End of training

README.md

@@ -4,6 +4,7 @@ license: apache-2.0
 base_model: Qwen/Qwen2.5-7B-Instruct
 tags:
 - llama-factory
+- full
 - generated_from_trainer
 model-index:
 - name: b2_code_fasttext_pos_codeforces_neg_codereview_10k
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # b2_code_fasttext_pos_codeforces_neg_codereview_10k
 
-This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on an unknown dataset.
+This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on the mlfoundations-dev/b2_code_fasttext_pos_codeforces_neg_codereview_10k dataset.
 
 ## Model description
 

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.993610223642173,
+    "total_flos": 1.2901911719428751e+18,
+    "train_loss": 0.14476858102358303,
+    "train_runtime": 3885.8859,
+    "train_samples_per_second": 12.867,
+    "train_steps_per_second": 0.1
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.993610223642173,
+    "total_flos": 1.2901911719428751e+18,
+    "train_loss": 0.14476858102358303,
+    "train_runtime": 3885.8859,
+    "train_samples_per_second": 12.867,
+    "train_steps_per_second": 0.1
+}

trainer_state.json

@@ -0,0 +1,2772 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.993610223642173,
+  "eval_steps": 500,
+  "global_step": 390,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.012779552715654952,
+      "grad_norm": 6.606391299183944,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 1.0439,
+      "step": 1
+    },
+    {
+      "epoch": 0.025559105431309903,
+      "grad_norm": 6.580557712601049,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 1.0507,
+      "step": 2
+    },
+    {
+      "epoch": 0.038338658146964855,
+      "grad_norm": 6.509005644872208,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 1.0409,
+      "step": 3
+    },
+    {
+      "epoch": 0.051118210862619806,
+      "grad_norm": 5.947065826738325,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 1.0366,
+      "step": 4
+    },
+    {
+      "epoch": 0.06389776357827476,
+      "grad_norm": 4.4097960543741825,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 0.9889,
+      "step": 5
+    },
+    {
+      "epoch": 0.07667731629392971,
+      "grad_norm": 2.818415172846207,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.9628,
+      "step": 6
+    },
+    {
+      "epoch": 0.08945686900958466,
+      "grad_norm": 2.6496653931802,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 0.9587,
+      "step": 7
+    },
+    {
+      "epoch": 0.10223642172523961,
+      "grad_norm": 4.336870837964605,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 0.9602,
+      "step": 8
+    },
+    {
+      "epoch": 0.11501597444089456,
+      "grad_norm": 4.298942553498102,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.9378,
+      "step": 9
+    },
+    {
+      "epoch": 0.12779552715654952,
+      "grad_norm": 4.186284730314603,
+      "learning_rate": 1.0256410256410256e-05,
+      "loss": 0.912,
+      "step": 10
+    },
+    {
+      "epoch": 0.14057507987220447,
+      "grad_norm": 3.8257868758679257,
+      "learning_rate": 1.1282051282051283e-05,
+      "loss": 0.8958,
+      "step": 11
+    },
+    {
+      "epoch": 0.15335463258785942,
+      "grad_norm": 2.5816728496761097,
+      "learning_rate": 1.230769230769231e-05,
+      "loss": 0.8622,
+      "step": 12
+    },
+    {
+      "epoch": 0.16613418530351437,
+      "grad_norm": 1.8503542682907892,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.8189,
+      "step": 13
+    },
+    {
+      "epoch": 0.17891373801916932,
+      "grad_norm": 1.9534484580657967,
+      "learning_rate": 1.435897435897436e-05,
+      "loss": 0.8372,
+      "step": 14
+    },
+    {
+      "epoch": 0.19169329073482427,
+      "grad_norm": 1.516048850182833,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 0.8026,
+      "step": 15
+    },
+    {
+      "epoch": 0.20447284345047922,
+      "grad_norm": 1.0874051240964258,
+      "learning_rate": 1.641025641025641e-05,
+      "loss": 0.7636,
+      "step": 16
+    },
+    {
+      "epoch": 0.21725239616613418,
+      "grad_norm": 1.1631866901137267,
+      "learning_rate": 1.7435897435897438e-05,
+      "loss": 0.7816,
+      "step": 17
+    },
+    {
+      "epoch": 0.23003194888178913,
+      "grad_norm": 1.2171907306151695,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 0.7546,
+      "step": 18
+    },
+    {
+      "epoch": 0.24281150159744408,
+      "grad_norm": 0.9490832264240321,
+      "learning_rate": 1.9487179487179488e-05,
+      "loss": 0.7234,
+      "step": 19
+    },
+    {
+      "epoch": 0.25559105431309903,
+      "grad_norm": 0.9462270136925807,
+      "learning_rate": 2.0512820512820512e-05,
+      "loss": 0.7333,
+      "step": 20
+    },
+    {
+      "epoch": 0.268370607028754,
+      "grad_norm": 1.1591760710159509,
+      "learning_rate": 2.153846153846154e-05,
+      "loss": 0.748,
+      "step": 21
+    },
+    {
+      "epoch": 0.28115015974440893,
+      "grad_norm": 0.7918148886565669,
+      "learning_rate": 2.2564102564102566e-05,
+      "loss": 0.7159,
+      "step": 22
+    },
+    {
+      "epoch": 0.2939297124600639,
+      "grad_norm": 0.9632675527150978,
+      "learning_rate": 2.3589743589743593e-05,
+      "loss": 0.6958,
+      "step": 23
+    },
+    {
+      "epoch": 0.30670926517571884,
+      "grad_norm": 0.9408213867843294,
+      "learning_rate": 2.461538461538462e-05,
+      "loss": 0.7061,
+      "step": 24
+    },
+    {
+      "epoch": 0.3194888178913738,
+      "grad_norm": 0.7998105188303717,
+      "learning_rate": 2.5641025641025646e-05,
+      "loss": 0.7056,
+      "step": 25
+    },
+    {
+      "epoch": 0.33226837060702874,
+      "grad_norm": 0.9885203953583884,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.7043,
+      "step": 26
+    },
+    {
+      "epoch": 0.3450479233226837,
+      "grad_norm": 0.8586191987847381,
+      "learning_rate": 2.7692307692307694e-05,
+      "loss": 0.689,
+      "step": 27
+    },
+    {
+      "epoch": 0.35782747603833864,
+      "grad_norm": 0.8572793221628082,
+      "learning_rate": 2.871794871794872e-05,
+      "loss": 0.6885,
+      "step": 28
+    },
+    {
+      "epoch": 0.3706070287539936,
+      "grad_norm": 0.703637939749516,
+      "learning_rate": 2.9743589743589747e-05,
+      "loss": 0.6829,
+      "step": 29
+    },
+    {
+      "epoch": 0.38338658146964855,
+      "grad_norm": 0.9933804760621775,
+      "learning_rate": 3.0769230769230774e-05,
+      "loss": 0.6938,
+      "step": 30
+    },
+    {
+      "epoch": 0.3961661341853035,
+      "grad_norm": 0.8925386250116185,
+      "learning_rate": 3.1794871794871795e-05,
+      "loss": 0.6863,
+      "step": 31
+    },
+    {
+      "epoch": 0.40894568690095845,
+      "grad_norm": 1.0350535569861223,
+      "learning_rate": 3.282051282051282e-05,
+      "loss": 0.6893,
+      "step": 32
+    },
+    {
+      "epoch": 0.4217252396166134,
+      "grad_norm": 1.1445914450631147,
+      "learning_rate": 3.384615384615385e-05,
+      "loss": 0.6708,
+      "step": 33
+    },
+    {
+      "epoch": 0.43450479233226835,
+      "grad_norm": 0.8753384928477117,
+      "learning_rate": 3.4871794871794875e-05,
+      "loss": 0.6939,
+      "step": 34
+    },
+    {
+      "epoch": 0.4472843450479233,
+      "grad_norm": 0.8393311315022303,
+      "learning_rate": 3.58974358974359e-05,
+      "loss": 0.6661,
+      "step": 35
+    },
+    {
+      "epoch": 0.46006389776357826,
+      "grad_norm": 1.1565786569988288,
+      "learning_rate": 3.692307692307693e-05,
+      "loss": 0.6669,
+      "step": 36
+    },
+    {
+      "epoch": 0.4728434504792332,
+      "grad_norm": 0.6967547302370791,
+      "learning_rate": 3.794871794871795e-05,
+      "loss": 0.6626,
+      "step": 37
+    },
+    {
+      "epoch": 0.48562300319488816,
+      "grad_norm": 0.8751186173570553,
+      "learning_rate": 3.8974358974358976e-05,
+      "loss": 0.6393,
+      "step": 38
+    },
+    {
+      "epoch": 0.4984025559105431,
+      "grad_norm": 1.2770961083050876,
+      "learning_rate": 4e-05,
+      "loss": 0.6676,
+      "step": 39
+    },
+    {
+      "epoch": 0.5111821086261981,
+      "grad_norm": 1.2772025124958417,
+      "learning_rate": 3.9999198907597046e-05,
+      "loss": 0.6667,
+      "step": 40
+    },
+    {
+      "epoch": 0.5239616613418531,
+      "grad_norm": 0.8066890893797722,
+      "learning_rate": 3.9996795694563096e-05,
+      "loss": 0.6499,
+      "step": 41
+    },
+    {
+      "epoch": 0.536741214057508,
+      "grad_norm": 1.579755868796052,
+      "learning_rate": 3.999279055341771e-05,
+      "loss": 0.6569,
+      "step": 42
+    },
+    {
+      "epoch": 0.549520766773163,
+      "grad_norm": 0.706322047158653,
+      "learning_rate": 3.998718380500971e-05,
+      "loss": 0.6541,
+      "step": 43
+    },
+    {
+      "epoch": 0.5623003194888179,
