{
  "best_metric": 0.9961777329444885,
  "best_model_checkpoint": "Action_agent/checkpoint-1300",
  "epoch": 10.0,
  "eval_steps": 100,
  "global_step": 1340,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.11,
      "grad_norm": 1.2644492387771606,
      "learning_rate": 9.888059701492538e-06,
      "loss": 2.3069,
      "step": 15
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.0855854749679565,
      "learning_rate": 9.776119402985076e-06,
      "loss": 2.2883,
      "step": 30
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.0014735460281372,
      "learning_rate": 9.664179104477612e-06,
      "loss": 2.2674,
      "step": 45
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.1784591674804688,
      "learning_rate": 9.552238805970149e-06,
      "loss": 2.2511,
      "step": 60
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.3332306146621704,
      "learning_rate": 9.440298507462688e-06,
      "loss": 2.2346,
      "step": 75
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.2581124305725098,
      "learning_rate": 9.328358208955226e-06,
      "loss": 2.1982,
      "step": 90
    },
    {
      "epoch": 0.75,
      "eval_accuracy": 0.4850615114235501,
      "eval_classification_report": "              precision    recall  f1-score   support\n\n           0     0.3333    0.0385    0.0690        52\n           1     0.6420    0.8667    0.7376        60\n           2     0.3261    0.2941    0.3093        51\n           3     0.6667    0.2182    0.3288        55\n           4     0.4839    0.5357    0.5085        56\n           5     0.6667    0.2857    0.4000        56\n           6     0.5000    0.4286    0.4615        63\n           7     0.4060    0.9643    0.5714        56\n           8     0.4559    0.5167    0.4844        60\n           9     0.4805    0.6167    0.5401        60\n\n    accuracy                         0.4851       569\n   macro avg     0.4961    0.4765    0.4411       569\nweighted avg     0.4991    0.4851    0.4484       569\n",
      "eval_confusion_matrix": [
        [
          2,
          3,
          2,
          1,
          3,
          1,
          7,
          15,
          10,
          8
        ],
        [
          1,
          52,
          0,
          0,
          2,
          0,
          0,
          2,
          2,
          1
        ],
        [
          1,
          0,
          15,
          0,
          5,
          0,
          3,
          23,
          3,
          1
        ],
        [
          2,
          1,
          8,
          12,
          5,
          0,
          6,
          6,
          1,
          14
        ],
        [
          0,
          2,
          9,
          1,
          30,
          2,
          2,
          3,
          2,
          5
        ],
        [
          0,
          2,
          6,
          2,
          5,
          16,
          2,
          16,
          4,
          3
        ],
        [
          0,
          7,
          0,
          1,
          5,
          2,
          27,
          1,
          12,
          8
        ],
        [
          0,
          0,
          1,
          0,
          0,
          0,
          1,
          54,
          0,
          0
        ],
        [
          0,
          11,
          1,
          0,
          3,
          2,
          5,
          7,
          31,
          0
        ],
        [
          0,
          3,
          4,
          1,
          4,
          1,
          1,
          6,
          3,
          37
        ]
      ],
      "eval_loss": 2.158277988433838,
      "eval_runtime": 14.259,
      "eval_samples_per_second": 39.905,
      "eval_steps_per_second": 5.049,
      "step": 100
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.18204927444458,
      "learning_rate": 9.216417910447763e-06,
      "loss": 2.1795,
      "step": 105
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.2949470281600952,
      "learning_rate": 9.104477611940299e-06,
      "loss": 2.1617,
      "step": 120
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.331998586654663,
      "learning_rate": 8.992537313432836e-06,
      "loss": 2.1171,
      "step": 135
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.3272285461425781,
      "learning_rate": 8.880597014925374e-06,
      "loss": 2.1022,
      "step": 150
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.6193878650665283,
      "learning_rate": 8.768656716417911e-06,
      "loss": 2.0411,
      "step": 165
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.4247101545333862,
      "learning_rate": 8.656716417910447e-06,
      "loss": 1.9998,
      "step": 180
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.438057541847229,
      "learning_rate": 8.544776119402986e-06,
      "loss": 1.988,
      "step": 195
    },
    {
      "epoch": 1.49,
      "eval_accuracy": 0.6256590509666081,
      "eval_classification_report": "              precision    recall  f1-score   support\n\n           0     0.5789    0.2115    0.3099        52\n           1     0.6667    0.9667    0.7891        60\n           2     0.4524    0.3725    0.4086        51\n           3     0.8000    0.2909    0.4267        55\n           4     0.6494    0.8929    0.7519        56\n           5     0.8621    0.4464    0.5882        56\n           6     0.6909    0.6032    0.6441        63\n           7     0.4701    0.9821    0.6358        56\n           8     0.7234    0.5667    0.6355        60\n           9     0.6579    0.8333    0.7353        60\n\n    accuracy                         0.6257       569\n   macro avg     0.6552    0.6166    0.5925       569\nweighted avg     0.6583    0.6257    0.5997       569\n",
      "eval_confusion_matrix": [
        [
          11,
          6,
          2,
          0,
          7,
          1,
          3,
          10,
          7,
          5
        ],
        [
          0,
          58,
          0,
          0,
          1,
          0,
          0,
          0,
          1,
          0
        ],
        [
          1,
          1,
          19,
          0,
          4,
          1,
          1,
          24,
          0,
          0
        ],
        [
          1,
          1,
          5,
          16,
          3,
          0,
          6,
          7,
          0,
          16
        ],
        [
          1,
          1,
