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: no_pipeline_10k
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # no_pipeline_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/no_pipeline_10k dataset.
 
 ## Model description
 

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.984025559105431,
+    "total_flos": 4.8627183877370675e+17,
+    "train_loss": 0.3829498979716729,
+    "train_runtime": 7298.0864,
+    "train_samples_per_second": 6.851,
+    "train_steps_per_second": 0.053
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.984025559105431,
+    "total_flos": 4.8627183877370675e+17,
+    "train_loss": 0.3829498979716729,
+    "train_runtime": 7298.0864,
+    "train_samples_per_second": 6.851,
+    "train_steps_per_second": 0.053
+}

trainer_state.json

@@ -0,0 +1,2772 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.984025559105431,
+  "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": 6.89919949032927,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 1.1213,
+      "step": 1
+    },
+    {
+      "epoch": 0.025559105431309903,
+      "grad_norm": 7.112361262613927,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 1.134,
+      "step": 2
+    },
+    {
+      "epoch": 0.038338658146964855,
+      "grad_norm": 6.95945151642214,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 1.13,
+      "step": 3
+    },
+    {
+      "epoch": 0.051118210862619806,
+      "grad_norm": 6.237212844536889,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 1.0941,
+      "step": 4
+    },
+    {
+      "epoch": 0.06389776357827476,
+      "grad_norm": 4.9408704126707,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 1.0488,
+      "step": 5
+    },
+    {
+      "epoch": 0.07667731629392971,
+      "grad_norm": 2.713688934853278,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.9524,
+      "step": 6
+    },
+    {
+      "epoch": 0.08945686900958466,
+      "grad_norm": 2.236428015821853,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 0.917,
+      "step": 7
+    },
+    {
+      "epoch": 0.10223642172523961,
+      "grad_norm": 3.8378428129535105,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 0.9332,
+      "step": 8
+    },
+    {
+      "epoch": 0.11501597444089456,
+      "grad_norm": 4.063657979438273,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.9907,
+      "step": 9
+    },
+    {
+      "epoch": 0.12779552715654952,
+      "grad_norm": 4.143920952041562,
+      "learning_rate": 1.0256410256410256e-05,
+      "loss": 0.8763,
+      "step": 10
+    },
+    {
+      "epoch": 0.14057507987220447,
+      "grad_norm": 3.72170509093155,
+      "learning_rate": 1.1282051282051283e-05,
+      "loss": 0.8948,
+      "step": 11
+    },
+    {
+      "epoch": 0.15335463258785942,
+      "grad_norm": 2.770591627238587,
+      "learning_rate": 1.230769230769231e-05,
+      "loss": 0.8711,
+      "step": 12
+    },
+    {
+      "epoch": 0.16613418530351437,
+      "grad_norm": 2.4675123902750484,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.8195,
+      "step": 13
+    },
+    {
+      "epoch": 0.17891373801916932,
+      "grad_norm": 2.7527888817288444,
+      "learning_rate": 1.435897435897436e-05,
+      "loss": 0.8337,
+      "step": 14
+    },
+    {
+      "epoch": 0.19169329073482427,
+      "grad_norm": 2.2787439636344735,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 0.8374,
+      "step": 15
+    },
+    {
+      "epoch": 0.20447284345047922,
+      "grad_norm": 1.8909598331839135,
+      "learning_rate": 1.641025641025641e-05,
+      "loss": 0.7903,
+      "step": 16
+    },
+    {
+      "epoch": 0.21725239616613418,
+      "grad_norm": 1.657610544673612,
+      "learning_rate": 1.7435897435897438e-05,
+      "loss": 0.8262,
+      "step": 17
+    },
+    {
+      "epoch": 0.23003194888178913,
+      "grad_norm": 1.6415560009359647,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 0.7333,
+      "step": 18
+    },
+    {
+      "epoch": 0.24281150159744408,
+      "grad_norm": 1.4039315397426195,
+      "learning_rate": 1.9487179487179488e-05,
+      "loss": 0.7788,
+      "step": 19
+    },
+    {
+      "epoch": 0.25559105431309903,
+      "grad_norm": 1.2891286220119198,
+      "learning_rate": 2.0512820512820512e-05,
+      "loss": 0.7508,
+      "step": 20
+    },
+    {
+      "epoch": 0.268370607028754,
+      "grad_norm": 0.975724200928363,
+      "learning_rate": 2.153846153846154e-05,
+      "loss": 0.7405,
+      "step": 21
+    },
+    {
+      "epoch": 0.28115015974440893,
+      "grad_norm": 1.148383190119256,
+      "learning_rate": 2.2564102564102566e-05,
+      "loss": 0.7858,
+      "step": 22
+    },
+    {
+      "epoch": 0.2939297124600639,
+      "grad_norm": 0.9435222656914478,
+      "learning_rate": 2.3589743589743593e-05,
+      "loss": 0.7713,
+      "step": 23
+    },
+    {
+      "epoch": 0.30670926517571884,
+      "grad_norm": 0.8129609831560625,
+      "learning_rate": 2.461538461538462e-05,
+      "loss": 0.7211,
+      "step": 24
+    },
+    {
+      "epoch": 0.3194888178913738,
+      "grad_norm": 0.9523055151882185,
+      "learning_rate": 2.5641025641025646e-05,
+      "loss": 0.768,
+      "step": 25
+    },
+    {
+      "epoch": 0.33226837060702874,
+      "grad_norm": 0.7446867982723262,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.7338,
+      "step": 26
+    },
+    {
+      "epoch": 0.3450479233226837,
+      "grad_norm": 0.825665662190159,
+      "learning_rate": 2.7692307692307694e-05,
+      "loss": 0.7016,
+      "step": 27
+    },
+    {
+      "epoch": 0.35782747603833864,
+      "grad_norm": 0.9302071194411955,
+      "learning_rate": 2.871794871794872e-05,
+      "loss": 0.7046,
+      "step": 28
+    },
+    {
+      "epoch": 0.3706070287539936,
+      "grad_norm": 2.2820678794917657,
+      "learning_rate": 2.9743589743589747e-05,
+      "loss": 0.7409,
+      "step": 29
+    },
+    {
+      "epoch": 0.38338658146964855,
+      "grad_norm": 3.797748981047407,
+      "learning_rate": 3.0769230769230774e-05,
+      "loss": 0.7237,
+      "step": 30
+    },
+    {
+      "epoch": 0.3961661341853035,
+      "grad_norm": 1.8038524783041952,
+      "learning_rate": 3.1794871794871795e-05,
+      "loss": 0.7242,
+      "step": 31
+    },
+    {
+      "epoch": 0.40894568690095845,
+      "grad_norm": 0.8920441428713441,
+      "learning_rate": 3.282051282051282e-05,
+      "loss": 0.7189,
+      "step": 32
+    },
+    {
+      "epoch": 0.4217252396166134,
+      "grad_norm": 1.1984598828972841,
+      "learning_rate": 3.384615384615385e-05,
+      "loss": 0.6688,
+      "step": 33
+    },
+    {
+      "epoch": 0.43450479233226835,
+      "grad_norm": 1.0209161984060695,
+      "learning_rate": 3.4871794871794875e-05,
+      "loss": 0.7375,
+      "step": 34
+    },
+    {
+      "epoch": 0.4472843450479233,
+      "grad_norm": 1.0099951602767552,
+      "learning_rate": 3.58974358974359e-05,
+      "loss": 0.7443,
+      "step": 35
+    },
+    {
+      "epoch": 0.46006389776357826,
+      "grad_norm": 1.283116041948978,
+      "learning_rate": 3.692307692307693e-05,
+      "loss": 0.7327,
+      "step": 36
+    },
+    {
+      "epoch": 0.4728434504792332,
+      "grad_norm": 1.0301415862942631,
+      "learning_rate": 3.794871794871795e-05,
+      "loss": 0.717,
+      "step": 37
+    },
+    {
+      "epoch": 0.48562300319488816,
+      "grad_norm": 1.1146486639658266,
+      "learning_rate": 3.8974358974358976e-05,
+      "loss": 0.6944,
+      "step": 38
+    },
+    {
+      "epoch": 0.4984025559105431,
+      "grad_norm": 1.0128108377293754,
+      "learning_rate": 4e-05,
+      "loss": 0.7231,
+      "step": 39
+    },
+    {
+      "epoch": 0.5111821086261981,
+      "grad_norm": 0.9841003348246676,
+      "learning_rate": 3.9999198907597046e-05,
+      "loss": 0.6969,
+      "step": 40
+    },
+    {
+      "epoch": 0.5239616613418531,
+      "grad_norm": 1.0436172433676487,
+      "learning_rate": 3.9996795694563096e-05,
+      "loss": 0.6827,
+      "step": 41
+    },
+    {
+      "epoch": 0.536741214057508,
+      "grad_norm": 0.7410593224584431,
+      "learning_rate": 3.999279055341771e-05,
+      "loss": 0.7134,
+      "step": 42
+    },
+    {
+      "epoch": 0.549520766773163,
+      "grad_norm": 0.9370742862883669,
+      "learning_rate": 3.998718380500971e-05,
+      "loss": 0.6764,
+      "step": 43
+    },
+    {
+      "epoch": 0.5623003194888179,
+      "grad_norm": 0.8162957683591703,
+      "learning_rate": 3.997997589849145e-05,
+      "loss": 0.7141,
+      "step": 44
+    },
+    {
+      "epoch": 0.5750798722044729,
+      "grad_norm": 0.9370802243597423,
+      "learning_rate": 3.9971167411282835e-05,
+      "loss": 0.7471,
+      "step": 45
+    },
+    {
+      "epoch": 0.5878594249201278,
+      "grad_norm": 1.0913996850120742,
