{
  "best_metric": 1.3938915799371898e-07,
  "best_model_checkpoint": "Models/t5-base-class-gen/checkpoint-11600",
  "epoch": 4.868913857677903,
  "eval_steps": 100,
  "global_step": 13000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03745318352059925,
      "grad_norm": 69268.25,
      "learning_rate": 3.970037453183521e-05,
      "loss": 0.5911,
      "step": 100
    },
    {
      "epoch": 0.03745318352059925,
      "eval_loss": 0.11300157010555267,
      "eval_runtime": 0.271,
      "eval_samples_per_second": 110.716,
      "eval_steps_per_second": 7.381,
      "step": 100
    },
    {
      "epoch": 0.0749063670411985,
      "grad_norm": 40263.34765625,
      "learning_rate": 3.9400749063670414e-05,
      "loss": 0.1556,
      "step": 200
    },
    {
      "epoch": 0.0749063670411985,
      "eval_loss": 0.044787291437387466,
      "eval_runtime": 0.2726,
      "eval_samples_per_second": 110.063,
      "eval_steps_per_second": 7.338,
      "step": 200
    },
    {
      "epoch": 0.11235955056179775,
      "grad_norm": 54568.01171875,
      "learning_rate": 3.910112359550562e-05,
      "loss": 0.0855,
      "step": 300
    },
    {
      "epoch": 0.11235955056179775,
      "eval_loss": 0.028027888387441635,
      "eval_runtime": 0.2713,
      "eval_samples_per_second": 110.579,
      "eval_steps_per_second": 7.372,
      "step": 300
    },
    {
      "epoch": 0.149812734082397,
      "grad_norm": 31384.4140625,
      "learning_rate": 3.880149812734083e-05,
      "loss": 0.0672,
      "step": 400
    },
    {
      "epoch": 0.149812734082397,
      "eval_loss": 0.02276856079697609,
      "eval_runtime": 0.2731,
      "eval_samples_per_second": 109.855,
      "eval_steps_per_second": 7.324,
      "step": 400
    },
    {
      "epoch": 0.18726591760299627,
      "grad_norm": 47479.01953125,
      "learning_rate": 3.850187265917604e-05,
      "loss": 0.0427,
      "step": 500
    },
    {
      "epoch": 0.18726591760299627,
      "eval_loss": 0.01850857213139534,
      "eval_runtime": 0.2701,
      "eval_samples_per_second": 111.089,
      "eval_steps_per_second": 7.406,
      "step": 500
    },
    {
      "epoch": 0.2247191011235955,
      "grad_norm": 39500.9140625,
      "learning_rate": 3.820224719101124e-05,
      "loss": 0.0385,
      "step": 600
    },
    {
      "epoch": 0.2247191011235955,
      "eval_loss": 0.00761110894382,
      "eval_runtime": 0.2742,
      "eval_samples_per_second": 109.412,
      "eval_steps_per_second": 7.294,
      "step": 600
    },
    {
      "epoch": 0.26217228464419473,
      "grad_norm": 75871.453125,
      "learning_rate": 3.790262172284645e-05,
      "loss": 0.0356,
      "step": 700
    },
    {
      "epoch": 0.26217228464419473,
      "eval_loss": 0.007164248265326023,
      "eval_runtime": 0.2703,
      "eval_samples_per_second": 110.979,
      "eval_steps_per_second": 7.399,
      "step": 700
    },
    {
      "epoch": 0.299625468164794,
      "grad_norm": 41190.5390625,
      "learning_rate": 3.7602996254681646e-05,
      "loss": 0.0268,
      "step": 800
    },
    {
      "epoch": 0.299625468164794,
      "eval_loss": 0.0059220679104328156,
      "eval_runtime": 0.2696,
      "eval_samples_per_second": 111.27,
      "eval_steps_per_second": 7.418,
      "step": 800
    },
    {
      "epoch": 0.33707865168539325,
      "grad_norm": 6835.3095703125,
      "learning_rate": 3.730337078651685e-05,
      "loss": 0.0276,
      "step": 900
    },
    {
      "epoch": 0.33707865168539325,
      "eval_loss": 0.004661493003368378,
      "eval_runtime": 0.2701,
      "eval_samples_per_second": 111.086,
      "eval_steps_per_second": 7.406,
      "step": 900
    },
    {
      "epoch": 0.37453183520599254,
      "grad_norm": 47556.41015625,
      "learning_rate": 3.7003745318352064e-05,
      "loss": 0.0234,
      "step": 1000
    },
    {
      "epoch": 0.37453183520599254,
      "eval_loss": 0.005712361074984074,
      "eval_runtime": 0.2702,
      "eval_samples_per_second": 111.024,
      "eval_steps_per_second": 7.402,
      "step": 1000
    },
    {
      "epoch": 0.41198501872659177,
      "grad_norm": 20349.455078125,
      "learning_rate": 3.670411985018727e-05,
      "loss": 0.0288,
      "step": 1100
    },
    {
      "epoch": 0.41198501872659177,
      "eval_loss": 0.0040772343054413795,
      "eval_runtime": 0.2695,
      "eval_samples_per_second": 111.298,
      "eval_steps_per_second": 7.42,
      "step": 1100
    },
    {
      "epoch": 0.449438202247191,
      "grad_norm": 37560.2734375,
      "learning_rate": 3.6404494382022475e-05,
      "loss": 0.0213,
      "step": 1200
    },
    {
      "epoch": 0.449438202247191,
      "eval_loss": 0.003308301791548729,
      "eval_runtime": 0.2703,
      "eval_samples_per_second": 110.979,
      "eval_steps_per_second": 7.399,
      "step": 1200
    },
    {
      "epoch": 0.4868913857677903,
      "grad_norm": 23133.609375,
      "learning_rate": 3.610486891385768e-05,
      "loss": 0.0194,
      "step": 1300
    },
    {
      "epoch": 0.4868913857677903,
      "eval_loss": 0.003378539811819792,
      "eval_runtime": 0.2712,
      "eval_samples_per_second": 110.612,
      "eval_steps_per_second": 7.374,
      "step": 1300
    },
    {
      "epoch": 0.5243445692883895,
      "grad_norm": 8233.9853515625,
      "learning_rate": 3.5805243445692885e-05,
      "loss": 0.02,
      "step": 1400
    },
    {
      "epoch": 0.5243445692883895,
      "eval_loss": 0.0034168362617492676,
      "eval_runtime": 0.2693,
      "eval_samples_per_second": 111.389,
      "eval_steps_per_second": 7.426,
      "step": 1400
    },
    {
      "epoch": 0.5617977528089888,
      "grad_norm": 10800.7392578125,
      "learning_rate": 3.550561797752809e-05,
      "loss": 0.0201,
      "step": 1500
    },
    {
      "epoch": 0.5617977528089888,
      "eval_loss": 0.004822230897843838,
      "eval_runtime": 0.2698,
      "eval_samples_per_second": 111.205,
      "eval_steps_per_second": 7.414,
      "step": 1500
    },
    {
      "epoch": 0.599250936329588,
      "grad_norm": 18013.958984375,
      "learning_rate": 3.52059925093633e-05,
      "loss": 0.0175,
      "step": 1600
    },
    {
      "epoch": 0.599250936329588,
      "eval_loss": 0.0028443075716495514,
      "eval_runtime": 0.2713,
      "eval_samples_per_second": 110.566,
      "eval_steps_per_second": 7.371,
