Training in progress, epoch 3

all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 3.0,
+    "eval_accuracy": 0.9825637339578237,
+    "eval_loss": 0.04691937565803528,
+    "eval_runtime": 172.6572,
+    "eval_samples_per_second": 499.585,
+    "eval_steps_per_second": 62.453,
+    "train_loss": 0.17630483790067725,
+    "train_runtime": 837.6448,
+    "train_samples_per_second": 308.927,
+    "train_steps_per_second": 4.828
+}

config.json

@@ -0,0 +1,61 @@
+{
+  "_name_or_path": "microsoft/swin-tiny-patch4-window7-224",
+  "architectures": [
+    "SwinForImageClassification"
+  ],
+  "attention_probs_dropout_prob": 0.0,
+  "depths": [
+    2,
+    2,
+    6,
+    2
+  ],
+  "drop_path_rate": 0.1,
+  "embed_dim": 96,
+  "encoder_stride": 32,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.0,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "Fake",
+    "1": "Real"
+  },
+  "image_size": 224,
+  "initializer_range": 0.02,
+  "label2id": {
+    "Fake": 0,
+    "Real": 1
+  },
+  "layer_norm_eps": 1e-05,
+  "mlp_ratio": 4.0,
+  "model_type": "swin",
+  "num_channels": 3,
+  "num_heads": [
+    3,
+    6,
+    12,
+    24
+  ],
+  "num_layers": 4,
+  "out_features": [
+    "stage4"
+  ],
+  "out_indices": [
+    4
+  ],
+  "patch_size": 4,
+  "path_norm": true,
+  "problem_type": "single_label_classification",
+  "qkv_bias": true,
+  "stage_names": [
+    "stem",
+    "stage1",
+    "stage2",
+    "stage3",
+    "stage4"
+  ],
+  "torch_dtype": "float32",
+  "transformers_version": "4.39.2",
+  "use_absolute_embeddings": false,
+  "window_size": 7
+}

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 3.0,
+    "eval_accuracy": 0.9825637339578237,
+    "eval_loss": 0.04691937565803528,
+    "eval_runtime": 172.6572,
+    "eval_samples_per_second": 499.585,
+    "eval_steps_per_second": 62.453
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f863a9b52df79593998927c4e0cead411ba5265003573772677e855c32ec36e
+size 110342832

preprocessor_config.json

@@ -0,0 +1,36 @@
+{
+  "_valid_processor_keys": [
+    "images",
+    "do_resize",
+    "size",
+    "resample",
+    "do_rescale",
+    "rescale_factor",
+    "do_normalize",
+    "image_mean",
+    "image_std",
+    "return_tensors",
+    "data_format",
+    "input_data_format"
+  ],
+  "do_normalize": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.485,
+    0.456,
+    0.406
+  ],
+  "image_processor_type": "ViTImageProcessor",
+  "image_std": [
+    0.229,
+    0.224,
+    0.225
+  ],
+  "resample": 3,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "height": 224,
+    "width": 224
+  }
+}

runs/Mar29_19-09-46_DESKTOP-604-AI/events.out.tfevents.1711710586.DESKTOP-604-AI.8320.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:27e1ae2af3b84910c616e207ea78fa9d8795e61468fec66cedd9fff0ecb8af62
+size 4830

runs/Mar29_19-14-23_DESKTOP-604-AI/events.out.tfevents.1711710863.DESKTOP-604-AI.6844.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89aa9fe870f013aa52cd412e204b3108955f824f855a7fb4559e46d02db0fd2d
+size 22070

runs/Mar29_19-40-45_DESKTOP-604-AI/events.out.tfevents.1711712445.DESKTOP-604-AI.15316.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c2e161a46bf08b9ecf8d1901f969523fadbd0adf848b802977425f4a3044bc6d
+size 11394

runs/Mar29_21-42-36_DESKTOP-604-AI/events.out.tfevents.1711719756.DESKTOP-604-AI.14908.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5930a266cb54b7a6e7f2493f517aa502b54536119a419a9e8ceb083933e6c91
+size 4984

runs/Mar29_21-45-36_DESKTOP-604-AI/events.out.tfevents.1711719936.DESKTOP-604-AI.16272.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e3ee1d08eb0f2b1ade4e3a902d2f41ba3d66d929994bd728a8165a541332415
+size 4830

runs/Mar29_22-12-05_DESKTOP-604-AI/events.out.tfevents.1711721526.DESKTOP-604-AI.17172.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24b3e87f37d1dae2b9439996d41a02e6f2fc5ceb2ae45be54d40a0e716387473
+size 5754

runs/Mar29_22-15-48_DESKTOP-604-AI/events.out.tfevents.1711721748.DESKTOP-604-AI.15064.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f1390f8aa8f2acf4d70d70823357aef3c191656de67eac8f43e2ce8961adf99
+size 6216

runs/Mar29_22-19-05_DESKTOP-604-AI/events.out.tfevents.1711721946.DESKTOP-604-AI.16132.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:207de74bcc976165e75237ed1b0ce569ad9fc47341bdda6cc3f2043a6291c399
+size 4830

runs/Mar29_22-21-28_DESKTOP-604-AI/events.out.tfevents.1711722089.DESKTOP-604-AI.15564.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7fd5c9dfdc039d744ed4181ad742fe9d389ba9c989498481e2b9594551b3c253
+size 4522

runs/Mar29_23-49-26_DESKTOP-604-AI/events.out.tfevents.1711727385.DESKTOP-604-AI.15724.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aa2bee095460115ec7190ae7607292eed7ed85ed9099cc4c4379f6985d3901bd
+size 4568

runs/Mar29_23-52-52_DESKTOP-604-AI/events.out.tfevents.1711727573.DESKTOP-604-AI.1548.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:369653e51aba643a97ad5e80e9faa52d8cc7276a41661b158cfdbab4f829ac56
+size 4568

runs/Mar30_00-00-53_DESKTOP-604-AI/events.out.tfevents.1711728098.DESKTOP-604-AI.16088.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:93a981c2809dbacf9d7c7bd2f9ca5bc00e5fd7cae74b2b6b25ebde45f76992d8
+size 5028

runs/Mar30_00-08-29_DESKTOP-604-AI/events.out.tfevents.1711728522.DESKTOP-604-AI.9004.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:650f672b3770a15b32759f8c46c47a3b11ccb0c19d92a4a97c767c25e967da22
+size 90578

runs/Mar30_00-08-29_DESKTOP-604-AI/events.out.tfevents.1711729774.DESKTOP-604-AI.9004.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1e3b2e39f4fdb7bb815e479f2ac101284fd205e33ef0dfb3433d4e9ac6bf4ef
+size 346

runs/Mar30_00-33-25_DESKTOP-604-AI/events.out.tfevents.1711730020.DESKTOP-604-AI.13792.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed05c0acc98f88473fc605394540038a46cf215041f89020de53f13de8e59208
+size 10677

runs/Mar30_00-38-29_DESKTOP-604-AI/events.out.tfevents.1711730374.DESKTOP-604-AI.14788.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:185ea5e9f607f40e0868efc0821b793fefc17990298621290325b07d57262a16
+size 5856

runs/Mar30_00-42-00_DESKTOP-604-AI/events.out.tfevents.1711730520.DESKTOP-604-AI.15800.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:162aed65d213d0b17720852bc20fc54d988df92984a01fe21f740030cf3254d2
+size 6684

runs/Mar30_00-45-27_DESKTOP-604-AI/events.out.tfevents.1711730728.DESKTOP-604-AI.16276.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50a6b9b0cf6812aec250c7b02645a0ccf5580169d8e687a528b943ffc6c4710c
+size 5028

runs/Mar30_00-54-45_DESKTOP-604-AI/events.out.tfevents.1711731311.DESKTOP-604-AI.16272.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d9daa89124bc99b5824855e6f36ddbeb26ac4a0ab1f1d43decb733e935d9c24b
+size 5030

