{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 1,
  "global_step": 80,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.125,
      "grad_norm": 21.28069305419922,
      "learning_rate": 2.5e-05,
      "loss": 1.2067,
      "step": 1
    },
    {
      "epoch": 0.125,
      "eval_accuracy": 0.3548387096774194,
      "eval_f1": 0.2145748987854251,
      "eval_loss": 1.2299962043762207,
      "eval_runtime": 1.8781,
      "eval_samples_per_second": 33.012,
      "eval_steps_per_second": 2.13,
      "step": 1
    },
    {
      "epoch": 0.25,
      "grad_norm": 16.937559127807617,
      "learning_rate": 5e-05,
      "loss": 1.3485,
      "step": 2
    },
    {
      "epoch": 0.25,
      "eval_accuracy": 0.3709677419354839,
      "eval_f1": 0.18930041152263377,
      "eval_loss": 1.1593687534332275,
      "eval_runtime": 1.874,
      "eval_samples_per_second": 33.084,
      "eval_steps_per_second": 2.134,
      "step": 2
    },
    {
      "epoch": 0.375,
      "grad_norm": 21.755922317504883,
      "learning_rate": 4.935897435897436e-05,
      "loss": 1.2559,
      "step": 3
    },
    {
      "epoch": 0.375,
      "eval_accuracy": 0.46774193548387094,
      "eval_f1": 0.23272633346248225,
      "eval_loss": 1.0302419662475586,
      "eval_runtime": 1.875,
      "eval_samples_per_second": 33.067,
      "eval_steps_per_second": 2.133,
      "step": 3
    },
    {
      "epoch": 0.5,
      "grad_norm": 10.690485954284668,
      "learning_rate": 4.871794871794872e-05,
      "loss": 1.0408,
      "step": 4
    },
    {
      "epoch": 0.5,
      "eval_accuracy": 0.46774193548387094,
      "eval_f1": 0.24990730441231,
      "eval_loss": 0.9755544066429138,
      "eval_runtime": 1.8757,
      "eval_samples_per_second": 33.055,
      "eval_steps_per_second": 2.133,
      "step": 4
    },
    {
      "epoch": 0.625,
      "grad_norm": 8.635258674621582,
      "learning_rate": 4.8076923076923084e-05,
      "loss": 0.8126,
      "step": 5
    },
    {
      "epoch": 0.625,
      "eval_accuracy": 0.46774193548387094,
      "eval_f1": 0.2485632183908046,
      "eval_loss": 0.9575037956237793,
      "eval_runtime": 1.9274,
      "eval_samples_per_second": 32.168,
      "eval_steps_per_second": 2.075,
      "step": 5
    },
    {
      "epoch": 0.75,
      "grad_norm": 11.213544845581055,
      "learning_rate": 4.7435897435897435e-05,
      "loss": 0.9848,
      "step": 6
    },
    {
      "epoch": 0.75,
      "eval_accuracy": 0.5161290322580645,
      "eval_f1": 0.28421309872922773,
      "eval_loss": 0.951880693435669,
      "eval_runtime": 1.9311,
      "eval_samples_per_second": 32.106,
      "eval_steps_per_second": 2.071,
      "step": 6
    },
    {
      "epoch": 0.875,
      "grad_norm": 7.83177375793457,
      "learning_rate": 4.67948717948718e-05,
      "loss": 0.8319,
      "step": 7
    },
    {
      "epoch": 0.875,
      "eval_accuracy": 0.532258064516129,
      "eval_f1": 0.3174603174603175,
      "eval_loss": 0.962181806564331,
      "eval_runtime": 1.8805,
      "eval_samples_per_second": 32.97,
      "eval_steps_per_second": 2.127,
      "step": 7
    },
    {
      "epoch": 1.0,
      "grad_norm": 12.740473747253418,
      "learning_rate": 4.615384615384616e-05,
      "loss": 1.0211,
      "step": 8
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5,
      "eval_f1": 0.3313632781717888,
      "eval_loss": 0.9630481600761414,
      "eval_runtime": 1.8802,
      "eval_samples_per_second": 32.975,
      "eval_steps_per_second": 2.127,
      "step": 8
    },
    {
      "epoch": 1.125,
      "grad_norm": 12.164949417114258,
      "learning_rate": 4.5512820512820516e-05,
      "loss": 0.8899,
      "step": 9
    },
    {
      "epoch": 1.125,
      "eval_accuracy": 0.5,
      "eval_f1": 0.3108974358974359,
      "eval_loss": 0.9681987166404724,
      "eval_runtime": 1.8773,
      "eval_samples_per_second": 33.026,
      "eval_steps_per_second": 2.131,
      "step": 9
    },
    {
      "epoch": 1.25,
      "grad_norm": 6.146948337554932,
      "learning_rate": 4.4871794871794874e-05,
      "loss": 0.7234,
      "step": 10
    },
    {
      "epoch": 1.25,
      "eval_accuracy": 0.5,
      "eval_f1": 0.3108974358974359,
