{ "best_metric": null, "best_model_checkpoint": null, "epoch": 11.178537208559566, "global_step": 35000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.03, "learning_rate": 2.8744809964867454e-07, "loss": 91.176, "step": 100 }, { "epoch": 0.06, "learning_rate": 6.068348770360907e-07, "loss": 64.0684, "step": 200 }, { "epoch": 0.1, "learning_rate": 9.26221654423507e-07, "loss": 27.1061, "step": 300 }, { "epoch": 0.13, "learning_rate": 1.242414564037049e-06, "loss": 8.0198, "step": 400 }, { "epoch": 0.16, "learning_rate": 1.561801341424465e-06, "loss": 4.7691, "step": 500 }, { "epoch": 0.19, "learning_rate": 1.8811881188118813e-06, "loss": 3.4387, "step": 600 }, { "epoch": 0.22, "learning_rate": 2.2005748961992977e-06, "loss": 2.7027, "step": 700 }, { "epoch": 0.26, "learning_rate": 2.5199616735867137e-06, "loss": 2.4431, "step": 800 }, { "epoch": 0.29, "learning_rate": 2.8393484509741297e-06, "loss": 2.0534, "step": 900 }, { "epoch": 0.32, "learning_rate": 3.1587352283615465e-06, "loss": 1.8537, "step": 1000 }, { "epoch": 0.35, "learning_rate": 3.4781220057489625e-06, "loss": 1.642, "step": 1100 }, { "epoch": 0.38, "learning_rate": 3.797508783136378e-06, "loss": 1.5092, "step": 1200 }, { "epoch": 0.42, "learning_rate": 4.1168955605237945e-06, "loss": 1.4302, "step": 1300 }, { "epoch": 0.45, "learning_rate": 4.436282337911211e-06, "loss": 1.3527, "step": 1400 }, { "epoch": 0.48, "learning_rate": 4.755669115298627e-06, "loss": 1.2495, "step": 1500 }, { "epoch": 0.51, "learning_rate": 5.075055892686043e-06, "loss": 1.2039, "step": 1600 }, { "epoch": 0.54, "learning_rate": 5.39444267007346e-06, "loss": 1.1438, "step": 1700 }, { "epoch": 0.57, "learning_rate": 5.713829447460876e-06, "loss": 1.1056, "step": 1800 }, { "epoch": 0.61, "learning_rate": 6.033216224848291e-06, "loss": 1.0436, "step": 1900 }, { "epoch": 0.64, "learning_rate": 6.352603002235708e-06, "loss": 1.0225, "step": 2000 }, { "epoch": 0.67, "learning_rate": 6.671989779623123e-06, "loss": 0.9856, "step": 2100 }, { "epoch": 0.7, "learning_rate": 6.9913765570105406e-06, "loss": 0.9412, "step": 2200 }, { "epoch": 0.73, "learning_rate": 7.310763334397957e-06, "loss": 0.9075, "step": 2300 }, { "epoch": 0.77, "learning_rate": 7.630150111785373e-06, "loss": 0.8752, "step": 2400 }, { "epoch": 0.8, "learning_rate": 7.949536889172788e-06, "loss": 0.854, "step": 2500 }, { "epoch": 0.83, "learning_rate": 8.268923666560205e-06, "loss": 0.8282, "step": 2600 }, { "epoch": 0.86, "learning_rate": 8.588310443947621e-06, "loss": 0.8074, "step": 2700 }, { "epoch": 0.89, "learning_rate": 8.907697221335038e-06, "loss": 0.7976, "step": 2800 }, { "epoch": 0.93, "learning_rate": 9.227083998722454e-06, "loss": 0.7955, "step": 2900 }, { "epoch": 0.96, "learning_rate": 9.54647077610987e-06, "loss": 0.7502, "step": 3000 }, { "epoch": 0.99, "learning_rate": 9.865857553497285e-06, "loss": 0.7455, "step": 3100 }, { "epoch": 1.02, "learning_rate": 1.0185244330884702e-05, "loss": 0.7177, "step": 3200 }, { "epoch": 1.05, "learning_rate": 1.050463110827212e-05, "loss": 0.724, "step": 3300 }, { "epoch": 1.09, "learning_rate": 1.0824017885659535e-05, "loss": 0.6975, "step": 3400 }, { "epoch": 1.12, "learning_rate": 1.114340466304695e-05, "loss": 0.7054, "step": 3500 }, { "epoch": 1.15, "learning_rate": 1.1462791440434366e-05, "loss": 0.71, "step": 3600 }, { "epoch": 1.18, "learning_rate": 1.1782178217821782e-05, "loss": 0.6825, "step": 3700 }, { "epoch": 1.21, "learning_rate": 1.2101564995209199e-05, "loss": 0.6792, "step": 3800 }, { "epoch": 1.25, "learning_rate": 1.2420951772596615e-05, "loss": 0.6794, "step": 3900 }, { "epoch": 1.28, "learning_rate": 1.2740338549984032e-05, "loss": 0.6672, "step": 4000 }, { "epoch": 1.31, "learning_rate": 1.3059725327371449e-05, "loss": 0.6356, "step": 4100 }, { "epoch": 1.34, "learning_rate": 1.3379112104758865e-05, "loss": 0.6503, "step": 4200 }, { "epoch": 1.37, "learning_rate": 1.369849888214628e-05, "loss": 0.6495, "step": 4300 }, { "epoch": 1.41, "learning_rate": 1.4017885659533696e-05, "loss": 0.6428, "step": 4400 }, { "epoch": 1.44, "learning_rate": 1.4337272436921113e-05, "loss": 0.6171, "step": 4500 }, { "epoch": 1.47, "learning_rate": 1.4656659214308529e-05, "loss": 0.6084, "step": 4600 }, { "epoch": 1.5, "learning_rate": 1.4976045991695944e-05, "loss": 0.6222, "step": 4700 }, { "epoch": 1.53, "learning_rate": 1.529543276908336e-05, "loss": 0.5971, "step": 4800 }, { "epoch": 1.56, "learning_rate": 1.5614819546470777e-05, "loss": 0.5745, "step": 4900 }, { "epoch": 1.6, "learning_rate": 1.5934206323858196e-05, "loss": 0.5753, "step": 5000 }, { "epoch": 1.63, "learning_rate": 1.625359310124561e-05, "loss": 0.6045, "step": 5100 }, { "epoch": 1.66, "learning_rate": 1.6572979878633027e-05, "loss": 0.6006, "step": 5200 }, { "epoch": 1.69, "learning_rate": 1.6892366656020443e-05, "loss": 0.6003, "step": 5300 }, { "epoch": 1.72, "learning_rate": 1.7211753433407858e-05, "loss": 0.5688, "step": 5400 }, { "epoch": 1.76, "learning_rate": 1.7531140210795274e-05, "loss": 0.5662, "step": 5500 }, { "epoch": 1.79, "learning_rate": 1.785052698818269e-05, "loss": 0.5788, "step": 5600 }, { "epoch": 1.82, "learning_rate": 1.8169913765570105e-05, "loss": 0.5541, "step": 5700 }, { "epoch": 1.85, "learning_rate": 1.8489300542957524e-05, "loss": 0.5656, "step": 5800 }, { "epoch": 1.88, "learning_rate": 1.880868732034494e-05, "loss": 0.5466, "step": 5900 }, { "epoch": 1.92, "learning_rate": 1.9128074097732355e-05, "loss": 0.563, "step": 6000 }, { "epoch": 1.95, "learning_rate": 1.9447460875119774e-05, "loss": 0.5603, "step": 6100 }, { "epoch": 1.98, "learning_rate": 1.976684765250719e-05, "loss": 0.5579, "step": 6200 }, { "epoch": 2.01, "learning_rate": 2.0086234429894605e-05, "loss": 0.5346, "step": 6300 }, { "epoch": 2.04, "learning_rate": 2.0405621207282017e-05, "loss": 0.5038, "step": 6400 }, { "epoch": 2.08, "learning_rate": 2.0725007984669436e-05, "loss": 0.5121, "step": 6500 }, { "epoch": 2.11, "learning_rate": 2.1044394762056855e-05, "loss": 0.5135, "step": 6600 }, { "epoch": 2.14, "learning_rate": 2.1363781539444267e-05, "loss": 0.504, "step": 6700 }, { "epoch": 2.17, "learning_rate": 2.1683168316831686e-05, "loss": 0.5119, "step": 6800 }, { "epoch": 2.2, "learning_rate": 2.20025550942191e-05, "loss": 0.4986, "step": 6900 }, { "epoch": 2.24, "learning_rate": 2.2321941871606518e-05, "loss": 0.504, "step": 7000 }, { "epoch": 2.27, "learning_rate": 2.2641328648993933e-05, "loss": 0.5024, "step": 7100 }, { "epoch": 2.3, "learning_rate": 2.296071542638135e-05, "loss": 0.479, "step": 7200 }, { "epoch": 2.33, "learning_rate": 2.3280102203768764e-05, "loss": 0.492, "step": 7300 }, { "epoch": 2.36, "learning_rate": 2.3599488981156183e-05, "loss": 0.4928, "step": 7400 }, { "epoch": 2.4, "learning_rate": 2.39188757585436e-05, "loss": 0.4862, "step": 7500 }, { "epoch": 2.43, "learning_rate": 2.4238262535931014e-05, "loss": 0.5148, "step": 7600 }, { "epoch": 2.46, "learning_rate": 2.4557649313318433e-05, "loss": 0.4922, "step": 7700 }, { "epoch": 2.49, "learning_rate": 2.4877036090705846e-05, "loss": 0.4882, "step": 7800 }, { "epoch": 2.52, "learning_rate": 2.5196422868093264e-05, "loss": 0.4871, "step": 7900 }, { "epoch": 2.56, "learning_rate": 2.5515809645480677e-05, "loss": 0.4967, "step": 8000 }, { "epoch": 2.59, "learning_rate": 2.5835196422868096e-05, "loss": 0.4834, "step": 8100 }, { "epoch": 2.62, "learning_rate": 2.615458320025551e-05, "loss": 0.4769, "step": 8200 }, { "epoch": 2.65, "learning_rate": 2.6473969977642927e-05, "loss": 0.4776, "step": 8300 }, { "epoch": 2.68, "learning_rate": 2.6793356755030342e-05, "loss": 0.4861, "step": 8400 }, { "epoch": 2.71, "learning_rate": 2.711274353241776e-05, "loss": 0.486, "step": 8500 }, { "epoch": 2.75, "learning_rate": 2.7432130309805177e-05, "loss": 0.4745, "step": 8600 }, { "epoch": 2.78, "learning_rate": 2.7751517087192592e-05, "loss": 0.4805, "step": 8700 }, { "epoch": 2.81, "learning_rate": 