{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 3241,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.000308546744831842,
      "grad_norm": 3.8197367191314697,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 4.3327,
      "step": 1
    },
    {
      "epoch": 0.000617093489663684,
      "grad_norm": 3.611845016479492,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 4.2669,
      "step": 2
    },
    {
      "epoch": 0.0009256402344955261,
      "grad_norm": 5.70314359664917,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 5.4971,
      "step": 3
    },
    {
      "epoch": 0.001234186979327368,
      "grad_norm": 3.392024517059326,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 4.1838,
      "step": 4
    },
    {
      "epoch": 0.0015427337241592102,
      "grad_norm": 4.388611316680908,
      "learning_rate": 5e-07,
      "loss": 4.8521,
      "step": 5
    },
    {
      "epoch": 0.0018512804689910522,
      "grad_norm": 3.260829210281372,
      "learning_rate": 6.000000000000001e-07,
      "loss": 4.0387,
      "step": 6
    },
    {
      "epoch": 0.0021598272138228943,
      "grad_norm": 6.02992057800293,
      "learning_rate": 7.000000000000001e-07,
      "loss": 5.9947,
      "step": 7
    },
    {
      "epoch": 0.002468373958654736,
      "grad_norm": 4.1300554275512695,
      "learning_rate": 8.000000000000001e-07,
      "loss": 4.6317,
      "step": 8
    },
    {
      "epoch": 0.002776920703486578,
      "grad_norm": 4.799147605895996,
      "learning_rate": 9e-07,
      "loss": 4.9666,
      "step": 9
    },
    {
      "epoch": 0.0030854674483184203,
      "grad_norm": 3.797393798828125,
      "learning_rate": 1e-06,
      "loss": 4.4269,
      "step": 10
    },
    {
      "epoch": 0.003394014193150262,
      "grad_norm": 3.5223026275634766,
      "learning_rate": 1.1e-06,
      "loss": 4.1057,
      "step": 11
    },
    {
      "epoch": 0.0037025609379821045,
      "grad_norm": 4.609493255615234,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 5.0402,
      "step": 12
    },
    {
      "epoch": 0.004011107682813947,
      "grad_norm": 2.8431596755981445,
      "learning_rate": 1.3e-06,
      "loss": 3.7792,
      "step": 13
    },
    {
      "epoch": 0.004319654427645789,
      "grad_norm": 5.449502944946289,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 5.3713,
      "step": 14
    },
    {
      "epoch": 0.0046282011724776305,
      "grad_norm": 4.320559501647949,
      "learning_rate": 1.5e-06,
      "loss": 4.6332,
      "step": 15
    },
    {
      "epoch": 0.004936747917309472,
      "grad_norm": 4.168860912322998,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 4.5831,
      "step": 16
    },
    {
      "epoch": 0.005245294662141314,
      "grad_norm": 3.160491943359375,
      "learning_rate": 1.7e-06,
      "loss": 4.0278,
      "step": 17
    },
    {
      "epoch": 0.005553841406973156,
      "grad_norm": 4.066011428833008,
      "learning_rate": 1.8e-06,
      "loss": 4.7029,
      "step": 18
    },
    {
      "epoch": 0.005862388151804999,
      "grad_norm": 5.592798233032227,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 5.9086,
      "step": 19
    },
    {
      "epoch": 0.006170934896636841,
      "grad_norm": 2.722200870513916,
      "learning_rate": 2e-06,
      "loss": 3.5714,
      "step": 20
    },
    {
      "epoch": 0.0064794816414686825,
      "grad_norm": 6.213608741760254,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 5.5262,
      "step": 21
    },
    {
      "epoch": 0.006788028386300524,
      "grad_norm": 6.304937839508057,
      "learning_rate": 2.2e-06,
      "loss": 5.9214,
      "step": 22
    },
    {
      "epoch": 0.007096575131132366,
      "grad_norm": 4.180171012878418,
      "learning_rate": 2.3e-06,
      "loss": 4.5205,
      "step": 23
    },
    {
      "epoch": 0.007405121875964209,
      "grad_norm": 3.755070209503174,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 4.2664,
      "step": 24
    },
    {
      "epoch": 0.007713668620796051,
      "grad_norm": 2.4497368335723877,
      "learning_rate": 2.4999999999999998e-06,
      "loss": 3.497,
      "step": 25
    },
    {
      "epoch": 0.008022215365627893,
      "grad_norm": 4.276626110076904,
      "learning_rate": 2.6e-06,
      "loss": 4.7956,
      "step": 26
    },
    {
      "epoch": 0.008330762110459735,
      "grad_norm": 4.274627208709717,
      "learning_rate": 2.7e-06,
      "loss": 4.9595,
      "step": 27
    },
    {
      "epoch": 0.008639308855291577,
      "grad_norm": 2.989254951477051,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 4.0622,
      "step": 28
    },
    {
      "epoch": 0.008947855600123419,
      "grad_norm": 4.647172451019287,
      "learning_rate": 2.9e-06,
      "loss": 5.7777,
      "step": 29
    },
    {
      "epoch": 0.009256402344955261,
      "grad_norm": 3.569882869720459,
      "learning_rate": 3e-06,
      "loss": 4.9304,
      "step": 30
    },
    {
      "epoch": 0.009564949089787103,
      "grad_norm": 2.1073899269104004,
      "learning_rate": 3.1e-06,
      "loss": 3.8024,
      "step": 31
    },
    {
      "epoch": 0.009873495834618945,
      "grad_norm": 1.0976536273956299,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 2.873,
      "step": 32
    },
    {
      "epoch": 0.010182042579450786,
      "grad_norm": 2.685044527053833,
      "learning_rate": 3.3e-06,
      "loss": 4.6695,
      "step": 33
    },
    {
      "epoch": 0.010490589324282628,
      "grad_norm": 1.5810956954956055,
      "learning_rate": 3.4e-06,
      "loss": 3.5091,
      "step": 34
    },
    {
      "epoch": 0.01079913606911447,
      "grad_norm": 1.168277621269226,
      "learning_rate": 3.5e-06,
      "loss": 3.1201,
      "step": 35
    },
    {
      "epoch": 0.011107682813946312,
      "grad_norm": 1.4001338481903076,
      "learning_rate": 3.6e-06,
      "loss": 3.4514,
      "step": 36
    },
    {
      "epoch": 0.011416229558778156,
      "grad_norm": 1.0257220268249512,
      "learning_rate": 3.7e-06,
      "loss": 2.9701,
      "step": 37
    },
    {
      "epoch": 0.011724776303609998,
      "grad_norm": 1.1348458528518677,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 3.2321,
      "step": 38
    },
    {
      "epoch": 0.01203332304844184,
      "grad_norm": 1.5109179019927979,
      "learning_rate": 3.9e-06,
      "loss": 3.8248,
      "step": 39
    },
    {
      "epoch": 0.012341869793273681,
      "grad_norm": 1.1383541822433472,
      "learning_rate": 4e-06,
      "loss": 3.2195,
      "step": 40
    },
    {
      "epoch": 0.012650416538105523,
      "grad_norm": 1.4988147020339966,
      "learning_rate": 4.1e-06,
      "loss": 3.7383,
      "step": 41
    },
    {
      "epoch": 0.012958963282937365,
      "grad_norm": 0.7626239061355591,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 2.8485,
      "step": 42
    },
    {
      "epoch": 0.013267510027769207,
      "grad_norm": 1.1359997987747192,
      "learning_rate": 4.3e-06,
      "loss": 3.3959,
      "step": 43
    },
    {
      "epoch": 0.013576056772601049,
      "grad_norm": 1.991447925567627,
      "learning_rate": 4.4e-06,
      "loss": 4.5384,
      "step": 44
    },
    {
      "epoch": 0.01388460351743289,
      "grad_norm": 1.2268764972686768,
      "learning_rate": 4.5e-06,
      "loss": 3.5075,
      "step": 45
    },
    {
      "epoch": 0.014193150262264732,
      "grad_norm": 0.9016637802124023,
      "learning_rate": 4.6e-06,
      "loss": 3.2113,
      "step": 46
    },
    {
      "epoch": 0.014501697007096576,
      "grad_norm": 1.3222975730895996,
      "learning_rate": 4.700000000000001e-06,
      "loss": 3.9122,
      "step": 47
    },
    {
      "epoch": 0.014810243751928418,
      "grad_norm": 0.5912073850631714,
      "learning_rate": 4.800000000000001e-06,
      "loss": 2.8042,
      "step": 48
    },
    {
      "epoch": 0.01511879049676026,
      "grad_norm": 1.091262936592102,
      "learning_rate": 4.9e-06,
      "loss": 3.5044,
      "step": 49
    },
    {
      "epoch": 0.015427337241592102,
      "grad_norm": 1.0276159048080444,
      "learning_rate": 4.9999999999999996e-06,
      "loss": 3.5583,
      "step": 50
    },
    {
      "epoch": 0.01573588398642394,
      "grad_norm": 1.0583761930465698,
      "learning_rate": 5.1e-06,
      "loss": 3.8359,
      "step": 51
    },
    {
      "epoch": 0.016044430731255787,
      "grad_norm": 1.0619677305221558,
      "learning_rate": 5.2e-06,
      "loss": 3.8607,
      "step": 52
    },
    {
      "epoch": 0.01635297747608763,
      "grad_norm": 0.8387205004692078,
      "learning_rate": 5.3e-06,
      "loss": 3.232,
      "step": 53
    },
    {
      "epoch": 0.01666152422091947,
      "grad_norm": 1.1671696901321411,
      "learning_rate": 5.4e-06,
      "loss": 4.1826,
      "step": 54
    },
    {
      "epoch": 0.016970070965751313,
      "grad_norm": 0.8083593249320984,
      "learning_rate": 5.5e-06,
      "loss": 3.4237,
      "step": 55
    },
    {
      "epoch": 0.017278617710583154,
      "grad_norm": 0.99895179271698,
      "learning_rate": 5.600000000000001e-06,
      "loss": 3.8469,
      "step": 56
    },
    {
      "epoch": 0.017587164455414996,
      "grad_norm": 0.7100948095321655,
      "learning_rate": 5.7000000000000005e-06,
      "loss": 3.2944,
      "step": 57
    },
    {
      "epoch": 0.017895711200246838,
      "grad_norm": 0.6428288221359253,
      "learning_rate": 5.8e-06,
      "loss": 3.0704,
      "step": 58
    },
    {
      "epoch": 0.01820425794507868,
      "grad_norm": 0.8618065118789673,
      "learning_rate": 5.899999999999999e-06,
      "loss": 3.8168,
      "step": 59
    },
    {
      "epoch": 0.018512804689910522,
      "grad_norm": 0.7153589725494385,
      "learning_rate": 6e-06,
      "loss": 3.5421,
      "step": 60
    },
    {
      "epoch": 0.018821351434742364,
      "grad_norm": 0.5318284630775452,
      "learning_rate": 6.1e-06,
      "loss": 3.1603,
      "step": 61
    },
    {
      "epoch": 0.019129898179574206,
      "grad_norm": 0.348753422498703,
      "learning_rate": 6.2e-06,
      "loss": 2.692,
      "step": 62
    },
    {
      "epoch": 0.019438444924406047,
      "grad_norm": 0.5460414290428162,
      "learning_rate": 6.3e-06,
      "loss": 3.08,
      "step": 63
    },
    {
      "epoch": 0.01974699166923789,
      "grad_norm": 0.9770793318748474,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 3.4058,
      "step": 64
    },
    {
      "epoch": 0.02005553841406973,
      "grad_norm": 0.55307936668396,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 2.9789,
      "step": 65
    },
    {
      "epoch": 0.020364085158901573,
      "grad_norm": 0.7234558463096619,
      "learning_rate": 6.6e-06,
      "loss": 3.3505,
      "step": 66
    },
    {
      "epoch": 0.020672631903733415,
      "grad_norm": 0.5592244863510132,
      "learning_rate": 6.7e-06,
      "loss": 3.1349,
      "step": 67
    },
    {
      "epoch": 0.020981178648565257,
      "grad_norm": 0.8060568571090698,
      "learning_rate": 6.8e-06,
      "loss": 3.7522,
      "step": 68
    },
    {
      "epoch": 0.0212897253933971,
      "grad_norm": 0.6033865809440613,
      "learning_rate": 6.900000000000001e-06,
      "loss": 3.5573,
      "step": 69
    },
    {
      "epoch": 0.02159827213822894,