      "eval_loss": 0.9844852685928345,
      "eval_runtime": 1.8795,
      "eval_samples_per_second": 32.988,
      "eval_steps_per_second": 2.128,
      "step": 10
    },
    {
      "epoch": 1.375,
      "grad_norm": 8.822149276733398,
      "learning_rate": 4.423076923076923e-05,
      "loss": 1.0579,
      "step": 11
    },
    {
      "epoch": 1.375,
      "eval_accuracy": 0.5,
      "eval_f1": 0.3108974358974359,
      "eval_loss": 0.9911479353904724,
      "eval_runtime": 1.8777,
      "eval_samples_per_second": 33.019,
      "eval_steps_per_second": 2.13,
      "step": 11
    },
    {
      "epoch": 1.5,
      "grad_norm": 9.482677459716797,
      "learning_rate": 4.358974358974359e-05,
      "loss": 0.8232,
      "step": 12
    },
    {
      "epoch": 1.5,
      "eval_accuracy": 0.5,
      "eval_f1": 0.3108974358974359,
      "eval_loss": 0.9866904020309448,
      "eval_runtime": 1.9302,
      "eval_samples_per_second": 32.121,
      "eval_steps_per_second": 2.072,
      "step": 12
    },
    {
      "epoch": 1.625,
      "grad_norm": 17.120052337646484,
      "learning_rate": 4.294871794871795e-05,
      "loss": 1.063,
      "step": 13
    },
    {
      "epoch": 1.625,
      "eval_accuracy": 0.5,
      "eval_f1": 0.3108974358974359,
      "eval_loss": 0.9797441959381104,
      "eval_runtime": 1.9281,
      "eval_samples_per_second": 32.157,
      "eval_steps_per_second": 2.075,
      "step": 13
    },
    {
      "epoch": 1.75,
      "grad_norm": 8.349448204040527,
      "learning_rate": 4.230769230769231e-05,
      "loss": 0.9684,
      "step": 14
    },
    {
      "epoch": 1.75,
      "eval_accuracy": 0.46774193548387094,
      "eval_f1": 0.2941421721909527,
      "eval_loss": 0.9708881974220276,
      "eval_runtime": 1.9314,
      "eval_samples_per_second": 32.1,
      "eval_steps_per_second": 2.071,
      "step": 14
    },
    {
      "epoch": 1.875,
      "grad_norm": 4.90641975402832,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.9342,
      "step": 15
    },
    {
      "epoch": 1.875,
      "eval_accuracy": 0.4838709677419355,
      "eval_f1": 0.31024531024531027,
      "eval_loss": 0.9690807461738586,
      "eval_runtime": 1.8766,
      "eval_samples_per_second": 33.038,
      "eval_steps_per_second": 2.131,
      "step": 15
    },
    {
      "epoch": 2.0,
      "grad_norm": 9.04134750366211,
      "learning_rate": 4.1025641025641023e-05,
      "loss": 0.7387,
      "step": 16
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.4838709677419355,
      "eval_f1": 0.31024531024531027,
      "eval_loss": 0.9698801636695862,
      "eval_runtime": 1.8773,
      "eval_samples_per_second": 33.026,
      "eval_steps_per_second": 2.131,
      "step": 16
    },
    {
      "epoch": 2.125,
      "grad_norm": 5.919072151184082,
      "learning_rate": 4.038461538461539e-05,
      "loss": 0.8603,
      "step": 17
    },
    {
      "epoch": 2.125,
      "eval_accuracy": 0.4838709677419355,
      "eval_f1": 0.30253164556962026,
      "eval_loss": 0.9661668539047241,
      "eval_runtime": 1.928,
      "eval_samples_per_second": 32.158,
      "eval_steps_per_second": 2.075,
      "step": 17
    },
    {
      "epoch": 2.25,
      "grad_norm": 7.429778575897217,
      "learning_rate": 3.974358974358974e-05,
      "loss": 0.8114,
      "step": 18
    },
    {
      "epoch": 2.25,
      "eval_accuracy": 0.5,
      "eval_f1": 0.3189493433395872,
      "eval_loss": 0.9636289477348328,
      "eval_runtime": 1.9287,
      "eval_samples_per_second": 32.145,
      "eval_steps_per_second": 2.074,
      "step": 18
    },
    {
      "epoch": 2.375,
      "grad_norm": 6.09760046005249,
      "learning_rate": 3.9102564102564105e-05,
      "loss": 0.9508,
      "step": 19
    },
    {
      "epoch": 2.375,
      "eval_accuracy": 0.5,
      "eval_f1": 0.3189493433395872,
      "eval_loss": 0.9564090967178345,
      "eval_runtime": 1.8792,
      "eval_samples_per_second": 32.992,
      "eval_steps_per_second": 2.129,
      "step": 19
    },
    {
      "epoch": 2.5,
      "grad_norm": 4.588860988616943,
      "learning_rate": 3.846153846153846e-05,
      "loss": 0.7554,
      "step": 20
    },
    {
      "epoch": 2.5,
      "eval_accuracy": 0.5161290322580645,