+      "grad_norm": 1.586782594427505,
+      "learning_rate": 3.997997589849145e-05,
+      "loss": 0.652,
+      "step": 44
+    },
+    {
+      "epoch": 0.5750798722044729,
+      "grad_norm": 1.0807329202383447,
+      "learning_rate": 3.9971167411282835e-05,
+      "loss": 0.6556,
+      "step": 45
+    },
+    {
+      "epoch": 0.5878594249201278,
+      "grad_norm": 1.6536678088892875,
+      "learning_rate": 3.99607590490251e-05,
+      "loss": 0.6444,
+      "step": 46
+    },
+    {
+      "epoch": 0.6006389776357828,
+      "grad_norm": 1.1989223290936704,
+      "learning_rate": 3.9948751645524235e-05,
+      "loss": 0.6581,
+      "step": 47
+    },
+    {
+      "epoch": 0.6134185303514377,
+      "grad_norm": 1.5873983130186309,
+      "learning_rate": 3.9935146162684206e-05,
+      "loss": 0.634,
+      "step": 48
+    },
+    {
+      "epoch": 0.6261980830670927,
+      "grad_norm": 1.2371929427150843,
+      "learning_rate": 3.9919943690429906e-05,
+      "loss": 0.6423,
+      "step": 49
+    },
+    {
+      "epoch": 0.6389776357827476,
+      "grad_norm": 1.373873539863577,
+      "learning_rate": 3.9903145446619837e-05,
+      "loss": 0.6462,
+      "step": 50
+    },
+    {
+      "epoch": 0.6517571884984026,
+      "grad_norm": 1.1016235774139902,
+      "learning_rate": 3.9884752776948564e-05,
+      "loss": 0.6401,
+      "step": 51
+    },
+    {
+      "epoch": 0.6645367412140575,
+      "grad_norm": 1.1209870974953269,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.6367,
+      "step": 52
+    },
+    {
+      "epoch": 0.6773162939297125,
+      "grad_norm": 1.0860208778406302,
+      "learning_rate": 3.9843190181323744e-05,
+      "loss": 0.6417,
+      "step": 53
+    },
+    {
+      "epoch": 0.6900958466453674,
+      "grad_norm": 1.1698853570343006,
+      "learning_rate": 3.982002358491817e-05,
+      "loss": 0.6434,
+      "step": 54
+    },
+    {
+      "epoch": 0.7028753993610224,
+      "grad_norm": 0.9936615094214557,
+      "learning_rate": 3.979526922148058e-05,
+      "loss": 0.6428,
+      "step": 55
+    },
+    {
+      "epoch": 0.7156549520766773,
+      "grad_norm": 1.0541150791659504,
+      "learning_rate": 3.9768929074064206e-05,
+      "loss": 0.6509,
+      "step": 56
+    },
+    {
+      "epoch": 0.7284345047923323,
+      "grad_norm": 0.9556652153918601,
+      "learning_rate": 3.9741005252758255e-05,
+      "loss": 0.6287,
+      "step": 57
+    },
+    {
+      "epoch": 0.7412140575079872,
+      "grad_norm": 0.954056106286679,
+      "learning_rate": 3.971149999451886e-05,
+      "loss": 0.6293,
+      "step": 58
+    },
+    {
+      "epoch": 0.7539936102236422,
+      "grad_norm": 1.2607602622746636,
+      "learning_rate": 3.9680415662989806e-05,
+      "loss": 0.6451,
+      "step": 59
+    },
+    {
+      "epoch": 0.7667731629392971,
+      "grad_norm": 0.638676278829799,
+      "learning_rate": 3.9647754748313294e-05,
+      "loss": 0.6356,
+      "step": 60
+    },
+    {
+      "epoch": 0.7795527156549521,
+      "grad_norm": 1.0681234854801602,
+      "learning_rate": 3.96135198669304e-05,
+      "loss": 0.6368,
+      "step": 61
+    },
+    {
+      "epoch": 0.792332268370607,
+      "grad_norm": 1.033023863786734,
+      "learning_rate": 3.957771376137144e-05,
+      "loss": 0.6062,
+      "step": 62
+    },
+    {
+      "epoch": 0.805111821086262,
+      "grad_norm": 0.9542423560838191,
+      "learning_rate": 3.954033930003634e-05,
+      "loss": 0.6273,
+      "step": 63
+    },
+    {
+      "epoch": 0.8178913738019169,
+      "grad_norm": 0.9254432594288557,
+      "learning_rate": 3.9501399476964806e-05,
+      "loss": 0.6228,
+      "step": 64
+    },
+    {
+      "epoch": 0.8306709265175719,
+      "grad_norm": 0.7617576652045225,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.6106,
+      "step": 65
+    },
+    {
+      "epoch": 0.8434504792332268,
+      "grad_norm": 0.6249220755858457,
+      "learning_rate": 3.9418836348521045e-05,
+      "loss": 0.6333,
+      "step": 66
+    },
+    {
+      "epoch": 0.8562300319488818,
+      "grad_norm": 0.977881184781753,
+      "learning_rate": 3.937521965721831e-05,
+      "loss": 0.6158,
+      "step": 67
+    },
+    {
+      "epoch": 0.8690095846645367,
+      "grad_norm": 1.0978122758989277,
+      "learning_rate": 3.933005083178828e-05,
+      "loss": 0.627,
+      "step": 68
+    },
+    {
+      "epoch": 0.8817891373801917,
+      "grad_norm": 0.657146186705749,
+      "learning_rate": 3.928333349067125e-05,
+      "loss": 0.6355,
+      "step": 69
+    },
+    {
+      "epoch": 0.8945686900958466,
+      "grad_norm": 0.6445634396508156,
+      "learning_rate": 3.923507137635792e-05,
+      "loss": 0.6269,
+      "step": 70
+    },
+    {
+      "epoch": 0.9073482428115016,
+      "grad_norm": 0.8852782221046644,
+      "learning_rate": 3.9185268355089606e-05,
+      "loss": 0.6358,
+      "step": 71
+    },
+    {
+      "epoch": 0.9201277955271565,
+      "grad_norm": 0.8747591576573158,
+      "learning_rate": 3.913392841654851e-05,
+      "loss": 0.6365,
+      "step": 72
+    },
+    {
+      "epoch": 0.9329073482428115,
+      "grad_norm": 0.7450313154115504,
+      "learning_rate": 3.9081055673538093e-05,
+      "loss": 0.6213,
+      "step": 73
+    },
+    {
+      "epoch": 0.9456869009584664,
+      "grad_norm": 0.7884996548731228,
+      "learning_rate": 3.902665436165364e-05,
+      "loss": 0.6176,
+      "step": 74
+    },
+    {
+      "epoch": 0.9584664536741214,
+      "grad_norm": 0.7688130173646047,
+      "learning_rate": 3.897072883894291e-05,
+      "loss": 0.6165,
+      "step": 75
+    },
+    {
+      "epoch": 0.9712460063897763,
+      "grad_norm": 0.6191938351479404,
+      "learning_rate": 3.8913283585557054e-05,
+      "loss": 0.6137,
+      "step": 76
+    },
+    {
+      "epoch": 0.9840255591054313,
+      "grad_norm": 0.7872267173045574,
+      "learning_rate": 3.885432320339167e-05,
+      "loss": 0.6186,
+      "step": 77
+    },
+    {
+      "epoch": 0.9968051118210862,
+      "grad_norm": 0.895477964352014,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.6167,
+      "step": 78
+    },
+    {
+      "epoch": 1.0095846645367412,
+      "grad_norm": 0.8308432476526806,
+      "learning_rate": 3.873187606680543e-05,
+      "loss": 0.5632,
+      "step": 79
+    },
+    {
+      "epoch": 1.0223642172523961,
+      "grad_norm": 0.5713434732452398,
+      "learning_rate": 3.866839912153168e-05,
+      "loss": 0.5521,
+      "step": 80
+    },
+    {
+      "epoch": 1.035143769968051,
+      "grad_norm": 0.7142369781512922,
+      "learning_rate": 3.860342666498677e-05,
+      "loss": 0.5572,
+      "step": 81
+    },
+    {
+      "epoch": 1.0479233226837061,
+      "grad_norm": 0.799601323958914,
+      "learning_rate": 3.853696390206484e-05,
+      "loss": 0.5544,
+      "step": 82
+    },
+    {
+      "epoch": 1.060702875399361,
+      "grad_norm": 0.7455938478793749,
+      "learning_rate": 3.846901615704734e-05,
+      "loss": 0.5552,
+      "step": 83
+    },
+    {
+      "epoch": 1.073482428115016,
+      "grad_norm": 0.6034988797649778,
+      "learning_rate": 3.839958887317649e-05,
+      "loss": 0.5532,
+      "step": 84
+    },
+    {
+      "epoch": 1.0862619808306708,
+      "grad_norm": 0.6770389869828726,
+      "learning_rate": 3.832868761221926e-05,
+      "loss": 0.5451,
+      "step": 85
+    },
+    {
+      "epoch": 1.099041533546326,
+      "grad_norm": 0.954540528268312,
+      "learning_rate": 3.825631805402182e-05,
+      "loss": 0.5503,
+      "step": 86
+    },
+    {
+      "epoch": 1.1118210862619808,
+      "grad_norm": 0.8213749729805485,
+      "learning_rate": 3.818248599605448e-05,
+      "loss": 0.5601,
+      "step": 87
+    },
+    {
+      "epoch": 1.1246006389776357,
+      "grad_norm": 0.741351276087373,
+      "learning_rate": 3.810719735294731e-05,
+      "loss": 0.5654,
+      "step": 88
+    },
+    {
+      "epoch": 1.1373801916932909,
+      "grad_norm": 0.8481794308869871,
+      "learning_rate": 3.8030458156016326e-05,
+      "loss": 0.5565,
+      "step": 89
+    },
+    {
+      "epoch": 1.1501597444089458,
+      "grad_norm": 0.7594004526515341,
+      "learning_rate": 3.795227455278029e-05,
+      "loss": 0.5545,
+      "step": 90
+    },
+    {
+      "epoch": 1.1629392971246006,
+      "grad_norm": 0.632168169185403,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.5457,
+      "step": 91
+    },
+    {
+      "epoch": 1.1757188498402555,
+      "grad_norm": 0.7031486638124088,
+      "learning_rate": 3.7791599295517825e-05,
+      "loss": 0.5468,
+      "step": 92
+    },
+    {
+      "epoch": 1.1884984025559104,
+      "grad_norm": 0.6808588700454326,
+      "learning_rate": 3.7709120513064196e-05,
+      "loss": 0.5536,
+      "step": 93
+    },
+    {
+      "epoch": 1.2012779552715656,
+      "grad_norm": 0.6921396282319281,