runs/Mar30_00-57-52_DESKTOP-604-AI/events.out.tfevents.1711731472.DESKTOP-604-AI.15912.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d94403a4934d0fc2b4e2b26faa430b3c8a8b7ef2f0d5872e4ca9b69d6ce68e46
+size 27348

runs/Mar30_01-05-25_DESKTOP-604-AI/events.out.tfevents.1711731926.DESKTOP-604-AI.14892.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee0c665721945c6cbc4354abdd4fb57496d62750a34c3e054d09abf8e63b0c01
+size 91549

runs/Mar30_01-05-25_DESKTOP-604-AI/events.out.tfevents.1711733426.DESKTOP-604-AI.14892.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8eebaf58f07e57f2c2e1bf22d3107588d6af4810efb900c3ed2326e4e7fdd09d
+size 411

runs/Mar30_07-26-13_DESKTOP-604-AI/events.out.tfevents.1711754785.DESKTOP-604-AI.3688.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b00990c5e8eb8e412e2c7195d73121d5601693d19ef36067e9a1537501c92988
+size 91549

runs/Mar30_07-26-13_DESKTOP-604-AI/events.out.tfevents.1711756191.DESKTOP-604-AI.3688.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6a030b075787c898e480840f3d86db1b56d2d799be24dacd5e6cf4f672dd582a
+size 411

runs/Mar30_08-06-42_DESKTOP-604-AI/events.out.tfevents.1711757227.DESKTOP-604-AI.9104.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:137e11e11a384ed1d22717f2f7185158dd98f8d88b339ebb377f4b2bd26fd5b8
+size 90578

runs/Mar30_08-06-42_DESKTOP-604-AI/events.out.tfevents.1711758237.DESKTOP-604-AI.9104.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac2942bc2652b5eebbe6469da0dd55183e42ae0ad89886cfb6906401a45a48d9
+size 411

runs/Mar30_08-31-05_DESKTOP-604-AI/events.out.tfevents.1711758686.DESKTOP-604-AI.8592.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe011a11d44d4d71d7ba827f668576028e0cf0d64f1c7472f31ebc8ac57b2e81
+size 91548

train_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 3.0,
+    "train_loss": 0.17630483790067725,
+    "train_runtime": 837.6448,
+    "train_samples_per_second": 308.927,
+    "train_steps_per_second": 4.828
+}

