302,
+      "learning_rate": 3.281281092355293e-08,
+      "loss": 0.5644,
+      "step": 3480
+    },
+    {
+      "epoch": 2.7518674251793223,
+      "grad_norm": 0.23448978825284153,
+      "learning_rate": 3.2601933672225634e-08,
+      "loss": 0.5481,
+      "step": 3481
+    },
+    {
+      "epoch": 2.752658916646055,
+      "grad_norm": 0.23242910389211896,
+      "learning_rate": 3.239172499746645e-08,
+      "loss": 0.5507,
+      "step": 3482
+    },
+    {
+      "epoch": 2.7534504081127875,
+      "grad_norm": 0.2327606911525051,
+      "learning_rate": 3.2182185044552166e-08,
+      "loss": 0.5408,
+      "step": 3483
+    },
+    {
+      "epoch": 2.75424189957952,
+      "grad_norm": 0.2391192422876664,
+      "learning_rate": 3.1973313958297496e-08,
+      "loss": 0.5402,
+      "step": 3484
+    },
+    {
+      "epoch": 2.7550333910462528,
+      "grad_norm": 0.2604482409561624,
+      "learning_rate": 3.176511188305475e-08,
+      "loss": 0.5471,
+      "step": 3485
+    },
+    {
+      "epoch": 2.7558248825129854,
+      "grad_norm": 0.23540208883453675,
+      "learning_rate": 3.155757896271427e-08,
+      "loss": 0.5567,
+      "step": 3486
+    },
+    {
+      "epoch": 2.756616373979718,
+      "grad_norm": 0.23058161967027863,
+      "learning_rate": 3.1350715340703656e-08,
+      "loss": 0.5371,
+      "step": 3487
+    },
+    {
+      "epoch": 2.7574078654464507,
+      "grad_norm": 0.242007068792223,
+      "learning_rate": 3.1144521159988114e-08,
+      "loss": 0.5471,
+      "step": 3488
+    },
+    {
+      "epoch": 2.7581993569131833,
+      "grad_norm": 0.2354541265211657,
+      "learning_rate": 3.0938996563069865e-08,
+      "loss": 0.5578,
+      "step": 3489
+    },
+    {
+      "epoch": 2.758990848379916,
+      "grad_norm": 0.23229659189411686,
+      "learning_rate": 3.0734141691988845e-08,
+      "loss": 0.5353,
+      "step": 3490
+    },
+    {
+      "epoch": 2.7597823398466486,
+      "grad_norm": 0.2291565180226149,
+      "learning_rate": 3.0529956688321587e-08,
+      "loss": 0.5517,
+      "step": 3491
+    },
+    {
+      "epoch": 2.760573831313381,
+      "grad_norm": 0.23328504089921456,
+      "learning_rate": 3.032644169318244e-08,
+      "loss": 0.5435,
+      "step": 3492
+    },
+    {
+      "epoch": 2.761365322780114,
+      "grad_norm": 0.2377236397176347,
+      "learning_rate": 3.012359684722199e-08,
+      "loss": 0.5384,
+      "step": 3493
+    },
+    {
+      "epoch": 2.7621568142468464,
+      "grad_norm": 0.23127161748830127,
+      "learning_rate": 2.9921422290627773e-08,
+      "loss": 0.5551,
+      "step": 3494
+    },
+    {
+      "epoch": 2.762948305713579,
+      "grad_norm": 0.23847532948448244,
+      "learning_rate": 2.971991816312458e-08,
+      "loss": 0.5488,
+      "step": 3495
+    },
+    {
+      "epoch": 2.7637397971803117,
+      "grad_norm": 0.22552587630689253,
+      "learning_rate": 2.9519084603973344e-08,
+      "loss": 0.5414,
+      "step": 3496
+    },
+    {
+      "epoch": 2.7645312886470443,
+      "grad_norm": 0.23767561808758256,
+      "learning_rate": 2.9318921751971813e-08,
+      "loss": 0.5354,
+      "step": 3497
+    },
+    {
+      "epoch": 2.765322780113777,
+      "grad_norm": 0.23387258295826305,
+      "learning_rate": 2.9119429745454115e-08,
+      "loss": 0.5425,
+      "step": 3498
+    },
+    {
+      "epoch": 2.7661142715805096,
+      "grad_norm": 0.23093274846171483,
+      "learning_rate": 2.892060872229074e-08,
+      "loss": 0.5475,
+      "step": 3499
+    },
+    {
+      "epoch": 2.7669057630472422,
+      "grad_norm": 0.23556135428259767,
+      "learning_rate": 2.872245881988855e-08,
+      "loss": 0.548,
+      "step": 3500
+    },
+    {
+      "epoch": 2.767697254513975,
+      "grad_norm": 0.23018517015742726,
+      "learning_rate": 2.852498017519045e-08,
+      "loss": 0.5444,
+      "step": 3501
+    },
+    {
+      "epoch": 2.7684887459807075,
+      "grad_norm": 0.23840627780612114,
+      "learning_rate": 2.8328172924675596e-08,
+      "loss": 0.5336,
+      "step": 3502
+    },
+    {
+      "epoch": 2.76928023744744,
+      "grad_norm": 0.23468436703558332,
+      "learning_rate": 2.8132037204359082e-08,
+      "loss": 0.5435,
+      "step": 3503
+    },
+    {
+      "epoch": 2.7700717289141727,
+      "grad_norm": 0.23643967159386806,
+      "learning_rate": 2.793657314979192e-08,
+      "loss": 0.5495,
+      "step": 3504
+    },
+    {
+      "epoch": 2.7708632203809054,
+      "grad_norm": 0.23623426572835918,
+      "learning_rate": 2.7741780896060717e-08,
+      "loss": 0.5319,
+      "step": 3505
+    },
+    {
+      "epoch": 2.771654711847638,
+      "grad_norm": 0.23190503062035855,
+      "learning_rate": 2.7547660577788123e-08,
+      "loss": 0.5307,
+      "step": 3506
+    },
+    {
+      "epoch": 2.7724462033143706,
+      "grad_norm": 0.23768799013678277,
+      "learning_rate": 2.7354212329131932e-08,
+      "loss": 0.541,
+      "step": 3507
+    },
+    {
+      "epoch": 2.7732376947811033,
+      "grad_norm": 0.23000439972524941,
+      "learning_rate": 2.716143628378631e-08,
+      "loss": 0.5617,
+      "step": 3508
+    },
+    {
+      "epoch": 2.774029186247836,
+      "grad_norm": 0.23595121150464654,
+      "learning_rate": 2.696933257497991e-08,
+      "loss": 0.5572,
+      "step": 3509
+    },
+    {
+      "epoch": 2.7748206777145685,
+      "grad_norm": 0.23560222839548514,
+      "learning_rate": 2.6777901335477303e-08,
+      "loss": 0.5574,
+      "step": 3510
+    },
+    {
+      "epoch": 2.775612169181301,
+      "grad_norm": 0.23403609430669742,
+      "learning_rate": 2.6587142697578114e-08,
+      "loss": 0.5513,
+      "step": 3511
+    },
+    {
+      "epoch": 2.776403660648034,
+      "grad_norm": 0.22425397228693494,
+      "learning_rate": 2.6397056793117212e-08,
+      "loss": 0.5346,
+      "step": 3512
+    },
+    {
+      "epoch": 2.7771951521147664,
+      "grad_norm": 0.2376508180433101,
+      "learning_rate": 2.6207643753464405e-08,
+      "loss": 0.5391,
+      "step": 3513
+    },
+    {
+      "epoch": 2.777986643581499,
+      "grad_norm": 0.23130205544977794,
+      "learning_rate": 2.6018903709524753e-08,
+      "loss": 0.5409,
+      "step": 3514
+    },
+    {
+      "epoch": 2.7787781350482312,
+      "grad_norm": 0.23409277352689428,
+      "learning_rate": 2.5830836791737808e-08,
+      "loss": 0.5289,
+      "step": 3515
+    },
+    {
+      "epoch": 2.779569626514964,
+      "grad_norm": 0.23395289069440428,
+      "learning_rate": 2.564344313007827e-08,
+      "loss": 0.5559,
+      "step": 3516
+    },
+    {
+      "epoch": 2.7803611179816965,
+      "grad_norm": 0.22712398663529615,
+      "learning_rate": 2.5456722854055203e-08,
+      "loss": 0.5468,
+      "step": 3517
+    },
+    {
+      "epoch": 2.781152609448429,
+      "grad_norm": 0.23634599671534784,
+      "learning_rate": 2.5270676092712717e-08,
+      "loss": 0.5349,
+      "step": 3518
+    },
+    {
+      "epoch": 2.7819441009151618,
