{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.783969847313565,
  "eval_steps": 500,
  "global_step": 2900,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.006153609476558594,
      "grad_norm": 25468.8046875,
      "learning_rate": 1.9999999999999996e-07,
      "loss": 8.7082,
      "step": 10
    },
    {
      "epoch": 0.012307218953117188,
      "grad_norm": 1228.4501953125,
      "learning_rate": 1.2e-06,
      "loss": 8.0776,
      "step": 20
    },
    {
      "epoch": 0.018460828429675783,
      "grad_norm": 355.1081237792969,
      "learning_rate": 2.1999999999999997e-06,
      "loss": 7.2372,
      "step": 30
    },
    {
      "epoch": 0.024614437906234375,
      "grad_norm": 255.58433532714844,
      "learning_rate": 3.1999999999999994e-06,
      "loss": 6.9725,
      "step": 40
    },
    {
      "epoch": 0.03076804738279297,
      "grad_norm": 253.23995971679688,
      "learning_rate": 4.2e-06,
      "loss": 6.8585,
      "step": 50
    },
    {
      "epoch": 0.036921656859351566,
      "grad_norm": 241.82064819335938,
      "learning_rate": 5.199999999999999e-06,
      "loss": 6.7367,
      "step": 60
    },
    {
      "epoch": 0.043075266335910155,
      "grad_norm": 229.0261993408203,
      "learning_rate": 6.199999999999999e-06,
      "loss": 6.6534,
      "step": 70
    },
    {
      "epoch": 0.04922887581246875,
      "grad_norm": 321.6871337890625,
      "learning_rate": 7.2e-06,
      "loss": 6.6305,
      "step": 80
    },
    {
      "epoch": 0.055382485289027346,
      "grad_norm": 465.00543212890625,
      "learning_rate": 8.2e-06,
      "loss": 6.6253,
      "step": 90
    },
    {
      "epoch": 0.06153609476558594,
      "grad_norm": 566.6929321289062,
      "learning_rate": 9.199999999999998e-06,
      "loss": 6.5724,
      "step": 100
    },
    {
      "epoch": 0.06768970424214453,
      "grad_norm": 432.230712890625,
      "learning_rate": 1.02e-05,
      "loss": 6.5328,
      "step": 110
    },
    {
      "epoch": 0.07384331371870313,
      "grad_norm": 289.2374267578125,
      "learning_rate": 1.12e-05,
      "loss": 6.5547,
      "step": 120
    },
    {
      "epoch": 0.07999692319526172,
      "grad_norm": 595.7386474609375,
      "learning_rate": 1.2199999999999998e-05,
      "loss": 6.5651,
      "step": 130
    },
    {
      "epoch": 0.08615053267182031,
      "grad_norm": 861.4376220703125,
      "learning_rate": 1.3199999999999997e-05,
      "loss": 6.5337,
      "step": 140
    },
    {
      "epoch": 0.09230414214837891,
      "grad_norm": 689.42919921875,
      "learning_rate": 1.4199999999999998e-05,
      "loss": 6.4927,
      "step": 150
    },
    {
      "epoch": 0.0984577516249375,
      "grad_norm": 380.9496765136719,
      "learning_rate": 1.5199999999999998e-05,
      "loss": 6.4812,
      "step": 160
    },
    {
      "epoch": 0.1046113611014961,
      "grad_norm": 221.21005249023438,
      "learning_rate": 1.6199999999999997e-05,
      "loss": 6.4508,
      "step": 170
    },
    {
      "epoch": 0.11076497057805469,
      "grad_norm": 199.30874633789062,
      "learning_rate": 1.7199999999999998e-05,
      "loss": 6.4642,
      "step": 180
    },
    {
      "epoch": 0.11691858005461328,
      "grad_norm": 252.44459533691406,
      "learning_rate": 1.82e-05,
      "loss": 6.45,
      "step": 190
    },
    {
      "epoch": 0.12307218953117188,
      "grad_norm": 121.32320404052734,
      "learning_rate": 1.92e-05,
      "loss": 6.4115,
      "step": 200
    },
    {
      "epoch": 0.12922579900773049,
      "grad_norm": 207.04344177246094,
      "learning_rate": 2.0199999999999996e-05,
      "loss": 6.4532,
      "step": 210
    },
    {
      "epoch": 0.13537940848428906,
      "grad_norm": 296.6274719238281,
      "learning_rate": 2.12e-05,
      "loss": 6.3985,
      "step": 220
    },
    {
      "epoch": 0.14153301796084766,
      "grad_norm": 140.3836212158203,
      "learning_rate": 2.2199999999999998e-05,
      "loss": 6.4152,
      "step": 230
    },
    {
      "epoch": 0.14768662743740626,
      "grad_norm": 168.14443969726562,
      "learning_rate": 2.3199999999999998e-05,
      "loss": 6.3686,
      "step": 240
    },
    {
      "epoch": 0.15384023691396484,
      "grad_norm": 141.07810974121094,
      "learning_rate": 2.42e-05,
      "loss": 6.3809,
      "step": 250
    },
    {
      "epoch": 0.15999384639052344,
      "grad_norm": 117.08052062988281,
      "learning_rate": 2.52e-05,
      "loss": 6.3423,
      "step": 260
    },
    {
      "epoch": 0.16614745586708204,
      "grad_norm": 122.97789764404297,
      "learning_rate": 2.6199999999999996e-05,
      "loss": 6.3486,
      "step": 270
    },
    {
      "epoch": 0.17230106534364062,
      "grad_norm": 73.85701751708984,
      "learning_rate": 2.72e-05,
      "loss": 6.329,
      "step": 280
    },
    {
      "epoch": 0.17845467482019922,
      "grad_norm": 78.00408935546875,
      "learning_rate": 2.8199999999999998e-05,
      "loss": 6.3096,
      "step": 290
    },
    {
      "epoch": 0.18460828429675782,
      "grad_norm": 150.3341827392578,
      "learning_rate": 2.9199999999999995e-05,
      "loss": 6.3063,
      "step": 300
    },
    {
      "epoch": 0.1907618937733164,
      "grad_norm": 61.75702667236328,
      "learning_rate": 3.02e-05,
      "loss": 6.3004,
      "step": 310
    },
    {
      "epoch": 0.196915503249875,
      "grad_norm": 83.836181640625,
      "learning_rate": 3.119999999999999e-05,
      "loss": 6.2534,
      "step": 320
    },
    {
      "epoch": 0.2030691127264336,
      "grad_norm": 190.2839813232422,
      "learning_rate": 3.22e-05,
      "loss": 6.2603,
      "step": 330
    },
    {
      "epoch": 0.2092227222029922,
      "grad_norm": 52.196563720703125,
      "learning_rate": 3.3199999999999994e-05,
      "loss": 6.2439,
      "step": 340
    },
    {
      "epoch": 0.21537633167955078,
      "grad_norm": 65.71736907958984,
      "learning_rate": 3.42e-05,
      "loss": 6.227,
      "step": 350
    },
    {
      "epoch": 0.22152994115610938,
      "grad_norm": 66.71468353271484,
      "learning_rate": 3.5199999999999995e-05,
