From ef64093785d45c44bab061594f5bb2db9e55b09d Mon Sep 17 00:00:00 2001
From: YYJ-aaaa <1058387004@qq.com>
Date: Wed, 25 Dec 2024 17:39:32 +0800
Subject: [PATCH] first commit

---
 .gitignore               |    1 +
 all_results.json         |   13 +
 config.json              |   34 +
 eval_results.json        |    8 +
 preprocessor_config.json |   22 +
 pytorch_model.bin        |    3 +
 train_results.json       |    8 +
 trainer_state.json       | 3475 ++++++++++++++++++++++++++++++++++++++
 training_args.bin        |    3 +
 9 files changed, 3567 insertions(+)
 create mode 100644 .gitignore
 create mode 100644 all_results.json
 create mode 100644 config.json
 create mode 100644 eval_results.json
 create mode 100644 preprocessor_config.json
 create mode 100644 pytorch_model.bin
 create mode 100644 train_results.json
 create mode 100644 trainer_state.json
 create mode 100644 training_args.bin

diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..0348ea9
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1 @@
+checkpoint-*/
\ No newline at end of file
diff --git a/all_results.json b/all_results.json
new file mode 100644
index 0000000..449120a
--- /dev/null
+++ b/all_results.json
@@ -0,0 +1,13 @@
+{
+    "epoch": 2.0,
+    "eval_accuracy": 0.9043314126763362,
+    "eval_loss": 0.23999236524105072,
+    "eval_runtime": 93.6052,
+    "eval_samples_per_second": 107.537,
+    "eval_steps_per_second": 3.365,
+    "total_flos": 5.679462069216204e+19,
+    "train_loss": 0.1848815517786954,
+    "train_runtime": 16137.6204,
+    "train_samples_per_second": 45.426,
+    "train_steps_per_second": 0.355
+}
\ No newline at end of file
diff --git a/config.json b/config.json
new file mode 100644
index 0000000..73e514e
--- /dev/null
+++ b/config.json
@@ -0,0 +1,34 @@
+{
+  "_name_or_path": "google/vit-base-patch16-224-in21k",
+  "architectures": [
+    "ViTForImageClassification"
+  ],
+  "attention_probs_dropout_prob": 0.0,
+  "encoder_stride": 16,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.0,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "nude",
+    "1": "safe",
+    "2": "sexy"
+  },
+  "image_size": 224,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "nude": 0,
+    "safe": 1,
+    "sexy": 2
+  },
+  "layer_norm_eps": 1e-12,
+  "model_type": "vit",
+  "num_attention_heads": 12,
+  "num_channels": 3,
+  "num_hidden_layers": 12,
+  "patch_size": 16,
+  "problem_type": "single_label_classification",
+  "qkv_bias": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.28.0"
+}
diff --git a/eval_results.json b/eval_results.json
new file mode 100644
index 0000000..294988d
--- /dev/null
+++ b/eval_results.json
@@ -0,0 +1,8 @@
+{
+    "epoch": 2.0,
+    "eval_accuracy": 0.9043314126763362,
+    "eval_loss": 0.23999236524105072,
+    "eval_runtime": 93.6052,
+    "eval_samples_per_second": 107.537,
+    "eval_steps_per_second": 3.365
+}
\ No newline at end of file
diff --git a/preprocessor_config.json b/preprocessor_config.json
new file mode 100644
index 0000000..273d249
--- /dev/null
+++ b/preprocessor_config.json
@@ -0,0 +1,22 @@
+{
+  "do_normalize": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "image_processor_type": "ViTImageProcessor",
+  "image_std": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "resample": 2,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "height": 224,
+    "width": 224
+  }
+}
diff --git a/pytorch_model.bin b/pytorch_model.bin
new file mode 100644
index 0000000..2a217d0
--- /dev/null
+++ b/pytorch_model.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:167f2eb003e1b00a5c186be8feac4feb01b92d762212cf101a2efc51cc867a6e
+size 343271789
diff --git a/train_results.json b/train_results.json
new file mode 100644
index 0000000..c56de64
--- /dev/null
+++ b/train_results.json
@@ -0,0 +1,8 @@
+{
+    "epoch": 2.0,
+    "total_flos": 5.679462069216204e+19,
+    "train_loss": 0.1848815517786954,
+    "train_runtime": 16137.6204,
+    "train_samples_per_second": 45.426,
+    "train_steps_per_second": 0.355
+}
\ No newline at end of file
diff --git a/trainer_state.json b/trainer_state.json
new file mode 100644
index 0000000..2896c68
--- /dev/null
+++ b/trainer_state.json
@@ -0,0 +1,3475 @@
+{
+  "best_metric": 0.9043314126763362,
+  "best_model_checkpoint": "vit-base-patch16-224-in21k-finetuned-moderation/checkpoint-5726",
+  "epoch": 1.999476211261458,
+  "global_step": 5726,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.726003490401396e-07,
+      "loss": 1.0828,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7452006980802793e-06,
+      "loss": 1.081,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.617801047120419e-06,
+      "loss": 1.062,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4904013961605585e-06,
+      "loss": 1.0373,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.363001745200698e-06,
+      "loss": 0.9988,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.235602094240838e-06,
+      "loss": 0.9547,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.108202443280978e-06,
+      "loss": 0.8933,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.980802792321117e-06,
+      "loss": 0.8187,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.853403141361257e-06,
+      "loss": 0.7599,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.726003490401396e-06,
+      "loss": 0.7437,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.598603839441536e-06,
+      "loss": 0.6794,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.0471204188481676e-05,
+      "loss": 0.6599,
+      "step": 120
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1343804537521815e-05,
+      "loss": 0.6129,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.2216404886561955e-05,
+      "loss": 0.5633,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.3089005235602096e-05,
+      "loss": 0.535,
+      "step": 150
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3961605584642234e-05,
+      "loss": 0.4764,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.4834205933682374e-05,
+      "loss": 0.4337,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.5706806282722515e-05,
+      "loss": 0.4468,
+      "step": 180
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.6579406631762653e-05,
+      "loss": 0.4088,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.7452006980802792e-05,
+      "loss": 0.4164,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.8324607329842934e-05,
+      "loss": 0.3755,
+      "step": 210
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9197207678883072e-05,
+      "loss": 0.3639,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.006980802792321e-05,
+      "loss": 0.3615,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.0942408376963353e-05,
+      "loss": 0.3518,
+      "step": 240
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.181500872600349e-05,
+      "loss": 0.3331,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.268760907504363e-05,
+      "loss": 0.3374,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.3560209424083772e-05,
+      "loss": 0.3198,
+      "step": 270
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.443280977312391e-05,
+      "loss": 0.3014,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.5305410122164053e-05,
+      "loss": 0.3203,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.617801047120419e-05,
+      "loss": 0.317,
+      "step": 300
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.7050610820244333e-05,
+      "loss": 0.3067,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.792321116928447e-05,
+      "loss": 0.2929,
+      "step": 320
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.879581151832461e-05,
+      "loss": 0.2768,
+      "step": 330
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.966841186736475e-05,
+      "loss": 0.3025,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 3.054101221640489e-05,
+      "loss": 0.2868,
+      "step": 350
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 3.141361256544503e-05,
+      "loss": 0.3051,
+      "step": 360
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 3.228621291448517e-05,
+      "loss": 0.2929,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 3.3158813263525307e-05,
+      "loss": 0.2818,
+      "step": 380
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 3.403141361256545e-05,
+      "loss": 0.2739,
+      "step": 390
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 3.4904013961605584e-05,
+      "loss": 0.237,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 3.5776614310645726e-05,
+      "loss": 0.2697,
+      "step": 410
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 3.664921465968587e-05,
+      "loss": 0.3051,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 3.752181500872601e-05,
+      "loss": 0.2436,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 3.8394415357766145e-05,
+      "loss": 0.2459,
+      "step": 440