+      "learning_rate": 3.99607590490251e-05,
+      "loss": 0.7083,
+      "step": 46
+    },
+    {
+      "epoch": 0.6006389776357828,
+      "grad_norm": 0.7041500064159922,
+      "learning_rate": 3.9948751645524235e-05,
+      "loss": 0.7299,
+      "step": 47
+    },
+    {
+      "epoch": 0.6134185303514377,
+      "grad_norm": 1.0405539404112774,
+      "learning_rate": 3.9935146162684206e-05,
+      "loss": 0.6966,
+      "step": 48
+    },
+    {
+      "epoch": 0.6261980830670927,
+      "grad_norm": 0.6860353533522638,
+      "learning_rate": 3.9919943690429906e-05,
+      "loss": 0.7067,
+      "step": 49
+    },
+    {
+      "epoch": 0.6389776357827476,
+      "grad_norm": 0.9183145594548415,
+      "learning_rate": 3.9903145446619837e-05,
+      "loss": 0.7182,
+      "step": 50
+    },
+    {
+      "epoch": 0.6517571884984026,
+      "grad_norm": 0.8951375890184433,
+      "learning_rate": 3.9884752776948564e-05,
+      "loss": 0.6904,
+      "step": 51
+    },
+    {
+      "epoch": 0.6645367412140575,
+      "grad_norm": 0.9222228048252842,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.6986,
+      "step": 52
+    },
+    {
+      "epoch": 0.6773162939297125,
+      "grad_norm": 0.8584052162906096,
+      "learning_rate": 3.9843190181323744e-05,
+      "loss": 0.6819,
+      "step": 53
+    },
+    {
+      "epoch": 0.6900958466453674,
+      "grad_norm": 0.9372514098490873,
+      "learning_rate": 3.982002358491817e-05,
+      "loss": 0.7004,
+      "step": 54
+    },
+    {
+      "epoch": 0.7028753993610224,
+      "grad_norm": 0.8769696318933147,
+      "learning_rate": 3.979526922148058e-05,
+      "loss": 0.6889,
+      "step": 55
+    },
+    {
+      "epoch": 0.7156549520766773,
+      "grad_norm": 1.019831098979255,
+      "learning_rate": 3.9768929074064206e-05,
+      "loss": 0.6942,
+      "step": 56
+    },
+    {
+      "epoch": 0.7284345047923323,
+      "grad_norm": 0.7902771544209148,
+      "learning_rate": 3.9741005252758255e-05,
+      "loss": 0.6973,
+      "step": 57
+    },
+    {
+      "epoch": 0.7412140575079872,
+      "grad_norm": 0.7841232010952441,
+      "learning_rate": 3.971149999451886e-05,
+      "loss": 0.6962,
+      "step": 58
+    },
+    {
+      "epoch": 0.7539936102236422,
+      "grad_norm": 0.7570268765749476,
+      "learning_rate": 3.9680415662989806e-05,
+      "loss": 0.7047,
+      "step": 59
+    },
+    {
+      "epoch": 0.7667731629392971,
+      "grad_norm": 0.8473296414632037,
+      "learning_rate": 3.9647754748313294e-05,
+      "loss": 0.7143,
+      "step": 60
+    },
+    {
+      "epoch": 0.7795527156549521,
+      "grad_norm": 0.6866268169972648,
+      "learning_rate": 3.96135198669304e-05,
+      "loss": 0.6959,
+      "step": 61
+    },
+    {
+      "epoch": 0.792332268370607,
+      "grad_norm": 0.7119010577983071,
+      "learning_rate": 3.957771376137144e-05,
+      "loss": 0.6725,
+      "step": 62
+    },
+    {
+      "epoch": 0.805111821086262,
+      "grad_norm": 0.6789067799830576,
+      "learning_rate": 3.954033930003634e-05,
+      "loss": 0.6981,
+      "step": 63
+    },
+    {
+      "epoch": 0.8178913738019169,
+      "grad_norm": 0.6845822826739288,
+      "learning_rate": 3.9501399476964806e-05,
+      "loss": 0.6922,
+      "step": 64
+    },
+    {
+      "epoch": 0.8306709265175719,
+      "grad_norm": 0.7345864697984222,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.6954,
+      "step": 65
+    },
+    {
+      "epoch": 0.8434504792332268,
+      "grad_norm": 0.7756273926598104,
+      "learning_rate": 3.9418836348521045e-05,
+      "loss": 0.6754,
+      "step": 66
+    },
+    {
+      "epoch": 0.8562300319488818,
+      "grad_norm": 0.7515145798465848,
+      "learning_rate": 3.937521965721831e-05,
+      "loss": 0.6948,
+      "step": 67
+    },
+    {
+      "epoch": 0.8690095846645367,
+      "grad_norm": 0.9922587833641502,
+      "learning_rate": 3.933005083178828e-05,
+      "loss": 0.7235,
+      "step": 68
+    },
+    {
+      "epoch": 0.8817891373801917,
+      "grad_norm": 0.8759668934087257,
+      "learning_rate": 3.928333349067125e-05,
+      "loss": 0.6575,
+      "step": 69
+    },
+    {
+      "epoch": 0.8945686900958466,
+      "grad_norm": 2.72537223637405,
+      "learning_rate": 3.923507137635792e-05,
+      "loss": 0.653,
+      "step": 70
+    },
+    {
+      "epoch": 0.9073482428115016,
+      "grad_norm": 1.1067180024716368,
+      "learning_rate": 3.9185268355089606e-05,
+      "loss": 0.7038,
+      "step": 71
+    },
+    {
+      "epoch": 0.9201277955271565,
+      "grad_norm": 0.6693766946587052,
+      "learning_rate": 3.913392841654851e-05,
+      "loss": 0.6844,
+      "step": 72
+    },
+    {
+      "epoch": 0.9329073482428115,
+      "grad_norm": 0.8507328914824696,
+      "learning_rate": 3.9081055673538093e-05,
+      "loss": 0.6849,
+      "step": 73
+    },
+    {
+      "epoch": 0.9456869009584664,
+      "grad_norm": 0.7361291800204529,
+      "learning_rate": 3.902665436165364e-05,
+      "loss": 0.6673,
+      "step": 74
+    },
+    {
+      "epoch": 0.9584664536741214,
+      "grad_norm": 0.753388426257767,
+      "learning_rate": 3.897072883894291e-05,
+      "loss": 0.7367,
+      "step": 75
+    },
+    {
+      "epoch": 0.9712460063897763,
+      "grad_norm": 0.7881475535248718,
+      "learning_rate": 3.8913283585557054e-05,
+      "loss": 0.7094,
+      "step": 76
+    },
+    {
+      "epoch": 0.9840255591054313,
+      "grad_norm": 0.6439451381651863,
+      "learning_rate": 3.885432320339167e-05,
+      "loss": 0.6698,
+      "step": 77
+    },
+    {
+      "epoch": 0.9968051118210862,
+      "grad_norm": 0.8045843499463019,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.691,
+      "step": 78
+    },
+    {
+      "epoch": 1.0095846645367412,
+      "grad_norm": 0.7558149368747471,
+      "learning_rate": 3.873187606680543e-05,
+      "loss": 0.5566,
+      "step": 79
+    },
+    {
+      "epoch": 1.0223642172523961,
+      "grad_norm": 0.7490512478402147,
+      "learning_rate": 3.866839912153168e-05,
+      "loss": 0.5138,
+      "step": 80
+    },
+    {
+      "epoch": 1.035143769968051,
+      "grad_norm": 0.7550409847525211,
+      "learning_rate": 3.860342666498677e-05,
+      "loss": 0.4905,
+      "step": 81
+    },
+    {
+      "epoch": 1.0479233226837061,
+      "grad_norm": 0.8804964484828116,
+      "learning_rate": 3.853696390206484e-05,
+      "loss": 0.4799,
+      "step": 82
+    },
+    {
+      "epoch": 1.060702875399361,
+      "grad_norm": 0.8041590797927656,
+      "learning_rate": 3.846901615704734e-05,
+      "loss": 0.4996,
+      "step": 83
+    },
+    {
+      "epoch": 1.073482428115016,
+      "grad_norm": 0.8927936842593411,
+      "learning_rate": 3.839958887317649e-05,
+      "loss": 0.5083,
+      "step": 84
+    },
+    {
+      "epoch": 1.0862619808306708,
+      "grad_norm": 0.9319116509219216,
+      "learning_rate": 3.832868761221926e-05,
+      "loss": 0.543,
+      "step": 85
+    },
+    {
+      "epoch": 1.099041533546326,
+      "grad_norm": 0.670006971832743,
+      "learning_rate": 3.825631805402182e-05,
+      "loss": 0.448,
+      "step": 86
+    },
+    {
+      "epoch": 1.1118210862619808,
+      "grad_norm": 0.9064315152691065,
+      "learning_rate": 3.818248599605448e-05,
+      "loss": 0.5171,
+      "step": 87
+    },
+    {
+      "epoch": 1.1246006389776357,
+      "grad_norm": 0.758642809473103,
+      "learning_rate": 3.810719735294731e-05,
+      "loss": 0.5038,
+      "step": 88
+    },
+    {
+      "epoch": 1.1373801916932909,
+      "grad_norm": 0.764907407711914,
+      "learning_rate": 3.8030458156016326e-05,
+      "loss": 0.5037,
+      "step": 89
+    },
+    {
+      "epoch": 1.1501597444089458,
+      "grad_norm": 0.7627860020820714,
+      "learning_rate": 3.795227455278029e-05,
+      "loss": 0.4985,
+      "step": 90
+    },
+    {
+      "epoch": 1.1629392971246006,
+      "grad_norm": 0.7017412920263736,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.4911,
+      "step": 91
+    },
+    {
+      "epoch": 1.1757188498402555,
+      "grad_norm": 0.7318639654061692,
+      "learning_rate": 3.7791599295517825e-05,
+      "loss": 0.4683,
+      "step": 92
+    },
+    {
+      "epoch": 1.1884984025559104,
+      "grad_norm": 0.7419770013134663,
+      "learning_rate": 3.7709120513064196e-05,
+      "loss": 0.4905,
+      "step": 93
+    },
+    {
+      "epoch": 1.2012779552715656,
+      "grad_norm": 0.7029633343880348,
+      "learning_rate": 3.762522306641998e-05,
+      "loss": 0.4608,
+      "step": 94
+    },
+    {
+      "epoch": 1.2140575079872205,
+      "grad_norm": 0.6637712265613218,
+      "learning_rate": 3.7539913676545874e-05,
+      "loss": 0.4975,
+      "step": 95
+    },
+    {
