trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.18467220683287167,
  "eval_steps": 500,
  "global_step": 300,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0030778701138811943,
      "grad_norm": 0.47482776641845703,
      "learning_rate": 4.9999916942231176e-05,
      "loss": 1.1992,
      "num_input_tokens_seen": 42384,
      "step": 5,
      "train_runtime": 19.849,
      "train_tokens_per_second": 2135.325
    },
    {
      "epoch": 0.006155740227762389,
      "grad_norm": 0.24747319519519806,
      "learning_rate": 4.999957952099119e-05,
      "loss": 1.0247,
      "num_input_tokens_seen": 86000,
      "step": 10,
      "train_runtime": 39.0811,
      "train_tokens_per_second": 2200.554
    },
    {
      "epoch": 0.009233610341643583,
      "grad_norm": 0.10441822558641434,
      "learning_rate": 4.9998982548669973e-05,
      "loss": 0.8631,
      "num_input_tokens_seen": 124784,
      "step": 15,
      "train_runtime": 57.1682,
      "train_tokens_per_second": 2182.751
    },
    {
      "epoch": 0.012311480455524777,
      "grad_norm": 0.11058320850133896,
      "learning_rate": 4.999812603146542e-05,
      "loss": 0.9173,
      "num_input_tokens_seen": 165952,
      "step": 20,
      "train_runtime": 75.5733,
      "train_tokens_per_second": 2195.909
    },
    {
      "epoch": 0.015389350569405972,
      "grad_norm": 0.09879708290100098,
      "learning_rate": 4.9997009978270085e-05,
      "loss": 0.8722,
      "num_input_tokens_seen": 201584,
      "step": 25,
      "train_runtime": 92.5588,
      "train_tokens_per_second": 2177.903
    },
    {
      "epoch": 0.018467220683287166,
      "grad_norm": 0.11391397565603256,
      "learning_rate": 4.999563440067108e-05,
      "loss": 0.866,
      "num_input_tokens_seen": 240656,
      "step": 30,
      "train_runtime": 111.1373,
      "train_tokens_per_second": 2165.393
    },
    {
      "epoch": 0.02154509079716836,
      "grad_norm": 0.09722574800252914,
      "learning_rate": 4.9993999312949936e-05,
      "loss": 0.853,
      "num_input_tokens_seen": 287248,
      "step": 35,
      "train_runtime": 130.6684,
      "train_tokens_per_second": 2198.297
    },
    {
      "epoch": 0.024622960911049555,
      "grad_norm": 0.12133017182350159,
      "learning_rate": 4.99921047320825e-05,
      "loss": 0.88,
      "num_input_tokens_seen": 328448,
      "step": 40,
      "train_runtime": 149.193,
      "train_tokens_per_second": 2201.498
    },
    {
      "epoch": 0.027700831024930747,
      "grad_norm": 0.12174953520298004,
      "learning_rate": 4.9989950677738726e-05,
      "loss": 0.8503,
      "num_input_tokens_seen": 370512,
      "step": 45,
      "train_runtime": 168.2236,
      "train_tokens_per_second": 2202.497
    },
    {
      "epoch": 0.030778701138811943,
      "grad_norm": 0.12440948933362961,
      "learning_rate": 4.9987537172282466e-05,
      "loss": 0.8707,
      "num_input_tokens_seen": 409936,
      "step": 50,
      "train_runtime": 186.7223,
      "train_tokens_per_second": 2195.432
    },
    {
      "epoch": 0.03385657125269314,
      "grad_norm": 0.12797962129116058,
      "learning_rate": 4.998486424077127e-05,
      "loss": 0.8475,
      "num_input_tokens_seen": 450928,
      "step": 55,
      "train_runtime": 205.1107,
      "train_tokens_per_second": 2198.462
    },
    {
      "epoch": 0.03693444136657433,
      "grad_norm": 0.09793120622634888,
      "learning_rate": 4.9981931910956086e-05,
      "loss": 0.8409,
      "num_input_tokens_seen": 491152,
      "step": 60,