+      "grad_norm": 0.2312066239681208,
+      "learning_rate": 2.5085302974629074e-08,
+      "loss": 0.5543,
+      "step": 3519
+    },
+    {
+      "epoch": 2.7827355923818944,
+      "grad_norm": 0.24353576828579954,
+      "learning_rate": 2.4900603627917238e-08,
+      "loss": 0.5347,
+      "step": 3520
+    },
+    {
+      "epoch": 2.783527083848627,
+      "grad_norm": 0.23467857738704168,
+      "learning_rate": 2.4716578180224434e-08,
+      "loss": 0.5488,
+      "step": 3521
+    },
+    {
+      "epoch": 2.7843185753153596,
+      "grad_norm": 0.23502749131964953,
+      "learning_rate": 2.453322675873182e-08,
+      "loss": 0.5477,
+      "step": 3522
+    },
+    {
+      "epoch": 2.7851100667820923,
+      "grad_norm": 0.23564631541599618,
+      "learning_rate": 2.435054949015547e-08,
+      "loss": 0.5348,
+      "step": 3523
+    },
+    {
+      "epoch": 2.785901558248825,
+      "grad_norm": 0.23526267646175178,
+      "learning_rate": 2.4168546500744845e-08,
+      "loss": 0.5409,
+      "step": 3524
+    },
+    {
+      "epoch": 2.7866930497155575,
+      "grad_norm": 0.23494317308400237,
+      "learning_rate": 2.3987217916283662e-08,
+      "loss": 0.5331,
+      "step": 3525
+    },
+    {
+      "epoch": 2.78748454118229,
+      "grad_norm": 0.2542467653741271,
+      "learning_rate": 2.3806563862089678e-08,
+      "loss": 0.5422,
+      "step": 3526
+    },
+    {
+      "epoch": 2.788276032649023,
+      "grad_norm": 0.23253837434076613,
+      "learning_rate": 2.3626584463014242e-08,
+      "loss": 0.5475,
+      "step": 3527
+    },
+    {
+      "epoch": 2.7890675241157554,
+      "grad_norm": 0.23967383421535285,
+      "learning_rate": 2.3447279843442637e-08,
+      "loss": 0.5588,
+      "step": 3528
+    },
+    {
+      "epoch": 2.789859015582488,
+      "grad_norm": 0.23653332543355077,
+      "learning_rate": 2.326865012729373e-08,
+      "loss": 0.5461,
+      "step": 3529
+    },
+    {
+      "epoch": 2.7906505070492207,
+      "grad_norm": 0.23499861737507716,
+      "learning_rate": 2.3090695438020004e-08,
+      "loss": 0.5351,
+      "step": 3530
+    },
+    {
+      "epoch": 2.7914419985159533,
+      "grad_norm": 0.2305686362074388,
+      "learning_rate": 2.2913415898607293e-08,
+      "loss": 0.5366,
+      "step": 3531
+    },
+    {
+      "epoch": 2.792233489982686,
+      "grad_norm": 0.2297406533202431,
+      "learning_rate": 2.2736811631574814e-08,
+      "loss": 0.544,
+      "step": 3532
+    },
+    {
+      "epoch": 2.7930249814494186,
+      "grad_norm": 0.23216831218152165,
+      "learning_rate": 2.2560882758975387e-08,
+      "loss": 0.5571,
+      "step": 3533
+    },
+    {
+      "epoch": 2.793816472916151,
+      "grad_norm": 0.2345578836500067,
+      "learning_rate": 2.2385629402394745e-08,
+      "loss": 0.5452,
+      "step": 3534
+    },
+    {
+      "epoch": 2.794607964382884,
+      "grad_norm": 0.23321595494458838,
+      "learning_rate": 2.221105168295201e-08,
+      "loss": 0.5452,
+      "step": 3535
+    },
+    {
+      "epoch": 2.7953994558496165,
+      "grad_norm": 0.23098645497815196,
+      "learning_rate": 2.2037149721299107e-08,
+      "loss": 0.5294,
+      "step": 3536
+    },
+    {
+      "epoch": 2.796190947316349,
+      "grad_norm": 0.2314132114243511,
+      "learning_rate": 2.1863923637621017e-08,
+      "loss": 0.547,
+      "step": 3537
+    },
+    {
+      "epoch": 2.7969824387830817,
+      "grad_norm": 0.22801882532163015,
+      "learning_rate": 2.169137355163586e-08,
+      "loss": 0.5371,
+      "step": 3538
+    },
+    {
+      "epoch": 2.7977739302498144,
+      "grad_norm": 0.233720369303306,
+      "learning_rate": 2.151949958259436e-08,
+      "loss": 0.5471,
+      "step": 3539
+    },
+    {
+      "epoch": 2.798565421716547,
+      "grad_norm": 0.23423377714684682,
+      "learning_rate": 2.1348301849279825e-08,
+      "loss": 0.5586,
+      "step": 3540
+    },
+    {
+      "epoch": 2.7993569131832796,
+      "grad_norm": 0.23850635298398823,
+      "learning_rate": 2.117778047000873e-08,
+      "loss": 0.546,
+      "step": 3541
+    },
+    {
+      "epoch": 2.8001484046500122,
+      "grad_norm": 0.2316721329812288,
+      "learning_rate": 2.1007935562629364e-08,
+      "loss": 0.5475,
+      "step": 3542
+    },
+    {
+      "epoch": 2.800939896116745,
+      "grad_norm": 0.23374530840302743,
+      "learning_rate": 2.083876724452316e-08,
+      "loss": 0.5538,
+      "step": 3543
+    },
+    {
+      "epoch": 2.8017313875834775,
+      "grad_norm": 0.23499009395978418,
+      "learning_rate": 2.0670275632603705e-08,
+      "loss": 0.5332,
+      "step": 3544
+    },
+    {
+      "epoch": 2.80252287905021,
+      "grad_norm": 0.2401152049814733,
+      "learning_rate": 2.0502460843316638e-08,
+      "loss": 0.5325,
+      "step": 3545
+    },
+    {
+      "epoch": 2.8033143705169428,
+      "grad_norm": 0.2283248348034765,
+      "learning_rate": 2.0335322992640515e-08,
+      "loss": 0.5422,
+      "step": 3546
+    },
+    {
+      "epoch": 2.8041058619836754,
+      "grad_norm": 0.23694216114420952,
+      "learning_rate": 2.0168862196085268e-08,
+      "loss": 0.5555,
+      "step": 3547
+    },
+    {
+      "epoch": 2.804897353450408,
+      "grad_norm": 0.2287060180999378,
+      "learning_rate": 2.0003078568693542e-08,
+      "loss": 0.5304,
+      "step": 3548
+    },
+    {
+      "epoch": 2.8056888449171407,
+      "grad_norm": 0.23742244512905672,
+      "learning_rate": 1.983797222503969e-08,
+      "loss": 0.5409,
+      "step": 3549
+    },
+    {
+      "epoch": 2.8064803363838733,
+      "grad_norm": 0.24173773252938405,
+      "learning_rate": 1.9673543279229766e-08,
+      "loss": 0.5438,
+      "step": 3550
+    },
+    {
+      "epoch": 2.807271827850606,
+      "grad_norm": 0.23258151407290656,
+      "learning_rate": 1.9509791844902425e-08,
+      "loss": 0.5503,
+      "step": 3551
+    },
+    {
+      "epoch": 2.8080633193173385,
+      "grad_norm": 0.23040099587075968,
+      "learning_rate": 1.9346718035227138e-08,
+      "loss": 0.5417,
+      "step": 3552
+    },
+    {
+      "epoch": 2.808854810784071,
+      "grad_norm": 0.23511468624766385,
+      "learning_rate": 1.918432196290576e-08,
+      "loss": 0.533,
+      "step": 3553
+    },
+    {
+      "epoch": 2.809646302250804,
+      "grad_norm": 0.23219468499841275,
+      "learning_rate": 1.9022603740171506e-08,
+      "loss": 0.5421,
+      "step": 3554
+    },
+    {
+      "epoch": 2.8104377937175364,
+      "grad_norm": 0.23301189976458736,
+      "learning_rate": 1.8861563478788977e-08,
+      "loss": 0.5305,
+      "step": 3555
+    },
+    {
+      "epoch": 2.811229285184269,
+      "grad_norm": 0.23340596839688357,
+      "learning_rate": 1.87012012900547e-08,
+      "loss": 0.5381,
+      "step": 3556
+    },
+    {
+      "epoch": 2.8120207766510017,
+      "grad_norm": 0.23783691954664868,
+      "learning_rate": 1.8541517284795916e-08,
+      "loss": 0.5458,
+      "step": 3557
+    },