trainer_state.json

@@ -0,0 +1,2858 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 4044,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.081072807312012,
+      "learning_rate": 1.234567901234568e-06,
+      "loss": 0.6066,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.491064548492432,
+      "learning_rate": 2.469135802469136e-06,
+      "loss": 0.5882,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.853269338607788,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 0.555,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.870471239089966,
+      "learning_rate": 4.938271604938272e-06,
+      "loss": 0.5146,
+      "step": 40
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8400750160217285,
+      "learning_rate": 6.172839506172839e-06,
+      "loss": 0.5181,
+      "step": 50
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.394155263900757,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.4959,
+      "step": 60
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4633309841156006,
+      "learning_rate": 8.641975308641975e-06,
+      "loss": 0.5463,
+      "step": 70
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.766242504119873,
+      "learning_rate": 9.876543209876543e-06,
+      "loss": 0.5051,
+      "step": 80
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.002878189086914,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.5439,
+      "step": 90
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.458632469177246,
+      "learning_rate": 1.2345679012345678e-05,
+      "loss": 0.5019,
+      "step": 100
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1725046634674072,
+      "learning_rate": 1.3580246913580247e-05,
+      "loss": 0.5181,
+      "step": 110
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.216794490814209,
+      "learning_rate": 1.4814814814814815e-05,
+      "loss": 0.5341,
+      "step": 120
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1110751628875732,
+      "learning_rate": 1.604938271604938e-05,
+      "loss": 0.5146,
+      "step": 130
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0247414112091064,
+      "learning_rate": 1.728395061728395e-05,
+      "loss": 0.5027,
+      "step": 140
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.48282527923584,
+      "learning_rate": 1.8518518518518518e-05,
+      "loss": 0.4811,
+      "step": 150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1287665367126465,
+      "learning_rate": 1.9753086419753087e-05,
+      "loss": 0.5032,
+      "step": 160
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.911570072174072,
+      "learning_rate": 2.0987654320987655e-05,
+      "loss": 0.5049,
+      "step": 170
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5687508583068848,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.481,
+      "step": 180
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.535050392150879,
+      "learning_rate": 2.345679012345679e-05,
+      "loss": 0.485,
+      "step": 190
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.1198344230651855,
+      "learning_rate": 2.4691358024691357e-05,
+      "loss": 0.4441,
+      "step": 200
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.31569242477417,
+      "learning_rate": 2.5925925925925925e-05,
+      "loss": 0.4448,
+      "step": 210
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.231945514678955,
+      "learning_rate": 2.7160493827160493e-05,
+      "loss": 0.4745,
+      "step": 220
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.131611824035645,
+      "learning_rate": 2.839506172839506e-05,
+      "loss": 0.4313,
+      "step": 230
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.373456954956055,
+      "learning_rate": 2.962962962962963e-05,
+      "loss": 0.4343,
+      "step": 240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.445795059204102,
+      "learning_rate": 3.08641975308642e-05,
+      "loss": 0.3897,
+      "step": 250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 28.160856246948242,
+      "learning_rate": 3.209876543209876e-05,
+      "loss": 0.3948,
+      "step": 260
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.549301624298096,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.3807,
+      "step": 270
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.183242797851562,
+      "learning_rate": 3.45679012345679e-05,
+      "loss": 0.3427,
+      "step": 280
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.87475299835205,
+      "learning_rate": 3.580246913580247e-05,
+      "loss": 0.3912,
+      "step": 290
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.063522338867188,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 0.378,
+      "step": 300
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.313444137573242,
+      "learning_rate": 3.82716049382716e-05,
+      "loss": 0.3162,
+      "step": 310
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.576651573181152,
+      "learning_rate": 3.950617283950617e-05,
+      "loss": 0.339,
+      "step": 320
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.328672409057617,
+      "learning_rate": 4.074074074074074e-05,
+      "loss": 0.3427,
+      "step": 330
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 19.89008903503418,
+      "learning_rate": 4.197530864197531e-05,
+      "loss": 0.3345,
+      "step": 340
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.535983085632324,
+      "learning_rate": 4.3209876543209875e-05,
+      "loss": 0.3687,
+      "step": 350
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.986384391784668,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.3328,
+      "step": 360
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.05379581451416,
+      "learning_rate": 4.567901234567901e-05,
+      "loss": 0.3589,
+      "step": 370
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.419095516204834,
+      "learning_rate": 4.691358024691358e-05,
+      "loss": 0.3198,
+      "step": 380
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.914067268371582,
+      "learning_rate": 4.814814814814815e-05,
+      "loss": 0.3297,
+      "step": 390
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.265212535858154,
+      "learning_rate": 4.938271604938271e-05,
+      "loss": 0.3169,
+      "step": 400
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.57672119140625,
+      "learning_rate": 4.993129980763946e-05,
+      "loss": 0.2985,
+      "step": 410
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 25.84030532836914,
+      "learning_rate": 4.979389942291839e-05,
+      "loss": 0.3351,
+      "step": 420
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 24.027511596679688,
+      "learning_rate": 4.965649903819731e-05,
+      "loss": 0.3001,
+      "step": 430
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.898484230041504,
+      "learning_rate": 4.951909865347623e-05,
+      "loss": 0.319,
+      "step": 440
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.98867416381836,
+      "learning_rate": 4.9381698268755155e-05,
+      "loss": 0.3136,
+      "step": 450
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.969843864440918,
+      "learning_rate": 4.924429788403408e-05,
+      "loss": 0.3499,
+      "step": 460
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.836222171783447,
+      "learning_rate": 4.9106897499313e-05,
+      "loss": 0.4036,
+      "step": 470
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.171021461486816,
+      "learning_rate": 4.896949711459192e-05,
+      "loss": 0.2802,
+      "step": 480
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 14.70108413696289,
+      "learning_rate": 4.883209672987085e-05,
+      "loss": 0.2991,
+      "step": 490
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.405062675476074,
+      "learning_rate": 4.8694696345149774e-05,
+      "loss": 0.2726,
+      "step": 500
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.101396083831787,
+      "learning_rate": 4.8557295960428687e-05,
+      "loss": 0.2956,
+      "step": 510
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.017485618591309,
+      "learning_rate": 4.841989557570761e-05,
+      "loss": 0.3091,
+      "step": 520
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.195255279541016,
+      "learning_rate": 4.828249519098654e-05,
+      "loss": 0.2839,
+      "step": 530
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.744250774383545,
+      "learning_rate": 4.814509480626546e-05,
+      "loss": 0.3139,
+      "step": 540
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.792461395263672,
+      "learning_rate": 4.800769442154438e-05,
+      "loss": 0.2226,
+      "step": 550
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.129609107971191,
+      "learning_rate": 4.7870294036823306e-05,
+      "loss": 0.2469,
+      "step": 560
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.897442817687988,
+      "learning_rate": 4.7732893652102225e-05,
+      "loss": 0.2906,
+      "step": 570
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.450972557067871,
+      "learning_rate": 4.759549326738115e-05,
+      "loss": 0.2839,
+      "step": 580
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.978741645812988,
+      "learning_rate": 4.745809288266007e-05,
+      "loss": 0.28,
+      "step": 590
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.651639938354492,
+      "learning_rate": 4.7320692497939e-05,
+      "loss": 0.2546,
+      "step": 600
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 20.035005569458008,
+      "learning_rate": 4.718329211321792e-05,
+      "loss": 0.3578,
+      "step": 610
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 18.722013473510742,
+      "learning_rate": 4.7045891728496844e-05,
+      "loss": 0.3548,
+      "step": 620
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.33369255065918,
+      "learning_rate": 4.6908491343775764e-05,
+      "loss": 0.3182,
+      "step": 630
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.095462799072266,
+      "learning_rate": 4.6771090959054684e-05,
+      "loss": 0.2787,
+      "step": 640
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.798605918884277,
+      "learning_rate": 4.663369057433361e-05,
+      "loss": 0.2777,
+      "step": 650
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.711633682250977,
+      "learning_rate": 4.649629018961254e-05,
+      "loss": 0.266,
+      "step": 660
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.158576011657715,
+      "learning_rate": 4.6358889804891456e-05,
+      "loss": 0.2659,
+      "step": 670
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 29.281648635864258,
+      "learning_rate": 4.6221489420170376e-05,
+      "loss": 0.3339,
+      "step": 680
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.407151222229004,
+      "learning_rate": 4.60840890354493e-05,
+      "loss": 0.3152,