      "step": 1600
    },
    {
      "epoch": 0.6367041198501873,
      "grad_norm": 4363.4677734375,
      "learning_rate": 3.490636704119851e-05,
      "loss": 0.0157,
      "step": 1700
    },
    {
      "epoch": 0.6367041198501873,
      "eval_loss": 0.003480424638837576,
      "eval_runtime": 0.2687,
      "eval_samples_per_second": 111.651,
      "eval_steps_per_second": 7.443,
      "step": 1700
    },
    {
      "epoch": 0.6741573033707865,
      "grad_norm": 20063.248046875,
      "learning_rate": 3.4606741573033714e-05,
      "loss": 0.0158,
      "step": 1800
    },
    {
      "epoch": 0.6741573033707865,
      "eval_loss": 0.0031679689418524504,
      "eval_runtime": 0.2697,
      "eval_samples_per_second": 111.228,
      "eval_steps_per_second": 7.415,
      "step": 1800
    },
    {
      "epoch": 0.7116104868913857,
      "grad_norm": 41422.4765625,
      "learning_rate": 3.430711610486892e-05,
      "loss": 0.0171,
      "step": 1900
    },
    {
      "epoch": 0.7116104868913857,
      "eval_loss": 0.0027744669932872057,
      "eval_runtime": 0.2732,
      "eval_samples_per_second": 109.799,
      "eval_steps_per_second": 7.32,
      "step": 1900
    },
    {
      "epoch": 0.7490636704119851,
      "grad_norm": 32689.869140625,
      "learning_rate": 3.4007490636704125e-05,
      "loss": 0.0157,
      "step": 2000
    },
    {
      "epoch": 0.7490636704119851,
      "eval_loss": 0.006064592860639095,
      "eval_runtime": 0.2699,
      "eval_samples_per_second": 111.173,
      "eval_steps_per_second": 7.412,
      "step": 2000
    },
    {
      "epoch": 0.7865168539325843,
      "grad_norm": 8271.3828125,
      "learning_rate": 3.370786516853933e-05,
      "loss": 0.012,
      "step": 2100
    },
    {
      "epoch": 0.7865168539325843,
      "eval_loss": 0.0077110580168664455,
      "eval_runtime": 0.2699,
      "eval_samples_per_second": 111.144,
      "eval_steps_per_second": 7.41,
      "step": 2100
    },
    {
      "epoch": 0.8239700374531835,
      "grad_norm": 81113.7265625,
      "learning_rate": 3.3408239700374535e-05,
      "loss": 0.0119,
      "step": 2200
    },
    {
      "epoch": 0.8239700374531835,
      "eval_loss": 0.0007900427444837987,
      "eval_runtime": 0.2804,
      "eval_samples_per_second": 106.996,
      "eval_steps_per_second": 7.133,
      "step": 2200
    },
    {
      "epoch": 0.8614232209737828,
      "grad_norm": 48094.25390625,
      "learning_rate": 3.310861423220974e-05,
      "loss": 0.0112,
      "step": 2300
    },
    {
      "epoch": 0.8614232209737828,
      "eval_loss": 0.0006070551462471485,
      "eval_runtime": 0.2712,
      "eval_samples_per_second": 110.615,
      "eval_steps_per_second": 7.374,
      "step": 2300
    },
    {
      "epoch": 0.898876404494382,
      "grad_norm": 13731.2001953125,
      "learning_rate": 3.2808988764044946e-05,
      "loss": 0.0168,
      "step": 2400
    },
    {
      "epoch": 0.898876404494382,
      "eval_loss": 0.0011749672703444958,
      "eval_runtime": 0.2697,
      "eval_samples_per_second": 111.246,
      "eval_steps_per_second": 7.416,
      "step": 2400
    },
    {
      "epoch": 0.9363295880149812,
      "grad_norm": 13338.21484375,
      "learning_rate": 3.250936329588015e-05,
      "loss": 0.014,
      "step": 2500
    },
    {
      "epoch": 0.9363295880149812,
      "eval_loss": 0.003979253116995096,
      "eval_runtime": 0.2689,
      "eval_samples_per_second": 111.55,
      "eval_steps_per_second": 7.437,
      "step": 2500
    },
    {
      "epoch": 0.9737827715355806,
      "grad_norm": 9933.720703125,
      "learning_rate": 3.220973782771536e-05,
      "loss": 0.0153,
      "step": 2600
    },
    {
      "epoch": 0.9737827715355806,
      "eval_loss": 0.0011219431180506945,
      "eval_runtime": 0.2694,
      "eval_samples_per_second": 111.371,
      "eval_steps_per_second": 7.425,
      "step": 2600
    },
    {
      "epoch": 1.0112359550561798,
      "grad_norm": 13659.6982421875,
      "learning_rate": 3.191011235955056e-05,
      "loss": 0.0107,
      "step": 2700
    },
    {
      "epoch": 1.0112359550561798,
      "eval_loss": 0.0038317707367241383,
      "eval_runtime": 0.2722,
      "eval_samples_per_second": 110.194,
      "eval_steps_per_second": 7.346,
      "step": 2700
    },
    {
      "epoch": 1.048689138576779,
      "grad_norm": 9676.2880859375,
      "learning_rate": 3.1610486891385774e-05,
      "loss": 0.0129,
      "step": 2800
    },
    {
      "epoch": 1.048689138576779,
      "eval_loss": 0.0013198582455515862,
      "eval_runtime": 0.269,
      "eval_samples_per_second": 111.536,
      "eval_steps_per_second": 7.436,
      "step": 2800
    },
    {
      "epoch": 1.0861423220973783,
      "grad_norm": 4065.046142578125,
      "learning_rate": 3.131086142322098e-05,
      "loss": 0.0114,
      "step": 2900
    },
    {
      "epoch": 1.0861423220973783,
      "eval_loss": 0.0012054108083248138,
      "eval_runtime": 0.27,
      "eval_samples_per_second": 111.106,
      "eval_steps_per_second": 7.407,
      "step": 2900
    },
    {
      "epoch": 1.1235955056179776,
      "grad_norm": 2907.2998046875,
      "learning_rate": 3.1011235955056185e-05,
      "loss": 0.0094,
      "step": 3000
    },
    {
      "epoch": 1.1235955056179776,
      "eval_loss": 0.001887617981992662,
      "eval_runtime": 0.2702,
      "eval_samples_per_second": 111.034,
      "eval_steps_per_second": 7.402,
      "step": 3000
    },
    {
      "epoch": 1.1610486891385767,
      "grad_norm": 8905.38671875,
      "learning_rate": 3.071161048689139e-05,
      "loss": 0.0116,
      "step": 3100
    },
    {
      "epoch": 1.1610486891385767,
      "eval_loss": 0.0023780567571520805,
      "eval_runtime": 0.2704,
      "eval_samples_per_second": 110.96,
      "eval_steps_per_second": 7.397,
      "step": 3100
    },
    {
      "epoch": 1.198501872659176,
      "grad_norm": 35338.76953125,
      "learning_rate": 3.0411985018726593e-05,
      "loss": 0.014,
      "step": 3200
    },
    {
      "epoch": 1.198501872659176,
      "eval_loss": 0.0016208798624575138,
      "eval_runtime": 0.2702,
      "eval_samples_per_second": 111.03,
      "eval_steps_per_second": 7.402,
      "step": 3200
    },
    {
      "epoch": 1.2359550561797752,
      "grad_norm": 1403.9569091796875,
      "learning_rate": 3.0112359550561798e-05,
      "loss": 0.0103,
