{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.5135802469135803,
  "eval_steps": 500,
  "global_step": 468,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0010973936899862826,
      "grad_norm": 0.27108055353164673,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 1.9861,
      "step": 1
    },
    {
      "epoch": 0.0021947873799725653,
      "grad_norm": 0.27781054377555847,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 2.0261,
      "step": 2
    },
    {
      "epoch": 0.0032921810699588477,
      "grad_norm": 0.2786555290222168,
      "learning_rate": 6.000000000000001e-07,
      "loss": 2.0242,
      "step": 3
    },
    {
      "epoch": 0.0043895747599451305,
      "grad_norm": 0.3020860552787781,
      "learning_rate": 8.000000000000001e-07,
      "loss": 2.0951,
      "step": 4
    },
    {
      "epoch": 0.0054869684499314125,
      "grad_norm": 0.2764291763305664,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 1.9623,
      "step": 5
    },
    {
      "epoch": 0.006584362139917695,
      "grad_norm": 0.3049551248550415,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 2.0732,
      "step": 6
    },
    {
      "epoch": 0.007681755829903978,
      "grad_norm": 0.2732066512107849,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 2.015,
      "step": 7
    },
    {
      "epoch": 0.008779149519890261,
      "grad_norm": 0.3337247371673584,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 2.1519,
      "step": 8
    },
    {
      "epoch": 0.009876543209876543,
      "grad_norm": 0.3307493329048157,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 2.0968,
      "step": 9
    },
    {
      "epoch": 0.010973936899862825,
      "grad_norm": 0.29627102613449097,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.0329,
      "step": 10
    },
    {
      "epoch": 0.012071330589849109,
      "grad_norm": 0.2832229733467102,
      "learning_rate": 2.2e-06,
      "loss": 2.0597,
      "step": 11
    },
    {
      "epoch": 0.01316872427983539,
      "grad_norm": 0.2993783950805664,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 1.9716,
      "step": 12
    },
    {
      "epoch": 0.014266117969821674,
      "grad_norm": 0.3148089051246643,
      "learning_rate": 2.6e-06,
      "loss": 2.1158,
      "step": 13
    },
    {
      "epoch": 0.015363511659807956,
      "grad_norm": 0.3562306761741638,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 2.1904,
      "step": 14
    },
    {
      "epoch": 0.01646090534979424,
      "grad_norm": 0.2843739688396454,
      "learning_rate": 3e-06,
      "loss": 1.9782,
      "step": 15
    },
    {
      "epoch": 0.017558299039780522,
      "grad_norm": 0.2809521555900574,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 1.9743,
      "step": 16
    },
    {
      "epoch": 0.018655692729766804,
      "grad_norm": 0.30157962441444397,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 1.9993,
      "step": 17
    },
    {
      "epoch": 0.019753086419753086,
      "grad_norm": 0.3049308657646179,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 2.0293,
      "step": 18
    },
    {
      "epoch": 0.020850480109739368,
      "grad_norm": 0.31477564573287964,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 2.0614,
      "step": 19
    },
    {
      "epoch": 0.02194787379972565,
      "grad_norm": 0.34430813789367676,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.082,
      "step": 20
    },
    {
      "epoch": 0.023045267489711935,
      "grad_norm": 0.29869017004966736,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 1.9798,
      "step": 21
    },
    {
      "epoch": 0.024142661179698217,
      "grad_norm": 0.28752100467681885,
      "learning_rate": 4.4e-06,
      "loss": 1.9871,
      "step": 22
    },
    {
      "epoch": 0.0252400548696845,
      "grad_norm": 0.33573031425476074,
      "learning_rate": 4.600000000000001e-06,
      "loss": 2.0437,
      "step": 23
    },
    {
      "epoch": 0.02633744855967078,
      "grad_norm": 0.29683223366737366,
      "learning_rate": 4.800000000000001e-06,
      "loss": 2.037,
      "step": 24
    },
    {
      "epoch": 0.027434842249657063,
      "grad_norm": 0.319722056388855,
      "learning_rate": 5e-06,
      "loss": 2.0975,
      "step": 25
    },
    {
      "epoch": 0.02853223593964335,
      "grad_norm": 0.3013472855091095,
      "learning_rate": 5.2e-06,
      "loss": 2.1016,
      "step": 26
    },
    {
      "epoch": 0.02962962962962963,
      "grad_norm": 0.28419122099876404,
      "learning_rate": 5.400000000000001e-06,
      "loss": 1.9477,
      "step": 27
    },
    {
      "epoch": 0.030727023319615913,
      "grad_norm": 0.2823992669582367,
      "learning_rate": 5.600000000000001e-06,
      "loss": 1.9544,
      "step": 28
    },
    {
      "epoch": 0.031824417009602195,
      "grad_norm": 0.2990231215953827,
      "learning_rate": 5.8e-06,
      "loss": 1.9982,
      "step": 29
    },
    {
      "epoch": 0.03292181069958848,
      "grad_norm": 0.28071701526641846,
      "learning_rate": 6e-06,
      "loss": 1.9735,
      "step": 30
    },
    {
      "epoch": 0.03401920438957476,
      "grad_norm": 0.28065022826194763,
      "learning_rate": 6.200000000000001e-06,
      "loss": 2.0191,
      "step": 31
    },
    {
      "epoch": 0.035116598079561044,
      "grad_norm": 0.2757718563079834,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 1.9187,
      "step": 32
    },
    {
      "epoch": 0.03621399176954732,
      "grad_norm": 0.2993408739566803,
      "learning_rate": 6.600000000000001e-06,
      "loss": 1.9989,
      "step": 33
    },
    {
      "epoch": 0.03731138545953361,
      "grad_norm": 0.28326326608657837,
      "learning_rate": 6.800000000000001e-06,
      "loss": 2.0218,
      "step": 34
    },
    {
      "epoch": 0.038408779149519894,
      "grad_norm": 0.2843347489833832,
      "learning_rate": 7e-06,
      "loss": 1.9994,
      "step": 35
    },
    {
      "epoch": 0.03950617283950617,
      "grad_norm": 0.28945305943489075,
      "learning_rate": 7.2000000000000005e-06,
      "loss": 2.0353,
      "step": 36
    },
    {
      "epoch": 0.04060356652949246,
      "grad_norm": 0.26997682452201843,
      "learning_rate": 7.4e-06,
      "loss": 2.0038,
      "step": 37
    },
    {
      "epoch": 0.041700960219478736,
      "grad_norm": 0.3185621201992035,
      "learning_rate": 7.600000000000001e-06,
      "loss": 1.9736,
      "step": 38
    },
    {
      "epoch": 0.04279835390946502,
      "grad_norm": 0.32522234320640564,
      "learning_rate": 7.800000000000002e-06,
      "loss": 2.1093,
      "step": 39
    },
    {
      "epoch": 0.0438957475994513,
      "grad_norm": 0.3031747043132782,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.0228,
      "step": 40
    },
    {
      "epoch": 0.044993141289437585,
      "grad_norm": 0.31074920296669006,
      "learning_rate": 8.2e-06,
      "loss": 2.0363,
      "step": 41
    },
    {
      "epoch": 0.04609053497942387,
      "grad_norm": 0.3501761257648468,
      "learning_rate": 8.400000000000001e-06,
      "loss": 2.0894,
      "step": 42
    },
    {
      "epoch": 0.04718792866941015,
      "grad_norm": 0.25342920422554016,
      "learning_rate": 8.6e-06,
      "loss": 1.919,
      "step": 43
    },
    {
      "epoch": 0.048285322359396435,
      "grad_norm": 0.2695287764072418,
      "learning_rate": 8.8e-06,
      "loss": 1.9496,
      "step": 44
    },
    {
      "epoch": 0.04938271604938271,
      "grad_norm": 0.26249709725379944,
      "learning_rate": 9e-06,
      "loss": 1.9096,
      "step": 45
    },
    {
      "epoch": 0.050480109739369,
      "grad_norm": 0.2559400498867035,
      "learning_rate": 9.200000000000002e-06,
      "loss": 1.9609,
      "step": 46
    },
    {
      "epoch": 0.051577503429355284,
      "grad_norm": 0.26113536953926086,
      "learning_rate": 9.4e-06,
      "loss": 1.9484,
      "step": 47
    },
    {
      "epoch": 0.05267489711934156,
      "grad_norm": 0.2483452707529068,
      "learning_rate": 9.600000000000001e-06,
      "loss": 1.9271,
      "step": 48
    },
    {
      "epoch": 0.05377229080932785,
      "grad_norm": 0.24584588408470154,
      "learning_rate": 9.800000000000001e-06,
      "loss": 1.9134,
      "step": 49
    },
    {
      "epoch": 0.05486968449931413,
      "grad_norm": 0.23456168174743652,
      "learning_rate": 1e-05,
      "loss": 1.8595,
      "step": 50
    },
    {
      "epoch": 0.05596707818930041,
      "grad_norm": 0.24329470098018646,
      "learning_rate": 1.02e-05,
      "loss": 1.8626,
      "step": 51
    },
    {
      "epoch": 0.0570644718792867,
      "grad_norm": 0.22366423904895782,
      "learning_rate": 1.04e-05,
      "loss": 1.9075,
      "step": 52
    },
    {
      "epoch": 0.058161865569272976,
      "grad_norm": 0.2408476620912552,
      "learning_rate": 1.0600000000000002e-05,
      "loss": 1.9073,
      "step": 53
    },
    {
      "epoch": 0.05925925925925926,
      "grad_norm": 0.26551100611686707,
      "learning_rate": 1.0800000000000002e-05,
      "loss": 1.9767,
      "step": 54
    },
    {
      "epoch": 0.06035665294924554,
      "grad_norm": 0.2730308771133423,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 1.9087,
      "step": 55
    },
    {
      "epoch": 0.061454046639231825,
      "grad_norm": 0.2597392797470093,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 1.975,
      "step": 56
    },
    {
      "epoch": 0.06255144032921811,
      "grad_norm": 0.24098357558250427,
      "learning_rate": 1.14e-05,
      "loss": 1.8624,
      "step": 57
    },
    {
      "epoch": 0.06364883401920439,
      "grad_norm": 0.23562097549438477,
      "learning_rate": 1.16e-05,
      "loss": 1.9068,
      "step": 58
    },
    {
      "epoch": 0.06474622770919067,
      "grad_norm": 0.24404938519001007,
      "learning_rate": 1.18e-05,
      "loss": 1.9516,
      "step": 59
    },
    {
      "epoch": 0.06584362139917696,
      "grad_norm": 0.245197594165802,
      "learning_rate": 1.2e-05,
      "loss": 1.9105,
      "step": 60
    },
    {
      "epoch": 0.06694101508916324,