+      "learning_rate": 3.762522306641998e-05,
+      "loss": 0.5506,
+      "step": 94
+    },
+    {
+      "epoch": 1.2140575079872205,
+      "grad_norm": 0.5121947510357499,
+      "learning_rate": 3.7539913676545874e-05,
+      "loss": 0.5522,
+      "step": 95
+    },
+    {
+      "epoch": 1.2268370607028753,
+      "grad_norm": 0.6392685476020531,
+      "learning_rate": 3.745319917751229e-05,
+      "loss": 0.5558,
+      "step": 96
+    },
+    {
+      "epoch": 1.2396166134185305,
+      "grad_norm": 0.5837442702174773,
+      "learning_rate": 3.736508651595188e-05,
+      "loss": 0.5578,
+      "step": 97
+    },
+    {
+      "epoch": 1.2523961661341854,
+      "grad_norm": 0.4530666164198027,
+      "learning_rate": 3.727558275050301e-05,
+      "loss": 0.5458,
+      "step": 98
+    },
+    {
+      "epoch": 1.2651757188498403,
+      "grad_norm": 0.602372240522267,
+      "learning_rate": 3.718469505124434e-05,
+      "loss": 0.5501,
+      "step": 99
+    },
+    {
+      "epoch": 1.2779552715654952,
+      "grad_norm": 0.5644538160733871,
+      "learning_rate": 3.709243069912041e-05,
+      "loss": 0.5443,
+      "step": 100
+    },
+    {
+      "epoch": 1.29073482428115,
+      "grad_norm": 0.6022811499264725,
+      "learning_rate": 3.699879708535838e-05,
+      "loss": 0.5487,
+      "step": 101
+    },
+    {
+      "epoch": 1.3035143769968052,
+      "grad_norm": 0.550906743250385,
+      "learning_rate": 3.69038017108759e-05,
+      "loss": 0.5509,
+      "step": 102
+    },
+    {
+      "epoch": 1.31629392971246,
+      "grad_norm": 0.6233956386597138,
+      "learning_rate": 3.680745218568026e-05,
+      "loss": 0.5534,
+      "step": 103
+    },
+    {
+      "epoch": 1.329073482428115,
+      "grad_norm": 0.5182024410893715,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.556,
+      "step": 104
+    },
+    {
+      "epoch": 1.34185303514377,
+      "grad_norm": 0.564945200040499,
+      "learning_rate": 3.6610721664960236e-05,
+      "loss": 0.5512,
+      "step": 105
+    },
+    {
+      "epoch": 1.354632587859425,
+      "grad_norm": 0.612513184636645,
+      "learning_rate": 3.65103564293684e-05,
+      "loss": 0.5408,
+      "step": 106
+    },
+    {
+      "epoch": 1.3674121405750799,
+      "grad_norm": 0.5942673792674769,
+      "learning_rate": 3.640866856166601e-05,
+      "loss": 0.5451,
+      "step": 107
+    },
+    {
+      "epoch": 1.3801916932907348,
+      "grad_norm": 0.588459067623318,
+      "learning_rate": 3.6305666207990886e-05,
+      "loss": 0.5485,
+      "step": 108
+    },
+    {
+      "epoch": 1.3929712460063897,
+      "grad_norm": 0.4140189623419551,
+      "learning_rate": 3.6201357619783336e-05,
+      "loss": 0.5438,
+      "step": 109
+    },
+    {
+      "epoch": 1.4057507987220448,
+      "grad_norm": 0.5300715964850526,
+      "learning_rate": 3.609575115312511e-05,
+      "loss": 0.5539,
+      "step": 110
+    },
+    {
+      "epoch": 1.4185303514376997,
+      "grad_norm": 0.5406759521793079,
+      "learning_rate": 3.598885526807003e-05,
+      "loss": 0.5466,
+      "step": 111
+    },
+    {
+      "epoch": 1.4313099041533546,
+      "grad_norm": 0.47685129470426335,
+      "learning_rate": 3.5880678527966224e-05,
+      "loss": 0.5438,
+      "step": 112
+    },
+    {
+      "epoch": 1.4440894568690097,
+      "grad_norm": 0.5398592076049669,
+      "learning_rate": 3.577122959877017e-05,
+      "loss": 0.5419,
+      "step": 113
+    },
+    {
+      "epoch": 1.4568690095846646,
+      "grad_norm": 0.5490690544479012,
+      "learning_rate": 3.566051724835245e-05,
+      "loss": 0.5507,
+      "step": 114
+    },
+    {
+      "epoch": 1.4696485623003195,
+      "grad_norm": 0.47169320271914167,
+      "learning_rate": 3.554855034579532e-05,
+      "loss": 0.5346,
+      "step": 115
+    },
+    {
+      "epoch": 1.4824281150159744,
+      "grad_norm": 0.5414566322709992,
+      "learning_rate": 3.5435337860682304e-05,
+      "loss": 0.5423,
+      "step": 116
+    },
+    {
+      "epoch": 1.4952076677316293,
+      "grad_norm": 0.4186816440677675,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.5655,
+      "step": 117
+    },
+    {
+      "epoch": 1.5079872204472844,
+      "grad_norm": 0.5111754842170994,
+      "learning_rate": 3.520521251930941e-05,
+      "loss": 0.5567,
+      "step": 118
+    },
+    {
+      "epoch": 1.5207667731629393,
+      "grad_norm": 0.4889729745568838,
+      "learning_rate": 3.5088318098215805e-05,
+      "loss": 0.5494,
+      "step": 119
+    },
+    {
+      "epoch": 1.5335463258785942,
+      "grad_norm": 0.5970983261404058,
+      "learning_rate": 3.497021496342203e-05,
+      "loss": 0.5485,
+      "step": 120
+    },
+    {
+      "epoch": 1.5463258785942493,
+      "grad_norm": 0.6244174589560836,
+      "learning_rate": 3.485091257608047e-05,
+      "loss": 0.5521,
+      "step": 121
+    },
+    {
+      "epoch": 1.5591054313099042,
+      "grad_norm": 0.5150332399753115,
+      "learning_rate": 3.473042049341474e-05,
+      "loss": 0.5263,
+      "step": 122
+    },
+    {
+      "epoch": 1.571884984025559,
+      "grad_norm": 0.5062479419125213,
+      "learning_rate": 3.4608748367954064e-05,
+      "loss": 0.5662,
+      "step": 123
+    },
+    {
+      "epoch": 1.5846645367412142,
+      "grad_norm": 0.5795778408417265,
+      "learning_rate": 3.4485905946759965e-05,
+      "loss": 0.5388,
+      "step": 124
+    },
+    {
+      "epoch": 1.5974440894568689,
+      "grad_norm": 0.6101565234254713,
+      "learning_rate": 3.4361903070645484e-05,
+      "loss": 0.5468,
+      "step": 125
+    },
+    {
+      "epoch": 1.610223642172524,
+      "grad_norm": 0.5637863425267695,
+      "learning_rate": 3.423674967338681e-05,
+      "loss": 0.5675,
+      "step": 126
+    },
+    {
+      "epoch": 1.623003194888179,
+      "grad_norm": 0.5353085598782966,
+      "learning_rate": 3.411045578092754e-05,
+      "loss": 0.5318,
+      "step": 127
+    },
+    {
+      "epoch": 1.6357827476038338,
+      "grad_norm": 0.6536912429537135,
+      "learning_rate": 3.398303151057543e-05,
+      "loss": 0.5605,
+      "step": 128
+    },
+    {
+      "epoch": 1.648562300319489,
+      "grad_norm": 0.4899086385884318,
+      "learning_rate": 3.385448707019199e-05,
+      "loss": 0.5538,
+      "step": 129
+    },
+    {
+      "epoch": 1.6613418530351438,
+      "grad_norm": 0.5325946949888003,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.5588,
+      "step": 130
+    },
+    {
+      "epoch": 1.6741214057507987,
+      "grad_norm": 0.5403278546775407,
+      "learning_rate": 3.359407895863199e-05,
+      "loss": 0.5379,
+      "step": 131
+    },
+    {
+      "epoch": 1.6869009584664538,
+      "grad_norm": 0.5139765621340018,
+      "learning_rate": 3.34622361485514e-05,
+      "loss": 0.5425,
+      "step": 132
+    },
+    {
+      "epoch": 1.6996805111821085,
+      "grad_norm": 0.5582667520807697,
+      "learning_rate": 3.332931488896029e-05,
+      "loss": 0.5561,
+      "step": 133
+    },
+    {
+      "epoch": 1.7124600638977636,
+      "grad_norm": 0.5725496792559248,
+      "learning_rate": 3.319532582807977e-05,
+      "loss": 0.538,
+      "step": 134
+    },
+    {
+      "epoch": 1.7252396166134185,
+      "grad_norm": 0.5796487796988573,
+      "learning_rate": 3.30602796996717e-05,
+      "loss": 0.5461,
+      "step": 135
+    },
+    {
+      "epoch": 1.7380191693290734,
+      "grad_norm": 0.5264811040173926,
+      "learning_rate": 3.2924187322178865e-05,
+      "loss": 0.5297,
+      "step": 136
+    },
+    {
+      "epoch": 1.7507987220447285,
+      "grad_norm": 0.5094443596891554,
+      "learning_rate": 3.278705959785821e-05,
+      "loss": 0.5425,
+      "step": 137
+    },
+    {
+      "epoch": 1.7635782747603834,
+      "grad_norm": 0.560498854774084,
+      "learning_rate": 3.2648907511907544e-05,
+      "loss": 0.5288,
+      "step": 138
+    },
+    {
+      "epoch": 1.7763578274760383,
+      "grad_norm": 0.5650237995626031,
+      "learning_rate": 3.250974213158555e-05,
+      "loss": 0.5402,
+      "step": 139
+    },
+    {
+      "epoch": 1.7891373801916934,
+      "grad_norm": 0.5736887246028277,
+      "learning_rate": 3.23695746053251e-05,
+      "loss": 0.5462,
+      "step": 140
+    },
+    {
+      "epoch": 1.8019169329073481,
+      "grad_norm": 0.6448216840943815,
+      "learning_rate": 3.222841616184025e-05,
+      "loss": 0.5457,
+      "step": 141
+    },
+    {
+      "epoch": 1.8146964856230032,
+      "grad_norm": 0.5409547316064893,
+      "learning_rate": 3.208627810922665e-05,
+      "loss": 0.5467,
+      "step": 142
+    },
+    {
+      "epoch": 1.8274760383386581,
+      "grad_norm": 0.5281660531292864,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.5584,
+      "step": 143
+    },
+    {
+      "epoch": 1.840255591054313,
+      "grad_norm": 0.5642907834669127,