2.807090386458001e-05, "loss": 0.5003, "step": 8800 }, { "epoch": 2.84, "learning_rate": 2.8390290641967424e-05, "loss": 0.4639, "step": 8900 }, { "epoch": 2.87, "learning_rate": 2.8709677419354843e-05, "loss": 0.4763, "step": 9000 }, { "epoch": 2.91, "learning_rate": 2.9029064196742255e-05, "loss": 0.4727, "step": 9100 }, { "epoch": 2.94, "learning_rate": 2.9348450974129674e-05, "loss": 0.4839, "step": 9200 }, { "epoch": 2.97, "learning_rate": 2.966783775151709e-05, "loss": 0.4769, "step": 9300 }, { "epoch": 3.0, "learning_rate": 2.9987224528904505e-05, "loss": 0.4641, "step": 9400 }, { "epoch": 3.03, "learning_rate": 3.030661130629192e-05, "loss": 0.4282, "step": 9500 }, { "epoch": 3.07, "learning_rate": 3.0625998083679336e-05, "loss": 0.4057, "step": 9600 }, { "epoch": 3.1, "learning_rate": 3.094538486106676e-05, "loss": 0.4192, "step": 9700 }, { "epoch": 3.13, "learning_rate": 3.126477163845417e-05, "loss": 0.4186, "step": 9800 }, { "epoch": 3.16, "learning_rate": 3.158415841584159e-05, "loss": 0.4086, "step": 9900 }, { "epoch": 3.19, "learning_rate": 3.1903545193229005e-05, "loss": 0.4194, "step": 10000 }, { "epoch": 3.23, "learning_rate": 3.221973810284254e-05, "loss": 0.4443, "step": 10100 }, { "epoch": 3.26, "learning_rate": 3.2539124880229965e-05, "loss": 0.4217, "step": 10200 }, { "epoch": 3.29, "learning_rate": 3.2858511657617374e-05, "loss": 0.4275, "step": 10300 }, { "epoch": 3.32, "learning_rate": 3.3177898435004796e-05, "loss": 0.4422, "step": 10400 }, { "epoch": 3.35, "learning_rate": 3.3497285212392205e-05, "loss": 0.4227, "step": 10500 }, { "epoch": 3.39, "learning_rate": 3.381667198977963e-05, "loss": 0.4191, "step": 10600 }, { "epoch": 3.42, "learning_rate": 3.413605876716704e-05, "loss": 0.4309, "step": 10700 }, { "epoch": 3.45, "learning_rate": 3.445544554455446e-05, "loss": 0.4079, "step": 10800 }, { "epoch": 3.48, "learning_rate": 3.4774832321941874e-05, "loss": 0.4281, "step": 10900 }, { "epoch": 3.51, "learning_rate": 3.509421909932929e-05, "loss": 0.4137, "step": 11000 }, { "epoch": 3.55, "learning_rate": 3.5413605876716705e-05, "loss": 0.4168, "step": 11100 }, { "epoch": 3.58, "learning_rate": 3.573299265410412e-05, "loss": 0.4186, "step": 11200 }, { "epoch": 3.61, "learning_rate": 3.605237943149154e-05, "loss": 0.4325, "step": 11300 }, { "epoch": 3.64, "learning_rate": 3.637176620887895e-05, "loss": 0.4422, "step": 11400 }, { "epoch": 3.67, "learning_rate": 3.6691152986266374e-05, "loss": 0.4269, "step": 11500 }, { "epoch": 3.7, "learning_rate": 3.701053976365379e-05, "loss": 0.4368, "step": 11600 }, { "epoch": 3.74, "learning_rate": 3.7329926541041205e-05, "loss": 0.4207, "step": 11700 }, { "epoch": 3.77, "learning_rate": 3.764931331842862e-05, "loss": 0.4365, "step": 11800 }, { "epoch": 3.8, "learning_rate": 3.7968700095816036e-05, "loss": 0.4394, "step": 11900 }, { "epoch": 3.83, "learning_rate": 3.828808687320345e-05, "loss": 0.4351, "step": 12000 }, { "epoch": 3.86, "learning_rate": 3.860747365059087e-05, "loss": 0.4113, "step": 12100 }, { "epoch": 3.9, "learning_rate": 3.892686042797828e-05, "loss": 0.4261, "step": 12200 }, { "epoch": 3.93, "learning_rate": 3.92462472053657e-05, "loss": 0.4311, "step": 12300 }, { "epoch": 3.96, "learning_rate": 3.956563398275312e-05, "loss": 0.4246, "step": 12400 }, { "epoch": 3.99, "learning_rate": 3.988502076014053e-05, "loss": 0.4294, "step": 12500 }, { "epoch": 4.02, "learning_rate": 3.997728805138579e-05, "loss": 0.3854, "step": 12600 }, { "epoch": 4.06, "learning_rate": 3.994180063167608e-05, "loss": 0.3656, "step": 12700 }, { "epoch": 4.09, "learning_rate": 3.9906313211966366e-05, "loss": 0.3836, "step": 12800 }, { "epoch": 4.12, "learning_rate": 3.987082579225665e-05, "loss": 0.3664, "step": 12900 }, { "epoch": 4.15, "learning_rate": 3.9835338372546935e-05, "loss": 0.3884, "step": 13000 }, { "epoch": 4.18, "learning_rate": 3.979985095283722e-05, "loss": 0.3764, "step": 13100 }, { "epoch": 4.22, "learning_rate": 