      "train_runtime": 223.3671,
      "train_tokens_per_second": 2198.856
    },
    {
      "epoch": 0.040012311480455524,
      "grad_norm": 0.10957301408052444,
      "learning_rate": 4.9978740213281005e-05,
      "loss": 0.8524,
      "num_input_tokens_seen": 532160,
      "step": 65,
      "train_runtime": 241.5682,
      "train_tokens_per_second": 2202.939
    },
    {
      "epoch": 0.04309018159433672,
      "grad_norm": 0.12283050268888474,
      "learning_rate": 4.9975289180882936e-05,
      "loss": 0.8944,
      "num_input_tokens_seen": 573392,
      "step": 70,
      "train_runtime": 259.8567,
      "train_tokens_per_second": 2206.57
    },
    {
      "epoch": 0.046168051708217916,
      "grad_norm": 0.1338459849357605,
      "learning_rate": 4.9971578849591225e-05,
      "loss": 0.8478,
      "num_input_tokens_seen": 613488,
      "step": 75,
      "train_runtime": 278.0456,
      "train_tokens_per_second": 2206.429
    },
    {
      "epoch": 0.04924592182209911,
      "grad_norm": 0.10969562828540802,
      "learning_rate": 4.996760925792737e-05,
      "loss": 0.8331,
      "num_input_tokens_seen": 656480,
      "step": 80,
      "train_runtime": 296.3508,
      "train_tokens_per_second": 2215.212
    },
    {
      "epoch": 0.0523237919359803,
      "grad_norm": 0.1411936730146408,
      "learning_rate": 4.996338044710452e-05,
      "loss": 0.8306,
      "num_input_tokens_seen": 698272,
      "step": 85,
      "train_runtime": 315.096,
      "train_tokens_per_second": 2216.061
    },
    {
      "epoch": 0.055401662049861494,
      "grad_norm": 0.1435951143503189,
      "learning_rate": 4.995889246102711e-05,
      "loss": 0.8318,
      "num_input_tokens_seen": 735376,
      "step": 90,
      "train_runtime": 332.4774,
      "train_tokens_per_second": 2211.807
    },
    {
      "epoch": 0.05847953216374269,
      "grad_norm": 0.15133805572986603,
      "learning_rate": 4.995414534629039e-05,
      "loss": 0.8674,
      "num_input_tokens_seen": 772544,
      "step": 95,
      "train_runtime": 350.9252,
      "train_tokens_per_second": 2201.449
    },
    {
      "epoch": 0.061557402277623886,
      "grad_norm": 0.10615513473749161,
      "learning_rate": 4.9949139152179944e-05,
      "loss": 0.901,
      "num_input_tokens_seen": 814928,
      "step": 100,
      "train_runtime": 369.5452,
      "train_tokens_per_second": 2205.219
    },
    {
      "epoch": 0.06463527239150507,
      "grad_norm": 0.10441838204860687,
      "learning_rate": 4.994387393067117e-05,
      "loss": 0.8522,
      "num_input_tokens_seen": 855104,
      "step": 105,
      "train_runtime": 388.7482,
      "train_tokens_per_second": 2199.635
    },
    {
      "epoch": 0.06771314250538628,
      "grad_norm": 0.1405644714832306,
      "learning_rate": 4.993834973642874e-05,
      "loss": 0.8614,
      "num_input_tokens_seen": 896272,
      "step": 110,
      "train_runtime": 406.8129,
      "train_tokens_per_second": 2203.155
    },
    {
      "epoch": 0.07079101261926747,
      "grad_norm": 0.11736899614334106,
      "learning_rate": 4.993256662680604e-05,
      "loss": 0.8371,
      "num_input_tokens_seen": 940864,
      "step": 115,
      "train_runtime": 425.836,
      "train_tokens_per_second": 2209.452
    },
    {
      "epoch": 0.07386888273314866,
      "grad_norm": 0.1269819736480713,
      "learning_rate": 4.992652466184458e-05,
      "loss": 0.9,
      "num_input_tokens_seen": 983360,
      "step": 120,
      "train_runtime": 444.5802,
      "train_tokens_per_second": 2211.884
    },
    {