          1,
          0,
          50,
          0,
          2,
          0,
          0,
          1
        ],
        [
          1,
          0,
          11,
          0,
          6,
          25,
          0,
          11,
          0,
          2
        ],
        [
          2,
          8,
          1,
          1,
          3,
          1,
          38,
          2,
          5,
          2
        ],
        [
          0,
          0,
          1,
          0,
          0,
          0,
          0,
          55,
          0,
          0
        ],
        [
          1,
          12,
          0,
          0,
          1,
          1,
          5,
          6,
          34,
          0
        ],
        [
          1,
          0,
          2,
          3,
          2,
          0,
          0,
          2,
          0,
          50
        ]
      ],
      "eval_loss": 1.9349839687347412,
      "eval_runtime": 7.8827,
      "eval_samples_per_second": 72.184,
      "eval_steps_per_second": 9.134,
      "step": 200
    },
    {
      "epoch": 1.57,
      "grad_norm": 1.5711833238601685,
      "learning_rate": 8.432835820895524e-06,
      "loss": 1.9794,
      "step": 210
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.5019315481185913,
      "learning_rate": 8.320895522388061e-06,
      "loss": 1.9272,
      "step": 225
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.6470316648483276,
      "learning_rate": 8.208955223880599e-06,
      "loss": 1.8712,
      "step": 240
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.5267831087112427,
      "learning_rate": 8.097014925373135e-06,
      "loss": 1.8437,
      "step": 255
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.6073447465896606,
      "learning_rate": 7.985074626865672e-06,
      "loss": 1.8157,
      "step": 270
    },
    {
      "epoch": 2.13,
      "grad_norm": 1.7841838598251343,
      "learning_rate": 7.87313432835821e-06,
      "loss": 1.7875,
      "step": 285
    },
    {
      "epoch": 2.24,
      "grad_norm": 2.123765707015991,
      "learning_rate": 7.761194029850747e-06,
      "loss": 1.7347,
      "step": 300
    },
    {
      "epoch": 2.24,
      "eval_accuracy": 0.7223198594024605,
      "eval_classification_report": "              precision    recall  f1-score   support\n\n           0     0.6087    0.5385    0.5714        52\n           1     0.7160    0.9667    0.8227        60\n           2     0.6667    0.5490    0.6022        51\n           3     0.8788    0.5273    0.6591        55\n           4     0.8167    0.8750    0.8448        56\n           5     0.9459    0.6250    0.7527        56\n           6     0.7600    0.6032    0.6726        63\n           7     0.6437    1.0000    0.7832        56\n           8     0.6731    0.5833    0.6250        60\n           9     0.6790    0.9167    0.7801        60\n\n    accuracy                         0.7223       569\n   macro avg     0.7389    0.7185    0.7114       569\nweighted avg     0.7394    0.7223    0.7136       569\n",
      "eval_confusion_matrix": [
        [
          28,
          4,
          2,
          1,
          4,
          1,
          1,
          1,
          6,
          4
        ],
        [
          0,
          58,
          0,
          0,
          0,
          0,
          1,
          0,
          1,
          0
        ],
        [
          3,
          0,
          28,
          0,
          1,
          1,
          1,
          16,
          0,
          1
        ],
        [
          2,
          2,
          2,
          29,
          1,
          0,
          2,
          2,
          0,
          15
        ],
        [
          2,
          1,
          1,
          0,
          49,
          0,
          1,
          0,
          0,
          2
        ],
        [
          1,
          0,
          6,
          0,
          3,
          35,
          1,
          8,
          0,
          2
        ],
        [
          4,
          5,
          1,
          1,
          1,
          0,
          38,
          1,
          10,
          2
        ],
        [
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          56,
          0,
          0
        ],
        [
          6,
          11,
          0,
          0,
          1,
          0,
          5,
          2,
          35,
          0
        ],
        [
          0,
          0,
          2,
          2,
          0,
          0,
          0,
          1,
          0,
          55
        ]
      ],
      "eval_loss": 1.693671464920044,
      "eval_runtime": 7.858,
      "eval_samples_per_second": 72.411,
      "eval_steps_per_second": 9.163,
      "step": 300
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.736540675163269,
      "learning_rate": 7.649253731343284e-06,
      "loss": 1.6738,
      "step": 315
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.6560536623001099,
      "learning_rate": 7.537313432835821e-06,
      "loss": 1.685,
      "step": 330
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.7724891901016235,
      "learning_rate": 7.4253731343283585e-06,
      "loss": 1.6482,
      "step": 345
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.9107571840286255,
      "learning_rate": 7.313432835820896e-06,
      "loss": 1.6371,
      "step": 360
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.8733954429626465,
      "learning_rate": 7.2014925373134335e-06,
      "loss": 1.5774,
      "step": 375
    },
    {
      "epoch": 2.91,
      "grad_norm": 1.8130593299865723,
      "learning_rate": 7.089552238805971e-06,
      "loss": 1.5713,
      "step": 390
    },
    {
      "epoch": 2.99,
      "eval_accuracy": 0.7434094903339191,
      "eval_classification_report": "              precision    recall  f1-score   support\n\n           0     0.6500    0.5000    0.5652        52\n           1     0.7037    0.9500    0.8085        60\n           2     0.6444    0.5686    0.6042        51\n           3     0.9143    0.5818    0.7111        55\n           4     0.7903    0.8750    0.8305        56\n           5     0.9111    0.7321    0.8119        56\n           6     0.7778    0.6667    0.7179        63\n           7     0.7237    0.9821    0.8333        56\n           8     0.7143    0.5833    0.6422        60\n           9     0.6951    0.9500    0.8028        60\n\n    accuracy                         0.7434       569\n   macro avg     0.7525    0.7390    0.7328       569\nweighted avg     0.7532    0.7434    0.7353       569\n",