+      "epoch": 1.2268370607028753,
+      "grad_norm": 0.7420195475320207,
+      "learning_rate": 3.745319917751229e-05,
+      "loss": 0.5011,
+      "step": 96
+    },
+    {
+      "epoch": 1.2396166134185305,
+      "grad_norm": 0.7277392676480745,
+      "learning_rate": 3.736508651595188e-05,
+      "loss": 0.4965,
+      "step": 97
+    },
+    {
+      "epoch": 1.2523961661341854,
+      "grad_norm": 0.7162202220379209,
+      "learning_rate": 3.727558275050301e-05,
+      "loss": 0.5123,
+      "step": 98
+    },
+    {
+      "epoch": 1.2651757188498403,
+      "grad_norm": 0.6995610700257406,
+      "learning_rate": 3.718469505124434e-05,
+      "loss": 0.4782,
+      "step": 99
+    },
+    {
+      "epoch": 1.2779552715654952,
+      "grad_norm": 0.6647278023915858,
+      "learning_rate": 3.709243069912041e-05,
+      "loss": 0.4865,
+      "step": 100
+    },
+    {
+      "epoch": 1.29073482428115,
+      "grad_norm": 0.7167704751603956,
+      "learning_rate": 3.699879708535838e-05,
+      "loss": 0.4856,
+      "step": 101
+    },
+    {
+      "epoch": 1.3035143769968052,
+      "grad_norm": 0.545823163550541,
+      "learning_rate": 3.69038017108759e-05,
+      "loss": 0.4916,
+      "step": 102
+    },
+    {
+      "epoch": 1.31629392971246,
+      "grad_norm": 0.7160209336189851,
+      "learning_rate": 3.680745218568026e-05,
+      "loss": 0.5093,
+      "step": 103
+    },
+    {
+      "epoch": 1.329073482428115,
+      "grad_norm": 0.5793516549733936,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.4885,
+      "step": 104
+    },
+    {
+      "epoch": 1.34185303514377,
+      "grad_norm": 0.7336936481288561,
+      "learning_rate": 3.6610721664960236e-05,
+      "loss": 0.5164,
+      "step": 105
+    },
+    {
+      "epoch": 1.354632587859425,
+      "grad_norm": 0.6579506936715739,
+      "learning_rate": 3.65103564293684e-05,
+      "loss": 0.4899,
+      "step": 106
+    },
+    {
+      "epoch": 1.3674121405750799,
+      "grad_norm": 0.6614801100344516,
+      "learning_rate": 3.640866856166601e-05,
+      "loss": 0.4868,
+      "step": 107
+    },
+    {
+      "epoch": 1.3801916932907348,
+      "grad_norm": 0.6961251504569803,
+      "learning_rate": 3.6305666207990886e-05,
+      "loss": 0.5223,
+      "step": 108
+    },
+    {
+      "epoch": 1.3929712460063897,
+      "grad_norm": 0.5709110993698499,
+      "learning_rate": 3.6201357619783336e-05,
+      "loss": 0.4938,
+      "step": 109
+    },
+    {
+      "epoch": 1.4057507987220448,
+      "grad_norm": 0.6166361881217528,
+      "learning_rate": 3.609575115312511e-05,
+      "loss": 0.494,
+      "step": 110
+    },
+    {
+      "epoch": 1.4185303514376997,
+      "grad_norm": 0.6325395561407832,
+      "learning_rate": 3.598885526807003e-05,
+      "loss": 0.503,
+      "step": 111
+    },
+    {
+      "epoch": 1.4313099041533546,
+      "grad_norm": 0.5568640986348814,
+      "learning_rate": 3.5880678527966224e-05,
+      "loss": 0.4722,
+      "step": 112
+    },
+    {
+      "epoch": 1.4440894568690097,
+      "grad_norm": 0.595346327028961,
+      "learning_rate": 3.577122959877017e-05,
+      "loss": 0.4868,
+      "step": 113
+    },
+    {
+      "epoch": 1.4568690095846646,
+      "grad_norm": 0.5776621695803134,
+      "learning_rate": 3.566051724835245e-05,
+      "loss": 0.4967,
+      "step": 114
+    },
+    {
+      "epoch": 1.4696485623003195,
+      "grad_norm": 0.5921383544308383,
+      "learning_rate": 3.554855034579532e-05,
+      "loss": 0.488,
+      "step": 115
+    },
+    {
+      "epoch": 1.4824281150159744,
+      "grad_norm": 0.6100675505905345,
+      "learning_rate": 3.5435337860682304e-05,
+      "loss": 0.4966,
+      "step": 116
+    },
+    {
+      "epoch": 1.4952076677316293,
+      "grad_norm": 0.5711838553131381,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.5069,
+      "step": 117
+    },
+    {
+      "epoch": 1.5079872204472844,
+      "grad_norm": 0.5584013011485947,
+      "learning_rate": 3.520521251930941e-05,
+      "loss": 0.5035,
+      "step": 118
+    },
+    {
+      "epoch": 1.5207667731629393,
+      "grad_norm": 0.5432605967081409,
+      "learning_rate": 3.5088318098215805e-05,
+      "loss": 0.4946,
+      "step": 119
+    },
+    {
+      "epoch": 1.5335463258785942,
+      "grad_norm": 0.6623357724108133,
+      "learning_rate": 3.497021496342203e-05,
+      "loss": 0.4656,
+      "step": 120
+    },
+    {
+      "epoch": 1.5463258785942493,
+      "grad_norm": 0.7150767034699087,
+      "learning_rate": 3.485091257608047e-05,
+      "loss": 0.5106,
+      "step": 121
+    },
+    {
+      "epoch": 1.5591054313099042,
+      "grad_norm": 0.5327141830212154,
+      "learning_rate": 3.473042049341474e-05,
+      "loss": 0.4631,
+      "step": 122
+    },
+    {
+      "epoch": 1.571884984025559,
+      "grad_norm": 0.6586966611579816,
+      "learning_rate": 3.4608748367954064e-05,
+      "loss": 0.5059,
+      "step": 123
+    },
+    {
+      "epoch": 1.5846645367412142,
+      "grad_norm": 0.5751625836627946,
+      "learning_rate": 3.4485905946759965e-05,
+      "loss": 0.4678,
+      "step": 124
+    },
+    {
+      "epoch": 1.5974440894568689,
+      "grad_norm": 0.6117292557498856,
+      "learning_rate": 3.4361903070645484e-05,
+      "loss": 0.4854,
+      "step": 125
+    },
+    {
+      "epoch": 1.610223642172524,
+      "grad_norm": 0.6775775117936489,
+      "learning_rate": 3.423674967338681e-05,
+      "loss": 0.4832,
+      "step": 126
+    },
+    {
+      "epoch": 1.623003194888179,
+      "grad_norm": 0.5511411522344801,
+      "learning_rate": 3.411045578092754e-05,
+      "loss": 0.4673,
+      "step": 127
+    },
+    {
+      "epoch": 1.6357827476038338,
+      "grad_norm": 0.6606420585843323,
+      "learning_rate": 3.398303151057543e-05,
+      "loss": 0.503,
+      "step": 128
+    },
+    {
+      "epoch": 1.648562300319489,
+      "grad_norm": 0.5373855468282581,
+      "learning_rate": 3.385448707019199e-05,
+      "loss": 0.5026,
+      "step": 129
+    },
+    {
+      "epoch": 1.6613418530351438,
+      "grad_norm": 0.5417547145036586,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.5145,
+      "step": 130
+    },
+    {
+      "epoch": 1.6741214057507987,
+      "grad_norm": 0.5806081785102679,
+      "learning_rate": 3.359407895863199e-05,
+      "loss": 0.4778,
+      "step": 131
+    },
+    {
+      "epoch": 1.6869009584664538,
+      "grad_norm": 0.5871900686328172,
+      "learning_rate": 3.34622361485514e-05,
+      "loss": 0.5124,
+      "step": 132
+    },
+    {
+      "epoch": 1.6996805111821085,
+      "grad_norm": 0.5785566556272356,
+      "learning_rate": 3.332931488896029e-05,
+      "loss": 0.5153,
+      "step": 133
+    },
+    {
+      "epoch": 1.7124600638977636,
+      "grad_norm": 0.6250812725124909,
+      "learning_rate": 3.319532582807977e-05,
+      "loss": 0.494,
+      "step": 134
+    },
+    {
+      "epoch": 1.7252396166134185,
+      "grad_norm": 0.5729582627902413,
+      "learning_rate": 3.30602796996717e-05,
+      "loss": 0.5149,
+      "step": 135
+    },
+    {
+      "epoch": 1.7380191693290734,
+      "grad_norm": 0.6346433097401114,
+      "learning_rate": 3.2924187322178865e-05,
+      "loss": 0.4963,
+      "step": 136
+    },
+    {
+      "epoch": 1.7507987220447285,
+      "grad_norm": 0.688285438797911,
+      "learning_rate": 3.278705959785821e-05,
+      "loss": 0.5089,
+      "step": 137
+    },
+    {
+      "epoch": 1.7635782747603834,
+      "grad_norm": 0.5815336055495459,
+      "learning_rate": 3.2648907511907544e-05,
+      "loss": 0.4949,
+      "step": 138
+    },
+    {
+      "epoch": 1.7763578274760383,
+      "grad_norm": 0.5867302206761721,
+      "learning_rate": 3.250974213158555e-05,
+      "loss": 0.4929,
+      "step": 139
+    },
+    {
+      "epoch": 1.7891373801916934,
+      "grad_norm": 0.5669606233314102,
+      "learning_rate": 3.23695746053251e-05,
+      "loss": 0.5209,
+      "step": 140
+    },
+    {
+      "epoch": 1.8019169329073481,
+      "grad_norm": 0.6247727690264819,
+      "learning_rate": 3.222841616184025e-05,
+      "loss": 0.5244,
+      "step": 141
+    },
+    {
+      "epoch": 1.8146964856230032,
+      "grad_norm": 0.500931589877546,
+      "learning_rate": 3.208627810922665e-05,
+      "loss": 0.4874,
+      "step": 142
+    },
+    {
+      "epoch": 1.8274760383386581,
+      "grad_norm": 0.5652575596502337,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.4547,
+      "step": 143
+    },
+    {
+      "epoch": 1.840255591054313,
+      "grad_norm": 0.5809131471995411,
+      "learning_rate": 3.1799108800462466e-05,
+      "loss": 0.4782,
+      "step": 144
+    },
+    {
+      "epoch": 1.8530351437699681,
+      "grad_norm": 0.6300123083087871,
+      "learning_rate": 3.1654100549227024e-05,