      "grad_norm": 0.5775837302207947,
      "learning_rate": 7e-06,
      "loss": 3.3392,
      "step": 70
    },
    {
      "epoch": 0.021906818883060782,
      "grad_norm": 0.591890811920166,
      "learning_rate": 7.1e-06,
      "loss": 3.1473,
      "step": 71
    },
    {
      "epoch": 0.022215365627892624,
      "grad_norm": 0.6378189325332642,
      "learning_rate": 7.2e-06,
      "loss": 3.2234,
      "step": 72
    },
    {
      "epoch": 0.02252391237272447,
      "grad_norm": 0.5335432887077332,
      "learning_rate": 7.3e-06,
      "loss": 3.0607,
      "step": 73
    },
    {
      "epoch": 0.02283245911755631,
      "grad_norm": 0.4598855674266815,
      "learning_rate": 7.4e-06,
      "loss": 3.0668,
      "step": 74
    },
    {
      "epoch": 0.023141005862388153,
      "grad_norm": 0.5637675523757935,
      "learning_rate": 7.5e-06,
      "loss": 3.3607,
      "step": 75
    },
    {
      "epoch": 0.023449552607219995,
      "grad_norm": 0.6915042996406555,
      "learning_rate": 7.600000000000001e-06,
      "loss": 3.1566,
      "step": 76
    },
    {
      "epoch": 0.023758099352051837,
      "grad_norm": 0.44307759404182434,
      "learning_rate": 7.699999999999999e-06,
      "loss": 3.1364,
      "step": 77
    },
    {
      "epoch": 0.02406664609688368,
      "grad_norm": 0.43005770444869995,
      "learning_rate": 7.8e-06,
      "loss": 3.1683,
      "step": 78
    },
    {
      "epoch": 0.02437519284171552,
      "grad_norm": 0.6191526055335999,
      "learning_rate": 7.899999999999999e-06,
      "loss": 3.3363,
      "step": 79
    },
    {
      "epoch": 0.024683739586547362,
      "grad_norm": 0.35434824228286743,
      "learning_rate": 8e-06,
      "loss": 2.5193,
      "step": 80
    },
    {
      "epoch": 0.024992286331379204,
      "grad_norm": 0.656644344329834,
      "learning_rate": 8.1e-06,
      "loss": 3.7158,
      "step": 81
    },
    {
      "epoch": 0.025300833076211046,
      "grad_norm": 0.40206533670425415,
      "learning_rate": 8.2e-06,
      "loss": 2.7845,
      "step": 82
    },
    {
      "epoch": 0.025609379821042888,
      "grad_norm": 0.5501867532730103,
      "learning_rate": 8.3e-06,
      "loss": 2.7171,
      "step": 83
    },
    {
      "epoch": 0.02591792656587473,
      "grad_norm": 0.4792475998401642,
      "learning_rate": 8.400000000000001e-06,
      "loss": 3.0822,
      "step": 84
    },
    {
      "epoch": 0.02622647331070657,
      "grad_norm": 0.3702377676963806,
      "learning_rate": 8.5e-06,
      "loss": 2.6305,
      "step": 85
    },
    {
      "epoch": 0.026535020055538414,
      "grad_norm": 0.37643131613731384,
      "learning_rate": 8.6e-06,
      "loss": 2.6588,
      "step": 86
    },
    {
      "epoch": 0.026843566800370255,
      "grad_norm": 0.604988157749176,
      "learning_rate": 8.7e-06,
      "loss": 3.3621,
      "step": 87
    },
    {
      "epoch": 0.027152113545202097,
      "grad_norm": 1.1603400707244873,
      "learning_rate": 8.8e-06,
      "loss": 2.8903,
      "step": 88
    },
    {
      "epoch": 0.02746066029003394,
      "grad_norm": 0.6672317981719971,
      "learning_rate": 8.900000000000001e-06,
      "loss": 3.202,
      "step": 89
    },
    {
      "epoch": 0.02776920703486578,
      "grad_norm": 0.404778391122818,
      "learning_rate": 9e-06,
      "loss": 3.0092,
      "step": 90
    },
    {
      "epoch": 0.028077753779697623,
      "grad_norm": 0.7324615716934204,
      "learning_rate": 9.100000000000001e-06,
      "loss": 3.3472,
      "step": 91
    },
    {
      "epoch": 0.028386300524529465,
      "grad_norm": 0.5391788482666016,
      "learning_rate": 9.2e-06,
      "loss": 2.9677,
      "step": 92
    },
    {
      "epoch": 0.028694847269361307,
      "grad_norm": 0.4934414327144623,
      "learning_rate": 9.3e-06,
      "loss": 3.1551,
      "step": 93
    },
    {
      "epoch": 0.029003394014193152,
      "grad_norm": 0.6892746686935425,
      "learning_rate": 9.400000000000001e-06,
      "loss": 3.6399,
      "step": 94
    },
    {
      "epoch": 0.029311940759024994,
      "grad_norm": 0.40343931317329407,
      "learning_rate": 9.5e-06,
      "loss": 2.6643,
      "step": 95
    },
    {
      "epoch": 0.029620487503856836,
      "grad_norm": 0.37016206979751587,
      "learning_rate": 9.600000000000001e-06,
      "loss": 2.8084,
      "step": 96
    },
    {
      "epoch": 0.029929034248688677,
      "grad_norm": 0.31692177057266235,
      "learning_rate": 9.699999999999999e-06,
      "loss": 2.5649,
      "step": 97
    },
    {
      "epoch": 0.03023758099352052,
      "grad_norm": 0.48820602893829346,
      "learning_rate": 9.8e-06,
      "loss": 2.7648,
      "step": 98
    },
    {
      "epoch": 0.03054612773835236,
      "grad_norm": 0.4864860773086548,
      "learning_rate": 9.9e-06,
      "loss": 2.778,
      "step": 99
    },
    {
      "epoch": 0.030854674483184203,
      "grad_norm": 0.4217410087585449,
      "learning_rate": 9.999999999999999e-06,
      "loss": 2.6541,
      "step": 100
    },
    {
      "epoch": 0.031163221228016045,
      "grad_norm": 0.41729119420051575,
      "learning_rate": 1.01e-05,
      "loss": 2.5602,
      "step": 101
    },
    {
      "epoch": 0.03147176797284788,
      "grad_norm": 0.5560382604598999,
      "learning_rate": 1.02e-05,
      "loss": 3.1223,
      "step": 102
    },
    {
      "epoch": 0.031780314717679725,
      "grad_norm": 0.3572539985179901,
      "learning_rate": 1.03e-05,
      "loss": 2.706,
      "step": 103
    },
    {
      "epoch": 0.032088861462511574,
      "grad_norm": 0.49619024991989136,
      "learning_rate": 1.04e-05,
      "loss": 3.0865,
      "step": 104
    },
    {
      "epoch": 0.032397408207343416,
      "grad_norm": 0.4121740758419037,
      "learning_rate": 1.05e-05,
      "loss": 2.9064,
      "step": 105
    },
    {
      "epoch": 0.03270595495217526,
      "grad_norm": 0.5392472743988037,
      "learning_rate": 1.06e-05,
      "loss": 3.0245,
      "step": 106
    },
    {
      "epoch": 0.0330145016970071,
      "grad_norm": 0.37481802701950073,
      "learning_rate": 1.0700000000000001e-05,
      "loss": 3.0453,
      "step": 107
    },
    {
      "epoch": 0.03332304844183894,
      "grad_norm": 0.6298946142196655,
      "learning_rate": 1.08e-05,
      "loss": 2.9881,
      "step": 108
    },
    {
      "epoch": 0.03363159518667078,
      "grad_norm": 0.495768666267395,
      "learning_rate": 1.09e-05,
      "loss": 3.3919,
      "step": 109
    },
    {
      "epoch": 0.033940141931502625,
      "grad_norm": 0.7116994261741638,
      "learning_rate": 1.1e-05,
      "loss": 3.188,
      "step": 110
    },
    {
      "epoch": 0.03424868867633447,
      "grad_norm": 0.6338837146759033,
      "learning_rate": 1.11e-05,
      "loss": 3.1988,
      "step": 111
    },
    {
      "epoch": 0.03455723542116631,
      "grad_norm": 0.6400672793388367,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 3.3323,
      "step": 112
    },
    {
      "epoch": 0.03486578216599815,
      "grad_norm": 0.48958930373191833,
      "learning_rate": 1.13e-05,
      "loss": 2.7639,
      "step": 113
    },
    {
      "epoch": 0.03517432891082999,
      "grad_norm": 0.7853876948356628,
      "learning_rate": 1.1400000000000001e-05,
      "loss": 3.216,
      "step": 114
    },
    {
      "epoch": 0.035482875655661834,
      "grad_norm": 0.4167158603668213,
      "learning_rate": 1.1500000000000002e-05,
      "loss": 2.5009,
      "step": 115
    },
    {
      "epoch": 0.035791422400493676,
      "grad_norm": 0.43414178490638733,
      "learning_rate": 1.16e-05,
      "loss": 2.9166,
      "step": 116
    },
    {
      "epoch": 0.03609996914532552,
      "grad_norm": 0.5888383984565735,
      "learning_rate": 1.1700000000000001e-05,
      "loss": 3.4765,
      "step": 117
    },
    {
      "epoch": 0.03640851589015736,
      "grad_norm": 0.4863530099391937,
      "learning_rate": 1.1799999999999999e-05,
      "loss": 2.6866,
      "step": 118
    },
    {
      "epoch": 0.0367170626349892,
      "grad_norm": 0.5108136534690857,
      "learning_rate": 1.19e-05,
      "loss": 3.0799,
      "step": 119
    },
    {
      "epoch": 0.037025609379821044,
      "grad_norm": 0.72972172498703,
      "learning_rate": 1.2e-05,
      "loss": 3.3096,
      "step": 120
    },
    {
      "epoch": 0.037334156124652886,
      "grad_norm": 0.4917832016944885,
      "learning_rate": 1.21e-05,
      "loss": 3.3209,
      "step": 121
    },
    {
      "epoch": 0.03764270286948473,
      "grad_norm": 0.5651227831840515,
      "learning_rate": 1.22e-05,
      "loss": 3.2901,
      "step": 122
    },
    {
      "epoch": 0.03795124961431657,
      "grad_norm": 0.4422552287578583,
      "learning_rate": 1.2299999999999999e-05,
      "loss": 2.7225,
      "step": 123
    },
    {
      "epoch": 0.03825979635914841,
      "grad_norm": 0.4900197982788086,
      "learning_rate": 1.24e-05,
      "loss": 2.676,
      "step": 124
    },
    {
      "epoch": 0.03856834310398025,
      "grad_norm": 0.5758654475212097,
      "learning_rate": 1.25e-05,
      "loss": 2.7869,
      "step": 125
    },
    {
      "epoch": 0.038876889848812095,
      "grad_norm": 0.4216962456703186,
      "learning_rate": 1.26e-05,
      "loss": 2.5913,
      "step": 126
    },
    {
      "epoch": 0.03918543659364394,
      "grad_norm": 0.5779691934585571,
      "learning_rate": 1.27e-05,
      "loss": 2.93,
      "step": 127
    },
    {
      "epoch": 0.03949398333847578,
      "grad_norm": 0.5078710317611694,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 2.873,
      "step": 128
    },
    {
      "epoch": 0.03980253008330762,
      "grad_norm": 0.7252984046936035,
      "learning_rate": 1.29e-05,
      "loss": 3.2651,
      "step": 129
    },
    {
      "epoch": 0.04011107682813946,
      "grad_norm": 0.30763328075408936,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 2.5202,
      "step": 130
    },
    {
      "epoch": 0.040419623572971304,
      "grad_norm": 0.5326187610626221,
      "learning_rate": 1.31e-05,
      "loss": 2.7101,
      "step": 131
    },
    {
      "epoch": 0.040728170317803146,
      "grad_norm": 0.5274989604949951,
      "learning_rate": 1.32e-05,
      "loss": 2.8888,
      "step": 132
    },
    {
      "epoch": 0.04103671706263499,
      "grad_norm": 0.4721854031085968,
      "learning_rate": 1.3300000000000001e-05,
      "loss": 3.1288,
      "step": 133
    },
    {
      "epoch": 0.04134526380746683,
      "grad_norm": 0.4977727234363556,
      "learning_rate": 1.34e-05,
      "loss": 2.7822,
      "step": 134
    },
    {
      "epoch": 0.04165381055229867,
      "grad_norm": 0.5609750747680664,
      "learning_rate": 1.3500000000000001e-05,
      "loss": 2.9943,
      "step": 135
    },
    {
      "epoch": 0.04196235729713051,
      "grad_norm": 0.5416237711906433,
      "learning_rate": 1.36e-05,
      "loss": 2.9631,
      "step": 136
    },
    {
      "epoch": 0.042270904041962355,
      "grad_norm": 0.7931625247001648,
      "learning_rate": 1.3700000000000001e-05,
      "loss": 3.2442,
      "step": 137
    },
    {
      "epoch": 0.0425794507867942,
      "grad_norm": 0.41734540462493896,
      "learning_rate": 1.3800000000000002e-05,
      "loss": 2.7296,
      "step": 138
    },
    {