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 3.926701570680629e-05,
+      "loss": 0.2642,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.013961605584642e-05,
+      "loss": 0.2364,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1012216404886564e-05,
+      "loss": 0.2587,
+      "step": 470
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1884816753926706e-05,
+      "loss": 0.255,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.275741710296685e-05,
+      "loss": 0.2729,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.363001745200698e-05,
+      "loss": 0.2853,
+      "step": 500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.4502617801047125e-05,
+      "loss": 0.2502,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.537521815008726e-05,
+      "loss": 0.2967,
+      "step": 520
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.62478184991274e-05,
+      "loss": 0.2541,
+      "step": 530
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.7120418848167544e-05,
+      "loss": 0.2287,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.7993019197207686e-05,
+      "loss": 0.2518,
+      "step": 550
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.886561954624782e-05,
+      "loss": 0.2516,
+      "step": 560
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.973821989528796e-05,
+      "loss": 0.2355,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.99320784009315e-05,
+      "loss": 0.2299,
+      "step": 580
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.983504754511935e-05,
+      "loss": 0.2436,
+      "step": 590
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.973801668930721e-05,
+      "loss": 0.2382,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.964098583349505e-05,
+      "loss": 0.2596,
+      "step": 610
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.954395497768291e-05,
+      "loss": 0.2374,
+      "step": 620
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.9446924121870756e-05,
+      "loss": 0.1841,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.934989326605861e-05,
+      "loss": 0.2296,
+      "step": 640
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.9252862410246456e-05,
+      "loss": 0.243,
+      "step": 650
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.915583155443431e-05,
+      "loss": 0.197,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.905880069862216e-05,
+      "loss": 0.2505,
+      "step": 670
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.896176984281002e-05,
+      "loss": 0.2096,
+      "step": 680
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.886473898699787e-05,
+      "loss": 0.2395,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.876770813118572e-05,
+      "loss": 0.2487,
+      "step": 700
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.8670677275373574e-05,
+      "loss": 0.2439,
+      "step": 710
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.8573646419561424e-05,
+      "loss": 0.2214,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.8476615563749273e-05,
+      "loss": 0.1975,
+      "step": 730
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.837958470793712e-05,
+      "loss": 0.208,
+      "step": 740
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.828255385212498e-05,
+      "loss": 0.2596,
+      "step": 750
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.818552299631283e-05,
+      "loss": 0.2518,
+      "step": 760
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.808849214050068e-05,
+      "loss": 0.2186,
+      "step": 770
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.7991461284688535e-05,
+      "loss": 0.2157,
+      "step": 780
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.7894430428876385e-05,
+      "loss": 0.2105,
+      "step": 790
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.779739957306424e-05,
+      "loss": 0.2297,
+      "step": 800
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.7700368717252084e-05,
+      "loss": 0.2305,
+      "step": 810
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.760333786143994e-05,
+      "loss": 0.2191,
+      "step": 820
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.750630700562779e-05,
+      "loss": 0.234,
+      "step": 830
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.740927614981565e-05,
+      "loss": 0.2062,
+      "step": 840
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.731224529400349e-05,
+      "loss": 0.1978,
+      "step": 850
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.7215214438191346e-05,
+      "loss": 0.1953,
+      "step": 860
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.71181835823792e-05,
+      "loss": 0.2064,
+      "step": 870
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.702115272656705e-05,
+      "loss": 0.2081,
+      "step": 880
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.69241218707549e-05,
+      "loss": 0.2331,
+      "step": 890
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.682709101494275e-05,
+      "loss": 0.2039,
+      "step": 900
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.673006015913061e-05,
+      "loss": 0.2131,
+      "step": 910
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.663302930331846e-05,
+      "loss": 0.2211,
+      "step": 920
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.653599844750631e-05,
+      "loss": 0.2298,
+      "step": 930
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.643896759169416e-05,
+      "loss": 0.2118,
+      "step": 940
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.6341936735882014e-05,
+      "loss": 0.2067,
+      "step": 950
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.6244905880069863e-05,
+      "loss": 0.2129,
+      "step": 960
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.614787502425771e-05,
+      "loss": 0.1932,
+      "step": 970
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.605084416844557e-05,
+      "loss": 0.2309,
+      "step": 980
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.595381331263342e-05,
+      "loss": 0.2,
+      "step": 990
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.5856782456821276e-05,
+      "loss": 0.1866,
+      "step": 1000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.575975160100912e-05,
+      "loss": 0.1885,
+      "step": 1010
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.5662720745196975e-05,
+      "loss": 0.1993,
+      "step": 1020
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.5565689889384825e-05,
+      "loss": 0.1995,
+      "step": 1030
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.546865903357268e-05,
+      "loss": 0.1797,
+      "step": 1040
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.537162817776053e-05,
+      "loss": 0.2113,
+      "step": 1050
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.527459732194838e-05,
+      "loss": 0.2124,
+      "step": 1060
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.517756646613624e-05,
+      "loss": 0.2083,
+      "step": 1070
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.508053561032409e-05,
+      "loss": 0.1916,
+      "step": 1080
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.4983504754511936e-05,
+      "loss": 0.2094,
+      "step": 1090
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.4886473898699786e-05,
+      "loss": 0.1616,
+      "step": 1100
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.478944304288764e-05,
+      "loss": 0.1995,
+      "step": 1110
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.469241218707549e-05,
+      "loss": 0.2199,
+      "step": 1120
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.459538133126334e-05,
+      "loss": 0.2027,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.449835047545119e-05,
+      "loss": 0.1915,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.440131961963905e-05,
+      "loss": 0.1921,
+      "step": 1150
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.4304288763826904e-05,
+      "loss": 0.2095,
+      "step": 1160
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.420725790801475e-05,
+      "loss": 0.1852,
+      "step": 1170
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.4110227052202604e-05,
+      "loss": 0.2018,
+      "step": 1180
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.4013196196390453e-05,
+      "loss": 0.2073,
+      "step": 1190
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.391616534057831e-05,
+      "loss": 0.1873,
+      "step": 1200
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.381913448476615e-05,
+      "loss": 0.1941,
+      "step": 1210
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.372210362895401e-05,
+      "loss": 0.1835,
+      "step": 1220
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.362507277314186e-05,
+      "loss": 0.1873,
+      "step": 1230