      "epoch": 0.07694675284702986,
      "grad_norm": 0.10944321751594543,
      "learning_rate": 4.992022390427335e-05,
      "loss": 0.8629,
      "num_input_tokens_seen": 1025120,
      "step": 125,
      "train_runtime": 463.3805,
      "train_tokens_per_second": 2212.264
    },
    {
      "epoch": 0.08002462296091105,
      "grad_norm": 0.1438760608434677,
      "learning_rate": 4.99136644195082e-05,
      "loss": 0.8797,
      "num_input_tokens_seen": 1065360,
      "step": 130,
      "train_runtime": 480.9152,
      "train_tokens_per_second": 2215.276
    },
    {
      "epoch": 0.08310249307479224,
      "grad_norm": 0.13859188556671143,
      "learning_rate": 4.9906846275651125e-05,
      "loss": 0.8317,
      "num_input_tokens_seen": 1106112,
      "step": 135,
      "train_runtime": 498.8557,
      "train_tokens_per_second": 2217.299
    },
    {
      "epoch": 0.08618036318867343,
      "grad_norm": 0.1412304937839508,
      "learning_rate": 4.989976954348958e-05,
      "loss": 0.8422,
      "num_input_tokens_seen": 1149632,
      "step": 140,
      "train_runtime": 517.2481,
      "train_tokens_per_second": 2222.593
    },
    {
      "epoch": 0.08925823330255463,
      "grad_norm": 0.1513959765434265,
      "learning_rate": 4.989243429649573e-05,
      "loss": 0.8403,
      "num_input_tokens_seen": 1187824,
      "step": 145,
      "train_runtime": 535.2196,
      "train_tokens_per_second": 2219.321
    },
    {
      "epoch": 0.09233610341643583,
      "grad_norm": 0.13685107231140137,
      "learning_rate": 4.988484061082572e-05,
      "loss": 0.8961,
      "num_input_tokens_seen": 1229088,
      "step": 150,
      "train_runtime": 553.4777,
      "train_tokens_per_second": 2220.664
    },
    {
      "epoch": 0.09541397353031703,
      "grad_norm": 0.12999333441257477,
      "learning_rate": 4.987698856531884e-05,
      "loss": 0.814,
      "num_input_tokens_seen": 1273056,
      "step": 155,
      "train_runtime": 572.9621,
      "train_tokens_per_second": 2221.885
    },
    {
      "epoch": 0.09849184364419822,
      "grad_norm": 0.1481194943189621,
      "learning_rate": 4.986887824149674e-05,
      "loss": 0.8542,
      "num_input_tokens_seen": 1310288,
      "step": 160,
      "train_runtime": 589.8405,
      "train_tokens_per_second": 2221.427
    },
    {
      "epoch": 0.10156971375807941,
      "grad_norm": 0.1396579146385193,
      "learning_rate": 4.9860509723562573e-05,
      "loss": 0.8722,
      "num_input_tokens_seen": 1351312,
      "step": 165,
      "train_runtime": 607.8908,
      "train_tokens_per_second": 2222.952
    },
    {
      "epoch": 0.1046475838719606,
      "grad_norm": 0.12688206136226654,
      "learning_rate": 4.985188309840012e-05,
      "loss": 0.8587,
      "num_input_tokens_seen": 1396336,
      "step": 170,
      "train_runtime": 626.8902,
      "train_tokens_per_second": 2227.401
    },
    {
      "epoch": 0.1077254539858418,
      "grad_norm": 0.13825534284114838,
      "learning_rate": 4.984299845557287e-05,
      "loss": 0.8014,
      "num_input_tokens_seen": 1437984,
      "step": 175,
      "train_runtime": 645.2717,
      "train_tokens_per_second": 2228.494
    },
    {
      "epoch": 0.11080332409972299,
      "grad_norm": 0.12385495752096176,
      "learning_rate": 4.983385588732312e-05,
      "loss": 0.8385,
      "num_input_tokens_seen": 1475360,
      "step": 180,
      "train_runtime": 662.6866,
      "train_tokens_per_second": 2226.332
    },
    {
      "epoch": 0.11388119421360418,
      "grad_norm": 0.12246419489383698,