+      "learning_rate": 3.1799108800462466e-05,
+      "loss": 0.5397,
+      "step": 144
+    },
+    {
+      "epoch": 1.8530351437699681,
+      "grad_norm": 0.6958435076312217,
+      "learning_rate": 3.1654100549227024e-05,
+      "loss": 0.5619,
+      "step": 145
+    },
+    {
+      "epoch": 1.865814696485623,
+      "grad_norm": 0.44915446473186066,
+      "learning_rate": 3.1508158696850275e-05,
+      "loss": 0.5371,
+      "step": 146
+    },
+    {
+      "epoch": 1.878594249201278,
+      "grad_norm": 0.7415779882754461,
+      "learning_rate": 3.136129493462312e-05,
+      "loss": 0.5459,
+      "step": 147
+    },
+    {
+      "epoch": 1.891373801916933,
+      "grad_norm": 0.5590435801827939,
+      "learning_rate": 3.121352102768998e-05,
+      "loss": 0.5629,
+      "step": 148
+    },
+    {
+      "epoch": 1.9041533546325877,
+      "grad_norm": 0.6262150274271875,
+      "learning_rate": 3.106484881410628e-05,
+      "loss": 0.5475,
+      "step": 149
+    },
+    {
+      "epoch": 1.9169329073482428,
+      "grad_norm": 0.7182303942682138,
+      "learning_rate": 3.091529020389009e-05,
+      "loss": 0.5446,
+      "step": 150
+    },
+    {
+      "epoch": 1.9297124600638977,
+      "grad_norm": 0.4340720332975557,
+      "learning_rate": 3.076485717806808e-05,
+      "loss": 0.5435,
+      "step": 151
+    },
+    {
+      "epoch": 1.9424920127795526,
+      "grad_norm": 0.572990989223066,
+      "learning_rate": 3.061356178771564e-05,
+      "loss": 0.5419,
+      "step": 152
+    },
+    {
+      "epoch": 1.9552715654952078,
+      "grad_norm": 0.5761722054308671,
+      "learning_rate": 3.0461416152991555e-05,
+      "loss": 0.545,
+      "step": 153
+    },
+    {
+      "epoch": 1.9680511182108626,
+      "grad_norm": 0.47715951155810205,
+      "learning_rate": 3.0308432462167045e-05,
+      "loss": 0.5408,
+      "step": 154
+    },
+    {
+      "epoch": 1.9808306709265175,
+      "grad_norm": 0.44500377408250574,
+      "learning_rate": 3.015462297064936e-05,
+      "loss": 0.5281,
+      "step": 155
+    },
+    {
+      "epoch": 1.9936102236421727,
+      "grad_norm": 0.5298218047198479,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.5289,
+      "step": 156
+    },
+    {
+      "epoch": 2.0063897763578273,
+      "grad_norm": 0.486757795281033,
+      "learning_rate": 2.98445759369477e-05,
+      "loss": 0.5006,
+      "step": 157
+    },
+    {
+      "epoch": 2.0191693290734825,
+      "grad_norm": 0.46748405855369674,
+      "learning_rate": 2.9688363232396056e-05,
+      "loss": 0.4602,
+      "step": 158
+    },
+    {
+      "epoch": 2.0319488817891376,
+      "grad_norm": 0.9032320953813431,
+      "learning_rate": 2.9531374400426158e-05,
+      "loss": 0.4938,
+      "step": 159
+    },
+    {
+      "epoch": 2.0447284345047922,
+      "grad_norm": 0.7141637648465875,
+      "learning_rate": 2.9373622017294075e-05,
+      "loss": 0.4683,
+      "step": 160
+    },
+    {
+      "epoch": 2.0575079872204474,
+      "grad_norm": 0.6638457299888288,
+      "learning_rate": 2.9215118720423375e-05,
+      "loss": 0.4655,
+      "step": 161
+    },
+    {
+      "epoch": 2.070287539936102,
+      "grad_norm": 0.5740802765073745,
+      "learning_rate": 2.9055877207392752e-05,
+      "loss": 0.4664,
+      "step": 162
+    },
+    {
+      "epoch": 2.083067092651757,
+      "grad_norm": 0.541980004503229,
+      "learning_rate": 2.8895910234918828e-05,
+      "loss": 0.4594,
+      "step": 163
+    },
+    {
+      "epoch": 2.0958466453674123,
+      "grad_norm": 0.7230973797959268,
+      "learning_rate": 2.873523061783426e-05,
+      "loss": 0.4535,
+      "step": 164
+    },
+    {
+      "epoch": 2.108626198083067,
+      "grad_norm": 0.5775309097959617,
+      "learning_rate": 2.8573851228061084e-05,
+      "loss": 0.4469,
+      "step": 165
+    },
+    {
+      "epoch": 2.121405750798722,
+      "grad_norm": 0.7114460224550682,
+      "learning_rate": 2.8411784993579633e-05,
+      "loss": 0.4453,
+      "step": 166
+    },
+    {
+      "epoch": 2.134185303514377,
+      "grad_norm": 0.5544199950880329,
+      "learning_rate": 2.8249044897392814e-05,
+      "loss": 0.4447,
+      "step": 167
+    },
+    {
+      "epoch": 2.146964856230032,
+      "grad_norm": 0.7751203169810186,
+      "learning_rate": 2.80856439764861e-05,
+      "loss": 0.4627,
+      "step": 168
+    },
+    {
+      "epoch": 2.159744408945687,
+      "grad_norm": 0.6633500729778636,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.4621,
+      "step": 169
+    },
+    {
+      "epoch": 2.1725239616613417,
+      "grad_norm": 0.5933550753958191,
+      "learning_rate": 2.77569120720971e-05,
+      "loss": 0.4568,
+      "step": 170
+    },
+    {
+      "epoch": 2.1853035143769968,
+      "grad_norm": 0.7185282353498863,
+      "learning_rate": 2.7591607423077932e-05,
+      "loss": 0.4669,
+      "step": 171
+    },
+    {
+      "epoch": 2.198083067092652,
+      "grad_norm": 0.4752269249605016,
+      "learning_rate": 2.7425694616155474e-05,
+      "loss": 0.4418,
+      "step": 172
+    },
+    {
+      "epoch": 2.2108626198083066,
+      "grad_norm": 0.5405713655934162,
+      "learning_rate": 2.7259186942478656e-05,
+      "loss": 0.4496,
+      "step": 173
+    },
+    {
+      "epoch": 2.2236421725239617,
+      "grad_norm": 0.5215249693639501,
+      "learning_rate": 2.7092097740850712e-05,
+      "loss": 0.4599,
+      "step": 174
+    },
+    {
+      "epoch": 2.236421725239617,
+      "grad_norm": 0.42509149953729447,
+      "learning_rate": 2.692444039666066e-05,
+      "loss": 0.4538,
+      "step": 175
+    },
+    {
+      "epoch": 2.2492012779552715,
+      "grad_norm": 0.443104585356185,
+      "learning_rate": 2.6756228340810946e-05,
+      "loss": 0.4423,
+      "step": 176
+    },
+    {
+      "epoch": 2.2619808306709266,
+      "grad_norm": 0.4244498485086766,
+      "learning_rate": 2.6587475048641596e-05,
+      "loss": 0.4581,
+      "step": 177
+    },
+    {
+      "epoch": 2.2747603833865817,
+      "grad_norm": 0.4427678121202283,
+      "learning_rate": 2.6418194038850634e-05,
+      "loss": 0.4622,
+      "step": 178
+    },
+    {
+      "epoch": 2.2875399361022364,
+      "grad_norm": 0.44158722004763423,
+      "learning_rate": 2.624839887241115e-05,
+      "loss": 0.4666,
+      "step": 179
+    },
+    {
+      "epoch": 2.3003194888178915,
+      "grad_norm": 0.4314748854864371,
+      "learning_rate": 2.607810315148494e-05,
+      "loss": 0.4681,
+      "step": 180
+    },
+    {
+      "epoch": 2.313099041533546,
+      "grad_norm": 0.41655120357135617,
+      "learning_rate": 2.5907320518332827e-05,
+      "loss": 0.4539,
+      "step": 181
+    },
+    {
+      "epoch": 2.3258785942492013,
+      "grad_norm": 0.4203925750652448,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.4427,
+      "step": 182
+    },
+    {
+      "epoch": 2.3386581469648564,
+      "grad_norm": 0.3893882507171406,
+      "learning_rate": 2.5564349278329056e-05,
+      "loss": 0.4575,
+      "step": 183
+    },
+    {
+      "epoch": 2.351437699680511,
+      "grad_norm": 0.39085497976651834,
+      "learning_rate": 2.539218814664288e-05,
+      "loss": 0.4463,
+      "step": 184
+    },
+    {
+      "epoch": 2.364217252396166,
+      "grad_norm": 0.4703088592774784,
+      "learning_rate": 2.521959505086075e-05,
+      "loss": 0.4521,
+      "step": 185
+    },
+    {
+      "epoch": 2.376996805111821,
+      "grad_norm": 0.48500913272959334,
+      "learning_rate": 2.5046583817284437e-05,
+      "loss": 0.4608,
+      "step": 186
+    },
+    {
+      "epoch": 2.389776357827476,
+      "grad_norm": 0.43887176139587314,
+      "learning_rate": 2.487316830571244e-05,
+      "loss": 0.4835,
+      "step": 187
+    },
+    {
+      "epoch": 2.402555910543131,
+      "grad_norm": 0.464584961974617,
+      "learning_rate": 2.4699362408329646e-05,
+      "loss": 0.4604,
+      "step": 188
+    },
+    {
+      "epoch": 2.415335463258786,
+      "grad_norm": 0.4154167662123907,
+      "learning_rate": 2.4525180048594452e-05,
+      "loss": 0.4446,
+      "step": 189
+    },
+    {
+      "epoch": 2.428115015974441,
+      "grad_norm": 0.46994597432826285,
+      "learning_rate": 2.435063518012335e-05,
+      "loss": 0.4723,
+      "step": 190
+    },
+    {
+      "epoch": 2.440894568690096,
+      "grad_norm": 0.39642847919224067,
+      "learning_rate": 2.4175741785573177e-05,
+      "loss": 0.4441,
+      "step": 191
+    },
+    {
+      "epoch": 2.4536741214057507,
+      "grad_norm": 0.42331312684198885,
+      "learning_rate": 2.4000513875520892e-05,
+      "loss": 0.4751,
+      "step": 192
+    },
+    {
+      "epoch": 2.466453674121406,
+      "grad_norm": 0.42637579603720094,
+      "learning_rate": 2.3824965487341247e-05,
+      "loss": 0.4503,
+      "step": 193
+    },
+    {