      "eval_f1": 0.33455732354141166,
      "eval_loss": 0.9543712735176086,
      "eval_runtime": 1.9285,
      "eval_samples_per_second": 32.15,
      "eval_steps_per_second": 2.074,
      "step": 20
    },
    {
      "epoch": 2.625,
      "grad_norm": 4.015089988708496,
      "learning_rate": 3.782051282051282e-05,
      "loss": 0.8772,
      "step": 21
    },
    {
      "epoch": 2.625,
      "eval_accuracy": 0.5483870967741935,
      "eval_f1": 0.36436436436436437,
      "eval_loss": 0.9526702165603638,
      "eval_runtime": 1.9277,
      "eval_samples_per_second": 32.162,
      "eval_steps_per_second": 2.075,
      "step": 21
    },
    {
      "epoch": 2.75,
      "grad_norm": 7.522365570068359,
      "learning_rate": 3.717948717948718e-05,
      "loss": 0.848,
      "step": 22
    },
    {
      "epoch": 2.75,
      "eval_accuracy": 0.5806451612903226,
      "eval_f1": 0.3924349881796691,
      "eval_loss": 0.9473089575767517,
      "eval_runtime": 1.927,
      "eval_samples_per_second": 32.175,
      "eval_steps_per_second": 2.076,
      "step": 22
    },
    {
      "epoch": 2.875,
      "grad_norm": 10.20969295501709,
      "learning_rate": 3.653846153846154e-05,
      "loss": 0.6632,
      "step": 23
    },
    {
      "epoch": 2.875,
      "eval_accuracy": 0.6451612903225806,
      "eval_f1": 0.4470854555744199,
      "eval_loss": 0.9435995817184448,
      "eval_runtime": 1.9274,
      "eval_samples_per_second": 32.168,
      "eval_steps_per_second": 2.075,
      "step": 23
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.576776027679443,
      "learning_rate": 3.58974358974359e-05,
      "loss": 0.7259,
      "step": 24
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.47088750706613897,
      "eval_loss": 0.9455369710922241,
      "eval_runtime": 1.9275,
      "eval_samples_per_second": 32.166,
      "eval_steps_per_second": 2.075,
      "step": 24
    },
    {
      "epoch": 3.125,
      "grad_norm": 8.878484725952148,
      "learning_rate": 3.525641025641026e-05,
      "loss": 0.7974,
      "step": 25
    },
    {
      "epoch": 3.125,
      "eval_accuracy": 0.6935483870967742,
      "eval_f1": 0.48200757575757575,
      "eval_loss": 0.9411699771881104,
      "eval_runtime": 1.9284,
      "eval_samples_per_second": 32.151,
      "eval_steps_per_second": 2.074,
      "step": 25
    },
    {
      "epoch": 3.25,
      "grad_norm": 7.3488450050354,
      "learning_rate": 3.461538461538462e-05,
      "loss": 0.7728,
      "step": 26
    },
    {
      "epoch": 3.25,
      "eval_accuracy": 0.6612903225806451,
      "eval_f1": 0.45350669818754924,
      "eval_loss": 0.9431821703910828,
      "eval_runtime": 1.928,
      "eval_samples_per_second": 32.157,
      "eval_steps_per_second": 2.075,
      "step": 26
    },
    {
      "epoch": 3.375,
      "grad_norm": 4.182816028594971,
      "learning_rate": 3.397435897435898e-05,
      "loss": 0.9005,
      "step": 27
    },
    {
      "epoch": 3.375,
      "eval_accuracy": 0.5806451612903226,
      "eval_f1": 0.3848238482384824,
      "eval_loss": 0.9425324201583862,
      "eval_runtime": 1.9307,
      "eval_samples_per_second": 32.112,
      "eval_steps_per_second": 2.072,
      "step": 27
    },
    {
      "epoch": 3.5,
      "grad_norm": 6.473806858062744,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.8324,
      "step": 28
    },
    {
      "epoch": 3.5,
      "eval_accuracy": 0.5483870967741935,
      "eval_f1": 0.35341880341880344,
      "eval_loss": 0.9440051913261414,
      "eval_runtime": 1.9291,
      "eval_samples_per_second": 32.139,
      "eval_steps_per_second": 2.073,
      "step": 28
    },
    {
      "epoch": 3.625,
      "grad_norm": 7.45881986618042,
      "learning_rate": 3.269230769230769e-05,
      "loss": 0.7577,
      "step": 29
    },
    {
      "epoch": 3.625,
      "eval_accuracy": 0.532258064516129,
      "eval_f1": 0.33679878709118477,
      "eval_loss": 0.9446666836738586,
      "eval_runtime": 1.929,
      "eval_samples_per_second": 32.142,
      "eval_steps_per_second": 2.074,
      "step": 29
    },
    {
      "epoch": 3.75,
      "grad_norm": 5.7591633796691895,
      "learning_rate": 3.205128205128206e-05,
      "loss": 0.7887,
      "step": 30