      "eval_confusion_matrix": [
        [
          26,
          6,
          2,
          1,
          5,
          1,
          0,
          2,
          5,
          4
        ],
        [
          0,
          57,
          0,
          0,
          0,
          0,
          1,
          0,
          1,
          1
        ],
        [
          2,
          0,
          29,
          1,
          2,
          2,
          2,
          13,
          0,
          0
        ],
        [
          3,
          1,
          4,
          32,
          1,
          1,
          0,
          1,
          0,
          12
        ],
        [
          1,
          1,
          1,
          0,
          49,
          0,
          1,
          0,
          0,
          3
        ],
        [
          1,
          0,
          6,
          0,
          4,
          41,
          0,
          2,
          0,
          2
        ],
        [
          3,
          5,
          1,
          0,
          1,
          0,
          42,
          0,
          8,
          3
        ],
        [
          0,
          0,
          0,
          1,
          0,
          0,
          0,
          55,
          0,
          0
        ],
        [
          4,
          11,
          0,
          0,
          0,
          0,
          8,
          2,
          35,
          0
        ],
        [
          0,
          0,
          2,
          0,
          0,
          0,
          0,
          1,
          0,
          57
        ]
      ],
      "eval_loss": 1.485671043395996,
      "eval_runtime": 8.1555,
      "eval_samples_per_second": 69.769,
      "eval_steps_per_second": 8.828,
      "step": 400
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.5968806743621826,
      "learning_rate": 6.9776119402985076e-06,
      "loss": 1.5017,
      "step": 405
    },
    {
      "epoch": 3.13,
      "grad_norm": 1.815771460533142,
      "learning_rate": 6.865671641791045e-06,
      "loss": 1.5338,
      "step": 420
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.734825849533081,
      "learning_rate": 6.7537313432835825e-06,
      "loss": 1.4918,
      "step": 435
    },
    {
      "epoch": 3.36,
      "grad_norm": 2.4256784915924072,
      "learning_rate": 6.64179104477612e-06,
      "loss": 1.467,
      "step": 450
    },
    {
      "epoch": 3.47,
      "grad_norm": 2.3504505157470703,
      "learning_rate": 6.5298507462686575e-06,
      "loss": 1.4156,
      "step": 465
    },
    {
      "epoch": 3.58,
      "grad_norm": 2.3143842220306396,
      "learning_rate": 6.417910447761194e-06,
      "loss": 1.4313,
      "step": 480
    },
    {
      "epoch": 3.69,
      "grad_norm": 2.124675989151001,
      "learning_rate": 6.3059701492537316e-06,
      "loss": 1.3821,
      "step": 495
    },
    {
      "epoch": 3.73,
      "eval_accuracy": 0.7574692442882249,
      "eval_classification_report": "              precision    recall  f1-score   support\n\n           0     0.6818    0.5769    0.6250        52\n           1     0.7600    0.9500    0.8444        60\n           2     0.7143    0.5882    0.6452        51\n           3     0.7500    0.4909    0.5934        55\n           4     0.8750    0.8750    0.8750        56\n           5     0.9167    0.7857    0.8462        56\n           6     0.7778    0.7778    0.7778        63\n           7     0.7465    0.9464    0.8346        56\n           8     0.8095    0.5667    0.6667        60\n           9     0.6304    0.9667    0.7632        60\n\n    accuracy                         0.7575       569\n   macro avg     0.7662    0.7524    0.7471       569\nweighted avg     0.7667    0.7575    0.7498       569\n",
      "eval_confusion_matrix": [
        [
          30,
          4,
          0,
          3,
          4,
          1,
          0,
          2,
          4,
          4
        ],
        [
          0,
          57,
          0,
          0,
          0,
          0,
          1,
          0,
          1,
          1
        ],
        [
          2,
          0,
          30,
          4,
          1,
          2,
          1,
          10,
          0,
          1
        ],
        [
          3,
          2,
          2,
          27,
          0,
          1,
          0,
          2,
          0,
          18
        ],
        [
          1,
          1,
          1,
          0,
          49,
          0,
          1,
          0,
          0,
          3
        ],
        [
          1,
          0,
          5,
          0,
          1,
          44,
          1,
          1,
          0,
          3
        ],
        [
          4,
          0,
          1,
          1,
          1,
          0,
          49,
          0,
          3,
          4
        ],
        [
          0,
          0,
          2,
          1,
          0,
          0,
          0,
          53,
          0,
          0
        ],
        [
          3,
          11,
          0,
          0,
          0,
          0,
          10,
          2,
          34,
          0
        ],
        [
          0,
          0,
          1,
          0,
          0,
          0,
          0,
          1,
          0,
          58
        ]
      ],
      "eval_loss": 1.347652554512024,
      "eval_runtime": 7.8905,
      "eval_samples_per_second": 72.112,
      "eval_steps_per_second": 9.125,
      "step": 500
    },
    {
      "epoch": 3.81,
      "grad_norm": 2.028282880783081,
      "learning_rate": 6.194029850746269e-06,
      "loss": 1.3733,
      "step": 510
    },
    {
      "epoch": 3.92,
      "grad_norm": 2.1104109287261963,
      "learning_rate": 6.0820895522388065e-06,
      "loss": 1.3978,
      "step": 525
    },
    {
      "epoch": 4.03,
      "grad_norm": 2.8287768363952637,
      "learning_rate": 5.970149253731343e-06,
      "loss": 1.3507,
      "step": 540
    },
    {
      "epoch": 4.14,
      "grad_norm": 2.149505853652954,
      "learning_rate": 5.858208955223881e-06,
      "loss": 1.3286,
      "step": 555
    },
    {
      "epoch": 4.25,
      "grad_norm": 1.904295563697815,
      "learning_rate": 5.746268656716418e-06,
      "loss": 1.3065,
      "step": 570
    },