+      "epoch": 2.479233226837061,
+      "grad_norm": 0.4176770140541874,
+      "learning_rate": 2.3649110684082258e-05,
+      "loss": 0.4614,
+      "step": 194
+    },
+    {
+      "epoch": 2.4920127795527156,
+      "grad_norm": 0.4383167963140633,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.4475,
+      "step": 195
+    },
+    {
+      "epoch": 2.5047923322683707,
+      "grad_norm": 0.3533765206549483,
+      "learning_rate": 2.3296538206123134e-05,
+      "loss": 0.4717,
+      "step": 196
+    },
+    {
+      "epoch": 2.5175718849840254,
+      "grad_norm": 0.4802148000025239,
+      "learning_rate": 2.311984877573636e-05,
+      "loss": 0.4587,
+      "step": 197
+    },
+    {
+      "epoch": 2.5303514376996805,
+      "grad_norm": 0.4884867524733879,
+      "learning_rate": 2.2942909416634326e-05,
+      "loss": 0.471,
+      "step": 198
+    },
+    {
+      "epoch": 2.543130990415335,
+      "grad_norm": 0.38302934162963237,
+      "learning_rate": 2.2765734303294666e-05,
+      "loss": 0.4663,
+      "step": 199
+    },
+    {
+      "epoch": 2.5559105431309903,
+      "grad_norm": 0.4919891722403176,
+      "learning_rate": 2.2588337629081107e-05,
+      "loss": 0.4624,
+      "step": 200
+    },
+    {
+      "epoch": 2.5686900958466454,
+      "grad_norm": 0.42439763122441787,
+      "learning_rate": 2.2410733605106462e-05,
+      "loss": 0.4733,
+      "step": 201
+    },
+    {
+      "epoch": 2.5814696485623,
+      "grad_norm": 0.455600248616356,
+      "learning_rate": 2.2232936459094158e-05,
+      "loss": 0.4719,
+      "step": 202
+    },
+    {
+      "epoch": 2.594249201277955,
+      "grad_norm": 0.37772141842234497,
+      "learning_rate": 2.205496043423849e-05,
+      "loss": 0.4643,
+      "step": 203
+    },
+    {
+      "epoch": 2.6070287539936103,
+      "grad_norm": 0.39487928087355095,
+      "learning_rate": 2.1876819788063586e-05,
+      "loss": 0.4599,
+      "step": 204
+    },
+    {
+      "epoch": 2.619808306709265,
+      "grad_norm": 0.37752084464288793,
+      "learning_rate": 2.16985287912813e-05,
+      "loss": 0.4611,
+      "step": 205
+    },
+    {
+      "epoch": 2.63258785942492,
+      "grad_norm": 0.41690300860275786,
+      "learning_rate": 2.1520101726647922e-05,
+      "loss": 0.4656,
+      "step": 206
+    },
+    {
+      "epoch": 2.6453674121405752,
+      "grad_norm": 0.35278905850290787,
+      "learning_rate": 2.1341552887820048e-05,
+      "loss": 0.4509,
+      "step": 207
+    },
+    {
+      "epoch": 2.65814696485623,
+      "grad_norm": 0.3952082192254116,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.4552,
+      "step": 208
+    },
+    {
+      "epoch": 2.670926517571885,
+      "grad_norm": 0.37698129545807996,
+      "learning_rate": 2.0984147109837564e-05,
+      "loss": 0.4587,
+      "step": 209
+    },
+    {
+      "epoch": 2.68370607028754,
+      "grad_norm": 0.3385957900251669,
+      "learning_rate": 2.0805318802188307e-05,
+      "loss": 0.4708,
+      "step": 210
+    },
+    {
+      "epoch": 2.696485623003195,
+      "grad_norm": 0.3715034095313426,
+      "learning_rate": 2.0626425981061608e-05,
+      "loss": 0.4494,
+      "step": 211
+    },
+    {
+      "epoch": 2.70926517571885,
+      "grad_norm": 0.3251290710767365,
+      "learning_rate": 2.0447482977425465e-05,
+      "loss": 0.4563,
+      "step": 212
+    },
+    {
+      "epoch": 2.722044728434505,
+      "grad_norm": 0.37877042575127295,
+      "learning_rate": 2.0268504126267952e-05,
+      "loss": 0.4495,
+      "step": 213
+    },
+    {
+      "epoch": 2.7348242811501597,
+      "grad_norm": 0.3513613274662141,
+      "learning_rate": 2.008950376544887e-05,
+      "loss": 0.4681,
+      "step": 214
+    },
+    {
+      "epoch": 2.747603833865815,
+      "grad_norm": 0.3673333816439615,
+      "learning_rate": 1.9910496234551132e-05,
+      "loss": 0.4579,
+      "step": 215
+    },
+    {
+      "epoch": 2.7603833865814695,
+      "grad_norm": 0.4002781230499196,
+      "learning_rate": 1.9731495873732055e-05,
+      "loss": 0.4631,
+      "step": 216
+    },
+    {
+      "epoch": 2.7731629392971247,
+      "grad_norm": 0.4144225996048057,
+      "learning_rate": 1.9552517022574542e-05,
+      "loss": 0.4588,
+      "step": 217
+    },
+    {
+      "epoch": 2.7859424920127793,
+      "grad_norm": 0.38483992779543585,
+      "learning_rate": 1.93735740189384e-05,
+      "loss": 0.4554,
+      "step": 218
+    },
+    {
+      "epoch": 2.7987220447284344,
+      "grad_norm": 0.39775925575971094,
+      "learning_rate": 1.9194681197811703e-05,
+      "loss": 0.4379,
+      "step": 219
+    },
+    {
+      "epoch": 2.8115015974440896,
+      "grad_norm": 0.38104448687547743,
+      "learning_rate": 1.901585289016244e-05,
+      "loss": 0.4536,
+      "step": 220
+    },
+    {
+      "epoch": 2.8242811501597442,
+      "grad_norm": 0.37218167438181177,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.4546,
+      "step": 221
+    },
+    {
+      "epoch": 2.8370607028753994,
+      "grad_norm": 0.3238264848569907,
+      "learning_rate": 1.8658447112179952e-05,
+      "loss": 0.4971,
+      "step": 222
+    },
+    {
+      "epoch": 2.8498402555910545,
+      "grad_norm": 0.3669356917599918,
+      "learning_rate": 1.8479898273352084e-05,
+      "loss": 0.4636,
+      "step": 223
+    },
+    {
+      "epoch": 2.862619808306709,
+      "grad_norm": 0.34897969771117815,
+      "learning_rate": 1.83014712087187e-05,
+      "loss": 0.4454,
+      "step": 224
+    },
+    {
+      "epoch": 2.8753993610223643,
+      "grad_norm": 0.3531428513148952,
+      "learning_rate": 1.8123180211936417e-05,
+      "loss": 0.4321,
+      "step": 225
+    },
+    {
+      "epoch": 2.8881789137380194,
+      "grad_norm": 0.3155910003073882,
+      "learning_rate": 1.794503956576152e-05,
+      "loss": 0.4737,
+      "step": 226
+    },
+    {
+      "epoch": 2.900958466453674,
+      "grad_norm": 0.3514454764533909,
+      "learning_rate": 1.776706354090585e-05,
+      "loss": 0.4699,
+      "step": 227
+    },
+    {
+      "epoch": 2.913738019169329,
+      "grad_norm": 0.32261320466049764,
+      "learning_rate": 1.758926639489354e-05,
+      "loss": 0.449,
+      "step": 228
+    },
+    {
+      "epoch": 2.9265175718849843,
+      "grad_norm": 0.31699325291611036,
+      "learning_rate": 1.7411662370918893e-05,
+      "loss": 0.441,
+      "step": 229
+    },
+    {
+      "epoch": 2.939297124600639,
+      "grad_norm": 0.3465501067615838,
+      "learning_rate": 1.7234265696705344e-05,
+      "loss": 0.4747,
+      "step": 230
+    },
+    {
+      "epoch": 2.952076677316294,
+      "grad_norm": 0.3390277128709767,
+      "learning_rate": 1.7057090583365678e-05,
+      "loss": 0.4598,
+      "step": 231
+    },
+    {
+      "epoch": 2.9648562300319488,
+      "grad_norm": 0.36925982516132494,
+      "learning_rate": 1.6880151224263646e-05,
+      "loss": 0.4641,
+      "step": 232
+    },
+    {
+      "epoch": 2.977635782747604,
+      "grad_norm": 0.32500966228489814,
+      "learning_rate": 1.6703461793876876e-05,
+      "loss": 0.4586,
+      "step": 233
+    },
+    {
+      "epoch": 2.9904153354632586,
+      "grad_norm": 0.34908654244155446,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.4662,
+      "step": 234
+    },
+    {
+      "epoch": 3.012779552715655,
+      "grad_norm": 0.5128836169014913,
+      "learning_rate": 1.635088931591775e-05,
+      "loss": 0.3811,
+      "step": 235
+    },
+    {
+      "epoch": 3.02555910543131,
+      "grad_norm": 0.3546127617977104,
+      "learning_rate": 1.6175034512658753e-05,
+      "loss": 0.3924,
+      "step": 236
+    },
+    {
+      "epoch": 3.038338658146965,
+      "grad_norm": 0.6442307260367379,
+      "learning_rate": 1.5999486124479115e-05,
+      "loss": 0.393,
+      "step": 237
+    },
+    {
+      "epoch": 3.0511182108626196,
+      "grad_norm": 0.5003594645494083,
+      "learning_rate": 1.5824258214426833e-05,
+      "loss": 0.3805,
+      "step": 238
+    },
+    {
+      "epoch": 3.0638977635782747,
+      "grad_norm": 0.4609554964301076,
+      "learning_rate": 1.5649364819876655e-05,
+      "loss": 0.4057,
+      "step": 239
+    },
+    {
+      "epoch": 3.07667731629393,
+      "grad_norm": 0.4976572991284983,
+      "learning_rate": 1.547481995140556e-05,
+      "loss": 0.3879,
+      "step": 240
+    },
+    {
+      "epoch": 3.0894568690095845,
+      "grad_norm": 0.38362452674206254,
+      "learning_rate": 1.5300637591670357e-05,
+      "loss": 0.3733,
+      "step": 241
+    },
+    {
+      "epoch": 3.1022364217252396,
+      "grad_norm": 0.45482995584383695,
+      "learning_rate": 1.5126831694287564e-05,
+      "loss": 0.3696,
+      "step": 242
+    },
+    {
+      "epoch": 3.1150159744408947,
+      "grad_norm": 0.44666813959914614,
+      "learning_rate": 1.4953416182715566e-05,
+      "loss": 0.3736,