    },
    {
      "epoch": 3.75,
      "eval_accuracy": 0.532258064516129,
      "eval_f1": 0.33679878709118477,
      "eval_loss": 0.9500220417976379,
      "eval_runtime": 1.9285,
      "eval_samples_per_second": 32.149,
      "eval_steps_per_second": 2.074,
      "step": 30
    },
    {
      "epoch": 3.875,
      "grad_norm": 11.719704627990723,
      "learning_rate": 3.141025641025641e-05,
      "loss": 0.6706,
      "step": 31
    },
    {
      "epoch": 3.875,
      "eval_accuracy": 0.5806451612903226,
      "eval_f1": 0.38984088127294986,
      "eval_loss": 0.9507701992988586,
      "eval_runtime": 1.9302,
      "eval_samples_per_second": 32.121,
      "eval_steps_per_second": 2.072,
      "step": 31
    },
    {
      "epoch": 4.0,
      "grad_norm": 6.659142017364502,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 0.7348,
      "step": 32
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4682125603864735,
      "eval_loss": 0.9422016739845276,
      "eval_runtime": 1.9309,
      "eval_samples_per_second": 32.109,
      "eval_steps_per_second": 2.072,
      "step": 32
    },
    {
      "epoch": 4.125,
      "grad_norm": 8.99232006072998,
      "learning_rate": 3.012820512820513e-05,
      "loss": 0.797,
      "step": 33
    },
    {
      "epoch": 4.125,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.46957520091848454,
      "eval_loss": 0.9406993985176086,
      "eval_runtime": 1.8808,
      "eval_samples_per_second": 32.965,
      "eval_steps_per_second": 2.127,
      "step": 33
    },
    {
      "epoch": 4.25,
      "grad_norm": 15.037948608398438,
      "learning_rate": 2.948717948717949e-05,
      "loss": 0.7472,
      "step": 34
    },
    {
      "epoch": 4.25,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4708994708994709,
      "eval_loss": 0.9398488402366638,
      "eval_runtime": 1.9301,
      "eval_samples_per_second": 32.122,
      "eval_steps_per_second": 2.072,
      "step": 34
    },
    {
      "epoch": 4.375,
      "grad_norm": 9.858637809753418,
      "learning_rate": 2.8846153846153845e-05,
      "loss": 0.7184,
      "step": 35
    },
    {
      "epoch": 4.375,
      "eval_accuracy": 0.6612903225806451,
      "eval_f1": 0.45972458026787394,
      "eval_loss": 0.9320343732833862,
      "eval_runtime": 1.9306,
      "eval_samples_per_second": 32.115,
      "eval_steps_per_second": 2.072,
      "step": 35
    },
    {
      "epoch": 4.5,
      "grad_norm": 9.198653221130371,
      "learning_rate": 2.8205128205128207e-05,
      "loss": 0.8042,
      "step": 36
    },
    {
      "epoch": 4.5,
      "eval_accuracy": 0.6612903225806451,
      "eval_f1": 0.45972458026787394,
      "eval_loss": 0.9294925928115845,
      "eval_runtime": 1.9308,
      "eval_samples_per_second": 32.111,
      "eval_steps_per_second": 2.072,
      "step": 36
    },
    {
      "epoch": 4.625,
      "grad_norm": 5.223246097564697,
      "learning_rate": 2.756410256410257e-05,
      "loss": 0.7408,
      "step": 37
    },
    {
      "epoch": 4.625,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4708994708994709,
      "eval_loss": 0.9262655973434448,
      "eval_runtime": 1.9307,
      "eval_samples_per_second": 32.113,
      "eval_steps_per_second": 2.072,
      "step": 37
    },
    {
      "epoch": 4.75,
      "grad_norm": 6.609065532684326,
      "learning_rate": 2.6923076923076923e-05,
      "loss": 0.5936,
      "step": 38
    },
    {
      "epoch": 4.75,
      "eval_accuracy": 0.6935483870967742,
      "eval_f1": 0.48192090395480225,
      "eval_loss": 0.9254209399223328,
      "eval_runtime": 1.9339,
      "eval_samples_per_second": 32.059,
      "eval_steps_per_second": 2.068,
      "step": 38
    },
    {
      "epoch": 4.875,
      "grad_norm": 6.2056803703308105,
      "learning_rate": 2.6282051282051285e-05,
      "loss": 0.5937,
      "step": 39
    },
    {
      "epoch": 4.875,
      "eval_accuracy": 0.6935483870967742,
      "eval_f1": 0.48192090395480225,
      "eval_loss": 0.9212784171104431,
      "eval_runtime": 1.9303,
      "eval_samples_per_second": 32.12,
      "eval_steps_per_second": 2.072,
      "step": 39
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.507566452026367,