+      "loss": 0.4965,
+      "step": 145
+    },
+    {
+      "epoch": 1.865814696485623,
+      "grad_norm": 0.609470650733157,
+      "learning_rate": 3.1508158696850275e-05,
+      "loss": 0.4733,
+      "step": 146
+    },
+    {
+      "epoch": 1.878594249201278,
+      "grad_norm": 0.610271135194894,
+      "learning_rate": 3.136129493462312e-05,
+      "loss": 0.5021,
+      "step": 147
+    },
+    {
+      "epoch": 1.891373801916933,
+      "grad_norm": 0.5636011850098515,
+      "learning_rate": 3.121352102768998e-05,
+      "loss": 0.4812,
+      "step": 148
+    },
+    {
+      "epoch": 1.9041533546325877,
+      "grad_norm": 0.5899920985011103,
+      "learning_rate": 3.106484881410628e-05,
+      "loss": 0.5059,
+      "step": 149
+    },
+    {
+      "epoch": 1.9169329073482428,
+      "grad_norm": 0.5672932820176009,
+      "learning_rate": 3.091529020389009e-05,
+      "loss": 0.5208,
+      "step": 150
+    },
+    {
+      "epoch": 1.9297124600638977,
+      "grad_norm": 0.6018476018283119,
+      "learning_rate": 3.076485717806808e-05,
+      "loss": 0.4998,
+      "step": 151
+    },
+    {
+      "epoch": 1.9424920127795526,
+      "grad_norm": 0.6137563331005131,
+      "learning_rate": 3.061356178771564e-05,
+      "loss": 0.5024,
+      "step": 152
+    },
+    {
+      "epoch": 1.9552715654952078,
+      "grad_norm": 0.5993609686887109,
+      "learning_rate": 3.0461416152991555e-05,
+      "loss": 0.4846,
+      "step": 153
+    },
+    {
+      "epoch": 1.9680511182108626,
+      "grad_norm": 0.6105196573112612,
+      "learning_rate": 3.0308432462167045e-05,
+      "loss": 0.5265,
+      "step": 154
+    },
+    {
+      "epoch": 1.9808306709265175,
+      "grad_norm": 0.6312136376908368,
+      "learning_rate": 3.015462297064936e-05,
+      "loss": 0.492,
+      "step": 155
+    },
+    {
+      "epoch": 1.9936102236421727,
+      "grad_norm": 0.6093802641437902,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.4764,
+      "step": 156
+    },
+    {
+      "epoch": 2.0063897763578273,
+      "grad_norm": 0.5497290620839131,
+      "learning_rate": 2.98445759369477e-05,
+      "loss": 0.4175,
+      "step": 157
+    },
+    {
+      "epoch": 2.0191693290734825,
+      "grad_norm": 0.5194954967165948,
+      "learning_rate": 2.9688363232396056e-05,
+      "loss": 0.3484,
+      "step": 158
+    },
+    {
+      "epoch": 2.0319488817891376,
+      "grad_norm": 0.8303878718435693,
+      "learning_rate": 2.9531374400426158e-05,
+      "loss": 0.3083,
+      "step": 159
+    },
+    {
+      "epoch": 2.0447284345047922,
+      "grad_norm": 0.527659885783403,
+      "learning_rate": 2.9373622017294075e-05,
+      "loss": 0.3055,
+      "step": 160
+    },
+    {
+      "epoch": 2.0575079872204474,
+      "grad_norm": 0.6871640506299266,
+      "learning_rate": 2.9215118720423375e-05,
+      "loss": 0.3142,
+      "step": 161
+    },
+    {
+      "epoch": 2.070287539936102,
+      "grad_norm": 0.5862877959565796,
+      "learning_rate": 2.9055877207392752e-05,
+      "loss": 0.314,
+      "step": 162
+    },
+    {
+      "epoch": 2.083067092651757,
+      "grad_norm": 0.5360908239769426,
+      "learning_rate": 2.8895910234918828e-05,
+      "loss": 0.2916,
+      "step": 163
+    },
+    {
+      "epoch": 2.0958466453674123,
+      "grad_norm": 0.5981152051226198,
+      "learning_rate": 2.873523061783426e-05,
+      "loss": 0.3314,
+      "step": 164
+    },
+    {
+      "epoch": 2.108626198083067,
+      "grad_norm": 0.5177116035040744,
+      "learning_rate": 2.8573851228061084e-05,
+      "loss": 0.3283,
+      "step": 165
+    },
+    {
+      "epoch": 2.121405750798722,
+      "grad_norm": 0.5302662559149075,
+      "learning_rate": 2.8411784993579633e-05,
+      "loss": 0.321,
+      "step": 166
+    },
+    {
+      "epoch": 2.134185303514377,
+      "grad_norm": 0.5247610465655934,
+      "learning_rate": 2.8249044897392814e-05,
+      "loss": 0.3271,
+      "step": 167
+    },
+    {
+      "epoch": 2.146964856230032,
+      "grad_norm": 0.5182183307540772,
+      "learning_rate": 2.80856439764861e-05,
+      "loss": 0.298,
+      "step": 168
+    },
+    {
+      "epoch": 2.159744408945687,
+      "grad_norm": 0.5730957852554377,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.3201,
+      "step": 169
+    },
+    {
+      "epoch": 2.1725239616613417,
+      "grad_norm": 0.5013737772756125,
+      "learning_rate": 2.77569120720971e-05,
+      "loss": 0.3172,
+      "step": 170
+    },
+    {
+      "epoch": 2.1853035143769968,
+      "grad_norm": 0.48416794574263505,
+      "learning_rate": 2.7591607423077932e-05,
+      "loss": 0.3136,
+      "step": 171
+    },
+    {
+      "epoch": 2.198083067092652,
+      "grad_norm": 0.5286301172716432,
+      "learning_rate": 2.7425694616155474e-05,
+      "loss": 0.3012,
+      "step": 172
+    },
+    {
+      "epoch": 2.2108626198083066,
+      "grad_norm": 0.4759081452581543,
+      "learning_rate": 2.7259186942478656e-05,
+      "loss": 0.3091,
+      "step": 173
+    },
+    {
+      "epoch": 2.2236421725239617,
+      "grad_norm": 0.5140042094885082,
+      "learning_rate": 2.7092097740850712e-05,
+      "loss": 0.3281,
+      "step": 174
+    },
+    {
+      "epoch": 2.236421725239617,
+      "grad_norm": 0.4949027908863432,
+      "learning_rate": 2.692444039666066e-05,
+      "loss": 0.3008,
+      "step": 175
+    },
+    {
+      "epoch": 2.2492012779552715,
+      "grad_norm": 0.47051300241587324,
+      "learning_rate": 2.6756228340810946e-05,
+      "loss": 0.3168,
+      "step": 176
+    },
+    {
+      "epoch": 2.2619808306709266,
+      "grad_norm": 0.5419794202695959,
+      "learning_rate": 2.6587475048641596e-05,
+      "loss": 0.3305,
+      "step": 177
+    },
+    {
+      "epoch": 2.2747603833865817,
+      "grad_norm": 0.4940772116685351,
+      "learning_rate": 2.6418194038850634e-05,
+      "loss": 0.3047,
+      "step": 178
+    },
+    {
+      "epoch": 2.2875399361022364,
+      "grad_norm": 0.5374467110462928,
+      "learning_rate": 2.624839887241115e-05,
+      "loss": 0.3002,
+      "step": 179
+    },
+    {
+      "epoch": 2.3003194888178915,
+      "grad_norm": 0.47134978160686386,
+      "learning_rate": 2.607810315148494e-05,
+      "loss": 0.2963,
+      "step": 180
+    },
+    {
+      "epoch": 2.313099041533546,
+      "grad_norm": 0.5307945108018245,
+      "learning_rate": 2.5907320518332827e-05,
+      "loss": 0.3274,
+      "step": 181
+    },
+    {
+      "epoch": 2.3258785942492013,
+      "grad_norm": 0.45100791135081314,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.3318,
+      "step": 182
+    },
+    {
+      "epoch": 2.3386581469648564,
+      "grad_norm": 0.4866481702987381,
+      "learning_rate": 2.5564349278329056e-05,
+      "loss": 0.3145,
+      "step": 183
+    },
+    {
+      "epoch": 2.351437699680511,
+      "grad_norm": 0.4506406384787984,
+      "learning_rate": 2.539218814664288e-05,
+      "loss": 0.3226,
+      "step": 184
+    },
+    {
+      "epoch": 2.364217252396166,
+      "grad_norm": 0.4616113795442529,
+      "learning_rate": 2.521959505086075e-05,
+      "loss": 0.3079,
+      "step": 185
+    },
+    {
+      "epoch": 2.376996805111821,
+      "grad_norm": 0.4827300656713729,
+      "learning_rate": 2.5046583817284437e-05,
+      "loss": 0.3263,
+      "step": 186
+    },
+    {
+      "epoch": 2.389776357827476,
+      "grad_norm": 0.48666448945809476,
+      "learning_rate": 2.487316830571244e-05,
+      "loss": 0.2992,
+      "step": 187
+    },
+    {
+      "epoch": 2.402555910543131,
+      "grad_norm": 0.4793217861651193,
+      "learning_rate": 2.4699362408329646e-05,
+      "loss": 0.3085,
+      "step": 188
+    },
+    {
+      "epoch": 2.415335463258786,
+      "grad_norm": 0.4771997294308319,
+      "learning_rate": 2.4525180048594452e-05,
+      "loss": 0.3089,
+      "step": 189
+    },
+    {
+      "epoch": 2.428115015974441,
+      "grad_norm": 0.5013482619776147,
+      "learning_rate": 2.435063518012335e-05,
+      "loss": 0.3102,
+      "step": 190
+    },
+    {
+      "epoch": 2.440894568690096,
+      "grad_norm": 0.47104565066334697,
+      "learning_rate": 2.4175741785573177e-05,
+      "loss": 0.2989,
+      "step": 191
+    },
+    {
+      "epoch": 2.4536741214057507,
+      "grad_norm": 0.4850566508591137,
+      "learning_rate": 2.4000513875520892e-05,
+      "loss": 0.301,
+      "step": 192
+    },
+    {
+      "epoch": 2.466453674121406,
+      "grad_norm": 0.47472228454385534,
+      "learning_rate": 2.3824965487341247e-05,
+      "loss": 0.3177,
+      "step": 193
+    },
+    {
+      "epoch": 2.479233226837061,
+      "grad_norm": 0.4851559792848614,
+      "learning_rate": 2.3649110684082258e-05,
+      "loss": 0.3308,
+      "step": 194
+    },
+    {
+      "epoch": 2.4920127795527156,