      "loss": 6.2164,
      "step": 360
    },
    {
      "epoch": 0.22768355063266799,
      "grad_norm": 37.26243209838867,
      "learning_rate": 3.62e-05,
      "loss": 6.212,
      "step": 370
    },
    {
      "epoch": 0.23383716010922656,
      "grad_norm": 190.98501586914062,
      "learning_rate": 3.7199999999999996e-05,
      "loss": 6.1833,
      "step": 380
    },
    {
      "epoch": 0.23999076958578516,
      "grad_norm": 51.51064682006836,
      "learning_rate": 3.8199999999999993e-05,
      "loss": 6.1796,
      "step": 390
    },
    {
      "epoch": 0.24614437906234377,
      "grad_norm": 71.14685821533203,
      "learning_rate": 3.919999999999999e-05,
      "loss": 6.1614,
      "step": 400
    },
    {
      "epoch": 0.25229798853890234,
      "grad_norm": 47.48910140991211,
      "learning_rate": 4.02e-05,
      "loss": 6.1277,
      "step": 410
    },
    {
      "epoch": 0.25845159801546097,
      "grad_norm": 49.60618209838867,
      "learning_rate": 4.12e-05,
      "loss": 6.107,
      "step": 420
    },
    {
      "epoch": 0.26460520749201955,
      "grad_norm": 38.03694534301758,
      "learning_rate": 4.2199999999999996e-05,
      "loss": 6.0939,
      "step": 430
    },
    {
      "epoch": 0.2707588169685781,
      "grad_norm": 38.81101989746094,
      "learning_rate": 4.319999999999999e-05,
      "loss": 6.118,
      "step": 440
    },
    {
      "epoch": 0.27691242644513675,
      "grad_norm": 45.36625671386719,
      "learning_rate": 4.42e-05,
      "loss": 6.0641,
      "step": 450
    },
    {
      "epoch": 0.2830660359216953,
      "grad_norm": 44.09037780761719,
      "learning_rate": 4.5199999999999994e-05,
      "loss": 6.0417,
      "step": 460
    },
    {
      "epoch": 0.2892196453982539,
      "grad_norm": 39.48929214477539,
      "learning_rate": 4.62e-05,
      "loss": 6.0371,
      "step": 470
    },
    {
      "epoch": 0.29537325487481253,
      "grad_norm": 44.747039794921875,
      "learning_rate": 4.7199999999999995e-05,
      "loss": 6.0157,
      "step": 480
    },
    {
      "epoch": 0.3015268643513711,
      "grad_norm": 51.372249603271484,
      "learning_rate": 4.82e-05,
      "loss": 6.0378,
      "step": 490
    },
    {
      "epoch": 0.3076804738279297,
      "grad_norm": 46.31863021850586,
      "learning_rate": 4.9199999999999997e-05,
      "loss": 5.9815,
      "step": 500
    },
    {
      "epoch": 0.3138340833044883,
      "grad_norm": 74.35610961914062,
      "learning_rate": 5.0199999999999994e-05,
      "loss": 6.0108,
      "step": 510
    },
    {
      "epoch": 0.3199876927810469,
      "grad_norm": 31.986116409301758,
      "learning_rate": 5.119999999999999e-05,
      "loss": 5.9452,
      "step": 520
    },
    {
      "epoch": 0.32614130225760546,
      "grad_norm": 62.82748794555664,
      "learning_rate": 5.2199999999999995e-05,
      "loss": 5.9396,
      "step": 530
    },
    {
      "epoch": 0.3322949117341641,
      "grad_norm": 36.623085021972656,
      "learning_rate": 5.32e-05,
      "loss": 5.9078,
      "step": 540
    },
    {
      "epoch": 0.33844852121072266,
      "grad_norm": 66.77005004882812,
      "learning_rate": 5.4199999999999996e-05,
      "loss": 5.9127,
      "step": 550
    },
    {
      "epoch": 0.34460213068728124,
      "grad_norm": 42.2696533203125,
      "learning_rate": 5.519999999999999e-05,
      "loss": 5.8751,
      "step": 560
    },
    {
      "epoch": 0.35075574016383987,
      "grad_norm": 36.03275680541992,
      "learning_rate": 5.619999999999999e-05,
      "loss": 5.8704,
      "step": 570
    },
    {
      "epoch": 0.35690934964039844,
      "grad_norm": 34.99599838256836,
      "learning_rate": 5.72e-05,
      "loss": 5.8357,
      "step": 580
    },
    {
      "epoch": 0.363062959116957,
      "grad_norm": 37.03520584106445,
      "learning_rate": 5.82e-05,
      "loss": 5.8248,
      "step": 590
    },
    {
      "epoch": 0.36921656859351565,
      "grad_norm": 39.70176696777344,
      "learning_rate": 5.9199999999999996e-05,
      "loss": 5.782,
      "step": 600
    },
    {
      "epoch": 0.3753701780700742,
      "grad_norm": 39.41380310058594,
      "learning_rate": 6.019999999999999e-05,
      "loss": 5.7875,
      "step": 610
    },
    {
      "epoch": 0.3815237875466328,
      "grad_norm": 36.758148193359375,
      "learning_rate": 6.12e-05,
      "loss": 5.7705,
      "step": 620
    },
    {
      "epoch": 0.38767739702319143,
      "grad_norm": 47.42232894897461,
      "learning_rate": 6.22e-05,
      "loss": 5.7891,
      "step": 630
    },
    {
      "epoch": 0.39383100649975,
      "grad_norm": 41.998783111572266,
      "learning_rate": 6.319999999999999e-05,
      "loss": 5.735,
      "step": 640
    },
    {
      "epoch": 0.3999846159763086,
      "grad_norm": 35.05141830444336,
      "learning_rate": 6.419999999999999e-05,
      "loss": 5.7239,
      "step": 650
    },
    {
      "epoch": 0.4061382254528672,
      "grad_norm": 38.990840911865234,
      "learning_rate": 6.519999999999999e-05,
      "loss": 5.7106,
      "step": 660
    },
    {
      "epoch": 0.4122918349294258,
      "grad_norm": 38.231101989746094,
      "learning_rate": 6.62e-05,
      "loss": 5.6812,
      "step": 670
    },
    {
      "epoch": 0.4184454444059844,
      "grad_norm": 35.416473388671875,
      "learning_rate": 6.72e-05,
      "loss": 5.6728,
      "step": 680
    },
    {
      "epoch": 0.424599053882543,
      "grad_norm": 45.61351776123047,
      "learning_rate": 6.819999999999999e-05,
      "loss": 5.6484,
      "step": 690
    },
    {
      "epoch": 0.43075266335910156,
      "grad_norm": 41.97138595581055,
      "learning_rate": 6.919999999999999e-05,
      "loss": 5.6279,
      "step": 700
    },
    {
      "epoch": 0.4369062728356602,
      "grad_norm": 35.231204986572266,
      "learning_rate": 7.02e-05,
      "loss": 5.6131,
      "step": 710
    },
    {
      "epoch": 0.44305988231221877,
      "grad_norm": 43.76620864868164,
      "learning_rate": 7.12e-05,
      "loss": 5.576,
      "step": 720
    },
    {
      "epoch": 0.44921349178877734,