      "step": 3300
    },
    {
      "epoch": 1.2359550561797752,
      "eval_loss": 0.0008052433258853853,
      "eval_runtime": 0.2843,
      "eval_samples_per_second": 105.529,
      "eval_steps_per_second": 7.035,
      "step": 3300
    },
    {
      "epoch": 1.2734082397003745,
      "grad_norm": 2095.459716796875,
      "learning_rate": 2.9812734082397007e-05,
      "loss": 0.0109,
      "step": 3400
    },
    {
      "epoch": 1.2734082397003745,
      "eval_loss": 0.0006734432536177337,
      "eval_runtime": 0.2718,
      "eval_samples_per_second": 110.379,
      "eval_steps_per_second": 7.359,
      "step": 3400
    },
    {
      "epoch": 1.3108614232209739,
      "grad_norm": 39442.0703125,
      "learning_rate": 2.9513108614232212e-05,
      "loss": 0.0121,
      "step": 3500
    },
    {
      "epoch": 1.3108614232209739,
      "eval_loss": 0.0014634514227509499,
      "eval_runtime": 0.2902,
      "eval_samples_per_second": 103.377,
      "eval_steps_per_second": 6.892,
      "step": 3500
    },
    {
      "epoch": 1.348314606741573,
      "grad_norm": 4135.88671875,
      "learning_rate": 2.9213483146067417e-05,
      "loss": 0.0089,
      "step": 3600
    },
    {
      "epoch": 1.348314606741573,
      "eval_loss": 0.001179067650809884,
      "eval_runtime": 0.2756,
      "eval_samples_per_second": 108.856,
      "eval_steps_per_second": 7.257,
      "step": 3600
    },
    {
      "epoch": 1.3857677902621723,
      "grad_norm": 8729.4716796875,
      "learning_rate": 2.8913857677902623e-05,
      "loss": 0.0101,
      "step": 3700
    },
    {
      "epoch": 1.3857677902621723,
      "eval_loss": 0.0001626112061785534,
      "eval_runtime": 0.27,
      "eval_samples_per_second": 111.124,
      "eval_steps_per_second": 7.408,
      "step": 3700
    },
    {
      "epoch": 1.4232209737827715,
      "grad_norm": 56410.66015625,
      "learning_rate": 2.8614232209737828e-05,
      "loss": 0.0114,
      "step": 3800
    },
    {
      "epoch": 1.4232209737827715,
      "eval_loss": 9.747984586283565e-05,
      "eval_runtime": 0.2718,
      "eval_samples_per_second": 110.357,
      "eval_steps_per_second": 7.357,
      "step": 3800
    },
    {
      "epoch": 1.4606741573033708,
      "grad_norm": 35145.53125,
      "learning_rate": 2.8314606741573034e-05,
      "loss": 0.0105,
      "step": 3900
    },
    {
      "epoch": 1.4606741573033708,
      "eval_loss": 0.00023929693270474672,
      "eval_runtime": 0.2713,
      "eval_samples_per_second": 110.563,
      "eval_steps_per_second": 7.371,
      "step": 3900
    },
    {
      "epoch": 1.4981273408239701,
      "grad_norm": 626.2529907226562,
      "learning_rate": 2.8014981273408242e-05,
      "loss": 0.0093,
      "step": 4000
    },
    {
      "epoch": 1.4981273408239701,
      "eval_loss": 0.00014960776024963707,
      "eval_runtime": 0.2703,
      "eval_samples_per_second": 110.999,
      "eval_steps_per_second": 7.4,
      "step": 4000
    },
    {
      "epoch": 1.5355805243445693,
      "grad_norm": 6510.17578125,
      "learning_rate": 2.7715355805243448e-05,
      "loss": 0.0093,
      "step": 4100
    },
    {
      "epoch": 1.5355805243445693,
      "eval_loss": 0.00028584664687514305,
      "eval_runtime": 0.2706,
      "eval_samples_per_second": 110.846,
      "eval_steps_per_second": 7.39,
      "step": 4100
    },
    {
      "epoch": 1.5730337078651684,
      "grad_norm": 576.2495727539062,
      "learning_rate": 2.7415730337078653e-05,
      "loss": 0.0089,
      "step": 4200
    },
    {
      "epoch": 1.5730337078651684,
      "eval_loss": 0.0004513164167292416,
      "eval_runtime": 0.271,
      "eval_samples_per_second": 110.71,
      "eval_steps_per_second": 7.381,
      "step": 4200
    },
    {
      "epoch": 1.6104868913857677,
      "grad_norm": 36006.58984375,
      "learning_rate": 2.711610486891386e-05,
      "loss": 0.0106,
      "step": 4300
    },
    {
      "epoch": 1.6104868913857677,
      "eval_loss": 0.002157117472961545,
      "eval_runtime": 0.2707,
      "eval_samples_per_second": 110.83,
      "eval_steps_per_second": 7.389,
      "step": 4300
    },
    {
      "epoch": 1.647940074906367,
      "grad_norm": 8769.625,
      "learning_rate": 2.6816479400749064e-05,
      "loss": 0.0115,
      "step": 4400
    },
    {
      "epoch": 1.647940074906367,
      "eval_loss": 0.0011888875160366297,
      "eval_runtime": 0.2704,
      "eval_samples_per_second": 110.955,
      "eval_steps_per_second": 7.397,
      "step": 4400
    },
    {
      "epoch": 1.6853932584269664,
      "grad_norm": 4818.69091796875,
      "learning_rate": 2.651685393258427e-05,
      "loss": 0.0094,
      "step": 4500
    },
    {
      "epoch": 1.6853932584269664,
      "eval_loss": 0.0001289191859541461,
      "eval_runtime": 0.2704,
      "eval_samples_per_second": 110.931,
      "eval_steps_per_second": 7.395,
      "step": 4500
    },
    {
      "epoch": 1.7228464419475655,
      "grad_norm": 31683.6875,
      "learning_rate": 2.6217228464419478e-05,
      "loss": 0.0101,
      "step": 4600
    },
    {
      "epoch": 1.7228464419475655,
      "eval_loss": 0.00011979354167124256,
      "eval_runtime": 0.2694,
      "eval_samples_per_second": 111.374,
      "eval_steps_per_second": 7.425,
      "step": 4600
    },
    {
      "epoch": 1.7602996254681647,
      "grad_norm": 7775.08984375,
      "learning_rate": 2.5917602996254683e-05,
      "loss": 0.0123,
      "step": 4700
    },
    {
      "epoch": 1.7602996254681647,
      "eval_loss": 0.00030495377723127604,
      "eval_runtime": 0.2695,
      "eval_samples_per_second": 111.3,
      "eval_steps_per_second": 7.42,
      "step": 4700
    },
    {
      "epoch": 1.797752808988764,
      "grad_norm": 101875.0234375,
      "learning_rate": 2.561797752808989e-05,
      "loss": 0.0114,
      "step": 4800
    },
    {
      "epoch": 1.797752808988764,
      "eval_loss": 0.00019709841581061482,
      "eval_runtime": 0.2754,
      "eval_samples_per_second": 108.946,
      "eval_steps_per_second": 7.263,
      "step": 4800
    },
    {
      "epoch": 1.8352059925093633,
      "grad_norm": 92537.5703125,
      "learning_rate": 2.5318352059925094e-05,
      "loss": 0.0093,
      "step": 4900
    },
    {
      "epoch": 1.8352059925093633,
      "eval_loss": 0.00014086294686421752,
      "eval_runtime": 0.2698,