      "learning_rate": 2.564102564102564e-05,
      "loss": 0.7497,
      "step": 40
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9186185002326965,
      "eval_runtime": 1.9335,
      "eval_samples_per_second": 32.066,
      "eval_steps_per_second": 2.069,
      "step": 40
    },
    {
      "epoch": 5.125,
      "grad_norm": 13.483494758605957,
      "learning_rate": 2.5e-05,
      "loss": 1.135,
      "step": 41
    },
    {
      "epoch": 5.125,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9170433878898621,
      "eval_runtime": 1.9361,
      "eval_samples_per_second": 32.023,
      "eval_steps_per_second": 2.066,
      "step": 41
    },
    {
      "epoch": 5.25,
      "grad_norm": 8.313968658447266,
      "learning_rate": 2.435897435897436e-05,
      "loss": 0.6677,
      "step": 42
    },
    {
      "epoch": 5.25,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9127059578895569,
      "eval_runtime": 1.9338,
      "eval_samples_per_second": 32.061,
      "eval_steps_per_second": 2.068,
      "step": 42
    },
    {
      "epoch": 5.375,
      "grad_norm": 7.541636943817139,
      "learning_rate": 2.3717948717948718e-05,
      "loss": 0.7515,
      "step": 43
    },
    {
      "epoch": 5.375,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9163720011711121,
      "eval_runtime": 1.9345,
      "eval_samples_per_second": 32.049,
      "eval_steps_per_second": 2.068,
      "step": 43
    },
    {
      "epoch": 5.5,
      "grad_norm": 9.568167686462402,
      "learning_rate": 2.307692307692308e-05,
      "loss": 0.6488,
      "step": 44
    },
    {
      "epoch": 5.5,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9121409058570862,
      "eval_runtime": 1.8843,
      "eval_samples_per_second": 32.904,
      "eval_steps_per_second": 2.123,
      "step": 44
    },
    {
      "epoch": 5.625,
      "grad_norm": 3.9321157932281494,
      "learning_rate": 2.2435897435897437e-05,
      "loss": 0.7113,
      "step": 45
    },
    {
      "epoch": 5.625,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9119479060173035,
      "eval_runtime": 1.9363,
      "eval_samples_per_second": 32.02,
      "eval_steps_per_second": 2.066,
      "step": 45
    },
    {
      "epoch": 5.75,
      "grad_norm": 4.554017543792725,
      "learning_rate": 2.1794871794871795e-05,
      "loss": 0.6476,
      "step": 46
    },
    {
      "epoch": 5.75,
      "eval_accuracy": 0.6935483870967742,
      "eval_f1": 0.48192090395480225,
      "eval_loss": 0.9131627082824707,
      "eval_runtime": 1.8815,
      "eval_samples_per_second": 32.953,
      "eval_steps_per_second": 2.126,
      "step": 46
    },
    {
      "epoch": 5.875,
      "grad_norm": 4.490871429443359,
      "learning_rate": 2.1153846153846154e-05,
      "loss": 0.5748,
      "step": 47
    },
    {
      "epoch": 5.875,
      "eval_accuracy": 0.6935483870967742,
      "eval_f1": 0.48192090395480225,
      "eval_loss": 0.9163798689842224,
      "eval_runtime": 1.8811,
      "eval_samples_per_second": 32.959,
      "eval_steps_per_second": 2.126,
      "step": 47
    },
    {
      "epoch": 6.0,
      "grad_norm": 4.915560722351074,
      "learning_rate": 2.0512820512820512e-05,
      "loss": 0.7456,
      "step": 48
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6935483870967742,
      "eval_f1": 0.48192090395480225,
      "eval_loss": 0.9137986898422241,
      "eval_runtime": 1.9333,
      "eval_samples_per_second": 32.069,
      "eval_steps_per_second": 2.069,
      "step": 48
    },
    {
      "epoch": 6.125,
      "grad_norm": 5.666485786437988,
      "learning_rate": 1.987179487179487e-05,
      "loss": 0.7035,
      "step": 49
    },
    {
      "epoch": 6.125,
      "eval_accuracy": 0.6935483870967742,
      "eval_f1": 0.48192090395480225,
      "eval_loss": 0.9176655411720276,
      "eval_runtime": 1.8845,
      "eval_samples_per_second": 32.9,
      "eval_steps_per_second": 2.123,
      "step": 49
    },
    {
      "epoch": 6.25,
      "grad_norm": 4.38871431350708,
      "learning_rate": 1.923076923076923e-05,
      "loss": 0.6157,
      "step": 50
    },
    {
      "epoch": 6.25,
      "eval_accuracy": 0.6935483870967742,
      "eval_f1": 0.48192090395480225,