      "grad_norm": 34.20133972167969,
      "learning_rate": 7.219999999999999e-05,
      "loss": 5.5689,
      "step": 730
    },
    {
      "epoch": 0.45536710126533597,
      "grad_norm": 37.153480529785156,
      "learning_rate": 7.319999999999999e-05,
      "loss": 5.5499,
      "step": 740
    },
    {
      "epoch": 0.46152071074189455,
      "grad_norm": 55.655784606933594,
      "learning_rate": 7.419999999999999e-05,
      "loss": 5.5353,
      "step": 750
    },
    {
      "epoch": 0.4676743202184531,
      "grad_norm": 33.66733169555664,
      "learning_rate": 7.519999999999998e-05,
      "loss": 5.5018,
      "step": 760
    },
    {
      "epoch": 0.47382792969501175,
      "grad_norm": 33.522857666015625,
      "learning_rate": 7.62e-05,
      "loss": 5.483,
      "step": 770
    },
    {
      "epoch": 0.4799815391715703,
      "grad_norm": 35.31947326660156,
      "learning_rate": 7.72e-05,
      "loss": 5.4719,
      "step": 780
    },
    {
      "epoch": 0.4861351486481289,
      "grad_norm": 34.84073257446289,
      "learning_rate": 7.819999999999999e-05,
      "loss": 5.4676,
      "step": 790
    },
    {
      "epoch": 0.49228875812468753,
      "grad_norm": 31.39910316467285,
      "learning_rate": 7.92e-05,
      "loss": 5.4421,
      "step": 800
    },
    {
      "epoch": 0.4984423676012461,
      "grad_norm": 33.14794158935547,
      "learning_rate": 8.019999999999998e-05,
      "loss": 5.3994,
      "step": 810
    },
    {
      "epoch": 0.5045959770778047,
      "grad_norm": 36.15346145629883,
      "learning_rate": 8.12e-05,
      "loss": 5.3823,
      "step": 820
    },
    {
      "epoch": 0.5107495865543633,
      "grad_norm": 32.183895111083984,
      "learning_rate": 8.22e-05,
      "loss": 5.3428,
      "step": 830
    },
    {
      "epoch": 0.5169031960309219,
      "grad_norm": 32.388916015625,
      "learning_rate": 8.319999999999999e-05,
      "loss": 5.3514,
      "step": 840
    },
    {
      "epoch": 0.5230568055074805,
      "grad_norm": 31.176891326904297,
      "learning_rate": 8.42e-05,
      "loss": 5.3256,
      "step": 850
    },
    {
      "epoch": 0.5292104149840391,
      "grad_norm": 33.613014221191406,
      "learning_rate": 8.519999999999998e-05,
      "loss": 5.3117,
      "step": 860
    },
    {
      "epoch": 0.5353640244605977,
      "grad_norm": 30.59576416015625,
      "learning_rate": 8.62e-05,
      "loss": 5.2776,
      "step": 870
    },
    {
      "epoch": 0.5415176339371562,
      "grad_norm": 34.864505767822266,
      "learning_rate": 8.719999999999999e-05,
      "loss": 5.2651,
      "step": 880
    },
    {
      "epoch": 0.5476712434137149,
      "grad_norm": 33.03851318359375,
      "learning_rate": 8.819999999999999e-05,
      "loss": 5.2467,
      "step": 890
    },
    {
      "epoch": 0.5538248528902735,
      "grad_norm": 36.94324493408203,
      "learning_rate": 8.92e-05,
      "loss": 5.2287,
      "step": 900
    },
    {
      "epoch": 0.559978462366832,
      "grad_norm": 32.03374481201172,
      "learning_rate": 9.02e-05,
      "loss": 5.197,
      "step": 910
    },
    {
      "epoch": 0.5661320718433906,
      "grad_norm": 31.21794891357422,
      "learning_rate": 9.12e-05,
      "loss": 5.1565,
      "step": 920
    },
    {
      "epoch": 0.5722856813199493,
      "grad_norm": 29.664560317993164,
      "learning_rate": 9.219999999999999e-05,
      "loss": 5.1325,
      "step": 930
    },
    {
      "epoch": 0.5784392907965078,
      "grad_norm": 29.547014236450195,
      "learning_rate": 9.319999999999999e-05,
      "loss": 5.1322,
      "step": 940
    },
    {
      "epoch": 0.5845929002730664,
      "grad_norm": 30.9229679107666,
      "learning_rate": 9.419999999999999e-05,
      "loss": 5.1176,
      "step": 950
    },
    {
      "epoch": 0.5907465097496251,
      "grad_norm": 32.148712158203125,
      "learning_rate": 9.52e-05,
      "loss": 5.0631,
      "step": 960
    },
    {
      "epoch": 0.5969001192261836,
      "grad_norm": 30.888355255126953,
      "learning_rate": 9.62e-05,
      "loss": 5.0937,
      "step": 970
    },
    {
      "epoch": 0.6030537287027422,
      "grad_norm": 37.31905746459961,
      "learning_rate": 9.719999999999999e-05,
      "loss": 5.0572,
      "step": 980
    },
    {
      "epoch": 0.6092073381793008,
      "grad_norm": 27.809349060058594,
      "learning_rate": 9.819999999999999e-05,
      "loss": 5.0144,
      "step": 990
    },
    {
      "epoch": 0.6153609476558594,
      "grad_norm": 34.41085433959961,
      "learning_rate": 9.919999999999999e-05,
      "loss": 5.0256,
      "step": 1000
    },
    {
      "epoch": 0.621514557132418,
      "grad_norm": 29.413028717041016,
      "learning_rate": 0.0001002,
      "loss": 4.9648,
      "step": 1010
    },
    {
      "epoch": 0.6276681666089766,
      "grad_norm": 30.73712158203125,
      "learning_rate": 0.0001012,
      "loss": 4.9744,
      "step": 1020
    },
    {
      "epoch": 0.6338217760855351,
      "grad_norm": 30.328628540039062,
      "learning_rate": 0.00010219999999999999,
      "loss": 4.9165,
      "step": 1030
    },
    {
      "epoch": 0.6399753855620938,
      "grad_norm": 29.132736206054688,
      "learning_rate": 0.00010319999999999999,
      "loss": 4.9574,
      "step": 1040
    },
    {
      "epoch": 0.6461289950386524,
      "grad_norm": 28.395519256591797,
      "learning_rate": 0.00010419999999999998,
      "loss": 4.8759,
      "step": 1050
    },
    {
      "epoch": 0.6522826045152109,
      "grad_norm": 26.395824432373047,
      "learning_rate": 0.0001052,
      "loss": 4.9002,
      "step": 1060
    },
    {
      "epoch": 0.6584362139917695,
      "grad_norm": 27.33197593688965,
      "learning_rate": 0.00010619999999999998,
      "loss": 4.8743,
      "step": 1070
    },
    {
      "epoch": 0.6645898234683282,
      "grad_norm": 25.773096084594727,
      "learning_rate": 0.00010719999999999999,
      "loss": 4.8093,
      "step": 1080
    },
    {
      "epoch": 0.6707434329448867,
      "grad_norm": 27.487308502197266,
      "learning_rate": 0.0001082,
      "loss": 4.8178,
      "step": 1090