      "eval_samples_per_second": 111.203,
      "eval_steps_per_second": 7.414,
      "step": 4900
    },
    {
      "epoch": 1.8726591760299627,
      "grad_norm": 4901.89892578125,
      "learning_rate": 2.50187265917603e-05,
      "loss": 0.0103,
      "step": 5000
    },
    {
      "epoch": 1.8726591760299627,
      "eval_loss": 0.00010910825221799314,
      "eval_runtime": 0.2701,
      "eval_samples_per_second": 111.059,
      "eval_steps_per_second": 7.404,
      "step": 5000
    },
    {
      "epoch": 1.9101123595505618,
      "grad_norm": 13060.30078125,
      "learning_rate": 2.4719101123595505e-05,
      "loss": 0.0099,
      "step": 5100
    },
    {
      "epoch": 1.9101123595505618,
      "eval_loss": 3.3620919566601515e-05,
      "eval_runtime": 0.272,
      "eval_samples_per_second": 110.304,
      "eval_steps_per_second": 7.354,
      "step": 5100
    },
    {
      "epoch": 1.947565543071161,
      "grad_norm": 168.86660766601562,
      "learning_rate": 2.4419475655430717e-05,
      "loss": 0.0099,
      "step": 5200
    },
    {
      "epoch": 1.947565543071161,
      "eval_loss": 0.00021029771596658975,
      "eval_runtime": 0.2705,
      "eval_samples_per_second": 110.885,
      "eval_steps_per_second": 7.392,
      "step": 5200
    },
    {
      "epoch": 1.9850187265917603,
      "grad_norm": 14831.890625,
      "learning_rate": 2.411985018726592e-05,
      "loss": 0.009,
      "step": 5300
    },
    {
      "epoch": 1.9850187265917603,
      "eval_loss": 1.2607515600393526e-05,
      "eval_runtime": 0.2708,
      "eval_samples_per_second": 110.79,
      "eval_steps_per_second": 7.386,
      "step": 5300
    },
    {
      "epoch": 2.0224719101123596,
      "grad_norm": 24069.68359375,
      "learning_rate": 2.3820224719101125e-05,
      "loss": 0.007,
      "step": 5400
    },
    {
      "epoch": 2.0224719101123596,
      "eval_loss": 0.00012969045201316476,
      "eval_runtime": 0.2727,
      "eval_samples_per_second": 110.006,
      "eval_steps_per_second": 7.334,
      "step": 5400
    },
    {
      "epoch": 2.059925093632959,
      "grad_norm": 6022.14599609375,
      "learning_rate": 2.352059925093633e-05,
      "loss": 0.0099,
      "step": 5500
    },
    {
      "epoch": 2.059925093632959,
      "eval_loss": 1.658714609220624e-05,
      "eval_runtime": 0.2735,
      "eval_samples_per_second": 109.696,
      "eval_steps_per_second": 7.313,
      "step": 5500
    },
    {
      "epoch": 2.097378277153558,
      "grad_norm": 14537.3671875,
      "learning_rate": 2.3220973782771535e-05,
      "loss": 0.0069,
      "step": 5600
    },
    {
      "epoch": 2.097378277153558,
      "eval_loss": 9.241083171218634e-05,
      "eval_runtime": 0.2711,
      "eval_samples_per_second": 110.646,
      "eval_steps_per_second": 7.376,
      "step": 5600
    },
    {
      "epoch": 2.134831460674157,
      "grad_norm": 3323.2705078125,
      "learning_rate": 2.292134831460674e-05,
      "loss": 0.008,
      "step": 5700
    },
    {
      "epoch": 2.134831460674157,
      "eval_loss": 0.001304484554566443,
      "eval_runtime": 0.2724,
      "eval_samples_per_second": 110.15,
      "eval_steps_per_second": 7.343,
      "step": 5700
    },
    {
      "epoch": 2.1722846441947565,
      "grad_norm": 30671.177734375,
      "learning_rate": 2.2621722846441953e-05,
      "loss": 0.0084,
      "step": 5800
    },
    {
      "epoch": 2.1722846441947565,
      "eval_loss": 0.00012851983774453402,
      "eval_runtime": 0.2707,
      "eval_samples_per_second": 110.817,
      "eval_steps_per_second": 7.388,
      "step": 5800
    },
    {
      "epoch": 2.209737827715356,
      "grad_norm": 21532.076171875,
      "learning_rate": 2.2322097378277158e-05,
      "loss": 0.0097,
      "step": 5900
    },
    {
      "epoch": 2.209737827715356,
      "eval_loss": 0.0001036034373100847,
      "eval_runtime": 0.2707,
      "eval_samples_per_second": 110.829,
      "eval_steps_per_second": 7.389,
      "step": 5900
    },
    {
      "epoch": 2.247191011235955,
      "grad_norm": 12783.5244140625,
      "learning_rate": 2.202247191011236e-05,
      "loss": 0.0092,
      "step": 6000
    },
    {
      "epoch": 2.247191011235955,
      "eval_loss": 0.0001524754916317761,
      "eval_runtime": 0.2696,
      "eval_samples_per_second": 111.263,
      "eval_steps_per_second": 7.418,
      "step": 6000
    },
    {
      "epoch": 2.284644194756554,
      "grad_norm": 12.399551391601562,
      "learning_rate": 2.1722846441947566e-05,
      "loss": 0.008,
      "step": 6100
    },
    {
      "epoch": 2.284644194756554,
      "eval_loss": 6.522914918605238e-05,
      "eval_runtime": 0.2688,
      "eval_samples_per_second": 111.598,
      "eval_steps_per_second": 7.44,
      "step": 6100
    },
    {
      "epoch": 2.3220973782771535,
      "grad_norm": 1171.171630859375,
      "learning_rate": 2.142322097378277e-05,
      "loss": 0.0083,
      "step": 6200
    },
    {
      "epoch": 2.3220973782771535,
      "eval_loss": 2.3407532353303395e-05,
      "eval_runtime": 0.2711,
      "eval_samples_per_second": 110.676,
      "eval_steps_per_second": 7.378,
      "step": 6200
    },
    {
      "epoch": 2.359550561797753,
      "grad_norm": 14939.6650390625,
      "learning_rate": 2.1123595505617976e-05,
      "loss": 0.0084,
      "step": 6300
    },
    {
      "epoch": 2.359550561797753,
      "eval_loss": 7.250668204505928e-06,
      "eval_runtime": 0.2736,
      "eval_samples_per_second": 109.63,
      "eval_steps_per_second": 7.309,
      "step": 6300
    },
    {
      "epoch": 2.397003745318352,
      "grad_norm": 31581.005859375,
      "learning_rate": 2.082397003745319e-05,
      "loss": 0.0093,
      "step": 6400
    },
    {
      "epoch": 2.397003745318352,
      "eval_loss": 0.00011670963431242853,
      "eval_runtime": 0.2702,
      "eval_samples_per_second": 111.025,
      "eval_steps_per_second": 7.402,
      "step": 6400
    },
    {
      "epoch": 2.4344569288389515,
      "grad_norm": 1787.995849609375,
      "learning_rate": 2.0524344569288394e-05,
      "loss": 0.0095,
      "step": 6500
    },
    {
      "epoch": 2.4344569288389515,
      "eval_loss": 0.0032277393620461226,
      "eval_runtime": 0.2705,
      "eval_samples_per_second": 110.89,
      "eval_steps_per_second": 7.393,
      "step": 6500
    },
    {
      "epoch": 2.4719101123595504,