+      "grad_norm": 0.45957374345892266,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.2942,
+      "step": 195
+    },
+    {
+      "epoch": 2.5047923322683707,
+      "grad_norm": 0.478046394348993,
+      "learning_rate": 2.3296538206123134e-05,
+      "loss": 0.3007,
+      "step": 196
+    },
+    {
+      "epoch": 2.5175718849840254,
+      "grad_norm": 0.4500421760261728,
+      "learning_rate": 2.311984877573636e-05,
+      "loss": 0.3154,
+      "step": 197
+    },
+    {
+      "epoch": 2.5303514376996805,
+      "grad_norm": 0.45135851042932,
+      "learning_rate": 2.2942909416634326e-05,
+      "loss": 0.3101,
+      "step": 198
+    },
+    {
+      "epoch": 2.543130990415335,
+      "grad_norm": 0.43566336256690485,
+      "learning_rate": 2.2765734303294666e-05,
+      "loss": 0.3157,
+      "step": 199
+    },
+    {
+      "epoch": 2.5559105431309903,
+      "grad_norm": 0.4708370950757383,
+      "learning_rate": 2.2588337629081107e-05,
+      "loss": 0.3141,
+      "step": 200
+    },
+    {
+      "epoch": 2.5686900958466454,
+      "grad_norm": 0.48710715768801055,
+      "learning_rate": 2.2410733605106462e-05,
+      "loss": 0.2981,
+      "step": 201
+    },
+    {
+      "epoch": 2.5814696485623,
+      "grad_norm": 0.4552285451346172,
+      "learning_rate": 2.2232936459094158e-05,
+      "loss": 0.306,
+      "step": 202
+    },
+    {
+      "epoch": 2.594249201277955,
+      "grad_norm": 0.484036112610832,
+      "learning_rate": 2.205496043423849e-05,
+      "loss": 0.3111,
+      "step": 203
+    },
+    {
+      "epoch": 2.6070287539936103,
+      "grad_norm": 0.4313179941844075,
+      "learning_rate": 2.1876819788063586e-05,
+      "loss": 0.299,
+      "step": 204
+    },
+    {
+      "epoch": 2.619808306709265,
+      "grad_norm": 0.44752600322752145,
+      "learning_rate": 2.16985287912813e-05,
+      "loss": 0.32,
+      "step": 205
+    },
+    {
+      "epoch": 2.63258785942492,
+      "grad_norm": 0.47156813687100724,
+      "learning_rate": 2.1520101726647922e-05,
+      "loss": 0.2955,
+      "step": 206
+    },
+    {
+      "epoch": 2.6453674121405752,
+      "grad_norm": 0.4671296524582926,
+      "learning_rate": 2.1341552887820048e-05,
+      "loss": 0.3256,
+      "step": 207
+    },
+    {
+      "epoch": 2.65814696485623,
+      "grad_norm": 0.477325601348818,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.3093,
+      "step": 208
+    },
+    {
+      "epoch": 2.670926517571885,
+      "grad_norm": 0.4343001880175009,
+      "learning_rate": 2.0984147109837564e-05,
+      "loss": 0.3213,
+      "step": 209
+    },
+    {
+      "epoch": 2.68370607028754,
+      "grad_norm": 0.5021855054737505,
+      "learning_rate": 2.0805318802188307e-05,
+      "loss": 0.3411,
+      "step": 210
+    },
+    {
+      "epoch": 2.696485623003195,
+      "grad_norm": 0.4446138549976679,
+      "learning_rate": 2.0626425981061608e-05,
+      "loss": 0.3308,
+      "step": 211
+    },
+    {
+      "epoch": 2.70926517571885,
+      "grad_norm": 0.47462483959874785,
+      "learning_rate": 2.0447482977425465e-05,
+      "loss": 0.2839,
+      "step": 212
+    },
+    {
+      "epoch": 2.722044728434505,
+      "grad_norm": 0.4880795535243631,
+      "learning_rate": 2.0268504126267952e-05,
+      "loss": 0.3141,
+      "step": 213
+    },
+    {
+      "epoch": 2.7348242811501597,
+      "grad_norm": 0.49550522053315904,
+      "learning_rate": 2.008950376544887e-05,
+      "loss": 0.306,
+      "step": 214
+    },
+    {
+      "epoch": 2.747603833865815,
+      "grad_norm": 0.4773403431106275,
+      "learning_rate": 1.9910496234551132e-05,
+      "loss": 0.3273,
+      "step": 215
+    },
+    {
+      "epoch": 2.7603833865814695,
+      "grad_norm": 0.4664219846205084,
+      "learning_rate": 1.9731495873732055e-05,
+      "loss": 0.3224,
+      "step": 216
+    },
+    {
+      "epoch": 2.7731629392971247,
+      "grad_norm": 0.5332860342603946,
+      "learning_rate": 1.9552517022574542e-05,
+      "loss": 0.2983,
+      "step": 217
+    },
+    {
+      "epoch": 2.7859424920127793,
+      "grad_norm": 0.45159746691880526,
+      "learning_rate": 1.93735740189384e-05,
+      "loss": 0.3037,
+      "step": 218
+    },
+    {
+      "epoch": 2.7987220447284344,
+      "grad_norm": 0.43415754973661796,
+      "learning_rate": 1.9194681197811703e-05,
+      "loss": 0.3344,
+      "step": 219
+    },
+    {
+      "epoch": 2.8115015974440896,
+      "grad_norm": 0.4460053985485829,
+      "learning_rate": 1.901585289016244e-05,
+      "loss": 0.268,
+      "step": 220
+    },
+    {
+      "epoch": 2.8242811501597442,
+      "grad_norm": 0.4687470067793797,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.3099,
+      "step": 221
+    },
+    {
+      "epoch": 2.8370607028753994,
+      "grad_norm": 0.4464233911002488,
+      "learning_rate": 1.8658447112179952e-05,
+      "loss": 0.3306,
+      "step": 222
+    },
+    {
+      "epoch": 2.8498402555910545,
+      "grad_norm": 0.45790496471353626,
+      "learning_rate": 1.8479898273352084e-05,
+      "loss": 0.313,
+      "step": 223
+    },
+    {
+      "epoch": 2.862619808306709,
+      "grad_norm": 0.4477831178911786,
+      "learning_rate": 1.83014712087187e-05,
+      "loss": 0.3073,
+      "step": 224
+    },
+    {
+      "epoch": 2.8753993610223643,
+      "grad_norm": 0.4205151220694784,
+      "learning_rate": 1.8123180211936417e-05,
+      "loss": 0.279,
+      "step": 225
+    },
+    {
+      "epoch": 2.8881789137380194,
+      "grad_norm": 0.45621627231742085,
+      "learning_rate": 1.794503956576152e-05,
+      "loss": 0.3086,
+      "step": 226
+    },
+    {
+      "epoch": 2.900958466453674,
+      "grad_norm": 0.43966286624150863,
+      "learning_rate": 1.776706354090585e-05,
+      "loss": 0.3263,
+      "step": 227
+    },
+    {
+      "epoch": 2.913738019169329,
+      "grad_norm": 0.43712314258828755,
+      "learning_rate": 1.758926639489354e-05,
+      "loss": 0.3111,
+      "step": 228
+    },
+    {
+      "epoch": 2.9265175718849843,
+      "grad_norm": 0.44098349807935044,
+      "learning_rate": 1.7411662370918893e-05,
+      "loss": 0.3201,
+      "step": 229
+    },
+    {
+      "epoch": 2.939297124600639,
+      "grad_norm": 0.44012128305086873,
+      "learning_rate": 1.7234265696705344e-05,
+      "loss": 0.3105,
+      "step": 230
+    },
+    {
+      "epoch": 2.952076677316294,
+      "grad_norm": 0.4261544954967208,
+      "learning_rate": 1.7057090583365678e-05,
+      "loss": 0.3155,
+      "step": 231
+    },
+    {
+      "epoch": 2.9648562300319488,
+      "grad_norm": 0.39501010773563255,
+      "learning_rate": 1.6880151224263646e-05,
+      "loss": 0.3249,
+      "step": 232
+    },
+    {
+      "epoch": 2.977635782747604,
+      "grad_norm": 0.44673652811460274,
+      "learning_rate": 1.6703461793876876e-05,
+      "loss": 0.3347,
+      "step": 233
+    },
+    {
+      "epoch": 2.9904153354632586,
+      "grad_norm": 0.41778124617712703,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.3218,
+      "step": 234
+    },
+    {
+      "epoch": 3.0031948881789137,
+      "grad_norm": 0.4118846954185027,
+      "learning_rate": 1.635088931591775e-05,
+      "loss": 0.2839,
+      "step": 235
+    },
+    {
+      "epoch": 3.015974440894569,
+      "grad_norm": 0.4898464181866644,
+      "learning_rate": 1.6175034512658753e-05,
+      "loss": 0.2162,
+      "step": 236
+    },
+    {
+      "epoch": 3.0287539936102235,
+      "grad_norm": 0.3475153734923242,
+      "learning_rate": 1.5999486124479115e-05,
+      "loss": 0.2219,
+      "step": 237
+    },
+    {
+      "epoch": 3.0415335463258786,
+      "grad_norm": 0.5275561594821107,
+      "learning_rate": 1.5824258214426833e-05,
+      "loss": 0.2289,
+      "step": 238
+    },
+    {
+      "epoch": 3.0543130990415337,
+      "grad_norm": 0.5905626407321699,
+      "learning_rate": 1.5649364819876655e-05,
+      "loss": 0.2155,
+      "step": 239
+    },
+    {
+      "epoch": 3.0670926517571884,
+      "grad_norm": 0.43846638538855603,
+      "learning_rate": 1.547481995140556e-05,
+      "loss": 0.1985,
+      "step": 240
+    },
+    {
+      "epoch": 3.0798722044728435,
+      "grad_norm": 0.4614750070286066,
+      "learning_rate": 1.5300637591670357e-05,
+      "loss": 0.217,
+      "step": 241
+    },
+    {
+      "epoch": 3.0926517571884986,
+      "grad_norm": 0.4216214980650696,
+      "learning_rate": 1.5126831694287564e-05,
+      "loss": 0.1696,
+      "step": 242
+    },
+    {
+      "epoch": 3.1054313099041533,
+      "grad_norm": 0.42013283916246674,
+      "learning_rate": 1.4953416182715566e-05,
+      "loss": 0.1992,
+      "step": 243
+    },
+    {
+      "epoch": 3.1182108626198084,
+      "grad_norm": 0.3648504156967987,