+    {
+      "epoch": 2.8128122681177343,
+      "grad_norm": 0.2282605799176566,
+      "learning_rate": 1.8382511573371785e-08,
+      "loss": 0.556,
+      "step": 3558
+    },
+    {
+      "epoch": 2.813603759584467,
+      "grad_norm": 0.23753014772684272,
+      "learning_rate": 1.8224184265672295e-08,
+      "loss": 0.5404,
+      "step": 3559
+    },
+    {
+      "epoch": 2.8143952510511996,
+      "grad_norm": 0.23031513978317342,
+      "learning_rate": 1.8066535471118693e-08,
+      "loss": 0.5415,
+      "step": 3560
+    },
+    {
+      "epoch": 2.815186742517932,
+      "grad_norm": 0.23667671620117772,
+      "learning_rate": 1.79095652986635e-08,
+      "loss": 0.5353,
+      "step": 3561
+    },
+    {
+      "epoch": 2.815978233984665,
+      "grad_norm": 0.23722248155222286,
+      "learning_rate": 1.775327385678993e-08,
+      "loss": 0.5547,
+      "step": 3562
+    },
+    {
+      "epoch": 2.8167697254513975,
+      "grad_norm": 0.23049729316446924,
+      "learning_rate": 1.7597661253512363e-08,
+      "loss": 0.5403,
+      "step": 3563
+    },
+    {
+      "epoch": 2.81756121691813,
+      "grad_norm": 0.22592288027231294,
+      "learning_rate": 1.744272759637566e-08,
+      "loss": 0.5304,
+      "step": 3564
+    },
+    {
+      "epoch": 2.8183527083848627,
+      "grad_norm": 0.32681943286543735,
+      "learning_rate": 1.7288472992455948e-08,
+      "loss": 0.5392,
+      "step": 3565
+    },
+    {
+      "epoch": 2.8191441998515954,
+      "grad_norm": 0.23437810226232614,
+      "learning_rate": 1.713489754835984e-08,
+      "loss": 0.5307,
+      "step": 3566
+    },
+    {
+      "epoch": 2.819935691318328,
+      "grad_norm": 0.24144825637446976,
+      "learning_rate": 1.698200137022443e-08,
+      "loss": 0.5509,
+      "step": 3567
+    },
+    {
+      "epoch": 2.8207271827850606,
+      "grad_norm": 0.2313386721507604,
+      "learning_rate": 1.6829784563717532e-08,
+      "loss": 0.5278,
+      "step": 3568
+    },
+    {
+      "epoch": 2.8215186742517933,
+      "grad_norm": 0.23995629981802358,
+      "learning_rate": 1.6678247234037435e-08,
+      "loss": 0.5416,
+      "step": 3569
+    },
+    {
+      "epoch": 2.822310165718526,
+      "grad_norm": 0.23027681319850457,
+      "learning_rate": 1.6527389485912813e-08,
+      "loss": 0.5315,
+      "step": 3570
+    },
+    {
+      "epoch": 2.8231016571852585,
+      "grad_norm": 0.24034734751644593,
+      "learning_rate": 1.63772114236026e-08,
+      "loss": 0.5506,
+      "step": 3571
+    },
+    {
+      "epoch": 2.823893148651991,
+      "grad_norm": 0.23562823586904702,
+      "learning_rate": 1.6227713150896106e-08,
+      "loss": 0.5496,
+      "step": 3572
+    },
+    {
+      "epoch": 2.8246846401187238,
+      "grad_norm": 0.23697594867551047,
+      "learning_rate": 1.6078894771112796e-08,
+      "loss": 0.5433,
+      "step": 3573
+    },
+    {
+      "epoch": 2.8254761315854564,
+      "grad_norm": 0.23202930551790968,
+      "learning_rate": 1.5930756387102174e-08,
+      "loss": 0.5446,
+      "step": 3574
+    },
+    {
+      "epoch": 2.826267623052189,
+      "grad_norm": 0.23437474430100744,
+      "learning_rate": 1.57832981012439e-08,
+      "loss": 0.5516,
+      "step": 3575
+    },
+    {
+      "epoch": 2.8270591145189217,
+      "grad_norm": 0.22678494836197086,
+      "learning_rate": 1.5636520015447686e-08,
+      "loss": 0.5349,
+      "step": 3576
+    },
+    {
+      "epoch": 2.8278506059856543,
+      "grad_norm": 0.23084951474263085,
+      "learning_rate": 1.5490422231152933e-08,
+      "loss": 0.5527,
+      "step": 3577
+    },
+    {
+      "epoch": 2.828642097452387,
+      "grad_norm": 0.2415577245295724,
+      "learning_rate": 1.534500484932899e-08,
+      "loss": 0.5289,
+      "step": 3578
+    },
+    {
+      "epoch": 2.8294335889191196,
+      "grad_norm": 0.2322913620283281,
+      "learning_rate": 1.5200267970475022e-08,
+      "loss": 0.5262,
+      "step": 3579
+    },
+    {
+      "epoch": 2.830225080385852,
+      "grad_norm": 0.22879348301674715,
+      "learning_rate": 1.505621169461979e-08,
+      "loss": 0.5439,
+      "step": 3580
+    },
+    {
+      "epoch": 2.831016571852585,
+      "grad_norm": 0.23081602018863082,
+      "learning_rate": 1.4912836121321882e-08,
+      "loss": 0.5217,
+      "step": 3581
+    },
+    {
+      "epoch": 2.8318080633193174,
+      "grad_norm": 0.23953489128778746,
+      "learning_rate": 1.4770141349669141e-08,
+      "loss": 0.5644,
+      "step": 3582
+    },
+    {
+      "epoch": 2.83259955478605,
+      "grad_norm": 0.23594273303660432,
+      "learning_rate": 1.462812747827924e-08,
+      "loss": 0.5422,
+      "step": 3583
+    },
+    {
+      "epoch": 2.8333910462527827,
+      "grad_norm": 0.23107491008936532,
+      "learning_rate": 1.4486794605298891e-08,
+      "loss": 0.5421,
+      "step": 3584
+    },
+    {
+      "epoch": 2.8341825377195153,
+      "grad_norm": 0.23846821821463982,
+      "learning_rate": 1.4346142828404407e-08,
+      "loss": 0.5416,
+      "step": 3585
+    },
+    {
+      "epoch": 2.834974029186248,
+      "grad_norm": 0.22986602265930947,
+      "learning_rate": 1.4206172244801585e-08,
+      "loss": 0.5347,
+      "step": 3586
+    },
+    {
+      "epoch": 2.8357655206529806,
+      "grad_norm": 0.2296823559896836,
+      "learning_rate": 1.4066882951225046e-08,
+      "loss": 0.5466,
+      "step": 3587
+    },
+    {
+      "epoch": 2.8365570121197132,
+      "grad_norm": 0.2301570053150726,
+      "learning_rate": 1.3928275043938786e-08,
+      "loss": 0.5564,
+      "step": 3588
+    },
+    {
+      "epoch": 2.837348503586446,
+      "grad_norm": 0.2296514803800265,
+      "learning_rate": 1.3790348618735848e-08,
+      "loss": 0.5325,
+      "step": 3589
+    },
+    {
+      "epoch": 2.8381399950531785,
+      "grad_norm": 0.22901689070204145,
+      "learning_rate": 1.3653103770938091e-08,
+      "loss": 0.5418,
+      "step": 3590
+    },
+    {
+      "epoch": 2.838931486519911,
+      "grad_norm": 0.23699501479345256,
+      "learning_rate": 1.3516540595396864e-08,
+      "loss": 0.5492,
+      "step": 3591
+    },
+    {
+      "epoch": 2.8397229779866437,
+      "grad_norm": 0.23808576468403075,
+      "learning_rate": 1.3380659186491894e-08,
+      "loss": 0.5423,
+      "step": 3592
+    },
+    {
+      "epoch": 2.8405144694533764,
+      "grad_norm": 0.23491662688542603,
+      "learning_rate": 1.324545963813195e-08,
+      "loss": 0.5513,
+      "step": 3593
+    },
+    {
+      "epoch": 2.841305960920109,
+      "grad_norm": 0.22997445500928412,
+      "learning_rate": 1.3110942043754513e-08,
+      "loss": 0.5275,
+      "step": 3594
+    },
+    {
+      "epoch": 2.8420974523868416,
+      "grad_norm": 0.23387300453408702,
+      "learning_rate": 1.2977106496325662e-08,
+      "loss": 0.5463,
+      "step": 3595
+    },
+    {
+      "epoch": 2.8428889438535743,
+      "grad_norm": 0.2336068330709784,
+      "learning_rate": 1.2843953088340409e-08,