      "epoch": 0.04288799753162604,
      "grad_norm": 0.5195217132568359,
      "learning_rate": 1.39e-05,
      "loss": 3.0406,
      "step": 139
    },
    {
      "epoch": 0.04319654427645788,
      "grad_norm": 0.43490850925445557,
      "learning_rate": 1.4e-05,
      "loss": 2.7431,
      "step": 140
    },
    {
      "epoch": 0.04350509102128972,
      "grad_norm": 0.6502295732498169,
      "learning_rate": 1.4099999999999999e-05,
      "loss": 3.0238,
      "step": 141
    },
    {
      "epoch": 0.043813637766121565,
      "grad_norm": 0.44002243876457214,
      "learning_rate": 1.42e-05,
      "loss": 2.9071,
      "step": 142
    },
    {
      "epoch": 0.044122184510953406,
      "grad_norm": 0.578681468963623,
      "learning_rate": 1.43e-05,
      "loss": 3.1955,
      "step": 143
    },
    {
      "epoch": 0.04443073125578525,
      "grad_norm": 0.5037577152252197,
      "learning_rate": 1.44e-05,
      "loss": 2.6283,
      "step": 144
    },
    {
      "epoch": 0.04473927800061709,
      "grad_norm": 0.46894946694374084,
      "learning_rate": 1.45e-05,
      "loss": 2.5679,
      "step": 145
    },
    {
      "epoch": 0.04504782474544894,
      "grad_norm": 0.7387483716011047,
      "learning_rate": 1.46e-05,
      "loss": 3.3251,
      "step": 146
    },
    {
      "epoch": 0.04535637149028078,
      "grad_norm": 0.4696255624294281,
      "learning_rate": 1.47e-05,
      "loss": 2.7964,
      "step": 147
    },
    {
      "epoch": 0.04566491823511262,
      "grad_norm": 0.6008668541908264,
      "learning_rate": 1.48e-05,
      "loss": 2.7697,
      "step": 148
    },
    {
      "epoch": 0.045973464979944464,
      "grad_norm": 0.8156336545944214,
      "learning_rate": 1.49e-05,
      "loss": 3.2155,
      "step": 149
    },
    {
      "epoch": 0.046282011724776306,
      "grad_norm": 0.37692368030548096,
      "learning_rate": 1.5e-05,
      "loss": 2.7007,
      "step": 150
    },
    {
      "epoch": 0.04659055846960815,
      "grad_norm": 0.7642585635185242,
      "learning_rate": 1.51e-05,
      "loss": 3.6619,
      "step": 151
    },
    {
      "epoch": 0.04689910521443999,
      "grad_norm": 0.6754474639892578,
      "learning_rate": 1.5200000000000002e-05,
      "loss": 3.1602,
      "step": 152
    },
    {
      "epoch": 0.04720765195927183,
      "grad_norm": 0.6416701674461365,
      "learning_rate": 1.53e-05,
      "loss": 3.3756,
      "step": 153
    },
    {
      "epoch": 0.047516198704103674,
      "grad_norm": 0.654334306716919,
      "learning_rate": 1.5399999999999998e-05,
      "loss": 3.168,
      "step": 154
    },
    {
      "epoch": 0.047824745448935516,
      "grad_norm": 0.6520740389823914,
      "learning_rate": 1.55e-05,
      "loss": 3.2226,
      "step": 155
    },
    {
      "epoch": 0.04813329219376736,
      "grad_norm": 0.384902685880661,
      "learning_rate": 1.56e-05,
      "loss": 2.5777,
      "step": 156
    },
    {
      "epoch": 0.0484418389385992,
      "grad_norm": 0.5706138014793396,
      "learning_rate": 1.57e-05,
      "loss": 3.2847,
      "step": 157
    },
    {
      "epoch": 0.04875038568343104,
      "grad_norm": 0.5485690832138062,
      "learning_rate": 1.5799999999999998e-05,
      "loss": 3.1548,
      "step": 158
    },
    {
      "epoch": 0.04905893242826288,
      "grad_norm": 0.5101959705352783,
      "learning_rate": 1.59e-05,
      "loss": 2.8384,
      "step": 159
    },
    {
      "epoch": 0.049367479173094725,
      "grad_norm": 0.4596351683139801,
      "learning_rate": 1.6e-05,
      "loss": 2.9464,
      "step": 160
    },
    {
      "epoch": 0.04967602591792657,
      "grad_norm": 0.6397862434387207,
      "learning_rate": 1.61e-05,
      "loss": 3.2851,
      "step": 161
    },
    {
      "epoch": 0.04998457266275841,
      "grad_norm": 0.3608058989048004,
      "learning_rate": 1.62e-05,
      "loss": 2.43,
      "step": 162
    },
    {
      "epoch": 0.05029311940759025,
      "grad_norm": 0.7656980752944946,
      "learning_rate": 1.63e-05,
      "loss": 3.5049,
      "step": 163
    },
    {
      "epoch": 0.05060166615242209,
      "grad_norm": 0.4537852108478546,
      "learning_rate": 1.64e-05,
      "loss": 2.6858,
      "step": 164
    },
    {
      "epoch": 0.050910212897253934,
      "grad_norm": 0.46467283368110657,
      "learning_rate": 1.65e-05,
      "loss": 2.8103,
      "step": 165
    },
    {
      "epoch": 0.051218759642085776,
      "grad_norm": 0.6642769575119019,
      "learning_rate": 1.66e-05,
      "loss": 3.2552,
      "step": 166
    },
    {
      "epoch": 0.05152730638691762,
      "grad_norm": 0.707314133644104,
      "learning_rate": 1.67e-05,
      "loss": 3.5667,
      "step": 167
    },
    {
      "epoch": 0.05183585313174946,
      "grad_norm": 0.4103851020336151,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 2.8039,
      "step": 168
    },
    {
      "epoch": 0.0521443998765813,
      "grad_norm": 0.978696882724762,
      "learning_rate": 1.69e-05,
      "loss": 2.8174,
      "step": 169
    },
    {
      "epoch": 0.05245294662141314,
      "grad_norm": 0.4563823640346527,
      "learning_rate": 1.7e-05,
      "loss": 3.1365,
      "step": 170
    },
    {
      "epoch": 0.052761493366244985,
      "grad_norm": 0.4271509051322937,
      "learning_rate": 1.71e-05,
      "loss": 2.6638,
      "step": 171
    },
    {
      "epoch": 0.05307004011107683,
      "grad_norm": 0.570706844329834,
      "learning_rate": 1.72e-05,
      "loss": 2.9269,
      "step": 172
    },
    {
      "epoch": 0.05337858685590867,
      "grad_norm": 0.38479119539260864,
      "learning_rate": 1.73e-05,
      "loss": 2.1779,
      "step": 173
    },
    {
      "epoch": 0.05368713360074051,
      "grad_norm": 0.5205680727958679,
      "learning_rate": 1.74e-05,
      "loss": 2.8916,
      "step": 174
    },
    {
      "epoch": 0.05399568034557235,
      "grad_norm": 0.5419281721115112,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 2.8412,
      "step": 175
    },
    {
      "epoch": 0.054304227090404195,
      "grad_norm": 0.5400970578193665,
      "learning_rate": 1.76e-05,
      "loss": 3.1398,
      "step": 176
    },
    {
      "epoch": 0.054612773835236036,
      "grad_norm": 0.4808214604854584,
      "learning_rate": 1.77e-05,
      "loss": 2.9332,
      "step": 177
    },
    {
      "epoch": 0.05492132058006788,
      "grad_norm": 0.5370422005653381,
      "learning_rate": 1.7800000000000002e-05,
      "loss": 2.6539,
      "step": 178
    },
    {
      "epoch": 0.05522986732489972,
      "grad_norm": 0.5261011719703674,
      "learning_rate": 1.79e-05,
      "loss": 2.6934,
      "step": 179
    },
    {
      "epoch": 0.05553841406973156,
      "grad_norm": 0.4394947588443756,
      "learning_rate": 1.8e-05,
      "loss": 2.6768,
      "step": 180
    },
    {
      "epoch": 0.055846960814563404,
      "grad_norm": 0.3960530459880829,
      "learning_rate": 1.8100000000000003e-05,
      "loss": 2.7511,
      "step": 181
    },
    {
      "epoch": 0.056155507559395246,
      "grad_norm": 0.4033471941947937,
      "learning_rate": 1.8200000000000002e-05,
      "loss": 2.6748,
      "step": 182
    },
    {
      "epoch": 0.05646405430422709,
      "grad_norm": 0.46211346983909607,
      "learning_rate": 1.83e-05,
      "loss": 2.5696,
      "step": 183
    },
    {
      "epoch": 0.05677260104905893,
      "grad_norm": 0.48176342248916626,
      "learning_rate": 1.84e-05,
      "loss": 2.6858,
      "step": 184
    },
    {
      "epoch": 0.05708114779389077,
      "grad_norm": 0.5655994415283203,
      "learning_rate": 1.8500000000000002e-05,
      "loss": 3.2977,
      "step": 185
    },
    {
      "epoch": 0.05738969453872261,
      "grad_norm": 0.7159984111785889,
      "learning_rate": 1.86e-05,
      "loss": 3.7336,
      "step": 186
    },
    {
      "epoch": 0.05769824128355446,
      "grad_norm": 0.5511050224304199,
      "learning_rate": 1.87e-05,
      "loss": 2.7869,
      "step": 187
    },
    {
      "epoch": 0.058006788028386304,
      "grad_norm": 0.5503979921340942,
      "learning_rate": 1.8800000000000003e-05,
      "loss": 2.7981,
      "step": 188
    },
    {
      "epoch": 0.058315334773218146,
      "grad_norm": 0.4197019040584564,
      "learning_rate": 1.8900000000000002e-05,
      "loss": 2.5163,
      "step": 189
    },
    {
      "epoch": 0.05862388151804999,
      "grad_norm": 0.5114262104034424,
      "learning_rate": 1.9e-05,
      "loss": 2.9493,
      "step": 190
    },
    {
      "epoch": 0.05893242826288183,
      "grad_norm": 0.8843671679496765,
      "learning_rate": 1.9100000000000003e-05,
      "loss": 3.295,
      "step": 191
    },
    {
      "epoch": 0.05924097500771367,
      "grad_norm": 0.3419288396835327,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 2.3456,
      "step": 192
    },
    {
      "epoch": 0.05954952175254551,
      "grad_norm": 0.5502027869224548,
      "learning_rate": 1.9299999999999998e-05,
      "loss": 2.3674,
      "step": 193
    },
    {
      "epoch": 0.059858068497377355,
      "grad_norm": 0.7232564091682434,
      "learning_rate": 1.9399999999999997e-05,
      "loss": 3.4577,
      "step": 194
    },
    {
      "epoch": 0.0601666152422092,
      "grad_norm": 0.4990410804748535,
      "learning_rate": 1.95e-05,
      "loss": 2.5966,
      "step": 195
    },
    {
      "epoch": 0.06047516198704104,
      "grad_norm": 0.5583244562149048,
      "learning_rate": 1.96e-05,
      "loss": 2.58,
      "step": 196
    },
    {
      "epoch": 0.06078370873187288,
      "grad_norm": 0.8695673942565918,
      "learning_rate": 1.9699999999999998e-05,
      "loss": 3.292,
      "step": 197
    },
    {
      "epoch": 0.06109225547670472,
      "grad_norm": 0.47713199257850647,
      "learning_rate": 1.98e-05,
      "loss": 2.5254,
      "step": 198
    },
    {
      "epoch": 0.061400802221536564,
      "grad_norm": 0.7606462836265564,
      "learning_rate": 1.99e-05,
      "loss": 2.9146,
      "step": 199
    },
    {
      "epoch": 0.061709348966368406,
      "grad_norm": 0.6246155500411987,
      "learning_rate": 1.9999999999999998e-05,
      "loss": 3.1394,
      "step": 200
    },
    {
      "epoch": 0.06201789571120025,
      "grad_norm": 0.741635262966156,
      "learning_rate": 2.01e-05,
      "loss": 2.8496,
      "step": 201
    },
    {
      "epoch": 0.06232644245603209,
      "grad_norm": 0.5045515894889832,
      "learning_rate": 2.02e-05,
      "loss": 2.7577,
      "step": 202
    },
    {
      "epoch": 0.06263498920086392,
      "grad_norm": 0.6223952174186707,
      "learning_rate": 2.03e-05,
      "loss": 2.7214,
      "step": 203
    },
    {
      "epoch": 0.06294353594569577,
      "grad_norm": 0.5237772464752197,
      "learning_rate": 2.04e-05,
      "loss": 2.9416,
      "step": 204
    },
    {
      "epoch": 0.06325208269052761,
      "grad_norm": 1.014267086982727,
      "learning_rate": 2.05e-05,
      "loss": 3.0007,
      "step": 205
    },
    {
      "epoch": 0.06356062943535945,
      "grad_norm": 0.6386702060699463,
      "learning_rate": 2.06e-05,
      "loss": 2.8931,
      "step": 206
    },
    {
      "epoch": 0.06386917618019129,
      "grad_norm": 0.6047970056533813,
      "learning_rate": 2.07e-05,
      "loss": 2.4792,
      "step": 207
    },
    {
      "epoch": 0.06417772292502315,