      "learning_rate": 4.982445548857102e-05,
      "loss": 0.8372,
      "num_input_tokens_seen": 1513152,
      "step": 185,
      "train_runtime": 680.3316,
      "train_tokens_per_second": 2224.139
    },
    {
      "epoch": 0.11695906432748537,
      "grad_norm": 0.12281708419322968,
      "learning_rate": 4.981479735691354e-05,
      "loss": 0.8446,
      "num_input_tokens_seen": 1556256,
      "step": 190,
      "train_runtime": 698.7652,
      "train_tokens_per_second": 2227.152
    },
    {
      "epoch": 0.12003693444136658,
      "grad_norm": 0.13810615241527557,
      "learning_rate": 4.980488159262353e-05,
      "loss": 0.8141,
      "num_input_tokens_seen": 1598832,
      "step": 195,
      "train_runtime": 717.4389,
      "train_tokens_per_second": 2228.527
    },
    {
      "epoch": 0.12311480455524777,
      "grad_norm": 0.15435761213302612,
      "learning_rate": 4.97947082986486e-05,
      "loss": 0.8398,
      "num_input_tokens_seen": 1641712,
      "step": 200,
      "train_runtime": 736.4476,
      "train_tokens_per_second": 2229.231
    },
    {
      "epoch": 0.12619267466912895,
      "grad_norm": 0.15266934037208557,
      "learning_rate": 4.978427758061009e-05,
      "loss": 0.8714,
      "num_input_tokens_seen": 1684544,
      "step": 205,
      "train_runtime": 756.7894,
      "train_tokens_per_second": 2225.908
    },
    {
      "epoch": 0.12927054478301014,
      "grad_norm": 0.13250574469566345,
      "learning_rate": 4.977358954680202e-05,
      "loss": 0.8116,
      "num_input_tokens_seen": 1723120,
      "step": 210,
      "train_runtime": 774.2541,
      "train_tokens_per_second": 2225.523
    },
    {
      "epoch": 0.13234841489689136,
      "grad_norm": 0.11579591035842896,
      "learning_rate": 4.9762644308189865e-05,
      "loss": 0.8204,
      "num_input_tokens_seen": 1768096,
      "step": 215,
      "train_runtime": 793.0535,
      "train_tokens_per_second": 2229.479
    },
    {
      "epoch": 0.13542628501077256,
      "grad_norm": 0.12654677033424377,
      "learning_rate": 4.9751441978409484e-05,
      "loss": 0.8334,
      "num_input_tokens_seen": 1808176,
      "step": 220,
      "train_runtime": 810.905,
      "train_tokens_per_second": 2229.825
    },
    {
      "epoch": 0.13850415512465375,
      "grad_norm": 0.15607228875160217,
      "learning_rate": 4.97399826737659e-05,
      "loss": 0.8177,
      "num_input_tokens_seen": 1848480,
      "step": 225,
      "train_runtime": 829.3126,
      "train_tokens_per_second": 2228.93
    },
    {
      "epoch": 0.14158202523853494,
      "grad_norm": 0.1535269021987915,
      "learning_rate": 4.972826651323211e-05,
      "loss": 0.807,
      "num_input_tokens_seen": 1889168,
      "step": 230,
      "train_runtime": 847.0449,
      "train_tokens_per_second": 2230.304
    },
    {
      "epoch": 0.14465989535241613,
      "grad_norm": 0.13663041591644287,
      "learning_rate": 4.971629361844785e-05,
      "loss": 0.8557,
      "num_input_tokens_seen": 1930784,
      "step": 235,
      "train_runtime": 865.4457,
      "train_tokens_per_second": 2230.971
    },
    {
      "epoch": 0.14773776546629733,
      "grad_norm": 0.1719619184732437,
      "learning_rate": 4.9704064113718326e-05,
      "loss": 0.8436,
      "num_input_tokens_seen": 1974176,
      "step": 240,
      "train_runtime": 883.9962,
      "train_tokens_per_second": 2233.24
    },
    {
      "epoch": 0.15081563558017852,
      "grad_norm": 0.15830345451831818,
      "learning_rate": 4.9691578126012905e-05,
      "loss": 0.853,