    {
      "epoch": 4.37,
      "grad_norm": 2.056488513946533,
      "learning_rate": 5.6343283582089556e-06,
      "loss": 1.3238,
      "step": 585
    },
    {
      "epoch": 4.48,
      "grad_norm": 2.633546829223633,
      "learning_rate": 5.522388059701493e-06,
      "loss": 1.3065,
      "step": 600
    },
    {
      "epoch": 4.48,
      "eval_accuracy": 0.7855887521968365,
      "eval_classification_report": "              precision    recall  f1-score   support\n\n           0     0.7333    0.6346    0.6804        52\n           1     0.7887    0.9333    0.8550        60\n           2     0.8286    0.5686    0.6744        51\n           3     0.7660    0.6545    0.7059        55\n           4     0.8929    0.8929    0.8929        56\n           5     0.8571    0.7500    0.8000        56\n           6     0.8154    0.8413    0.8281        63\n           7     0.7143    0.9821    0.8271        56\n           8     0.8043    0.6167    0.6981        60\n           9     0.7179    0.9333    0.8116        60\n\n    accuracy                         0.7856       569\n   macro avg     0.7919    0.7807    0.7773       569\nweighted avg     0.7918    0.7856    0.7799       569\n",
      "eval_confusion_matrix": [
        [
          33,
          4,
          0,
          1,
          3,
          1,
          0,
          2,
          4,
          4
        ],
        [
          0,
          56,
          0,
          0,
          0,
          0,
          1,
          0,
          2,
          1
        ],
        [
          1,
          0,
          29,
          5,
          1,
          2,
          1,
          12,
          0,
          0
        ],
        [
          2,
          1,
          1,
          36,
          0,
          3,
          0,
          2,
          0,
          10
        ],
        [
          1,
          1,
          1,
          1,
          50,
          0,
          0,
          0,
          0,
          2
        ],
        [
          1,
          0,
          4,
          1,
          1,
          42,
          1,
          4,
          0,
          2
        ],
        [
          3,
          0,
          0,
          0,
          1,
          0,
          53,
          0,
          3,
          3
        ],
        [
          0,
          0,
          0,
          1,
          0,
          0,
          0,
          55,
          0,
          0
        ],
        [
          4,
          9,
          0,
          0,
          0,
          0,
          9,
          1,
          37,
          0
        ],
        [
          0,
          0,
          0,
          2,
          0,
          1,
          0,
          1,
          0,
          56
        ]
      ],
      "eval_loss": 1.2436854839324951,
      "eval_runtime": 7.8957,
      "eval_samples_per_second": 72.064,
      "eval_steps_per_second": 9.119,
      "step": 600
    },
    {
      "epoch": 4.59,
      "grad_norm": 2.528501272201538,
      "learning_rate": 5.41044776119403e-06,
      "loss": 1.2892,
      "step": 615
    },
    {
      "epoch": 4.7,
      "grad_norm": 1.980857014656067,
      "learning_rate": 5.298507462686567e-06,
      "loss": 1.261,
      "step": 630
    },
    {
      "epoch": 4.81,
      "grad_norm": 2.3212943077087402,
      "learning_rate": 5.186567164179105e-06,
      "loss": 1.2991,
      "step": 645
    },
    {
      "epoch": 4.93,
      "grad_norm": 2.138998031616211,
      "learning_rate": 5.074626865671642e-06,
      "loss": 1.271,
      "step": 660
    },
    {
      "epoch": 5.04,
      "grad_norm": 2.3111581802368164,
      "learning_rate": 4.9626865671641796e-06,
      "loss": 1.2146,
      "step": 675
    },
    {
      "epoch": 5.15,
      "grad_norm": 2.4816582202911377,
      "learning_rate": 4.850746268656717e-06,
      "loss": 1.2329,
      "step": 690
    },
    {
      "epoch": 5.22,
      "eval_accuracy": 0.7908611599297012,
      "eval_classification_report": "              precision    recall  f1-score   support\n\n           0     0.7727    0.6538    0.7083        52\n           1     0.7703    0.9500    0.8507        60\n           2     0.7500    0.6471    0.6947        51\n           3     0.7561    0.5636    0.6458        55\n           4     0.8621    0.8929    0.8772        56\n           5     0.8600    0.7679    0.8113        56\n           6     0.8235    0.8889    0.8550        63\n           7     0.8281    0.9464    0.8833        56\n           8     0.8571    0.6000    0.7059        60\n           9     0.6786    0.9500    0.7917        60\n\n    accuracy                         0.7909       569\n   macro avg     0.7959    0.7861    0.7824       569\nweighted avg     0.7963    0.7909    0.7848       569\n",
      "eval_confusion_matrix": [
        [
          34,
          4,
          0,
          1,
          3,
          1,
          0,
          1,
          4,
          4
        ],
        [
          0,
          57,
          0,
          0,
          0,
          0,
          1,
          0,
          1,
          1
        ],
        [
          1,
          0,
          33,
          5,
          1,
          3,
          1,
          7,
          0,
          0
        ],
        [
          3,
          1,
          1,
          31,
          1,
          2,
          0,
          1,
          0,
          15
        ],
        [
          1,
          1,
          1,
          1,
          50,
          0,
          0,
          0,
          0,
          2
        ],
        [
          1,
          0,
          7,
          1,
          2,
          43,
          0,
          0,
          0,
          2
        ],
        [
          2,
          0,
          0,
          0,
          1,
          0,
          56,
          0,
          1,
          3
        ],
        [
          0,
          0,
          2,
          1,
          0,
          0,
          0,
          53,
          0,
          0
        ],
        [
          2,
          11,
          0,
          0,
          0,
          0,
          10,
          1,
          36,
          0
        ],
        [
          0,
          0,
          0,
          1,
          0,
          1,
          0,
          1,
          0,
          57
        ]
      ],
      "eval_loss": 1.1645153760910034,
      "eval_runtime": 7.9519,