      "grad_norm": 0.5924057364463806,
      "learning_rate": 2.08e-05,
      "loss": 2.9367,
      "step": 208
    },
    {
      "epoch": 0.06448626966985499,
      "grad_norm": 0.5173370242118835,
      "learning_rate": 2.09e-05,
      "loss": 2.499,
      "step": 209
    },
    {
      "epoch": 0.06479481641468683,
      "grad_norm": 0.9348816275596619,
      "learning_rate": 2.1e-05,
      "loss": 2.9291,
      "step": 210
    },
    {
      "epoch": 0.06510336315951867,
      "grad_norm": 0.500065267086029,
      "learning_rate": 2.11e-05,
      "loss": 3.0239,
      "step": 211
    },
    {
      "epoch": 0.06541190990435052,
      "grad_norm": 0.6069537401199341,
      "learning_rate": 2.12e-05,
      "loss": 2.9213,
      "step": 212
    },
    {
      "epoch": 0.06572045664918236,
      "grad_norm": 0.8060280084609985,
      "learning_rate": 2.13e-05,
      "loss": 2.9865,
      "step": 213
    },
    {
      "epoch": 0.0660290033940142,
      "grad_norm": 0.5638638138771057,
      "learning_rate": 2.1400000000000002e-05,
      "loss": 3.1433,
      "step": 214
    },
    {
      "epoch": 0.06633755013884604,
      "grad_norm": 0.9096368551254272,
      "learning_rate": 2.15e-05,
      "loss": 3.3942,
      "step": 215
    },
    {
      "epoch": 0.06664609688367788,
      "grad_norm": 0.4683953523635864,
      "learning_rate": 2.16e-05,
      "loss": 2.5365,
      "step": 216
    },
    {
      "epoch": 0.06695464362850972,
      "grad_norm": 0.6851359009742737,
      "learning_rate": 2.1700000000000002e-05,
      "loss": 2.6351,
      "step": 217
    },
    {
      "epoch": 0.06726319037334157,
      "grad_norm": 0.5846662521362305,
      "learning_rate": 2.18e-05,
      "loss": 2.8887,
      "step": 218
    },
    {
      "epoch": 0.06757173711817341,
      "grad_norm": 0.42902109026908875,
      "learning_rate": 2.19e-05,
      "loss": 2.5048,
      "step": 219
    },
    {
      "epoch": 0.06788028386300525,
      "grad_norm": 0.5018395185470581,
      "learning_rate": 2.2e-05,
      "loss": 2.854,
      "step": 220
    },
    {
      "epoch": 0.06818883060783709,
      "grad_norm": 0.5672473311424255,
      "learning_rate": 2.2100000000000002e-05,
      "loss": 2.6505,
      "step": 221
    },
    {
      "epoch": 0.06849737735266893,
      "grad_norm": 0.4538055956363678,
      "learning_rate": 2.22e-05,
      "loss": 2.6871,
      "step": 222
    },
    {
      "epoch": 0.06880592409750078,
      "grad_norm": 0.4177427589893341,
      "learning_rate": 2.23e-05,
      "loss": 2.5534,
      "step": 223
    },
    {
      "epoch": 0.06911447084233262,
      "grad_norm": 0.5405595898628235,
      "learning_rate": 2.2400000000000002e-05,
      "loss": 2.8893,
      "step": 224
    },
    {
      "epoch": 0.06942301758716446,
      "grad_norm": 0.5323094725608826,
      "learning_rate": 2.25e-05,
      "loss": 2.5933,
      "step": 225
    },
    {
      "epoch": 0.0697315643319963,
      "grad_norm": 0.6036468744277954,
      "learning_rate": 2.26e-05,
      "loss": 2.8331,
      "step": 226
    },
    {
      "epoch": 0.07004011107682814,
      "grad_norm": 0.545752763748169,
      "learning_rate": 2.2700000000000003e-05,
      "loss": 2.6558,
      "step": 227
    },
    {
      "epoch": 0.07034865782165999,
      "grad_norm": 0.3827911913394928,
      "learning_rate": 2.2800000000000002e-05,
      "loss": 2.3423,
      "step": 228
    },
    {
      "epoch": 0.07065720456649183,
      "grad_norm": 0.4605464041233063,
      "learning_rate": 2.29e-05,
      "loss": 2.7205,
      "step": 229
    },
    {
      "epoch": 0.07096575131132367,
      "grad_norm": 0.5113175511360168,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 2.5991,
      "step": 230
    },
    {
      "epoch": 0.07127429805615551,
      "grad_norm": 1.0119787454605103,
      "learning_rate": 2.3100000000000002e-05,
      "loss": 3.7089,
      "step": 231
    },
    {
      "epoch": 0.07158284480098735,
      "grad_norm": 0.9721714854240417,
      "learning_rate": 2.32e-05,
      "loss": 2.7576,
      "step": 232
    },
    {
      "epoch": 0.0718913915458192,
      "grad_norm": 0.7502716779708862,
      "learning_rate": 2.33e-05,
      "loss": 2.5823,
      "step": 233
    },
    {
      "epoch": 0.07219993829065104,
      "grad_norm": 0.46226513385772705,
      "learning_rate": 2.3400000000000003e-05,
      "loss": 2.4788,
      "step": 234
    },
    {
      "epoch": 0.07250848503548288,
      "grad_norm": 0.6776145696640015,
      "learning_rate": 2.3500000000000002e-05,
      "loss": 2.7757,
      "step": 235
    },
    {
      "epoch": 0.07281703178031472,
      "grad_norm": 1.051665186882019,
      "learning_rate": 2.3599999999999998e-05,
      "loss": 3.2042,
      "step": 236
    },
    {
      "epoch": 0.07312557852514656,
      "grad_norm": 0.4878885746002197,
      "learning_rate": 2.37e-05,
      "loss": 2.6037,
      "step": 237
    },
    {
      "epoch": 0.0734341252699784,
      "grad_norm": 1.0094560384750366,
      "learning_rate": 2.38e-05,
      "loss": 3.0565,
      "step": 238
    },
    {
      "epoch": 0.07374267201481025,
      "grad_norm": 0.6273016929626465,
      "learning_rate": 2.3899999999999998e-05,
      "loss": 2.6155,
      "step": 239
    },
    {
      "epoch": 0.07405121875964209,
      "grad_norm": 0.7550265192985535,
      "learning_rate": 2.4e-05,
      "loss": 2.7951,
      "step": 240
    },
    {
      "epoch": 0.07435976550447393,
      "grad_norm": 0.9445892572402954,
      "learning_rate": 2.41e-05,
      "loss": 2.8222,
      "step": 241
    },
    {
      "epoch": 0.07466831224930577,
      "grad_norm": 0.7777389883995056,
      "learning_rate": 2.42e-05,
      "loss": 3.1079,
      "step": 242
    },
    {
      "epoch": 0.07497685899413761,
      "grad_norm": 0.6355181932449341,
      "learning_rate": 2.43e-05,
      "loss": 2.6854,
      "step": 243
    },
    {
      "epoch": 0.07528540573896945,
      "grad_norm": 1.261012315750122,
      "learning_rate": 2.44e-05,
      "loss": 2.6395,
      "step": 244
    },
    {
      "epoch": 0.0755939524838013,
      "grad_norm": 0.659038245677948,
      "learning_rate": 2.45e-05,
      "loss": 2.5814,
      "step": 245
    },
    {
      "epoch": 0.07590249922863314,
      "grad_norm": 0.873461127281189,
      "learning_rate": 2.4599999999999998e-05,
      "loss": 3.2166,
      "step": 246
    },
    {
      "epoch": 0.07621104597346498,
      "grad_norm": 0.5640694499015808,
      "learning_rate": 2.47e-05,
      "loss": 2.6797,
      "step": 247
    },
    {
      "epoch": 0.07651959271829682,
      "grad_norm": 0.8357604742050171,
      "learning_rate": 2.48e-05,
      "loss": 3.1053,
      "step": 248
    },
    {
      "epoch": 0.07682813946312866,
      "grad_norm": 0.6787241697311401,
      "learning_rate": 2.49e-05,
      "loss": 3.0971,
      "step": 249
    },
    {
      "epoch": 0.0771366862079605,
      "grad_norm": 1.1458957195281982,
      "learning_rate": 2.5e-05,
      "loss": 3.6878,
      "step": 250
    },
    {
      "epoch": 0.07744523295279235,
      "grad_norm": 0.5414482355117798,
      "learning_rate": 2.51e-05,
      "loss": 2.7741,
      "step": 251
    },
    {
      "epoch": 0.07775377969762419,
      "grad_norm": 0.6754476428031921,
      "learning_rate": 2.52e-05,
      "loss": 2.9239,
      "step": 252
    },
    {
      "epoch": 0.07806232644245603,
      "grad_norm": 0.6492197513580322,
      "learning_rate": 2.5300000000000002e-05,
      "loss": 2.9217,
      "step": 253
    },
    {
      "epoch": 0.07837087318728787,
      "grad_norm": 0.4921364188194275,
      "learning_rate": 2.54e-05,
      "loss": 2.4709,
      "step": 254
    },
    {
      "epoch": 0.07867941993211972,
      "grad_norm": 0.535729169845581,
      "learning_rate": 2.55e-05,
      "loss": 2.4383,
      "step": 255
    },
    {
      "epoch": 0.07898796667695156,
      "grad_norm": 0.6355189085006714,
      "learning_rate": 2.5600000000000002e-05,
      "loss": 2.8076,
      "step": 256
    },
    {
      "epoch": 0.0792965134217834,
      "grad_norm": 0.5118688344955444,
      "learning_rate": 2.57e-05,
      "loss": 2.8182,
      "step": 257
    },
    {
      "epoch": 0.07960506016661524,
      "grad_norm": 0.49077171087265015,
      "learning_rate": 2.58e-05,
      "loss": 2.4566,
      "step": 258
    },
    {
      "epoch": 0.07991360691144708,
      "grad_norm": 0.698340654373169,
      "learning_rate": 2.59e-05,
      "loss": 2.8407,
      "step": 259
    },
    {
      "epoch": 0.08022215365627892,
      "grad_norm": 0.6589793562889099,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 2.7079,
      "step": 260
    },
    {
      "epoch": 0.08053070040111077,
      "grad_norm": 0.4942406117916107,
      "learning_rate": 2.61e-05,
      "loss": 2.7984,
      "step": 261
    },
    {
      "epoch": 0.08083924714594261,
      "grad_norm": 0.5735039710998535,
      "learning_rate": 2.62e-05,
      "loss": 2.7639,
      "step": 262
    },
    {
      "epoch": 0.08114779389077445,
      "grad_norm": 0.5074241757392883,
      "learning_rate": 2.6300000000000002e-05,
      "loss": 2.8036,
      "step": 263
    },
    {
      "epoch": 0.08145634063560629,
      "grad_norm": 0.4602759778499603,
      "learning_rate": 2.64e-05,
      "loss": 2.257,
      "step": 264
    },
    {
      "epoch": 0.08176488738043813,
      "grad_norm": 0.8142499327659607,
      "learning_rate": 2.65e-05,
      "loss": 3.1813,
      "step": 265
    },
    {
      "epoch": 0.08207343412526998,
      "grad_norm": 0.44730252027511597,
      "learning_rate": 2.6600000000000003e-05,
      "loss": 2.6948,
      "step": 266
    },
    {
      "epoch": 0.08238198087010182,
      "grad_norm": 0.3990883529186249,
      "learning_rate": 2.6700000000000002e-05,
      "loss": 2.5447,
      "step": 267
    },
    {
      "epoch": 0.08269052761493366,
      "grad_norm": 0.7322868704795837,
      "learning_rate": 2.68e-05,
      "loss": 2.9178,
      "step": 268
    },
    {
      "epoch": 0.0829990743597655,
      "grad_norm": 0.7331518530845642,
      "learning_rate": 2.69e-05,
      "loss": 2.953,
      "step": 269
    },
    {
      "epoch": 0.08330762110459734,
      "grad_norm": 0.44720855355262756,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 2.6966,
      "step": 270
    },
    {
      "epoch": 0.08361616784942918,
      "grad_norm": 0.4966452717781067,
      "learning_rate": 2.71e-05,
      "loss": 2.9062,
      "step": 271
    },
    {
      "epoch": 0.08392471459426103,
      "grad_norm": 0.6715177893638611,
      "learning_rate": 2.72e-05,
      "loss": 3.301,
      "step": 272
    },
    {
      "epoch": 0.08423326133909287,
      "grad_norm": 0.5892104506492615,
      "learning_rate": 2.7300000000000003e-05,
      "loss": 3.0694,
      "step": 273
    },
    {
      "epoch": 0.08454180808392471,
      "grad_norm": 0.7469586730003357,
      "learning_rate": 2.7400000000000002e-05,
      "loss": 3.3209,
      "step": 274
    },
    {
      "epoch": 0.08485035482875655,
      "grad_norm": 0.41894596815109253,
      "learning_rate": 2.75e-05,
      "loss": 2.5125,
      "step": 275
    },
    {
      "epoch": 0.0851589015735884,
      "grad_norm": 0.7170995473861694,
      "learning_rate": 2.7600000000000003e-05,
      "loss": 2.9459,
      "step": 276
    },
    {