    },
    {
      "epoch": 0.6768970424214453,
      "grad_norm": 26.95354652404785,
      "learning_rate": 0.00010919999999999998,
      "loss": 4.7463,
      "step": 1100
    },
    {
      "epoch": 0.683050651898004,
      "grad_norm": 30.955183029174805,
      "learning_rate": 0.0001102,
      "loss": 4.7438,
      "step": 1110
    },
    {
      "epoch": 0.6892042613745625,
      "grad_norm": 29.22182846069336,
      "learning_rate": 0.00011119999999999998,
      "loss": 4.739,
      "step": 1120
    },
    {
      "epoch": 0.6953578708511211,
      "grad_norm": 25.642513275146484,
      "learning_rate": 0.00011219999999999999,
      "loss": 4.6881,
      "step": 1130
    },
    {
      "epoch": 0.7015114803276797,
      "grad_norm": 28.86591148376465,
      "learning_rate": 0.0001132,
      "loss": 4.6853,
      "step": 1140
    },
    {
      "epoch": 0.7076650898042383,
      "grad_norm": 28.62666130065918,
      "learning_rate": 0.00011419999999999998,
      "loss": 4.6572,
      "step": 1150
    },
    {
      "epoch": 0.7138186992807969,
      "grad_norm": 27.118507385253906,
      "learning_rate": 0.0001152,
      "loss": 4.6118,
      "step": 1160
    },
    {
      "epoch": 0.7199723087573555,
      "grad_norm": 28.83325958251953,
      "learning_rate": 0.00011619999999999998,
      "loss": 4.6325,
      "step": 1170
    },
    {
      "epoch": 0.726125918233914,
      "grad_norm": 27.323766708374023,
      "learning_rate": 0.00011719999999999999,
      "loss": 4.5598,
      "step": 1180
    },
    {
      "epoch": 0.7322795277104727,
      "grad_norm": 28.03806495666504,
      "learning_rate": 0.0001182,
      "loss": 4.5404,
      "step": 1190
    },
    {
      "epoch": 0.7384331371870313,
      "grad_norm": 30.81092071533203,
      "learning_rate": 0.00011919999999999998,
      "loss": 4.5604,
      "step": 1200
    },
    {
      "epoch": 0.7445867466635898,
      "grad_norm": 26.44489288330078,
      "learning_rate": 0.0001202,
      "loss": 4.5447,
      "step": 1210
    },
    {
      "epoch": 0.7507403561401484,
      "grad_norm": 27.878007888793945,
      "learning_rate": 0.00012119999999999999,
      "loss": 4.5259,
      "step": 1220
    },
    {
      "epoch": 0.7568939656167071,
      "grad_norm": 29.378477096557617,
      "learning_rate": 0.0001222,
      "loss": 4.4411,
      "step": 1230
    },
    {
      "epoch": 0.7630475750932656,
      "grad_norm": 25.017864227294922,
      "learning_rate": 0.00012319999999999999,
      "loss": 4.4634,
      "step": 1240
    },
    {
      "epoch": 0.7692011845698242,
      "grad_norm": 26.963115692138672,
      "learning_rate": 0.00012419999999999998,
      "loss": 4.4299,
      "step": 1250
    },
    {
      "epoch": 0.7753547940463829,
      "grad_norm": 28.014739990234375,
      "learning_rate": 0.00012519999999999998,
      "loss": 4.4148,
      "step": 1260
    },
    {
      "epoch": 0.7815084035229414,
      "grad_norm": 27.028656005859375,
      "learning_rate": 0.0001262,
      "loss": 4.4078,
      "step": 1270
    },
    {
      "epoch": 0.7876620129995,
      "grad_norm": 28.67365264892578,
      "learning_rate": 0.00012719999999999997,
      "loss": 4.3355,
      "step": 1280
    },
    {
      "epoch": 0.7938156224760586,
      "grad_norm": 26.66254997253418,
      "learning_rate": 0.0001282,
      "loss": 4.3166,
      "step": 1290
    },
    {
      "epoch": 0.7999692319526172,
      "grad_norm": 27.120641708374023,
      "learning_rate": 0.00012919999999999997,
      "loss": 4.2582,
      "step": 1300
    },
    {
      "epoch": 0.8061228414291758,
      "grad_norm": 24.28302001953125,
      "learning_rate": 0.0001302,
      "loss": 4.2967,
      "step": 1310
    },
    {
      "epoch": 0.8122764509057344,
      "grad_norm": 27.492734909057617,
      "learning_rate": 0.0001312,
      "loss": 4.263,
      "step": 1320
    },
    {
      "epoch": 0.8184300603822929,
      "grad_norm": 22.861173629760742,
      "learning_rate": 0.0001322,
      "loss": 4.2387,
      "step": 1330
    },
    {
      "epoch": 0.8245836698588516,
      "grad_norm": 24.373743057250977,
      "learning_rate": 0.00013319999999999999,
      "loss": 4.2242,
      "step": 1340
    },
    {
      "epoch": 0.8307372793354102,
      "grad_norm": 28.591249465942383,
      "learning_rate": 0.0001342,
      "loss": 4.1666,
      "step": 1350
    },
    {
      "epoch": 0.8368908888119688,
      "grad_norm": 24.015104293823242,
      "learning_rate": 0.00013519999999999998,
      "loss": 4.1729,
      "step": 1360
    },
    {
      "epoch": 0.8430444982885273,
      "grad_norm": 28.374052047729492,
      "learning_rate": 0.0001362,
      "loss": 4.1558,
      "step": 1370
    },
    {
      "epoch": 0.849198107765086,
      "grad_norm": 24.058826446533203,
      "learning_rate": 0.00013719999999999997,
      "loss": 4.086,
      "step": 1380
    },
    {
      "epoch": 0.8553517172416446,
      "grad_norm": 24.95467185974121,
      "learning_rate": 0.0001382,
      "loss": 4.0687,
      "step": 1390
    },
    {
      "epoch": 0.8615053267182031,
      "grad_norm": 26.37350082397461,
      "learning_rate": 0.0001392,
      "loss": 4.0863,
      "step": 1400
    },
    {
      "epoch": 0.8676589361947618,
      "grad_norm": 25.355846405029297,
      "learning_rate": 0.0001402,
      "loss": 4.0348,
      "step": 1410
    },
    {
      "epoch": 0.8738125456713204,
      "grad_norm": 24.209243774414062,
      "learning_rate": 0.0001412,
      "loss": 4.0416,
      "step": 1420
    },
    {
      "epoch": 0.8799661551478789,
      "grad_norm": 27.465925216674805,
      "learning_rate": 0.0001422,
      "loss": 3.9939,
      "step": 1430
    },
    {
      "epoch": 0.8861197646244375,
      "grad_norm": 25.085203170776367,
      "learning_rate": 0.00014319999999999998,
      "loss": 4.0192,
      "step": 1440
    },
    {
      "epoch": 0.8922733741009962,
      "grad_norm": 23.539377212524414,
      "learning_rate": 0.00014419999999999998,
      "loss": 3.9236,
      "step": 1450
    },
    {
      "epoch": 0.8984269835775547,
      "grad_norm": 26.392433166503906,