      "eval_samples_per_second": 71.555,
      "eval_steps_per_second": 9.054,
      "step": 700
    },
    {
      "epoch": 5.26,
      "grad_norm": 2.4168813228607178,
      "learning_rate": 4.738805970149254e-06,
      "loss": 1.2091,
      "step": 705
    },
    {
      "epoch": 5.37,
      "grad_norm": 2.246443033218384,
      "learning_rate": 4.626865671641791e-06,
      "loss": 1.221,
      "step": 720
    },
    {
      "epoch": 5.49,
      "grad_norm": 2.505394458770752,
      "learning_rate": 4.514925373134329e-06,
      "loss": 1.2529,
      "step": 735
    },
    {
      "epoch": 5.6,
      "grad_norm": 2.8784241676330566,
      "learning_rate": 4.402985074626866e-06,
      "loss": 1.1616,
      "step": 750
    },
    {
      "epoch": 5.71,
      "grad_norm": 2.4807000160217285,
      "learning_rate": 4.2910447761194036e-06,
      "loss": 1.1851,
      "step": 765
    },
    {
      "epoch": 5.82,
      "grad_norm": 2.8072807788848877,
      "learning_rate": 4.17910447761194e-06,
      "loss": 1.1799,
      "step": 780
    },
    {
      "epoch": 5.93,
      "grad_norm": 2.5767550468444824,
      "learning_rate": 4.067164179104478e-06,
      "loss": 1.1736,
      "step": 795
    },
    {
      "epoch": 5.97,
      "eval_accuracy": 0.789103690685413,
      "eval_classification_report": "              precision    recall  f1-score   support\n\n           0     0.7143    0.6731    0.6931        52\n           1     0.7808    0.9500    0.8571        60\n           2     0.7447    0.6863    0.7143        51\n           3     0.8095    0.6182    0.7010        55\n           4     0.9074    0.8750    0.8909        56\n           5     0.8600    0.7679    0.8113        56\n           6     0.8095    0.8095    0.8095        63\n           7     0.8254    0.9286    0.8739        56\n           8     0.8043    0.6167    0.6981        60\n           9     0.6829    0.9333    0.7887        60\n\n    accuracy                         0.7891       569\n   macro avg     0.7939    0.7858    0.7838       569\nweighted avg     0.7942    0.7891    0.7855       569\n",
      "eval_confusion_matrix": [
        [
          35,
          4,
          0,
          0,
          2,
          1,
          1,
          1,
          4,
          4
        ],
        [
          0,
          57,
          0,
          0,
          0,
          0,
          1,
          0,
          1,
          1
        ],
        [
          2,
          0,
          35,
          2,
          1,
          3,
          1,
          7,
          0,
          0
        ],
        [
          3,
          1,
          0,
          34,
          0,
          3,
          0,
          1,
          0,
          13
        ],
        [
          1,
          1,
          2,
          1,
          49,
          0,
          0,
          0,
          0,
          2
        ],
        [
          1,
          0,
          7,
          1,
          1,
          43,
          1,
          0,
          0,
          2
        ],
        [
          3,
          0,
          0,
          0,
          1,
          0,
          51,
          0,
          4,
          4
        ],
        [
          0,
          0,
          3,
          1,
          0,
          0,
          0,
          52,
          0,
          0
        ],
        [
          4,
          10,
          0,
          0,
          0,
          0,
          8,
          1,
          37,
          0
        ],
        [
          0,
          0,
          0,
          3,
          0,
          0,
          0,
          1,
          0,
          56
        ]
      ],
      "eval_loss": 1.1158992052078247,
      "eval_runtime": 7.7475,
      "eval_samples_per_second": 73.443,
      "eval_steps_per_second": 9.293,
      "step": 800
    },
    {
      "epoch": 6.04,
      "grad_norm": 2.618880271911621,
      "learning_rate": 3.955223880597015e-06,
      "loss": 1.1539,
      "step": 810
    },
    {
      "epoch": 6.16,
      "grad_norm": 3.0116899013519287,
      "learning_rate": 3.843283582089553e-06,
      "loss": 1.1117,
      "step": 825
    },
    {
      "epoch": 6.27,
      "grad_norm": 2.360898971557617,
      "learning_rate": 3.73134328358209e-06,
      "loss": 1.122,
      "step": 840
    },
    {
      "epoch": 6.38,
      "grad_norm": 2.760197162628174,
      "learning_rate": 3.619402985074627e-06,
      "loss": 1.1269,
      "step": 855
    },
    {
      "epoch": 6.49,
      "grad_norm": 2.6149590015411377,
      "learning_rate": 3.5074626865671646e-06,
      "loss": 1.1327,
      "step": 870
    },
    {
      "epoch": 6.6,
      "grad_norm": 2.9123787879943848,
      "learning_rate": 3.3955223880597017e-06,
      "loss": 1.1544,
      "step": 885
    },
    {
      "epoch": 6.72,
      "grad_norm": 2.19488263130188,
      "learning_rate": 3.283582089552239e-06,
      "loss": 1.1396,
      "step": 900
    },
    {
      "epoch": 6.72,
      "eval_accuracy": 0.8066783831282952,
      "eval_classification_report": "              precision    recall  f1-score   support\n\n           0     0.7222    0.7500    0.7358        52\n           1     0.8000    0.9333    0.8615        60\n           2     0.7451    0.7451    0.7451        51\n           3     0.8684    0.6000    0.7097        55\n           4     0.9259    0.8929    0.9091        56\n           5     0.8462    0.7857    0.8148        56\n           6     0.8413    0.8413    0.8413        63\n           7     0.8254    0.9286    0.8739        56\n           8     0.8605    0.6167    0.7184        60\n           9     0.7037    0.9500    0.8085        60\n\n    accuracy                         0.8067       569\n   macro avg     0.8139    0.8044    0.8018       569\nweighted avg     0.8148    0.8067    0.8033       569\n",
      "eval_confusion_matrix": [
        [
          39,
          3,
          0,
          0,
          1,
          1,
          0,
          2,
          3,
          3
        ],
        [
          1,
          56,
          0,
          0,
          0,
          0,
          1,
          0,
          1,
          1
        ],
        [
          2,
          0,
          38,
          1,
          1,
          3,
          0,
          6,
          0,
          0
        ],
        [
          3,
          1,
          1,
          33,
          0,
          3,
          0,
          1,
          0,