      "grad_norm": 0.22279669344425201,
      "learning_rate": 1.22e-05,
      "loss": 1.9019,
      "step": 61
    },
    {
      "epoch": 0.06803840877914952,
      "grad_norm": 0.2473766803741455,
      "learning_rate": 1.2400000000000002e-05,
      "loss": 1.8418,
      "step": 62
    },
    {
      "epoch": 0.0691358024691358,
      "grad_norm": 0.2862042486667633,
      "learning_rate": 1.2600000000000001e-05,
      "loss": 1.9353,
      "step": 63
    },
    {
      "epoch": 0.07023319615912209,
      "grad_norm": 0.2588101625442505,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 1.9448,
      "step": 64
    },
    {
      "epoch": 0.07133058984910837,
      "grad_norm": 0.2884417772293091,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 1.8889,
      "step": 65
    },
    {
      "epoch": 0.07242798353909465,
      "grad_norm": 0.23664578795433044,
      "learning_rate": 1.3200000000000002e-05,
      "loss": 1.8244,
      "step": 66
    },
    {
      "epoch": 0.07352537722908094,
      "grad_norm": 0.25291287899017334,
      "learning_rate": 1.3400000000000002e-05,
      "loss": 1.8071,
      "step": 67
    },
    {
      "epoch": 0.07462277091906722,
      "grad_norm": 0.2404923290014267,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 1.8677,
      "step": 68
    },
    {
      "epoch": 0.0757201646090535,
      "grad_norm": 0.24635937809944153,
      "learning_rate": 1.38e-05,
      "loss": 1.8015,
      "step": 69
    },
    {
      "epoch": 0.07681755829903979,
      "grad_norm": 0.24646110832691193,
      "learning_rate": 1.4e-05,
      "loss": 1.7901,
      "step": 70
    },
    {
      "epoch": 0.07791495198902607,
      "grad_norm": 0.243991419672966,
      "learning_rate": 1.4200000000000001e-05,
      "loss": 1.7483,
      "step": 71
    },
    {
      "epoch": 0.07901234567901234,
      "grad_norm": 0.26060330867767334,
      "learning_rate": 1.4400000000000001e-05,
      "loss": 1.7065,
      "step": 72
    },
    {
      "epoch": 0.08010973936899862,
      "grad_norm": 0.2618904709815979,
      "learning_rate": 1.46e-05,
      "loss": 1.8377,
      "step": 73
    },
    {
      "epoch": 0.08120713305898491,
      "grad_norm": 0.26802942156791687,
      "learning_rate": 1.48e-05,
      "loss": 1.7947,
      "step": 74
    },
    {
      "epoch": 0.0823045267489712,
      "grad_norm": 0.28539785742759705,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 1.7596,
      "step": 75
    },
    {
      "epoch": 0.08340192043895747,
      "grad_norm": 0.25785282254219055,
      "learning_rate": 1.5200000000000002e-05,
      "loss": 1.6622,
      "step": 76
    },
    {
      "epoch": 0.08449931412894376,
      "grad_norm": 0.2746061682701111,
      "learning_rate": 1.54e-05,
      "loss": 1.6833,
      "step": 77
    },
    {
      "epoch": 0.08559670781893004,
      "grad_norm": 0.2993699610233307,
      "learning_rate": 1.5600000000000003e-05,
      "loss": 1.728,
      "step": 78
    },
    {
      "epoch": 0.08669410150891632,
      "grad_norm": 0.31998202204704285,
      "learning_rate": 1.58e-05,
      "loss": 1.7735,
      "step": 79
    },
    {
      "epoch": 0.0877914951989026,
      "grad_norm": 0.2985467314720154,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 1.6394,
      "step": 80
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 0.31099262833595276,
      "learning_rate": 1.62e-05,
      "loss": 1.7466,
      "step": 81
    },
    {
      "epoch": 0.08998628257887517,
      "grad_norm": 0.2854881286621094,
      "learning_rate": 1.64e-05,
      "loss": 1.7291,
      "step": 82
    },
    {
      "epoch": 0.09108367626886145,
      "grad_norm": 0.3132914900779724,
      "learning_rate": 1.66e-05,
      "loss": 1.7431,
      "step": 83
    },
    {
      "epoch": 0.09218106995884774,
      "grad_norm": 0.3053690791130066,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 1.6816,
      "step": 84
    },
    {
      "epoch": 0.09327846364883402,
      "grad_norm": 0.2843291759490967,
      "learning_rate": 1.7e-05,
      "loss": 1.7334,
      "step": 85
    },
    {
      "epoch": 0.0943758573388203,
      "grad_norm": 0.31299757957458496,
      "learning_rate": 1.72e-05,
      "loss": 1.6034,
      "step": 86
    },
    {
      "epoch": 0.09547325102880659,
      "grad_norm": 0.29502251744270325,
      "learning_rate": 1.7400000000000003e-05,
      "loss": 1.6195,
      "step": 87
    },
    {
      "epoch": 0.09657064471879287,
      "grad_norm": 0.32556185126304626,
      "learning_rate": 1.76e-05,
      "loss": 1.6456,
      "step": 88
    },
    {
      "epoch": 0.09766803840877915,
      "grad_norm": 0.3797527253627777,
      "learning_rate": 1.7800000000000002e-05,
      "loss": 1.5675,
      "step": 89
    },
    {
      "epoch": 0.09876543209876543,
      "grad_norm": 0.4491126239299774,
      "learning_rate": 1.8e-05,
      "loss": 1.6092,
      "step": 90
    },
    {
      "epoch": 0.09986282578875172,
      "grad_norm": 0.3437753915786743,
      "learning_rate": 1.8200000000000002e-05,
      "loss": 1.5709,
      "step": 91
    },
    {
      "epoch": 0.100960219478738,
      "grad_norm": 0.33262452483177185,
      "learning_rate": 1.8400000000000003e-05,
      "loss": 1.5559,
      "step": 92
    },
    {
      "epoch": 0.10205761316872428,
      "grad_norm": 0.4088902771472931,
      "learning_rate": 1.86e-05,
      "loss": 1.5929,
      "step": 93
    },
    {
      "epoch": 0.10315500685871057,
      "grad_norm": 0.42855969071388245,
      "learning_rate": 1.88e-05,
      "loss": 1.481,
      "step": 94
    },
    {
      "epoch": 0.10425240054869685,
      "grad_norm": 0.38623425364494324,
      "learning_rate": 1.9e-05,
      "loss": 1.4687,
      "step": 95
    },
    {
      "epoch": 0.10534979423868313,
      "grad_norm": 0.3554774522781372,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 1.4739,
      "step": 96
    },
    {
      "epoch": 0.1064471879286694,
      "grad_norm": 0.4467147886753082,
      "learning_rate": 1.94e-05,
      "loss": 1.4996,
      "step": 97
    },
    {
      "epoch": 0.1075445816186557,
      "grad_norm": 0.4558533728122711,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 1.463,
      "step": 98
    },
    {
      "epoch": 0.10864197530864197,
      "grad_norm": 0.3612457513809204,
      "learning_rate": 1.98e-05,
      "loss": 1.4667,
      "step": 99
    },
    {
      "epoch": 0.10973936899862825,
      "grad_norm": 0.3397855758666992,
      "learning_rate": 2e-05,
      "loss": 1.41,
      "step": 100
    },
    {
      "epoch": 0.11083676268861455,
      "grad_norm": 0.42120879888534546,
      "learning_rate": 1.997533908754624e-05,
      "loss": 1.3884,
      "step": 101
    },
    {
      "epoch": 0.11193415637860082,
      "grad_norm": 0.34555450081825256,
      "learning_rate": 1.995067817509248e-05,
      "loss": 1.4158,
      "step": 102
    },
    {
      "epoch": 0.1130315500685871,
      "grad_norm": 0.30095019936561584,
      "learning_rate": 1.992601726263872e-05,
      "loss": 1.4731,
      "step": 103
    },
    {
      "epoch": 0.1141289437585734,
      "grad_norm": 0.3604261577129364,
      "learning_rate": 1.990135635018496e-05,
      "loss": 1.4059,
      "step": 104
    },
    {
      "epoch": 0.11522633744855967,
      "grad_norm": 0.3028746545314789,
      "learning_rate": 1.9876695437731196e-05,
      "loss": 1.4234,
      "step": 105
    },
    {
      "epoch": 0.11632373113854595,
      "grad_norm": 0.343404084444046,
      "learning_rate": 1.985203452527744e-05,
      "loss": 1.319,
      "step": 106
    },
    {
      "epoch": 0.11742112482853223,
      "grad_norm": 0.26763787865638733,
      "learning_rate": 1.9827373612823677e-05,
      "loss": 1.3013,
      "step": 107
    },
    {
      "epoch": 0.11851851851851852,
      "grad_norm": 0.2645047903060913,
      "learning_rate": 1.9802712700369916e-05,
      "loss": 1.3119,
      "step": 108
    },
    {
      "epoch": 0.1196159122085048,
      "grad_norm": 0.24199865758419037,
      "learning_rate": 1.9778051787916155e-05,
      "loss": 1.3058,
      "step": 109
    },
    {
      "epoch": 0.12071330589849108,
      "grad_norm": 0.2630648910999298,
      "learning_rate": 1.9753390875462394e-05,
      "loss": 1.3024,
      "step": 110
    },
    {
      "epoch": 0.12181069958847737,
      "grad_norm": 0.29971155524253845,
      "learning_rate": 1.9728729963008633e-05,
      "loss": 1.2853,
      "step": 111
    },
    {
      "epoch": 0.12290809327846365,
      "grad_norm": 0.23841412365436554,
      "learning_rate": 1.9704069050554872e-05,
      "loss": 1.3528,
      "step": 112
    },
    {
      "epoch": 0.12400548696844993,
      "grad_norm": 0.18174125254154205,
      "learning_rate": 1.967940813810111e-05,
      "loss": 1.3534,
      "step": 113
    },
    {
      "epoch": 0.12510288065843622,
      "grad_norm": 0.23740167915821075,
      "learning_rate": 1.965474722564735e-05,
      "loss": 1.2486,
      "step": 114
    },
    {
      "epoch": 0.1262002743484225,
      "grad_norm": 0.317221999168396,
      "learning_rate": 1.9630086313193592e-05,
      "loss": 1.2452,
      "step": 115
    },
    {
      "epoch": 0.12729766803840878,
      "grad_norm": 0.17076945304870605,
      "learning_rate": 1.9605425400739828e-05,
      "loss": 1.282,
      "step": 116
    },
    {
      "epoch": 0.12839506172839507,
      "grad_norm": 0.19480882585048676,
      "learning_rate": 1.9580764488286066e-05,
      "loss": 1.3408,
      "step": 117
    },
    {
      "epoch": 0.12949245541838134,
      "grad_norm": 0.21473167836666107,
      "learning_rate": 1.955610357583231e-05,
      "loss": 1.2389,
      "step": 118
    },
    {
      "epoch": 0.13058984910836763,
      "grad_norm": 0.21171647310256958,
      "learning_rate": 1.9531442663378544e-05,
      "loss": 1.2866,
      "step": 119
    },
    {
      "epoch": 0.13168724279835392,
      "grad_norm": 0.16918618977069855,
      "learning_rate": 1.9506781750924787e-05,
      "loss": 1.3065,
      "step": 120
    },
    {
      "epoch": 0.13278463648834019,
      "grad_norm": 0.17470093071460724,