      "epoch": 0.08546744831842024,
      "grad_norm": 0.4706147015094757,
      "learning_rate": 2.7700000000000002e-05,
      "loss": 2.3327,
      "step": 277
    },
    {
      "epoch": 0.08577599506325208,
      "grad_norm": 0.6144742965698242,
      "learning_rate": 2.78e-05,
      "loss": 2.6478,
      "step": 278
    },
    {
      "epoch": 0.08608454180808392,
      "grad_norm": 0.5859728455543518,
      "learning_rate": 2.79e-05,
      "loss": 2.6092,
      "step": 279
    },
    {
      "epoch": 0.08639308855291576,
      "grad_norm": 0.5228575468063354,
      "learning_rate": 2.8e-05,
      "loss": 2.8229,
      "step": 280
    },
    {
      "epoch": 0.0867016352977476,
      "grad_norm": 0.7771852612495422,
      "learning_rate": 2.81e-05,
      "loss": 3.0858,
      "step": 281
    },
    {
      "epoch": 0.08701018204257945,
      "grad_norm": 0.6817198395729065,
      "learning_rate": 2.8199999999999998e-05,
      "loss": 2.5982,
      "step": 282
    },
    {
      "epoch": 0.08731872878741129,
      "grad_norm": 0.8393194079399109,
      "learning_rate": 2.83e-05,
      "loss": 3.2083,
      "step": 283
    },
    {
      "epoch": 0.08762727553224313,
      "grad_norm": 0.5846584439277649,
      "learning_rate": 2.84e-05,
      "loss": 2.6578,
      "step": 284
    },
    {
      "epoch": 0.08793582227707497,
      "grad_norm": 0.7105833292007446,
      "learning_rate": 2.8499999999999998e-05,
      "loss": 2.3854,
      "step": 285
    },
    {
      "epoch": 0.08824436902190681,
      "grad_norm": 0.5590451955795288,
      "learning_rate": 2.86e-05,
      "loss": 2.7077,
      "step": 286
    },
    {
      "epoch": 0.08855291576673865,
      "grad_norm": 0.3551070988178253,
      "learning_rate": 2.87e-05,
      "loss": 2.3749,
      "step": 287
    },
    {
      "epoch": 0.0888614625115705,
      "grad_norm": 0.6778650283813477,
      "learning_rate": 2.88e-05,
      "loss": 2.5384,
      "step": 288
    },
    {
      "epoch": 0.08917000925640234,
      "grad_norm": 0.5005120635032654,
      "learning_rate": 2.89e-05,
      "loss": 2.4651,
      "step": 289
    },
    {
      "epoch": 0.08947855600123418,
      "grad_norm": 0.7873884439468384,
      "learning_rate": 2.9e-05,
      "loss": 3.0269,
      "step": 290
    },
    {
      "epoch": 0.08978710274606604,
      "grad_norm": 0.8063921928405762,
      "learning_rate": 2.91e-05,
      "loss": 2.8679,
      "step": 291
    },
    {
      "epoch": 0.09009564949089788,
      "grad_norm": 0.75018310546875,
      "learning_rate": 2.92e-05,
      "loss": 2.6179,
      "step": 292
    },
    {
      "epoch": 0.09040419623572972,
      "grad_norm": 0.3952403962612152,
      "learning_rate": 2.93e-05,
      "loss": 2.3271,
      "step": 293
    },
    {
      "epoch": 0.09071274298056156,
      "grad_norm": 0.49011290073394775,
      "learning_rate": 2.94e-05,
      "loss": 2.8457,
      "step": 294
    },
    {
      "epoch": 0.0910212897253934,
      "grad_norm": 0.720333456993103,
      "learning_rate": 2.95e-05,
      "loss": 2.5857,
      "step": 295
    },
    {
      "epoch": 0.09132983647022525,
      "grad_norm": 0.4137563705444336,
      "learning_rate": 2.96e-05,
      "loss": 2.4882,
      "step": 296
    },
    {
      "epoch": 0.09163838321505709,
      "grad_norm": 0.5216336846351624,
      "learning_rate": 2.97e-05,
      "loss": 2.7284,
      "step": 297
    },
    {
      "epoch": 0.09194692995988893,
      "grad_norm": 0.7018035054206848,
      "learning_rate": 2.98e-05,
      "loss": 3.4228,
      "step": 298
    },
    {
      "epoch": 0.09225547670472077,
      "grad_norm": 0.5406361818313599,
      "learning_rate": 2.9900000000000002e-05,
      "loss": 3.1286,
      "step": 299
    },
    {
      "epoch": 0.09256402344955261,
      "grad_norm": 0.48260602355003357,
      "learning_rate": 3e-05,
      "loss": 2.9251,
      "step": 300
    },
    {
      "epoch": 0.09287257019438445,
      "grad_norm": 0.599492073059082,
      "learning_rate": 3e-05,
      "loss": 2.9756,
      "step": 301
    },
    {
      "epoch": 0.0931811169392163,
      "grad_norm": 0.4386523962020874,
      "learning_rate": 3e-05,
      "loss": 2.8846,
      "step": 302
    },
    {
      "epoch": 0.09348966368404814,
      "grad_norm": 0.45703089237213135,
      "learning_rate": 3e-05,
      "loss": 2.8327,
      "step": 303
    },
    {
      "epoch": 0.09379821042887998,
      "grad_norm": 0.5618659257888794,
      "learning_rate": 3e-05,
      "loss": 2.442,
      "step": 304
    },
    {
      "epoch": 0.09410675717371182,
      "grad_norm": 0.5637118220329285,
      "learning_rate": 3e-05,
      "loss": 3.2367,
      "step": 305
    },
    {
      "epoch": 0.09441530391854366,
      "grad_norm": 0.7520857453346252,
      "learning_rate": 3e-05,
      "loss": 2.925,
      "step": 306
    },
    {
      "epoch": 0.0947238506633755,
      "grad_norm": 0.5939932465553284,
      "learning_rate": 3e-05,
      "loss": 3.1811,
      "step": 307
    },
    {
      "epoch": 0.09503239740820735,
      "grad_norm": 0.3925078809261322,
      "learning_rate": 3e-05,
      "loss": 2.8408,
      "step": 308
    },
    {
      "epoch": 0.09534094415303919,
      "grad_norm": 0.415669322013855,
      "learning_rate": 3e-05,
      "loss": 2.5195,
      "step": 309
    },
    {
      "epoch": 0.09564949089787103,
      "grad_norm": 0.6630593538284302,
      "learning_rate": 3e-05,
      "loss": 3.2672,
      "step": 310
    },
    {
      "epoch": 0.09595803764270287,
      "grad_norm": 0.4107421934604645,
      "learning_rate": 3e-05,
      "loss": 2.4262,
      "step": 311
    },
    {
      "epoch": 0.09626658438753471,
      "grad_norm": 0.5593919157981873,
      "learning_rate": 3e-05,
      "loss": 2.5541,
      "step": 312
    },
    {
      "epoch": 0.09657513113236656,
      "grad_norm": 0.42898955941200256,
      "learning_rate": 3e-05,
      "loss": 2.5179,
      "step": 313
    },
    {
      "epoch": 0.0968836778771984,
      "grad_norm": 0.31650015711784363,
      "learning_rate": 3e-05,
      "loss": 2.3117,
      "step": 314
    },
    {
      "epoch": 0.09719222462203024,
      "grad_norm": 0.6258980631828308,
      "learning_rate": 3e-05,
      "loss": 2.6524,
      "step": 315
    },
    {
      "epoch": 0.09750077136686208,
      "grad_norm": 0.45299839973449707,
      "learning_rate": 3e-05,
      "loss": 2.689,
      "step": 316
    },
    {
      "epoch": 0.09780931811169392,
      "grad_norm": 0.4562253952026367,
      "learning_rate": 3e-05,
      "loss": 2.5171,
      "step": 317
    },
    {
      "epoch": 0.09811786485652577,
      "grad_norm": 0.5021069049835205,
      "learning_rate": 3e-05,
      "loss": 2.4173,
      "step": 318
    },
    {
      "epoch": 0.09842641160135761,
      "grad_norm": 0.664849042892456,
      "learning_rate": 3e-05,
      "loss": 3.1166,
      "step": 319
    },
    {
      "epoch": 0.09873495834618945,
      "grad_norm": 0.45309123396873474,
      "learning_rate": 3e-05,
      "loss": 2.7596,
      "step": 320
    },
    {
      "epoch": 0.09904350509102129,
      "grad_norm": 0.7038612365722656,
      "learning_rate": 3e-05,
      "loss": 2.7522,
      "step": 321
    },
    {
      "epoch": 0.09935205183585313,
      "grad_norm": 1.059171199798584,
      "learning_rate": 3e-05,
      "loss": 3.175,
      "step": 322
    },
    {
      "epoch": 0.09966059858068498,
      "grad_norm": 0.5896137356758118,
      "learning_rate": 3e-05,
      "loss": 2.9216,
      "step": 323
    },
    {
      "epoch": 0.09996914532551682,
      "grad_norm": 0.637228786945343,
      "learning_rate": 3e-05,
      "loss": 2.8087,
      "step": 324
    },
    {
      "epoch": 0.10027769207034866,
      "grad_norm": 0.8929901123046875,
      "learning_rate": 3e-05,
      "loss": 2.948,
      "step": 325
    },
    {
      "epoch": 0.1005862388151805,
      "grad_norm": 0.48769569396972656,
      "learning_rate": 3e-05,
      "loss": 2.7344,
      "step": 326
    },
    {
      "epoch": 0.10089478556001234,
      "grad_norm": 0.507175862789154,
      "learning_rate": 3e-05,
      "loss": 2.5428,
      "step": 327
    },
    {
      "epoch": 0.10120333230484418,
      "grad_norm": 1.1506534814834595,
      "learning_rate": 3e-05,
      "loss": 2.9784,
      "step": 328
    },
    {
      "epoch": 0.10151187904967603,
      "grad_norm": 0.8921215534210205,
      "learning_rate": 3e-05,
      "loss": 3.1315,
      "step": 329
    },
    {
      "epoch": 0.10182042579450787,
      "grad_norm": 0.9674636125564575,
      "learning_rate": 3e-05,
      "loss": 2.9945,
      "step": 330
    },
    {
      "epoch": 0.10212897253933971,
      "grad_norm": 0.7912929058074951,
      "learning_rate": 3e-05,
      "loss": 2.7517,
      "step": 331
    },
    {
      "epoch": 0.10243751928417155,
      "grad_norm": 0.6095521450042725,
      "learning_rate": 3e-05,
      "loss": 3.4603,
      "step": 332
    },
    {
      "epoch": 0.1027460660290034,
      "grad_norm": 0.3913911283016205,
      "learning_rate": 3e-05,
      "loss": 2.3178,
      "step": 333
    },
    {
      "epoch": 0.10305461277383524,
      "grad_norm": 0.5703157782554626,
      "learning_rate": 3e-05,
      "loss": 2.7245,
      "step": 334
    },
    {
      "epoch": 0.10336315951866708,
      "grad_norm": 0.4117796719074249,
      "learning_rate": 3e-05,
      "loss": 2.2675,
      "step": 335
    },
    {
      "epoch": 0.10367170626349892,
      "grad_norm": 0.4819619357585907,
      "learning_rate": 3e-05,
      "loss": 2.5655,
      "step": 336
    },
    {
      "epoch": 0.10398025300833076,
      "grad_norm": 0.579757809638977,
      "learning_rate": 3e-05,
      "loss": 2.8346,
      "step": 337
    },
    {
      "epoch": 0.1042887997531626,
      "grad_norm": 0.5951722860336304,
      "learning_rate": 3e-05,
      "loss": 3.3011,
      "step": 338
    },
    {
      "epoch": 0.10459734649799445,
      "grad_norm": 0.37846288084983826,
      "learning_rate": 3e-05,
      "loss": 2.4172,
      "step": 339
    },
    {
      "epoch": 0.10490589324282629,
      "grad_norm": 0.5596647262573242,
      "learning_rate": 3e-05,
      "loss": 2.8481,
      "step": 340
    },
    {
      "epoch": 0.10521443998765813,
      "grad_norm": 0.4432225227355957,
      "learning_rate": 3e-05,
      "loss": 2.9247,
      "step": 341
    },
    {
      "epoch": 0.10552298673248997,
      "grad_norm": 0.5610328316688538,
      "learning_rate": 3e-05,
      "loss": 3.0414,
      "step": 342
    },
    {
      "epoch": 0.10583153347732181,
      "grad_norm": 0.6313160061836243,
      "learning_rate": 3e-05,
      "loss": 2.7949,
      "step": 343
    },
    {
      "epoch": 0.10614008022215365,
      "grad_norm": 0.5100327730178833,
      "learning_rate": 3e-05,
      "loss": 2.9491,
      "step": 344
    },
    {
      "epoch": 0.1064486269669855,
      "grad_norm": 0.4908035099506378,
      "learning_rate": 3e-05,
      "loss": 2.727,
      "step": 345
    },
    {
      "epoch": 0.10675717371181734,
      "grad_norm": 0.5060238242149353,
      "learning_rate": 3e-05,
      "loss": 2.4032,
      "step": 346
    },
    {
      "epoch": 0.10706572045664918,
      "grad_norm": 0.5218526124954224,
      "learning_rate": 3e-05,
      "loss": 2.8393,
      "step": 347
    },
    {
      "epoch": 0.10737426720148102,