          13
        ],
        [
          1,
          1,
          2,
          1,
          50,
          0,
          0,
          0,
          0,
          1
        ],
        [
          0,
          0,
          7,
          1,
          1,
          44,
          1,
          0,
          0,
          2
        ],
        [
          3,
          0,
          0,
          0,
          1,
          0,
          53,
          0,
          2,
          4
        ],
        [
          0,
          0,
          3,
          1,
          0,
          0,
          0,
          52,
          0,
          0
        ],
        [
          5,
          9,
          0,
          0,
          0,
          0,
          8,
          1,
          37,
          0
        ],
        [
          0,
          0,
          0,
          1,
          0,
          1,
          0,
          1,
          0,
          57
        ]
      ],
      "eval_loss": 1.0748533010482788,
      "eval_runtime": 8.0459,
      "eval_samples_per_second": 70.72,
      "eval_steps_per_second": 8.949,
      "step": 900
    },
    {
      "epoch": 6.83,
      "grad_norm": 2.578913688659668,
      "learning_rate": 3.1716417910447766e-06,
      "loss": 1.1102,
      "step": 915
    },
    {
      "epoch": 6.94,
      "grad_norm": 2.287675619125366,
      "learning_rate": 3.0597014925373137e-06,
      "loss": 1.0931,
      "step": 930
    },
    {
      "epoch": 7.05,
      "grad_norm": 2.465820074081421,
      "learning_rate": 2.947761194029851e-06,
      "loss": 1.0911,
      "step": 945
    },
    {
      "epoch": 7.16,
      "grad_norm": 2.1009232997894287,
      "learning_rate": 2.835820895522388e-06,
      "loss": 1.0785,
      "step": 960
    },
    {
      "epoch": 7.28,
      "grad_norm": 2.86567759513855,
      "learning_rate": 2.7238805970149257e-06,
      "loss": 1.0793,
      "step": 975
    },
    {
      "epoch": 7.39,
      "grad_norm": 2.230086088180542,
      "learning_rate": 2.6119402985074627e-06,
      "loss": 1.0577,
      "step": 990
    },
    {
      "epoch": 7.46,
      "eval_accuracy": 0.8154657293497364,
      "eval_classification_report": "              precision    recall  f1-score   support\n\n           0     0.7708    0.7115    0.7400        52\n           1     0.8028    0.9500    0.8702        60\n           2     0.7755    0.7451    0.7600        51\n           3     0.7547    0.7273    0.7407        55\n           4     0.9259    0.8929    0.9091        56\n           5     0.8333    0.8036    0.8182        56\n           6     0.8154    0.8413    0.8281        63\n           7     0.8667    0.9286    0.8966        56\n           8     0.8444    0.6333    0.7238        60\n           9     0.7714    0.9000    0.8308        60\n\n    accuracy                         0.8155       569\n   macro avg     0.8161    0.8134    0.8117       569\nweighted avg     0.8167    0.8155    0.8130       569\n",
      "eval_confusion_matrix": [
        [
          37,
          3,
          0,
          0,
          1,
          1,
          1,
          2,
          4,
          3
        ],
        [
          0,
          57,
          0,
          0,
          0,
          0,
          1,
          0,
          1,
          1
        ],
        [
          1,
          0,
          38,
          4,
          1,
          4,
          0,
          3,
          0,
          0
        ],
        [
          3,
          1,
          0,
          40,
          0,
          3,
          0,
          1,
          0,
          7
        ],
        [
          1,
          1,
          2,
          1,
          50,
          0,
          0,
          0,
          0,
          1
        ],
        [
          0,
          0,
          6,
          1,
          1,
          45,
          1,
          0,
          0,
          2
        ],
        [
          3,
          0,
          0,
          2,
          1,
          0,
          53,
          0,
          2,
          2
        ],
        [
          0,
          0,
          3,
          1,
          0,
          0,
          0,
          52,
          0,
          0
        ],
        [
          3,
          9,
          0,
          0,
          0,
          0,
          9,
          1,
          38,
          0
        ],
        [
          0,
          0,
          0,
          4,
          0,
          1,
          0,
          1,
          0,
          54
        ]
      ],
      "eval_loss": 1.039870023727417,
      "eval_runtime": 7.9559,
      "eval_samples_per_second": 71.52,
      "eval_steps_per_second": 9.05,
      "step": 1000
    },
    {
      "epoch": 7.5,
      "grad_norm": 3.3555350303649902,
      "learning_rate": 2.5e-06,
      "loss": 1.0715,
      "step": 1005
    },
    {
      "epoch": 7.61,
      "grad_norm": 2.3127570152282715,
      "learning_rate": 2.3880597014925373e-06,
      "loss": 1.0708,
      "step": 1020
    },
    {
      "epoch": 7.72,
      "grad_norm": 2.815685987472534,
      "learning_rate": 2.2761194029850747e-06,
      "loss": 1.0413,
      "step": 1035
    },
    {
      "epoch": 7.84,
      "grad_norm": 2.632782459259033,
      "learning_rate": 2.1641791044776118e-06,
      "loss": 1.1223,
      "step": 1050
    },
    {
      "epoch": 7.95,
      "grad_norm": 3.01796293258667,
      "learning_rate": 2.0522388059701497e-06,
      "loss": 1.1029,
      "step": 1065
    },
    {
      "epoch": 8.06,
      "grad_norm": 2.525937080383301,
      "learning_rate": 1.9402985074626867e-06,
      "loss": 1.0888,
      "step": 1080
    },
    {
      "epoch": 8.17,
      "grad_norm": 3.033353328704834,
      "learning_rate": 1.828358208955224e-06,
      "loss": 0.9935,
      "step": 1095
    },
    {
      "epoch": 8.21,
      "eval_accuracy": 0.8189806678383128,
      "eval_classification_report": "              precision    recall  f1-score   support\n\n           0     0.7917    0.7308    0.7600        52\n           1     0.7808    0.9500    0.8571        60\n           2     0.7755    0.7451    0.7600        51\n           3     0.8085    0.6909    0.7451        55\n           4     0.9091    0.8929    0.9009        56\n           5     0.8462    0.7857    0.8148        56\n           6     0.8438    0.8571    0.8504        63\n           7     0.8281    0.9464    0.8833        56\n           8     0.8636    0.6333    0.7308        60\n           9     0.7671    0.9333    0.8421        60\n\n    accuracy                         0.8190       569\n   macro avg     0.8214    0.8166    0.8145       569\nweighted avg     0.8220    0.8190    0.8158       569\n",