      "eval_loss": 0.9223140478134155,
      "eval_runtime": 1.8794,
      "eval_samples_per_second": 32.99,
      "eval_steps_per_second": 2.128,
      "step": 50
    },
    {
      "epoch": 6.375,
      "grad_norm": 11.442863464355469,
      "learning_rate": 1.858974358974359e-05,
      "loss": 0.561,
      "step": 51
    },
    {
      "epoch": 6.375,
      "eval_accuracy": 0.6935483870967742,
      "eval_f1": 0.48192090395480225,
      "eval_loss": 0.9202437996864319,
      "eval_runtime": 1.9299,
      "eval_samples_per_second": 32.125,
      "eval_steps_per_second": 2.073,
      "step": 51
    },
    {
      "epoch": 6.5,
      "grad_norm": 7.4318928718566895,
      "learning_rate": 1.794871794871795e-05,
      "loss": 0.7139,
      "step": 52
    },
    {
      "epoch": 6.5,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9179559350013733,
      "eval_runtime": 1.8777,
      "eval_samples_per_second": 33.02,
      "eval_steps_per_second": 2.13,
      "step": 52
    },
    {
      "epoch": 6.625,
      "grad_norm": 5.853055477142334,
      "learning_rate": 1.730769230769231e-05,
      "loss": 0.6748,
      "step": 53
    },
    {
      "epoch": 6.625,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9139246940612793,
      "eval_runtime": 1.9266,
      "eval_samples_per_second": 32.182,
      "eval_steps_per_second": 2.076,
      "step": 53
    },
    {
      "epoch": 6.75,
      "grad_norm": 5.003907680511475,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.5338,
      "step": 54
    },
    {
      "epoch": 6.75,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9166889786720276,
      "eval_runtime": 1.9286,
      "eval_samples_per_second": 32.148,
      "eval_steps_per_second": 2.074,
      "step": 54
    },
    {
      "epoch": 6.875,
      "grad_norm": 4.901527404785156,
      "learning_rate": 1.602564102564103e-05,
      "loss": 0.6889,
      "step": 55
    },
    {
      "epoch": 6.875,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9157655835151672,
      "eval_runtime": 1.9276,
      "eval_samples_per_second": 32.164,
      "eval_steps_per_second": 2.075,
      "step": 55
    },
    {
      "epoch": 7.0,
      "grad_norm": 5.2783918380737305,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.8556,
      "step": 56
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.921044647693634,
      "eval_runtime": 1.9306,
      "eval_samples_per_second": 32.115,
      "eval_steps_per_second": 2.072,
      "step": 56
    },
    {
      "epoch": 7.125,
      "grad_norm": 10.69826602935791,
      "learning_rate": 1.4743589743589745e-05,
      "loss": 0.8448,
      "step": 57
    },
    {
      "epoch": 7.125,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.916961669921875,
      "eval_runtime": 1.9308,
      "eval_samples_per_second": 32.111,
      "eval_steps_per_second": 2.072,
      "step": 57
    },
    {
      "epoch": 7.25,
      "grad_norm": 6.701515197753906,
      "learning_rate": 1.4102564102564104e-05,
      "loss": 0.6535,
      "step": 58
    },
    {
      "epoch": 7.25,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9177826642990112,
      "eval_runtime": 1.9295,
      "eval_samples_per_second": 32.133,
      "eval_steps_per_second": 2.073,
      "step": 58
    },
    {
      "epoch": 7.375,
      "grad_norm": 5.344858646392822,
      "learning_rate": 1.3461538461538462e-05,
      "loss": 0.6185,
      "step": 59
    },
    {
      "epoch": 7.375,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9164468050003052,
      "eval_runtime": 1.9345,
      "eval_samples_per_second": 32.049,
      "eval_steps_per_second": 2.068,
      "step": 59
    },
    {
      "epoch": 7.5,
      "grad_norm": 5.061964511871338,
      "learning_rate": 1.282051282051282e-05,
      "loss": 0.7128,
      "step": 60
    },
    {
      "epoch": 7.5,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9154215455055237,
      "eval_runtime": 1.9313,
      "eval_samples_per_second": 32.103,
      "eval_steps_per_second": 2.071,
      "step": 60
    },
    {
      "epoch": 7.625,
      "grad_norm": 6.167932033538818,
      "learning_rate": 1.217948717948718e-05,
      "loss": 0.5839,
      "step": 61
    },
    {