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.3528041917329715e-05,
+      "loss": 0.1943,
+      "step": 1240
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.3431011061517565e-05,
+      "loss": 0.1754,
+      "step": 1250
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.3333980205705415e-05,
+      "loss": 0.2008,
+      "step": 1260
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.323694934989327e-05,
+      "loss": 0.1908,
+      "step": 1270
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.313991849408112e-05,
+      "loss": 0.2015,
+      "step": 1280
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.304288763826897e-05,
+      "loss": 0.1948,
+      "step": 1290
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.294585678245682e-05,
+      "loss": 0.1714,
+      "step": 1300
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.284882592664468e-05,
+      "loss": 0.191,
+      "step": 1310
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.2751795070832526e-05,
+      "loss": 0.1686,
+      "step": 1320
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.2654764215020376e-05,
+      "loss": 0.1814,
+      "step": 1330
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.255773335920823e-05,
+      "loss": 0.2003,
+      "step": 1340
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.246070250339608e-05,
+      "loss": 0.1778,
+      "step": 1350
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.236367164758394e-05,
+      "loss": 0.1706,
+      "step": 1360
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.226664079177178e-05,
+      "loss": 0.2011,
+      "step": 1370
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.216960993595964e-05,
+      "loss": 0.1996,
+      "step": 1380
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.207257908014749e-05,
+      "loss": 0.2055,
+      "step": 1390
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.1975548224335344e-05,
+      "loss": 0.2081,
+      "step": 1400
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.187851736852319e-05,
+      "loss": 0.1686,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.178148651271104e-05,
+      "loss": 0.1796,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.168445565689889e-05,
+      "loss": 0.1765,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.158742480108675e-05,
+      "loss": 0.152,
+      "step": 1440
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.14903939452746e-05,
+      "loss": 0.1917,
+      "step": 1450
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.139336308946245e-05,
+      "loss": 0.1845,
+      "step": 1460
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.1296332233650305e-05,
+      "loss": 0.1704,
+      "step": 1470
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.1199301377838155e-05,
+      "loss": 0.1952,
+      "step": 1480
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.1102270522026005e-05,
+      "loss": 0.1907,
+      "step": 1490
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.1005239666213854e-05,
+      "loss": 0.1732,
+      "step": 1500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.090820881040171e-05,
+      "loss": 0.175,
+      "step": 1510
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.081117795458956e-05,
+      "loss": 0.1638,
+      "step": 1520
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.071414709877741e-05,
+      "loss": 0.1775,
+      "step": 1530
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.0617116242965267e-05,
+      "loss": 0.171,
+      "step": 1540
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.0520085387153116e-05,
+      "loss": 0.1805,
+      "step": 1550
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.042305453134097e-05,
+      "loss": 0.1716,
+      "step": 1560
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.0326023675528816e-05,
+      "loss": 0.1651,
+      "step": 1570
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.022899281971667e-05,
+      "loss": 0.1722,
+      "step": 1580
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.013196196390452e-05,
+      "loss": 0.1779,
+      "step": 1590
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.003493110809238e-05,
+      "loss": 0.1594,
+      "step": 1600
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 3.993790025228023e-05,
+      "loss": 0.1814,
+      "step": 1610
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 3.984086939646808e-05,
+      "loss": 0.1671,
+      "step": 1620
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 3.9743838540655934e-05,
+      "loss": 0.2036,
+      "step": 1630
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 3.9646807684843784e-05,
+      "loss": 0.1786,
+      "step": 1640
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.954977682903163e-05,
+      "loss": 0.1923,
+      "step": 1650
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.945274597321948e-05,
+      "loss": 0.186,
+      "step": 1660
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.935571511740734e-05,
+      "loss": 0.1776,
+      "step": 1670
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.925868426159519e-05,
+      "loss": 0.1803,
+      "step": 1680
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.916165340578304e-05,
+      "loss": 0.1813,
+      "step": 1690
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.906462254997089e-05,
+      "loss": 0.166,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.8967591694158745e-05,
+      "loss": 0.1538,
+      "step": 1710
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.8870560838346595e-05,
+      "loss": 0.1866,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.8773529982534444e-05,
+      "loss": 0.1952,
+      "step": 1730
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.86764991267223e-05,
+      "loss": 0.1801,
+      "step": 1740
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.857946827091015e-05,
+      "loss": 0.1753,
+      "step": 1750
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.848243741509801e-05,
+      "loss": 0.1811,
+      "step": 1760
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.838540655928585e-05,
+      "loss": 0.1809,
+      "step": 1770
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.8288375703473706e-05,
+      "loss": 0.1734,
+      "step": 1780
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.8191344847661556e-05,
+      "loss": 0.1625,
+      "step": 1790
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.809431399184941e-05,
+      "loss": 0.1781,
+      "step": 1800
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.799728313603726e-05,
+      "loss": 0.1679,
+      "step": 1810
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.790025228022511e-05,
+      "loss": 0.1781,
+      "step": 1820
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.780322142441297e-05,
+      "loss": 0.189,
+      "step": 1830
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.770619056860082e-05,
+      "loss": 0.1679,
+      "step": 1840
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.760915971278867e-05,
+      "loss": 0.2142,
+      "step": 1850
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.751212885697652e-05,
+      "loss": 0.175,
+      "step": 1860
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.7415098001164374e-05,
+      "loss": 0.172,
+      "step": 1870
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.731806714535222e-05,
+      "loss": 0.1889,
+      "step": 1880
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.722103628954007e-05,
+      "loss": 0.1787,
+      "step": 1890
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.712400543372793e-05,
+      "loss": 0.1739,
+      "step": 1900
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.702697457791578e-05,
+      "loss": 0.1479,
+      "step": 1910
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.6929943722103636e-05,
+      "loss": 0.1857,
+      "step": 1920
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.683291286629148e-05,
+      "loss": 0.1719,
+      "step": 1930
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.6735882010479335e-05,
+      "loss": 0.1562,
+      "step": 1940
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.6638851154667185e-05,
+      "loss": 0.1666,
+      "step": 1950
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.654182029885504e-05,
+      "loss": 0.1932,
+      "step": 1960
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.6444789443042884e-05,
+      "loss": 0.1807,
+      "step": 1970
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.634775858723074e-05,
+      "loss": 0.1675,
+      "step": 1980
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.625072773141859e-05,
+      "loss": 0.1641,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.6153696875606447e-05,
+      "loss": 0.1436,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.6056666019794296e-05,