      "num_input_tokens_seen": 2013008,
      "step": 245,
      "train_runtime": 902.1681,
      "train_tokens_per_second": 2231.3
    },
    {
      "epoch": 0.1538935056940597,
      "grad_norm": 0.14472956955432892,
      "learning_rate": 4.967883578496385e-05,
      "loss": 0.874,
      "num_input_tokens_seen": 2053328,
      "step": 250,
      "train_runtime": 919.7822,
      "train_tokens_per_second": 2232.407
    },
    {
      "epoch": 0.1569713758079409,
      "grad_norm": 0.13329900801181793,
      "learning_rate": 4.966583722286491e-05,
      "loss": 0.8333,
      "num_input_tokens_seen": 2094992,
      "step": 255,
      "train_runtime": 937.9537,
      "train_tokens_per_second": 2233.577
    },
    {
      "epoch": 0.1600492459218221,
      "grad_norm": 0.1551639437675476,
      "learning_rate": 4.9652582574669995e-05,
      "loss": 0.864,
      "num_input_tokens_seen": 2135376,
      "step": 260,
      "train_runtime": 955.8955,
      "train_tokens_per_second": 2233.901
    },
    {
      "epoch": 0.1631271160357033,
      "grad_norm": 0.12460612505674362,
      "learning_rate": 4.963907197799174e-05,
      "loss": 0.8321,
      "num_input_tokens_seen": 2176624,
      "step": 265,
      "train_runtime": 974.3283,
      "train_tokens_per_second": 2233.974
    },
    {
      "epoch": 0.16620498614958448,
      "grad_norm": 0.13981439173221588,
      "learning_rate": 4.9625305573100115e-05,
      "loss": 0.8349,
      "num_input_tokens_seen": 2220496,
      "step": 270,
      "train_runtime": 992.5551,
      "train_tokens_per_second": 2237.151
    },
    {
      "epoch": 0.16928285626346568,
      "grad_norm": 0.15424658358097076,
      "learning_rate": 4.961128350292094e-05,
      "loss": 0.8248,
      "num_input_tokens_seen": 2263808,
      "step": 275,
      "train_runtime": 1010.9218,
      "train_tokens_per_second": 2239.35
    },
    {
      "epoch": 0.17236072637734687,
      "grad_norm": 0.14288848638534546,
      "learning_rate": 4.959700591303438e-05,
      "loss": 0.8488,
      "num_input_tokens_seen": 2311696,
      "step": 280,
      "train_runtime": 1030.8143,
      "train_tokens_per_second": 2242.592
    },
    {
      "epoch": 0.17543859649122806,
      "grad_norm": 0.15639737248420715,
      "learning_rate": 4.95824729516735e-05,
      "loss": 0.8602,
      "num_input_tokens_seen": 2353200,
      "step": 285,
      "train_runtime": 1049.2401,
      "train_tokens_per_second": 2242.766
    },
    {
      "epoch": 0.17851646660510925,
      "grad_norm": 0.1268603354692459,
      "learning_rate": 4.9567684769722664e-05,
      "loss": 0.8987,
      "num_input_tokens_seen": 2396000,
      "step": 290,
      "train_runtime": 1067.7324,
      "train_tokens_per_second": 2244.008
    },
    {
      "epoch": 0.18159433671899045,
      "grad_norm": 0.14782211184501648,
      "learning_rate": 4.9552641520716014e-05,
      "loss": 0.7794,
      "num_input_tokens_seen": 2438032,
      "step": 295,
      "train_runtime": 1085.9257,
      "train_tokens_per_second": 2245.119
    },
    {
      "epoch": 0.18467220683287167,
      "grad_norm": 0.15259359776973724,
      "learning_rate": 4.953734336083583e-05,
      "loss": 0.8648,
      "num_input_tokens_seen": 2478048,
      "step": 300,
      "train_runtime": 1103.5538,
      "train_tokens_per_second": 2245.516
    }
  ],
  "logging_steps": 5,
  "max_steps": 4875,
  "num_input_tokens_seen": 2478048,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.765645728959693e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}