      "epoch": 7.625,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9095370173454285,
      "eval_runtime": 1.9307,
      "eval_samples_per_second": 32.112,
      "eval_steps_per_second": 2.072,
      "step": 61
    },
    {
      "epoch": 7.75,
      "grad_norm": 2.6592514514923096,
      "learning_rate": 1.153846153846154e-05,
      "loss": 0.5118,
      "step": 62
    },
    {
      "epoch": 7.75,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9109383821487427,
      "eval_runtime": 1.9289,
      "eval_samples_per_second": 32.143,
      "eval_steps_per_second": 2.074,
      "step": 62
    },
    {
      "epoch": 7.875,
      "grad_norm": 10.835193634033203,
      "learning_rate": 1.0897435897435898e-05,
      "loss": 0.6146,
      "step": 63
    },
    {
      "epoch": 7.875,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9067746996879578,
      "eval_runtime": 1.9278,
      "eval_samples_per_second": 32.161,
      "eval_steps_per_second": 2.075,
      "step": 63
    },
    {
      "epoch": 8.0,
      "grad_norm": 11.179938316345215,
      "learning_rate": 1.0256410256410256e-05,
      "loss": 0.9724,
      "step": 64
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9018973112106323,
      "eval_runtime": 1.9274,
      "eval_samples_per_second": 32.168,
      "eval_steps_per_second": 2.075,
      "step": 64
    },
    {
      "epoch": 8.125,
      "grad_norm": 6.455613136291504,
      "learning_rate": 9.615384615384616e-06,
      "loss": 0.6459,
      "step": 65
    },
    {
      "epoch": 8.125,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9056165218353271,
      "eval_runtime": 1.9292,
      "eval_samples_per_second": 32.138,
      "eval_steps_per_second": 2.073,
      "step": 65
    },
    {
      "epoch": 8.25,
      "grad_norm": 7.206265449523926,
      "learning_rate": 8.974358974358976e-06,
      "loss": 0.7838,
      "step": 66
    },
    {
      "epoch": 8.25,
      "eval_accuracy": 0.6612903225806451,
      "eval_f1": 0.4595103578154425,
      "eval_loss": 0.909091591835022,
      "eval_runtime": 1.9295,
      "eval_samples_per_second": 32.132,
      "eval_steps_per_second": 2.073,
      "step": 66
    },
    {
      "epoch": 8.375,
      "grad_norm": 5.377885341644287,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.7472,
      "step": 67
    },
    {
      "epoch": 8.375,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9005382657051086,
      "eval_runtime": 1.8794,
      "eval_samples_per_second": 32.99,
      "eval_steps_per_second": 2.128,
      "step": 67
    },
    {
      "epoch": 8.5,
      "grad_norm": 5.025684356689453,
      "learning_rate": 7.692307692307694e-06,
      "loss": 0.4137,
      "step": 68
    },
    {
      "epoch": 8.5,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9060821533203125,
      "eval_runtime": 1.8832,
      "eval_samples_per_second": 32.923,
      "eval_steps_per_second": 2.124,
      "step": 68
    },
    {
      "epoch": 8.625,
      "grad_norm": 6.750956058502197,
      "learning_rate": 7.051282051282052e-06,
      "loss": 0.5145,
      "step": 69
    },
    {
      "epoch": 8.625,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.8989046216011047,
      "eval_runtime": 1.9282,
      "eval_samples_per_second": 32.155,
      "eval_steps_per_second": 2.075,
      "step": 69
    },
    {
      "epoch": 8.75,
      "grad_norm": 5.707816123962402,
      "learning_rate": 6.41025641025641e-06,
      "loss": 0.7129,
      "step": 70
    },
    {
      "epoch": 8.75,
      "eval_accuracy": 0.6612903225806451,
      "eval_f1": 0.4595103578154425,
      "eval_loss": 0.8961211442947388,
      "eval_runtime": 1.929,
      "eval_samples_per_second": 32.14,
      "eval_steps_per_second": 2.074,
      "step": 70
    },
    {
      "epoch": 8.875,
      "grad_norm": 3.8523380756378174,
      "learning_rate": 5.76923076923077e-06,
      "loss": 0.554,
      "step": 71
    },
    {
      "epoch": 8.875,
      "eval_accuracy": 0.6612903225806451,
      "eval_f1": 0.4595103578154425,
      "eval_loss": 0.8937899470329285,
      "eval_runtime": 1.9304,
      "eval_samples_per_second": 32.117,
      "eval_steps_per_second": 2.072,
      "step": 71
    },
    {
      "epoch": 9.0,
      "grad_norm": 7.517840385437012,