+      "loss": 0.1525,
+      "step": 2010
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.5959635163982146e-05,
+      "loss": 0.1736,
+      "step": 2020
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.586260430817e-05,
+      "loss": 0.1538,
+      "step": 2030
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.576557345235785e-05,
+      "loss": 0.1819,
+      "step": 2040
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.56685425965457e-05,
+      "loss": 0.1531,
+      "step": 2050
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.557151174073355e-05,
+      "loss": 0.1644,
+      "step": 2060
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.547448088492141e-05,
+      "loss": 0.1665,
+      "step": 2070
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.537745002910926e-05,
+      "loss": 0.1295,
+      "step": 2080
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.528041917329711e-05,
+      "loss": 0.1561,
+      "step": 2090
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.5183388317484964e-05,
+      "loss": 0.1988,
+      "step": 2100
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.508635746167281e-05,
+      "loss": 0.1848,
+      "step": 2110
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.498932660586067e-05,
+      "loss": 0.1825,
+      "step": 2120
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.489229575004851e-05,
+      "loss": 0.1895,
+      "step": 2130
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.479526489423637e-05,
+      "loss": 0.1784,
+      "step": 2140
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.469823403842422e-05,
+      "loss": 0.1624,
+      "step": 2150
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.4601203182612075e-05,
+      "loss": 0.1735,
+      "step": 2160
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.4504172326799925e-05,
+      "loss": 0.1705,
+      "step": 2170
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.4407141470987775e-05,
+      "loss": 0.1667,
+      "step": 2180
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.431011061517563e-05,
+      "loss": 0.1635,
+      "step": 2190
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.421307975936348e-05,
+      "loss": 0.1938,
+      "step": 2200
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.411604890355133e-05,
+      "loss": 0.1608,
+      "step": 2210
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.401901804773918e-05,
+      "loss": 0.1671,
+      "step": 2220
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.3921987191927037e-05,
+      "loss": 0.1666,
+      "step": 2230
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.3824956336114886e-05,
+      "loss": 0.1382,
+      "step": 2240
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.3727925480302736e-05,
+      "loss": 0.1573,
+      "step": 2250
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.3630894624490586e-05,
+      "loss": 0.1664,
+      "step": 2260
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.353386376867844e-05,
+      "loss": 0.1631,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.343683291286629e-05,
+      "loss": 0.1537,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.333980205705414e-05,
+      "loss": 0.1482,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.3242771201242e-05,
+      "loss": 0.1735,
+      "step": 2300
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.314574034542985e-05,
+      "loss": 0.1796,
+      "step": 2310
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.3048709489617704e-05,
+      "loss": 0.1426,
+      "step": 2320
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.295167863380555e-05,
+      "loss": 0.1716,
+      "step": 2330
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.28546477779934e-05,
+      "loss": 0.1631,
+      "step": 2340
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.275761692218125e-05,
+      "loss": 0.193,
+      "step": 2350
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.266058606636911e-05,
+      "loss": 0.165,
+      "step": 2360
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.256355521055696e-05,
+      "loss": 0.1655,
+      "step": 2370
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.246652435474481e-05,
+      "loss": 0.1994,
+      "step": 2380
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.2369493498932665e-05,
+      "loss": 0.1655,
+      "step": 2390
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.2272462643120515e-05,
+      "loss": 0.1543,
+      "step": 2400
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.2175431787308365e-05,
+      "loss": 0.1596,
+      "step": 2410
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.2078400931496214e-05,
+      "loss": 0.1665,
+      "step": 2420
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.198137007568407e-05,
+      "loss": 0.1687,
+      "step": 2430
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.188433921987192e-05,
+      "loss": 0.1689,
+      "step": 2440
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.178730836405977e-05,
+      "loss": 0.1698,
+      "step": 2450
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.1690277508247626e-05,
+      "loss": 0.1682,
+      "step": 2460
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.1593246652435476e-05,
+      "loss": 0.1768,
+      "step": 2470
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.149621579662333e-05,
+      "loss": 0.142,
+      "step": 2480
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.1399184940811176e-05,
+      "loss": 0.1688,
+      "step": 2490
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.130215408499903e-05,
+      "loss": 0.1472,
+      "step": 2500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.120512322918688e-05,
+      "loss": 0.1714,
+      "step": 2510
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.110809237337474e-05,
+      "loss": 0.1522,
+      "step": 2520
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.101106151756258e-05,
+      "loss": 0.1856,
+      "step": 2530
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.091403066175044e-05,
+      "loss": 0.1556,
+      "step": 2540
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.081699980593829e-05,
+      "loss": 0.1502,
+      "step": 2550
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.0719968950126144e-05,
+      "loss": 0.1358,
+      "step": 2560
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.062293809431399e-05,
+      "loss": 0.1574,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.052590723850184e-05,
+      "loss": 0.1469,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.0428876382689696e-05,
+      "loss": 0.1933,
+      "step": 2590
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.033184552687755e-05,
+      "loss": 0.1413,
+      "step": 2600
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.02348146710654e-05,
+      "loss": 0.1595,
+      "step": 2610
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.0137783815253252e-05,
+      "loss": 0.1734,
+      "step": 2620
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.0040752959441105e-05,
+      "loss": 0.1558,
+      "step": 2630
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.9943722103628958e-05,
+      "loss": 0.1594,
+      "step": 2640
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.9846691247816804e-05,
+      "loss": 0.1411,
+      "step": 2650
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.9749660392004657e-05,
+      "loss": 0.1858,
+      "step": 2660
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.965262953619251e-05,
+      "loss": 0.1753,
+      "step": 2670
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.9555598680380363e-05,
+      "loss": 0.1601,
+      "step": 2680
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.9458567824568213e-05,
+      "loss": 0.155,
+      "step": 2690
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.9361536968756066e-05,
+      "loss": 0.1487,
+      "step": 2700
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.926450611294392e-05,
+      "loss": 0.1452,
+      "step": 2710
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.9167475257131772e-05,
+      "loss": 0.1514,
+      "step": 2720
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.907044440131962e-05,
+      "loss": 0.1595,
+      "step": 2730
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.897341354550747e-05,
+      "loss": 0.1733,
+      "step": 2740
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.8876382689695325e-05,
+      "loss": 0.1536,
+      "step": 2750
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.8779351833883178e-05,
+      "loss": 0.1614,
+      "step": 2760
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.8682320978071024e-05,
+      "loss": 0.1642,
+      "step": 2770
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.8585290122258877e-05,
+      "loss": 0.1609,