      "grad_norm": 8077.27978515625,
      "learning_rate": 2.02247191011236e-05,
      "loss": 0.0087,
      "step": 6600
    },
    {
      "epoch": 2.4719101123595504,
      "eval_loss": 7.2363211074844e-05,
      "eval_runtime": 0.2711,
      "eval_samples_per_second": 110.656,
      "eval_steps_per_second": 7.377,
      "step": 6600
    },
    {
      "epoch": 2.5093632958801497,
      "grad_norm": 47643.1484375,
      "learning_rate": 1.99250936329588e-05,
      "loss": 0.008,
      "step": 6700
    },
    {
      "epoch": 2.5093632958801497,
      "eval_loss": 0.0006091619143262506,
      "eval_runtime": 0.2716,
      "eval_samples_per_second": 110.449,
      "eval_steps_per_second": 7.363,
      "step": 6700
    },
    {
      "epoch": 2.546816479400749,
      "grad_norm": 23858.97265625,
      "learning_rate": 1.9625468164794007e-05,
      "loss": 0.0089,
      "step": 6800
    },
    {
      "epoch": 2.546816479400749,
      "eval_loss": 0.00017142061551567167,
      "eval_runtime": 0.2735,
      "eval_samples_per_second": 109.688,
      "eval_steps_per_second": 7.313,
      "step": 6800
    },
    {
      "epoch": 2.5842696629213484,
      "grad_norm": 49.6945915222168,
      "learning_rate": 1.9325842696629215e-05,
      "loss": 0.0079,
      "step": 6900
    },
    {
      "epoch": 2.5842696629213484,
      "eval_loss": 3.809615009231493e-05,
      "eval_runtime": 0.2699,
      "eval_samples_per_second": 111.159,
      "eval_steps_per_second": 7.411,
      "step": 6900
    },
    {
      "epoch": 2.6217228464419478,
      "grad_norm": 294.8653564453125,
      "learning_rate": 1.902621722846442e-05,
      "loss": 0.008,
      "step": 7000
    },
    {
      "epoch": 2.6217228464419478,
      "eval_loss": 5.702438556909328e-06,
      "eval_runtime": 0.273,
      "eval_samples_per_second": 109.906,
      "eval_steps_per_second": 7.327,
      "step": 7000
    },
    {
      "epoch": 2.6591760299625467,
      "grad_norm": 929.7149658203125,
      "learning_rate": 1.8726591760299626e-05,
      "loss": 0.0097,
      "step": 7100
    },
    {
      "epoch": 2.6591760299625467,
      "eval_loss": 3.306344524389715e-06,
      "eval_runtime": 0.2695,
      "eval_samples_per_second": 111.315,
      "eval_steps_per_second": 7.421,
      "step": 7100
    },
    {
      "epoch": 2.696629213483146,
      "grad_norm": 28578.640625,
      "learning_rate": 1.8426966292134835e-05,
      "loss": 0.0093,
      "step": 7200
    },
    {
      "epoch": 2.696629213483146,
      "eval_loss": 7.41585245123133e-05,
      "eval_runtime": 0.2718,
      "eval_samples_per_second": 110.387,
      "eval_steps_per_second": 7.359,
      "step": 7200
    },
    {
      "epoch": 2.7340823970037453,
      "grad_norm": 18150.853515625,
      "learning_rate": 1.812734082397004e-05,
      "loss": 0.009,
      "step": 7300
    },
    {
      "epoch": 2.7340823970037453,
      "eval_loss": 0.00015120919852051884,
      "eval_runtime": 0.2704,
      "eval_samples_per_second": 110.966,
      "eval_steps_per_second": 7.398,
      "step": 7300
    },
    {
      "epoch": 2.7715355805243447,
      "grad_norm": 17.830102920532227,
      "learning_rate": 1.7827715355805242e-05,
      "loss": 0.0077,
      "step": 7400
    },
    {
      "epoch": 2.7715355805243447,
      "eval_loss": 0.000870275020133704,
      "eval_runtime": 0.2752,
      "eval_samples_per_second": 109.007,
      "eval_steps_per_second": 7.267,
      "step": 7400
    },
    {
      "epoch": 2.808988764044944,
      "grad_norm": 16.789941787719727,
      "learning_rate": 1.752808988764045e-05,
      "loss": 0.0088,
      "step": 7500
    },
    {
      "epoch": 2.808988764044944,
      "eval_loss": 6.32848750683479e-05,
      "eval_runtime": 0.2693,
      "eval_samples_per_second": 111.418,
      "eval_steps_per_second": 7.428,
      "step": 7500
    },
    {
      "epoch": 2.846441947565543,
      "grad_norm": 17693.96875,
      "learning_rate": 1.7228464419475657e-05,
      "loss": 0.0093,
      "step": 7600
    },
    {
      "epoch": 2.846441947565543,
      "eval_loss": 3.997669409727678e-05,
      "eval_runtime": 0.2694,
      "eval_samples_per_second": 111.364,
      "eval_steps_per_second": 7.424,
      "step": 7600
    },
    {
      "epoch": 2.8838951310861423,
      "grad_norm": 56.93388366699219,
      "learning_rate": 1.6928838951310862e-05,
      "loss": 0.0071,
      "step": 7700
    },
    {
      "epoch": 2.8838951310861423,
      "eval_loss": 4.744604211737169e-06,
      "eval_runtime": 0.2702,
      "eval_samples_per_second": 111.018,
      "eval_steps_per_second": 7.401,
      "step": 7700
    },
    {
      "epoch": 2.9213483146067416,
      "grad_norm": 7.296459674835205,
      "learning_rate": 1.662921348314607e-05,
      "loss": 0.0094,
      "step": 7800
    },
    {
      "epoch": 2.9213483146067416,
      "eval_loss": 3.2065245250123553e-06,
      "eval_runtime": 0.2713,
      "eval_samples_per_second": 110.58,
      "eval_steps_per_second": 7.372,
      "step": 7800
    },
    {
      "epoch": 2.958801498127341,
      "grad_norm": 3018.1708984375,
      "learning_rate": 1.6329588014981276e-05,
      "loss": 0.0093,
      "step": 7900
    },
    {
      "epoch": 2.958801498127341,
      "eval_loss": 1.5507857824559323e-06,
      "eval_runtime": 0.2707,
      "eval_samples_per_second": 110.815,
      "eval_steps_per_second": 7.388,
      "step": 7900
    },
    {
      "epoch": 2.9962546816479403,
      "grad_norm": 1547.920166015625,
      "learning_rate": 1.602996254681648e-05,
      "loss": 0.0103,
      "step": 8000
    },
    {
      "epoch": 2.9962546816479403,
      "eval_loss": 2.826002173605957e-06,
      "eval_runtime": 0.2719,
      "eval_samples_per_second": 110.33,
      "eval_steps_per_second": 7.355,
      "step": 8000
    },
    {
      "epoch": 3.033707865168539,
      "grad_norm": 688.6261596679688,
      "learning_rate": 1.5730337078651687e-05,
      "loss": 0.009,
      "step": 8100
    },
    {
      "epoch": 3.033707865168539,
      "eval_loss": 5.406631771620596e-06,
      "eval_runtime": 0.2693,
      "eval_samples_per_second": 111.416,
      "eval_steps_per_second": 7.428,
      "step": 8100
    },
    {
      "epoch": 3.0711610486891385,
      "grad_norm": 64.58126831054688,
      "learning_rate": 1.5430711610486892e-05,
      "loss": 0.0074,
      "step": 8200
    },
    {