+      "step": 243
+    },
+    {
+      "epoch": 3.1277955271565494,
+      "grad_norm": 0.4292571016117774,
+      "learning_rate": 1.478040494913926e-05,
+      "loss": 0.3827,
+      "step": 244
+    },
+    {
+      "epoch": 3.1405750798722045,
+      "grad_norm": 0.46347737412533785,
+      "learning_rate": 1.460781185335713e-05,
+      "loss": 0.3923,
+      "step": 245
+    },
+    {
+      "epoch": 3.1533546325878596,
+      "grad_norm": 0.4775852522273919,
+      "learning_rate": 1.443565072167095e-05,
+      "loss": 0.4088,
+      "step": 246
+    },
+    {
+      "epoch": 3.1661341853035143,
+      "grad_norm": 0.41083533606012235,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.3801,
+      "step": 247
+    },
+    {
+      "epoch": 3.1789137380191694,
+      "grad_norm": 0.4484362081329354,
+      "learning_rate": 1.409267948166718e-05,
+      "loss": 0.376,
+      "step": 248
+    },
+    {
+      "epoch": 3.191693290734824,
+      "grad_norm": 0.3906442765094229,
+      "learning_rate": 1.3921896848515064e-05,
+      "loss": 0.3815,
+      "step": 249
+    },
+    {
+      "epoch": 3.2044728434504792,
+      "grad_norm": 0.3700225581587479,
+      "learning_rate": 1.3751601127588849e-05,
+      "loss": 0.3826,
+      "step": 250
+    },
+    {
+      "epoch": 3.2172523961661343,
+      "grad_norm": 0.38606672559398486,
+      "learning_rate": 1.3581805961149371e-05,
+      "loss": 0.3779,
+      "step": 251
+    },
+    {
+      "epoch": 3.230031948881789,
+      "grad_norm": 0.3643687933394165,
+      "learning_rate": 1.341252495135841e-05,
+      "loss": 0.3734,
+      "step": 252
+    },
+    {
+      "epoch": 3.242811501597444,
+      "grad_norm": 0.3859244143097149,
+      "learning_rate": 1.324377165918906e-05,
+      "loss": 0.3828,
+      "step": 253
+    },
+    {
+      "epoch": 3.255591054313099,
+      "grad_norm": 0.387692164875178,
+      "learning_rate": 1.3075559603339354e-05,
+      "loss": 0.3811,
+      "step": 254
+    },
+    {
+      "epoch": 3.268370607028754,
+      "grad_norm": 0.35085934782850087,
+      "learning_rate": 1.2907902259149287e-05,
+      "loss": 0.3861,
+      "step": 255
+    },
+    {
+      "epoch": 3.281150159744409,
+      "grad_norm": 0.4166197346329323,
+      "learning_rate": 1.274081305752135e-05,
+      "loss": 0.3934,
+      "step": 256
+    },
+    {
+      "epoch": 3.2939297124600637,
+      "grad_norm": 0.3418519825744561,
+      "learning_rate": 1.2574305383844528e-05,
+      "loss": 0.3877,
+      "step": 257
+    },
+    {
+      "epoch": 3.306709265175719,
+      "grad_norm": 0.36354782411908043,
+      "learning_rate": 1.2408392576922075e-05,
+      "loss": 0.3849,
+      "step": 258
+    },
+    {
+      "epoch": 3.319488817891374,
+      "grad_norm": 0.37242419363296814,
+      "learning_rate": 1.2243087927902905e-05,
+      "loss": 0.3937,
+      "step": 259
+    },
+    {
+      "epoch": 3.3322683706070286,
+      "grad_norm": 0.40415705465213253,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.3826,
+      "step": 260
+    },
+    {
+      "epoch": 3.3450479233226837,
+      "grad_norm": 0.3178335493857135,
+      "learning_rate": 1.1914356023513904e-05,
+      "loss": 0.3834,
+      "step": 261
+    },
+    {
+      "epoch": 3.357827476038339,
+      "grad_norm": 0.3648587053326533,
+      "learning_rate": 1.1750955102607193e-05,
+      "loss": 0.3699,
+      "step": 262
+    },
+    {
+      "epoch": 3.3706070287539935,
+      "grad_norm": 0.32908420449476,
+      "learning_rate": 1.1588215006420374e-05,
+      "loss": 0.406,
+      "step": 263
+    },
+    {
+      "epoch": 3.3833865814696487,
+      "grad_norm": 0.3230655828560371,
+      "learning_rate": 1.1426148771938915e-05,
+      "loss": 0.4007,
+      "step": 264
+    },
+    {
+      "epoch": 3.3961661341853033,
+      "grad_norm": 0.364247334085622,
+      "learning_rate": 1.1264769382165748e-05,
+      "loss": 0.3912,
+      "step": 265
+    },
+    {
+      "epoch": 3.4089456869009584,
+      "grad_norm": 0.34854917848030975,
+      "learning_rate": 1.110408976508118e-05,
+      "loss": 0.3981,
+      "step": 266
+    },
+    {
+      "epoch": 3.4217252396166136,
+      "grad_norm": 0.32853287611588194,
+      "learning_rate": 1.094412279260726e-05,
+      "loss": 0.3846,
+      "step": 267
+    },
+    {
+      "epoch": 3.4345047923322682,
+      "grad_norm": 0.3407376822824476,
+      "learning_rate": 1.0784881279576635e-05,
+      "loss": 0.3877,
+      "step": 268
+    },
+    {
+      "epoch": 3.4472843450479234,
+      "grad_norm": 0.3161855560360877,
+      "learning_rate": 1.0626377982705929e-05,
+      "loss": 0.3635,
+      "step": 269
+    },
+    {
+      "epoch": 3.460063897763578,
+      "grad_norm": 0.29388908626785254,
+      "learning_rate": 1.0468625599573842e-05,
+      "loss": 0.3946,
+      "step": 270
+    },
+    {
+      "epoch": 3.472843450479233,
+      "grad_norm": 0.31849610999090494,
+      "learning_rate": 1.0311636767603952e-05,
+      "loss": 0.3731,
+      "step": 271
+    },
+    {
+      "epoch": 3.4856230031948883,
+      "grad_norm": 0.3208723917598162,
+      "learning_rate": 1.0155424063052306e-05,
+      "loss": 0.3673,
+      "step": 272
+    },
+    {
+      "epoch": 3.498402555910543,
+      "grad_norm": 0.29245116811873584,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.3796,
+      "step": 273
+    },
+    {
+      "epoch": 3.511182108626198,
+      "grad_norm": 0.33746616232506677,
+      "learning_rate": 9.84537702935065e-06,
+      "loss": 0.386,
+      "step": 274
+    },
+    {
+      "epoch": 3.523961661341853,
+      "grad_norm": 0.29309254314881333,
+      "learning_rate": 9.691567537832964e-06,
+      "loss": 0.4085,
+      "step": 275
+    },
+    {
+      "epoch": 3.536741214057508,
+      "grad_norm": 0.28232085496123377,
+      "learning_rate": 9.538583847008452e-06,
+      "loss": 0.3766,
+      "step": 276
+    },
+    {
+      "epoch": 3.549520766773163,
+      "grad_norm": 0.30542807138023287,
+      "learning_rate": 9.386438212284372e-06,
+      "loss": 0.3713,
+      "step": 277
+    },
+    {
+      "epoch": 3.562300319488818,
+      "grad_norm": 0.3001693318285487,
+      "learning_rate": 9.235142821931928e-06,
+      "loss": 0.3621,
+      "step": 278
+    },
+    {
+      "epoch": 3.5750798722044728,
+      "grad_norm": 0.2706474019112891,
+      "learning_rate": 9.084709796109907e-06,
+      "loss": 0.4121,
+      "step": 279
+    },
+    {
+      "epoch": 3.587859424920128,
+      "grad_norm": 0.2895527171945164,
+      "learning_rate": 8.93515118589373e-06,
+      "loss": 0.3948,
+      "step": 280
+    },
+    {
+      "epoch": 3.600638977635783,
+      "grad_norm": 0.3117817510807347,
+      "learning_rate": 8.786478972310023e-06,
+      "loss": 0.401,
+      "step": 281
+    },
+    {
+      "epoch": 3.6134185303514377,
+      "grad_norm": 0.26216491972350675,
+      "learning_rate": 8.638705065376887e-06,
+      "loss": 0.3952,
+      "step": 282
+    },
+    {
+      "epoch": 3.626198083067093,
+      "grad_norm": 0.284049361201387,
+      "learning_rate": 8.491841303149728e-06,
+      "loss": 0.3943,
+      "step": 283
+    },
+    {
+      "epoch": 3.6389776357827475,
+      "grad_norm": 0.2942161552846955,
+      "learning_rate": 8.345899450772975e-06,
+      "loss": 0.4198,
+      "step": 284
+    },
+    {
+      "epoch": 3.6517571884984026,
+      "grad_norm": 0.2815277755618436,
+      "learning_rate": 8.200891199537549e-06,
+      "loss": 0.3775,
+      "step": 285
+    },
+    {
+      "epoch": 3.6645367412140573,
+      "grad_norm": 0.27855689397961697,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.3701,
+      "step": 286
+    },
+    {
+      "epoch": 3.6773162939297124,
+      "grad_norm": 0.29191871129995184,
+      "learning_rate": 7.913721890773354e-06,
+      "loss": 0.3736,
+      "step": 287
+    },
+    {
+      "epoch": 3.6900958466453675,
+      "grad_norm": 0.26869731691732596,
+      "learning_rate": 7.771583838159756e-06,
+      "loss": 0.3831,
+      "step": 288
+    },
+    {
+      "epoch": 3.702875399361022,
+      "grad_norm": 0.2671559120158343,
+      "learning_rate": 7.630425394674903e-06,
+      "loss": 0.366,
+      "step": 289
+    },
+    {
+      "epoch": 3.7156549520766773,
+      "grad_norm": 0.29113928428290825,
+      "learning_rate": 7.49025786841445e-06,
+      "loss": 0.378,
+      "step": 290
+    },
+    {
+      "epoch": 3.7284345047923324,
+      "grad_norm": 0.2994031778893088,
+      "learning_rate": 7.3510924880924575e-06,
+      "loss": 0.3774,
+      "step": 291
+    },
+    {
+      "epoch": 3.741214057507987,
+      "grad_norm": 0.2738974817232764,
+      "learning_rate": 7.212940402141808e-06,
+      "loss": 0.3846,
+      "step": 292
+    },
+    {
+      "epoch": 3.753993610223642,
+      "grad_norm": 0.2826921730840266,
+      "learning_rate": 7.075812677821145e-06,