+      "step": 2780
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.848825926644673e-05,
+      "loss": 0.1469,
+      "step": 2790
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.8391228410634587e-05,
+      "loss": 0.1546,
+      "step": 2800
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.8294197554822433e-05,
+      "loss": 0.1542,
+      "step": 2810
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.8197166699010286e-05,
+      "loss": 0.1557,
+      "step": 2820
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.810013584319814e-05,
+      "loss": 0.1644,
+      "step": 2830
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.8003104987385992e-05,
+      "loss": 0.1604,
+      "step": 2840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.790607413157384e-05,
+      "loss": 0.1812,
+      "step": 2850
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.780904327576169e-05,
+      "loss": 0.1295,
+      "step": 2860
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.8736340154977151,
+      "eval_loss": 0.31398746371269226,
+      "eval_runtime": 82.6529,
+      "eval_samples_per_second": 121.786,
+      "eval_steps_per_second": 3.811,
+      "step": 2863
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7712012419949544e-05,
+      "loss": 0.1444,
+      "step": 2870
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7614981564137398e-05,
+      "loss": 0.1578,
+      "step": 2880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7517950708325247e-05,
+      "loss": 0.1343,
+      "step": 2890
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.74209198525131e-05,
+      "loss": 0.1427,
+      "step": 2900
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7323888996700953e-05,
+      "loss": 0.1587,
+      "step": 2910
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7226858140888806e-05,
+      "loss": 0.1475,
+      "step": 2920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7129827285076653e-05,
+      "loss": 0.1569,
+      "step": 2930
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.7032796429264506e-05,
+      "loss": 0.1379,
+      "step": 2940
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.693576557345236e-05,
+      "loss": 0.1468,
+      "step": 2950
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.6838734717640212e-05,
+      "loss": 0.1327,
+      "step": 2960
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.674170386182806e-05,
+      "loss": 0.1824,
+      "step": 2970
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.6644673006015915e-05,
+      "loss": 0.1753,
+      "step": 2980
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.6547642150203768e-05,
+      "loss": 0.1506,
+      "step": 2990
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.645061129439162e-05,
+      "loss": 0.1308,
+      "step": 3000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.6353580438579467e-05,
+      "loss": 0.156,
+      "step": 3010
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.625654958276732e-05,
+      "loss": 0.1319,
+      "step": 3020
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.6159518726955173e-05,
+      "loss": 0.1453,
+      "step": 3030
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.6062487871143026e-05,
+      "loss": 0.1413,
+      "step": 3040
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.5965457015330873e-05,
+      "loss": 0.17,
+      "step": 3050
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.5868426159518726e-05,
+      "loss": 0.1385,
+      "step": 3060
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.577139530370658e-05,
+      "loss": 0.1408,
+      "step": 3070
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.5674364447894432e-05,
+      "loss": 0.1635,
+      "step": 3080
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.557733359208228e-05,
+      "loss": 0.1494,
+      "step": 3090
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.5480302736270134e-05,
+      "loss": 0.1407,
+      "step": 3100
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.5383271880457988e-05,
+      "loss": 0.119,
+      "step": 3110
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.528624102464584e-05,
+      "loss": 0.128,
+      "step": 3120
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.5189210168833687e-05,
+      "loss": 0.1274,
+      "step": 3130
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.509217931302154e-05,
+      "loss": 0.1338,
+      "step": 3140
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.4995148457209393e-05,
+      "loss": 0.1252,
+      "step": 3150
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.4898117601397246e-05,
+      "loss": 0.1344,
+      "step": 3160
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.48010867455851e-05,
+      "loss": 0.1404,
+      "step": 3170
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.470405588977295e-05,
+      "loss": 0.1537,
+      "step": 3180
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.4607025033960802e-05,
+      "loss": 0.1468,
+      "step": 3190
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.450999417814865e-05,
+      "loss": 0.1842,
+      "step": 3200
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.4412963322336505e-05,
+      "loss": 0.1441,
+      "step": 3210
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.4315932466524354e-05,
+      "loss": 0.1595,
+      "step": 3220
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.4218901610712207e-05,
+      "loss": 0.147,
+      "step": 3230
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.4121870754900057e-05,
+      "loss": 0.1612,
+      "step": 3240
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.402483989908791e-05,
+      "loss": 0.126,
+      "step": 3250
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.3927809043275763e-05,
+      "loss": 0.1325,
+      "step": 3260
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.3830778187463616e-05,
+      "loss": 0.1584,
+      "step": 3270
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.3733747331651466e-05,
+      "loss": 0.1427,
+      "step": 3280
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.363671647583932e-05,
+      "loss": 0.1404,
+      "step": 3290
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.353968562002717e-05,
+      "loss": 0.1567,
+      "step": 3300
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.344265476421502e-05,
+      "loss": 0.1912,
+      "step": 3310
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.334562390840287e-05,
+      "loss": 0.1254,
+      "step": 3320
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.3248593052590724e-05,
+      "loss": 0.1318,
+      "step": 3330
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.3151562196778574e-05,
+      "loss": 0.1511,
+      "step": 3340
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.3054531340966427e-05,
+      "loss": 0.1472,
+      "step": 3350
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.295750048515428e-05,
+      "loss": 0.1294,
+      "step": 3360
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.2860469629342133e-05,
+      "loss": 0.1453,
+      "step": 3370
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.2763438773529983e-05,
+      "loss": 0.1367,
+      "step": 3380
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.2666407917717836e-05,
+      "loss": 0.1589,
+      "step": 3390
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.2569377061905686e-05,
+      "loss": 0.1571,
+      "step": 3400
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.247234620609354e-05,
+      "loss": 0.1475,
+      "step": 3410
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.237531535028139e-05,
+      "loss": 0.1419,
+      "step": 3420
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.227828449446924e-05,
+      "loss": 0.1393,
+      "step": 3430
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.2181253638657095e-05,
+      "loss": 0.1433,
+      "step": 3440
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.2084222782844948e-05,
+      "loss": 0.1599,
+      "step": 3450
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.1987191927032797e-05,
+      "loss": 0.15,
+      "step": 3460
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.189016107122065e-05,
+      "loss": 0.1552,
+      "step": 3470
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.17931302154085e-05,
+      "loss": 0.1229,
+      "step": 3480
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.1696099359596353e-05,
+      "loss": 0.1247,
+      "step": 3490
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.1599068503784203e-05,
+      "loss": 0.1413,
+      "step": 3500
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.1502037647972056e-05,
+      "loss": 0.1348,
+      "step": 3510
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.1405006792159906e-05,
+      "loss": 0.1272,