      "epoch": 3.0711610486891385,
      "eval_loss": 5.987918029859429e-06,
      "eval_runtime": 0.2706,
      "eval_samples_per_second": 110.848,
      "eval_steps_per_second": 7.39,
      "step": 8200
    },
    {
      "epoch": 3.108614232209738,
      "grad_norm": 27061.69921875,
      "learning_rate": 1.5131086142322098e-05,
      "loss": 0.0057,
      "step": 8300
    },
    {
      "epoch": 3.108614232209738,
      "eval_loss": 9.855740245257039e-06,
      "eval_runtime": 0.2693,
      "eval_samples_per_second": 111.4,
      "eval_steps_per_second": 7.427,
      "step": 8300
    },
    {
      "epoch": 3.146067415730337,
      "grad_norm": 111.40906524658203,
      "learning_rate": 1.4831460674157305e-05,
      "loss": 0.0059,
      "step": 8400
    },
    {
      "epoch": 3.146067415730337,
      "eval_loss": 2.333690645173192e-06,
      "eval_runtime": 0.2709,
      "eval_samples_per_second": 110.749,
      "eval_steps_per_second": 7.383,
      "step": 8400
    },
    {
      "epoch": 3.1835205992509366,
      "grad_norm": 40.62784957885742,
      "learning_rate": 1.453183520599251e-05,
      "loss": 0.0083,
      "step": 8500
    },
    {
      "epoch": 3.1835205992509366,
      "eval_loss": 1.2001773029624019e-05,
      "eval_runtime": 0.2696,
      "eval_samples_per_second": 111.271,
      "eval_steps_per_second": 7.418,
      "step": 8500
    },
    {
      "epoch": 3.2209737827715355,
      "grad_norm": 964.7347412109375,
      "learning_rate": 1.4232209737827715e-05,
      "loss": 0.0072,
      "step": 8600
    },
    {
      "epoch": 3.2209737827715355,
      "eval_loss": 2.515283449611161e-05,
      "eval_runtime": 0.2704,
      "eval_samples_per_second": 110.931,
      "eval_steps_per_second": 7.395,
      "step": 8600
    },
    {
      "epoch": 3.258426966292135,
      "grad_norm": 20970.783203125,
      "learning_rate": 1.3932584269662923e-05,
      "loss": 0.0077,
      "step": 8700
    },
    {
      "epoch": 3.258426966292135,
      "eval_loss": 1.2501574019552208e-05,
      "eval_runtime": 0.2702,
      "eval_samples_per_second": 111.039,
      "eval_steps_per_second": 7.403,
      "step": 8700
    },
    {
      "epoch": 3.295880149812734,
      "grad_norm": 22375.5546875,
      "learning_rate": 1.3632958801498128e-05,
      "loss": 0.0077,
      "step": 8800
    },
    {
      "epoch": 3.295880149812734,
      "eval_loss": 0.00010068963456433266,
      "eval_runtime": 0.2687,
      "eval_samples_per_second": 111.63,
      "eval_steps_per_second": 7.442,
      "step": 8800
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 17.80713653564453,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.0087,
      "step": 8900
    },
    {
      "epoch": 3.3333333333333335,
      "eval_loss": 4.862140485784039e-06,
      "eval_runtime": 0.2711,
      "eval_samples_per_second": 110.641,
      "eval_steps_per_second": 7.376,
      "step": 8900
    },
    {
      "epoch": 3.370786516853933,
      "grad_norm": 311.2126770019531,
      "learning_rate": 1.303370786516854e-05,
      "loss": 0.0065,
      "step": 9000
    },
    {
      "epoch": 3.370786516853933,
      "eval_loss": 1.9432780391070992e-05,
      "eval_runtime": 0.2685,
      "eval_samples_per_second": 111.729,
      "eval_steps_per_second": 7.449,
      "step": 9000
    },
    {
      "epoch": 3.4082397003745317,
      "grad_norm": 74.10047912597656,
      "learning_rate": 1.2734082397003746e-05,
      "loss": 0.0085,
      "step": 9100
    },
    {
      "epoch": 3.4082397003745317,
      "eval_loss": 1.4302119097919785e-06,
      "eval_runtime": 0.2727,
      "eval_samples_per_second": 110.01,
      "eval_steps_per_second": 7.334,
      "step": 9100
    },
    {
      "epoch": 3.445692883895131,
      "grad_norm": 156.77000427246094,
      "learning_rate": 1.2434456928838951e-05,
      "loss": 0.0075,
      "step": 9200
    },
    {
      "epoch": 3.445692883895131,
      "eval_loss": 3.0803469144302653e-06,
      "eval_runtime": 0.2691,
      "eval_samples_per_second": 111.479,
      "eval_steps_per_second": 7.432,
      "step": 9200
    },
    {
      "epoch": 3.4831460674157304,
      "grad_norm": 537.3113403320312,
      "learning_rate": 1.213483146067416e-05,
      "loss": 0.0062,
      "step": 9300
    },
    {
      "epoch": 3.4831460674157304,
      "eval_loss": 7.492248550988734e-05,
      "eval_runtime": 0.2694,
      "eval_samples_per_second": 111.353,
      "eval_steps_per_second": 7.424,
      "step": 9300
    },
    {
      "epoch": 3.5205992509363297,
      "grad_norm": 31979.20703125,
      "learning_rate": 1.1835205992509364e-05,
      "loss": 0.0078,
      "step": 9400
    },
    {
      "epoch": 3.5205992509363297,
      "eval_loss": 6.174228474264964e-05,
      "eval_runtime": 0.2699,
      "eval_samples_per_second": 111.165,
      "eval_steps_per_second": 7.411,
      "step": 9400
    },
    {
      "epoch": 3.558052434456929,
      "grad_norm": 17425.19921875,
      "learning_rate": 1.1535580524344569e-05,
      "loss": 0.0096,
      "step": 9500
    },
    {
      "epoch": 3.558052434456929,
      "eval_loss": 0.00016342249000445008,
      "eval_runtime": 0.2705,
      "eval_samples_per_second": 110.905,
      "eval_steps_per_second": 7.394,
      "step": 9500
    },
    {
      "epoch": 3.595505617977528,
      "grad_norm": 47.60939407348633,
      "learning_rate": 1.1235955056179778e-05,
      "loss": 0.0049,
      "step": 9600
    },
    {
      "epoch": 3.595505617977528,
      "eval_loss": 8.28147130960133e-06,
      "eval_runtime": 0.2779,
      "eval_samples_per_second": 107.94,
      "eval_steps_per_second": 7.196,
      "step": 9600
    },
    {
      "epoch": 3.6329588014981273,
      "grad_norm": 3124.997314453125,
      "learning_rate": 1.0936329588014981e-05,
      "loss": 0.0096,
      "step": 9700
    },
    {
      "epoch": 3.6329588014981273,
      "eval_loss": 2.2524438918480882e-06,
      "eval_runtime": 0.2702,
      "eval_samples_per_second": 111.015,
      "eval_steps_per_second": 7.401,
      "step": 9700
    },
    {
      "epoch": 3.6704119850187267,
      "grad_norm": 22.917911529541016,
      "learning_rate": 1.0636704119850187e-05,
      "loss": 0.0098,
      "step": 9800
    },
    {
      "epoch": 3.6704119850187267,
      "eval_loss": 2.805942585837329e-06,
      "eval_runtime": 0.2717,
      "eval_samples_per_second": 110.402,