      "learning_rate": 0.00014519999999999998,
      "loss": 3.9303,
      "step": 1460
    },
    {
      "epoch": 0.9045805930541133,
      "grad_norm": 23.23287582397461,
      "learning_rate": 0.0001462,
      "loss": 3.9163,
      "step": 1470
    },
    {
      "epoch": 0.9107342025306719,
      "grad_norm": 24.316932678222656,
      "learning_rate": 0.00014719999999999997,
      "loss": 3.868,
      "step": 1480
    },
    {
      "epoch": 0.9168878120072305,
      "grad_norm": 23.149797439575195,
      "learning_rate": 0.0001482,
      "loss": 3.8912,
      "step": 1490
    },
    {
      "epoch": 0.9230414214837891,
      "grad_norm": 25.656810760498047,
      "learning_rate": 0.0001492,
      "loss": 3.7883,
      "step": 1500
    },
    {
      "epoch": 0.9291950309603477,
      "grad_norm": 24.43971824645996,
      "learning_rate": 0.0001502,
      "loss": 3.8087,
      "step": 1510
    },
    {
      "epoch": 0.9353486404369062,
      "grad_norm": 24.23967170715332,
      "learning_rate": 0.0001512,
      "loss": 3.7772,
      "step": 1520
    },
    {
      "epoch": 0.9415022499134649,
      "grad_norm": 33.623390197753906,
      "learning_rate": 0.00015219999999999999,
      "loss": 3.7577,
      "step": 1530
    },
    {
      "epoch": 0.9476558593900235,
      "grad_norm": 21.980667114257812,
      "learning_rate": 0.0001532,
      "loss": 3.7378,
      "step": 1540
    },
    {
      "epoch": 0.953809468866582,
      "grad_norm": 22.828100204467773,
      "learning_rate": 0.00015419999999999998,
      "loss": 3.701,
      "step": 1550
    },
    {
      "epoch": 0.9599630783431407,
      "grad_norm": 22.391563415527344,
      "learning_rate": 0.00015519999999999998,
      "loss": 3.6866,
      "step": 1560
    },
    {
      "epoch": 0.9661166878196993,
      "grad_norm": 21.836578369140625,
      "learning_rate": 0.0001562,
      "loss": 3.6618,
      "step": 1570
    },
    {
      "epoch": 0.9722702972962578,
      "grad_norm": 24.34965705871582,
      "learning_rate": 0.0001572,
      "loss": 3.6196,
      "step": 1580
    },
    {
      "epoch": 0.9784239067728164,
      "grad_norm": 23.68498420715332,
      "learning_rate": 0.00015819999999999997,
      "loss": 3.5995,
      "step": 1590
    },
    {
      "epoch": 0.9845775162493751,
      "grad_norm": 23.469165802001953,
      "learning_rate": 0.00015919999999999997,
      "loss": 3.602,
      "step": 1600
    },
    {
      "epoch": 0.9907311257259336,
      "grad_norm": 22.707569122314453,
      "learning_rate": 0.0001602,
      "loss": 3.548,
      "step": 1610
    },
    {
      "epoch": 0.9968847352024922,
      "grad_norm": 23.006725311279297,
      "learning_rate": 0.0001612,
      "loss": 3.5996,
      "step": 1620
    },
    {
      "epoch": 1.0024614437906234,
      "grad_norm": 24.067211151123047,
      "learning_rate": 0.00016219999999999999,
      "loss": 3.5286,
      "step": 1630
    },
    {
      "epoch": 1.008615053267182,
      "grad_norm": 23.17717933654785,
      "learning_rate": 0.0001632,
      "loss": 3.4575,
      "step": 1640
    },
    {
      "epoch": 1.0147686627437407,
      "grad_norm": 21.80594825744629,
      "learning_rate": 0.00016419999999999998,
      "loss": 3.4402,
      "step": 1650
    },
    {
      "epoch": 1.0209222722202993,
      "grad_norm": 23.56287384033203,
      "learning_rate": 0.00016519999999999998,
      "loss": 3.418,
      "step": 1660
    },
    {
      "epoch": 1.0270758816968577,
      "grad_norm": 22.383155822753906,
      "learning_rate": 0.0001662,
      "loss": 3.4202,
      "step": 1670
    },
    {
      "epoch": 1.0332294911734163,
      "grad_norm": 22.10567855834961,
      "learning_rate": 0.0001672,
      "loss": 3.3831,
      "step": 1680
    },
    {
      "epoch": 1.039383100649975,
      "grad_norm": 21.35976219177246,
      "learning_rate": 0.00016819999999999997,
      "loss": 3.3813,
      "step": 1690
    },
    {
      "epoch": 1.0455367101265336,
      "grad_norm": 22.603801727294922,
      "learning_rate": 0.00016919999999999997,
      "loss": 3.3525,
      "step": 1700
    },
    {
      "epoch": 1.0516903196030922,
      "grad_norm": 20.92445945739746,
      "learning_rate": 0.0001702,
      "loss": 3.3775,
      "step": 1710
    },
    {
      "epoch": 1.0578439290796509,
      "grad_norm": 22.412351608276367,
      "learning_rate": 0.0001712,
      "loss": 3.3344,
      "step": 1720
    },
    {
      "epoch": 1.0639975385562095,
      "grad_norm": 20.877464294433594,
      "learning_rate": 0.00017219999999999998,
      "loss": 3.3102,
      "step": 1730
    },
    {
      "epoch": 1.070151148032768,
      "grad_norm": 20.799846649169922,
      "learning_rate": 0.00017319999999999998,
      "loss": 3.251,
      "step": 1740
    },
    {
      "epoch": 1.0763047575093265,
      "grad_norm": 19.017059326171875,
      "learning_rate": 0.00017419999999999998,
      "loss": 3.3153,
      "step": 1750
    },
    {
      "epoch": 1.0824583669858852,
      "grad_norm": 24.539533615112305,
      "learning_rate": 0.00017519999999999998,
      "loss": 3.2335,
      "step": 1760
    },
    {
      "epoch": 1.0886119764624438,
      "grad_norm": 20.221553802490234,
      "learning_rate": 0.0001762,
      "loss": 3.2131,
      "step": 1770
    },
    {
      "epoch": 1.0947655859390024,
      "grad_norm": 19.945335388183594,
      "learning_rate": 0.0001772,
      "loss": 3.1835,
      "step": 1780
    },
    {
      "epoch": 1.1009191954155608,
      "grad_norm": 19.56667709350586,
      "learning_rate": 0.00017819999999999997,
      "loss": 3.2059,
      "step": 1790
    },
    {
      "epoch": 1.1070728048921195,
      "grad_norm": 18.606618881225586,
      "learning_rate": 0.0001792,
      "loss": 3.2107,
      "step": 1800
    },
    {
      "epoch": 1.113226414368678,
      "grad_norm": 19.815696716308594,
      "learning_rate": 0.0001802,
      "loss": 3.131,
      "step": 1810
    },
    {
      "epoch": 1.1193800238452367,
      "grad_norm": 19.790706634521484,
      "learning_rate": 0.00018119999999999999,
      "loss": 3.1329,
      "step": 1820
    },
    {