3.976436353312751e-05, "loss": 0.3733, "step": 13200 }, { "epoch": 4.25, "learning_rate": 3.97288761134178e-05, "loss": 0.3682, "step": 13300 }, { "epoch": 4.28, "learning_rate": 3.969338869370809e-05, "loss": 0.371, "step": 13400 }, { "epoch": 4.31, "learning_rate": 3.965790127399837e-05, "loss": 0.3741, "step": 13500 }, { "epoch": 4.34, "learning_rate": 3.9622413854288656e-05, "loss": 0.3757, "step": 13600 }, { "epoch": 4.38, "learning_rate": 3.9586926434578944e-05, "loss": 0.3652, "step": 13700 }, { "epoch": 4.41, "learning_rate": 3.955143901486923e-05, "loss": 0.3892, "step": 13800 }, { "epoch": 4.44, "learning_rate": 3.951630646935662e-05, "loss": 0.3893, "step": 13900 }, { "epoch": 4.47, "learning_rate": 3.948081904964691e-05, "loss": 0.3823, "step": 14000 }, { "epoch": 4.5, "learning_rate": 3.944533162993719e-05, "loss": 0.3654, "step": 14100 }, { "epoch": 4.54, "learning_rate": 3.9409844210227476e-05, "loss": 0.3791, "step": 14200 }, { "epoch": 4.57, "learning_rate": 3.9374356790517764e-05, "loss": 0.3918, "step": 14300 }, { "epoch": 4.6, "learning_rate": 3.933886937080805e-05, "loss": 0.3713, "step": 14400 }, { "epoch": 4.63, "learning_rate": 3.930338195109834e-05, "loss": 0.376, "step": 14500 }, { "epoch": 4.66, "learning_rate": 3.926789453138863e-05, "loss": 0.3682, "step": 14600 }, { "epoch": 4.69, "learning_rate": 3.9232407111678915e-05, "loss": 0.374, "step": 14700 }, { "epoch": 4.73, "learning_rate": 3.91969196919692e-05, "loss": 0.3774, "step": 14800 }, { "epoch": 4.76, "learning_rate": 3.9161432272259484e-05, "loss": 0.3806, "step": 14900 }, { "epoch": 4.79, "learning_rate": 3.912594485254977e-05, "loss": 0.3886, "step": 15000 }, { "epoch": 4.82, "learning_rate": 3.909045743284006e-05, "loss": 0.3742, "step": 15100 }, { "epoch": 4.85, "learning_rate": 3.905497001313035e-05, "loss": 0.3714, "step": 15200 }, { "epoch": 4.89, "learning_rate": 3.9019482593420636e-05, "loss": 0.3721, "step": 15300 }, { "epoch": 4.92, "learning_rate": 3.8983995173710924e-05, "loss": 0.3716, "step": 15400 }, { "epoch": 4.95, "learning_rate": 3.8948507754001205e-05, "loss": 0.3728, "step": 15500 }, { "epoch": 4.98, "learning_rate": 3.891302033429149e-05, "loss": 0.384, "step": 15600 }, { "epoch": 5.01, "learning_rate": 3.887753291458178e-05, "loss": 0.3415, "step": 15700 }, { "epoch": 5.05, "learning_rate": 3.884204549487207e-05, "loss": 0.3064, "step": 15800 }, { "epoch": 5.08, "learning_rate": 3.880655807516236e-05, "loss": 0.3164, "step": 15900 }, { "epoch": 5.11, "learning_rate": 3.8771070655452645e-05, "loss": 0.3215, "step": 16000 }, { "epoch": 5.14, "learning_rate": 3.873558323574293e-05, "loss": 0.3234, "step": 16100 }, { "epoch": 5.17, "learning_rate": 3.870009581603322e-05, "loss": 0.3119, "step": 16200 }, { "epoch": 5.21, "learning_rate": 3.866460839632351e-05, "loss": 0.3145, "step": 16300 }, { "epoch": 5.24, "learning_rate": 3.8629120976613796e-05, "loss": 0.3333, "step": 16400 }, { "epoch": 5.27, "learning_rate": 3.8593633556904084e-05, "loss": 0.3195, "step": 16500 }, { "epoch": 5.3, "learning_rate": 3.855814613719437e-05, "loss": 0.3159, "step": 16600 }, { "epoch": 5.33, "learning_rate": 3.852265871748465e-05, "loss": 0.3189, "step": 16700 }, { "epoch": 5.37, "learning_rate": 3.848717129777494e-05, "loss": 0.317, "step": 16800 }, { "epoch": 5.4, "learning_rate": 3.845168387806523e-05, "loss": 0.3231, "step": 16900 }, { "epoch": 5.43, "learning_rate": 3.841619645835552e-05, "loss": 0.3328, "step": 17000 }, { "epoch": 5.46, "learning_rate": 3.8380709038645805e-05, "loss": 0.3282, "step": 17100 }, { "epoch": 5.49, "learning_rate": 3.8345221618936086e-05, "loss": 0.3169, "step": 17200 }, { "epoch": 5.53, "learning_rate": 3.8309734199226374e-05, "loss": 0.3156, "step": 17300 }, { "epoch": 5.56, "learning_rate": 3.827424677951666e-05, "loss": 0.3313, "step": 17400 }, { "epoch": 5.59, "learning_rate": 3.823875935980695e-05, "loss": 0.3181, "step": 17500 }, { "epoch": 5.62, "learning_rate": 