+      "loss": 0.5452,
+      "step": 3596
+    },
+    {
+      "epoch": 2.843680435320307,
+      "grad_norm": 0.23790914023782178,
+      "learning_rate": 1.271148191182203e-08,
+      "loss": 0.5532,
+      "step": 3597
+    },
+    {
+      "epoch": 2.8444719267870395,
+      "grad_norm": 0.2393653332050634,
+      "learning_rate": 1.2579693058322515e-08,
+      "loss": 0.5384,
+      "step": 3598
+    },
+    {
+      "epoch": 2.845263418253772,
+      "grad_norm": 0.2348288395855586,
+      "learning_rate": 1.2448586618922119e-08,
+      "loss": 0.5328,
+      "step": 3599
+    },
+    {
+      "epoch": 2.846054909720505,
+      "grad_norm": 0.27919785487916793,
+      "learning_rate": 1.2318162684229471e-08,
+      "loss": 0.5369,
+      "step": 3600
+    },
+    {
+      "epoch": 2.8468464011872374,
+      "grad_norm": 0.2278029862920757,
+      "learning_rate": 1.2188421344381805e-08,
+      "loss": 0.5624,
+      "step": 3601
+    },
+    {
+      "epoch": 2.84763789265397,
+      "grad_norm": 0.23543883959177841,
+      "learning_rate": 1.2059362689044172e-08,
+      "loss": 0.5435,
+      "step": 3602
+    },
+    {
+      "epoch": 2.8484293841207027,
+      "grad_norm": 0.2326574508053554,
+      "learning_rate": 1.1930986807410337e-08,
+      "loss": 0.5397,
+      "step": 3603
+    },
+    {
+      "epoch": 2.8492208755874353,
+      "grad_norm": 0.2345907080681373,
+      "learning_rate": 1.1803293788201662e-08,
+      "loss": 0.5377,
+      "step": 3604
+    },
+    {
+      "epoch": 2.850012367054168,
+      "grad_norm": 0.23047322101162068,
+      "learning_rate": 1.1676283719667778e-08,
+      "loss": 0.5431,
+      "step": 3605
+    },
+    {
+      "epoch": 2.8508038585209006,
+      "grad_norm": 0.2337950098843311,
+      "learning_rate": 1.1549956689586582e-08,
+      "loss": 0.531,
+      "step": 3606
+    },
+    {
+      "epoch": 2.851595349987633,
+      "grad_norm": 0.23696401034968725,
+      "learning_rate": 1.1424312785263568e-08,
+      "loss": 0.541,
+      "step": 3607
+    },
+    {
+      "epoch": 2.852386841454366,
+      "grad_norm": 0.23381401620065134,
+      "learning_rate": 1.1299352093532166e-08,
+      "loss": 0.5456,
+      "step": 3608
+    },
+    {
+      "epoch": 2.8531783329210985,
+      "grad_norm": 0.2328559920772804,
+      "learning_rate": 1.1175074700753962e-08,
+      "loss": 0.5426,
+      "step": 3609
+    },
+    {
+      "epoch": 2.853969824387831,
+      "grad_norm": 0.24369332847811104,
+      "learning_rate": 1.1051480692817804e-08,
+      "loss": 0.54,
+      "step": 3610
+    },
+    {
+      "epoch": 2.8547613158545637,
+      "grad_norm": 0.23404016134285346,
+      "learning_rate": 1.0928570155140593e-08,
+      "loss": 0.5428,
+      "step": 3611
+    },
+    {
+      "epoch": 2.855552807321296,
+      "grad_norm": 0.23350951551608884,
+      "learning_rate": 1.0806343172666931e-08,
+      "loss": 0.5437,
+      "step": 3612
+    },
+    {
+      "epoch": 2.8563442987880285,
+      "grad_norm": 0.23588925655037235,
+      "learning_rate": 1.0684799829868585e-08,
+      "loss": 0.543,
+      "step": 3613
+    },
+    {
+      "epoch": 2.857135790254761,
+      "grad_norm": 0.2349753036718382,
+      "learning_rate": 1.056394021074547e-08,
+      "loss": 0.542,
+      "step": 3614
+    },
+    {
+      "epoch": 2.857927281721494,
+      "grad_norm": 0.23506426952632153,
+      "learning_rate": 1.0443764398824328e-08,
+      "loss": 0.5463,
+      "step": 3615
+    },
+    {
+      "epoch": 2.8587187731882264,
+      "grad_norm": 0.2344535064572982,
+      "learning_rate": 1.0324272477159834e-08,
+      "loss": 0.5452,
+      "step": 3616
+    },
+    {
+      "epoch": 2.859510264654959,
+      "grad_norm": 0.22766830797611373,
+      "learning_rate": 1.0205464528333596e-08,
+      "loss": 0.5568,
+      "step": 3617
+    },
+    {
+      "epoch": 2.8603017561216917,
+      "grad_norm": 0.2327774942747983,
+      "learning_rate": 1.0087340634454934e-08,
+      "loss": 0.5516,
+      "step": 3618
+    },
+    {
+      "epoch": 2.8610932475884243,
+      "grad_norm": 0.23033824478521248,
+      "learning_rate": 9.96990087716032e-09,
+      "loss": 0.5462,
+      "step": 3619
+    },
+    {
+      "epoch": 2.861884739055157,
+      "grad_norm": 0.23527791127801123,
+      "learning_rate": 9.853145337613056e-09,
+      "loss": 0.5577,
+      "step": 3620
+    },
+    {
+      "epoch": 2.8626762305218896,
+      "grad_norm": 0.23043840008569028,
+      "learning_rate": 9.737074096503928e-09,
+      "loss": 0.5471,
+      "step": 3621
+    },
+    {
+      "epoch": 2.863467721988622,
+      "grad_norm": 0.23488737054786746,
+      "learning_rate": 9.621687234050658e-09,
+      "loss": 0.5405,
+      "step": 3622
+    },
+    {
+      "epoch": 2.864259213455355,
+      "grad_norm": 0.2324017817548189,
+      "learning_rate": 9.50698482999812e-09,
+      "loss": 0.552,
+      "step": 3623
+    },
+    {
+      "epoch": 2.8650507049220875,
+      "grad_norm": 0.23832437701855722,
+      "learning_rate": 9.392966963618131e-09,
+      "loss": 0.5381,
+      "step": 3624
+    },
+    {
+      "epoch": 2.86584219638882,
+      "grad_norm": 0.2356875120153845,
+      "learning_rate": 9.279633713709212e-09,
+      "loss": 0.5425,
+      "step": 3625
+    },
+    {
+      "epoch": 2.8666336878555527,
+      "grad_norm": 0.23554714723924744,
+      "learning_rate": 9.166985158596818e-09,
+      "loss": 0.5449,
+      "step": 3626
+    },
+    {
+      "epoch": 2.8674251793222854,
+      "grad_norm": 0.22964814443532963,
+      "learning_rate": 9.055021376133453e-09,
+      "loss": 0.5374,
+      "step": 3627
+    },
+    {
+      "epoch": 2.868216670789018,
+      "grad_norm": 0.2360445105076418,
+      "learning_rate": 8.943742443697999e-09,
+      "loss": 0.5478,
+      "step": 3628
+    },
+    {
+      "epoch": 2.8690081622557506,
+      "grad_norm": 0.23249939484838325,
+      "learning_rate": 8.833148438196492e-09,
+      "loss": 0.541,
+      "step": 3629
+    },
+    {
+      "epoch": 2.8697996537224832,
+      "grad_norm": 0.23685114299265833,
+      "learning_rate": 8.723239436061126e-09,
+      "loss": 0.5451,
+      "step": 3630
+    },
+    {
+      "epoch": 2.870591145189216,
+      "grad_norm": 0.23567758136093175,
+      "learning_rate": 8.614015513250916e-09,
+      "loss": 0.5385,
+      "step": 3631
+    },
+    {
+      "epoch": 2.8713826366559485,
+      "grad_norm": 0.2314050478292498,
+      "learning_rate": 8.505476745251483e-09,
+      "loss": 0.5362,
+      "step": 3632
+    },
+    {
+      "epoch": 2.872174128122681,
+      "grad_norm": 0.23804147753747956,
+      "learning_rate": 8.397623207074711e-09,
+      "loss": 0.5508,
+      "step": 3633
+    },
+    {
+      "epoch": 2.8729656195894138,
+      "grad_norm": 0.23344629198932168,
+      "learning_rate": 8.290454973259087e-09,
+      "loss": 0.5586,
+      "step": 3634
+    },
+    {
+      "epoch": 2.8737571110561464,
+      "grad_norm": 0.24088819059057526,
+      "learning_rate": 8.183972117869475e-09,