      "eval_steps_per_second": 7.36,
      "step": 9800
    },
    {
      "epoch": 3.7078651685393256,
      "grad_norm": 46.66259002685547,
      "learning_rate": 1.0337078651685396e-05,
      "loss": 0.0091,
      "step": 9900
    },
    {
      "epoch": 3.7078651685393256,
      "eval_loss": 1.3199253771745134e-05,
      "eval_runtime": 0.2717,
      "eval_samples_per_second": 110.403,
      "eval_steps_per_second": 7.36,
      "step": 9900
    },
    {
      "epoch": 3.7453183520599254,
      "grad_norm": 9.422457695007324,
      "learning_rate": 1.0037453183520601e-05,
      "loss": 0.0076,
      "step": 10000
    },
    {
      "epoch": 3.7453183520599254,
      "eval_loss": 2.4877342639229028e-06,
      "eval_runtime": 0.2712,
      "eval_samples_per_second": 110.604,
      "eval_steps_per_second": 7.374,
      "step": 10000
    },
    {
      "epoch": 3.7827715355805243,
      "grad_norm": 354.9677734375,
      "learning_rate": 9.737827715355806e-06,
      "loss": 0.0089,
      "step": 10100
    },
    {
      "epoch": 3.7827715355805243,
      "eval_loss": 7.319365522562293e-07,
      "eval_runtime": 0.2715,
      "eval_samples_per_second": 110.494,
      "eval_steps_per_second": 7.366,
      "step": 10100
    },
    {
      "epoch": 3.8202247191011236,
      "grad_norm": 967.5068359375,
      "learning_rate": 9.438202247191012e-06,
      "loss": 0.0074,
      "step": 10200
    },
    {
      "epoch": 3.8202247191011236,
      "eval_loss": 1.387107386108255e-05,
      "eval_runtime": 0.2802,
      "eval_samples_per_second": 107.063,
      "eval_steps_per_second": 7.138,
      "step": 10200
    },
    {
      "epoch": 3.857677902621723,
      "grad_norm": 24458.201171875,
      "learning_rate": 9.138576779026219e-06,
      "loss": 0.0084,
      "step": 10300
    },
    {
      "epoch": 3.857677902621723,
      "eval_loss": 1.0614428902044892e-05,
      "eval_runtime": 0.2693,
      "eval_samples_per_second": 111.408,
      "eval_steps_per_second": 7.427,
      "step": 10300
    },
    {
      "epoch": 3.895131086142322,
      "grad_norm": 32194.888671875,
      "learning_rate": 8.838951310861424e-06,
      "loss": 0.0104,
      "step": 10400
    },
    {
      "epoch": 3.895131086142322,
      "eval_loss": 2.7534890705283033e-06,
      "eval_runtime": 0.2706,
      "eval_samples_per_second": 110.869,
      "eval_steps_per_second": 7.391,
      "step": 10400
    },
    {
      "epoch": 3.932584269662921,
      "grad_norm": 18117.740234375,
      "learning_rate": 8.53932584269663e-06,
      "loss": 0.0058,
      "step": 10500
    },
    {
      "epoch": 3.932584269662921,
      "eval_loss": 2.031719077422167e-06,
      "eval_runtime": 0.2713,
      "eval_samples_per_second": 110.592,
      "eval_steps_per_second": 7.373,
      "step": 10500
    },
    {
      "epoch": 3.9700374531835205,
      "grad_norm": 14929.3486328125,
      "learning_rate": 8.239700374531837e-06,
      "loss": 0.0081,
      "step": 10600
    },
    {
      "epoch": 3.9700374531835205,
      "eval_loss": 1.8773675947159063e-06,
      "eval_runtime": 0.2688,
      "eval_samples_per_second": 111.598,
      "eval_steps_per_second": 7.44,
      "step": 10600
    },
    {
      "epoch": 4.007490636704119,
      "grad_norm": 1356.1688232421875,
      "learning_rate": 7.940074906367042e-06,
      "loss": 0.0058,
      "step": 10700
    },
    {
      "epoch": 4.007490636704119,
      "eval_loss": 2.515456287710549e-07,
      "eval_runtime": 0.2743,
      "eval_samples_per_second": 109.363,
      "eval_steps_per_second": 7.291,
      "step": 10700
    },
    {
      "epoch": 4.044943820224719,
      "grad_norm": 4.012579917907715,
      "learning_rate": 7.640449438202247e-06,
      "loss": 0.0073,
      "step": 10800
    },
    {
      "epoch": 4.044943820224719,
      "eval_loss": 3.1969119618224795e-07,
      "eval_runtime": 0.2694,
      "eval_samples_per_second": 111.353,
      "eval_steps_per_second": 7.424,
      "step": 10800
    },
    {
      "epoch": 4.082397003745318,
      "grad_norm": 22902.978515625,
      "learning_rate": 7.340823970037454e-06,
      "loss": 0.0077,
      "step": 10900
    },
    {
      "epoch": 4.082397003745318,
      "eval_loss": 7.97858433543297e-07,
      "eval_runtime": 0.2701,
      "eval_samples_per_second": 111.07,
      "eval_steps_per_second": 7.405,
      "step": 10900
    },
    {
      "epoch": 4.119850187265918,
      "grad_norm": 19.242652893066406,
      "learning_rate": 7.04119850187266e-06,
      "loss": 0.006,
      "step": 11000
    },
    {
      "epoch": 4.119850187265918,
      "eval_loss": 3.754879571715719e-07,
      "eval_runtime": 0.2696,
      "eval_samples_per_second": 111.279,
      "eval_steps_per_second": 7.419,
      "step": 11000
    },
    {
      "epoch": 4.157303370786517,
      "grad_norm": 594.0513916015625,
      "learning_rate": 6.741573033707865e-06,
      "loss": 0.0076,
      "step": 11100
    },
    {
      "epoch": 4.157303370786517,
      "eval_loss": 1.5144707958825165e-06,
      "eval_runtime": 0.2689,
      "eval_samples_per_second": 111.549,
      "eval_steps_per_second": 7.437,
      "step": 11100
    },
    {
      "epoch": 4.194756554307116,
      "grad_norm": 40.14280319213867,
      "learning_rate": 6.4419475655430715e-06,
      "loss": 0.0072,
      "step": 11200
    },
    {
      "epoch": 4.194756554307116,
      "eval_loss": 4.6455261326627806e-07,
      "eval_runtime": 0.2709,
      "eval_samples_per_second": 110.757,
      "eval_steps_per_second": 7.384,
      "step": 11200
    },
    {
      "epoch": 4.2322097378277155,
      "grad_norm": 11464.6328125,
      "learning_rate": 6.142322097378277e-06,
      "loss": 0.0066,
      "step": 11300
    },
    {
      "epoch": 4.2322097378277155,
      "eval_loss": 2.016280546968119e-07,
      "eval_runtime": 0.2784,
      "eval_samples_per_second": 107.752,
      "eval_steps_per_second": 7.183,
      "step": 11300
    },
    {
      "epoch": 4.269662921348314,
      "grad_norm": 25621.724609375,
      "learning_rate": 5.842696629213483e-06,
      "loss": 0.0076,
      "step": 11400
    },
    {
      "epoch": 4.269662921348314,
      "eval_loss": 2.0839000569594646e-07,
      "eval_runtime": 0.2707,
      "eval_samples_per_second": 110.843,
      "eval_steps_per_second": 7.39,
      "step": 11400
    },
    {
      "epoch": 4.307116104868914,