3.820327194009724e-05, "loss": 0.334, "step": 17600 }, { "epoch": 5.65, "learning_rate": 3.8167784520387526e-05, "loss": 0.3313, "step": 17700 }, { "epoch": 5.69, "learning_rate": 3.8132297100677814e-05, "loss": 0.3331, "step": 17800 }, { "epoch": 5.72, "learning_rate": 3.80968096809681e-05, "loss": 0.3405, "step": 17900 }, { "epoch": 5.75, "learning_rate": 3.806132226125839e-05, "loss": 0.3345, "step": 18000 }, { "epoch": 5.78, "learning_rate": 3.802583484154868e-05, "loss": 0.3295, "step": 18100 }, { "epoch": 5.81, "learning_rate": 3.7990347421838965e-05, "loss": 0.3244, "step": 18200 }, { "epoch": 5.84, "learning_rate": 3.795486000212925e-05, "loss": 0.3346, "step": 18300 }, { "epoch": 5.88, "learning_rate": 3.791972745661663e-05, "loss": 0.3317, "step": 18400 }, { "epoch": 5.91, "learning_rate": 3.788424003690692e-05, "loss": 0.3219, "step": 18500 }, { "epoch": 5.94, "learning_rate": 3.784875261719721e-05, "loss": 0.3167, "step": 18600 }, { "epoch": 5.97, "learning_rate": 3.781326519748749e-05, "loss": 0.33, "step": 18700 }, { "epoch": 6.0, "learning_rate": 3.777777777777778e-05, "loss": 0.3211, "step": 18800 }, { "epoch": 6.04, "learning_rate": 3.7742290358068066e-05, "loss": 0.2794, "step": 18900 }, { "epoch": 6.07, "learning_rate": 3.7706802938358354e-05, "loss": 0.2675, "step": 19000 }, { "epoch": 6.1, "learning_rate": 3.767131551864864e-05, "loss": 0.2676, "step": 19100 }, { "epoch": 6.13, "learning_rate": 3.763582809893893e-05, "loss": 0.2693, "step": 19200 }, { "epoch": 6.16, "learning_rate": 3.760034067922922e-05, "loss": 0.2669, "step": 19300 }, { "epoch": 6.2, "learning_rate": 3.7564853259519506e-05, "loss": 0.2744, "step": 19400 }, { "epoch": 6.23, "learning_rate": 3.7529365839809794e-05, "loss": 0.282, "step": 19500 }, { "epoch": 6.26, "learning_rate": 3.749387842010008e-05, "loss": 0.2848, "step": 19600 }, { "epoch": 6.29, "learning_rate": 3.745839100039037e-05, "loss": 0.2783, "step": 19700 }, { "epoch": 6.32, "learning_rate": 3.742290358068066e-05, "loss": 0.2843, "step": 19800 }, { "epoch": 6.36, "learning_rate": 3.738741616097094e-05, "loss": 0.2696, "step": 19900 }, { "epoch": 6.39, "learning_rate": 3.7351928741261226e-05, "loss": 0.2932, "step": 20000 }, { "epoch": 6.42, "learning_rate": 3.7316441321551514e-05, "loss": 0.2775, "step": 20100 }, { "epoch": 6.45, "learning_rate": 3.72809539018418e-05, "loss": 0.2893, "step": 20200 }, { "epoch": 6.48, "learning_rate": 3.724546648213209e-05, "loss": 0.286, "step": 20300 }, { "epoch": 6.52, "learning_rate": 3.720997906242237e-05, "loss": 0.2799, "step": 20400 }, { "epoch": 6.55, "learning_rate": 3.717449164271266e-05, "loss": 0.2827, "step": 20500 }, { "epoch": 6.58, "learning_rate": 3.713900422300295e-05, "loss": 0.2783, "step": 20600 }, { "epoch": 6.61, "learning_rate": 3.7103516803293235e-05, "loss": 0.2729, "step": 20700 }, { "epoch": 6.64, "learning_rate": 3.706802938358352e-05, "loss": 0.2747, "step": 20800 }, { "epoch": 6.68, "learning_rate": 3.703254196387381e-05, "loss": 0.2764, "step": 20900 }, { "epoch": 6.71, "learning_rate": 3.699705454416409e-05, "loss": 0.2804, "step": 21000 }, { "epoch": 6.74, "learning_rate": 3.696156712445438e-05, "loss": 0.2841, "step": 21100 }, { "epoch": 6.77, "learning_rate": 3.692607970474467e-05, "loss": 0.277, "step": 21200 }, { "epoch": 6.8, "learning_rate": 3.6890592285034956e-05, "loss": 0.284, "step": 21300 }, { "epoch": 6.83, "learning_rate": 3.6855104865325244e-05, "loss": 0.2778, "step": 21400 }, { "epoch": 6.87, "learning_rate": 3.681961744561553e-05, "loss": 0.2797, "step": 21500 }, { "epoch": 6.9, "learning_rate": 3.678413002590582e-05, "loss": 0.3036, "step": 21600 }, { "epoch": 6.93, "learning_rate": 3.674864260619611e-05, "loss": 0.2869, "step": 21700 }, { "epoch": 6.96, "learning_rate": 3.6713155186486395e-05, "loss": 0.2921, "step": 21800 }, { "epoch": 6.99, "learning_rate": 3.667766776677668e-05, "loss": 0.2844, "step": 21900 }, { "epoch": 7.03, "learning_rate": 