+      "step": 690
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.647674560546875,
+      "learning_rate": 4.594668865072822e-05,
+      "loss": 0.267,
+      "step": 700
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.17988109588623,
+      "learning_rate": 4.580928826600714e-05,
+      "loss": 0.3163,
+      "step": 710
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.614954948425293,
+      "learning_rate": 4.567188788128607e-05,
+      "loss": 0.2857,
+      "step": 720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.587752342224121,
+      "learning_rate": 4.5534487496564995e-05,
+      "loss": 0.245,
+      "step": 730
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.359344482421875,
+      "learning_rate": 4.5397087111843915e-05,
+      "loss": 0.215,
+      "step": 740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.509681701660156,
+      "learning_rate": 4.525968672712284e-05,
+      "loss": 0.2454,
+      "step": 750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.866887092590332,
+      "learning_rate": 4.512228634240176e-05,
+      "loss": 0.2479,
+      "step": 760
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 20.613985061645508,
+      "learning_rate": 4.498488595768068e-05,
+      "loss": 0.2819,
+      "step": 770
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.16818904876709,
+      "learning_rate": 4.484748557295961e-05,
+      "loss": 0.1846,
+      "step": 780
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.097874641418457,
+      "learning_rate": 4.4710085188238534e-05,
+      "loss": 0.2412,
+      "step": 790
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.144585609436035,
+      "learning_rate": 4.4572684803517453e-05,
+      "loss": 0.1982,
+      "step": 800
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.713954448699951,
+      "learning_rate": 4.443528441879637e-05,
+      "loss": 0.2237,
+      "step": 810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.307291030883789,
+      "learning_rate": 4.42978840340753e-05,
+      "loss": 0.2141,
+      "step": 820
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.887890815734863,
+      "learning_rate": 4.4160483649354226e-05,
+      "loss": 0.2256,
+      "step": 830
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 14.669983863830566,
+      "learning_rate": 4.402308326463314e-05,
+      "loss": 0.2602,
+      "step": 840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.341273307800293,
+      "learning_rate": 4.3885682879912066e-05,
+      "loss": 0.2225,
+      "step": 850
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 18.095508575439453,
+      "learning_rate": 4.374828249519099e-05,
+      "loss": 0.2188,
+      "step": 860
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.606022834777832,
+      "learning_rate": 4.361088211046991e-05,
+      "loss": 0.2103,
+      "step": 870
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 15.687580108642578,
+      "learning_rate": 4.347348172574883e-05,
+      "loss": 0.2387,
+      "step": 880
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 18.16533088684082,
+      "learning_rate": 4.333608134102776e-05,
+      "loss": 0.2452,
+      "step": 890
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.142931938171387,
+      "learning_rate": 4.319868095630668e-05,
+      "loss": 0.2081,
+      "step": 900
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.506136894226074,
+      "learning_rate": 4.3061280571585604e-05,
+      "loss": 0.2051,
+      "step": 910
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.4028706550598145,
+      "learning_rate": 4.2923880186864524e-05,
+      "loss": 0.2309,
+      "step": 920
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.455471038818359,
+      "learning_rate": 4.278647980214345e-05,
+      "loss": 0.228,
+      "step": 930
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.138572692871094,
+      "learning_rate": 4.264907941742237e-05,
+      "loss": 0.2219,
+      "step": 940
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.745438575744629,
+      "learning_rate": 4.25116790327013e-05,
+      "loss": 0.2162,
+      "step": 950
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.894223213195801,
+      "learning_rate": 4.2374278647980216e-05,
+      "loss": 0.1809,
+      "step": 960
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.754095077514648,
+      "learning_rate": 4.2236878263259136e-05,
+      "loss": 0.2358,
+      "step": 970
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.232985496520996,
+      "learning_rate": 4.209947787853806e-05,
+      "loss": 0.1953,
+      "step": 980
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.504615306854248,
+      "learning_rate": 4.196207749381699e-05,
+      "loss": 0.2461,
+      "step": 990
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.350137710571289,
+      "learning_rate": 4.18246771090959e-05,
+      "loss": 0.2208,
+      "step": 1000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.46402359008789,
+      "learning_rate": 4.168727672437483e-05,
+      "loss": 0.2037,
+      "step": 1010
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.952566146850586,
+      "learning_rate": 4.1549876339653755e-05,
+      "loss": 0.1881,
+      "step": 1020
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 27.129024505615234,
+      "learning_rate": 4.1412475954932675e-05,
+      "loss": 0.2177,
+      "step": 1030
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.786013603210449,
+      "learning_rate": 4.1275075570211595e-05,
+      "loss": 0.1875,
+      "step": 1040
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.231268882751465,
+      "learning_rate": 4.113767518549052e-05,
+      "loss": 0.1833,
+      "step": 1050
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.376607894897461,
+      "learning_rate": 4.100027480076945e-05,
+      "loss": 0.2072,
+      "step": 1060
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.239863395690918,
+      "learning_rate": 4.086287441604837e-05,
+      "loss": 0.185,
+      "step": 1070
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.10155725479126,
+      "learning_rate": 4.072547403132729e-05,
+      "loss": 0.188,
+      "step": 1080
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.190369606018066,
+      "learning_rate": 4.0588073646606214e-05,
+      "loss": 0.1936,
+      "step": 1090
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.407243728637695,
+      "learning_rate": 4.045067326188513e-05,
+      "loss": 0.2119,
+      "step": 1100
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.534134864807129,
+      "learning_rate": 4.031327287716406e-05,
+      "loss": 0.1915,
+      "step": 1110
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.6572346687316895,
+      "learning_rate": 4.017587249244298e-05,
+      "loss": 0.1655,
+      "step": 1120
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 18.50929832458496,
+      "learning_rate": 4.0038472107721906e-05,
+      "loss": 0.2091,
+      "step": 1130
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.284636497497559,
+      "learning_rate": 3.9901071723000826e-05,
+      "loss": 0.2185,
+      "step": 1140
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.729572296142578,
+      "learning_rate": 3.976367133827975e-05,
+      "loss": 0.1799,
+      "step": 1150
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8940351009368896,
+      "learning_rate": 3.962627095355867e-05,
+      "loss": 0.1992,
+      "step": 1160
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.741419792175293,
+      "learning_rate": 3.948887056883759e-05,
+      "loss": 0.1752,
+      "step": 1170
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.793862819671631,
+      "learning_rate": 3.935147018411652e-05,
+      "loss": 0.2016,
+      "step": 1180
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.699429988861084,
+      "learning_rate": 3.9214069799395445e-05,
+      "loss": 0.2257,
+      "step": 1190
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.716045379638672,
+      "learning_rate": 3.907666941467436e-05,
+      "loss": 0.2182,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.854546546936035,
+      "learning_rate": 3.8939269029953284e-05,
+      "loss": 0.1911,
+      "step": 1210
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.813031196594238,
+      "learning_rate": 3.880186864523221e-05,
+      "loss": 0.1997,
+      "step": 1220
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.75151252746582,
+      "learning_rate": 3.866446826051113e-05,
+      "loss": 0.2199,
+      "step": 1230
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.344423294067383,
+      "learning_rate": 3.852706787579005e-05,
+      "loss": 0.1976,
+      "step": 1240
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.043191909790039,
+      "learning_rate": 3.8389667491068977e-05,
+      "loss": 0.1968,
+      "step": 1250
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 17.982160568237305,
+      "learning_rate": 3.82522671063479e-05,
+      "loss": 0.1788,
+      "step": 1260
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 14.675708770751953,
+      "learning_rate": 3.811486672162682e-05,
+      "loss": 0.1839,
+      "step": 1270
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.255990982055664,
+      "learning_rate": 3.797746633690574e-05,
+      "loss": 0.1655,
+      "step": 1280
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.073582172393799,
+      "learning_rate": 3.784006595218467e-05,
+      "loss": 0.1873,
+      "step": 1290
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.076309204101562,
+      "learning_rate": 3.770266556746359e-05,
+      "loss": 0.165,
+      "step": 1300
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.572199821472168,
+      "learning_rate": 3.7565265182742515e-05,
+      "loss": 0.2013,
+      "step": 1310
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.93818473815918,
+      "learning_rate": 3.7427864798021435e-05,
+      "loss": 0.1843,
+      "step": 1320
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.742272853851318,
+      "learning_rate": 3.7290464413300355e-05,
+      "loss": 0.1641,
+      "step": 1330
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.64502477645874,
+      "learning_rate": 3.715306402857928e-05,
+      "loss": 0.1596,
+      "step": 1340
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.271630764007568,
+      "learning_rate": 3.701566364385821e-05,
+      "loss": 0.1956,
+      "step": 1350
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 23.52874755859375,
+      "learning_rate": 3.687826325913713e-05,
+      "loss": 0.1977,
+      "step": 1360
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 8.219966888427734,
+      "learning_rate": 3.674086287441605e-05,
+      "loss": 0.1911,
+      "step": 1370