      "learning_rate": 1.9482120838471025e-05,
      "loss": 1.3088,
      "step": 121
    },
    {
      "epoch": 0.13388203017832648,
      "grad_norm": 0.16515418887138367,
      "learning_rate": 1.9457459926017264e-05,
      "loss": 1.2865,
      "step": 122
    },
    {
      "epoch": 0.13497942386831277,
      "grad_norm": 0.17647692561149597,
      "learning_rate": 1.9432799013563503e-05,
      "loss": 1.2298,
      "step": 123
    },
    {
      "epoch": 0.13607681755829903,
      "grad_norm": 0.1819489300251007,
      "learning_rate": 1.9408138101109742e-05,
      "loss": 1.2641,
      "step": 124
    },
    {
      "epoch": 0.13717421124828533,
      "grad_norm": 0.16584061086177826,
      "learning_rate": 1.938347718865598e-05,
      "loss": 1.2872,
      "step": 125
    },
    {
      "epoch": 0.1382716049382716,
      "grad_norm": 0.15061640739440918,
      "learning_rate": 1.935881627620222e-05,
      "loss": 1.1875,
      "step": 126
    },
    {
      "epoch": 0.13936899862825788,
      "grad_norm": 0.15820416808128357,
      "learning_rate": 1.9334155363748462e-05,
      "loss": 1.2242,
      "step": 127
    },
    {
      "epoch": 0.14046639231824418,
      "grad_norm": 0.16064974665641785,
      "learning_rate": 1.9309494451294698e-05,
      "loss": 1.2635,
      "step": 128
    },
    {
      "epoch": 0.14156378600823044,
      "grad_norm": 0.1459810435771942,
      "learning_rate": 1.928483353884094e-05,
      "loss": 1.2475,
      "step": 129
    },
    {
      "epoch": 0.14266117969821673,
      "grad_norm": 0.18043562769889832,
      "learning_rate": 1.9260172626387176e-05,
      "loss": 1.3077,
      "step": 130
    },
    {
      "epoch": 0.14375857338820303,
      "grad_norm": 0.16961152851581573,
      "learning_rate": 1.9235511713933418e-05,
      "loss": 1.2155,
      "step": 131
    },
    {
      "epoch": 0.1448559670781893,
      "grad_norm": 0.16236276924610138,
      "learning_rate": 1.9210850801479657e-05,
      "loss": 1.1733,
      "step": 132
    },
    {
      "epoch": 0.14595336076817558,
      "grad_norm": 0.15118443965911865,
      "learning_rate": 1.9186189889025896e-05,
      "loss": 1.2129,
      "step": 133
    },
    {
      "epoch": 0.14705075445816188,
      "grad_norm": 0.15442046523094177,
      "learning_rate": 1.9161528976572135e-05,
      "loss": 1.2811,
      "step": 134
    },
    {
      "epoch": 0.14814814814814814,
      "grad_norm": 0.15228775143623352,
      "learning_rate": 1.9136868064118374e-05,
      "loss": 1.2877,
      "step": 135
    },
    {
      "epoch": 0.14924554183813443,
      "grad_norm": 0.20188400149345398,
      "learning_rate": 1.9112207151664612e-05,
      "loss": 1.2504,
      "step": 136
    },
    {
      "epoch": 0.15034293552812072,
      "grad_norm": 0.15376144647598267,
      "learning_rate": 1.908754623921085e-05,
      "loss": 1.2608,
      "step": 137
    },
    {
      "epoch": 0.151440329218107,
      "grad_norm": 0.18212908506393433,
      "learning_rate": 1.9062885326757094e-05,
      "loss": 1.2443,
      "step": 138
    },
    {
      "epoch": 0.15253772290809328,
      "grad_norm": 0.18315152823925018,
      "learning_rate": 1.903822441430333e-05,
      "loss": 1.2282,
      "step": 139
    },
    {
      "epoch": 0.15363511659807957,
      "grad_norm": 0.18566685914993286,
      "learning_rate": 1.901356350184957e-05,
      "loss": 1.2061,
      "step": 140
    },
    {
      "epoch": 0.15473251028806584,
      "grad_norm": 0.17911821603775024,
      "learning_rate": 1.898890258939581e-05,
      "loss": 1.2598,
      "step": 141
    },
    {
      "epoch": 0.15582990397805213,
      "grad_norm": 0.15996401011943817,
      "learning_rate": 1.896424167694205e-05,
      "loss": 1.3087,
      "step": 142
    },
    {
      "epoch": 0.1569272976680384,
      "grad_norm": 0.15476301312446594,
      "learning_rate": 1.8939580764488288e-05,
      "loss": 1.2545,
      "step": 143
    },
    {
      "epoch": 0.1580246913580247,
      "grad_norm": 0.22089847922325134,
      "learning_rate": 1.8914919852034527e-05,
      "loss": 1.1303,
      "step": 144
    },
    {
      "epoch": 0.15912208504801098,
      "grad_norm": 0.19699229300022125,
      "learning_rate": 1.8890258939580766e-05,
      "loss": 1.1878,
      "step": 145
    },
    {
      "epoch": 0.16021947873799725,
      "grad_norm": 0.1997714638710022,
      "learning_rate": 1.8865598027127005e-05,
      "loss": 1.2437,
      "step": 146
    },
    {
      "epoch": 0.16131687242798354,
      "grad_norm": 0.19294144213199615,
      "learning_rate": 1.8840937114673244e-05,
      "loss": 1.178,
      "step": 147
    },
    {
      "epoch": 0.16241426611796983,
      "grad_norm": 0.18544507026672363,
      "learning_rate": 1.8816276202219483e-05,
      "loss": 1.1617,
      "step": 148
    },
    {
      "epoch": 0.1635116598079561,
      "grad_norm": 0.1713118851184845,
      "learning_rate": 1.879161528976572e-05,
      "loss": 1.2141,
      "step": 149
    },
    {
      "epoch": 0.1646090534979424,
      "grad_norm": 0.1757712960243225,
      "learning_rate": 1.876695437731196e-05,
      "loss": 1.2505,
      "step": 150
    },
    {
      "epoch": 0.16570644718792868,
      "grad_norm": 0.25870320200920105,
      "learning_rate": 1.87422934648582e-05,
      "loss": 1.1955,
      "step": 151
    },
    {
      "epoch": 0.16680384087791494,
      "grad_norm": 0.197688490152359,
      "learning_rate": 1.8717632552404442e-05,
      "loss": 1.1598,
      "step": 152
    },
    {
      "epoch": 0.16790123456790124,
      "grad_norm": 0.21808716654777527,
      "learning_rate": 1.8692971639950677e-05,
      "loss": 1.1682,
      "step": 153
    },
    {
      "epoch": 0.16899862825788753,
      "grad_norm": 0.26828381419181824,
      "learning_rate": 1.866831072749692e-05,
      "loss": 1.2372,
      "step": 154
    },
    {
      "epoch": 0.1700960219478738,
      "grad_norm": 0.20964942872524261,
      "learning_rate": 1.864364981504316e-05,
      "loss": 1.1816,
      "step": 155
    },
    {
      "epoch": 0.17119341563786009,
      "grad_norm": 0.2113889753818512,
      "learning_rate": 1.8618988902589397e-05,
      "loss": 1.1749,
      "step": 156
    },
    {
      "epoch": 0.17229080932784638,
      "grad_norm": 0.2011781483888626,
      "learning_rate": 1.8594327990135636e-05,
      "loss": 1.2413,
      "step": 157
    },
    {
      "epoch": 0.17338820301783264,
      "grad_norm": 0.2024763971567154,
      "learning_rate": 1.8569667077681875e-05,
      "loss": 1.2371,
      "step": 158
    },
    {
      "epoch": 0.17448559670781894,
      "grad_norm": 0.21682558953762054,
      "learning_rate": 1.8545006165228114e-05,
      "loss": 1.2055,
      "step": 159
    },
    {
      "epoch": 0.1755829903978052,
      "grad_norm": 0.24796062707901,
      "learning_rate": 1.8520345252774353e-05,
      "loss": 1.2046,
      "step": 160
    },
    {
      "epoch": 0.1766803840877915,
      "grad_norm": 0.25501304864883423,
      "learning_rate": 1.8495684340320595e-05,
      "loss": 1.1452,
      "step": 161
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 0.24295510351657867,
      "learning_rate": 1.847102342786683e-05,
      "loss": 1.3494,
      "step": 162
    },
    {
      "epoch": 0.17887517146776405,
      "grad_norm": 0.23228970170021057,
      "learning_rate": 1.8446362515413073e-05,
      "loss": 1.2089,
      "step": 163
    },
    {
      "epoch": 0.17997256515775034,
      "grad_norm": 0.22333182394504547,
      "learning_rate": 1.8421701602959312e-05,
      "loss": 1.1953,
      "step": 164
    },
    {
      "epoch": 0.18106995884773663,
      "grad_norm": 0.28809961676597595,
      "learning_rate": 1.839704069050555e-05,
      "loss": 1.2015,
      "step": 165
    },
    {
      "epoch": 0.1821673525377229,
      "grad_norm": 0.24514758586883545,
      "learning_rate": 1.837237977805179e-05,
      "loss": 1.1897,
      "step": 166
    },
    {
      "epoch": 0.1832647462277092,
      "grad_norm": 0.23966249823570251,
      "learning_rate": 1.834771886559803e-05,
      "loss": 1.2381,
      "step": 167
    },
    {
      "epoch": 0.18436213991769548,
      "grad_norm": 0.2911086082458496,
      "learning_rate": 1.8323057953144268e-05,
      "loss": 1.2303,
      "step": 168
    },
    {
      "epoch": 0.18545953360768175,
      "grad_norm": 0.2177533060312271,
      "learning_rate": 1.8298397040690507e-05,
      "loss": 1.2319,
      "step": 169
    },
    {
      "epoch": 0.18655692729766804,
      "grad_norm": 0.2377919703722,
      "learning_rate": 1.8273736128236746e-05,
      "loss": 1.2014,
      "step": 170
    },
    {
      "epoch": 0.18765432098765433,
      "grad_norm": 0.21792137622833252,
      "learning_rate": 1.8249075215782984e-05,
      "loss": 1.1645,
      "step": 171
    },
    {
      "epoch": 0.1887517146776406,
      "grad_norm": 0.2608526051044464,
      "learning_rate": 1.8224414303329227e-05,
      "loss": 1.1807,
      "step": 172
    },
    {
      "epoch": 0.1898491083676269,
      "grad_norm": 0.14767280220985413,
      "learning_rate": 1.8199753390875462e-05,
      "loss": 1.2637,
      "step": 173
    },
    {
      "epoch": 0.19094650205761318,
      "grad_norm": 0.15476517379283905,
      "learning_rate": 1.8175092478421705e-05,
      "loss": 1.229,
      "step": 174
    },
    {
      "epoch": 0.19204389574759945,
      "grad_norm": 0.1727033257484436,
      "learning_rate": 1.8150431565967943e-05,
      "loss": 1.2646,
      "step": 175
    },
    {
      "epoch": 0.19314128943758574,
      "grad_norm": 0.17207948863506317,
      "learning_rate": 1.8125770653514182e-05,
      "loss": 1.1604,
      "step": 176
    },
    {
      "epoch": 0.194238683127572,
      "grad_norm": 0.19457168877124786,
      "learning_rate": 1.810110974106042e-05,
      "loss": 1.3146,
      "step": 177
    },
    {
      "epoch": 0.1953360768175583,
      "grad_norm": 0.2222977876663208,
      "learning_rate": 1.807644882860666e-05,
      "loss": 1.2462,
      "step": 178
    },
    {
      "epoch": 0.1964334705075446,
      "grad_norm": 0.21270394325256348,