+      "loss": 0.5441,
+      "step": 3635
+    },
+    {
+      "epoch": 2.874548602522879,
+      "grad_norm": 0.23362904142228713,
+      "learning_rate": 8.078174714497011e-09,
+      "loss": 0.5552,
+      "step": 3636
+    },
+    {
+      "epoch": 2.8753400939896117,
+      "grad_norm": 0.23195150006361642,
+      "learning_rate": 7.973062836259204e-09,
+      "loss": 0.5503,
+      "step": 3637
+    },
+    {
+      "epoch": 2.8761315854563443,
+      "grad_norm": 0.23281250790763872,
+      "learning_rate": 7.868636555799613e-09,
+      "loss": 0.5333,
+      "step": 3638
+    },
+    {
+      "epoch": 2.876923076923077,
+      "grad_norm": 0.23484417361113952,
+      "learning_rate": 7.764895945288064e-09,
+      "loss": 0.5471,
+      "step": 3639
+    },
+    {
+      "epoch": 2.8777145683898095,
+      "grad_norm": 0.2403679884476522,
+      "learning_rate": 7.66184107642065e-09,
+      "loss": 0.5277,
+      "step": 3640
+    },
+    {
+      "epoch": 2.878506059856542,
+      "grad_norm": 0.23633848873242225,
+      "learning_rate": 7.559472020419177e-09,
+      "loss": 0.5258,
+      "step": 3641
+    },
+    {
+      "epoch": 2.879297551323275,
+      "grad_norm": 0.23270832961159474,
+      "learning_rate": 7.457788848032054e-09,
+      "loss": 0.5356,
+      "step": 3642
+    },
+    {
+      "epoch": 2.8800890427900074,
+      "grad_norm": 0.235574059142225,
+      "learning_rate": 7.356791629533177e-09,
+      "loss": 0.5427,
+      "step": 3643
+    },
+    {
+      "epoch": 2.88088053425674,
+      "grad_norm": 0.2330963450443202,
+      "learning_rate": 7.2564804347223785e-09,
+      "loss": 0.5468,
+      "step": 3644
+    },
+    {
+      "epoch": 2.8816720257234727,
+      "grad_norm": 0.22829126129192814,
+      "learning_rate": 7.156855332925871e-09,
+      "loss": 0.5267,
+      "step": 3645
+    },
+    {
+      "epoch": 2.8824635171902053,
+      "grad_norm": 0.23199124614006095,
+      "learning_rate": 7.057916392995023e-09,
+      "loss": 0.5509,
+      "step": 3646
+    },
+    {
+      "epoch": 2.883255008656938,
+      "grad_norm": 0.23393639588493617,
+      "learning_rate": 6.9596636833075815e-09,
+      "loss": 0.5386,
+      "step": 3647
+    },
+    {
+      "epoch": 2.8840465001236706,
+      "grad_norm": 0.22419508574534858,
+      "learning_rate": 6.862097271766565e-09,
+      "loss": 0.5226,
+      "step": 3648
+    },
+    {
+      "epoch": 2.884837991590403,
+      "grad_norm": 0.2345720986651518,
+      "learning_rate": 6.765217225801145e-09,
+      "loss": 0.5378,
+      "step": 3649
+    },
+    {
+      "epoch": 2.885629483057136,
+      "grad_norm": 0.23287378541532291,
+      "learning_rate": 6.669023612365654e-09,
+      "loss": 0.5386,
+      "step": 3650
+    },
+    {
+      "epoch": 2.8864209745238685,
+      "grad_norm": 0.22496024331992462,
+      "learning_rate": 6.573516497940468e-09,
+      "loss": 0.5294,
+      "step": 3651
+    },
+    {
+      "epoch": 2.887212465990601,
+      "grad_norm": 0.23452122179342824,
+      "learning_rate": 6.478695948531232e-09,
+      "loss": 0.5335,
+      "step": 3652
+    },
+    {
+      "epoch": 2.8880039574573337,
+      "grad_norm": 0.22952083605403764,
+      "learning_rate": 6.384562029669193e-09,
+      "loss": 0.5346,
+      "step": 3653
+    },
+    {
+      "epoch": 2.8887954489240664,
+      "grad_norm": 0.2417314657178486,
+      "learning_rate": 6.29111480641098e-09,
+      "loss": 0.5588,
+      "step": 3654
+    },
+    {
+      "epoch": 2.889586940390799,
+      "grad_norm": 0.22907102186316622,
+      "learning_rate": 6.1983543433389295e-09,
+      "loss": 0.5357,
+      "step": 3655
+    },
+    {
+      "epoch": 2.8903784318575316,
+      "grad_norm": 0.2383233633695022,
+      "learning_rate": 6.106280704560318e-09,
+      "loss": 0.5432,
+      "step": 3656
+    },
+    {
+      "epoch": 2.8911699233242643,
+      "grad_norm": 0.23278374702529836,
+      "learning_rate": 6.014893953708134e-09,
+      "loss": 0.5557,
+      "step": 3657
+    },
+    {
+      "epoch": 2.891961414790997,
+      "grad_norm": 0.2362970080721708,
+      "learning_rate": 5.924194153940409e-09,
+      "loss": 0.5347,
+      "step": 3658
+    },
+    {
+      "epoch": 2.8927529062577295,
+      "grad_norm": 0.24076146219857836,
+      "learning_rate": 5.834181367940449e-09,
+      "loss": 0.5596,
+      "step": 3659
+    },
+    {
+      "epoch": 2.893544397724462,
+      "grad_norm": 0.2357068807951035,
+      "learning_rate": 5.744855657916936e-09,
+      "loss": 0.5504,
+      "step": 3660
+    },
+    {
+      "epoch": 2.8943358891911948,
+      "grad_norm": 0.2318661379144214,
+      "learning_rate": 5.656217085603598e-09,
+      "loss": 0.526,
+      "step": 3661
+    },
+    {
+      "epoch": 2.8951273806579274,
+      "grad_norm": 0.23135737972044004,
+      "learning_rate": 5.568265712259212e-09,
+      "loss": 0.5474,
+      "step": 3662
+    },
+    {
+      "epoch": 2.89591887212466,
+      "grad_norm": 0.2312055681180558,
+      "learning_rate": 5.481001598667601e-09,
+      "loss": 0.5405,
+      "step": 3663
+    },
+    {
+      "epoch": 2.8967103635913927,
+      "grad_norm": 0.22954336397695554,
+      "learning_rate": 5.394424805137743e-09,
+      "loss": 0.5408,
+      "step": 3664
+    },
+    {
+      "epoch": 2.8975018550581253,
+      "grad_norm": 0.2350687889999863,
+      "learning_rate": 5.3085353915035545e-09,
+      "loss": 0.5464,
+      "step": 3665
+    },
+    {
+      "epoch": 2.898293346524858,
+      "grad_norm": 0.22990848568273484,
+      "learning_rate": 5.223333417123999e-09,
+      "loss": 0.5457,
+      "step": 3666
+    },
+    {
+      "epoch": 2.8990848379915906,
+      "grad_norm": 0.2343502009069639,
+      "learning_rate": 5.13881894088275e-09,
+      "loss": 0.5481,
+      "step": 3667
+    },
+    {
+      "epoch": 2.899876329458323,
+      "grad_norm": 0.23345598474950643,
+      "learning_rate": 5.054992021188531e-09,
+      "loss": 0.5409,
+      "step": 3668
+    },
+    {
+      "epoch": 2.900667820925056,
+      "grad_norm": 0.23391372354504558,
+      "learning_rate": 4.971852715974556e-09,
+      "loss": 0.5615,
+      "step": 3669
+    },
+    {
+      "epoch": 2.901459312391788,
+      "grad_norm": 0.2333856026236808,
+      "learning_rate": 4.889401082699307e-09,
+      "loss": 0.5362,
+      "step": 3670
+    },
+    {
+      "epoch": 2.9022508038585206,
+      "grad_norm": 0.23740898652369022,
+      "learning_rate": 4.807637178345758e-09,
+      "loss": 0.5394,
+      "step": 3671
+    },
+    {
+      "epoch": 2.9030422953252533,
+      "grad_norm": 0.2351692934039835,
+      "learning_rate": 4.726561059421485e-09,
+      "loss": 0.5341,
+      "step": 3672
+    },
+    {
+      "epoch": 2.903833786791986,
+      "grad_norm": 0.23123807687856532,
+      "learning_rate": 4.64617278195889e-09,
+      "loss": 0.5528,
+      "step": 3673
+    },
+    {
+      "epoch": 2.9046252782587185,
+      "grad_norm": 0.23253440822820867,
+      "learning_rate": 4.566472401514976e-09,