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.677926540374756,
+      "learning_rate": 3.6603462489694974e-05,
+      "loss": 0.1718,
+      "step": 1380
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 12.710423469543457,
+      "learning_rate": 3.64660621049739e-05,
+      "loss": 0.1714,
+      "step": 1390
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 11.909724235534668,
+      "learning_rate": 3.632866172025282e-05,
+      "loss": 0.1563,
+      "step": 1400
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.356451988220215,
+      "learning_rate": 3.619126133553174e-05,
+      "loss": 0.1767,
+      "step": 1410
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.096663951873779,
+      "learning_rate": 3.6053860950810666e-05,
+      "loss": 0.2078,
+      "step": 1420
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 13.783403396606445,
+      "learning_rate": 3.5916460566089586e-05,
+      "loss": 0.2154,
+      "step": 1430
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.28662109375,
+      "learning_rate": 3.577906018136851e-05,
+      "loss": 0.2059,
+      "step": 1440
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 22.58875274658203,
+      "learning_rate": 3.564165979664743e-05,
+      "loss": 0.1753,
+      "step": 1450
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.86113452911377,
+      "learning_rate": 3.550425941192635e-05,
+      "loss": 0.1705,
+      "step": 1460
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.679204940795898,
+      "learning_rate": 3.536685902720528e-05,
+      "loss": 0.1485,
+      "step": 1470
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 12.120351791381836,
+      "learning_rate": 3.5229458642484205e-05,
+      "loss": 0.1955,
+      "step": 1480
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.561965942382812,
+      "learning_rate": 3.5092058257763125e-05,
+      "loss": 0.1358,
+      "step": 1490
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 11.117871284484863,
+      "learning_rate": 3.4954657873042044e-05,
+      "loss": 0.1895,
+      "step": 1500
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.080453872680664,
+      "learning_rate": 3.481725748832097e-05,
+      "loss": 0.1826,
+      "step": 1510
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.163525104522705,
+      "learning_rate": 3.46798571035999e-05,
+      "loss": 0.1514,
+      "step": 1520
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.087124824523926,
+      "learning_rate": 3.454245671887881e-05,
+      "loss": 0.1786,
+      "step": 1530
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.43030834197998,
+      "learning_rate": 3.440505633415774e-05,
+      "loss": 0.1511,
+      "step": 1540
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.804444313049316,
+      "learning_rate": 3.426765594943666e-05,
+      "loss": 0.1741,
+      "step": 1550
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 6.939291000366211,
+      "learning_rate": 3.413025556471558e-05,
+      "loss": 0.1643,
+      "step": 1560
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.169299125671387,
+      "learning_rate": 3.39928551799945e-05,
+      "loss": 0.1868,
+      "step": 1570
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.091197967529297,
+      "learning_rate": 3.385545479527343e-05,
+      "loss": 0.1785,
+      "step": 1580
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.629822731018066,
+      "learning_rate": 3.371805441055235e-05,
+      "loss": 0.1593,
+      "step": 1590
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.69659423828125,
+      "learning_rate": 3.3580654025831275e-05,
+      "loss": 0.1748,
+      "step": 1600
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.6397123336792,
+      "learning_rate": 3.3443253641110195e-05,
+      "loss": 0.1685,
+      "step": 1610
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.966397285461426,
+      "learning_rate": 3.330585325638912e-05,
+      "loss": 0.1493,
+      "step": 1620
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.841334342956543,
+      "learning_rate": 3.316845287166804e-05,
+      "loss": 0.1263,
+      "step": 1630
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.405898094177246,
+      "learning_rate": 3.303105248694697e-05,
+      "loss": 0.1668,
+      "step": 1640
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.914663791656494,
+      "learning_rate": 3.289365210222589e-05,
+      "loss": 0.1395,
+      "step": 1650
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 6.453075408935547,
+      "learning_rate": 3.275625171750481e-05,
+      "loss": 0.1282,
+      "step": 1660
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.115422248840332,
+      "learning_rate": 3.2618851332783734e-05,
+      "loss": 0.1569,
+      "step": 1670
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.41796588897705,
+      "learning_rate": 3.248145094806266e-05,
+      "loss": 0.1276,
+      "step": 1680
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.391621589660645,
+      "learning_rate": 3.234405056334158e-05,
+      "loss": 0.1462,
+      "step": 1690
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.930850982666016,
+      "learning_rate": 3.22066501786205e-05,
+      "loss": 0.1293,
+      "step": 1700
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.332276344299316,
+      "learning_rate": 3.2069249793899426e-05,
+      "loss": 0.1361,
+      "step": 1710
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.518467903137207,
+      "learning_rate": 3.193184940917835e-05,
+      "loss": 0.1934,
+      "step": 1720
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 4.080516338348389,
+      "learning_rate": 3.1794449024457266e-05,
+      "loss": 0.1816,
+      "step": 1730
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.780229568481445,
+      "learning_rate": 3.165704863973619e-05,
+      "loss": 0.1776,
+      "step": 1740
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.983511924743652,
+      "learning_rate": 3.151964825501512e-05,
+      "loss": 0.1745,
+      "step": 1750
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.346813201904297,
+      "learning_rate": 3.138224787029404e-05,
+      "loss": 0.1847,
+      "step": 1760
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.698652267456055,
+      "learning_rate": 3.124484748557296e-05,
+      "loss": 0.1398,
+      "step": 1770
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.913334846496582,
+      "learning_rate": 3.1107447100851885e-05,
+      "loss": 0.1128,
+      "step": 1780
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.111039161682129,
+      "learning_rate": 3.0970046716130804e-05,
+      "loss": 0.1507,
+      "step": 1790
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.033466339111328,
+      "learning_rate": 3.083264633140973e-05,
+      "loss": 0.1406,
+      "step": 1800
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 5.35683012008667,
+      "learning_rate": 3.069524594668865e-05,
+      "loss": 0.1157,
+      "step": 1810
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.188616752624512,
+      "learning_rate": 3.055784556196758e-05,
+      "loss": 0.1291,
+      "step": 1820
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 5.21785831451416,
+      "learning_rate": 3.0420445177246497e-05,
+      "loss": 0.1277,
+      "step": 1830
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 19.48537826538086,
+      "learning_rate": 3.0283044792525423e-05,
+      "loss": 0.182,
+      "step": 1840
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.281648635864258,
+      "learning_rate": 3.014564440780434e-05,
+      "loss": 0.1806,
+      "step": 1850
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 4.03902006149292,
+      "learning_rate": 3.0008244023083266e-05,
+      "loss": 0.1342,
+      "step": 1860
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.868727684020996,
+      "learning_rate": 2.987084363836219e-05,
+      "loss": 0.1153,
+      "step": 1870
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 14.366231918334961,
+      "learning_rate": 2.9733443253641112e-05,
+      "loss": 0.1657,
+      "step": 1880
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 5.684247016906738,
+      "learning_rate": 2.9596042868920032e-05,
+      "loss": 0.1473,
+      "step": 1890
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 20.257328033447266,
+      "learning_rate": 2.9458642484198955e-05,
+      "loss": 0.1679,
+      "step": 1900
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 6.537503719329834,
+      "learning_rate": 2.932124209947788e-05,
+      "loss": 0.1507,
+      "step": 1910
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 12.927555084228516,
+      "learning_rate": 2.9183841714756805e-05,
+      "loss": 0.1315,
+      "step": 1920
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 4.525102138519287,
+      "learning_rate": 2.9046441330035725e-05,
+      "loss": 0.1191,
+      "step": 1930
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 19.349470138549805,
+      "learning_rate": 2.8909040945314648e-05,
+      "loss": 0.1452,
+      "step": 1940
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.777169227600098,
+      "learning_rate": 2.877164056059357e-05,
+      "loss": 0.1601,
+      "step": 1950
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.030716896057129,
+      "learning_rate": 2.8634240175872494e-05,
+      "loss": 0.1621,
+      "step": 1960
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.78432846069336,
+      "learning_rate": 2.8496839791151414e-05,
+      "loss": 0.1157,
+      "step": 1970
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 4.950636386871338,
+      "learning_rate": 2.8359439406430337e-05,
+      "loss": 0.1395,
+      "step": 1980
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.61556625366211,
+      "learning_rate": 2.8222039021709263e-05,
+      "loss": 0.1555,
+      "step": 1990
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.099406242370605,
+      "learning_rate": 2.8084638636988186e-05,
+      "loss": 0.1071,
+      "step": 2000
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.480673789978027,
+      "learning_rate": 2.7947238252267106e-05,
+      "loss": 0.145,
+      "step": 2010
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.461160659790039,
+      "learning_rate": 2.780983786754603e-05,
+      "loss": 0.1923,
+      "step": 2020
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 4.666472434997559,
+      "learning_rate": 2.7672437482824952e-05,
+      "loss": 0.1391,
+      "step": 2030
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.440675735473633,
+      "learning_rate": 2.753503709810388e-05,
+      "loss": 0.1583,
+      "step": 2040
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 4.4610419273376465,
+      "learning_rate": 2.7397636713382795e-05,
+      "loss": 0.1152,
+      "step": 2050