      "learning_rate": 1.80517879161529e-05,
      "loss": 1.1648,
      "step": 179
    },
    {
      "epoch": 0.19753086419753085,
      "grad_norm": 0.22708119451999664,
      "learning_rate": 1.8027127003699138e-05,
      "loss": 1.1397,
      "step": 180
    },
    {
      "epoch": 0.19862825788751715,
      "grad_norm": 0.21450617909431458,
      "learning_rate": 1.8002466091245377e-05,
      "loss": 1.1892,
      "step": 181
    },
    {
      "epoch": 0.19972565157750344,
      "grad_norm": 0.16464580595493317,
      "learning_rate": 1.7977805178791616e-05,
      "loss": 1.0918,
      "step": 182
    },
    {
      "epoch": 0.2008230452674897,
      "grad_norm": 0.22942472994327545,
      "learning_rate": 1.7953144266337855e-05,
      "loss": 1.2338,
      "step": 183
    },
    {
      "epoch": 0.201920438957476,
      "grad_norm": 0.16418524086475372,
      "learning_rate": 1.7928483353884094e-05,
      "loss": 1.241,
      "step": 184
    },
    {
      "epoch": 0.2030178326474623,
      "grad_norm": 0.18531592190265656,
      "learning_rate": 1.7903822441430333e-05,
      "loss": 1.1884,
      "step": 185
    },
    {
      "epoch": 0.20411522633744855,
      "grad_norm": 0.16026315093040466,
      "learning_rate": 1.7879161528976575e-05,
      "loss": 1.1948,
      "step": 186
    },
    {
      "epoch": 0.20521262002743484,
      "grad_norm": 0.16926898062229156,
      "learning_rate": 1.785450061652281e-05,
      "loss": 1.1487,
      "step": 187
    },
    {
      "epoch": 0.20631001371742114,
      "grad_norm": 0.21820904314517975,
      "learning_rate": 1.7829839704069053e-05,
      "loss": 1.1851,
      "step": 188
    },
    {
      "epoch": 0.2074074074074074,
      "grad_norm": 0.16973170638084412,
      "learning_rate": 1.780517879161529e-05,
      "loss": 1.2141,
      "step": 189
    },
    {
      "epoch": 0.2085048010973937,
      "grad_norm": 0.15451133251190186,
      "learning_rate": 1.778051787916153e-05,
      "loss": 1.2001,
      "step": 190
    },
    {
      "epoch": 0.20960219478738,
      "grad_norm": 0.20399673283100128,
      "learning_rate": 1.775585696670777e-05,
      "loss": 1.1314,
      "step": 191
    },
    {
      "epoch": 0.21069958847736625,
      "grad_norm": 0.16214239597320557,
      "learning_rate": 1.7731196054254008e-05,
      "loss": 1.1438,
      "step": 192
    },
    {
      "epoch": 0.21179698216735254,
      "grad_norm": 0.1611064076423645,
      "learning_rate": 1.7706535141800247e-05,
      "loss": 1.1743,
      "step": 193
    },
    {
      "epoch": 0.2128943758573388,
      "grad_norm": 0.12813259661197662,
      "learning_rate": 1.7681874229346486e-05,
      "loss": 1.1954,
      "step": 194
    },
    {
      "epoch": 0.2139917695473251,
      "grad_norm": 0.1601381003856659,
      "learning_rate": 1.765721331689273e-05,
      "loss": 1.1613,
      "step": 195
    },
    {
      "epoch": 0.2150891632373114,
      "grad_norm": 0.13577936589717865,
      "learning_rate": 1.7632552404438964e-05,
      "loss": 1.1982,
      "step": 196
    },
    {
      "epoch": 0.21618655692729766,
      "grad_norm": 0.13865117728710175,
      "learning_rate": 1.7607891491985206e-05,
      "loss": 1.1725,
      "step": 197
    },
    {
      "epoch": 0.21728395061728395,
      "grad_norm": 0.13816504180431366,
      "learning_rate": 1.7583230579531445e-05,
      "loss": 1.205,
      "step": 198
    },
    {
      "epoch": 0.21838134430727024,
      "grad_norm": 0.155447855591774,
      "learning_rate": 1.7558569667077684e-05,
      "loss": 1.1324,
      "step": 199
    },
    {
      "epoch": 0.2194787379972565,
      "grad_norm": 0.10742881894111633,
      "learning_rate": 1.7533908754623923e-05,
      "loss": 1.2384,
      "step": 200
    },
    {
      "epoch": 0.2205761316872428,
      "grad_norm": 0.12242433428764343,
      "learning_rate": 1.7509247842170162e-05,
      "loss": 1.1908,
      "step": 201
    },
    {
      "epoch": 0.2216735253772291,
      "grad_norm": 0.1424039602279663,
      "learning_rate": 1.74845869297164e-05,
      "loss": 1.1699,
      "step": 202
    },
    {
      "epoch": 0.22277091906721536,
      "grad_norm": 0.16237859427928925,
      "learning_rate": 1.745992601726264e-05,
      "loss": 1.0864,
      "step": 203
    },
    {
      "epoch": 0.22386831275720165,
      "grad_norm": 0.22748763859272003,
      "learning_rate": 1.743526510480888e-05,
      "loss": 1.2065,
      "step": 204
    },
    {
      "epoch": 0.22496570644718794,
      "grad_norm": 0.16209690272808075,
      "learning_rate": 1.7410604192355117e-05,
      "loss": 1.2071,
      "step": 205
    },
    {
      "epoch": 0.2260631001371742,
      "grad_norm": 0.12627777457237244,
      "learning_rate": 1.738594327990136e-05,
      "loss": 1.2082,
      "step": 206
    },
    {
      "epoch": 0.2271604938271605,
      "grad_norm": 0.12761177122592926,
      "learning_rate": 1.7361282367447595e-05,
      "loss": 1.1662,
      "step": 207
    },
    {
      "epoch": 0.2282578875171468,
      "grad_norm": 0.16722379624843597,
      "learning_rate": 1.7336621454993838e-05,
      "loss": 1.3239,
      "step": 208
    },
    {
      "epoch": 0.22935528120713305,
      "grad_norm": 0.22056573629379272,
      "learning_rate": 1.7311960542540076e-05,
      "loss": 1.24,
      "step": 209
    },
    {
      "epoch": 0.23045267489711935,
      "grad_norm": 0.15545333921909332,
      "learning_rate": 1.7287299630086315e-05,
      "loss": 1.1569,
      "step": 210
    },
    {
      "epoch": 0.2315500685871056,
      "grad_norm": 0.15357747673988342,
      "learning_rate": 1.7262638717632554e-05,
      "loss": 1.1662,
      "step": 211
    },
    {
      "epoch": 0.2326474622770919,
      "grad_norm": 0.2370174080133438,
      "learning_rate": 1.7237977805178793e-05,
      "loss": 1.163,
      "step": 212
    },
    {
      "epoch": 0.2337448559670782,
      "grad_norm": 0.18010015785694122,
      "learning_rate": 1.7213316892725032e-05,
      "loss": 1.2003,
      "step": 213
    },
    {
      "epoch": 0.23484224965706446,
      "grad_norm": 0.12438640743494034,
      "learning_rate": 1.718865598027127e-05,
      "loss": 1.2254,
      "step": 214
    },
    {
      "epoch": 0.23593964334705075,
      "grad_norm": 0.11115527898073196,
      "learning_rate": 1.716399506781751e-05,
      "loss": 1.1531,
      "step": 215
    },
    {
      "epoch": 0.23703703703703705,
      "grad_norm": 0.19279608130455017,
      "learning_rate": 1.713933415536375e-05,
      "loss": 1.1805,
      "step": 216
    },
    {
      "epoch": 0.2381344307270233,
      "grad_norm": 0.18822820484638214,
      "learning_rate": 1.7114673242909988e-05,
      "loss": 1.2364,
      "step": 217
    },
    {
      "epoch": 0.2392318244170096,
      "grad_norm": 0.11501055210828781,
      "learning_rate": 1.709001233045623e-05,
      "loss": 1.2192,
      "step": 218
    },
    {
      "epoch": 0.2403292181069959,
      "grad_norm": 0.15815867483615875,
      "learning_rate": 1.7065351418002466e-05,
      "loss": 1.1756,
      "step": 219
    },
    {
      "epoch": 0.24142661179698216,
      "grad_norm": 0.11231168359518051,
      "learning_rate": 1.7040690505548708e-05,
      "loss": 1.1561,
      "step": 220
    },
    {
      "epoch": 0.24252400548696845,
      "grad_norm": 0.17530983686447144,
      "learning_rate": 1.7016029593094943e-05,
      "loss": 1.2097,
      "step": 221
    },
    {
      "epoch": 0.24362139917695474,
      "grad_norm": 0.22233904898166656,
      "learning_rate": 1.6991368680641186e-05,
      "loss": 1.0435,
      "step": 222
    },
    {
      "epoch": 0.244718792866941,
      "grad_norm": 0.16978253424167633,
      "learning_rate": 1.6966707768187425e-05,
      "loss": 1.1275,
      "step": 223
    },
    {
      "epoch": 0.2458161865569273,
      "grad_norm": 0.10136917978525162,
      "learning_rate": 1.6942046855733664e-05,
      "loss": 1.2177,
      "step": 224
    },
    {
      "epoch": 0.24691358024691357,
      "grad_norm": 0.14352397620677948,
      "learning_rate": 1.6917385943279902e-05,
      "loss": 1.178,
      "step": 225
    },
    {
      "epoch": 0.24801097393689986,
      "grad_norm": 0.16630198061466217,
      "learning_rate": 1.689272503082614e-05,
      "loss": 1.1465,
      "step": 226
    },
    {
      "epoch": 0.24910836762688615,
      "grad_norm": 0.1705196648836136,
      "learning_rate": 1.686806411837238e-05,
      "loss": 1.2136,
      "step": 227
    },
    {
      "epoch": 0.25020576131687244,
      "grad_norm": 0.14427916705608368,
      "learning_rate": 1.684340320591862e-05,
      "loss": 1.1775,
      "step": 228
    },
    {
      "epoch": 0.25130315500685874,
      "grad_norm": 0.14155656099319458,
      "learning_rate": 1.681874229346486e-05,
      "loss": 1.1594,
      "step": 229
    },
    {
      "epoch": 0.252400548696845,
      "grad_norm": 0.13098452985286713,
      "learning_rate": 1.6794081381011097e-05,
      "loss": 1.202,
      "step": 230
    },
    {
      "epoch": 0.25349794238683127,
      "grad_norm": 0.20060771703720093,
      "learning_rate": 1.676942046855734e-05,
      "loss": 1.1472,
      "step": 231
    },
    {
      "epoch": 0.25459533607681756,
      "grad_norm": 0.22307084500789642,
      "learning_rate": 1.6744759556103578e-05,
      "loss": 1.155,
      "step": 232
    },
    {
      "epoch": 0.25569272976680385,
      "grad_norm": 0.14008598029613495,
      "learning_rate": 1.6720098643649817e-05,
      "loss": 1.244,
      "step": 233
    },
    {
      "epoch": 0.25679012345679014,
      "grad_norm": 0.12508675456047058,
      "learning_rate": 1.6695437731196056e-05,
      "loss": 1.1144,
      "step": 234
    },
    {
      "epoch": 0.2578875171467764,
      "grad_norm": 0.1596231907606125,
      "learning_rate": 1.6670776818742295e-05,
      "loss": 1.0987,
      "step": 235
    },
    {
      "epoch": 0.25898491083676267,
      "grad_norm": 0.14945028722286224,
      "learning_rate": 1.6646115906288534e-05,
      "loss": 1.2141,
      "step": 236
    },
    {
      "epoch": 0.26008230452674896,
      "grad_norm": 0.15245778858661652,