      "learning_rate": 5.128205128205128e-06,
      "loss": 0.7014,
      "step": 72
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.6612903225806451,
      "eval_f1": 0.4595103578154425,
      "eval_loss": 0.9007149934768677,
      "eval_runtime": 1.9291,
      "eval_samples_per_second": 32.139,
      "eval_steps_per_second": 2.073,
      "step": 72
    },
    {
      "epoch": 9.125,
      "grad_norm": 7.3192362785339355,
      "learning_rate": 4.487179487179488e-06,
      "loss": 0.7076,
      "step": 73
    },
    {
      "epoch": 9.125,
      "eval_accuracy": 0.6612903225806451,
      "eval_f1": 0.4595103578154425,
      "eval_loss": 0.8964553475379944,
      "eval_runtime": 1.9292,
      "eval_samples_per_second": 32.137,
      "eval_steps_per_second": 2.073,
      "step": 73
    },
    {
      "epoch": 9.25,
      "grad_norm": 6.544466495513916,
      "learning_rate": 3.846153846153847e-06,
      "loss": 0.6754,
      "step": 74
    },
    {
      "epoch": 9.25,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.9004920125007629,
      "eval_runtime": 1.9283,
      "eval_samples_per_second": 32.153,
      "eval_steps_per_second": 2.074,
      "step": 74
    },
    {
      "epoch": 9.375,
      "grad_norm": 5.344063758850098,
      "learning_rate": 3.205128205128205e-06,
      "loss": 0.591,
      "step": 75
    },
    {
      "epoch": 9.375,
      "eval_accuracy": 0.6612903225806451,
      "eval_f1": 0.4595103578154425,
      "eval_loss": 0.8975746631622314,
      "eval_runtime": 1.8785,
      "eval_samples_per_second": 33.006,
      "eval_steps_per_second": 2.129,
      "step": 75
    },
    {
      "epoch": 9.5,
      "grad_norm": 7.665821075439453,
      "learning_rate": 2.564102564102564e-06,
      "loss": 0.6753,
      "step": 76
    },
    {
      "epoch": 9.5,
      "eval_accuracy": 0.6612903225806451,
      "eval_f1": 0.4595103578154425,
      "eval_loss": 0.8977981209754944,
      "eval_runtime": 1.9303,
      "eval_samples_per_second": 32.119,
      "eval_steps_per_second": 2.072,
      "step": 76
    },
    {
      "epoch": 9.625,
      "grad_norm": 6.320114612579346,
      "learning_rate": 1.9230769230769234e-06,
      "loss": 0.6099,
      "step": 77
    },
    {
      "epoch": 9.625,
      "eval_accuracy": 0.6612903225806451,
      "eval_f1": 0.4595103578154425,
      "eval_loss": 0.8969810009002686,
      "eval_runtime": 1.9299,
      "eval_samples_per_second": 32.126,
      "eval_steps_per_second": 2.073,
      "step": 77
    },
    {
      "epoch": 9.75,
      "grad_norm": 7.137789249420166,
      "learning_rate": 1.282051282051282e-06,
      "loss": 0.7863,
      "step": 78
    },
    {
      "epoch": 9.75,
      "eval_accuracy": 0.6774193548387096,
      "eval_f1": 0.4704331450094162,
      "eval_loss": 0.8956599235534668,
      "eval_runtime": 1.9278,
      "eval_samples_per_second": 32.161,
      "eval_steps_per_second": 2.075,
      "step": 78
    },
    {
      "epoch": 9.875,
      "grad_norm": 5.855066299438477,
      "learning_rate": 6.41025641025641e-07,
      "loss": 0.5154,
      "step": 79
    },
    {
      "epoch": 9.875,
      "eval_accuracy": 0.6612903225806451,
      "eval_f1": 0.4595103578154425,
      "eval_loss": 0.8908558487892151,
      "eval_runtime": 1.8786,
      "eval_samples_per_second": 33.003,
      "eval_steps_per_second": 2.129,
      "step": 79
    },
    {
      "epoch": 10.0,
      "grad_norm": 6.291125774383545,
      "learning_rate": 0.0,
      "loss": 0.6067,
      "step": 80
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.6612903225806451,
      "eval_f1": 0.4595103578154425,
      "eval_loss": 0.8924511075019836,
      "eval_runtime": 1.9287,
      "eval_samples_per_second": 32.145,
      "eval_steps_per_second": 2.074,
      "step": 80
    },
    {
      "epoch": 10.0,
      "step": 80,
      "total_flos": 1.4174270632493056e+16,
      "train_loss": 0.7646934121847153,
      "train_runtime": 412.0561,
      "train_samples_per_second": 5.922,
      "train_steps_per_second": 0.194
    }
  ],
  "logging_steps": 1,
  "max_steps": 80,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 1.4174270632493056e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}