+      "learning_rate": 1.478040494913926e-05,
+      "loss": 0.1906,
+      "step": 244
+    },
+    {
+      "epoch": 3.130990415335463,
+      "grad_norm": 0.3817479070143553,
+      "learning_rate": 1.460781185335713e-05,
+      "loss": 0.2129,
+      "step": 245
+    },
+    {
+      "epoch": 3.143769968051118,
+      "grad_norm": 0.4107900965741899,
+      "learning_rate": 1.443565072167095e-05,
+      "loss": 0.1978,
+      "step": 246
+    },
+    {
+      "epoch": 3.1565495207667733,
+      "grad_norm": 0.40705646221959346,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.1865,
+      "step": 247
+    },
+    {
+      "epoch": 3.169329073482428,
+      "grad_norm": 0.37445060369847183,
+      "learning_rate": 1.409267948166718e-05,
+      "loss": 0.2018,
+      "step": 248
+    },
+    {
+      "epoch": 3.182108626198083,
+      "grad_norm": 0.3663877408196603,
+      "learning_rate": 1.3921896848515064e-05,
+      "loss": 0.209,
+      "step": 249
+    },
+    {
+      "epoch": 3.194888178913738,
+      "grad_norm": 0.36673504033885207,
+      "learning_rate": 1.3751601127588849e-05,
+      "loss": 0.226,
+      "step": 250
+    },
+    {
+      "epoch": 3.207667731629393,
+      "grad_norm": 0.37834028988583124,
+      "learning_rate": 1.3581805961149371e-05,
+      "loss": 0.1875,
+      "step": 251
+    },
+    {
+      "epoch": 3.220447284345048,
+      "grad_norm": 0.3905748603905938,
+      "learning_rate": 1.341252495135841e-05,
+      "loss": 0.178,
+      "step": 252
+    },
+    {
+      "epoch": 3.2332268370607027,
+      "grad_norm": 0.3950966215122354,
+      "learning_rate": 1.324377165918906e-05,
+      "loss": 0.2084,
+      "step": 253
+    },
+    {
+      "epoch": 3.246006389776358,
+      "grad_norm": 0.3684648231126202,
+      "learning_rate": 1.3075559603339354e-05,
+      "loss": 0.1705,
+      "step": 254
+    },
+    {
+      "epoch": 3.258785942492013,
+      "grad_norm": 0.378284322778239,
+      "learning_rate": 1.2907902259149287e-05,
+      "loss": 0.2058,
+      "step": 255
+    },
+    {
+      "epoch": 3.2715654952076676,
+      "grad_norm": 0.368072266734711,
+      "learning_rate": 1.274081305752135e-05,
+      "loss": 0.1958,
+      "step": 256
+    },
+    {
+      "epoch": 3.2843450479233227,
+      "grad_norm": 0.3651821336479358,
+      "learning_rate": 1.2574305383844528e-05,
+      "loss": 0.2059,
+      "step": 257
+    },
+    {
+      "epoch": 3.297124600638978,
+      "grad_norm": 0.34927203165475584,
+      "learning_rate": 1.2408392576922075e-05,
+      "loss": 0.2098,
+      "step": 258
+    },
+    {
+      "epoch": 3.3099041533546325,
+      "grad_norm": 0.39934848760760827,
+      "learning_rate": 1.2243087927902905e-05,
+      "loss": 0.2376,
+      "step": 259
+    },
+    {
+      "epoch": 3.3226837060702876,
+      "grad_norm": 0.3510103557115404,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.1694,
+      "step": 260
+    },
+    {
+      "epoch": 3.3354632587859427,
+      "grad_norm": 0.32940972784569894,
+      "learning_rate": 1.1914356023513904e-05,
+      "loss": 0.1976,
+      "step": 261
+    },
+    {
+      "epoch": 3.3482428115015974,
+      "grad_norm": 0.35356566851137694,
+      "learning_rate": 1.1750955102607193e-05,
+      "loss": 0.2217,
+      "step": 262
+    },
+    {
+      "epoch": 3.3610223642172525,
+      "grad_norm": 0.3433245675706747,
+      "learning_rate": 1.1588215006420374e-05,
+      "loss": 0.1818,
+      "step": 263
+    },
+    {
+      "epoch": 3.373801916932907,
+      "grad_norm": 0.3772192103436919,
+      "learning_rate": 1.1426148771938915e-05,
+      "loss": 0.2124,
+      "step": 264
+    },
+    {
+      "epoch": 3.3865814696485623,
+      "grad_norm": 0.3536916172844722,
+      "learning_rate": 1.1264769382165748e-05,
+      "loss": 0.1761,
+      "step": 265
+    },
+    {
+      "epoch": 3.3993610223642174,
+      "grad_norm": 0.32618493991756353,
+      "learning_rate": 1.110408976508118e-05,
+      "loss": 0.1937,
+      "step": 266
+    },
+    {
+      "epoch": 3.412140575079872,
+      "grad_norm": 0.3242579981685326,
+      "learning_rate": 1.094412279260726e-05,
+      "loss": 0.1983,
+      "step": 267
+    },
+    {
+      "epoch": 3.4249201277955272,
+      "grad_norm": 0.3507572481745218,
+      "learning_rate": 1.0784881279576635e-05,
+      "loss": 0.1906,
+      "step": 268
+    },
+    {
+      "epoch": 3.437699680511182,
+      "grad_norm": 0.3511832840549672,
+      "learning_rate": 1.0626377982705929e-05,
+      "loss": 0.2081,
+      "step": 269
+    },
+    {
+      "epoch": 3.450479233226837,
+      "grad_norm": 0.3393476520653178,
+      "learning_rate": 1.0468625599573842e-05,
+      "loss": 0.1895,
+      "step": 270
+    },
+    {
+      "epoch": 3.463258785942492,
+      "grad_norm": 0.3400743743918658,
+      "learning_rate": 1.0311636767603952e-05,
+      "loss": 0.189,
+      "step": 271
+    },
+    {
+      "epoch": 3.476038338658147,
+      "grad_norm": 0.34521702139503097,
+      "learning_rate": 1.0155424063052306e-05,
+      "loss": 0.1834,
+      "step": 272
+    },
+    {
+      "epoch": 3.488817891373802,
+      "grad_norm": 0.3478326744060575,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.214,
+      "step": 273
+    },
+    {
+      "epoch": 3.501597444089457,
+      "grad_norm": 0.3311182191085231,
+      "learning_rate": 9.84537702935065e-06,
+      "loss": 0.1585,
+      "step": 274
+    },
+    {
+      "epoch": 3.5143769968051117,
+      "grad_norm": 0.3402617716899934,
+      "learning_rate": 9.691567537832964e-06,
+      "loss": 0.2098,
+      "step": 275
+    },
+    {
+      "epoch": 3.527156549520767,
+      "grad_norm": 0.32990975767100783,
+      "learning_rate": 9.538583847008452e-06,
+      "loss": 0.2042,
+      "step": 276
+    },
+    {
+      "epoch": 3.539936102236422,
+      "grad_norm": 0.34241938769940633,
+      "learning_rate": 9.386438212284372e-06,
+      "loss": 0.1823,
+      "step": 277
+    },
+    {
+      "epoch": 3.5527156549520766,
+      "grad_norm": 0.32040946109330926,
+      "learning_rate": 9.235142821931928e-06,
+      "loss": 0.1604,
+      "step": 278
+    },
+    {
+      "epoch": 3.5654952076677318,
+      "grad_norm": 0.3371299079090535,
+      "learning_rate": 9.084709796109907e-06,
+      "loss": 0.1696,
+      "step": 279
+    },
+    {
+      "epoch": 3.5782747603833864,
+      "grad_norm": 0.38134689984011066,
+      "learning_rate": 8.93515118589373e-06,
+      "loss": 0.2058,
+      "step": 280
+    },
+    {
+      "epoch": 3.5910543130990416,
+      "grad_norm": 0.37741904375380997,
+      "learning_rate": 8.786478972310023e-06,
+      "loss": 0.1898,
+      "step": 281
+    },
+    {
+      "epoch": 3.6038338658146962,
+      "grad_norm": 0.3184732845288458,
+      "learning_rate": 8.638705065376887e-06,
+      "loss": 0.2277,
+      "step": 282
+    },
+    {
+      "epoch": 3.6166134185303513,
+      "grad_norm": 0.3551673243357515,
+      "learning_rate": 8.491841303149728e-06,
+      "loss": 0.2038,
+      "step": 283
+    },
+    {
+      "epoch": 3.6293929712460065,
+      "grad_norm": 0.3200397792956637,
+      "learning_rate": 8.345899450772975e-06,
+      "loss": 0.1912,
+      "step": 284
+    },
+    {
+      "epoch": 3.642172523961661,
+      "grad_norm": 0.3165951666590773,
+      "learning_rate": 8.200891199537549e-06,
+      "loss": 0.1589,
+      "step": 285
+    },
+    {
+      "epoch": 3.6549520766773163,
+      "grad_norm": 0.3331867277874367,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.1978,
+      "step": 286
+    },
+    {
+      "epoch": 3.6677316293929714,
+      "grad_norm": 0.32956221422208387,
+      "learning_rate": 7.913721890773354e-06,
+      "loss": 0.1989,
+      "step": 287
+    },
+    {
+      "epoch": 3.680511182108626,
+      "grad_norm": 0.3089320627340336,
+      "learning_rate": 7.771583838159756e-06,
+      "loss": 0.2057,
+      "step": 288
+    },
+    {
+      "epoch": 3.693290734824281,
+      "grad_norm": 0.3414056292548639,
+      "learning_rate": 7.630425394674903e-06,
+      "loss": 0.2062,
+      "step": 289
+    },
+    {
+      "epoch": 3.7060702875399363,
+      "grad_norm": 0.345166863705804,
+      "learning_rate": 7.49025786841445e-06,
+      "loss": 0.2177,
+      "step": 290
+    },
+    {
+      "epoch": 3.718849840255591,
+      "grad_norm": 0.3325079476961589,
+      "learning_rate": 7.3510924880924575e-06,
+      "loss": 0.1732,
+      "step": 291
+    },
+    {
+      "epoch": 3.731629392971246,
+      "grad_norm": 0.3278845587133297,
+      "learning_rate": 7.212940402141808e-06,
+      "loss": 0.1873,
+      "step": 292
+    },
+    {
+      "epoch": 3.744408945686901,
+      "grad_norm": 0.33267417285846135,
+      "learning_rate": 7.075812677821145e-06,
+      "loss": 0.1837,
+      "step": 293
+    },
+    {