      "learning_rate": 1.6621454993834773e-05,
      "loss": 1.1507,
      "step": 237
    },
    {
      "epoch": 0.26117969821673526,
      "grad_norm": 0.13351744413375854,
      "learning_rate": 1.6596794081381015e-05,
      "loss": 1.1998,
      "step": 238
    },
    {
      "epoch": 0.26227709190672155,
      "grad_norm": 0.10289657860994339,
      "learning_rate": 1.657213316892725e-05,
      "loss": 1.1589,
      "step": 239
    },
    {
      "epoch": 0.26337448559670784,
      "grad_norm": 0.11290785670280457,
      "learning_rate": 1.6547472256473493e-05,
      "loss": 1.1987,
      "step": 240
    },
    {
      "epoch": 0.2644718792866941,
      "grad_norm": 0.1390170007944107,
      "learning_rate": 1.652281134401973e-05,
      "loss": 1.1691,
      "step": 241
    },
    {
      "epoch": 0.26556927297668037,
      "grad_norm": 0.12496920675039291,
      "learning_rate": 1.649815043156597e-05,
      "loss": 1.1337,
      "step": 242
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 0.11996202915906906,
      "learning_rate": 1.647348951911221e-05,
      "loss": 1.118,
      "step": 243
    },
    {
      "epoch": 0.26776406035665296,
      "grad_norm": 0.14449340105056763,
      "learning_rate": 1.644882860665845e-05,
      "loss": 1.1664,
      "step": 244
    },
    {
      "epoch": 0.26886145404663925,
      "grad_norm": 0.11750940978527069,
      "learning_rate": 1.6424167694204687e-05,
      "loss": 1.1768,
      "step": 245
    },
    {
      "epoch": 0.26995884773662554,
      "grad_norm": 0.11455260217189789,
      "learning_rate": 1.6399506781750926e-05,
      "loss": 1.1669,
      "step": 246
    },
    {
      "epoch": 0.2710562414266118,
      "grad_norm": 0.1612052172422409,
      "learning_rate": 1.6374845869297165e-05,
      "loss": 1.0813,
      "step": 247
    },
    {
      "epoch": 0.27215363511659807,
      "grad_norm": 0.1127680242061615,
      "learning_rate": 1.6350184956843404e-05,
      "loss": 1.1253,
      "step": 248
    },
    {
      "epoch": 0.27325102880658436,
      "grad_norm": 0.12000089883804321,
      "learning_rate": 1.6325524044389643e-05,
      "loss": 1.1451,
      "step": 249
    },
    {
      "epoch": 0.27434842249657065,
      "grad_norm": 0.10796555131673813,
      "learning_rate": 1.6300863131935882e-05,
      "loss": 1.2189,
      "step": 250
    },
    {
      "epoch": 0.27544581618655695,
      "grad_norm": 0.11138719320297241,
      "learning_rate": 1.627620221948212e-05,
      "loss": 1.1434,
      "step": 251
    },
    {
      "epoch": 0.2765432098765432,
      "grad_norm": 0.11853396147489548,
      "learning_rate": 1.6251541307028363e-05,
      "loss": 1.1962,
      "step": 252
    },
    {
      "epoch": 0.2776406035665295,
      "grad_norm": 0.11319135874509811,
      "learning_rate": 1.62268803945746e-05,
      "loss": 1.1238,
      "step": 253
    },
    {
      "epoch": 0.27873799725651577,
      "grad_norm": 0.11749457567930222,
      "learning_rate": 1.620221948212084e-05,
      "loss": 1.1505,
      "step": 254
    },
    {
      "epoch": 0.27983539094650206,
      "grad_norm": 0.11317736655473709,
      "learning_rate": 1.617755856966708e-05,
      "loss": 1.2262,
      "step": 255
    },
    {
      "epoch": 0.28093278463648835,
      "grad_norm": 0.12298385798931122,
      "learning_rate": 1.615289765721332e-05,
      "loss": 1.1338,
      "step": 256
    },
    {
      "epoch": 0.28203017832647465,
      "grad_norm": 0.10966061800718307,
      "learning_rate": 1.6128236744759558e-05,
      "loss": 1.1722,
      "step": 257
    },
    {
      "epoch": 0.2831275720164609,
      "grad_norm": 0.11691927909851074,
      "learning_rate": 1.6103575832305797e-05,
      "loss": 1.212,
      "step": 258
    },
    {
      "epoch": 0.2842249657064472,
      "grad_norm": 0.13233579695224762,
      "learning_rate": 1.6078914919852035e-05,
      "loss": 1.1113,
      "step": 259
    },
    {
      "epoch": 0.28532235939643347,
      "grad_norm": 0.10713832825422287,
      "learning_rate": 1.6054254007398274e-05,
      "loss": 1.1638,
      "step": 260
    },
    {
      "epoch": 0.28641975308641976,
      "grad_norm": 0.10320425778627396,
      "learning_rate": 1.6029593094944513e-05,
      "loss": 1.1553,
      "step": 261
    },
    {
      "epoch": 0.28751714677640605,
      "grad_norm": 0.14395684003829956,
      "learning_rate": 1.6004932182490752e-05,
      "loss": 1.1875,
      "step": 262
    },
    {
      "epoch": 0.28861454046639234,
      "grad_norm": 0.11742989718914032,
      "learning_rate": 1.5980271270036994e-05,
      "loss": 1.1434,
      "step": 263
    },
    {
      "epoch": 0.2897119341563786,
      "grad_norm": 0.12088185548782349,
      "learning_rate": 1.595561035758323e-05,
      "loss": 1.1365,
      "step": 264
    },
    {
      "epoch": 0.2908093278463649,
      "grad_norm": 0.12207765877246857,
      "learning_rate": 1.5930949445129472e-05,
      "loss": 1.0191,
      "step": 265
    },
    {
      "epoch": 0.29190672153635117,
      "grad_norm": 0.12891145050525665,
      "learning_rate": 1.590628853267571e-05,
      "loss": 1.1808,
      "step": 266
    },
    {
      "epoch": 0.29300411522633746,
      "grad_norm": 0.12283164262771606,
      "learning_rate": 1.588162762022195e-05,
      "loss": 1.1244,
      "step": 267
    },
    {
      "epoch": 0.29410150891632375,
      "grad_norm": 0.11997072398662567,
      "learning_rate": 1.585696670776819e-05,
      "loss": 1.1481,
      "step": 268
    },
    {
      "epoch": 0.29519890260631,
      "grad_norm": 0.13282917439937592,
      "learning_rate": 1.5832305795314428e-05,
      "loss": 1.1222,
      "step": 269
    },
    {
      "epoch": 0.2962962962962963,
      "grad_norm": 0.11063191294670105,
      "learning_rate": 1.5807644882860667e-05,
      "loss": 1.2377,
      "step": 270
    },
    {
      "epoch": 0.29739368998628257,
      "grad_norm": 0.16821421682834625,
      "learning_rate": 1.5782983970406906e-05,
      "loss": 1.111,
      "step": 271
    },
    {
      "epoch": 0.29849108367626886,
      "grad_norm": 0.1252758502960205,
      "learning_rate": 1.5758323057953148e-05,
      "loss": 1.1257,
      "step": 272
    },
    {
      "epoch": 0.29958847736625516,
      "grad_norm": 0.11989396065473557,
      "learning_rate": 1.5733662145499384e-05,
      "loss": 1.2035,
      "step": 273
    },
    {
      "epoch": 0.30068587105624145,
      "grad_norm": 0.122134268283844,
      "learning_rate": 1.5709001233045626e-05,
      "loss": 1.1897,
      "step": 274
    },
    {
      "epoch": 0.3017832647462277,
      "grad_norm": 0.1424403041601181,
      "learning_rate": 1.5684340320591865e-05,
      "loss": 1.136,
      "step": 275
    },
    {
      "epoch": 0.302880658436214,
      "grad_norm": 0.11523528397083282,
      "learning_rate": 1.5659679408138104e-05,
      "loss": 1.1721,
      "step": 276
    },
    {
      "epoch": 0.30397805212620027,
      "grad_norm": 0.1188586950302124,
      "learning_rate": 1.5635018495684343e-05,
      "loss": 1.196,
      "step": 277
    },
    {
      "epoch": 0.30507544581618656,
      "grad_norm": 0.09858889132738113,
      "learning_rate": 1.561035758323058e-05,
      "loss": 1.1256,
      "step": 278
    },
    {
      "epoch": 0.30617283950617286,
      "grad_norm": 0.13993050158023834,
      "learning_rate": 1.558569667077682e-05,
      "loss": 1.142,
      "step": 279
    },
    {
      "epoch": 0.30727023319615915,
      "grad_norm": 0.11930122971534729,
      "learning_rate": 1.556103575832306e-05,
      "loss": 1.1358,
      "step": 280
    },
    {
      "epoch": 0.3083676268861454,
      "grad_norm": 0.12060839682817459,
      "learning_rate": 1.5536374845869298e-05,
      "loss": 1.2019,
      "step": 281
    },
    {
      "epoch": 0.3094650205761317,
      "grad_norm": 0.11242841929197311,
      "learning_rate": 1.5511713933415537e-05,
      "loss": 1.0959,
      "step": 282
    },
    {
      "epoch": 0.31056241426611797,
      "grad_norm": 0.12150558829307556,
      "learning_rate": 1.5487053020961776e-05,
      "loss": 1.1794,
      "step": 283
    },
    {
      "epoch": 0.31165980795610426,
      "grad_norm": 0.13793016970157623,
      "learning_rate": 1.5462392108508015e-05,
      "loss": 1.0896,
      "step": 284
    },
    {
      "epoch": 0.31275720164609055,
      "grad_norm": 0.1277860403060913,
      "learning_rate": 1.5437731196054254e-05,
      "loss": 1.1495,
      "step": 285
    },
    {
      "epoch": 0.3138545953360768,
      "grad_norm": 0.12210772931575775,
      "learning_rate": 1.5413070283600496e-05,
      "loss": 1.108,
      "step": 286
    },
    {
      "epoch": 0.3149519890260631,
      "grad_norm": 0.11492680013179779,
      "learning_rate": 1.538840937114673e-05,
      "loss": 1.1468,
      "step": 287
    },
    {
      "epoch": 0.3160493827160494,
      "grad_norm": 0.10993566364049911,
      "learning_rate": 1.5363748458692974e-05,
      "loss": 1.0821,
      "step": 288
    },
    {
      "epoch": 0.31714677640603567,
      "grad_norm": 0.11138095706701279,
      "learning_rate": 1.5339087546239213e-05,
      "loss": 1.1024,
      "step": 289
    },
    {
      "epoch": 0.31824417009602196,
      "grad_norm": 0.12479337304830551,
      "learning_rate": 1.5314426633785452e-05,
      "loss": 1.1282,
      "step": 290
    },
    {
      "epoch": 0.31934156378600825,
      "grad_norm": 0.1219845563173294,
      "learning_rate": 1.528976572133169e-05,
      "loss": 1.1404,
      "step": 291
    },
    {
      "epoch": 0.3204389574759945,
      "grad_norm": 0.1182159036397934,
      "learning_rate": 1.526510480887793e-05,
      "loss": 1.173,
      "step": 292
    },
    {
      "epoch": 0.3215363511659808,
      "grad_norm": 0.11335504055023193,
      "learning_rate": 1.5240443896424168e-05,
      "loss": 1.1762,
      "step": 293
    },
    {
      "epoch": 0.3226337448559671,
      "grad_norm": 0.11732617020606995,
      "learning_rate": 1.5215782983970409e-05,
      "loss": 1.1762,
      "step": 294
    },
    {
      "epoch": 0.32373113854595337,
      "grad_norm": 0.1295761615037918,