      "eval_confusion_matrix": [
        [
          38,
          4,
          0,
          0,
          1,
          1,
          0,
          2,
          3,
          3
        ],
        [
          0,
          57,
          0,
          0,
          0,
          0,
          1,
          0,
          1,
          1
        ],
        [
          1,
          0,
          38,
          2,
          1,
          3,
          0,
          6,
          0,
          0
        ],
        [
          3,
          1,
          0,
          38,
          0,
          3,
          0,
          1,
          0,
          9
        ],
        [
          1,
          1,
          2,
          1,
          50,
          0,
          0,
          0,
          0,
          1
        ],
        [
          0,
          0,
          7,
          1,
          2,
          44,
          0,
          0,
          0,
          2
        ],
        [
          3,
          0,
          0,
          2,
          1,
          0,
          54,
          0,
          2,
          1
        ],
        [
          0,
          0,
          2,
          1,
          0,
          0,
          0,
          53,
          0,
          0
        ],
        [
          2,
          10,
          0,
          0,
          0,
          0,
          9,
          1,
          38,
          0
        ],
        [
          0,
          0,
          0,
          2,
          0,
          1,
          0,
          1,
          0,
          56
        ]
      ],
      "eval_loss": 1.0204936265945435,
      "eval_runtime": 8.9672,
      "eval_samples_per_second": 63.453,
      "eval_steps_per_second": 8.029,
      "step": 1100
    },
    {
      "epoch": 8.28,
      "grad_norm": 2.5986831188201904,
      "learning_rate": 1.7164179104477613e-06,
      "loss": 1.0598,
      "step": 1110
    },
    {
      "epoch": 8.4,
      "grad_norm": 2.732192039489746,
      "learning_rate": 1.6044776119402985e-06,
      "loss": 1.0134,
      "step": 1125
    },
    {
      "epoch": 8.51,
      "grad_norm": 2.673107385635376,
      "learning_rate": 1.4925373134328358e-06,
      "loss": 1.0408,
      "step": 1140
    },
    {
      "epoch": 8.62,
      "grad_norm": 3.2620863914489746,
      "learning_rate": 1.3805970149253733e-06,
      "loss": 1.0704,
      "step": 1155
    },
    {
      "epoch": 8.73,
      "grad_norm": 3.8948822021484375,
      "learning_rate": 1.2686567164179105e-06,
      "loss": 1.0194,
      "step": 1170
    },
    {
      "epoch": 8.84,
      "grad_norm": 3.380185127258301,
      "learning_rate": 1.1567164179104478e-06,
      "loss": 1.0534,
      "step": 1185
    },
    {
      "epoch": 8.96,
      "grad_norm": 2.3721847534179688,
      "learning_rate": 1.044776119402985e-06,
      "loss": 1.1058,
      "step": 1200
    },
    {
      "epoch": 8.96,
      "eval_accuracy": 0.8224956063268892,
      "eval_classification_report": "              precision    recall  f1-score   support\n\n           0     0.7600    0.7308    0.7451        52\n           1     0.8028    0.9500    0.8702        60\n           2     0.7708    0.7255    0.7475        51\n           3     0.8298    0.7091    0.7647        55\n           4     0.9091    0.8929    0.9009        56\n           5     0.8182    0.8036    0.8108        56\n           6     0.8209    0.8730    0.8462        63\n           7     0.8814    0.9286    0.9043        56\n           8     0.8667    0.6500    0.7429        60\n           9     0.7778    0.9333    0.8485        60\n\n    accuracy                         0.8225       569\n   macro avg     0.8237    0.8197    0.8181       569\nweighted avg     0.8244    0.8225    0.8197       569\n",
      "eval_confusion_matrix": [
        [
          38,
          3,
          0,
          0,
          2,
          1,
          1,
          1,
          3,
          3
        ],
        [
          0,
          57,
          0,
          0,
          0,
          0,
          1,
          0,
          1,
          1
        ],
        [
          1,
          0,
          37,
          2,
          1,
          5,
          0,
          5,
          0,
          0
        ],
        [
          4,
          1,
          0,
          39,
          0,
          3,
          0,
          0,
          0,
          8
        ],
        [
          1,
          1,
          2,
          1,
          50,
          0,
          0,
          0,
          0,
          1
        ],
        [
          0,
          0,
          6,
          1,
          1,
          45,
          1,
          0,
          0,
          2
        ],
        [
          3,
          0,
          0,
          1,
          1,
          0,
          55,
          0,
          2,
          1
        ],
        [
          0,
          0,
          3,
          1,
          0,
          0,
          0,
          52,
          0,
          0
        ],
        [
          3,
          9,
          0,
          0,
          0,
          0,
          9,
          0,
          39,
          0
        ],
        [
          0,
          0,
          0,
          2,
          0,
          1,
          0,
          1,
          0,
          56
        ]
      ],
      "eval_loss": 1.0022475719451904,
      "eval_runtime": 7.8225,
      "eval_samples_per_second": 72.739,
      "eval_steps_per_second": 9.204,
      "step": 1200
    },
    {
      "epoch": 9.07,
      "grad_norm": 2.744582176208496,
      "learning_rate": 9.328358208955225e-07,
      "loss": 1.0503,
      "step": 1215
    },
    {
      "epoch": 9.18,
      "grad_norm": 3.565460443496704,
      "learning_rate": 8.208955223880598e-07,
      "loss": 1.052,
      "step": 1230
    },
    {
      "epoch": 9.29,
      "grad_norm": 3.0185256004333496,
      "learning_rate": 7.08955223880597e-07,
      "loss": 1.0378,
      "step": 1245
    },
    {
      "epoch": 9.4,
      "grad_norm": 2.446512460708618,
      "learning_rate": 5.970149253731343e-07,
      "loss": 1.0401,
      "step": 1260
    },
    {
      "epoch": 9.51,
      "grad_norm": 2.392070770263672,