+      "step": 3520
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.130797593634776e-05,
+      "loss": 0.1389,
+      "step": 3530
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.121094508053561e-05,
+      "loss": 0.1388,
+      "step": 3540
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.1113914224723465e-05,
+      "loss": 0.1311,
+      "step": 3550
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.1016883368911314e-05,
+      "loss": 0.1388,
+      "step": 3560
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.0919852513099167e-05,
+      "loss": 0.1431,
+      "step": 3570
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.0822821657287017e-05,
+      "loss": 0.1371,
+      "step": 3580
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.072579080147487e-05,
+      "loss": 0.1369,
+      "step": 3590
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.062875994566272e-05,
+      "loss": 0.126,
+      "step": 3600
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.0531729089850573e-05,
+      "loss": 0.1305,
+      "step": 3610
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.0434698234038423e-05,
+      "loss": 0.1432,
+      "step": 3620
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.0337667378226276e-05,
+      "loss": 0.1464,
+      "step": 3630
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.024063652241413e-05,
+      "loss": 0.1377,
+      "step": 3640
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.0143605666601982e-05,
+      "loss": 0.1483,
+      "step": 3650
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.004657481078983e-05,
+      "loss": 0.1306,
+      "step": 3660
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.9949543954977685e-05,
+      "loss": 0.1434,
+      "step": 3670
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.9852513099165534e-05,
+      "loss": 0.1372,
+      "step": 3680
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.9755482243353387e-05,
+      "loss": 0.1385,
+      "step": 3690
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.9658451387541237e-05,
+      "loss": 0.1429,
+      "step": 3700
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.956142053172909e-05,
+      "loss": 0.1322,
+      "step": 3710
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.9464389675916943e-05,
+      "loss": 0.1453,
+      "step": 3720
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.9367358820104796e-05,
+      "loss": 0.1473,
+      "step": 3730
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.9270327964292646e-05,
+      "loss": 0.1607,
+      "step": 3740
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.91732971084805e-05,
+      "loss": 0.1463,
+      "step": 3750
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.907626625266835e-05,
+      "loss": 0.1261,
+      "step": 3760
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.89792353968562e-05,
+      "loss": 0.1277,
+      "step": 3770
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.888220454104405e-05,
+      "loss": 0.144,
+      "step": 3780
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.8785173685231904e-05,
+      "loss": 0.1285,
+      "step": 3790
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.8688142829419754e-05,
+      "loss": 0.1601,
+      "step": 3800
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.8591111973607607e-05,
+      "loss": 0.133,
+      "step": 3810
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.849408111779546e-05,
+      "loss": 0.1366,
+      "step": 3820
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.8397050261983313e-05,
+      "loss": 0.1363,
+      "step": 3830
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.8300019406171163e-05,
+      "loss": 0.1265,
+      "step": 3840
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.8202988550359016e-05,
+      "loss": 0.1344,
+      "step": 3850
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.8105957694546866e-05,
+      "loss": 0.1293,
+      "step": 3860
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.800892683873472e-05,
+      "loss": 0.1421,
+      "step": 3870
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.791189598292257e-05,
+      "loss": 0.1319,
+      "step": 3880
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.781486512711042e-05,
+      "loss": 0.1159,
+      "step": 3890
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.771783427129827e-05,
+      "loss": 0.1407,
+      "step": 3900
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.7620803415486124e-05,
+      "loss": 0.1261,
+      "step": 3910
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.7523772559673977e-05,
+      "loss": 0.112,
+      "step": 3920
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.742674170386183e-05,
+      "loss": 0.1174,
+      "step": 3930
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.732971084804968e-05,
+      "loss": 0.1175,
+      "step": 3940
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.7232679992237533e-05,
+      "loss": 0.132,
+      "step": 3950
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.7135649136425383e-05,
+      "loss": 0.125,
+      "step": 3960
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.7038618280613236e-05,
+      "loss": 0.1378,
+      "step": 3970
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.6941587424801085e-05,
+      "loss": 0.1457,
+      "step": 3980
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.684455656898894e-05,
+      "loss": 0.1508,
+      "step": 3990
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.6747525713176788e-05,
+      "loss": 0.1513,
+      "step": 4000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.6650494857364645e-05,
+      "loss": 0.1337,
+      "step": 4010
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.6553464001552494e-05,
+      "loss": 0.1344,
+      "step": 4020
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.6456433145740347e-05,
+      "loss": 0.1284,
+      "step": 4030
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.6359402289928197e-05,
+      "loss": 0.1586,
+      "step": 4040
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.626237143411605e-05,
+      "loss": 0.1337,
+      "step": 4050
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.61653405783039e-05,
+      "loss": 0.1225,
+      "step": 4060
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.6068309722491753e-05,
+      "loss": 0.14,
+      "step": 4070
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.5971278866679603e-05,
+      "loss": 0.1517,
+      "step": 4080
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.5874248010867456e-05,
+      "loss": 0.1352,
+      "step": 4090
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.577721715505531e-05,
+      "loss": 0.1445,
+      "step": 4100
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.5680186299243162e-05,
+      "loss": 0.154,
+      "step": 4110
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.558315544343101e-05,
+      "loss": 0.1368,
+      "step": 4120
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.5486124587618864e-05,
+      "loss": 0.1335,
+      "step": 4130
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.5389093731806714e-05,
+      "loss": 0.1507,
+      "step": 4140
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.5292062875994567e-05,
+      "loss": 0.1253,
+      "step": 4150
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.5195032020182417e-05,
+      "loss": 0.1289,
+      "step": 4160
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.509800116437027e-05,
+      "loss": 0.1371,
+      "step": 4170
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.5000970308558121e-05,
+      "loss": 0.1458,
+      "step": 4180
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.4903939452745974e-05,
+      "loss": 0.1188,
+      "step": 4190
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.4806908596933824e-05,
+      "loss": 0.1435,
+      "step": 4200
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.4709877741121677e-05,
+      "loss": 0.147,
+      "step": 4210
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.4612846885309529e-05,
+      "loss": 0.1371,
+      "step": 4220
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.4515816029497382e-05,
+      "loss": 0.1351,
+      "step": 4230
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.4418785173685231e-05,
+      "loss": 0.1641,
+      "step": 4240
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.4321754317873084e-05,
+      "loss": 0.126,
+      "step": 4250
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.4224723462060936e-05,
+      "loss": 0.1297,
+      "step": 4260
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.4127692606248789e-05,
+      "loss": 0.1347,
+      "step": 4270
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.4030661750436638e-05,
+      "loss": 0.1157,