      "learning_rate": 1.5191122071516646e-05,
      "loss": 1.0506,
      "step": 295
    },
    {
      "epoch": 0.32482853223593966,
      "grad_norm": 0.11054473370313644,
      "learning_rate": 1.5166461159062887e-05,
      "loss": 1.0934,
      "step": 296
    },
    {
      "epoch": 0.32592592592592595,
      "grad_norm": 0.11660335958003998,
      "learning_rate": 1.5141800246609126e-05,
      "loss": 1.0956,
      "step": 297
    },
    {
      "epoch": 0.3270233196159122,
      "grad_norm": 0.10848717391490936,
      "learning_rate": 1.5117139334155365e-05,
      "loss": 1.1147,
      "step": 298
    },
    {
      "epoch": 0.3281207133058985,
      "grad_norm": 0.11791523545980453,
      "learning_rate": 1.5092478421701604e-05,
      "loss": 1.1918,
      "step": 299
    },
    {
      "epoch": 0.3292181069958848,
      "grad_norm": 0.11942354589700699,
      "learning_rate": 1.5067817509247844e-05,
      "loss": 1.2093,
      "step": 300
    },
    {
      "epoch": 0.33031550068587107,
      "grad_norm": 0.11650928854942322,
      "learning_rate": 1.5043156596794081e-05,
      "loss": 1.1171,
      "step": 301
    },
    {
      "epoch": 0.33141289437585736,
      "grad_norm": 0.1390320360660553,
      "learning_rate": 1.5018495684340322e-05,
      "loss": 1.0934,
      "step": 302
    },
    {
      "epoch": 0.3325102880658436,
      "grad_norm": 0.1456214338541031,
      "learning_rate": 1.4993834771886561e-05,
      "loss": 1.0721,
      "step": 303
    },
    {
      "epoch": 0.3336076817558299,
      "grad_norm": 0.11829102784395218,
      "learning_rate": 1.49691738594328e-05,
      "loss": 1.1352,
      "step": 304
    },
    {
      "epoch": 0.3347050754458162,
      "grad_norm": 0.11520489305257797,
      "learning_rate": 1.4944512946979039e-05,
      "loss": 1.1205,
      "step": 305
    },
    {
      "epoch": 0.3358024691358025,
      "grad_norm": 0.1192188486456871,
      "learning_rate": 1.491985203452528e-05,
      "loss": 1.0658,
      "step": 306
    },
    {
      "epoch": 0.33689986282578877,
      "grad_norm": 0.12938235700130463,
      "learning_rate": 1.4895191122071517e-05,
      "loss": 1.2173,
      "step": 307
    },
    {
      "epoch": 0.33799725651577506,
      "grad_norm": 0.11106622219085693,
      "learning_rate": 1.4870530209617757e-05,
      "loss": 1.0982,
      "step": 308
    },
    {
      "epoch": 0.3390946502057613,
      "grad_norm": 0.11311810463666916,
      "learning_rate": 1.4845869297163998e-05,
      "loss": 1.0987,
      "step": 309
    },
    {
      "epoch": 0.3401920438957476,
      "grad_norm": 0.14787402749061584,
      "learning_rate": 1.4821208384710235e-05,
      "loss": 1.2011,
      "step": 310
    },
    {
      "epoch": 0.3412894375857339,
      "grad_norm": 0.12094119936227798,
      "learning_rate": 1.4796547472256476e-05,
      "loss": 1.0972,
      "step": 311
    },
    {
      "epoch": 0.34238683127572017,
      "grad_norm": 0.12174614518880844,
      "learning_rate": 1.4771886559802713e-05,
      "loss": 1.2196,
      "step": 312
    },
    {
      "epoch": 0.34348422496570646,
      "grad_norm": 0.12063402682542801,
      "learning_rate": 1.4747225647348953e-05,
      "loss": 1.0724,
      "step": 313
    },
    {
      "epoch": 0.34458161865569276,
      "grad_norm": 0.1335890144109726,
      "learning_rate": 1.4722564734895192e-05,
      "loss": 1.2439,
      "step": 314
    },
    {
      "epoch": 0.345679012345679,
      "grad_norm": 0.120571069419384,
      "learning_rate": 1.4697903822441431e-05,
      "loss": 1.1374,
      "step": 315
    },
    {
      "epoch": 0.3467764060356653,
      "grad_norm": 0.1246613934636116,
      "learning_rate": 1.467324290998767e-05,
      "loss": 1.1456,
      "step": 316
    },
    {
      "epoch": 0.3478737997256516,
      "grad_norm": 0.1251581907272339,
      "learning_rate": 1.464858199753391e-05,
      "loss": 1.1829,
      "step": 317
    },
    {
      "epoch": 0.34897119341563787,
      "grad_norm": 0.12001658231019974,
      "learning_rate": 1.4623921085080148e-05,
      "loss": 1.2,
      "step": 318
    },
    {
      "epoch": 0.35006858710562416,
      "grad_norm": 0.11571818590164185,
      "learning_rate": 1.4599260172626389e-05,
      "loss": 1.0829,
      "step": 319
    },
    {
      "epoch": 0.3511659807956104,
      "grad_norm": 0.11667678505182266,
      "learning_rate": 1.4574599260172627e-05,
      "loss": 1.1365,
      "step": 320
    },
    {
      "epoch": 0.3522633744855967,
      "grad_norm": 0.13005587458610535,
      "learning_rate": 1.4549938347718866e-05,
      "loss": 1.0685,
      "step": 321
    },
    {
      "epoch": 0.353360768175583,
      "grad_norm": 0.13064813613891602,
      "learning_rate": 1.4525277435265105e-05,
      "loss": 1.1665,
      "step": 322
    },
    {
      "epoch": 0.3544581618655693,
      "grad_norm": 0.11439865827560425,
      "learning_rate": 1.4500616522811346e-05,
      "loss": 1.0968,
      "step": 323
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 0.11633537709712982,
      "learning_rate": 1.4475955610357583e-05,
      "loss": 1.2706,
      "step": 324
    },
    {
      "epoch": 0.35665294924554186,
      "grad_norm": 0.11462391167879105,
      "learning_rate": 1.4451294697903824e-05,
      "loss": 1.2981,
      "step": 325
    },
    {
      "epoch": 0.3577503429355281,
      "grad_norm": 0.11040814965963364,
      "learning_rate": 1.4426633785450064e-05,
      "loss": 1.0935,
      "step": 326
    },
    {
      "epoch": 0.3588477366255144,
      "grad_norm": 0.14056402444839478,
      "learning_rate": 1.4401972872996302e-05,
      "loss": 1.1784,
      "step": 327
    },
    {
      "epoch": 0.3599451303155007,
      "grad_norm": 0.12376075237989426,
      "learning_rate": 1.4377311960542542e-05,
      "loss": 1.1781,
      "step": 328
    },
    {
      "epoch": 0.361042524005487,
      "grad_norm": 0.13468991219997406,
      "learning_rate": 1.4352651048088781e-05,
      "loss": 1.151,
      "step": 329
    },
    {
      "epoch": 0.36213991769547327,
      "grad_norm": 0.11974407732486725,
      "learning_rate": 1.432799013563502e-05,
      "loss": 1.0996,
      "step": 330
    },
    {
      "epoch": 0.36323731138545956,
      "grad_norm": 0.13505803048610687,
      "learning_rate": 1.4303329223181259e-05,
      "loss": 1.0323,
      "step": 331
    },
    {
      "epoch": 0.3643347050754458,
      "grad_norm": 0.13089002668857574,
      "learning_rate": 1.4278668310727498e-05,
      "loss": 1.0352,
      "step": 332
    },
    {
      "epoch": 0.3654320987654321,
      "grad_norm": 0.13586758077144623,
      "learning_rate": 1.4254007398273737e-05,
      "loss": 1.1415,
      "step": 333
    },
    {
      "epoch": 0.3665294924554184,
      "grad_norm": 0.13658414781093597,
      "learning_rate": 1.4229346485819977e-05,
      "loss": 1.2563,
      "step": 334
    },
    {
      "epoch": 0.3676268861454047,
      "grad_norm": 0.13994932174682617,
      "learning_rate": 1.4204685573366214e-05,
      "loss": 1.1559,
      "step": 335
    },
    {
      "epoch": 0.36872427983539097,
      "grad_norm": 0.11985601484775543,
      "learning_rate": 1.4180024660912455e-05,
      "loss": 1.1639,
      "step": 336
    },
    {
      "epoch": 0.3698216735253772,
      "grad_norm": 0.15504884719848633,
      "learning_rate": 1.4155363748458694e-05,
      "loss": 1.0932,
      "step": 337
    },
    {
      "epoch": 0.3709190672153635,
      "grad_norm": 0.11062606424093246,
      "learning_rate": 1.4130702836004933e-05,
      "loss": 1.1327,
      "step": 338
    },
    {
      "epoch": 0.3720164609053498,
      "grad_norm": 0.13901466131210327,
      "learning_rate": 1.4106041923551172e-05,
      "loss": 1.1924,
      "step": 339
    },
    {
      "epoch": 0.3731138545953361,
      "grad_norm": 0.1262989044189453,
      "learning_rate": 1.4081381011097412e-05,
      "loss": 1.1411,
      "step": 340
    },
    {
      "epoch": 0.3742112482853224,
      "grad_norm": 0.1387164443731308,
      "learning_rate": 1.405672009864365e-05,
      "loss": 1.116,
      "step": 341
    },
    {
      "epoch": 0.37530864197530867,
      "grad_norm": 0.13789041340351105,
      "learning_rate": 1.403205918618989e-05,
      "loss": 1.1003,
      "step": 342
    },
    {
      "epoch": 0.3764060356652949,
      "grad_norm": 0.12197256088256836,
      "learning_rate": 1.400739827373613e-05,
      "loss": 1.1363,
      "step": 343
    },
    {
      "epoch": 0.3775034293552812,
      "grad_norm": 0.11622656881809235,
      "learning_rate": 1.3982737361282368e-05,
      "loss": 1.2004,
      "step": 344
    },
    {
      "epoch": 0.3786008230452675,
      "grad_norm": 0.13322965800762177,
      "learning_rate": 1.3958076448828609e-05,
      "loss": 1.0868,
      "step": 345
    },
    {
      "epoch": 0.3796982167352538,
      "grad_norm": 0.13613948225975037,
      "learning_rate": 1.3933415536374848e-05,
      "loss": 1.0893,
      "step": 346
    },
    {
      "epoch": 0.38079561042524007,
      "grad_norm": 0.12749332189559937,
      "learning_rate": 1.3908754623921086e-05,
      "loss": 1.1242,
      "step": 347
    },
    {
      "epoch": 0.38189300411522636,
      "grad_norm": 0.13489745557308197,
      "learning_rate": 1.3884093711467325e-05,
      "loss": 1.0463,
      "step": 348
    },
    {
      "epoch": 0.3829903978052126,
      "grad_norm": 0.13293969631195068,
      "learning_rate": 1.3859432799013564e-05,
      "loss": 1.0682,
      "step": 349
    },
    {
      "epoch": 0.3840877914951989,
      "grad_norm": 0.13386596739292145,
      "learning_rate": 1.3834771886559803e-05,
      "loss": 1.1216,
      "step": 350
    },
    {
      "epoch": 0.3851851851851852,
      "grad_norm": 0.1273600161075592,
      "learning_rate": 1.3810110974106044e-05,
      "loss": 1.1638,
      "step": 351
    },
    {
      "epoch": 0.3862825788751715,
      "grad_norm": 0.113369882106781,
      "learning_rate": 1.3785450061652281e-05,
      "loss": 1.0776,
      "step": 352
    },
    {
      "epoch": 0.38737997256515777,
      "grad_norm": 0.13265594840049744,
      "learning_rate": 1.3760789149198522e-05,