      "epoch": 1.1255336333217953,
      "grad_norm": 22.01865005493164,
      "learning_rate": 0.00018219999999999996,
      "loss": 3.1121,
      "step": 1830
    },
    {
      "epoch": 1.131687242798354,
      "grad_norm": 18.54840850830078,
      "learning_rate": 0.00018319999999999998,
      "loss": 3.0815,
      "step": 1840
    },
    {
      "epoch": 1.1378408522749126,
      "grad_norm": 20.705862045288086,
      "learning_rate": 0.00018419999999999998,
      "loss": 3.0422,
      "step": 1850
    },
    {
      "epoch": 1.143994461751471,
      "grad_norm": 19.450159072875977,
      "learning_rate": 0.00018519999999999998,
      "loss": 3.0255,
      "step": 1860
    },
    {
      "epoch": 1.1501480712280296,
      "grad_norm": 19.23040199279785,
      "learning_rate": 0.0001862,
      "loss": 3.0009,
      "step": 1870
    },
    {
      "epoch": 1.1563016807045883,
      "grad_norm": 18.69268035888672,
      "learning_rate": 0.0001872,
      "loss": 3.0006,
      "step": 1880
    },
    {
      "epoch": 1.162455290181147,
      "grad_norm": 20.293420791625977,
      "learning_rate": 0.00018819999999999997,
      "loss": 2.9436,
      "step": 1890
    },
    {
      "epoch": 1.1686088996577055,
      "grad_norm": 18.272842407226562,
      "learning_rate": 0.0001892,
      "loss": 2.9509,
      "step": 1900
    },
    {
      "epoch": 1.174762509134264,
      "grad_norm": 18.34984016418457,
      "learning_rate": 0.0001902,
      "loss": 2.9567,
      "step": 1910
    },
    {
      "epoch": 1.1809161186108226,
      "grad_norm": 17.388120651245117,
      "learning_rate": 0.00019119999999999999,
      "loss": 2.9164,
      "step": 1920
    },
    {
      "epoch": 1.1870697280873812,
      "grad_norm": 18.276813507080078,
      "learning_rate": 0.0001922,
      "loss": 2.8638,
      "step": 1930
    },
    {
      "epoch": 1.1932233375639398,
      "grad_norm": 17.28123664855957,
      "learning_rate": 0.00019319999999999998,
      "loss": 2.9009,
      "step": 1940
    },
    {
      "epoch": 1.1993769470404985,
      "grad_norm": 16.94942855834961,
      "learning_rate": 0.00019419999999999998,
      "loss": 2.9128,
      "step": 1950
    },
    {
      "epoch": 1.205530556517057,
      "grad_norm": 18.007675170898438,
      "learning_rate": 0.00019519999999999997,
      "loss": 2.8609,
      "step": 1960
    },
    {
      "epoch": 1.2116841659936157,
      "grad_norm": 18.41819953918457,
      "learning_rate": 0.0001962,
      "loss": 2.8666,
      "step": 1970
    },
    {
      "epoch": 1.2178377754701741,
      "grad_norm": 17.115619659423828,
      "learning_rate": 0.00019719999999999997,
      "loss": 2.8805,
      "step": 1980
    },
    {
      "epoch": 1.2239913849467328,
      "grad_norm": 17.493419647216797,
      "learning_rate": 0.00019819999999999997,
      "loss": 2.8557,
      "step": 1990
    },
    {
      "epoch": 1.2301449944232914,
      "grad_norm": 18.269655227661133,
      "learning_rate": 0.0001992,
      "loss": 2.8357,
      "step": 2000
    },
    {
      "epoch": 1.23629860389985,
      "grad_norm": 16.41655731201172,
      "learning_rate": 0.0002002,
      "loss": 2.7481,
      "step": 2010
    },
    {
      "epoch": 1.2424522133764087,
      "grad_norm": 19.108936309814453,
      "learning_rate": 0.00020119999999999998,
      "loss": 2.7697,
      "step": 2020
    },
    {
      "epoch": 1.2486058228529673,
      "grad_norm": 17.761171340942383,
      "learning_rate": 0.0002022,
      "loss": 2.8015,
      "step": 2030
    },
    {
      "epoch": 1.254759432329526,
      "grad_norm": 17.668210983276367,
      "learning_rate": 0.00020319999999999998,
      "loss": 2.7173,
      "step": 2040
    },
    {
      "epoch": 1.2609130418060843,
      "grad_norm": 18.238393783569336,
      "learning_rate": 0.00020419999999999998,
      "loss": 2.7466,
      "step": 2050
    },
    {
      "epoch": 1.267066651282643,
      "grad_norm": 18.7877140045166,
      "learning_rate": 0.0002052,
      "loss": 2.7203,
      "step": 2060
    },
    {
      "epoch": 1.2732202607592016,
      "grad_norm": 15.977209091186523,
      "learning_rate": 0.0002062,
      "loss": 2.6307,
      "step": 2070
    },
    {
      "epoch": 1.2793738702357602,
      "grad_norm": 17.05807113647461,
      "learning_rate": 0.00020719999999999997,
      "loss": 2.6672,
      "step": 2080
    },
    {
      "epoch": 1.2855274797123188,
      "grad_norm": 19.629629135131836,
      "learning_rate": 0.00020819999999999996,
      "loss": 2.6817,
      "step": 2090
    },
    {
      "epoch": 1.2916810891888773,
      "grad_norm": 18.23768424987793,
      "learning_rate": 0.0002092,
      "loss": 2.6663,
      "step": 2100
    },
    {
      "epoch": 1.2978346986654359,
      "grad_norm": 16.28844451904297,
      "learning_rate": 0.0002102,
      "loss": 2.6131,
      "step": 2110
    },
    {
      "epoch": 1.3039883081419945,
      "grad_norm": 17.443166732788086,
      "learning_rate": 0.00021119999999999996,
      "loss": 2.6489,
      "step": 2120
    },
    {
      "epoch": 1.3101419176185531,
      "grad_norm": 16.729419708251953,
      "learning_rate": 0.00021219999999999998,
      "loss": 2.5928,
      "step": 2130
    },
    {
      "epoch": 1.3162955270951118,
      "grad_norm": 17.59943389892578,
      "learning_rate": 0.00021319999999999998,
      "loss": 2.6301,
      "step": 2140
    },
    {
      "epoch": 1.3224491365716704,
      "grad_norm": 16.870969772338867,
      "learning_rate": 0.00021419999999999998,
      "loss": 2.5733,
      "step": 2150
    },
    {
      "epoch": 1.328602746048229,
      "grad_norm": 15.78957462310791,
      "learning_rate": 0.0002152,
      "loss": 2.5533,
      "step": 2160
    },
    {
      "epoch": 1.3347563555247874,
      "grad_norm": 16.400348663330078,
      "learning_rate": 0.0002162,
      "loss": 2.554,
      "step": 2170
    },
    {
      "epoch": 1.340909965001346,
      "grad_norm": 16.348743438720703,
      "learning_rate": 0.00021719999999999997,
      "loss": 2.5985,
      "step": 2180
    },
    {
      "epoch": 1.3470635744779047,
      "grad_norm": 15.286760330200195,