+      "loss": 0.5415,
+      "step": 3674
+    },
+    {
+      "epoch": 2.905416769725451,
+      "grad_norm": 0.22763153955250545,
+      "learning_rate": 4.487459973171237e-09,
+      "loss": 0.5333,
+      "step": 3675
+    },
+    {
+      "epoch": 2.906208261192184,
+      "grad_norm": 0.23320624013952176,
+      "learning_rate": 4.409135551533993e-09,
+      "loss": 0.5524,
+      "step": 3676
+    },
+    {
+      "epoch": 2.9069997526589164,
+      "grad_norm": 0.23190345421383513,
+      "learning_rate": 4.331499190733723e-09,
+      "loss": 0.5374,
+      "step": 3677
+    },
+    {
+      "epoch": 2.907791244125649,
+      "grad_norm": 0.2383379936333827,
+      "learning_rate": 4.254550944425506e-09,
+      "loss": 0.55,
+      "step": 3678
+    },
+    {
+      "epoch": 2.9085827355923817,
+      "grad_norm": 0.22785390609910963,
+      "learning_rate": 4.178290865788914e-09,
+      "loss": 0.5402,
+      "step": 3679
+    },
+    {
+      "epoch": 2.9093742270591143,
+      "grad_norm": 0.2352625918619613,
+      "learning_rate": 4.1027190075280105e-09,
+      "loss": 0.5354,
+      "step": 3680
+    },
+    {
+      "epoch": 2.910165718525847,
+      "grad_norm": 0.23786703181770358,
+      "learning_rate": 4.027835421871018e-09,
+      "loss": 0.5428,
+      "step": 3681
+    },
+    {
+      "epoch": 2.9109572099925796,
+      "grad_norm": 0.23571169879494544,
+      "learning_rate": 3.953640160570648e-09,
+      "loss": 0.5354,
+      "step": 3682
+    },
+    {
+      "epoch": 2.911748701459312,
+      "grad_norm": 0.23769944504185458,
+      "learning_rate": 3.880133274903774e-09,
+      "loss": 0.5445,
+      "step": 3683
+    },
+    {
+      "epoch": 2.912540192926045,
+      "grad_norm": 0.23227835874811428,
+      "learning_rate": 3.8073148156715364e-09,
+      "loss": 0.5461,
+      "step": 3684
+    },
+    {
+      "epoch": 2.9133316843927775,
+      "grad_norm": 0.2355513808693547,
+      "learning_rate": 3.7351848331995674e-09,
+      "loss": 0.547,
+      "step": 3685
+    },
+    {
+      "epoch": 2.91412317585951,
+      "grad_norm": 0.23317426937749436,
+      "learning_rate": 3.663743377337325e-09,
+      "loss": 0.5467,
+      "step": 3686
+    },
+    {
+      "epoch": 2.9149146673262427,
+      "grad_norm": 0.23798269387358634,
+      "learning_rate": 3.592990497458537e-09,
+      "loss": 0.5407,
+      "step": 3687
+    },
+    {
+      "epoch": 2.9157061587929753,
+      "grad_norm": 0.2382784295768859,
+      "learning_rate": 3.5229262424611996e-09,
+      "loss": 0.5576,
+      "step": 3688
+    },
+    {
+      "epoch": 2.916497650259708,
+      "grad_norm": 0.23215209734361164,
+      "learning_rate": 3.453550660767135e-09,
+      "loss": 0.5409,
+      "step": 3689
+    },
+    {
+      "epoch": 2.9172891417264406,
+      "grad_norm": 0.23637942810673981,
+      "learning_rate": 3.384863800322546e-09,
+      "loss": 0.5255,
+      "step": 3690
+    },
+    {
+      "epoch": 2.9180806331931732,
+      "grad_norm": 0.23261767139674736,
+      "learning_rate": 3.3168657085974603e-09,
+      "loss": 0.5293,
+      "step": 3691
+    },
+    {
+      "epoch": 2.918872124659906,
+      "grad_norm": 0.23205218319756035,
+      "learning_rate": 3.2495564325857316e-09,
+      "loss": 0.5178,
+      "step": 3692
+    },
+    {
+      "epoch": 2.9196636161266385,
+      "grad_norm": 0.23363088780947577,
+      "learning_rate": 3.182936018805482e-09,
+      "loss": 0.5384,
+      "step": 3693
+    },
+    {
+      "epoch": 2.920455107593371,
+      "grad_norm": 0.23404640760239762,
+      "learning_rate": 3.117004513298438e-09,
+      "loss": 0.5354,
+      "step": 3694
+    },
+    {
+      "epoch": 2.9212465990601038,
+      "grad_norm": 0.22895226717897585,
+      "learning_rate": 3.0517619616307054e-09,
+      "loss": 0.5225,
+      "step": 3695
+    },
+    {
+      "epoch": 2.9220380905268364,
+      "grad_norm": 0.2360703490332348,
+      "learning_rate": 2.9872084088916617e-09,
+      "loss": 0.5417,
+      "step": 3696
+    },
+    {
+      "epoch": 2.922829581993569,
+      "grad_norm": 0.2333998971287613,
+      "learning_rate": 2.923343899694841e-09,
+      "loss": 0.5461,
+      "step": 3697
+    },
+    {
+      "epoch": 2.9236210734603016,
+      "grad_norm": 0.2378209196119549,
+      "learning_rate": 2.8601684781777156e-09,
+      "loss": 0.5459,
+      "step": 3698
+    },
+    {
+      "epoch": 2.9244125649270343,
+      "grad_norm": 0.23861138065538962,
+      "learning_rate": 2.797682188001027e-09,
+      "loss": 0.5447,
+      "step": 3699
+    },
+    {
+      "epoch": 2.925204056393767,
+      "grad_norm": 0.23404748186940524,
+      "learning_rate": 2.735885072349675e-09,
+      "loss": 0.5349,
+      "step": 3700
+    },
+    {
+      "epoch": 2.9259955478604995,
+      "grad_norm": 0.2364313283670403,
+      "learning_rate": 2.6747771739322745e-09,
+      "loss": 0.5438,
+      "step": 3701
+    },
+    {
+      "epoch": 2.926787039327232,
+      "grad_norm": 0.25078310404459,
+      "learning_rate": 2.6143585349808204e-09,
+      "loss": 0.5492,
+      "step": 3702
+    },
+    {
+      "epoch": 2.927578530793965,
+      "grad_norm": 0.2335519225837004,
+      "learning_rate": 2.5546291972511345e-09,
+      "loss": 0.5228,
+      "step": 3703
+    },
+    {
+      "epoch": 2.9283700222606974,
+      "grad_norm": 0.22623403321301705,
+      "learning_rate": 2.495589202022752e-09,
+      "loss": 0.5385,
+      "step": 3704
+    },
+    {
+      "epoch": 2.92916151372743,
+      "grad_norm": 0.2377080853338766,
+      "learning_rate": 2.4372385900984782e-09,
+      "loss": 0.5338,
+      "step": 3705
+    },
+    {
+      "epoch": 2.9299530051941627,
+      "grad_norm": 0.23171608354222004,
+      "learning_rate": 2.3795774018051663e-09,
+      "loss": 0.5405,
+      "step": 3706
+    },
+    {
+      "epoch": 2.9307444966608953,
+      "grad_norm": 0.23272173466659182,
+      "learning_rate": 2.322605676992717e-09,
+      "loss": 0.5422,
+      "step": 3707
+    },
+    {
+      "epoch": 2.931535988127628,
+      "grad_norm": 0.2304871167327762,
+      "learning_rate": 2.2663234550347464e-09,
+      "loss": 0.5464,
+      "step": 3708
+    },
+    {
+      "epoch": 2.9323274795943606,
+      "grad_norm": 0.24568605879888988,
+      "learning_rate": 2.210730774828473e-09,
+      "loss": 0.5369,
+      "step": 3709
+    },
+    {
+      "epoch": 2.933118971061093,
+      "grad_norm": 0.2297037081702141,
+      "learning_rate": 2.155827674794386e-09,
+      "loss": 0.5379,
+      "step": 3710
+    },
+    {
+      "epoch": 2.933910462527826,
+      "grad_norm": 0.2402732800167233,
+      "learning_rate": 2.1016141928765772e-09,
+      "loss": 0.5373,
+      "step": 3711
+    },
+    {
+      "epoch": 2.9347019539945585,
+      "grad_norm": 0.22853359580876342,
+      "learning_rate": 2.048090366542077e-09,
+      "loss": 0.5483,
+      "step": 3712
+    },
+    {
+      "epoch": 2.935493445461291,
+      "grad_norm": 0.23225064386070493,