3.664218034706697e-05, "loss": 0.2427, "step": 22000 }, { "epoch": 7.06, "learning_rate": 3.660669292735726e-05, "loss": 0.243, "step": 22100 }, { "epoch": 7.09, "learning_rate": 3.657120550764754e-05, "loss": 0.2311, "step": 22200 }, { "epoch": 7.12, "learning_rate": 3.653571808793783e-05, "loss": 0.2358, "step": 22300 }, { "epoch": 7.15, "learning_rate": 3.6500230668228116e-05, "loss": 0.2287, "step": 22400 }, { "epoch": 7.19, "learning_rate": 3.6464743248518404e-05, "loss": 0.2519, "step": 22500 }, { "epoch": 7.22, "learning_rate": 3.642925582880869e-05, "loss": 0.2528, "step": 22600 }, { "epoch": 7.25, "learning_rate": 3.639376840909898e-05, "loss": 0.2507, "step": 22700 }, { "epoch": 7.28, "learning_rate": 3.635828098938926e-05, "loss": 0.2496, "step": 22800 }, { "epoch": 7.31, "learning_rate": 3.632279356967955e-05, "loss": 0.2503, "step": 22900 }, { "epoch": 7.35, "learning_rate": 3.628730614996984e-05, "loss": 0.2383, "step": 23000 }, { "epoch": 7.38, "learning_rate": 3.6251818730260125e-05, "loss": 0.2513, "step": 23100 }, { "epoch": 7.41, "learning_rate": 3.621633131055041e-05, "loss": 0.2492, "step": 23200 }, { "epoch": 7.44, "learning_rate": 3.61808438908407e-05, "loss": 0.2473, "step": 23300 }, { "epoch": 7.47, "learning_rate": 3.614535647113099e-05, "loss": 0.2502, "step": 23400 }, { "epoch": 7.51, "learning_rate": 3.6109869051421276e-05, "loss": 0.2492, "step": 23500 }, { "epoch": 7.54, "learning_rate": 3.6074381631711564e-05, "loss": 0.2366, "step": 23600 }, { "epoch": 7.57, "learning_rate": 3.603889421200185e-05, "loss": 0.2393, "step": 23700 }, { "epoch": 7.6, "learning_rate": 3.600340679229214e-05, "loss": 0.2483, "step": 23800 }, { "epoch": 7.63, "learning_rate": 3.596791937258242e-05, "loss": 0.2428, "step": 23900 }, { "epoch": 7.67, "learning_rate": 3.593243195287271e-05, "loss": 0.2507, "step": 24000 }, { "epoch": 7.7, "learning_rate": 3.5896944533163e-05, "loss": 0.2491, "step": 24100 }, { "epoch": 7.73, "learning_rate": 3.5861457113453285e-05, "loss": 0.2549, "step": 24200 }, { "epoch": 7.76, "learning_rate": 3.582596969374357e-05, "loss": 0.251, "step": 24300 }, { "epoch": 7.79, "learning_rate": 3.579048227403386e-05, "loss": 0.2357, "step": 24400 }, { "epoch": 7.82, "learning_rate": 3.575499485432414e-05, "loss": 0.2441, "step": 24500 }, { "epoch": 7.86, "learning_rate": 3.571950743461443e-05, "loss": 0.2561, "step": 24600 }, { "epoch": 7.89, "learning_rate": 3.568402001490472e-05, "loss": 0.2524, "step": 24700 }, { "epoch": 7.92, "learning_rate": 3.5648532595195006e-05, "loss": 0.2467, "step": 24800 }, { "epoch": 7.95, "learning_rate": 3.5613045175485293e-05, "loss": 0.2416, "step": 24900 }, { "epoch": 7.98, "learning_rate": 3.557755775577558e-05, "loss": 0.2511, "step": 25000 }, { "epoch": 8.02, "learning_rate": 3.554207033606587e-05, "loss": 0.2323, "step": 25100 }, { "epoch": 8.05, "learning_rate": 3.550658291635615e-05, "loss": 0.2158, "step": 25200 }, { "epoch": 8.08, "learning_rate": 3.547109549664644e-05, "loss": 0.21, "step": 25300 }, { "epoch": 8.11, "learning_rate": 3.5435608076936726e-05, "loss": 0.2102, "step": 25400 }, { "epoch": 8.14, "learning_rate": 3.5400120657227014e-05, "loss": 0.2071, "step": 25500 }, { "epoch": 8.18, "learning_rate": 3.53646332375173e-05, "loss": 0.2198, "step": 25600 }, { "epoch": 8.21, "learning_rate": 3.532914581780759e-05, "loss": 0.2193, "step": 25700 }, { "epoch": 8.24, "learning_rate": 3.529365839809788e-05, "loss": 0.2166, "step": 25800 }, { "epoch": 8.27, "learning_rate": 3.5258170978388166e-05, "loss": 0.2237, "step": 25900 }, { "epoch": 8.3, "learning_rate": 3.5222683558678454e-05, "loss": 0.2241, "step": 26000 }, { "epoch": 8.34, "learning_rate": 3.518719613896874e-05, "loss": 0.2132, "step": 26100 }, { "epoch": 8.37, "learning_rate": 3.515170871925902e-05, "loss": 0.2101, "step": 26200 }, { "epoch": 8.4, "learning_rate": 3.511622129954931e-05, "loss": 0.227, "step": 26300 }, { "epoch": 8.43, "learning_rate": 3.50807338798396e-05, "loss": 