+      "loss": 0.3727,
+      "step": 293
+    },
+    {
+      "epoch": 3.7667731629392973,
+      "grad_norm": 0.26668241839234336,
+      "learning_rate": 6.939720300328303e-06,
+      "loss": 0.391,
+      "step": 294
+    },
+    {
+      "epoch": 3.779552715654952,
+      "grad_norm": 0.2529079411168859,
+      "learning_rate": 6.8046741719202385e-06,
+      "loss": 0.3945,
+      "step": 295
+    },
+    {
+      "epoch": 3.792332268370607,
+      "grad_norm": 0.26348898417120825,
+      "learning_rate": 6.67068511103971e-06,
+      "loss": 0.3764,
+      "step": 296
+    },
+    {
+      "epoch": 3.8051118210862622,
+      "grad_norm": 0.2766036475667321,
+      "learning_rate": 6.537763851448593e-06,
+      "loss": 0.3857,
+      "step": 297
+    },
+    {
+      "epoch": 3.817891373801917,
+      "grad_norm": 0.2682733264498869,
+      "learning_rate": 6.4059210413680175e-06,
+      "loss": 0.3873,
+      "step": 298
+    },
+    {
+      "epoch": 3.830670926517572,
+      "grad_norm": 0.2742368830967292,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.3766,
+      "step": 299
+    },
+    {
+      "epoch": 3.8434504792332267,
+      "grad_norm": 0.25114479786735716,
+      "learning_rate": 6.145512929808013e-06,
+      "loss": 0.3975,
+      "step": 300
+    },
+    {
+      "epoch": 3.856230031948882,
+      "grad_norm": 0.2630221067698174,
+      "learning_rate": 6.016968489424572e-06,
+      "loss": 0.3816,
+      "step": 301
+    },
+    {
+      "epoch": 3.8690095846645365,
+      "grad_norm": 0.2693660116982185,
+      "learning_rate": 5.889544219072465e-06,
+      "loss": 0.3916,
+      "step": 302
+    },
+    {
+      "epoch": 3.8817891373801916,
+      "grad_norm": 0.26878930310587784,
+      "learning_rate": 5.7632503266131925e-06,
+      "loss": 0.3929,
+      "step": 303
+    },
+    {
+      "epoch": 3.8945686900958467,
+      "grad_norm": 0.2576790616305518,
+      "learning_rate": 5.638096929354522e-06,
+      "loss": 0.3934,
+      "step": 304
+    },
+    {
+      "epoch": 3.9073482428115014,
+      "grad_norm": 0.2551594985684908,
+      "learning_rate": 5.514094053240035e-06,
+      "loss": 0.3971,
+      "step": 305
+    },
+    {
+      "epoch": 3.9201277955271565,
+      "grad_norm": 0.26282254693750906,
+      "learning_rate": 5.39125163204594e-06,
+      "loss": 0.3852,
+      "step": 306
+    },
+    {
+      "epoch": 3.9329073482428116,
+      "grad_norm": 0.2729750102242486,
+      "learning_rate": 5.269579506585259e-06,
+      "loss": 0.3792,
+      "step": 307
+    },
+    {
+      "epoch": 3.9456869009584663,
+      "grad_norm": 0.2746715339249571,
+      "learning_rate": 5.149087423919541e-06,
+      "loss": 0.3809,
+      "step": 308
+    },
+    {
+      "epoch": 3.9584664536741214,
+      "grad_norm": 0.2589108681555479,
+      "learning_rate": 5.029785036577976e-06,
+      "loss": 0.38,
+      "step": 309
+    },
+    {
+      "epoch": 3.9712460063897765,
+      "grad_norm": 0.2583977076620543,
+      "learning_rate": 4.911681901784198e-06,
+      "loss": 0.3823,
+      "step": 310
+    },
+    {
+      "epoch": 3.984025559105431,
+      "grad_norm": 0.26533553439913465,
+      "learning_rate": 4.794787480690597e-06,
+      "loss": 0.4095,
+      "step": 311
+    },
+    {
+      "epoch": 3.9968051118210863,
+      "grad_norm": 0.29179544366540294,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.3764,
+      "step": 312
+    },
+    {
+      "epoch": 4.0095846645367414,
+      "grad_norm": 0.37070974270993895,
+      "learning_rate": 4.5646621393177e-06,
+      "loss": 0.3437,
+      "step": 313
+    },
+    {
+      "epoch": 4.022364217252396,
+      "grad_norm": 0.3284947553106813,
+      "learning_rate": 4.451449654204685e-06,
+      "loss": 0.3662,
+      "step": 314
+    },
+    {
+      "epoch": 4.035143769968051,
+      "grad_norm": 0.2843402603459979,
+      "learning_rate": 4.339482751647557e-06,
+      "loss": 0.3355,
+      "step": 315
+    },
+    {
+      "epoch": 4.047923322683706,
+      "grad_norm": 0.33919450346249647,
+      "learning_rate": 4.228770401229824e-06,
+      "loss": 0.3394,
+      "step": 316
+    },
+    {
+      "epoch": 4.060702875399361,
+      "grad_norm": 0.4472299669588526,
+      "learning_rate": 4.119321472033779e-06,
+      "loss": 0.3384,
+      "step": 317
+    },
+    {
+      "epoch": 4.073482428115016,
+      "grad_norm": 0.3771224880409924,
+      "learning_rate": 4.011144731929981e-06,
+      "loss": 0.3434,
+      "step": 318
+    },
+    {
+      "epoch": 4.086261980830671,
+      "grad_norm": 0.2916380797164523,
+      "learning_rate": 3.904248846874894e-06,
+      "loss": 0.32,
+      "step": 319
+    },
+    {
+      "epoch": 4.099041533546326,
+      "grad_norm": 0.33504974166135404,
+      "learning_rate": 3.7986423802166705e-06,
+      "loss": 0.3235,
+      "step": 320
+    },
+    {
+      "epoch": 4.111821086261981,
+      "grad_norm": 0.33212341390770345,
+      "learning_rate": 3.694333792009115e-06,
+      "loss": 0.3452,
+      "step": 321
+    },
+    {
+      "epoch": 4.124600638977636,
+      "grad_norm": 0.31976825359898686,
+      "learning_rate": 3.5913314383339937e-06,
+      "loss": 0.3473,
+      "step": 322
+    },
+    {
+      "epoch": 4.137380191693291,
+      "grad_norm": 0.28590474501240015,
+      "learning_rate": 3.4896435706316e-06,
+      "loss": 0.319,
+      "step": 323
+    },
+    {
+      "epoch": 4.1501597444089455,
+      "grad_norm": 0.286464326555546,
+      "learning_rate": 3.3892783350397675e-06,
+      "loss": 0.3473,
+      "step": 324
+    },
+    {
+      "epoch": 4.1629392971246,
+      "grad_norm": 0.3082641016025165,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.3522,
+      "step": 325
+    },
+    {
+      "epoch": 4.175718849840256,
+      "grad_norm": 0.30101579695268377,
+      "learning_rate": 3.1925478143197418e-06,
+      "loss": 0.3364,
+      "step": 326
+    },
+    {
+      "epoch": 4.18849840255591,
+      "grad_norm": 0.30620894785433056,
+      "learning_rate": 3.0961982891241083e-06,
+      "loss": 0.3383,
+      "step": 327
+    },
+    {
+      "epoch": 4.201277955271565,
+      "grad_norm": 0.26164000104597274,
+      "learning_rate": 3.001202914641628e-06,
+      "loss": 0.3387,
+      "step": 328
+    },
+    {
+      "epoch": 4.214057507987221,
+      "grad_norm": 0.27396421835749335,
+      "learning_rate": 2.907569300879596e-06,
+      "loss": 0.3359,
+      "step": 329
+    },
+    {
+      "epoch": 4.226837060702875,
+      "grad_norm": 0.273485919076872,
+      "learning_rate": 2.815304948755664e-06,
+      "loss": 0.3116,
+      "step": 330
+    },
+    {
+      "epoch": 4.23961661341853,
+      "grad_norm": 0.2569560566217525,
+      "learning_rate": 2.7244172494969978e-06,
+      "loss": 0.3369,
+      "step": 331
+    },
+    {
+      "epoch": 4.252396166134186,
+      "grad_norm": 0.24962527429270018,
+      "learning_rate": 2.6349134840481294e-06,
+      "loss": 0.3716,
+      "step": 332
+    },
+    {
+      "epoch": 4.26517571884984,
+      "grad_norm": 0.24800106442570202,
+      "learning_rate": 2.546800822487714e-06,
+      "loss": 0.3573,
+      "step": 333
+    },
+    {
+      "epoch": 4.277955271565495,
+      "grad_norm": 0.25821807561951066,
+      "learning_rate": 2.4600863234541338e-06,
+      "loss": 0.3409,
+      "step": 334
+    },
+    {
+      "epoch": 4.2907348242811505,
+      "grad_norm": 0.25158547315449414,
+      "learning_rate": 2.374776933580025e-06,
+      "loss": 0.3146,
+      "step": 335
+    },
+    {
+      "epoch": 4.303514376996805,
+      "grad_norm": 0.24647625625675712,
+      "learning_rate": 2.2908794869358044e-06,
+      "loss": 0.3442,
+      "step": 336
+    },
+    {
+      "epoch": 4.31629392971246,
+      "grad_norm": 0.2534440552420663,
+      "learning_rate": 2.2084007044821764e-06,
+      "loss": 0.3352,
+      "step": 337
+    },
+    {
+      "epoch": 4.329073482428115,
+      "grad_norm": 0.23930760555111544,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.3388,
+      "step": 338
+    },
+    {
+      "epoch": 4.34185303514377,
+      "grad_norm": 0.23902114243628853,
+      "learning_rate": 2.0477254472197237e-06,
+      "loss": 0.3379,
+      "step": 339
+    },
+    {
+      "epoch": 4.354632587859425,
+      "grad_norm": 0.24839488860265907,
+      "learning_rate": 1.96954184398368e-06,
+      "loss": 0.3239,
+      "step": 340
+    },
+    {
+      "epoch": 4.36741214057508,
+      "grad_norm": 0.23307275325799628,
+      "learning_rate": 1.8928026470526917e-06,
+      "loss": 0.3671,
+      "step": 341
+    },
+    {
+      "epoch": 4.380191693290735,
+      "grad_norm": 0.2475348745152134,
+      "learning_rate": 1.817514003945524e-06,
+      "loss": 0.3573,
+      "step": 342
+    },
+    {
+      "epoch": 4.39297124600639,
+      "grad_norm": 0.2573337703222915,