      "grad_norm": 0.45432546734809875,
      "learning_rate": 3e-05,
      "loss": 2.3492,
      "step": 348
    },
    {
      "epoch": 0.10768281394631286,
      "grad_norm": 0.39769670367240906,
      "learning_rate": 3e-05,
      "loss": 2.4955,
      "step": 349
    },
    {
      "epoch": 0.1079913606911447,
      "grad_norm": 0.4049716591835022,
      "learning_rate": 3e-05,
      "loss": 2.5978,
      "step": 350
    },
    {
      "epoch": 0.10829990743597655,
      "grad_norm": 0.43709230422973633,
      "learning_rate": 3e-05,
      "loss": 2.4538,
      "step": 351
    },
    {
      "epoch": 0.10860845418080839,
      "grad_norm": 0.6582475900650024,
      "learning_rate": 3e-05,
      "loss": 2.7746,
      "step": 352
    },
    {
      "epoch": 0.10891700092564023,
      "grad_norm": 0.5041898488998413,
      "learning_rate": 3e-05,
      "loss": 2.8402,
      "step": 353
    },
    {
      "epoch": 0.10922554767047207,
      "grad_norm": 0.4751797318458557,
      "learning_rate": 3e-05,
      "loss": 2.5371,
      "step": 354
    },
    {
      "epoch": 0.10953409441530391,
      "grad_norm": 0.8427959084510803,
      "learning_rate": 3e-05,
      "loss": 2.6777,
      "step": 355
    },
    {
      "epoch": 0.10984264116013576,
      "grad_norm": 0.41446399688720703,
      "learning_rate": 3e-05,
      "loss": 2.5396,
      "step": 356
    },
    {
      "epoch": 0.1101511879049676,
      "grad_norm": 0.6712360978126526,
      "learning_rate": 3e-05,
      "loss": 3.3444,
      "step": 357
    },
    {
      "epoch": 0.11045973464979944,
      "grad_norm": 0.6713085174560547,
      "learning_rate": 3e-05,
      "loss": 3.2281,
      "step": 358
    },
    {
      "epoch": 0.11076828139463128,
      "grad_norm": 0.9303755164146423,
      "learning_rate": 3e-05,
      "loss": 3.2125,
      "step": 359
    },
    {
      "epoch": 0.11107682813946312,
      "grad_norm": 0.4587060511112213,
      "learning_rate": 3e-05,
      "loss": 2.786,
      "step": 360
    },
    {
      "epoch": 0.11138537488429497,
      "grad_norm": 0.7472328543663025,
      "learning_rate": 3e-05,
      "loss": 2.8521,
      "step": 361
    },
    {
      "epoch": 0.11169392162912681,
      "grad_norm": 0.502673864364624,
      "learning_rate": 3e-05,
      "loss": 2.7606,
      "step": 362
    },
    {
      "epoch": 0.11200246837395865,
      "grad_norm": 0.631190836429596,
      "learning_rate": 3e-05,
      "loss": 2.9696,
      "step": 363
    },
    {
      "epoch": 0.11231101511879049,
      "grad_norm": 0.6717808842658997,
      "learning_rate": 3e-05,
      "loss": 3.1333,
      "step": 364
    },
    {
      "epoch": 0.11261956186362233,
      "grad_norm": 0.8173141479492188,
      "learning_rate": 3e-05,
      "loss": 3.5001,
      "step": 365
    },
    {
      "epoch": 0.11292810860845418,
      "grad_norm": 0.5130720138549805,
      "learning_rate": 3e-05,
      "loss": 3.0127,
      "step": 366
    },
    {
      "epoch": 0.11323665535328602,
      "grad_norm": 0.5507920980453491,
      "learning_rate": 3e-05,
      "loss": 2.6635,
      "step": 367
    },
    {
      "epoch": 0.11354520209811786,
      "grad_norm": 0.5814347267150879,
      "learning_rate": 3e-05,
      "loss": 2.5873,
      "step": 368
    },
    {
      "epoch": 0.1138537488429497,
      "grad_norm": 0.4552783668041229,
      "learning_rate": 3e-05,
      "loss": 2.2747,
      "step": 369
    },
    {
      "epoch": 0.11416229558778154,
      "grad_norm": 0.48348626494407654,
      "learning_rate": 3e-05,
      "loss": 2.5721,
      "step": 370
    },
    {
      "epoch": 0.11447084233261338,
      "grad_norm": 0.7298465967178345,
      "learning_rate": 3e-05,
      "loss": 3.0346,
      "step": 371
    },
    {
      "epoch": 0.11477938907744523,
      "grad_norm": 0.4409029483795166,
      "learning_rate": 3e-05,
      "loss": 2.7849,
      "step": 372
    },
    {
      "epoch": 0.11508793582227707,
      "grad_norm": 0.47219765186309814,
      "learning_rate": 3e-05,
      "loss": 2.7278,
      "step": 373
    },
    {
      "epoch": 0.11539648256710892,
      "grad_norm": 0.6541340947151184,
      "learning_rate": 3e-05,
      "loss": 2.9001,
      "step": 374
    },
    {
      "epoch": 0.11570502931194077,
      "grad_norm": 0.696343183517456,
      "learning_rate": 3e-05,
      "loss": 2.6636,
      "step": 375
    },
    {
      "epoch": 0.11601357605677261,
      "grad_norm": 0.8152115941047668,
      "learning_rate": 3e-05,
      "loss": 2.7692,
      "step": 376
    },
    {
      "epoch": 0.11632212280160445,
      "grad_norm": 0.5533241033554077,
      "learning_rate": 3e-05,
      "loss": 2.6548,
      "step": 377
    },
    {
      "epoch": 0.11663066954643629,
      "grad_norm": 0.7412006855010986,
      "learning_rate": 3e-05,
      "loss": 2.847,
      "step": 378
    },
    {
      "epoch": 0.11693921629126813,
      "grad_norm": 0.3919863700866699,
      "learning_rate": 3e-05,
      "loss": 2.4736,
      "step": 379
    },
    {
      "epoch": 0.11724776303609998,
      "grad_norm": 1.019760251045227,
      "learning_rate": 3e-05,
      "loss": 3.2336,
      "step": 380
    },
    {
      "epoch": 0.11755630978093182,
      "grad_norm": 0.6420919895172119,
      "learning_rate": 3e-05,
      "loss": 2.887,
      "step": 381
    },
    {
      "epoch": 0.11786485652576366,
      "grad_norm": 0.3596365749835968,
      "learning_rate": 3e-05,
      "loss": 2.2894,
      "step": 382
    },
    {
      "epoch": 0.1181734032705955,
      "grad_norm": 0.5054413676261902,
      "learning_rate": 3e-05,
      "loss": 2.1737,
      "step": 383
    },
    {
      "epoch": 0.11848195001542734,
      "grad_norm": 0.6912227869033813,
      "learning_rate": 3e-05,
      "loss": 2.8684,
      "step": 384
    },
    {
      "epoch": 0.11879049676025918,
      "grad_norm": 0.5603222846984863,
      "learning_rate": 3e-05,
      "loss": 2.9749,
      "step": 385
    },
    {
      "epoch": 0.11909904350509103,
      "grad_norm": 0.4751299321651459,
      "learning_rate": 3e-05,
      "loss": 2.743,
      "step": 386
    },
    {
      "epoch": 0.11940759024992287,
      "grad_norm": 0.9485656023025513,
      "learning_rate": 3e-05,
      "loss": 3.1394,
      "step": 387
    },
    {
      "epoch": 0.11971613699475471,
      "grad_norm": 0.528243362903595,
      "learning_rate": 3e-05,
      "loss": 3.0556,
      "step": 388
    },
    {
      "epoch": 0.12002468373958655,
      "grad_norm": 0.48346731066703796,
      "learning_rate": 3e-05,
      "loss": 2.92,
      "step": 389
    },
    {
      "epoch": 0.1203332304844184,
      "grad_norm": 0.8272796869277954,
      "learning_rate": 3e-05,
      "loss": 2.78,
      "step": 390
    },
    {
      "epoch": 0.12064177722925024,
      "grad_norm": 0.5784688591957092,
      "learning_rate": 3e-05,
      "loss": 2.7566,
      "step": 391
    },
    {
      "epoch": 0.12095032397408208,
      "grad_norm": 0.5071232914924622,
      "learning_rate": 3e-05,
      "loss": 2.7876,
      "step": 392
    },
    {
      "epoch": 0.12125887071891392,
      "grad_norm": 0.4624871015548706,
      "learning_rate": 3e-05,
      "loss": 2.6045,
      "step": 393
    },
    {
      "epoch": 0.12156741746374576,
      "grad_norm": 0.7949879169464111,
      "learning_rate": 3e-05,
      "loss": 3.0567,
      "step": 394
    },
    {
      "epoch": 0.1218759642085776,
      "grad_norm": 0.4112931191921234,
      "learning_rate": 3e-05,
      "loss": 2.5882,
      "step": 395
    },
    {
      "epoch": 0.12218451095340944,
      "grad_norm": 0.508385956287384,
      "learning_rate": 3e-05,
      "loss": 2.8354,
      "step": 396
    },
    {
      "epoch": 0.12249305769824129,
      "grad_norm": 0.571725606918335,
      "learning_rate": 3e-05,
      "loss": 3.1637,
      "step": 397
    },
    {
      "epoch": 0.12280160444307313,
      "grad_norm": 0.4010452628135681,
      "learning_rate": 3e-05,
      "loss": 2.2202,
      "step": 398
    },
    {
      "epoch": 0.12311015118790497,
      "grad_norm": 0.6162316203117371,
      "learning_rate": 3e-05,
      "loss": 3.0857,
      "step": 399
    },
    {
      "epoch": 0.12341869793273681,
      "grad_norm": 0.5714832544326782,
      "learning_rate": 3e-05,
      "loss": 2.6672,
      "step": 400
    },
    {
      "epoch": 0.12372724467756865,
      "grad_norm": 0.7332398295402527,
      "learning_rate": 3e-05,
      "loss": 2.6827,
      "step": 401
    },
    {
      "epoch": 0.1240357914224005,
      "grad_norm": 0.8438281416893005,
      "learning_rate": 3e-05,
      "loss": 3.2429,
      "step": 402
    },
    {
      "epoch": 0.12434433816723234,
      "grad_norm": 0.5126581192016602,
      "learning_rate": 3e-05,
      "loss": 2.7263,
      "step": 403
    },
    {
      "epoch": 0.12465288491206418,
      "grad_norm": 1.201810598373413,
      "learning_rate": 3e-05,
      "loss": 2.9671,
      "step": 404
    },
    {
      "epoch": 0.12496143165689602,
      "grad_norm": 0.5391373634338379,
      "learning_rate": 3e-05,
      "loss": 2.596,
      "step": 405
    },
    {
      "epoch": 0.12526997840172785,
      "grad_norm": 0.49124547839164734,
      "learning_rate": 3e-05,
      "loss": 2.772,
      "step": 406
    },
    {
      "epoch": 0.1255785251465597,
      "grad_norm": 0.6671253442764282,
      "learning_rate": 3e-05,
      "loss": 2.4786,
      "step": 407
    },
    {
      "epoch": 0.12588707189139153,
      "grad_norm": 0.671453058719635,
      "learning_rate": 3e-05,
      "loss": 2.7801,
      "step": 408
    },
    {
      "epoch": 0.1261956186362234,
      "grad_norm": 0.5589339137077332,
      "learning_rate": 3e-05,
      "loss": 2.4786,
      "step": 409
    },
    {
      "epoch": 0.12650416538105522,
      "grad_norm": 0.48147571086883545,
      "learning_rate": 3e-05,
      "loss": 2.5111,
      "step": 410
    },
    {
      "epoch": 0.12681271212588707,
      "grad_norm": 0.7984548807144165,
      "learning_rate": 3e-05,
      "loss": 3.3761,
      "step": 411
    },
    {
      "epoch": 0.1271212588707189,
      "grad_norm": 0.4393971264362335,
      "learning_rate": 3e-05,
      "loss": 2.6791,
      "step": 412
    },
    {
      "epoch": 0.12742980561555076,
      "grad_norm": 0.3642028868198395,
      "learning_rate": 3e-05,
      "loss": 2.3538,
      "step": 413
    },
    {
      "epoch": 0.12773835236038258,
      "grad_norm": 0.4370051324367523,
      "learning_rate": 3e-05,
      "loss": 2.6664,
      "step": 414
    },
    {
      "epoch": 0.12804689910521444,
      "grad_norm": 0.4502509534358978,
      "learning_rate": 3e-05,
      "loss": 2.4552,
      "step": 415
    },
    {
      "epoch": 0.1283554458500463,
      "grad_norm": 0.8213755488395691,
      "learning_rate": 3e-05,
      "loss": 3.2743,
      "step": 416
    },
    {
      "epoch": 0.12866399259487812,
      "grad_norm": 0.3885883390903473,
      "learning_rate": 3e-05,
      "loss": 2.625,
      "step": 417
    },
    {
      "epoch": 0.12897253933970998,
      "grad_norm": 0.672049343585968,
      "learning_rate": 3e-05,
      "loss": 2.5793,
      "step": 418
    },
    {
      "epoch": 0.1292810860845418,
      "grad_norm": 0.8375455141067505,
      "learning_rate": 3e-05,
      "loss": 3.2976,
      "step": 419
    },
    {
      "epoch": 0.12958963282937366,