      "grad_norm": 18361.001953125,
      "learning_rate": 5.54307116104869e-06,
      "loss": 0.0063,
      "step": 11500
    },
    {
      "epoch": 4.307116104868914,
      "eval_loss": 3.3212629091394774e-07,
      "eval_runtime": 0.2696,
      "eval_samples_per_second": 111.296,
      "eval_steps_per_second": 7.42,
      "step": 11500
    },
    {
      "epoch": 4.344569288389513,
      "grad_norm": 68192.8984375,
      "learning_rate": 5.243445692883895e-06,
      "loss": 0.0078,
      "step": 11600
    },
    {
      "epoch": 4.344569288389513,
      "eval_loss": 1.3938915799371898e-07,
      "eval_runtime": 0.2711,
      "eval_samples_per_second": 110.662,
      "eval_steps_per_second": 7.377,
      "step": 11600
    },
    {
      "epoch": 4.382022471910112,
      "grad_norm": 62.034523010253906,
      "learning_rate": 4.943820224719101e-06,
      "loss": 0.0069,
      "step": 11700
    },
    {
      "epoch": 4.382022471910112,
      "eval_loss": 1.2977771746136568e-07,
      "eval_runtime": 0.271,
      "eval_samples_per_second": 110.697,
      "eval_steps_per_second": 7.38,
      "step": 11700
    },
    {
      "epoch": 4.419475655430712,
      "grad_norm": 27752.787109375,
      "learning_rate": 4.644194756554308e-06,
      "loss": 0.0081,
      "step": 11800
    },
    {
      "epoch": 4.419475655430712,
      "eval_loss": 1.483793425904878e-07,
      "eval_runtime": 0.2751,
      "eval_samples_per_second": 109.043,
      "eval_steps_per_second": 7.27,
      "step": 11800
    },
    {
      "epoch": 4.456928838951311,
      "grad_norm": 24.9007511138916,
      "learning_rate": 4.3445692883895135e-06,
      "loss": 0.0078,
      "step": 11900
    },
    {
      "epoch": 4.456928838951311,
      "eval_loss": 5.084881422590115e-07,
      "eval_runtime": 0.2713,
      "eval_samples_per_second": 110.57,
      "eval_steps_per_second": 7.371,
      "step": 11900
    },
    {
      "epoch": 4.49438202247191,
      "grad_norm": 7.775390148162842,
      "learning_rate": 4.04494382022472e-06,
      "loss": 0.0076,
      "step": 12000
    },
    {
      "epoch": 4.49438202247191,
      "eval_loss": 9.027158967001014e-07,
      "eval_runtime": 0.2721,
      "eval_samples_per_second": 110.24,
      "eval_steps_per_second": 7.349,
      "step": 12000
    },
    {
      "epoch": 4.531835205992509,
      "grad_norm": 3265.2392578125,
      "learning_rate": 3.7453183520599255e-06,
      "loss": 0.0061,
      "step": 12100
    },
    {
      "epoch": 4.531835205992509,
      "eval_loss": 1.5819171039765934e-06,
      "eval_runtime": 0.2698,
      "eval_samples_per_second": 111.177,
      "eval_steps_per_second": 7.412,
      "step": 12100
    },
    {
      "epoch": 4.569288389513108,
      "grad_norm": 12992.0361328125,
      "learning_rate": 3.4456928838951313e-06,
      "loss": 0.0079,
      "step": 12200
    },
    {
      "epoch": 4.569288389513108,
      "eval_loss": 7.184162882367673e-07,
      "eval_runtime": 0.2716,
      "eval_samples_per_second": 110.468,
      "eval_steps_per_second": 7.365,
      "step": 12200
    },
    {
      "epoch": 4.606741573033708,
      "grad_norm": 94.70529174804688,
      "learning_rate": 3.146067415730337e-06,
      "loss": 0.006,
      "step": 12300
    },
    {
      "epoch": 4.606741573033708,
      "eval_loss": 1.6477933968417346e-06,
      "eval_runtime": 0.2698,
      "eval_samples_per_second": 111.208,
      "eval_steps_per_second": 7.414,
      "step": 12300
    },
    {
      "epoch": 4.644194756554307,
      "grad_norm": 24.974130630493164,
      "learning_rate": 2.846441947565543e-06,
      "loss": 0.0061,
      "step": 12400
    },
    {
      "epoch": 4.644194756554307,
      "eval_loss": 6.807524073337845e-07,
      "eval_runtime": 0.2704,
      "eval_samples_per_second": 110.947,
      "eval_steps_per_second": 7.396,
      "step": 12400
    },
    {
      "epoch": 4.681647940074907,
      "grad_norm": 4.709781169891357,
      "learning_rate": 2.5468164794007496e-06,
      "loss": 0.0075,
      "step": 12500
    },
    {
      "epoch": 4.681647940074907,
      "eval_loss": 3.077086603298085e-07,
      "eval_runtime": 0.2692,
      "eval_samples_per_second": 111.445,
      "eval_steps_per_second": 7.43,
      "step": 12500
    },
    {
      "epoch": 4.719101123595506,
      "grad_norm": 61.4736442565918,
      "learning_rate": 2.2471910112359554e-06,
      "loss": 0.0054,
      "step": 12600
    },
    {
      "epoch": 4.719101123595506,
      "eval_loss": 2.621384282974759e-07,
      "eval_runtime": 0.2704,
      "eval_samples_per_second": 110.964,
      "eval_steps_per_second": 7.398,
      "step": 12600
    },
    {
      "epoch": 4.7565543071161045,
      "grad_norm": 279.9930114746094,
      "learning_rate": 1.947565543071161e-06,
      "loss": 0.0072,
      "step": 12700
    },
    {
      "epoch": 4.7565543071161045,
      "eval_loss": 3.2119106663230923e-07,
      "eval_runtime": 0.2696,
      "eval_samples_per_second": 111.29,
      "eval_steps_per_second": 7.419,
      "step": 12700
    },
    {
      "epoch": 4.794007490636704,
      "grad_norm": 1.977974772453308,
      "learning_rate": 1.647940074906367e-06,
      "loss": 0.0066,
      "step": 12800
    },
    {
      "epoch": 4.794007490636704,
      "eval_loss": 2.8463981038839847e-07,
      "eval_runtime": 0.2718,
      "eval_samples_per_second": 110.378,
      "eval_steps_per_second": 7.359,
      "step": 12800
    },
    {
      "epoch": 4.831460674157303,
      "grad_norm": 6.965720176696777,
      "learning_rate": 1.348314606741573e-06,
      "loss": 0.0076,
      "step": 12900
    },
    {
      "epoch": 4.831460674157303,
      "eval_loss": 5.141406518305303e-07,
      "eval_runtime": 0.27,
      "eval_samples_per_second": 111.115,
      "eval_steps_per_second": 7.408,
      "step": 12900
    },
    {
      "epoch": 4.868913857677903,
      "grad_norm": 68.03124237060547,
      "learning_rate": 1.0486891385767792e-06,
      "loss": 0.0076,
      "step": 13000
    },
    {
      "epoch": 4.868913857677903,
      "eval_loss": 5.566216145780345e-07,
      "eval_runtime": 0.2753,
      "eval_samples_per_second": 108.964,
      "eval_steps_per_second": 7.264,
      "step": 13000
    }
  ],
  "logging_steps": 100,
  "max_steps": 13350,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.242248406331392e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}