+      "step": 4280
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.3933630894624491e-05,
+      "loss": 0.139,
+      "step": 4290
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.3836600038812341e-05,
+      "loss": 0.1441,
+      "step": 4300
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.3739569183000194e-05,
+      "loss": 0.1313,
+      "step": 4310
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.3642538327188046e-05,
+      "loss": 0.1558,
+      "step": 4320
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.3545507471375899e-05,
+      "loss": 0.1474,
+      "step": 4330
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.3448476615563748e-05,
+      "loss": 0.1312,
+      "step": 4340
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.3351445759751601e-05,
+      "loss": 0.1168,
+      "step": 4350
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.3254414903939453e-05,
+      "loss": 0.1237,
+      "step": 4360
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.3157384048127306e-05,
+      "loss": 0.1377,
+      "step": 4370
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.3060353192315156e-05,
+      "loss": 0.1393,
+      "step": 4380
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.2963322336503009e-05,
+      "loss": 0.1377,
+      "step": 4390
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.286629148069086e-05,
+      "loss": 0.1233,
+      "step": 4400
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.2769260624878713e-05,
+      "loss": 0.123,
+      "step": 4410
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.2672229769066563e-05,
+      "loss": 0.1453,
+      "step": 4420
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.2575198913254416e-05,
+      "loss": 0.1328,
+      "step": 4430
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.2478168057442267e-05,
+      "loss": 0.1642,
+      "step": 4440
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.2381137201630118e-05,
+      "loss": 0.1237,
+      "step": 4450
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.2284106345817972e-05,
+      "loss": 0.1277,
+      "step": 4460
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.2187075490005823e-05,
+      "loss": 0.1282,
+      "step": 4470
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.2090044634193674e-05,
+      "loss": 0.1423,
+      "step": 4480
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.1993013778381526e-05,
+      "loss": 0.1222,
+      "step": 4490
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.1895982922569377e-05,
+      "loss": 0.131,
+      "step": 4500
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.179895206675723e-05,
+      "loss": 0.1208,
+      "step": 4510
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.1701921210945081e-05,
+      "loss": 0.1363,
+      "step": 4520
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.1604890355132933e-05,
+      "loss": 0.1361,
+      "step": 4530
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.1507859499320784e-05,
+      "loss": 0.1373,
+      "step": 4540
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.1410828643508637e-05,
+      "loss": 0.1404,
+      "step": 4550
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.1313797787696489e-05,
+      "loss": 0.1694,
+      "step": 4560
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.121676693188434e-05,
+      "loss": 0.1423,
+      "step": 4570
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.1119736076072191e-05,
+      "loss": 0.1543,
+      "step": 4580
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.1022705220260043e-05,
+      "loss": 0.1492,
+      "step": 4590
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.0925674364447896e-05,
+      "loss": 0.1359,
+      "step": 4600
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.0828643508635747e-05,
+      "loss": 0.1163,
+      "step": 4610
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.0731612652823599e-05,
+      "loss": 0.11,
+      "step": 4620
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.063458179701145e-05,
+      "loss": 0.1603,
+      "step": 4630
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.0537550941199301e-05,
+      "loss": 0.1326,
+      "step": 4640
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.0440520085387154e-05,
+      "loss": 0.121,
+      "step": 4650
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.0343489229575006e-05,
+      "loss": 0.1329,
+      "step": 4660
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.0246458373762857e-05,
+      "loss": 0.122,
+      "step": 4670
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.0149427517950708e-05,
+      "loss": 0.1226,
+      "step": 4680
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.0052396662138562e-05,
+      "loss": 0.1402,
+      "step": 4690
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.955365806326413e-06,
+      "loss": 0.1186,
+      "step": 4700
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.858334950514264e-06,
+      "loss": 0.1384,
+      "step": 4710
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.761304094702116e-06,
+      "loss": 0.1309,
+      "step": 4720
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.664273238889967e-06,
+      "loss": 0.1203,
+      "step": 4730
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 9.56724238307782e-06,
+      "loss": 0.1066,
+      "step": 4740
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 9.470211527265671e-06,
+      "loss": 0.1264,
+      "step": 4750
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 9.373180671453523e-06,
+      "loss": 0.1388,
+      "step": 4760
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 9.276149815641374e-06,
+      "loss": 0.1133,
+      "step": 4770
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 9.179118959829226e-06,
+      "loss": 0.1243,
+      "step": 4780
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 9.082088104017079e-06,
+      "loss": 0.1402,
+      "step": 4790
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.98505724820493e-06,
+      "loss": 0.1301,
+      "step": 4800
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.888026392392781e-06,
+      "loss": 0.1268,
+      "step": 4810
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.790995536580633e-06,
+      "loss": 0.1556,
+      "step": 4820
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.693964680768484e-06,
+      "loss": 0.1317,
+      "step": 4830
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.596933824956337e-06,
+      "loss": 0.1183,
+      "step": 4840
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.499902969144189e-06,
+      "loss": 0.1482,
+      "step": 4850
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.40287211333204e-06,
+      "loss": 0.1449,
+      "step": 4860
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.305841257519891e-06,
+      "loss": 0.1359,
+      "step": 4870
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.208810401707744e-06,
+      "loss": 0.1296,
+      "step": 4880
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.111779545895596e-06,
+      "loss": 0.1428,
+      "step": 4890
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.014748690083447e-06,
+      "loss": 0.1277,
+      "step": 4900
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 7.917717834271298e-06,
+      "loss": 0.115,
+      "step": 4910
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 7.82068697845915e-06,
+      "loss": 0.1279,
+      "step": 4920
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 7.723656122647003e-06,
+      "loss": 0.1163,
+      "step": 4930
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 7.626625266834854e-06,
+      "loss": 0.1416,
+      "step": 4940
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 7.529594411022706e-06,
+      "loss": 0.1425,
+      "step": 4950
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 7.432563555210558e-06,
+      "loss": 0.1374,
+      "step": 4960
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.335532699398409e-06,
+      "loss": 0.1234,
+      "step": 4970
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.2385018435862606e-06,
+      "loss": 0.1294,
+      "step": 4980
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.141470987774113e-06,
+      "loss": 0.0957,
+      "step": 4990
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.044440131961964e-06,
+      "loss": 0.1472,
+      "step": 5000
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 6.947409276149816e-06,
+      "loss": 0.1204,
+      "step": 5010
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 6.850378420337668e-06,
+      "loss": 0.1207,
+      "step": 5020
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 6.75334756452552e-06,
+      "loss": 0.1242,
+      "step": 5030
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 6.656316708713371e-06,
+      "loss": 0.1502,