      "loss": 1.1413,
      "step": 353
    },
    {
      "epoch": 0.388477366255144,
      "grad_norm": 0.10907372832298279,
      "learning_rate": 1.373612823674476e-05,
      "loss": 1.1685,
      "step": 354
    },
    {
      "epoch": 0.3895747599451303,
      "grad_norm": 0.14246472716331482,
      "learning_rate": 1.3711467324291e-05,
      "loss": 1.1737,
      "step": 355
    },
    {
      "epoch": 0.3906721536351166,
      "grad_norm": 0.11236773431301117,
      "learning_rate": 1.3686806411837238e-05,
      "loss": 1.2439,
      "step": 356
    },
    {
      "epoch": 0.3917695473251029,
      "grad_norm": 0.14173471927642822,
      "learning_rate": 1.3662145499383479e-05,
      "loss": 1.0831,
      "step": 357
    },
    {
      "epoch": 0.3928669410150892,
      "grad_norm": 0.13366690278053284,
      "learning_rate": 1.3637484586929716e-05,
      "loss": 1.1333,
      "step": 358
    },
    {
      "epoch": 0.39396433470507547,
      "grad_norm": 0.14148059487342834,
      "learning_rate": 1.3612823674475957e-05,
      "loss": 1.1519,
      "step": 359
    },
    {
      "epoch": 0.3950617283950617,
      "grad_norm": 0.11044170707464218,
      "learning_rate": 1.3588162762022197e-05,
      "loss": 1.255,
      "step": 360
    },
    {
      "epoch": 0.396159122085048,
      "grad_norm": 0.11801745742559433,
      "learning_rate": 1.3563501849568435e-05,
      "loss": 1.0922,
      "step": 361
    },
    {
      "epoch": 0.3972565157750343,
      "grad_norm": 0.11835184693336487,
      "learning_rate": 1.3538840937114675e-05,
      "loss": 1.1125,
      "step": 362
    },
    {
      "epoch": 0.3983539094650206,
      "grad_norm": 0.1362667679786682,
      "learning_rate": 1.3514180024660914e-05,
      "loss": 1.1118,
      "step": 363
    },
    {
      "epoch": 0.3994513031550069,
      "grad_norm": 0.1164252832531929,
      "learning_rate": 1.3489519112207153e-05,
      "loss": 1.0671,
      "step": 364
    },
    {
      "epoch": 0.40054869684499317,
      "grad_norm": 0.12106901407241821,
      "learning_rate": 1.3464858199753392e-05,
      "loss": 1.1901,
      "step": 365
    },
    {
      "epoch": 0.4016460905349794,
      "grad_norm": 0.1594018191099167,
      "learning_rate": 1.3440197287299632e-05,
      "loss": 1.0512,
      "step": 366
    },
    {
      "epoch": 0.4027434842249657,
      "grad_norm": 0.11015798151493073,
      "learning_rate": 1.341553637484587e-05,
      "loss": 1.1859,
      "step": 367
    },
    {
      "epoch": 0.403840877914952,
      "grad_norm": 0.1140451431274414,
      "learning_rate": 1.339087546239211e-05,
      "loss": 1.093,
      "step": 368
    },
    {
      "epoch": 0.4049382716049383,
      "grad_norm": 0.1487288475036621,
      "learning_rate": 1.3366214549938348e-05,
      "loss": 1.2052,
      "step": 369
    },
    {
      "epoch": 0.4060356652949246,
      "grad_norm": 0.12370628118515015,
      "learning_rate": 1.3341553637484588e-05,
      "loss": 1.0985,
      "step": 370
    },
    {
      "epoch": 0.4071330589849108,
      "grad_norm": 0.11730250716209412,
      "learning_rate": 1.3316892725030827e-05,
      "loss": 1.082,
      "step": 371
    },
    {
      "epoch": 0.4082304526748971,
      "grad_norm": 0.1203538104891777,
      "learning_rate": 1.3292231812577066e-05,
      "loss": 1.0728,
      "step": 372
    },
    {
      "epoch": 0.4093278463648834,
      "grad_norm": 0.11578691750764847,
      "learning_rate": 1.3267570900123305e-05,
      "loss": 1.1351,
      "step": 373
    },
    {
      "epoch": 0.4104252400548697,
      "grad_norm": 0.1101141944527626,
      "learning_rate": 1.3242909987669545e-05,
      "loss": 1.1311,
      "step": 374
    },
    {
      "epoch": 0.411522633744856,
      "grad_norm": 0.11373890191316605,
      "learning_rate": 1.3218249075215783e-05,
      "loss": 1.0773,
      "step": 375
    },
    {
      "epoch": 0.4126200274348423,
      "grad_norm": 0.14648625254631042,
      "learning_rate": 1.3193588162762023e-05,
      "loss": 1.2263,
      "step": 376
    },
    {
      "epoch": 0.4137174211248285,
      "grad_norm": 0.11010008305311203,
      "learning_rate": 1.3168927250308264e-05,
      "loss": 1.1445,
      "step": 377
    },
    {
      "epoch": 0.4148148148148148,
      "grad_norm": 0.12386184185743332,
      "learning_rate": 1.3144266337854501e-05,
      "loss": 1.0454,
      "step": 378
    },
    {
      "epoch": 0.4159122085048011,
      "grad_norm": 0.11810237169265747,
      "learning_rate": 1.3119605425400742e-05,
      "loss": 1.1877,
      "step": 379
    },
    {
      "epoch": 0.4170096021947874,
      "grad_norm": 0.1291743367910385,
      "learning_rate": 1.309494451294698e-05,
      "loss": 1.2161,
      "step": 380
    },
    {
      "epoch": 0.4181069958847737,
      "grad_norm": 0.115330770611763,
      "learning_rate": 1.307028360049322e-05,
      "loss": 1.1117,
      "step": 381
    },
    {
      "epoch": 0.41920438957476,
      "grad_norm": 0.15565675497055054,
      "learning_rate": 1.3045622688039458e-05,
      "loss": 1.1049,
      "step": 382
    },
    {
      "epoch": 0.4203017832647462,
      "grad_norm": 0.17110465466976166,
      "learning_rate": 1.3020961775585699e-05,
      "loss": 1.1047,
      "step": 383
    },
    {
      "epoch": 0.4213991769547325,
      "grad_norm": 0.12324660271406174,
      "learning_rate": 1.2996300863131936e-05,
      "loss": 1.091,
      "step": 384
    },
    {
      "epoch": 0.4224965706447188,
      "grad_norm": 0.14864134788513184,
      "learning_rate": 1.2971639950678177e-05,
      "loss": 1.1008,
      "step": 385
    },
    {
      "epoch": 0.4235939643347051,
      "grad_norm": 0.15296033024787903,
      "learning_rate": 1.2946979038224414e-05,
      "loss": 1.1242,
      "step": 386
    },
    {
      "epoch": 0.4246913580246914,
      "grad_norm": 0.16041865944862366,
      "learning_rate": 1.2922318125770655e-05,
      "loss": 1.1821,
      "step": 387
    },
    {
      "epoch": 0.4257887517146776,
      "grad_norm": 0.13168294727802277,
      "learning_rate": 1.2897657213316894e-05,
      "loss": 1.1585,
      "step": 388
    },
    {
      "epoch": 0.4268861454046639,
      "grad_norm": 0.11569740623235703,
      "learning_rate": 1.2872996300863132e-05,
      "loss": 1.1275,
      "step": 389
    },
    {
      "epoch": 0.4279835390946502,
      "grad_norm": 0.14467458426952362,
      "learning_rate": 1.2848335388409371e-05,
      "loss": 1.1436,
      "step": 390
    },
    {
      "epoch": 0.4290809327846365,
      "grad_norm": 0.11647368967533112,
      "learning_rate": 1.2823674475955612e-05,
      "loss": 1.1061,
      "step": 391
    },
    {
      "epoch": 0.4301783264746228,
      "grad_norm": 0.11563649028539658,
      "learning_rate": 1.279901356350185e-05,
      "loss": 1.1721,
      "step": 392
    },
    {
      "epoch": 0.4312757201646091,
      "grad_norm": 0.13032524287700653,
      "learning_rate": 1.277435265104809e-05,
      "loss": 1.1382,
      "step": 393
    },
    {
      "epoch": 0.4323731138545953,
      "grad_norm": 0.12809215486049652,
      "learning_rate": 1.274969173859433e-05,
      "loss": 1.1197,
      "step": 394
    },
    {
      "epoch": 0.4334705075445816,
      "grad_norm": 0.11773429811000824,
      "learning_rate": 1.2725030826140568e-05,
      "loss": 1.1025,
      "step": 395
    },
    {
      "epoch": 0.4345679012345679,
      "grad_norm": 0.12838852405548096,
      "learning_rate": 1.2700369913686808e-05,
      "loss": 1.1378,
      "step": 396
    },
    {
      "epoch": 0.4356652949245542,
      "grad_norm": 0.14228248596191406,
      "learning_rate": 1.2675709001233047e-05,
      "loss": 1.2118,
      "step": 397
    },
    {
      "epoch": 0.4367626886145405,
      "grad_norm": 0.14777442812919617,
      "learning_rate": 1.2651048088779286e-05,
      "loss": 1.129,
      "step": 398
    },
    {
      "epoch": 0.4378600823045268,
      "grad_norm": 0.11499614268541336,
      "learning_rate": 1.2626387176325525e-05,
      "loss": 1.1193,
      "step": 399
    },
    {
      "epoch": 0.438957475994513,
      "grad_norm": 0.1274246871471405,
      "learning_rate": 1.2601726263871766e-05,
      "loss": 1.1436,
      "step": 400
    },
    {
      "epoch": 0.4400548696844993,
      "grad_norm": 0.13293783366680145,
      "learning_rate": 1.2577065351418003e-05,
      "loss": 1.0427,
      "step": 401
    },
    {
      "epoch": 0.4411522633744856,
      "grad_norm": 0.12298054248094559,
      "learning_rate": 1.2552404438964243e-05,
      "loss": 1.1429,
      "step": 402
    },
    {
      "epoch": 0.4422496570644719,
      "grad_norm": 0.10508795082569122,
      "learning_rate": 1.2527743526510482e-05,
      "loss": 1.2339,
      "step": 403
    },
    {
      "epoch": 0.4433470507544582,
      "grad_norm": 0.14140458405017853,
      "learning_rate": 1.2503082614056721e-05,
      "loss": 1.0924,
      "step": 404
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 0.16051869094371796,
      "learning_rate": 1.247842170160296e-05,
      "loss": 1.055,
      "step": 405
    },
    {
      "epoch": 0.4455418381344307,
      "grad_norm": 0.12968482077121735,
      "learning_rate": 1.2453760789149199e-05,
      "loss": 1.1362,
      "step": 406
    },
    {
      "epoch": 0.446639231824417,
      "grad_norm": 0.12716621160507202,
      "learning_rate": 1.2429099876695438e-05,
      "loss": 1.0987,
      "step": 407
    },
    {
      "epoch": 0.4477366255144033,
      "grad_norm": 0.15174546837806702,
      "learning_rate": 1.2404438964241678e-05,
      "loss": 1.1899,
      "step": 408
    },
    {
      "epoch": 0.4488340192043896,
      "grad_norm": 0.1363244652748108,
      "learning_rate": 1.2379778051787916e-05,
      "loss": 1.1274,
      "step": 409
    },
    {
      "epoch": 0.4499314128943759,
      "grad_norm": 0.11740902811288834,
      "learning_rate": 1.2355117139334156e-05,
      "loss": 1.0766,
      "step": 410
    },
    {
      "epoch": 0.4510288065843621,
      "grad_norm": 0.1075834259390831,
      "learning_rate": 1.2330456226880397e-05,
      "loss": 1.1318,