      "grad_norm": 0.29910895228385925,
      "learning_rate": 3e-05,
      "loss": 2.1132,
      "step": 420
    },
    {
      "epoch": 0.1298981795742055,
      "grad_norm": 0.5657820701599121,
      "learning_rate": 3e-05,
      "loss": 3.0597,
      "step": 421
    },
    {
      "epoch": 0.13020672631903735,
      "grad_norm": 0.5639760494232178,
      "learning_rate": 3e-05,
      "loss": 2.8884,
      "step": 422
    },
    {
      "epoch": 0.13051527306386917,
      "grad_norm": 0.567143976688385,
      "learning_rate": 3e-05,
      "loss": 3.3492,
      "step": 423
    },
    {
      "epoch": 0.13082381980870103,
      "grad_norm": 0.42509710788726807,
      "learning_rate": 3e-05,
      "loss": 2.754,
      "step": 424
    },
    {
      "epoch": 0.13113236655353286,
      "grad_norm": 0.5575029850006104,
      "learning_rate": 3e-05,
      "loss": 2.7359,
      "step": 425
    },
    {
      "epoch": 0.13144091329836471,
      "grad_norm": 0.577087938785553,
      "learning_rate": 3e-05,
      "loss": 2.4267,
      "step": 426
    },
    {
      "epoch": 0.13174946004319654,
      "grad_norm": 0.6447242498397827,
      "learning_rate": 3e-05,
      "loss": 2.7924,
      "step": 427
    },
    {
      "epoch": 0.1320580067880284,
      "grad_norm": 0.5749005675315857,
      "learning_rate": 3e-05,
      "loss": 2.9875,
      "step": 428
    },
    {
      "epoch": 0.13236655353286023,
      "grad_norm": 0.5711660385131836,
      "learning_rate": 3e-05,
      "loss": 2.7619,
      "step": 429
    },
    {
      "epoch": 0.13267510027769208,
      "grad_norm": 0.9919552803039551,
      "learning_rate": 3e-05,
      "loss": 3.5351,
      "step": 430
    },
    {
      "epoch": 0.1329836470225239,
      "grad_norm": 0.5573298335075378,
      "learning_rate": 3e-05,
      "loss": 2.6573,
      "step": 431
    },
    {
      "epoch": 0.13329219376735577,
      "grad_norm": 0.6087166666984558,
      "learning_rate": 3e-05,
      "loss": 2.871,
      "step": 432
    },
    {
      "epoch": 0.1336007405121876,
      "grad_norm": 0.9867172241210938,
      "learning_rate": 3e-05,
      "loss": 2.6612,
      "step": 433
    },
    {
      "epoch": 0.13390928725701945,
      "grad_norm": 0.4709426760673523,
      "learning_rate": 3e-05,
      "loss": 2.494,
      "step": 434
    },
    {
      "epoch": 0.13421783400185128,
      "grad_norm": 0.8016806244850159,
      "learning_rate": 3e-05,
      "loss": 3.5046,
      "step": 435
    },
    {
      "epoch": 0.13452638074668313,
      "grad_norm": 0.8725690245628357,
      "learning_rate": 3e-05,
      "loss": 3.0465,
      "step": 436
    },
    {
      "epoch": 0.13483492749151496,
      "grad_norm": 0.4316865801811218,
      "learning_rate": 3e-05,
      "loss": 2.5318,
      "step": 437
    },
    {
      "epoch": 0.13514347423634682,
      "grad_norm": 0.5138392448425293,
      "learning_rate": 3e-05,
      "loss": 2.2298,
      "step": 438
    },
    {
      "epoch": 0.13545202098117864,
      "grad_norm": 0.5295316576957703,
      "learning_rate": 3e-05,
      "loss": 2.3288,
      "step": 439
    },
    {
      "epoch": 0.1357605677260105,
      "grad_norm": 0.3984488248825073,
      "learning_rate": 3e-05,
      "loss": 2.3586,
      "step": 440
    },
    {
      "epoch": 0.13606911447084233,
      "grad_norm": 0.8097387552261353,
      "learning_rate": 3e-05,
      "loss": 3.2978,
      "step": 441
    },
    {
      "epoch": 0.13637766121567418,
      "grad_norm": 0.6348075270652771,
      "learning_rate": 3e-05,
      "loss": 2.7686,
      "step": 442
    },
    {
      "epoch": 0.136686207960506,
      "grad_norm": 0.6103842854499817,
      "learning_rate": 3e-05,
      "loss": 2.8613,
      "step": 443
    },
    {
      "epoch": 0.13699475470533787,
      "grad_norm": 0.5461227297782898,
      "learning_rate": 3e-05,
      "loss": 2.8534,
      "step": 444
    },
    {
      "epoch": 0.1373033014501697,
      "grad_norm": 0.6284303069114685,
      "learning_rate": 3e-05,
      "loss": 2.6065,
      "step": 445
    },
    {
      "epoch": 0.13761184819500155,
      "grad_norm": 0.3893953561782837,
      "learning_rate": 3e-05,
      "loss": 2.599,
      "step": 446
    },
    {
      "epoch": 0.13792039493983338,
      "grad_norm": 0.34796831011772156,
      "learning_rate": 3e-05,
      "loss": 2.2519,
      "step": 447
    },
    {
      "epoch": 0.13822894168466524,
      "grad_norm": 0.4476280212402344,
      "learning_rate": 3e-05,
      "loss": 2.4558,
      "step": 448
    },
    {
      "epoch": 0.13853748842949706,
      "grad_norm": 0.5699636936187744,
      "learning_rate": 3e-05,
      "loss": 2.6878,
      "step": 449
    },
    {
      "epoch": 0.13884603517432892,
      "grad_norm": 0.2940067648887634,
      "learning_rate": 3e-05,
      "loss": 2.281,
      "step": 450
    },
    {
      "epoch": 0.13915458191916075,
      "grad_norm": 0.9075848460197449,
      "learning_rate": 3e-05,
      "loss": 3.2938,
      "step": 451
    },
    {
      "epoch": 0.1394631286639926,
      "grad_norm": 0.4113154113292694,
      "learning_rate": 3e-05,
      "loss": 2.5557,
      "step": 452
    },
    {
      "epoch": 0.13977167540882443,
      "grad_norm": 0.5481420755386353,
      "learning_rate": 3e-05,
      "loss": 2.5868,
      "step": 453
    },
    {
      "epoch": 0.1400802221536563,
      "grad_norm": 0.4203355312347412,
      "learning_rate": 3e-05,
      "loss": 2.5495,
      "step": 454
    },
    {
      "epoch": 0.14038876889848811,
      "grad_norm": 0.40559303760528564,
      "learning_rate": 3e-05,
      "loss": 2.5116,
      "step": 455
    },
    {
      "epoch": 0.14069731564331997,
      "grad_norm": 0.5221837162971497,
      "learning_rate": 3e-05,
      "loss": 2.8499,
      "step": 456
    },
    {
      "epoch": 0.1410058623881518,
      "grad_norm": 0.48368218541145325,
      "learning_rate": 3e-05,
      "loss": 2.6947,
      "step": 457
    },
    {
      "epoch": 0.14131440913298365,
      "grad_norm": 0.4436626732349396,
      "learning_rate": 3e-05,
      "loss": 2.5644,
      "step": 458
    },
    {
      "epoch": 0.14162295587781548,
      "grad_norm": 0.5758005380630493,
      "learning_rate": 3e-05,
      "loss": 2.5727,
      "step": 459
    },
    {
      "epoch": 0.14193150262264734,
      "grad_norm": 0.2992756962776184,
      "learning_rate": 3e-05,
      "loss": 2.1367,
      "step": 460
    },
    {
      "epoch": 0.14224004936747917,
      "grad_norm": 0.4820541739463806,
      "learning_rate": 3e-05,
      "loss": 2.8957,
      "step": 461
    },
    {
      "epoch": 0.14254859611231102,
      "grad_norm": 0.6373788118362427,
      "learning_rate": 3e-05,
      "loss": 2.9102,
      "step": 462
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 0.49387943744659424,
      "learning_rate": 3e-05,
      "loss": 2.9304,
      "step": 463
    },
    {
      "epoch": 0.1431656896019747,
      "grad_norm": 0.3972255289554596,
      "learning_rate": 3e-05,
      "loss": 2.5099,
      "step": 464
    },
    {
      "epoch": 0.14347423634680653,
      "grad_norm": 0.49275124073028564,
      "learning_rate": 3e-05,
      "loss": 2.6418,
      "step": 465
    },
    {
      "epoch": 0.1437827830916384,
      "grad_norm": 0.39669597148895264,
      "learning_rate": 3e-05,
      "loss": 2.3448,
      "step": 466
    },
    {
      "epoch": 0.14409132983647022,
      "grad_norm": 0.6432366967201233,
      "learning_rate": 3e-05,
      "loss": 2.9027,
      "step": 467
    },
    {
      "epoch": 0.14439987658130207,
      "grad_norm": 0.43117955327033997,
      "learning_rate": 3e-05,
      "loss": 2.7841,
      "step": 468
    },
    {
      "epoch": 0.1447084233261339,
      "grad_norm": 0.4256863296031952,
      "learning_rate": 3e-05,
      "loss": 2.5825,
      "step": 469
    },
    {
      "epoch": 0.14501697007096576,
      "grad_norm": 0.43717724084854126,
      "learning_rate": 3e-05,
      "loss": 2.6433,
      "step": 470
    },
    {
      "epoch": 0.14532551681579758,
      "grad_norm": 0.759493887424469,
      "learning_rate": 3e-05,
      "loss": 2.7323,
      "step": 471
    },
    {
      "epoch": 0.14563406356062944,
      "grad_norm": 0.33263617753982544,
      "learning_rate": 3e-05,
      "loss": 2.417,
      "step": 472
    },
    {
      "epoch": 0.14594261030546127,
      "grad_norm": 0.34977760910987854,
      "learning_rate": 3e-05,
      "loss": 2.3743,
      "step": 473
    },
    {
      "epoch": 0.14625115705029312,
      "grad_norm": 0.5000836253166199,
      "learning_rate": 3e-05,
      "loss": 2.5705,
      "step": 474
    },
    {
      "epoch": 0.14655970379512495,
      "grad_norm": 1.2762246131896973,
      "learning_rate": 3e-05,
      "loss": 3.133,
      "step": 475
    },
    {
      "epoch": 0.1468682505399568,
      "grad_norm": 0.6496342420578003,
      "learning_rate": 3e-05,
      "loss": 2.9256,
      "step": 476
    },
    {
      "epoch": 0.14717679728478864,
      "grad_norm": 0.716373085975647,
      "learning_rate": 3e-05,
      "loss": 2.5074,
      "step": 477
    },
    {
      "epoch": 0.1474853440296205,
      "grad_norm": 0.7680580019950867,
      "learning_rate": 3e-05,
      "loss": 2.8563,
      "step": 478
    },
    {
      "epoch": 0.14779389077445232,
      "grad_norm": 0.4760282337665558,
      "learning_rate": 3e-05,
      "loss": 2.4567,
      "step": 479
    },
    {
      "epoch": 0.14810243751928417,
      "grad_norm": 0.7085632681846619,
      "learning_rate": 3e-05,
      "loss": 3.3149,
      "step": 480
    },
    {
      "epoch": 0.148410984264116,
      "grad_norm": 0.48510226607322693,
      "learning_rate": 3e-05,
      "loss": 2.9825,
      "step": 481
    },
    {
      "epoch": 0.14871953100894786,
      "grad_norm": 0.5194237232208252,
      "learning_rate": 3e-05,
      "loss": 2.8049,
      "step": 482
    },
    {
      "epoch": 0.1490280777537797,
      "grad_norm": 0.6330525279045105,
      "learning_rate": 3e-05,
      "loss": 3.0564,
      "step": 483
    },
    {
      "epoch": 0.14933662449861154,
      "grad_norm": 0.38547223806381226,
      "learning_rate": 3e-05,
      "loss": 2.7878,
      "step": 484
    },
    {
      "epoch": 0.14964517124344337,
      "grad_norm": 0.5610853433609009,
      "learning_rate": 3e-05,
      "loss": 3.0895,
      "step": 485
    },
    {
      "epoch": 0.14995371798827523,
      "grad_norm": 0.6889973878860474,
      "learning_rate": 3e-05,
      "loss": 3.396,
      "step": 486
    },
    {
      "epoch": 0.15026226473310705,
      "grad_norm": 0.374765008687973,
      "learning_rate": 3e-05,
      "loss": 2.2703,
      "step": 487
    },
    {
      "epoch": 0.1505708114779389,
      "grad_norm": 0.5405288338661194,
      "learning_rate": 3e-05,
      "loss": 2.8522,
      "step": 488
    },
    {
      "epoch": 0.15087935822277074,
      "grad_norm": 0.4556944668292999,
      "learning_rate": 3e-05,
      "loss": 2.9866,
      "step": 489
    },
    {
      "epoch": 0.1511879049676026,
      "grad_norm": 0.326506108045578,
      "learning_rate": 3e-05,
      "loss": 2.2067,
      "step": 490
    },
    {
      "epoch": 0.15149645171243442,
      "grad_norm": 0.3182726204395294,
      "learning_rate": 3e-05,
      "loss": 2.134,
      "step": 491
    },
    {
      "epoch": 0.15180499845726628,
      "grad_norm": 0.3820968568325043,