+      "epoch": 3.757188498402556,
+      "grad_norm": 0.3366856452726616,
+      "learning_rate": 6.939720300328303e-06,
+      "loss": 0.2002,
+      "step": 294
+    },
+    {
+      "epoch": 3.769968051118211,
+      "grad_norm": 0.321208777278204,
+      "learning_rate": 6.8046741719202385e-06,
+      "loss": 0.1959,
+      "step": 295
+    },
+    {
+      "epoch": 3.7827476038338657,
+      "grad_norm": 0.3375970379867743,
+      "learning_rate": 6.67068511103971e-06,
+      "loss": 0.1789,
+      "step": 296
+    },
+    {
+      "epoch": 3.7955271565495208,
+      "grad_norm": 0.32158191602872427,
+      "learning_rate": 6.537763851448593e-06,
+      "loss": 0.1707,
+      "step": 297
+    },
+    {
+      "epoch": 3.8083067092651754,
+      "grad_norm": 0.3055621332789564,
+      "learning_rate": 6.4059210413680175e-06,
+      "loss": 0.1819,
+      "step": 298
+    },
+    {
+      "epoch": 3.8210862619808306,
+      "grad_norm": 0.3098397894666991,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.2039,
+      "step": 299
+    },
+    {
+      "epoch": 3.8338658146964857,
+      "grad_norm": 0.3350047904411531,
+      "learning_rate": 6.145512929808013e-06,
+      "loss": 0.1827,
+      "step": 300
+    },
+    {
+      "epoch": 3.8466453674121404,
+      "grad_norm": 0.31827064394272075,
+      "learning_rate": 6.016968489424572e-06,
+      "loss": 0.2014,
+      "step": 301
+    },
+    {
+      "epoch": 3.8594249201277955,
+      "grad_norm": 0.3366698689028845,
+      "learning_rate": 5.889544219072465e-06,
+      "loss": 0.1957,
+      "step": 302
+    },
+    {
+      "epoch": 3.8722044728434506,
+      "grad_norm": 0.29980922295653295,
+      "learning_rate": 5.7632503266131925e-06,
+      "loss": 0.1858,
+      "step": 303
+    },
+    {
+      "epoch": 3.8849840255591053,
+      "grad_norm": 0.3108974797098454,
+      "learning_rate": 5.638096929354522e-06,
+      "loss": 0.2017,
+      "step": 304
+    },
+    {
+      "epoch": 3.8977635782747604,
+      "grad_norm": 0.31273298348609424,
+      "learning_rate": 5.514094053240035e-06,
+      "loss": 0.1959,
+      "step": 305
+    },
+    {
+      "epoch": 3.9105431309904155,
+      "grad_norm": 0.3177550801502446,
+      "learning_rate": 5.39125163204594e-06,
+      "loss": 0.1862,
+      "step": 306
+    },
+    {
+      "epoch": 3.92332268370607,
+      "grad_norm": 0.327808048296833,
+      "learning_rate": 5.269579506585259e-06,
+      "loss": 0.2253,
+      "step": 307
+    },
+    {
+      "epoch": 3.9361022364217253,
+      "grad_norm": 0.31501899934860117,
+      "learning_rate": 5.149087423919541e-06,
+      "loss": 0.1802,
+      "step": 308
+    },
+    {
+      "epoch": 3.9488817891373804,
+      "grad_norm": 0.2945599502076206,
+      "learning_rate": 5.029785036577976e-06,
+      "loss": 0.1976,
+      "step": 309
+    },
+    {
+      "epoch": 3.961661341853035,
+      "grad_norm": 0.3095589565776998,
+      "learning_rate": 4.911681901784198e-06,
+      "loss": 0.184,
+      "step": 310
+    },
+    {
+      "epoch": 3.97444089456869,
+      "grad_norm": 0.30910384088987386,
+      "learning_rate": 4.794787480690597e-06,
+      "loss": 0.1925,
+      "step": 311
+    },
+    {
+      "epoch": 3.987220447284345,
+      "grad_norm": 0.290395337724728,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.2046,
+      "step": 312
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.32962339482294273,
+      "learning_rate": 4.5646621393177e-06,
+      "loss": 0.1694,
+      "step": 313
+    },
+    {
+      "epoch": 4.012779552715655,
+      "grad_norm": 0.37895492313335855,
+      "learning_rate": 4.451449654204685e-06,
+      "loss": 0.1533,
+      "step": 314
+    },
+    {
+      "epoch": 4.02555910543131,
+      "grad_norm": 0.35175078665302684,
+      "learning_rate": 4.339482751647557e-06,
+      "loss": 0.1668,
+      "step": 315
+    },
+    {
+      "epoch": 4.038338658146965,
+      "grad_norm": 0.3084303036064777,
+      "learning_rate": 4.228770401229824e-06,
+      "loss": 0.1443,
+      "step": 316
+    },
+    {
+      "epoch": 4.05111821086262,
+      "grad_norm": 0.24654455468032596,
+      "learning_rate": 4.119321472033779e-06,
+      "loss": 0.1268,
+      "step": 317
+    },
+    {
+      "epoch": 4.063897763578275,
+      "grad_norm": 0.24478097825177367,
+      "learning_rate": 4.011144731929981e-06,
+      "loss": 0.1425,
+      "step": 318
+    },
+    {
+      "epoch": 4.07667731629393,
+      "grad_norm": 0.29823705953007107,
+      "learning_rate": 3.904248846874894e-06,
+      "loss": 0.1484,
+      "step": 319
+    },
+    {
+      "epoch": 4.0894568690095845,
+      "grad_norm": 0.3513695808566353,
+      "learning_rate": 3.7986423802166705e-06,
+      "loss": 0.1675,
+      "step": 320
+    },
+    {
+      "epoch": 4.102236421725239,
+      "grad_norm": 0.33145804172476107,
+      "learning_rate": 3.694333792009115e-06,
+      "loss": 0.1514,
+      "step": 321
+    },
+    {
+      "epoch": 4.115015974440895,
+      "grad_norm": 0.3244682368477128,
+      "learning_rate": 3.5913314383339937e-06,
+      "loss": 0.1448,
+      "step": 322
+    },
+    {
+      "epoch": 4.127795527156549,
+      "grad_norm": 0.2912835832312697,
+      "learning_rate": 3.4896435706316e-06,
+      "loss": 0.139,
+      "step": 323
+    },
+    {
+      "epoch": 4.140575079872204,
+      "grad_norm": 0.25664533550038154,
+      "learning_rate": 3.3892783350397675e-06,
+      "loss": 0.1428,
+      "step": 324
+    },
+    {
+      "epoch": 4.15335463258786,
+      "grad_norm": 0.2485015346847118,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.1266,
+      "step": 325
+    },
+    {
+      "epoch": 4.166134185303514,
+      "grad_norm": 0.2667976785080385,
+      "learning_rate": 3.1925478143197418e-06,
+      "loss": 0.1491,
+      "step": 326
+    },
+    {
+      "epoch": 4.178913738019169,
+      "grad_norm": 0.2545505054473165,
+      "learning_rate": 3.0961982891241083e-06,
+      "loss": 0.1226,
+      "step": 327
+    },
+    {
+      "epoch": 4.1916932907348246,
+      "grad_norm": 0.26137003051603064,
+      "learning_rate": 3.001202914641628e-06,
+      "loss": 0.1258,
+      "step": 328
+    },
+    {
+      "epoch": 4.204472843450479,
+      "grad_norm": 0.24942552630422518,
+      "learning_rate": 2.907569300879596e-06,
+      "loss": 0.1498,
+      "step": 329
+    },
+    {
+      "epoch": 4.217252396166134,
+      "grad_norm": 0.23766244655852287,
+      "learning_rate": 2.815304948755664e-06,
+      "loss": 0.1538,
+      "step": 330
+    },
+    {
+      "epoch": 4.2300319488817895,
+      "grad_norm": 0.2448089304122396,
+      "learning_rate": 2.7244172494969978e-06,
+      "loss": 0.1462,
+      "step": 331
+    },
+    {
+      "epoch": 4.242811501597444,
+      "grad_norm": 0.2479082333540749,
+      "learning_rate": 2.6349134840481294e-06,
+      "loss": 0.1447,
+      "step": 332
+    },
+    {
+      "epoch": 4.255591054313099,
+      "grad_norm": 0.25087229695568564,
+      "learning_rate": 2.546800822487714e-06,
+      "loss": 0.1451,
+      "step": 333
+    },
+    {
+      "epoch": 4.268370607028754,
+      "grad_norm": 0.2482830226361139,
+      "learning_rate": 2.4600863234541338e-06,
+      "loss": 0.1562,
+      "step": 334
+    },
+    {
+      "epoch": 4.281150159744409,
+      "grad_norm": 0.25698774921447254,
+      "learning_rate": 2.374776933580025e-06,
+      "loss": 0.1619,
+      "step": 335
+    },
+    {
+      "epoch": 4.293929712460064,
+      "grad_norm": 0.2401487421066426,
+      "learning_rate": 2.2908794869358044e-06,
+      "loss": 0.1222,
+      "step": 336
+    },
+    {
+      "epoch": 4.306709265175719,
+      "grad_norm": 0.23443848991398947,
+      "learning_rate": 2.2084007044821764e-06,
+      "loss": 0.1446,
+      "step": 337
+    },
+    {
+      "epoch": 4.319488817891374,
+      "grad_norm": 0.25071970122976406,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.1359,
+      "step": 338
+    },
+    {
+      "epoch": 4.332268370607029,
+      "grad_norm": 0.23926322256510976,
+      "learning_rate": 2.0477254472197237e-06,
+      "loss": 0.1554,
+      "step": 339
+    },
+    {
+      "epoch": 4.345047923322683,
+      "grad_norm": 0.2489405396679404,
+      "learning_rate": 1.96954184398368e-06,
+      "loss": 0.149,
+      "step": 340
+    },
+    {
+      "epoch": 4.357827476038339,
+      "grad_norm": 0.24072103848564777,
+      "learning_rate": 1.8928026470526917e-06,
+      "loss": 0.128,
+      "step": 341
+    },
+    {
+      "epoch": 4.3706070287539935,
+      "grad_norm": 0.24424363634766746,
+      "learning_rate": 1.817514003945524e-06,
+      "loss": 0.1578,
+      "step": 342
+    },
+    {
+      "epoch": 4.383386581469648,
+      "grad_norm": 0.23447034458231758,
+      "learning_rate": 1.743681945978184e-06,