0.2221, "step": 26400 }, { "epoch": 8.46, "learning_rate": 3.504560133432698e-05, "loss": 0.2181, "step": 26500 }, { "epoch": 8.5, "learning_rate": 3.501011391461727e-05, "loss": 0.2208, "step": 26600 }, { "epoch": 8.53, "learning_rate": 3.4974626494907555e-05, "loss": 0.2192, "step": 26700 }, { "epoch": 8.56, "learning_rate": 3.493913907519784e-05, "loss": 0.2227, "step": 26800 }, { "epoch": 8.59, "learning_rate": 3.490365165548813e-05, "loss": 0.2308, "step": 26900 }, { "epoch": 8.62, "learning_rate": 3.486816423577842e-05, "loss": 0.22, "step": 27000 }, { "epoch": 8.66, "learning_rate": 3.4832676816068706e-05, "loss": 0.2247, "step": 27100 }, { "epoch": 8.69, "learning_rate": 3.4797189396358994e-05, "loss": 0.2148, "step": 27200 }, { "epoch": 8.72, "learning_rate": 3.476170197664928e-05, "loss": 0.2136, "step": 27300 }, { "epoch": 8.75, "learning_rate": 3.472621455693957e-05, "loss": 0.2282, "step": 27400 }, { "epoch": 8.78, "learning_rate": 3.469072713722986e-05, "loss": 0.2266, "step": 27500 }, { "epoch": 8.82, "learning_rate": 3.4655239717520146e-05, "loss": 0.2198, "step": 27600 }, { "epoch": 8.85, "learning_rate": 3.461975229781043e-05, "loss": 0.2243, "step": 27700 }, { "epoch": 8.88, "learning_rate": 3.4584264878100715e-05, "loss": 0.2358, "step": 27800 }, { "epoch": 8.91, "learning_rate": 3.4548777458391e-05, "loss": 0.2255, "step": 27900 }, { "epoch": 8.94, "learning_rate": 3.451329003868129e-05, "loss": 0.2258, "step": 28000 }, { "epoch": 8.97, "learning_rate": 3.447780261897158e-05, "loss": 0.2228, "step": 28100 }, { "epoch": 9.01, "learning_rate": 3.4442315199261866e-05, "loss": 0.2183, "step": 28200 }, { "epoch": 9.04, "learning_rate": 3.440682777955215e-05, "loss": 0.1988, "step": 28300 }, { "epoch": 9.07, "learning_rate": 3.4371340359842436e-05, "loss": 0.188, "step": 28400 }, { "epoch": 9.1, "learning_rate": 3.4335852940132723e-05, "loss": 0.1963, "step": 28500 }, { "epoch": 9.13, "learning_rate": 3.430036552042301e-05, "loss": 0.1919, "step": 28600 }, { "epoch": 9.17, "learning_rate": 3.42648781007133e-05, "loss": 0.1906, "step": 28700 }, { "epoch": 9.2, "learning_rate": 3.422939068100359e-05, "loss": 0.1911, "step": 28800 }, { "epoch": 9.23, "learning_rate": 3.4193903261293875e-05, "loss": 0.1971, "step": 28900 }, { "epoch": 9.26, "learning_rate": 3.415841584158416e-05, "loss": 0.1931, "step": 29000 }, { "epoch": 9.29, "learning_rate": 3.412292842187445e-05, "loss": 0.1951, "step": 29100 }, { "epoch": 9.33, "learning_rate": 3.408744100216474e-05, "loss": 0.2008, "step": 29200 }, { "epoch": 9.36, "learning_rate": 3.405195358245503e-05, "loss": 0.2053, "step": 29300 }, { "epoch": 9.39, "learning_rate": 3.401646616274531e-05, "loss": 0.2015, "step": 29400 }, { "epoch": 9.42, "learning_rate": 3.3980978743035596e-05, "loss": 0.193, "step": 29500 }, { "epoch": 9.45, "learning_rate": 3.3945491323325884e-05, "loss": 0.1955, "step": 29600 }, { "epoch": 9.49, "learning_rate": 3.391000390361617e-05, "loss": 0.1993, "step": 29700 }, { "epoch": 9.52, "learning_rate": 3.387451648390646e-05, "loss": 0.2043, "step": 29800 }, { "epoch": 9.55, "learning_rate": 3.383902906419675e-05, "loss": 0.1955, "step": 29900 }, { "epoch": 9.58, "learning_rate": 3.380354164448703e-05, "loss": 0.1961, "step": 30000 }, { "epoch": 9.61, "learning_rate": 3.3768054224777316e-05, "loss": 0.1961, "step": 30100 }, { "epoch": 9.65, "learning_rate": 3.3732566805067604e-05, "loss": 0.196, "step": 30200 }, { "epoch": 9.68, "learning_rate": 3.369707938535789e-05, "loss": 0.2088, "step": 30300 }, { "epoch": 9.71, "learning_rate": 3.366159196564818e-05, "loss": 0.2031, "step": 30400 }, { "epoch": 9.74, "learning_rate": 3.362610454593847e-05, "loss": 0.1949, "step": 30500 }, { "epoch": 9.77, "learning_rate": 3.3590617126228756e-05, "loss": 0.1988, "step": 30600 }, { "epoch": 9.81, "learning_rate": 3.3555129706519044e-05, "loss": 0.2061, "step": 30700 }, { "epoch": 9.84, "learning_rate": 3.351964228680933e-05, "loss": 