+      "learning_rate": 1.743681945978184e-06,
+      "loss": 0.324,
+      "step": 343
+    },
+    {
+      "epoch": 4.405750798722044,
+      "grad_norm": 0.2509537961622258,
+      "learning_rate": 1.6713123877807413e-06,
+      "loss": 0.3235,
+      "step": 344
+    },
+    {
+      "epoch": 4.4185303514377,
+      "grad_norm": 0.23972950347988456,
+      "learning_rate": 1.6004111268235156e-06,
+      "loss": 0.324,
+      "step": 345
+    },
+    {
+      "epoch": 4.431309904153355,
+      "grad_norm": 0.24276933283329488,
+      "learning_rate": 1.5309838429526714e-06,
+      "loss": 0.3324,
+      "step": 346
+    },
+    {
+      "epoch": 4.444089456869009,
+      "grad_norm": 0.23081639636114068,
+      "learning_rate": 1.4630360979351644e-06,
+      "loss": 0.3407,
+      "step": 347
+    },
+    {
+      "epoch": 4.456869009584665,
+      "grad_norm": 0.23886385168549315,
+      "learning_rate": 1.396573335013236e-06,
+      "loss": 0.3324,
+      "step": 348
+    },
+    {
+      "epoch": 4.4696485623003195,
+      "grad_norm": 0.23076136632315059,
+      "learning_rate": 1.3316008784683265e-06,
+      "loss": 0.3538,
+      "step": 349
+    },
+    {
+      "epoch": 4.482428115015974,
+      "grad_norm": 0.24404693679132852,
+      "learning_rate": 1.2681239331945695e-06,
+      "loss": 0.3309,
+      "step": 350
+    },
+    {
+      "epoch": 4.49520766773163,
+      "grad_norm": 0.2268383924961628,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.3299,
+      "step": 351
+    },
+    {
+      "epoch": 4.507987220447284,
+      "grad_norm": 0.22465135503054393,
+      "learning_rate": 1.1456767966083393e-06,
+      "loss": 0.324,
+      "step": 352
+    },
+    {
+      "epoch": 4.520766773162939,
+      "grad_norm": 0.2220366554395612,
+      "learning_rate": 1.086716414442952e-06,
+      "loss": 0.3676,
+      "step": 353
+    },
+    {
+      "epoch": 4.533546325878595,
+      "grad_norm": 0.21940448549398342,
+      "learning_rate": 1.0292711610570904e-06,
+      "loss": 0.334,
+      "step": 354
+    },
+    {
+      "epoch": 4.546325878594249,
+      "grad_norm": 0.2144078580002602,
+      "learning_rate": 9.733456383463658e-07,
+      "loss": 0.3481,
+      "step": 355
+    },
+    {
+      "epoch": 4.559105431309904,
+      "grad_norm": 0.2275935061530386,
+      "learning_rate": 9.189443264619102e-07,
+      "loss": 0.3256,
+      "step": 356
+    },
+    {
+      "epoch": 4.571884984025559,
+      "grad_norm": 0.226985551191106,
+      "learning_rate": 8.660715834514977e-07,
+      "loss": 0.3173,
+      "step": 357
+    },
+    {
+      "epoch": 4.584664536741214,
+      "grad_norm": 0.22485580253792856,
+      "learning_rate": 8.147316449103959e-07,
+      "loss": 0.3446,
+      "step": 358
+    },
+    {
+      "epoch": 4.597444089456869,
+      "grad_norm": 0.2340075357389187,
+      "learning_rate": 7.649286236420806e-07,
+      "loss": 0.3364,
+      "step": 359
+    },
+    {
+      "epoch": 4.6102236421725244,
+      "grad_norm": 0.22345846407489506,
+      "learning_rate": 7.166665093287539e-07,
+      "loss": 0.3367,
+      "step": 360
+    },
+    {
+      "epoch": 4.623003194888179,
+      "grad_norm": 0.22370903675134662,
+      "learning_rate": 6.69949168211721e-07,
+      "loss": 0.3602,
+      "step": 361
+    },
+    {
+      "epoch": 4.635782747603834,
+      "grad_norm": 0.22486085593015095,
+      "learning_rate": 6.247803427816945e-07,
+      "loss": 0.357,
+      "step": 362
+    },
+    {
+      "epoch": 4.6485623003194885,
+      "grad_norm": 0.22083816946876406,
+      "learning_rate": 5.811636514789598e-07,
+      "loss": 0.3407,
+      "step": 363
+    },
+    {
+      "epoch": 4.661341853035144,
+      "grad_norm": 0.2289657141053928,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.3283,
+      "step": 364
+    },
+    {
+      "epoch": 4.674121405750799,
+      "grad_norm": 0.2245238228806618,
+      "learning_rate": 4.986005230351954e-07,
+      "loss": 0.3358,
+      "step": 365
+    },
+    {
+      "epoch": 4.686900958466453,
+      "grad_norm": 0.22184337753951489,
+      "learning_rate": 4.5966069996365993e-07,
+      "loss": 0.3331,
+      "step": 366
+    },
+    {
+      "epoch": 4.699680511182109,
+      "grad_norm": 0.2304636950863329,
+      "learning_rate": 4.22286238628562e-07,
+      "loss": 0.3297,
+      "step": 367
+    },
+    {
+      "epoch": 4.712460063897764,
+      "grad_norm": 0.2168016391163561,
+      "learning_rate": 3.8648013306960664e-07,
+      "loss": 0.3407,
+      "step": 368
+    },
+    {
+      "epoch": 4.725239616613418,
+      "grad_norm": 0.21992239865456173,
+      "learning_rate": 3.522452516867048e-07,
+      "loss": 0.3278,
+      "step": 369
+    },
+    {
+      "epoch": 4.738019169329074,
+      "grad_norm": 0.22103634468553346,
+      "learning_rate": 3.1958433701019697e-07,
+      "loss": 0.3136,
+      "step": 370
+    },
+    {
+      "epoch": 4.7507987220447285,
+      "grad_norm": 0.221656009377628,
+      "learning_rate": 2.8850000548115155e-07,
+      "loss": 0.342,
+      "step": 371
+    },
+    {
+      "epoch": 4.763578274760383,
+      "grad_norm": 0.2195544732802204,
+      "learning_rate": 2.5899474724174313e-07,
+      "loss": 0.3507,
+      "step": 372
+    },
+    {
+      "epoch": 4.776357827476039,
+      "grad_norm": 0.23056922821652975,
+      "learning_rate": 2.3107092593579905e-07,
+      "loss": 0.3375,
+      "step": 373
+    },
+    {
+      "epoch": 4.789137380191693,
+      "grad_norm": 0.222492788995926,
+      "learning_rate": 2.0473077851942858e-07,
+      "loss": 0.3524,
+      "step": 374
+    },
+    {
+      "epoch": 4.801916932907348,
+      "grad_norm": 0.22232266738755696,
+      "learning_rate": 1.799764150818306e-07,
+      "loss": 0.3511,
+      "step": 375
+    },
+    {
+      "epoch": 4.814696485623003,
+      "grad_norm": 0.21962084479736094,
+      "learning_rate": 1.5680981867625566e-07,
+      "loss": 0.3334,
+      "step": 376
+    },
+    {
+      "epoch": 4.827476038338658,
+      "grad_norm": 0.21579569570483248,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.3578,
+      "step": 377
+    },
+    {
+      "epoch": 4.840255591054313,
+      "grad_norm": 0.21848361329456045,
+      "learning_rate": 1.1524722305144231e-07,
+      "loss": 0.3482,
+      "step": 378
+    },
+    {
+      "epoch": 4.853035143769968,
+      "grad_norm": 0.22295101336863904,
+      "learning_rate": 9.685455338016347e-08,
+      "loss": 0.344,
+      "step": 379
+    },
+    {
+      "epoch": 4.865814696485623,
+      "grad_norm": 0.22388599501739676,
+      "learning_rate": 8.005630957010014e-08,
+      "loss": 0.3507,
+      "step": 380
+    },
+    {
+      "epoch": 4.878594249201278,
+      "grad_norm": 0.21620839076719045,
+      "learning_rate": 6.485383731580142e-08,
+      "loss": 0.3338,
+      "step": 381
+    },
+    {
+      "epoch": 4.891373801916933,
+      "grad_norm": 0.21452810970623,
+      "learning_rate": 5.1248354475768034e-08,
+      "loss": 0.3269,
+      "step": 382
+    },
+    {
+      "epoch": 4.904153354632588,
+      "grad_norm": 0.22073444367795245,
+      "learning_rate": 3.924095097489922e-08,
+      "loss": 0.3343,
+      "step": 383
+    },
+    {
+      "epoch": 4.916932907348243,
+      "grad_norm": 0.2168939465061565,
+      "learning_rate": 2.8832588717164766e-08,
+      "loss": 0.3393,
+      "step": 384
+    },
+    {
+      "epoch": 4.9297124600638975,
+      "grad_norm": 0.22692258552939853,
+      "learning_rate": 2.0024101508555604e-08,
+      "loss": 0.3312,
+      "step": 385
+    },
+    {
+      "epoch": 4.942492012779553,
+      "grad_norm": 0.21678847192685047,
+      "learning_rate": 1.281619499029274e-08,
+      "loss": 0.3482,
+      "step": 386
+    },
+    {
+      "epoch": 4.955271565495208,
+      "grad_norm": 0.22738297907016455,
+      "learning_rate": 7.209446582292501e-09,
+      "loss": 0.3349,
+      "step": 387
+    },
+    {
+      "epoch": 4.968051118210862,
+      "grad_norm": 0.222159525238591,
+      "learning_rate": 3.2043054369057523e-09,
+      "loss": 0.3281,
+      "step": 388
+    },
+    {
+      "epoch": 4.980830670926517,
+      "grad_norm": 0.22624430672708137,
+      "learning_rate": 8.010924029533406e-10,
+      "loss": 0.3238,
+      "step": 389
+    },
+    {
+      "epoch": 4.993610223642173,
+      "grad_norm": 0.23797061131952146,
+      "learning_rate": 0.0,
+      "loss": 0.3224,
+      "step": 390
+    },
+    {
+      "epoch": 4.993610223642173,
+      "step": 390,
+      "total_flos": 1.2901911719428751e+18,
+      "train_loss": 0.14476858102358303,
+      "train_runtime": 3885.8859,
+      "train_samples_per_second": 12.867,
+      "train_steps_per_second": 0.1
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 390,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.2901911719428751e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}