+      "loss": 0.1594,
+      "step": 343
+    },
+    {
+      "epoch": 4.396166134185304,
+      "grad_norm": 0.24308934604128066,
+      "learning_rate": 1.6713123877807413e-06,
+      "loss": 0.15,
+      "step": 344
+    },
+    {
+      "epoch": 4.4089456869009584,
+      "grad_norm": 0.2384136053396971,
+      "learning_rate": 1.6004111268235156e-06,
+      "loss": 0.1495,
+      "step": 345
+    },
+    {
+      "epoch": 4.421725239616613,
+      "grad_norm": 0.2566703636838048,
+      "learning_rate": 1.5309838429526714e-06,
+      "loss": 0.1318,
+      "step": 346
+    },
+    {
+      "epoch": 4.434504792332269,
+      "grad_norm": 0.2405475065975282,
+      "learning_rate": 1.4630360979351644e-06,
+      "loss": 0.1383,
+      "step": 347
+    },
+    {
+      "epoch": 4.447284345047923,
+      "grad_norm": 0.22829935755735,
+      "learning_rate": 1.396573335013236e-06,
+      "loss": 0.1317,
+      "step": 348
+    },
+    {
+      "epoch": 4.460063897763578,
+      "grad_norm": 0.22379300993973641,
+      "learning_rate": 1.3316008784683265e-06,
+      "loss": 0.1433,
+      "step": 349
+    },
+    {
+      "epoch": 4.472843450479234,
+      "grad_norm": 0.23948442317999727,
+      "learning_rate": 1.2681239331945695e-06,
+      "loss": 0.1541,
+      "step": 350
+    },
+    {
+      "epoch": 4.485623003194888,
+      "grad_norm": 0.2378807940092141,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.1543,
+      "step": 351
+    },
+    {
+      "epoch": 4.498402555910543,
+      "grad_norm": 0.22991699876912358,
+      "learning_rate": 1.1456767966083393e-06,
+      "loss": 0.1626,
+      "step": 352
+    },
+    {
+      "epoch": 4.511182108626198,
+      "grad_norm": 0.23534690525704505,
+      "learning_rate": 1.086716414442952e-06,
+      "loss": 0.1393,
+      "step": 353
+    },
+    {
+      "epoch": 4.523961661341853,
+      "grad_norm": 0.22984471517651295,
+      "learning_rate": 1.0292711610570904e-06,
+      "loss": 0.1349,
+      "step": 354
+    },
+    {
+      "epoch": 4.536741214057508,
+      "grad_norm": 0.24717164483819987,
+      "learning_rate": 9.733456383463658e-07,
+      "loss": 0.1762,
+      "step": 355
+    },
+    {
+      "epoch": 4.549520766773163,
+      "grad_norm": 0.22851304628815633,
+      "learning_rate": 9.189443264619102e-07,
+      "loss": 0.141,
+      "step": 356
+    },
+    {
+      "epoch": 4.562300319488818,
+      "grad_norm": 0.24018082021800888,
+      "learning_rate": 8.660715834514977e-07,
+      "loss": 0.1305,
+      "step": 357
+    },
+    {
+      "epoch": 4.575079872204473,
+      "grad_norm": 0.23090582591157588,
+      "learning_rate": 8.147316449103959e-07,
+      "loss": 0.162,
+      "step": 358
+    },
+    {
+      "epoch": 4.587859424920127,
+      "grad_norm": 0.21323658350177105,
+      "learning_rate": 7.649286236420806e-07,
+      "loss": 0.1221,
+      "step": 359
+    },
+    {
+      "epoch": 4.600638977635783,
+      "grad_norm": 0.24556164593191016,
+      "learning_rate": 7.166665093287539e-07,
+      "loss": 0.1613,
+      "step": 360
+    },
+    {
+      "epoch": 4.613418530351438,
+      "grad_norm": 0.22903125199290664,
+      "learning_rate": 6.69949168211721e-07,
+      "loss": 0.1398,
+      "step": 361
+    },
+    {
+      "epoch": 4.626198083067092,
+      "grad_norm": 0.24584858293592662,
+      "learning_rate": 6.247803427816945e-07,
+      "loss": 0.1635,
+      "step": 362
+    },
+    {
+      "epoch": 4.638977635782748,
+      "grad_norm": 0.22174496211525754,
+      "learning_rate": 5.811636514789598e-07,
+      "loss": 0.1306,
+      "step": 363
+    },
+    {
+      "epoch": 4.651757188498403,
+      "grad_norm": 0.21490418012521362,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.1254,
+      "step": 364
+    },
+    {
+      "epoch": 4.664536741214057,
+      "grad_norm": 0.24337815202535223,
+      "learning_rate": 4.986005230351954e-07,
+      "loss": 0.1629,
+      "step": 365
+    },
+    {
+      "epoch": 4.677316293929713,
+      "grad_norm": 0.2347402767470565,
+      "learning_rate": 4.5966069996365993e-07,
+      "loss": 0.1477,
+      "step": 366
+    },
+    {
+      "epoch": 4.6900958466453675,
+      "grad_norm": 0.23582230301531468,
+      "learning_rate": 4.22286238628562e-07,
+      "loss": 0.1305,
+      "step": 367
+    },
+    {
+      "epoch": 4.702875399361022,
+      "grad_norm": 0.23362884394351063,
+      "learning_rate": 3.8648013306960664e-07,
+      "loss": 0.1652,
+      "step": 368
+    },
+    {
+      "epoch": 4.715654952076678,
+      "grad_norm": 0.23118875956783638,
+      "learning_rate": 3.522452516867048e-07,
+      "loss": 0.14,
+      "step": 369
+    },
+    {
+      "epoch": 4.728434504792332,
+      "grad_norm": 0.22885718753512982,
+      "learning_rate": 3.1958433701019697e-07,
+      "loss": 0.1581,
+      "step": 370
+    },
+    {
+      "epoch": 4.741214057507987,
+      "grad_norm": 0.21816020871681144,
+      "learning_rate": 2.8850000548115155e-07,
+      "loss": 0.1247,
+      "step": 371
+    },
+    {
+      "epoch": 4.753993610223642,
+      "grad_norm": 0.24112108226087545,
+      "learning_rate": 2.5899474724174313e-07,
+      "loss": 0.1604,
+      "step": 372
+    },
+    {
+      "epoch": 4.766773162939297,
+      "grad_norm": 0.2727904649496295,
+      "learning_rate": 2.3107092593579905e-07,
+      "loss": 0.1569,
+      "step": 373
+    },
+    {
+      "epoch": 4.779552715654952,
+      "grad_norm": 0.22501840756080654,
+      "learning_rate": 2.0473077851942858e-07,
+      "loss": 0.1537,
+      "step": 374
+    },
+    {
+      "epoch": 4.792332268370607,
+      "grad_norm": 0.23250124505030753,
+      "learning_rate": 1.799764150818306e-07,
+      "loss": 0.1526,
+      "step": 375
+    },
+    {
+      "epoch": 4.805111821086262,
+      "grad_norm": 0.2382431019703871,
+      "learning_rate": 1.5680981867625566e-07,
+      "loss": 0.1314,
+      "step": 376
+    },
+    {
+      "epoch": 4.817891373801917,
+      "grad_norm": 0.23282919857016401,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.1396,
+      "step": 377
+    },
+    {
+      "epoch": 4.830670926517572,
+      "grad_norm": 0.23493935207815764,
+      "learning_rate": 1.1524722305144231e-07,
+      "loss": 0.1502,
+      "step": 378
+    },
+    {
+      "epoch": 4.843450479233227,
+      "grad_norm": 0.23404071158283998,
+      "learning_rate": 9.685455338016347e-08,
+      "loss": 0.1216,
+      "step": 379
+    },
+    {
+      "epoch": 4.856230031948882,
+      "grad_norm": 0.22382988499816492,
+      "learning_rate": 8.005630957010014e-08,
+      "loss": 0.1472,
+      "step": 380
+    },
+    {
+      "epoch": 4.8690095846645365,
+      "grad_norm": 0.23097503960623259,
+      "learning_rate": 6.485383731580142e-08,
+      "loss": 0.1428,
+      "step": 381
+    },
+    {
+      "epoch": 4.881789137380192,
+      "grad_norm": 0.22768705230447,
+      "learning_rate": 5.1248354475768034e-08,
+      "loss": 0.1345,
+      "step": 382
+    },
+    {
+      "epoch": 4.894568690095847,
+      "grad_norm": 0.2258297215315307,
+      "learning_rate": 3.924095097489922e-08,
+      "loss": 0.1471,
+      "step": 383
+    },
+    {
+      "epoch": 4.907348242811501,
+      "grad_norm": 0.2283167583269427,
+      "learning_rate": 2.8832588717164766e-08,
+      "loss": 0.1423,
+      "step": 384
+    },
+    {
+      "epoch": 4.920127795527156,
+      "grad_norm": 0.2257318382233484,
+      "learning_rate": 2.0024101508555604e-08,
+      "loss": 0.1506,
+      "step": 385
+    },
+    {
+      "epoch": 4.932907348242812,
+      "grad_norm": 0.2200908120664162,
+      "learning_rate": 1.281619499029274e-08,
+      "loss": 0.1511,
+      "step": 386
+    },
+    {
+      "epoch": 4.945686900958466,
+      "grad_norm": 0.23881077866912595,
+      "learning_rate": 7.209446582292501e-09,
+      "loss": 0.1504,
+      "step": 387
+    },
+    {
+      "epoch": 4.958466453674122,
+      "grad_norm": 0.24375578451468075,
+      "learning_rate": 3.2043054369057523e-09,
+      "loss": 0.152,
+      "step": 388
+    },
+    {
+      "epoch": 4.9712460063897765,
+      "grad_norm": 0.23619042060971052,
+      "learning_rate": 8.010924029533406e-10,
+      "loss": 0.172,
+      "step": 389
+    },
+    {
+      "epoch": 4.984025559105431,
+      "grad_norm": 0.22988874661865163,
+      "learning_rate": 0.0,
+      "loss": 0.1405,
+      "step": 390
+    },
+    {
+      "epoch": 4.984025559105431,
+      "step": 390,
+      "total_flos": 4.8627183877370675e+17,
+      "train_loss": 0.3829498979716729,
+      "train_runtime": 7298.0864,
+      "train_samples_per_second": 6.851,
+      "train_steps_per_second": 0.053
+    }
+  ],
+  "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": 4.8627183877370675e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}