+      "learning_rate": 1.995256232781961e-09,
+      "loss": 0.5471,
+      "step": 3713
+    },
+    {
+      "epoch": 2.9362849369280237,
+      "grad_norm": 0.23148387471125012,
+      "learning_rate": 1.9431118281102443e-09,
+      "loss": 0.5357,
+      "step": 3714
+    },
+    {
+      "epoch": 2.9370764283947564,
+      "grad_norm": 0.23470038523800482,
+      "learning_rate": 1.891657188564322e-09,
+      "loss": 0.5525,
+      "step": 3715
+    },
+    {
+      "epoch": 2.937867919861489,
+      "grad_norm": 0.23264448419154526,
+      "learning_rate": 1.840892349704859e-09,
+      "loss": 0.5541,
+      "step": 3716
+    },
+    {
+      "epoch": 2.9386594113282216,
+      "grad_norm": 0.23209477039544965,
+      "learning_rate": 1.7908173466159026e-09,
+      "loss": 0.5307,
+      "step": 3717
+    },
+    {
+      "epoch": 2.9394509027949542,
+      "grad_norm": 0.22584066617976098,
+      "learning_rate": 1.741432213904659e-09,
+      "loss": 0.5395,
+      "step": 3718
+    },
+    {
+      "epoch": 2.940242394261687,
+      "grad_norm": 0.23676557988764327,
+      "learning_rate": 1.6927369857017149e-09,
+      "loss": 0.5453,
+      "step": 3719
+    },
+    {
+      "epoch": 2.9410338857284195,
+      "grad_norm": 0.23213111535032288,
+      "learning_rate": 1.644731695660484e-09,
+      "loss": 0.5364,
+      "step": 3720
+    },
+    {
+      "epoch": 2.941825377195152,
+      "grad_norm": 0.22968614395011067,
+      "learning_rate": 1.597416376957983e-09,
+      "loss": 0.5361,
+      "step": 3721
+    },
+    {
+      "epoch": 2.9426168686618848,
+      "grad_norm": 0.2382832272179857,
+      "learning_rate": 1.5507910622941655e-09,
+      "loss": 0.5554,
+      "step": 3722
+    },
+    {
+      "epoch": 2.9434083601286174,
+      "grad_norm": 0.23169949406280965,
+      "learning_rate": 1.5048557838921449e-09,
+      "loss": 0.533,
+      "step": 3723
+    },
+    {
+      "epoch": 2.94419985159535,
+      "grad_norm": 0.23665168804830689,
+      "learning_rate": 1.4596105734983043e-09,
+      "loss": 0.5272,
+      "step": 3724
+    },
+    {
+      "epoch": 2.9449913430620827,
+      "grad_norm": 0.23398965155126505,
+      "learning_rate": 1.4150554623817424e-09,
+      "loss": 0.5429,
+      "step": 3725
+    },
+    {
+      "epoch": 2.9457828345288153,
+      "grad_norm": 0.2399756658001373,
+      "learning_rate": 1.3711904813349384e-09,
+      "loss": 0.5411,
+      "step": 3726
+    },
+    {
+      "epoch": 2.946574325995548,
+      "grad_norm": 0.2409703242003032,
+      "learning_rate": 1.3280156606734206e-09,
+      "loss": 0.5551,
+      "step": 3727
+    },
+    {
+      "epoch": 2.9473658174622805,
+      "grad_norm": 0.23291663729754578,
+      "learning_rate": 1.2855310302355426e-09,
+      "loss": 0.5469,
+      "step": 3728
+    },
+    {
+      "epoch": 2.948157308929013,
+      "grad_norm": 0.23607926421418604,
+      "learning_rate": 1.2437366193828181e-09,
+      "loss": 0.545,
+      "step": 3729
+    },
+    {
+      "epoch": 2.948948800395746,
+      "grad_norm": 0.23271538265326955,
+      "learning_rate": 1.202632456999808e-09,
+      "loss": 0.5339,
+      "step": 3730
+    },
+    {
+      "epoch": 2.9497402918624784,
+      "grad_norm": 0.23455239752017618,
+      "learning_rate": 1.1622185714936783e-09,
+      "loss": 0.5434,
+      "step": 3731
+    },
+    {
+      "epoch": 2.950531783329211,
+      "grad_norm": 0.23122451040910008,
+      "learning_rate": 1.1224949907949754e-09,
+      "loss": 0.5476,
+      "step": 3732
+    },
+    {
+      "epoch": 2.9513232747959437,
+      "grad_norm": 0.2314625171421439,
+      "learning_rate": 1.0834617423569614e-09,
+      "loss": 0.5538,
+      "step": 3733
+    },
+    {
+      "epoch": 2.9521147662626763,
+      "grad_norm": 0.23516302668753056,
+      "learning_rate": 1.0451188531558353e-09,
+      "loss": 0.5371,
+      "step": 3734
+    },
+    {
+      "epoch": 2.952906257729409,
+      "grad_norm": 0.2344195709587897,
+      "learning_rate": 1.0074663496906222e-09,
+      "loss": 0.5282,
+      "step": 3735
+    },
+    {
+      "epoch": 2.9536977491961416,
+      "grad_norm": 0.22724554654759332,
+      "learning_rate": 9.705042579832845e-10,
+      "loss": 0.5472,
+      "step": 3736
+    },
+    {
+      "epoch": 2.954489240662874,
+      "grad_norm": 0.2299260794859835,
+      "learning_rate": 9.342326035787218e-10,
+      "loss": 0.5331,
+      "step": 3737
+    },
+    {
+      "epoch": 2.955280732129607,
+      "grad_norm": 0.23031166598159403,
+      "learning_rate": 8.986514115443266e-10,
+      "loss": 0.5343,
+      "step": 3738
+    },
+    {
+      "epoch": 2.9560722235963395,
+      "grad_norm": 0.23191019069221036,
+      "learning_rate": 8.63760706470873e-10,
+      "loss": 0.547,
+      "step": 3739
+    },
+    {
+      "epoch": 2.956863715063072,
+      "grad_norm": 0.2312162938804386,
+      "learning_rate": 8.295605124712946e-10,
+      "loss": 0.5384,
+      "step": 3740
+    },
+    {
+      "epoch": 2.9576552065298047,
+      "grad_norm": 0.22724776984982867,
+      "learning_rate": 7.960508531817955e-10,
+      "loss": 0.5631,
+      "step": 3741
+    },
+    {
+      "epoch": 2.9584466979965374,
+      "grad_norm": 0.2303256186878083,
+      "learning_rate": 7.632317517610731e-10,
+      "loss": 0.5357,
+      "step": 3742
+    },
+    {
+      "epoch": 2.95923818946327,
+      "grad_norm": 0.23279163047722276,
+      "learning_rate": 7.311032308908727e-10,
+      "loss": 0.5399,
+      "step": 3743
+    },
+    {
+      "epoch": 2.9600296809300026,
+      "grad_norm": 0.23305250004275926,
+      "learning_rate": 6.996653127752106e-10,
+      "loss": 0.5415,
+      "step": 3744
+    },
+    {
+      "epoch": 2.9608211723967353,
+      "grad_norm": 0.23375977755189584,
+      "learning_rate": 6.689180191412624e-10,
+      "loss": 0.5478,
+      "step": 3745
+    },
+    {
+      "epoch": 2.961612663863468,
+      "grad_norm": 0.24026216062993833,
+      "learning_rate": 6.388613712385859e-10,
+      "loss": 0.5385,
+      "step": 3746
+    },
+    {
+      "epoch": 2.9624041553302005,
+      "grad_norm": 0.23268894387393466,
+      "learning_rate": 6.09495389839787e-10,
+      "loss": 0.5351,
+      "step": 3747
+    },
+    {
+      "epoch": 2.963195646796933,
+      "grad_norm": 0.2289542926917948,
+      "learning_rate": 5.808200952396314e-10,
+      "loss": 0.5335,
+      "step": 3748
+    },
+    {
+      "epoch": 2.9639871382636658,
+      "grad_norm": 0.23523839503162927,
+      "learning_rate": 5.528355072561552e-10,
+      "loss": 0.5464,
+      "step": 3749
+    },
+    {
+      "epoch": 2.9647786297303984,
+      "grad_norm": 0.22840618491477813,
+      "learning_rate": 5.255416452295547e-10,
+      "loss": 0.5502,
+      "step": 3750
+    },
+    {
+      "epoch": 2.965570121197131,
+      "grad_norm": 0.23445275168731747,
+      "learning_rate": 4.98938528022852e-10,
+      "loss": 0.5404,
+      "step": 3751
+    },
+    {
+      "epoch": 2.9663616126638637,