0.2022, "step": 30800 }, { "epoch": 9.87, "learning_rate": 3.348450974129671e-05, "loss": 0.1965, "step": 30900 }, { "epoch": 9.9, "learning_rate": 3.3449022321587e-05, "loss": 0.1999, "step": 31000 }, { "epoch": 9.93, "learning_rate": 3.341353490187729e-05, "loss": 0.2065, "step": 31100 }, { "epoch": 9.96, "learning_rate": 3.3378047482167576e-05, "loss": 0.2031, "step": 31200 }, { "epoch": 10.0, "learning_rate": 3.3342560062457864e-05, "loss": 0.2078, "step": 31300 }, { "epoch": 10.03, "learning_rate": 3.330707264274815e-05, "loss": 0.1733, "step": 31400 }, { "epoch": 10.06, "learning_rate": 3.327158522303843e-05, "loss": 0.1654, "step": 31500 }, { "epoch": 10.09, "learning_rate": 3.323609780332872e-05, "loss": 0.1674, "step": 31600 }, { "epoch": 10.12, "learning_rate": 3.320061038361901e-05, "loss": 0.1725, "step": 31700 }, { "epoch": 10.16, "learning_rate": 3.3165122963909297e-05, "loss": 0.1761, "step": 31800 }, { "epoch": 10.19, "learning_rate": 3.3129635544199584e-05, "loss": 0.1757, "step": 31900 }, { "epoch": 10.22, "learning_rate": 3.309414812448987e-05, "loss": 0.1781, "step": 32000 }, { "epoch": 10.25, "learning_rate": 3.3058660704780153e-05, "loss": 0.1716, "step": 32100 }, { "epoch": 10.28, "learning_rate": 3.302317328507044e-05, "loss": 0.1717, "step": 32200 }, { "epoch": 10.32, "learning_rate": 3.298768586536073e-05, "loss": 0.179, "step": 32300 }, { "epoch": 10.35, "learning_rate": 3.295219844565102e-05, "loss": 0.174, "step": 32400 }, { "epoch": 10.38, "learning_rate": 3.2916711025941305e-05, "loss": 0.178, "step": 32500 }, { "epoch": 10.41, "learning_rate": 3.288122360623159e-05, "loss": 0.1765, "step": 32600 }, { "epoch": 10.44, "learning_rate": 3.284573618652188e-05, "loss": 0.1824, "step": 32700 }, { "epoch": 10.48, "learning_rate": 3.281024876681217e-05, "loss": 0.1806, "step": 32800 }, { "epoch": 10.51, "learning_rate": 3.277511622129955e-05, "loss": 0.1785, "step": 32900 }, { "epoch": 10.54, "learning_rate": 3.2739983675786936e-05, "loss": 0.1698, "step": 33000 }, { "epoch": 10.57, "learning_rate": 3.2704496256077224e-05, "loss": 0.1734, "step": 33100 }, { "epoch": 10.6, "learning_rate": 3.2669008836367505e-05, "loss": 0.1761, "step": 33200 }, { "epoch": 10.64, "learning_rate": 3.263352141665779e-05, "loss": 0.1786, "step": 33300 }, { "epoch": 10.67, "learning_rate": 3.259803399694808e-05, "loss": 0.1767, "step": 33400 }, { "epoch": 10.7, "learning_rate": 3.256254657723837e-05, "loss": 0.1881, "step": 33500 }, { "epoch": 10.73, "learning_rate": 3.252705915752866e-05, "loss": 0.1792, "step": 33600 }, { "epoch": 10.76, "learning_rate": 3.2491571737818945e-05, "loss": 0.1832, "step": 33700 }, { "epoch": 10.8, "learning_rate": 3.245608431810923e-05, "loss": 0.1889, "step": 33800 }, { "epoch": 10.83, "learning_rate": 3.242059689839952e-05, "loss": 0.1824, "step": 33900 }, { "epoch": 10.86, "learning_rate": 3.238510947868981e-05, "loss": 0.1746, "step": 34000 }, { "epoch": 10.89, "learning_rate": 3.2349622058980096e-05, "loss": 0.1735, "step": 34100 }, { "epoch": 10.92, "learning_rate": 3.2314134639270384e-05, "loss": 0.1822, "step": 34200 }, { "epoch": 10.95, "learning_rate": 3.227864721956067e-05, "loss": 0.1768, "step": 34300 }, { "epoch": 10.99, "learning_rate": 3.224315979985095e-05, "loss": 0.1812, "step": 34400 }, { "epoch": 11.02, "learning_rate": 3.220767238014124e-05, "loss": 0.1645, "step": 34500 }, { "epoch": 11.05, "learning_rate": 3.217218496043153e-05, "loss": 0.1483, "step": 34600 }, { "epoch": 11.08, "learning_rate": 3.213669754072182e-05, "loss": 0.1615, "step": 34700 }, { "epoch": 11.11, "learning_rate": 3.2101210121012105e-05, "loss": 0.1637, "step": 34800 }, { "epoch": 11.15, "learning_rate": 3.206572270130239e-05, "loss": 0.1497, "step": 34900 }, { "epoch": 11.18, "learning_rate": 3.2030235281592674e-05, "loss": 0.1524, "step": 35000 } ], "max_steps": 125240, "num_train_epochs": 40, "total_flos": 0.0, "trial_name": null, "trial_params": null }