      "learning_rate": 0.0002182,
      "loss": 2.4854,
      "step": 2190
    },
    {
      "epoch": 1.3532171839544633,
      "grad_norm": 18.324291229248047,
      "learning_rate": 0.0002192,
      "loss": 2.5129,
      "step": 2200
    },
    {
      "epoch": 1.359370793431022,
      "grad_norm": 15.46131420135498,
      "learning_rate": 0.00022019999999999999,
      "loss": 2.4918,
      "step": 2210
    },
    {
      "epoch": 1.3655244029075804,
      "grad_norm": 17.585025787353516,
      "learning_rate": 0.00022119999999999996,
      "loss": 2.492,
      "step": 2220
    },
    {
      "epoch": 1.371678012384139,
      "grad_norm": 17.009857177734375,
      "learning_rate": 0.00022219999999999998,
      "loss": 2.5181,
      "step": 2230
    },
    {
      "epoch": 1.3778316218606976,
      "grad_norm": 16.09518814086914,
      "learning_rate": 0.00022319999999999998,
      "loss": 2.4728,
      "step": 2240
    },
    {
      "epoch": 1.3839852313372563,
      "grad_norm": 17.182323455810547,
      "learning_rate": 0.00022419999999999997,
      "loss": 2.4996,
      "step": 2250
    },
    {
      "epoch": 1.390138840813815,
      "grad_norm": 18.38797950744629,
      "learning_rate": 0.0002252,
      "loss": 2.4687,
      "step": 2260
    },
    {
      "epoch": 1.3962924502903735,
      "grad_norm": 16.63583755493164,
      "learning_rate": 0.00022619999999999997,
      "loss": 2.4667,
      "step": 2270
    },
    {
      "epoch": 1.4024460597669322,
      "grad_norm": 16.635272979736328,
      "learning_rate": 0.00022719999999999997,
      "loss": 2.4259,
      "step": 2280
    },
    {
      "epoch": 1.4085996692434906,
      "grad_norm": 15.283349990844727,
      "learning_rate": 0.0002282,
      "loss": 2.447,
      "step": 2290
    },
    {
      "epoch": 1.4147532787200492,
      "grad_norm": 16.514480590820312,
      "learning_rate": 0.0002292,
      "loss": 2.401,
      "step": 2300
    },
    {
      "epoch": 1.4209068881966078,
      "grad_norm": 15.541825294494629,
      "learning_rate": 0.00023019999999999998,
      "loss": 2.3892,
      "step": 2310
    },
    {
      "epoch": 1.4270604976731665,
      "grad_norm": 16.92156219482422,
      "learning_rate": 0.0002312,
      "loss": 2.3836,
      "step": 2320
    },
    {
      "epoch": 1.433214107149725,
      "grad_norm": 17.072193145751953,
      "learning_rate": 0.00023219999999999998,
      "loss": 2.4165,
      "step": 2330
    },
    {
      "epoch": 1.4393677166262835,
      "grad_norm": 17.85093879699707,
      "learning_rate": 0.00023319999999999998,
      "loss": 2.402,
      "step": 2340
    },
    {
      "epoch": 1.4455213261028421,
      "grad_norm": 16.710161209106445,
      "learning_rate": 0.00023419999999999997,
      "loss": 2.3802,
      "step": 2350
    },
    {
      "epoch": 1.4516749355794007,
      "grad_norm": 16.6106014251709,
      "learning_rate": 0.0002352,
      "loss": 2.3677,
      "step": 2360
    },
    {
      "epoch": 1.4578285450559594,
      "grad_norm": 16.886838912963867,
      "learning_rate": 0.00023619999999999997,
      "loss": 2.3263,
      "step": 2370
    },
    {
      "epoch": 1.463982154532518,
      "grad_norm": 15.4957275390625,
      "learning_rate": 0.00023719999999999997,
      "loss": 2.3548,
      "step": 2380
    },
    {
      "epoch": 1.4701357640090766,
      "grad_norm": 16.018007278442383,
      "learning_rate": 0.0002382,
      "loss": 2.3748,
      "step": 2390
    },
    {
      "epoch": 1.4762893734856353,
      "grad_norm": 16.532024383544922,
      "learning_rate": 0.0002392,
      "loss": 2.3739,
      "step": 2400
    },
    {
      "epoch": 1.4824429829621937,
      "grad_norm": 17.2122745513916,
      "learning_rate": 0.00024019999999999996,
      "loss": 2.3544,
      "step": 2410
    },
    {
      "epoch": 1.4885965924387523,
      "grad_norm": 15.620635986328125,
      "learning_rate": 0.00024119999999999998,
      "loss": 2.3239,
      "step": 2420
    },
    {
      "epoch": 1.494750201915311,
      "grad_norm": 16.69614601135254,
      "learning_rate": 0.00024219999999999998,
      "loss": 2.3095,
      "step": 2430
    },
    {
      "epoch": 1.5009038113918696,
      "grad_norm": 15.723666191101074,
      "learning_rate": 0.00024319999999999998,
      "loss": 2.3264,
      "step": 2440
    },
    {
      "epoch": 1.5070574208684282,
      "grad_norm": 16.057676315307617,
      "learning_rate": 0.00024419999999999997,
      "loss": 2.323,
      "step": 2450
    },
    {
      "epoch": 1.5132110303449866,
      "grad_norm": 16.74864387512207,
      "learning_rate": 0.0002452,
      "loss": 2.2934,
      "step": 2460
    },
    {
      "epoch": 1.5193646398215455,
      "grad_norm": 16.351608276367188,
      "learning_rate": 0.00024619999999999997,
      "loss": 2.2719,
      "step": 2470
    },
    {
      "epoch": 1.5255182492981039,
      "grad_norm": 16.09889793395996,
      "learning_rate": 0.0002472,
      "loss": 2.2935,
      "step": 2480
    },
    {
      "epoch": 1.5316718587746625,
      "grad_norm": 17.544286727905273,
      "learning_rate": 0.00024819999999999996,
      "loss": 2.299,
      "step": 2490
    },
    {
      "epoch": 1.5378254682512211,
      "grad_norm": 15.155010223388672,
      "learning_rate": 0.0002492,
      "loss": 2.2966,
      "step": 2500
    },
    {
      "epoch": 1.5439790777277795,
      "grad_norm": 17.26972007751465,
      "learning_rate": 0.00025019999999999996,
      "loss": 2.2372,
      "step": 2510
    },
    {
      "epoch": 1.5501326872043384,
      "grad_norm": 17.257909774780273,
      "learning_rate": 0.0002512,
      "loss": 2.2774,
      "step": 2520
    },
    {
      "epoch": 1.5562862966808968,
      "grad_norm": 16.608787536621094,
      "learning_rate": 0.0002522,
      "loss": 2.2761,
      "step": 2530
    },
    {
      "epoch": 1.5624399061574554,
      "grad_norm": 18.28546905517578,
      "learning_rate": 0.0002532,
      "loss": 2.2391,
      "step": 2540
    },
    {
      "epoch": 1.568593515634014,
      "grad_norm": 15.028303146362305,
      "learning_rate": 0.0002542,