+      "grad_norm": 0.23208368631000384,
+      "learning_rate": 4.730261740217845e-10,
+      "loss": 0.5462,
+      "step": 3752
+    },
+    {
+      "epoch": 2.9671531041305963,
+      "grad_norm": 0.2374890121648241,
+      "learning_rate": 4.4780460113436056e-10,
+      "loss": 0.5404,
+      "step": 3753
+    },
+    {
+      "epoch": 2.967944595597329,
+      "grad_norm": 0.2386164219573942,
+      "learning_rate": 4.232738267916369e-10,
+      "loss": 0.536,
+      "step": 3754
+    },
+    {
+      "epoch": 2.9687360870640616,
+      "grad_norm": 0.24677232879508854,
+      "learning_rate": 3.9943386794694114e-10,
+      "loss": 0.5504,
+      "step": 3755
+    },
+    {
+      "epoch": 2.969527578530794,
+      "grad_norm": 0.2323060441295835,
+      "learning_rate": 3.762847410762049e-10,
+      "loss": 0.5517,
+      "step": 3756
+    },
+    {
+      "epoch": 2.970319069997527,
+      "grad_norm": 0.23257010792329827,
+      "learning_rate": 3.5382646217807513e-10,
+      "loss": 0.5475,
+      "step": 3757
+    },
+    {
+      "epoch": 2.9711105614642594,
+      "grad_norm": 0.2397127420431173,
+      "learning_rate": 3.3205904677358067e-10,
+      "loss": 0.5544,
+      "step": 3758
+    },
+    {
+      "epoch": 2.971902052930992,
+      "grad_norm": 0.2300831983679334,
+      "learning_rate": 3.109825099064656e-10,
+      "loss": 0.5289,
+      "step": 3759
+    },
+    {
+      "epoch": 2.9726935443977247,
+      "grad_norm": 0.2366547519309783,
+      "learning_rate": 2.905968661427449e-10,
+      "loss": 0.5452,
+      "step": 3760
+    },
+    {
+      "epoch": 2.9734850358644573,
+      "grad_norm": 0.23545772111242555,
+      "learning_rate": 2.709021295711489e-10,
+      "loss": 0.543,
+      "step": 3761
+    },
+    {
+      "epoch": 2.97427652733119,
+      "grad_norm": 0.23449394354914055,
+      "learning_rate": 2.518983138029007e-10,
+      "loss": 0.5454,
+      "step": 3762
+    },
+    {
+      "epoch": 2.9750680187979226,
+      "grad_norm": 0.23740749952846354,
+      "learning_rate": 2.3358543197171675e-10,
+      "loss": 0.5446,
+      "step": 3763
+    },
+    {
+      "epoch": 2.9758595102646552,
+      "grad_norm": 0.22915338657332054,
+      "learning_rate": 2.159634967336954e-10,
+      "loss": 0.5366,
+      "step": 3764
+    },
+    {
+      "epoch": 2.976651001731388,
+      "grad_norm": 0.23119515539409277,
+      "learning_rate": 1.9903252026753913e-10,
+      "loss": 0.5416,
+      "step": 3765
+    },
+    {
+      "epoch": 2.97744249319812,
+      "grad_norm": 0.23040765132410965,
+      "learning_rate": 1.827925142743325e-10,
+      "loss": 0.5487,
+      "step": 3766
+    },
+    {
+      "epoch": 2.9782339846648527,
+      "grad_norm": 0.23374006345544013,
+      "learning_rate": 1.6724348997776416e-10,
+      "loss": 0.5476,
+      "step": 3767
+    },
+    {
+      "epoch": 2.9790254761315853,
+      "grad_norm": 0.23645004763469699,
+      "learning_rate": 1.5238545812390478e-10,
+      "loss": 0.5391,
+      "step": 3768
+    },
+    {
+      "epoch": 2.979816967598318,
+      "grad_norm": 0.23381810614972207,
+      "learning_rate": 1.3821842898109614e-10,
+      "loss": 0.5394,
+      "step": 3769
+    },
+    {
+      "epoch": 2.9806084590650506,
+      "grad_norm": 0.24314135577603835,
+      "learning_rate": 1.2474241234039506e-10,
+      "loss": 0.5394,
+      "step": 3770
+    },
+    {
+      "epoch": 2.981399950531783,
+      "grad_norm": 0.22682123727042564,
+      "learning_rate": 1.1195741751524046e-10,
+      "loss": 0.5147,
+      "step": 3771
+    },
+    {
+      "epoch": 2.982191441998516,
+      "grad_norm": 0.2333728467115532,
+      "learning_rate": 9.986345334134227e-11,
+      "loss": 0.5564,
+      "step": 3772
+    },
+    {
+      "epoch": 2.9829829334652485,
+      "grad_norm": 0.2369145629920544,
+      "learning_rate": 8.846052817701455e-11,
+      "loss": 0.5519,
+      "step": 3773
+    },
+    {
+      "epoch": 2.983774424931981,
+      "grad_norm": 0.23765521298710948,
+      "learning_rate": 7.774864990284235e-11,
+      "loss": 0.5294,
+      "step": 3774
+    },
+    {
+      "epoch": 2.9845659163987137,
+      "grad_norm": 0.24207765137101736,
+      "learning_rate": 6.772782592201487e-11,
+      "loss": 0.5433,
+      "step": 3775
+    },
+    {
+      "epoch": 2.9853574078654463,
+      "grad_norm": 0.23325791923099654,
+      "learning_rate": 5.839806315977025e-11,
+      "loss": 0.5381,
+      "step": 3776
+    },
+    {
+      "epoch": 2.986148899332179,
+      "grad_norm": 0.23140694777700227,
+      "learning_rate": 4.9759368064283826e-11,
+      "loss": 0.5334,
+      "step": 3777
+    },
+    {
+      "epoch": 2.9869403907989116,
+      "grad_norm": 0.2316247147131433,
+      "learning_rate": 4.181174660555786e-11,
+      "loss": 0.5349,
+      "step": 3778
+    },
+    {
+      "epoch": 2.9877318822656442,
+      "grad_norm": 0.23038555873196936,
+      "learning_rate": 3.4555204276309756e-11,
+      "loss": 0.5313,
+      "step": 3779
+    },
+    {
+      "epoch": 2.988523373732377,
+      "grad_norm": 0.2323028047014364,
+      "learning_rate": 2.7989746091749976e-11,
+      "loss": 0.5479,
+      "step": 3780
+    },
+    {
+      "epoch": 2.9893148651991095,
+      "grad_norm": 0.2351028460363793,
+      "learning_rate": 2.2115376589249e-11,
+      "loss": 0.536,
+      "step": 3781
+    },
+    {
+      "epoch": 2.990106356665842,
+      "grad_norm": 0.23625805916792003,
+      "learning_rate": 1.6932099828448344e-11,
+      "loss": 0.5518,
+      "step": 3782
+    },
+    {
+      "epoch": 2.9908978481325748,
+      "grad_norm": 0.27357256852199685,
+      "learning_rate": 1.2439919391815657e-11,
+      "loss": 0.5439,
+      "step": 3783
+    },
+    {
+      "epoch": 2.9916893395993074,
+      "grad_norm": 0.22866341813556185,
+      "learning_rate": 8.638838383867586e-12,
+      "loss": 0.547,
+      "step": 3784
+    },
+    {
+      "epoch": 2.99248083106604,
+      "grad_norm": 0.25991542219861574,
+      "learning_rate": 5.528859431391808e-12,
+      "loss": 0.5617,
+      "step": 3785
+    },
+    {
+      "epoch": 2.9932723225327726,
+      "grad_norm": 0.2268350128674517,
+      "learning_rate": 3.1099846840021427e-12,
+      "loss": 0.5443,
+      "step": 3786
+    },
+    {
+      "epoch": 2.9940638139995053,
+      "grad_norm": 0.23165237079234247,
+      "learning_rate": 1.3822158131393536e-12,
+      "loss": 0.5305,
+      "step": 3787
+    },
+    {
+      "epoch": 2.994855305466238,
+      "grad_norm": 0.2496839263118096,
+      "learning_rate": 3.4555401295932593e-13,
+      "loss": 0.5444,
+      "step": 3788
+    },
+    {
+      "epoch": 2.9956467969329705,
+      "grad_norm": 0.2291287178701386,
+      "learning_rate": 0.0,
+      "loss": 0.5383,
+      "step": 3789
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3789,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1263,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.6928365495713792e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}