      "learning_rate": 4.850746268656717e-07,
      "loss": 1.0048,
      "step": 1275
    },
    {
      "epoch": 9.63,
      "grad_norm": 2.743360996246338,
      "learning_rate": 3.7313432835820895e-07,
      "loss": 1.0422,
      "step": 1290
    },
    {
      "epoch": 9.7,
      "eval_accuracy": 0.8242530755711776,
      "eval_classification_report": "              precision    recall  f1-score   support\n\n           0     0.7800    0.7500    0.7647        52\n           1     0.8028    0.9500    0.8702        60\n           2     0.7600    0.7451    0.7525        51\n           3     0.8298    0.7091    0.7647        55\n           4     0.9091    0.8929    0.9009        56\n           5     0.8302    0.7857    0.8073        56\n           6     0.8333    0.8730    0.8527        63\n           7     0.8667    0.9286    0.8966        56\n           8     0.8667    0.6500    0.7429        60\n           9     0.7778    0.9333    0.8485        60\n\n    accuracy                         0.8243       569\n   macro avg     0.8256    0.8218    0.8201       569\nweighted avg     0.8264    0.8243    0.8216       569\n",
      "eval_confusion_matrix": [
        [
          39,
          3,
          0,
          0,
          2,
          1,
          0,
          1,
          3,
          3
        ],
        [
          0,
          57,
          0,
          0,
          0,
          0,
          1,
          0,
          1,
          1
        ],
        [
          1,
          0,
          38,
          2,
          1,
          4,
          0,
          5,
          0,
          0
        ],
        [
          4,
          1,
          0,
          39,
          0,
          3,
          0,
          0,
          0,
          8
        ],
        [
          1,
          1,
          2,
          1,
          50,
          0,
          0,
          0,
          0,
          1
        ],
        [
          0,
          0,
          7,
          1,
          1,
          44,
          1,
          0,
          0,
          2
        ],
        [
          3,
          0,
          0,
          1,
          1,
          0,
          55,
          0,
          2,
          1
        ],
        [
          0,
          0,
          3,
          1,
          0,
          0,
          0,
          52,
          0,
          0
        ],
        [
          2,
          9,
          0,
          0,
          0,
          0,
          9,
          1,
          39,
          0
        ],
        [
          0,
          0,
          0,
          2,
          0,
          1,
          0,
          1,
          0,
          56
        ]
      ],
      "eval_loss": 0.9961777329444885,
      "eval_runtime": 8.2421,
      "eval_samples_per_second": 69.036,
      "eval_steps_per_second": 8.736,
      "step": 1300
    },
    {
      "epoch": 9.74,
      "grad_norm": 3.270967960357666,
      "learning_rate": 2.6119402985074626e-07,
      "loss": 1.0001,
      "step": 1305
    },
    {
      "epoch": 9.85,
      "grad_norm": 2.724989175796509,
      "learning_rate": 1.4925373134328358e-07,
      "loss": 1.0597,
      "step": 1320
    },
    {
      "epoch": 9.96,
      "grad_norm": 3.0584731101989746,
      "learning_rate": 3.7313432835820895e-08,
      "loss": 1.0519,
      "step": 1335
    },
    {
      "epoch": 10.0,
      "step": 1340,
      "total_flos": 3.3230947683690086e+18,
      "train_loss": 1.4043623358456057,
      "train_runtime": 1212.668,
      "train_samples_per_second": 35.36,
      "train_steps_per_second": 1.105
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8242530755711776,
      "eval_classification_report": "              precision    recall  f1-score   support\n\n           0     0.7800    0.7500    0.7647        52\n           1     0.8028    0.9500    0.8702        60\n           2     0.7600    0.7451    0.7525        51\n           3     0.8298    0.7091    0.7647        55\n           4     0.9091    0.8929    0.9009        56\n           5     0.8302    0.7857    0.8073        56\n           6     0.8333    0.8730    0.8527        63\n           7     0.8667    0.9286    0.8966        56\n           8     0.8667    0.6500    0.7429        60\n           9     0.7778    0.9333    0.8485        60\n\n    accuracy                         0.8243       569\n   macro avg     0.8256    0.8218    0.8201       569\nweighted avg     0.8264    0.8243    0.8216       569\n",
      "eval_confusion_matrix": [
        [
          39,
          3,
          0,
          0,
          2,
          1,
          0,
          1,
          3,
          3
        ],
        [
          0,
          57,
          0,
          0,
          0,
          0,
          1,
          0,
          1,
          1
        ],
        [
          1,
          0,
          38,
          2,
          1,
          4,
          0,
          5,
          0,
          0
        ],
        [
          4,
          1,
          0,
          39,
          0,
          3,
          0,
          0,
          0,
          8
        ],
        [
          1,
          1,
          2,
          1,
          50,
          0,
          0,
          0,
          0,
          1
        ],
        [
          0,
          0,
          7,
          1,
          1,
          44,
          1,
          0,
          0,
          2
        ],
        [
          3,
          0,
          0,
          1,
          1,
          0,
          55,
          0,
          2,
          1
        ],
        [
          0,
          0,
          3,
          1,
          0,
          0,
          0,
          52,
          0,
          0
        ],
        [
          2,
          9,
          0,
          0,
          0,
          0,
          9,
          1,
          39,
          0
        ],
        [
          0,
          0,
          0,
          2,
          0,
          1,
          0,
          1,
          0,
          56
        ]
      ],
      "eval_loss": 0.9961777329444885,
      "eval_runtime": 7.9684,
      "eval_samples_per_second": 71.407,
      "eval_steps_per_second": 9.036,
      "step": 1340
    }
  ],
  "logging_steps": 15,
  "max_steps": 1340,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 100,
  "total_flos": 3.3230947683690086e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}