+      "step": 5040
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 6.559285852901223e-06,
+      "loss": 0.1279,
+      "step": 5050
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 6.462254997089075e-06,
+      "loss": 0.1162,
+      "step": 5060
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 6.365224141276926e-06,
+      "loss": 0.1363,
+      "step": 5070
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 6.2681932854647785e-06,
+      "loss": 0.1194,
+      "step": 5080
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 6.17116242965263e-06,
+      "loss": 0.1361,
+      "step": 5090
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 6.074131573840482e-06,
+      "loss": 0.1195,
+      "step": 5100
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5.9771007180283334e-06,
+      "loss": 0.1314,
+      "step": 5110
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5.880069862216185e-06,
+      "loss": 0.1286,
+      "step": 5120
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5.783039006404037e-06,
+      "loss": 0.1134,
+      "step": 5130
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5.686008150591888e-06,
+      "loss": 0.1198,
+      "step": 5140
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5.588977294779741e-06,
+      "loss": 0.1128,
+      "step": 5150
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5.491946438967592e-06,
+      "loss": 0.1137,
+      "step": 5160
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5.394915583155443e-06,
+      "loss": 0.116,
+      "step": 5170
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5.2978847273432956e-06,
+      "loss": 0.1318,
+      "step": 5180
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5.200853871531147e-06,
+      "loss": 0.1215,
+      "step": 5190
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5.103823015718999e-06,
+      "loss": 0.1345,
+      "step": 5200
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5.0067921599068505e-06,
+      "loss": 0.1035,
+      "step": 5210
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 4.909761304094703e-06,
+      "loss": 0.117,
+      "step": 5220
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 4.812730448282554e-06,
+      "loss": 0.1229,
+      "step": 5230
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 4.7156995924704055e-06,
+      "loss": 0.1129,
+      "step": 5240
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 4.618668736658258e-06,
+      "loss": 0.1387,
+      "step": 5250
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 4.521637880846109e-06,
+      "loss": 0.1138,
+      "step": 5260
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 4.424607025033961e-06,
+      "loss": 0.1196,
+      "step": 5270
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 4.327576169221813e-06,
+      "loss": 0.1284,
+      "step": 5280
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 4.230545313409665e-06,
+      "loss": 0.131,
+      "step": 5290
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 4.133514457597516e-06,
+      "loss": 0.1207,
+      "step": 5300
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 4.036483601785368e-06,
+      "loss": 0.1424,
+      "step": 5310
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.93945274597322e-06,
+      "loss": 0.121,
+      "step": 5320
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.842421890161071e-06,
+      "loss": 0.1329,
+      "step": 5330
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.745391034348923e-06,
+      "loss": 0.1137,
+      "step": 5340
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.6483601785367748e-06,
+      "loss": 0.1178,
+      "step": 5350
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.5513293227246266e-06,
+      "loss": 0.1266,
+      "step": 5360
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.4542984669124784e-06,
+      "loss": 0.1185,
+      "step": 5370
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.35726761110033e-06,
+      "loss": 0.1433,
+      "step": 5380
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.260236755288182e-06,
+      "loss": 0.1204,
+      "step": 5390
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.1632058994760337e-06,
+      "loss": 0.1606,
+      "step": 5400
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.066175043663885e-06,
+      "loss": 0.1419,
+      "step": 5410
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 2.969144187851737e-06,
+      "loss": 0.099,
+      "step": 5420
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 2.8721133320395887e-06,
+      "loss": 0.1293,
+      "step": 5430
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 2.7750824762274405e-06,
+      "loss": 0.1112,
+      "step": 5440
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 2.6780516204152923e-06,
+      "loss": 0.1152,
+      "step": 5450
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.581020764603144e-06,
+      "loss": 0.1456,
+      "step": 5460
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.483989908790996e-06,
+      "loss": 0.1227,
+      "step": 5470
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.3869590529788472e-06,
+      "loss": 0.1134,
+      "step": 5480
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.289928197166699e-06,
+      "loss": 0.1279,
+      "step": 5490
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.192897341354551e-06,
+      "loss": 0.1351,
+      "step": 5500
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.0958664855424026e-06,
+      "loss": 0.1251,
+      "step": 5510
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.9988356297302544e-06,
+      "loss": 0.1343,
+      "step": 5520
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.901804773918106e-06,
+      "loss": 0.1276,
+      "step": 5530
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.8047739181059578e-06,
+      "loss": 0.144,
+      "step": 5540
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.7077430622938096e-06,
+      "loss": 0.1186,
+      "step": 5550
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.6107122064816612e-06,
+      "loss": 0.1531,
+      "step": 5560
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.513681350669513e-06,
+      "loss": 0.1327,
+      "step": 5570
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.4166504948573648e-06,
+      "loss": 0.1279,
+      "step": 5580
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.3196196390452165e-06,
+      "loss": 0.1133,
+      "step": 5590
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.2225887832330681e-06,
+      "loss": 0.1375,
+      "step": 5600
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.12555792742092e-06,
+      "loss": 0.1043,
+      "step": 5610
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.0285270716087717e-06,
+      "loss": 0.1294,
+      "step": 5620
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 9.314962157966234e-07,
+      "loss": 0.1025,
+      "step": 5630
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 8.344653599844751e-07,
+      "loss": 0.1274,
+      "step": 5640
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.374345041723269e-07,
+      "loss": 0.1357,
+      "step": 5650
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 6.404036483601786e-07,
+      "loss": 0.1316,
+      "step": 5660
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.433727925480303e-07,
+      "loss": 0.1215,
+      "step": 5670
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 4.4634193673588205e-07,
+      "loss": 0.1428,
+      "step": 5680
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.493110809237338e-07,
+      "loss": 0.1169,
+      "step": 5690
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.522802251115855e-07,
+      "loss": 0.1182,
+      "step": 5700
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.5524936929943724e-07,
+      "loss": 0.1338,
+      "step": 5710
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.821851348728896e-08,
+      "loss": 0.1181,
+      "step": 5720
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9043314126763362,
+      "eval_loss": 0.23999236524105072,
+      "eval_runtime": 105.6463,
+      "eval_samples_per_second": 95.28,
+      "eval_steps_per_second": 2.982,
+      "step": 5726
+    },
+    {
+      "epoch": 2.0,
+      "step": 5726,
+      "total_flos": 5.679462069216204e+19,
+      "train_loss": 0.1848815517786954,
+      "train_runtime": 16137.6204,
+      "train_samples_per_second": 45.426,
+      "train_steps_per_second": 0.355
+    }
+  ],
+  "max_steps": 5726,
+  "num_train_epochs": 2,
+  "total_flos": 5.679462069216204e+19,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/training_args.bin b/training_args.bin
new file mode 100644
index 0000000..456d41d
--- /dev/null
+++ b/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bedd40b1d33eccdfba2ab70bf80d2dcf1d6b2f55d34bfa6f0336e0811dd1028b
+size 3643