      "loss": 2.2567,
      "step": 2550
    },
    {
      "epoch": 1.5747471251105727,
      "grad_norm": 14.921636581420898,
      "learning_rate": 0.00025519999999999997,
      "loss": 2.2349,
      "step": 2560
    },
    {
      "epoch": 1.5809007345871313,
      "grad_norm": 15.385894775390625,
      "learning_rate": 0.0002562,
      "loss": 2.2452,
      "step": 2570
    },
    {
      "epoch": 1.5870543440636897,
      "grad_norm": 14.949525833129883,
      "learning_rate": 0.00025719999999999996,
      "loss": 2.2142,
      "step": 2580
    },
    {
      "epoch": 1.5932079535402486,
      "grad_norm": 16.841659545898438,
      "learning_rate": 0.0002582,
      "loss": 2.253,
      "step": 2590
    },
    {
      "epoch": 1.599361563016807,
      "grad_norm": 15.079891204833984,
      "learning_rate": 0.00025919999999999996,
      "loss": 2.2083,
      "step": 2600
    },
    {
      "epoch": 1.6055151724933656,
      "grad_norm": 15.801860809326172,
      "learning_rate": 0.0002602,
      "loss": 2.1959,
      "step": 2610
    },
    {
      "epoch": 1.6116687819699242,
      "grad_norm": 15.776305198669434,
      "learning_rate": 0.0002612,
      "loss": 2.1853,
      "step": 2620
    },
    {
      "epoch": 1.6178223914464829,
      "grad_norm": 16.545124053955078,
      "learning_rate": 0.0002622,
      "loss": 2.2363,
      "step": 2630
    },
    {
      "epoch": 1.6239760009230415,
      "grad_norm": 16.241558074951172,
      "learning_rate": 0.00026319999999999995,
      "loss": 2.2085,
      "step": 2640
    },
    {
      "epoch": 1.6301296103996,
      "grad_norm": 14.833049774169922,
      "learning_rate": 0.00026419999999999997,
      "loss": 2.1958,
      "step": 2650
    },
    {
      "epoch": 1.6362832198761588,
      "grad_norm": 15.921625137329102,
      "learning_rate": 0.0002652,
      "loss": 2.1964,
      "step": 2660
    },
    {
      "epoch": 1.6424368293527172,
      "grad_norm": 15.138516426086426,
      "learning_rate": 0.00026619999999999997,
      "loss": 2.1665,
      "step": 2670
    },
    {
      "epoch": 1.6485904388292758,
      "grad_norm": 14.26587200164795,
      "learning_rate": 0.0002672,
      "loss": 2.1491,
      "step": 2680
    },
    {
      "epoch": 1.6547440483058344,
      "grad_norm": 15.675385475158691,
      "learning_rate": 0.00026819999999999996,
      "loss": 2.1486,
      "step": 2690
    },
    {
      "epoch": 1.6608976577823928,
      "grad_norm": 15.808439254760742,
      "learning_rate": 0.0002692,
      "loss": 2.1817,
      "step": 2700
    },
    {
      "epoch": 1.6670512672589517,
      "grad_norm": 15.16848373413086,
      "learning_rate": 0.00027019999999999995,
      "loss": 2.1407,
      "step": 2710
    },
    {
      "epoch": 1.67320487673551,
      "grad_norm": 14.977187156677246,
      "learning_rate": 0.0002712,
      "loss": 2.17,
      "step": 2720
    },
    {
      "epoch": 1.6793584862120687,
      "grad_norm": 16.03354835510254,
      "learning_rate": 0.00027219999999999995,
      "loss": 2.1803,
      "step": 2730
    },
    {
      "epoch": 1.6855120956886274,
      "grad_norm": 55.20576095581055,
      "learning_rate": 0.00027319999999999997,
      "loss": 2.1433,
      "step": 2740
    },
    {
      "epoch": 1.691665705165186,
      "grad_norm": 15.511900901794434,
      "learning_rate": 0.0002742,
      "loss": 2.1162,
      "step": 2750
    },
    {
      "epoch": 1.6978193146417446,
      "grad_norm": 15.09477710723877,
      "learning_rate": 0.00027519999999999997,
      "loss": 2.0881,
      "step": 2760
    },
    {
      "epoch": 1.703972924118303,
      "grad_norm": 15.021590232849121,
      "learning_rate": 0.0002762,
      "loss": 2.1426,
      "step": 2770
    },
    {
      "epoch": 1.7101265335948619,
      "grad_norm": 15.347617149353027,
      "learning_rate": 0.0002772,
      "loss": 2.1274,
      "step": 2780
    },
    {
      "epoch": 1.7162801430714203,
      "grad_norm": 14.517563819885254,
      "learning_rate": 0.0002782,
      "loss": 2.1127,
      "step": 2790
    },
    {
      "epoch": 1.722433752547979,
      "grad_norm": 14.70672607421875,
      "learning_rate": 0.00027919999999999996,
      "loss": 2.1338,
      "step": 2800
    },
    {
      "epoch": 1.7285873620245376,
      "grad_norm": 14.38133716583252,
      "learning_rate": 0.0002802,
      "loss": 2.1007,
      "step": 2810
    },
    {
      "epoch": 1.734740971501096,
      "grad_norm": 14.325138092041016,
      "learning_rate": 0.0002812,
      "loss": 2.1358,
      "step": 2820
    },
    {
      "epoch": 1.7408945809776548,
      "grad_norm": 14.518105506896973,
      "learning_rate": 0.0002822,
      "loss": 2.1302,
      "step": 2830
    },
    {
      "epoch": 1.7470481904542132,
      "grad_norm": 14.69653606414795,
      "learning_rate": 0.00028319999999999994,
      "loss": 2.1173,
      "step": 2840
    },
    {
      "epoch": 1.7532017999307719,
      "grad_norm": 16.654796600341797,
      "learning_rate": 0.00028419999999999997,
      "loss": 2.0933,
      "step": 2850
    },
    {
      "epoch": 1.7593554094073305,
      "grad_norm": 15.500261306762695,
      "learning_rate": 0.0002852,
      "loss": 2.0765,
      "step": 2860
    },
    {
      "epoch": 1.7655090188838891,
      "grad_norm": 13.33629035949707,
      "learning_rate": 0.00028619999999999996,
      "loss": 2.0961,
      "step": 2870
    },
    {
      "epoch": 1.7716626283604477,
      "grad_norm": 15.36388874053955,
      "learning_rate": 0.0002872,
      "loss": 2.1031,
      "step": 2880
    },
    {
      "epoch": 1.7778162378370062,
      "grad_norm": 13.345674514770508,
      "learning_rate": 0.00028819999999999996,
      "loss": 2.0589,
      "step": 2890
    },
    {
      "epoch": 1.783969847313565,
      "grad_norm": 14.782434463500977,
      "learning_rate": 0.0002892,
      "loss": 2.0996,
      "step": 2900
    }
  ],
  "logging_steps": 10,
  "max_steps": 100000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 62,
  "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": 2.195286371893248e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}