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.9719390869140625,
+      "learning_rate": 2.726023632866172e-05,
+      "loss": 0.1262,
+      "step": 2060
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 18.907411575317383,
+      "learning_rate": 2.7122835943940645e-05,
+      "loss": 0.1407,
+      "step": 2070
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 4.285183906555176,
+      "learning_rate": 2.6985435559219568e-05,
+      "loss": 0.1365,
+      "step": 2080
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 12.173905372619629,
+      "learning_rate": 2.684803517449849e-05,
+      "loss": 0.1501,
+      "step": 2090
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 5.8501105308532715,
+      "learning_rate": 2.671063478977741e-05,
+      "loss": 0.1265,
+      "step": 2100
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.928159713745117,
+      "learning_rate": 2.6573234405056334e-05,
+      "loss": 0.1329,
+      "step": 2110
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.737021446228027,
+      "learning_rate": 2.643583402033526e-05,
+      "loss": 0.1083,
+      "step": 2120
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 16.16398811340332,
+      "learning_rate": 2.6298433635614183e-05,
+      "loss": 0.1667,
+      "step": 2130
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 5.912113189697266,
+      "learning_rate": 2.6161033250893103e-05,
+      "loss": 0.1451,
+      "step": 2140
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.185014724731445,
+      "learning_rate": 2.6023632866172026e-05,
+      "loss": 0.139,
+      "step": 2150
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.74959945678711,
+      "learning_rate": 2.588623248145095e-05,
+      "loss": 0.1468,
+      "step": 2160
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.38637638092041,
+      "learning_rate": 2.5748832096729876e-05,
+      "loss": 0.1313,
+      "step": 2170
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.455225944519043,
+      "learning_rate": 2.5611431712008792e-05,
+      "loss": 0.1507,
+      "step": 2180
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.756477355957031,
+      "learning_rate": 2.547403132728772e-05,
+      "loss": 0.1452,
+      "step": 2190
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.834283828735352,
+      "learning_rate": 2.5336630942566642e-05,
+      "loss": 0.1425,
+      "step": 2200
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 4.590889930725098,
+      "learning_rate": 2.5199230557845565e-05,
+      "loss": 0.1022,
+      "step": 2210
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.958761692047119,
+      "learning_rate": 2.5061830173124485e-05,
+      "loss": 0.1161,
+      "step": 2220
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.83637523651123,
+      "learning_rate": 2.4924429788403408e-05,
+      "loss": 0.1083,
+      "step": 2230
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.21489143371582,
+      "learning_rate": 2.478702940368233e-05,
+      "loss": 0.143,
+      "step": 2240
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.316998481750488,
+      "learning_rate": 2.4649629018961254e-05,
+      "loss": 0.1086,
+      "step": 2250
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.482443809509277,
+      "learning_rate": 2.4512228634240177e-05,
+      "loss": 0.1352,
+      "step": 2260
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.363387107849121,
+      "learning_rate": 2.43748282495191e-05,
+      "loss": 0.1204,
+      "step": 2270
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 5.618880271911621,
+      "learning_rate": 2.4237427864798023e-05,
+      "loss": 0.1257,
+      "step": 2280
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.133738040924072,
+      "learning_rate": 2.4100027480076946e-05,
+      "loss": 0.1277,
+      "step": 2290
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.639707565307617,
+      "learning_rate": 2.396262709535587e-05,
+      "loss": 0.1155,
+      "step": 2300
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.4000542163848877,
+      "learning_rate": 2.382522671063479e-05,
+      "loss": 0.1145,
+      "step": 2310
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 17.296470642089844,
+      "learning_rate": 2.3687826325913716e-05,
+      "loss": 0.1446,
+      "step": 2320
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 16.465299606323242,
+      "learning_rate": 2.3550425941192635e-05,
+      "loss": 0.1336,
+      "step": 2330
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.490622520446777,
+      "learning_rate": 2.341302555647156e-05,
+      "loss": 0.1473,
+      "step": 2340
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.074846267700195,
+      "learning_rate": 2.327562517175048e-05,
+      "loss": 0.1358,
+      "step": 2350
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.838306427001953,
+      "learning_rate": 2.3138224787029405e-05,
+      "loss": 0.1162,
+      "step": 2360
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 12.213912963867188,
+      "learning_rate": 2.3000824402308328e-05,
+      "loss": 0.1344,
+      "step": 2370
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.377532005310059,
+      "learning_rate": 2.286342401758725e-05,
+      "loss": 0.1264,
+      "step": 2380
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.104826927185059,
+      "learning_rate": 2.272602363286617e-05,
+      "loss": 0.0973,
+      "step": 2390
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.895303726196289,
+      "learning_rate": 2.2588623248145097e-05,
+      "loss": 0.1387,
+      "step": 2400
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.671829223632812,
+      "learning_rate": 2.2451222863424017e-05,
+      "loss": 0.1244,
+      "step": 2410
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.218570709228516,
+      "learning_rate": 2.2313822478702943e-05,
+      "loss": 0.1453,
+      "step": 2420
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.245031356811523,
+      "learning_rate": 2.2176422093981863e-05,
+      "loss": 0.1608,
+      "step": 2430
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.235481262207031,
+      "learning_rate": 2.2039021709260786e-05,
+      "loss": 0.1346,
+      "step": 2440
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 5.834726333618164,
+      "learning_rate": 2.190162132453971e-05,
+      "loss": 0.1034,
+      "step": 2450
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.327523231506348,
+      "learning_rate": 2.1764220939818633e-05,
+      "loss": 0.1543,
+      "step": 2460
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.398885726928711,
+      "learning_rate": 2.1626820555097556e-05,
+      "loss": 0.1395,
+      "step": 2470
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.02419900894165,
+      "learning_rate": 2.148942017037648e-05,
+      "loss": 0.1055,
+      "step": 2480
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.754438400268555,
+      "learning_rate": 2.13520197856554e-05,
+      "loss": 0.1464,
+      "step": 2490
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.242690086364746,
+      "learning_rate": 2.1214619400934325e-05,
+      "loss": 0.0901,
+      "step": 2500
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.369692325592041,
+      "learning_rate": 2.1077219016213245e-05,
+      "loss": 0.1085,
+      "step": 2510
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.569657802581787,
+      "learning_rate": 2.093981863149217e-05,
+      "loss": 0.0838,
+      "step": 2520
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.487847805023193,
+      "learning_rate": 2.080241824677109e-05,
+      "loss": 0.1121,
+      "step": 2530
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 12.342958450317383,
+      "learning_rate": 2.0665017862050014e-05,
+      "loss": 0.1124,
+      "step": 2540
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.757689476013184,
+      "learning_rate": 2.0527617477328937e-05,
+      "loss": 0.1579,
+      "step": 2550
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.211867332458496,
+      "learning_rate": 2.039021709260786e-05,
+      "loss": 0.1361,
+      "step": 2560
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 20.314220428466797,
+      "learning_rate": 2.0252816707886783e-05,
+      "loss": 0.0878,
+      "step": 2570
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.73823881149292,
+      "learning_rate": 2.0115416323165706e-05,
+      "loss": 0.0926,
+      "step": 2580
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 16.19614028930664,
+      "learning_rate": 1.9978015938444626e-05,
+      "loss": 0.1377,
+      "step": 2590
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.256732940673828,
+      "learning_rate": 1.9840615553723553e-05,
+      "loss": 0.119,
+      "step": 2600
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.039568901062012,
+      "learning_rate": 1.9703215169002472e-05,
+      "loss": 0.1266,
+      "step": 2610
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.308676719665527,
+      "learning_rate": 1.9565814784281396e-05,
+      "loss": 0.1175,
+      "step": 2620
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.749291896820068,
+      "learning_rate": 1.942841439956032e-05,
+      "loss": 0.0998,
+      "step": 2630
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.7631587982177734,
+      "learning_rate": 1.9291014014839242e-05,
+      "loss": 0.1354,
+      "step": 2640
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.042880058288574,
+      "learning_rate": 1.9153613630118165e-05,
+      "loss": 0.1219,
+      "step": 2650
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.238298416137695,
+      "learning_rate": 1.9016213245397088e-05,
+      "loss": 0.1127,
+      "step": 2660
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 12.992602348327637,
+      "learning_rate": 1.887881286067601e-05,
+      "loss": 0.1035,
+      "step": 2670
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.475639820098877,
+      "learning_rate": 1.8741412475954934e-05,
+      "loss": 0.1201,
+      "step": 2680
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.346205234527588,
+      "learning_rate": 1.8604012091233854e-05,
+      "loss": 0.1203,
+      "step": 2690
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.271622180938721,
+      "learning_rate": 1.846661170651278e-05,
+      "loss": 0.0947,
+      "step": 2700
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 14.107924461364746,
+      "learning_rate": 1.83292113217917e-05,
+      "loss": 0.1283,
+      "step": 2710
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.537943840026855,
+      "learning_rate": 1.8191810937070623e-05,
+      "loss": 0.1117,
+      "step": 2720
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 17.230525970458984,
+      "learning_rate": 1.805441055234955e-05,