      "step": 411
    },
    {
      "epoch": 0.4521262002743484,
      "grad_norm": 0.11810291558504105,
      "learning_rate": 1.2305795314426634e-05,
      "loss": 1.1657,
      "step": 412
    },
    {
      "epoch": 0.4532235939643347,
      "grad_norm": 0.1261415034532547,
      "learning_rate": 1.2281134401972875e-05,
      "loss": 1.2063,
      "step": 413
    },
    {
      "epoch": 0.454320987654321,
      "grad_norm": 0.14952872693538666,
      "learning_rate": 1.2256473489519114e-05,
      "loss": 1.0857,
      "step": 414
    },
    {
      "epoch": 0.4554183813443073,
      "grad_norm": 0.1363765001296997,
      "learning_rate": 1.2231812577065353e-05,
      "loss": 1.1038,
      "step": 415
    },
    {
      "epoch": 0.4565157750342936,
      "grad_norm": 0.1384081095457077,
      "learning_rate": 1.2207151664611591e-05,
      "loss": 1.1109,
      "step": 416
    },
    {
      "epoch": 0.4576131687242798,
      "grad_norm": 0.1204955130815506,
      "learning_rate": 1.2182490752157832e-05,
      "loss": 1.0567,
      "step": 417
    },
    {
      "epoch": 0.4587105624142661,
      "grad_norm": 0.14003603160381317,
      "learning_rate": 1.215782983970407e-05,
      "loss": 1.1006,
      "step": 418
    },
    {
      "epoch": 0.4598079561042524,
      "grad_norm": 0.18735840916633606,
      "learning_rate": 1.213316892725031e-05,
      "loss": 1.1884,
      "step": 419
    },
    {
      "epoch": 0.4609053497942387,
      "grad_norm": 0.16455943882465363,
      "learning_rate": 1.2108508014796549e-05,
      "loss": 1.0925,
      "step": 420
    },
    {
      "epoch": 0.462002743484225,
      "grad_norm": 0.14214913547039032,
      "learning_rate": 1.2083847102342788e-05,
      "loss": 1.0512,
      "step": 421
    },
    {
      "epoch": 0.4631001371742112,
      "grad_norm": 0.12036455422639847,
      "learning_rate": 1.2059186189889027e-05,
      "loss": 1.2031,
      "step": 422
    },
    {
      "epoch": 0.4641975308641975,
      "grad_norm": 0.13500386476516724,
      "learning_rate": 1.2034525277435265e-05,
      "loss": 1.1086,
      "step": 423
    },
    {
      "epoch": 0.4652949245541838,
      "grad_norm": 0.14389222860336304,
      "learning_rate": 1.2009864364981504e-05,
      "loss": 1.1124,
      "step": 424
    },
    {
      "epoch": 0.4663923182441701,
      "grad_norm": 0.14557717740535736,
      "learning_rate": 1.1985203452527745e-05,
      "loss": 1.1043,
      "step": 425
    },
    {
      "epoch": 0.4674897119341564,
      "grad_norm": 0.16723041236400604,
      "learning_rate": 1.1960542540073982e-05,
      "loss": 1.1054,
      "step": 426
    },
    {
      "epoch": 0.4685871056241427,
      "grad_norm": 0.11840852349996567,
      "learning_rate": 1.1935881627620223e-05,
      "loss": 1.2366,
      "step": 427
    },
    {
      "epoch": 0.4696844993141289,
      "grad_norm": 0.13276275992393494,
      "learning_rate": 1.1911220715166463e-05,
      "loss": 1.1297,
      "step": 428
    },
    {
      "epoch": 0.4707818930041152,
      "grad_norm": 0.12433144450187683,
      "learning_rate": 1.18865598027127e-05,
      "loss": 1.1457,
      "step": 429
    },
    {
      "epoch": 0.4718792866941015,
      "grad_norm": 0.1469450742006302,
      "learning_rate": 1.1861898890258941e-05,
      "loss": 1.1337,
      "step": 430
    },
    {
      "epoch": 0.4729766803840878,
      "grad_norm": 0.16274379193782806,
      "learning_rate": 1.183723797780518e-05,
      "loss": 1.1523,
      "step": 431
    },
    {
      "epoch": 0.4740740740740741,
      "grad_norm": 0.13927359879016876,
      "learning_rate": 1.1812577065351419e-05,
      "loss": 1.1721,
      "step": 432
    },
    {
      "epoch": 0.47517146776406033,
      "grad_norm": 0.11743171513080597,
      "learning_rate": 1.1787916152897658e-05,
      "loss": 1.0745,
      "step": 433
    },
    {
      "epoch": 0.4762688614540466,
      "grad_norm": 0.11775142699480057,
      "learning_rate": 1.1763255240443899e-05,
      "loss": 1.152,
      "step": 434
    },
    {
      "epoch": 0.4773662551440329,
      "grad_norm": 0.11214631050825119,
      "learning_rate": 1.1738594327990136e-05,
      "loss": 1.1544,
      "step": 435
    },
    {
      "epoch": 0.4784636488340192,
      "grad_norm": 0.15197938680648804,
      "learning_rate": 1.1713933415536376e-05,
      "loss": 1.0955,
      "step": 436
    },
    {
      "epoch": 0.4795610425240055,
      "grad_norm": 0.1254250705242157,
      "learning_rate": 1.1689272503082615e-05,
      "loss": 1.1479,
      "step": 437
    },
    {
      "epoch": 0.4806584362139918,
      "grad_norm": 0.15701737999916077,
      "learning_rate": 1.1664611590628854e-05,
      "loss": 1.0073,
      "step": 438
    },
    {
      "epoch": 0.48175582990397803,
      "grad_norm": 0.10863196849822998,
      "learning_rate": 1.1639950678175093e-05,
      "loss": 1.1565,
      "step": 439
    },
    {
      "epoch": 0.4828532235939643,
      "grad_norm": 0.11690645664930344,
      "learning_rate": 1.1615289765721334e-05,
      "loss": 1.1264,
      "step": 440
    },
    {
      "epoch": 0.4839506172839506,
      "grad_norm": 0.13278046250343323,
      "learning_rate": 1.1590628853267571e-05,
      "loss": 1.1141,
      "step": 441
    },
    {
      "epoch": 0.4850480109739369,
      "grad_norm": 0.13872136175632477,
      "learning_rate": 1.1565967940813812e-05,
      "loss": 1.1011,
      "step": 442
    },
    {
      "epoch": 0.4861454046639232,
      "grad_norm": 0.12824486196041107,
      "learning_rate": 1.1541307028360049e-05,
      "loss": 1.1151,
      "step": 443
    },
    {
      "epoch": 0.4872427983539095,
      "grad_norm": 0.12683530151844025,
      "learning_rate": 1.151664611590629e-05,
      "loss": 1.0178,
      "step": 444
    },
    {
      "epoch": 0.4883401920438957,
      "grad_norm": 0.12646201252937317,
      "learning_rate": 1.149198520345253e-05,
      "loss": 1.1747,
      "step": 445
    },
    {
      "epoch": 0.489437585733882,
      "grad_norm": 0.12779991328716278,
      "learning_rate": 1.1467324290998767e-05,
      "loss": 1.1784,
      "step": 446
    },
    {
      "epoch": 0.4905349794238683,
      "grad_norm": 0.13384082913398743,
      "learning_rate": 1.1442663378545008e-05,
      "loss": 1.076,
      "step": 447
    },
    {
      "epoch": 0.4916323731138546,
      "grad_norm": 0.13101617991924286,
      "learning_rate": 1.1418002466091247e-05,
      "loss": 1.0109,
      "step": 448
    },
    {
      "epoch": 0.4927297668038409,
      "grad_norm": 0.13373570144176483,
      "learning_rate": 1.1393341553637486e-05,
      "loss": 1.1277,
      "step": 449
    },
    {
      "epoch": 0.49382716049382713,
      "grad_norm": 0.13308392465114594,
      "learning_rate": 1.1368680641183724e-05,
      "loss": 1.0997,
      "step": 450
    },
    {
      "epoch": 0.4949245541838134,
      "grad_norm": 0.11568623036146164,
      "learning_rate": 1.1344019728729965e-05,
      "loss": 1.2082,
      "step": 451
    },
    {
      "epoch": 0.4960219478737997,
      "grad_norm": 0.12799036502838135,
      "learning_rate": 1.1319358816276202e-05,
      "loss": 1.1844,
      "step": 452
    },
    {
      "epoch": 0.497119341563786,
      "grad_norm": 0.12493016570806503,
      "learning_rate": 1.1294697903822443e-05,
      "loss": 1.2041,
      "step": 453
    },
    {
      "epoch": 0.4982167352537723,
      "grad_norm": 0.12631264328956604,
      "learning_rate": 1.1270036991368682e-05,
      "loss": 1.0397,
      "step": 454
    },
    {
      "epoch": 0.4993141289437586,
      "grad_norm": 0.12955130636692047,
      "learning_rate": 1.124537607891492e-05,
      "loss": 1.0504,
      "step": 455
    },
    {
      "epoch": 0.5004115226337449,
      "grad_norm": 0.12372354418039322,
      "learning_rate": 1.122071516646116e-05,
      "loss": 1.1982,
      "step": 456
    },
    {
      "epoch": 0.5015089163237312,
      "grad_norm": 0.13814988732337952,
      "learning_rate": 1.11960542540074e-05,
      "loss": 1.201,
      "step": 457
    },
    {
      "epoch": 0.5026063100137175,
      "grad_norm": 0.11566805094480515,
      "learning_rate": 1.1171393341553637e-05,
      "loss": 1.1334,
      "step": 458
    },
    {
      "epoch": 0.5037037037037037,
      "grad_norm": 0.11871378123760223,
      "learning_rate": 1.1146732429099878e-05,
      "loss": 1.1315,
      "step": 459
    },
    {
      "epoch": 0.50480109739369,
      "grad_norm": 0.12469706684350967,
      "learning_rate": 1.1122071516646115e-05,
      "loss": 1.1309,
      "step": 460
    },
    {
      "epoch": 0.5058984910836762,
      "grad_norm": 0.12486052513122559,
      "learning_rate": 1.1097410604192356e-05,
      "loss": 1.0966,
      "step": 461
    },
    {
      "epoch": 0.5069958847736625,
      "grad_norm": 0.12366752326488495,
      "learning_rate": 1.1072749691738596e-05,
      "loss": 1.1861,
      "step": 462
    },
    {
      "epoch": 0.5080932784636488,
      "grad_norm": 0.1204606145620346,
      "learning_rate": 1.1048088779284834e-05,
      "loss": 1.1,
      "step": 463
    },
    {
      "epoch": 0.5091906721536351,
      "grad_norm": 0.15034319460391998,
      "learning_rate": 1.1023427866831074e-05,
      "loss": 1.1683,
      "step": 464
    },
    {
      "epoch": 0.5102880658436214,
      "grad_norm": 0.1372024267911911,
      "learning_rate": 1.0998766954377313e-05,
      "loss": 1.1134,
      "step": 465
    },
    {
      "epoch": 0.5113854595336077,
      "grad_norm": 0.13857926428318024,
      "learning_rate": 1.097410604192355e-05,
      "loss": 1.1922,
      "step": 466
    },
    {
      "epoch": 0.512482853223594,
      "grad_norm": 0.1584538072347641,
      "learning_rate": 1.0949445129469791e-05,
      "loss": 1.139,
      "step": 467
    },
    {
      "epoch": 0.5135802469135803,
      "grad_norm": 0.14659465849399567,
      "learning_rate": 1.0924784217016032e-05,
      "loss": 1.1054,
      "step": 468
    }
  ],
  "logging_steps": 1,
  "max_steps": 911,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.861679479484129e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}