+      "loss": 0.1102,
+      "step": 2730
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 8.807158470153809,
+      "learning_rate": 1.791701016762847e-05,
+      "loss": 0.0836,
+      "step": 2740
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 8.520951271057129,
+      "learning_rate": 1.7779609782907393e-05,
+      "loss": 0.0793,
+      "step": 2750
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.132307529449463,
+      "learning_rate": 1.7642209398186316e-05,
+      "loss": 0.1162,
+      "step": 2760
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 7.422785758972168,
+      "learning_rate": 1.750480901346524e-05,
+      "loss": 0.1083,
+      "step": 2770
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 13.516892433166504,
+      "learning_rate": 1.7367408628744162e-05,
+      "loss": 0.1172,
+      "step": 2780
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 9.948054313659668,
+      "learning_rate": 1.7230008244023085e-05,
+      "loss": 0.1015,
+      "step": 2790
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 13.648367881774902,
+      "learning_rate": 1.7092607859302008e-05,
+      "loss": 0.0986,
+      "step": 2800
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.1467533111572266,
+      "learning_rate": 1.695520747458093e-05,
+      "loss": 0.1082,
+      "step": 2810
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 15.577645301818848,
+      "learning_rate": 1.681780708985985e-05,
+      "loss": 0.0911,
+      "step": 2820
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 10.578394889831543,
+      "learning_rate": 1.6680406705138778e-05,
+      "loss": 0.0642,
+      "step": 2830
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 9.884262084960938,
+      "learning_rate": 1.6543006320417697e-05,
+      "loss": 0.0907,
+      "step": 2840
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 6.715592384338379,
+      "learning_rate": 1.640560593569662e-05,
+      "loss": 0.1092,
+      "step": 2850
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.891101121902466,
+      "learning_rate": 1.6268205550975543e-05,
+      "loss": 0.0708,
+      "step": 2860
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 13.76768684387207,
+      "learning_rate": 1.6130805166254467e-05,
+      "loss": 0.1056,
+      "step": 2870
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 9.504161834716797,
+      "learning_rate": 1.599340478153339e-05,
+      "loss": 0.0998,
+      "step": 2880
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 9.887553215026855,
+      "learning_rate": 1.5856004396812313e-05,
+      "loss": 0.0812,
+      "step": 2890
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 6.6718220710754395,
+      "learning_rate": 1.5718604012091236e-05,
+      "loss": 0.0834,
+      "step": 2900
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 14.240887641906738,
+      "learning_rate": 1.558120362737016e-05,
+      "loss": 0.0785,
+      "step": 2910
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 14.264277458190918,
+      "learning_rate": 1.544380324264908e-05,
+      "loss": 0.1214,
+      "step": 2920
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 12.33563232421875,
+      "learning_rate": 1.5306402857928005e-05,
+      "loss": 0.0939,
+      "step": 2930
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.850363731384277,
+      "learning_rate": 1.5169002473206925e-05,
+      "loss": 0.1282,
+      "step": 2940
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 8.227639198303223,
+      "learning_rate": 1.503160208848585e-05,
+      "loss": 0.0581,
+      "step": 2950
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 14.434782981872559,
+      "learning_rate": 1.4894201703764771e-05,
+      "loss": 0.0839,
+      "step": 2960
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 9.338940620422363,
+      "learning_rate": 1.4756801319043694e-05,
+      "loss": 0.0965,
+      "step": 2970
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 11.376910209655762,
+      "learning_rate": 1.4619400934322616e-05,
+      "loss": 0.0996,
+      "step": 2980
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 7.160622596740723,
+      "learning_rate": 1.448200054960154e-05,
+      "loss": 0.1127,
+      "step": 2990
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 8.001399993896484,
+      "learning_rate": 1.4344600164880462e-05,
+      "loss": 0.0752,
+      "step": 3000
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 7.866940498352051,
+      "learning_rate": 1.4207199780159385e-05,
+      "loss": 0.0881,
+      "step": 3010
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 6.259250164031982,
+      "learning_rate": 1.4069799395438306e-05,
+      "loss": 0.0699,
+      "step": 3020
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.861302375793457,
+      "learning_rate": 1.3932399010717231e-05,
+      "loss": 0.0994,
+      "step": 3030
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 7.015869140625,
+      "learning_rate": 1.3794998625996153e-05,
+      "loss": 0.0922,
+      "step": 3040
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 18.333284378051758,
+      "learning_rate": 1.3657598241275078e-05,
+      "loss": 0.1123,
+      "step": 3050
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 13.079123497009277,
+      "learning_rate": 1.3520197856553999e-05,
+      "loss": 0.1094,
+      "step": 3060
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.898919582366943,
+      "learning_rate": 1.3382797471832922e-05,
+      "loss": 0.0788,
+      "step": 3070
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 10.12296199798584,
+      "learning_rate": 1.3245397087111843e-05,
+      "loss": 0.0853,
+      "step": 3080
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 8.511975288391113,
+      "learning_rate": 1.3107996702390768e-05,
+      "loss": 0.0798,
+      "step": 3090
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 10.737293243408203,
+      "learning_rate": 1.297059631766969e-05,
+      "loss": 0.0599,
+      "step": 3100
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 19.816972732543945,
+      "learning_rate": 1.2833195932948613e-05,
+      "loss": 0.1426,
+      "step": 3110
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 7.945023059844971,
+      "learning_rate": 1.2695795548227534e-05,
+      "loss": 0.0782,
+      "step": 3120
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 13.234269142150879,
+      "learning_rate": 1.2558395163506459e-05,
+      "loss": 0.0848,
+      "step": 3130
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 8.509943008422852,
+      "learning_rate": 1.2420994778785382e-05,
+      "loss": 0.0942,
+      "step": 3140
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 6.774186134338379,
+      "learning_rate": 1.2283594394064305e-05,
+      "loss": 0.103,
+      "step": 3150
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 8.81127643585205,
+      "learning_rate": 1.2146194009343227e-05,
+      "loss": 0.115,
+      "step": 3160
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.346981525421143,
+      "learning_rate": 1.200879362462215e-05,
+      "loss": 0.082,
+      "step": 3170
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 8.829075813293457,
+      "learning_rate": 1.1871393239901073e-05,
+      "loss": 0.0779,
+      "step": 3180
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 12.217733383178711,
+      "learning_rate": 1.1733992855179996e-05,
+      "loss": 0.0775,
+      "step": 3190
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.546054840087891,
+      "learning_rate": 1.1596592470458917e-05,
+      "loss": 0.0945,
+      "step": 3200
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.3229305744171143,
+      "learning_rate": 1.145919208573784e-05,
+      "loss": 0.103,
+      "step": 3210
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.730965614318848,
+      "learning_rate": 1.1321791701016764e-05,
+      "loss": 0.0687,
+      "step": 3220
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 7.760990142822266,
+      "learning_rate": 1.1184391316295687e-05,
+      "loss": 0.0924,
+      "step": 3230
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 15.921736717224121,
+      "learning_rate": 1.104699093157461e-05,
+      "loss": 0.1069,
+      "step": 3240
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 9.555643081665039,
+      "learning_rate": 1.0909590546853531e-05,
+      "loss": 0.0623,
+      "step": 3250
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 7.185321807861328,
+      "learning_rate": 1.0772190162132454e-05,
+      "loss": 0.0801,
+      "step": 3260
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 7.710333347320557,
+      "learning_rate": 1.0634789777411378e-05,
+      "loss": 0.128,
+      "step": 3270
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 8.165462493896484,
+      "learning_rate": 1.04973893926903e-05,
+      "loss": 0.0921,
+      "step": 3280
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.764280319213867,
+      "learning_rate": 1.0359989007969224e-05,
+      "loss": 0.0803,
+      "step": 3290
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 7.214282512664795,
+      "learning_rate": 1.0222588623248145e-05,
+      "loss": 0.0735,
+      "step": 3300
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 9.242117881774902,
+      "learning_rate": 1.0085188238527068e-05,
+      "loss": 0.0757,
+      "step": 3310
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 19.152359008789062,
+      "learning_rate": 9.947787853805991e-06,
+      "loss": 0.061,
+      "step": 3320
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.347075462341309,
+      "learning_rate": 9.810387469084915e-06,
+      "loss": 0.0939,
+      "step": 3330
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 8.318592071533203,
+      "learning_rate": 9.672987084363836e-06,
+      "loss": 0.1105,
+      "step": 3340
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 6.497332572937012,
+      "learning_rate": 9.535586699642759e-06,
+      "loss": 0.0843,
+      "step": 3350
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 7.793337345123291,
+      "learning_rate": 9.398186314921682e-06,
+      "loss": 0.1092,
+      "step": 3360
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.7840776443481445,
+      "learning_rate": 9.260785930200605e-06,
+      "loss": 0.0752,
+      "step": 3370
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.059652328491211,
+      "learning_rate": 9.123385545479528e-06,
+      "loss": 0.0853,
+      "step": 3380
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 11.957003593444824,
+      "learning_rate": 8.98598516075845e-06,
+      "loss": 0.0879,
+      "step": 3390
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 9.338296890258789,
+      "learning_rate": 8.848584776037373e-06,
+      "loss": 0.098,
+      "step": 3400
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 7.063011646270752,
+      "learning_rate": 8.711184391316296e-06,
+      "loss": 0.0933,
+      "step": 3410
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 12.401261329650879,
+      "learning_rate": 8.573784006595219e-06,
+      "loss": 0.0961,
+      "step": 3420
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 9.310322761535645,
+      "learning_rate": 8.436383621874142e-06,
+      "loss": 0.0942,
+      "step": 3430
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 7.748733043670654,
+      "learning_rate": 8.298983237153064e-06,
+      "loss": 0.106,
+      "step": 3440
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 9.764993667602539,
+      "learning_rate": 8.161582852431987e-06,
+      "loss": 0.0777,
+      "step": 3450
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.836904048919678,
+      "learning_rate": 8.02418246771091e-06,
+      "loss": 0.0868,
+      "step": 3460
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.264832973480225,
+      "learning_rate": 7.886782082989833e-06,
+      "loss": 0.085,
+      "step": 3470
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 11.601814270019531,
+      "learning_rate": 7.749381698268756e-06,
+      "loss": 0.0854,
+      "step": 3480
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 8.948418617248535,
+      "learning_rate": 7.611981313547678e-06,
+      "loss": 0.0769,
+      "step": 3490
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 2.6163172721862793,
+      "learning_rate": 7.474580928826601e-06,
+      "loss": 0.0798,
+      "step": 3500
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 6.531749725341797,
+      "learning_rate": 7.337180544105524e-06,
+      "loss": 0.0745,
+      "step": 3510
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 17.10733413696289,
+      "learning_rate": 7.199780159384446e-06,
+      "loss": 0.0901,
+      "step": 3520
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 58.79534912109375,
+      "learning_rate": 7.062379774663369e-06,
+      "loss": 0.0985,
+      "step": 3530
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 19.27933120727539,
+      "learning_rate": 6.924979389942292e-06,
+      "loss": 0.0786,
+      "step": 3540
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 10.52169418334961,
+      "learning_rate": 6.7875790052212145e-06,
+      "loss": 0.0853,
+      "step": 3550
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.8546366691589355,
+      "learning_rate": 6.650178620500138e-06,
+      "loss": 0.0854,
+      "step": 3560
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 7.258632659912109,
+      "learning_rate": 6.51277823577906e-06,
+      "loss": 0.1071,
+      "step": 3570
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 9.697196960449219,
+      "learning_rate": 6.375377851057983e-06,
+      "loss": 0.1049,
+      "step": 3580
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.308481216430664,
+      "learning_rate": 6.237977466336906e-06,
+      "loss": 0.0906,
+      "step": 3590
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.4868924617767334,
+      "learning_rate": 6.100577081615829e-06,
+      "loss": 0.0831,
+      "step": 3600
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 8.134696960449219,
+      "learning_rate": 5.9631766968947515e-06,
+      "loss": 0.0919,
+      "step": 3610
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 7.263360977172852,
+      "learning_rate": 5.825776312173675e-06,
+      "loss": 0.0886,
+      "step": 3620
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 7.543405532836914,
+      "learning_rate": 5.688375927452598e-06,
+      "loss": 0.0596,
+      "step": 3630
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 7.298725605010986,
+      "learning_rate": 5.55097554273152e-06,
+      "loss": 0.1053,
+      "step": 3640
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 10.91873836517334,
+      "learning_rate": 5.413575158010443e-06,
+      "loss": 0.0948,
+      "step": 3650
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 10.322714805603027,
+      "learning_rate": 5.276174773289365e-06,
+      "loss": 0.093,
+      "step": 3660
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.865463733673096,
+      "learning_rate": 5.1387743885682885e-06,
+      "loss": 0.0849,
+      "step": 3670
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 9.305975914001465,
+      "learning_rate": 5.001374003847211e-06,
+      "loss": 0.0914,
+      "step": 3680
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.644111156463623,
+      "learning_rate": 4.863973619126134e-06,
+      "loss": 0.0606,
+      "step": 3690
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 6.5188374519348145,
+      "learning_rate": 4.726573234405057e-06,
+      "loss": 0.0926,
+      "step": 3700
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.520944118499756,
+      "learning_rate": 4.589172849683979e-06,
+      "loss": 0.0714,
+      "step": 3710
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 15.523458480834961,
+      "learning_rate": 4.451772464962902e-06,
+      "loss": 0.0757,
+      "step": 3720
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.631608486175537,
+      "learning_rate": 4.314372080241825e-06,
+      "loss": 0.0824,
+      "step": 3730
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 6.842602729797363,
+      "learning_rate": 4.176971695520748e-06,
+      "loss": 0.078,
+      "step": 3740
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 9.014209747314453,
+      "learning_rate": 4.03957131079967e-06,
+      "loss": 0.087,
+      "step": 3750
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 8.050474166870117,
+      "learning_rate": 3.902170926078593e-06,
+      "loss": 0.1036,
+      "step": 3760
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 15.050945281982422,
+      "learning_rate": 3.7647705413575158e-06,
+      "loss": 0.0673,
+      "step": 3770
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 16.99614715576172,
+      "learning_rate": 3.6273701566364385e-06,
+      "loss": 0.0791,
+      "step": 3780
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 6.738255023956299,
+      "learning_rate": 3.489969771915361e-06,
+      "loss": 0.0891,
+      "step": 3790
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 11.865900039672852,
+      "learning_rate": 3.3525693871942843e-06,
+      "loss": 0.0668,
+      "step": 3800
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 8.240047454833984,
+      "learning_rate": 3.215169002473207e-06,
+      "loss": 0.0603,
+      "step": 3810
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 8.264572143554688,
+      "learning_rate": 3.07776861775213e-06,
+      "loss": 0.0736,
+      "step": 3820
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.329923152923584,
+      "learning_rate": 2.9403682330310528e-06,
+      "loss": 0.0638,
+      "step": 3830
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 13.339191436767578,
+      "learning_rate": 2.8029678483099755e-06,
+      "loss": 0.0904,
+      "step": 3840
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.455562591552734,
+      "learning_rate": 2.665567463588898e-06,
+      "loss": 0.05,
+      "step": 3850
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 2.050187110900879,
+      "learning_rate": 2.528167078867821e-06,
+      "loss": 0.0449,
+      "step": 3860
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 6.172808647155762,
+      "learning_rate": 2.390766694146744e-06,
+      "loss": 0.1031,
+      "step": 3870
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.912864685058594,
+      "learning_rate": 2.2533663094256666e-06,
+      "loss": 0.0684,
+      "step": 3880
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.280465126037598,
+      "learning_rate": 2.1159659247045893e-06,
+      "loss": 0.0538,
+      "step": 3890
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 14.526315689086914,
+      "learning_rate": 1.978565539983512e-06,
+      "loss": 0.067,
+      "step": 3900
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 10.776835441589355,
+      "learning_rate": 1.8411651552624347e-06,
+      "loss": 0.0734,
+      "step": 3910
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 11.692980766296387,
+      "learning_rate": 1.7037647705413576e-06,
+      "loss": 0.0915,
+      "step": 3920
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 9.303447723388672,
+      "learning_rate": 1.5663643858202803e-06,
+      "loss": 0.0685,
+      "step": 3930
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 7.01151704788208,
+      "learning_rate": 1.4289640010992032e-06,
+      "loss": 0.0851,
+      "step": 3940
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 8.348711013793945,
+      "learning_rate": 1.2915636163781259e-06,
+      "loss": 0.0597,
+      "step": 3950
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 5.007867813110352,
+      "learning_rate": 1.1541632316570488e-06,
+      "loss": 0.0852,
+      "step": 3960
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 11.604133605957031,
+      "learning_rate": 1.0167628469359715e-06,
+      "loss": 0.0728,
+      "step": 3970
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 12.828524589538574,
+      "learning_rate": 8.793624622148942e-07,
+      "loss": 0.0582,
+      "step": 3980
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 7.547138214111328,
+      "learning_rate": 7.419620774938171e-07,
+      "loss": 0.0695,
+      "step": 3990
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 8.606446266174316,
+      "learning_rate": 6.045616927727397e-07,
+      "loss": 0.0969,
+      "step": 4000
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 10.521493911743164,
+      "learning_rate": 4.6716130805166254e-07,
+      "loss": 0.0555,
+      "step": 4010
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.8085311651229858,
+      "learning_rate": 3.2976092333058533e-07,
+      "loss": 0.0662,
+      "step": 4020
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 3.2384657859802246,
+      "learning_rate": 1.9236053860950813e-07,
+      "loss": 0.0746,
+      "step": 4030
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 9.941252708435059,
+      "learning_rate": 5.4960153888430885e-08,
+      "loss": 0.0764,
+      "step": 4040
+    },
+    {
+      "epoch": 3.0,
+      "step": 4044,
+      "total_flos": 6.432009125858943e+18,
+      "train_loss": 0.17630483790067725,
+      "train_runtime": 837.6448,
+      "train_samples_per_second": 308.927,
+      "train_steps_per_second": 4.828
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 4044,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "total_flos": 6.432009125858943e+18,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b7daf8818538719d75675ac759b3f10732a1d427631d61ac3853dfaaaa40136
+size 4984