      "learning_rate": 3e-05,
      "loss": 2.5072,
      "step": 492
    },
    {
      "epoch": 0.1521135452020981,
      "grad_norm": 0.43182340264320374,
      "learning_rate": 3e-05,
      "loss": 2.5903,
      "step": 493
    },
    {
      "epoch": 0.15242209194692996,
      "grad_norm": 0.614032506942749,
      "learning_rate": 3e-05,
      "loss": 3.0159,
      "step": 494
    },
    {
      "epoch": 0.1527306386917618,
      "grad_norm": 0.34464317560195923,
      "learning_rate": 3e-05,
      "loss": 2.5565,
      "step": 495
    },
    {
      "epoch": 0.15303918543659364,
      "grad_norm": 0.7838470935821533,
      "learning_rate": 3e-05,
      "loss": 3.299,
      "step": 496
    },
    {
      "epoch": 0.15334773218142547,
      "grad_norm": 0.7292729616165161,
      "learning_rate": 3e-05,
      "loss": 3.1353,
      "step": 497
    },
    {
      "epoch": 0.15365627892625733,
      "grad_norm": 0.6598738431930542,
      "learning_rate": 3e-05,
      "loss": 3.0993,
      "step": 498
    },
    {
      "epoch": 0.15396482567108918,
      "grad_norm": 0.3278651535511017,
      "learning_rate": 3e-05,
      "loss": 2.4751,
      "step": 499
    },
    {
      "epoch": 0.154273372415921,
      "grad_norm": 0.3951056897640228,
      "learning_rate": 3e-05,
      "loss": 2.3225,
      "step": 500
    },
    {
      "epoch": 0.15458191916075287,
      "grad_norm": 0.5355321764945984,
      "learning_rate": 3e-05,
      "loss": 2.6751,
      "step": 501
    },
    {
      "epoch": 0.1548904659055847,
      "grad_norm": 0.4398941993713379,
      "learning_rate": 3e-05,
      "loss": 2.3192,
      "step": 502
    },
    {
      "epoch": 0.15519901265041655,
      "grad_norm": 0.9720609784126282,
      "learning_rate": 3e-05,
      "loss": 3.1218,
      "step": 503
    },
    {
      "epoch": 0.15550755939524838,
      "grad_norm": 0.5093293190002441,
      "learning_rate": 3e-05,
      "loss": 2.6208,
      "step": 504
    },
    {
      "epoch": 0.15581610614008023,
      "grad_norm": 0.5958328247070312,
      "learning_rate": 3e-05,
      "loss": 2.9366,
      "step": 505
    },
    {
      "epoch": 0.15612465288491206,
      "grad_norm": 1.0065058469772339,
      "learning_rate": 3e-05,
      "loss": 3.086,
      "step": 506
    },
    {
      "epoch": 0.15643319962974392,
      "grad_norm": 0.9630736708641052,
      "learning_rate": 3e-05,
      "loss": 2.7257,
      "step": 507
    },
    {
      "epoch": 0.15674174637457575,
      "grad_norm": 0.730649471282959,
      "learning_rate": 3e-05,
      "loss": 2.6617,
      "step": 508
    },
    {
      "epoch": 0.1570502931194076,
      "grad_norm": 0.9365665316581726,
      "learning_rate": 3e-05,
      "loss": 3.1122,
      "step": 509
    },
    {
      "epoch": 0.15735883986423943,
      "grad_norm": 0.6446232795715332,
      "learning_rate": 3e-05,
      "loss": 2.7233,
      "step": 510
    },
    {
      "epoch": 0.15766738660907129,
      "grad_norm": 0.37989816069602966,
      "learning_rate": 3e-05,
      "loss": 2.6314,
      "step": 511
    },
    {
      "epoch": 0.15797593335390311,
      "grad_norm": 0.7095012068748474,
      "learning_rate": 3e-05,
      "loss": 3.1006,
      "step": 512
    },
    {
      "epoch": 0.15828448009873497,
      "grad_norm": 0.7077836990356445,
      "learning_rate": 3e-05,
      "loss": 2.7112,
      "step": 513
    },
    {
      "epoch": 0.1585930268435668,
      "grad_norm": 0.6317113637924194,
      "learning_rate": 3e-05,
      "loss": 2.8544,
      "step": 514
    },
    {
      "epoch": 0.15890157358839865,
      "grad_norm": 0.53568035364151,
      "learning_rate": 3e-05,
      "loss": 2.8805,
      "step": 515
    },
    {
      "epoch": 0.15921012033323048,
      "grad_norm": 0.42564329504966736,
      "learning_rate": 3e-05,
      "loss": 2.4129,
      "step": 516
    },
    {
      "epoch": 0.15951866707806234,
      "grad_norm": 0.8857656121253967,
      "learning_rate": 3e-05,
      "loss": 3.2094,
      "step": 517
    },
    {
      "epoch": 0.15982721382289417,
      "grad_norm": 0.717786431312561,
      "learning_rate": 3e-05,
      "loss": 3.2728,
      "step": 518
    },
    {
      "epoch": 0.16013576056772602,
      "grad_norm": 0.5533412098884583,
      "learning_rate": 3e-05,
      "loss": 3.1927,
      "step": 519
    },
    {
      "epoch": 0.16044430731255785,
      "grad_norm": 0.5395249724388123,
      "learning_rate": 3e-05,
      "loss": 2.2287,
      "step": 520
    },
    {
      "epoch": 0.1607528540573897,
      "grad_norm": 1.1015528440475464,
      "learning_rate": 3e-05,
      "loss": 3.1084,
      "step": 521
    },
    {
      "epoch": 0.16106140080222153,
      "grad_norm": 0.4234045743942261,
      "learning_rate": 3e-05,
      "loss": 2.7718,
      "step": 522
    },
    {
      "epoch": 0.1613699475470534,
      "grad_norm": 0.5807662606239319,
      "learning_rate": 3e-05,
      "loss": 2.5903,
      "step": 523
    },
    {
      "epoch": 0.16167849429188522,
      "grad_norm": 0.6421242356300354,
      "learning_rate": 3e-05,
      "loss": 2.7322,
      "step": 524
    },
    {
      "epoch": 0.16198704103671707,
      "grad_norm": 0.5576215386390686,
      "learning_rate": 3e-05,
      "loss": 2.4455,
      "step": 525
    },
    {
      "epoch": 0.1622955877815489,
      "grad_norm": 0.3627455234527588,
      "learning_rate": 3e-05,
      "loss": 2.6403,
      "step": 526
    },
    {
      "epoch": 0.16260413452638076,
      "grad_norm": 0.6139925122261047,
      "learning_rate": 3e-05,
      "loss": 2.7791,
      "step": 527
    },
    {
      "epoch": 0.16291268127121258,
      "grad_norm": 0.475038081407547,
      "learning_rate": 3e-05,
      "loss": 2.3931,
      "step": 528
    },
    {
      "epoch": 0.16322122801604444,
      "grad_norm": 0.5573770999908447,
      "learning_rate": 3e-05,
      "loss": 3.0576,
      "step": 529
    },
    {
      "epoch": 0.16352977476087627,
      "grad_norm": 0.3340505361557007,
      "learning_rate": 3e-05,
      "loss": 2.427,
      "step": 530
    },
    {
      "epoch": 0.16383832150570812,
      "grad_norm": 0.3796028792858124,
      "learning_rate": 3e-05,
      "loss": 2.623,
      "step": 531
    },
    {
      "epoch": 0.16414686825053995,
      "grad_norm": 0.499763160943985,
      "learning_rate": 3e-05,
      "loss": 2.6086,
      "step": 532
    },
    {
      "epoch": 0.1644554149953718,
      "grad_norm": 0.5351141691207886,
      "learning_rate": 3e-05,
      "loss": 2.773,
      "step": 533
    },
    {
      "epoch": 0.16476396174020363,
      "grad_norm": 0.5103338956832886,
      "learning_rate": 3e-05,
      "loss": 2.8165,
      "step": 534
    },
    {
      "epoch": 0.1650725084850355,
      "grad_norm": 0.49195244908332825,
      "learning_rate": 3e-05,
      "loss": 2.8805,
      "step": 535
    },
    {
      "epoch": 0.16538105522986732,
      "grad_norm": 0.6390337347984314,
      "learning_rate": 3e-05,
      "loss": 2.556,
      "step": 536
    },
    {
      "epoch": 0.16568960197469917,
      "grad_norm": 0.6301625370979309,
      "learning_rate": 3e-05,
      "loss": 2.8134,
      "step": 537
    },
    {
      "epoch": 0.165998148719531,
      "grad_norm": 0.30376285314559937,
      "learning_rate": 3e-05,
      "loss": 2.1669,
      "step": 538
    },
    {
      "epoch": 0.16630669546436286,
      "grad_norm": 0.4260924756526947,
      "learning_rate": 3e-05,
      "loss": 2.578,
      "step": 539
    },
    {
      "epoch": 0.16661524220919469,
      "grad_norm": 0.9274812936782837,
      "learning_rate": 3e-05,
      "loss": 3.1846,
      "step": 540
    },
    {
      "epoch": 0.16692378895402654,
      "grad_norm": 0.31101086735725403,
      "learning_rate": 3e-05,
      "loss": 2.3578,
      "step": 541
    },
    {
      "epoch": 0.16723233569885837,
      "grad_norm": 0.4664568603038788,
      "learning_rate": 3e-05,
      "loss": 2.6042,
      "step": 542
    },
    {
      "epoch": 0.16754088244369023,
      "grad_norm": 0.41750529408454895,
      "learning_rate": 3e-05,
      "loss": 2.707,
      "step": 543
    },
    {
      "epoch": 0.16784942918852205,
      "grad_norm": 0.745094895362854,
      "learning_rate": 3e-05,
      "loss": 2.5031,
      "step": 544
    },
    {
      "epoch": 0.1681579759333539,
      "grad_norm": 0.53835129737854,
      "learning_rate": 3e-05,
      "loss": 2.6455,
      "step": 545
    },
    {
      "epoch": 0.16846652267818574,
      "grad_norm": 0.47485771775245667,
      "learning_rate": 3e-05,
      "loss": 2.8771,
      "step": 546
    },
    {
      "epoch": 0.1687750694230176,
      "grad_norm": 0.427839457988739,
      "learning_rate": 3e-05,
      "loss": 2.4684,
      "step": 547
    },
    {
      "epoch": 0.16908361616784942,
      "grad_norm": 0.459409236907959,
      "learning_rate": 3e-05,
      "loss": 2.834,
      "step": 548
    },
    {
      "epoch": 0.16939216291268128,
      "grad_norm": 0.594539225101471,
      "learning_rate": 3e-05,
      "loss": 2.8524,
      "step": 549
    },
    {
      "epoch": 0.1697007096575131,
      "grad_norm": 0.3290991485118866,
      "learning_rate": 3e-05,
      "loss": 2.4509,
      "step": 550
    },
    {
      "epoch": 0.17000925640234496,
      "grad_norm": 0.6383848786354065,
      "learning_rate": 3e-05,
      "loss": 3.06,
      "step": 551
    },
    {
      "epoch": 0.1703178031471768,
      "grad_norm": 0.4995775818824768,
      "learning_rate": 3e-05,
      "loss": 2.6543,
      "step": 552
    },
    {
      "epoch": 0.17062634989200864,
      "grad_norm": 0.2920311391353607,
      "learning_rate": 3e-05,
      "loss": 2.3851,
      "step": 553
    },
    {
      "epoch": 0.17093489663684047,
      "grad_norm": 0.37657687067985535,
      "learning_rate": 3e-05,
      "loss": 2.5885,
      "step": 554
    },
    {
      "epoch": 0.17124344338167233,
      "grad_norm": 0.3489648997783661,
      "learning_rate": 3e-05,
      "loss": 2.3189,
      "step": 555
    },
    {
      "epoch": 0.17155199012650416,
      "grad_norm": 0.5701825618743896,
      "learning_rate": 3e-05,
      "loss": 2.7273,
      "step": 556
    },
    {
      "epoch": 0.171860536871336,
      "grad_norm": 0.4302188456058502,
      "learning_rate": 3e-05,
      "loss": 2.7104,
      "step": 557
    },
    {
      "epoch": 0.17216908361616784,
      "grad_norm": 0.39768311381340027,
      "learning_rate": 3e-05,
      "loss": 2.7275,
      "step": 558
    },
    {
      "epoch": 0.1724776303609997,
      "grad_norm": 0.3526521623134613,
      "learning_rate": 3e-05,
      "loss": 2.4574,
      "step": 559
    },
    {
      "epoch": 0.17278617710583152,
      "grad_norm": 0.4580765962600708,
      "learning_rate": 3e-05,
      "loss": 2.6142,
      "step": 560
    },
    {
      "epoch": 0.17309472385066338,
      "grad_norm": 0.6257379055023193,
      "learning_rate": 3e-05,
      "loss": 2.6212,
      "step": 561
    },
    {
      "epoch": 0.1734032705954952,
      "grad_norm": 0.5225767493247986,
      "learning_rate": 3e-05,
      "loss": 3.0037,
      "step": 562
    },
    {
      "epoch": 0.17371181734032706,
      "grad_norm": 0.3681781589984894,
      "learning_rate": 3e-05,
      "loss": 2.4477,
      "step": 563
    },
    {
      "epoch": 0.1740203640851589,
      "grad_norm": 0.3147395849227905,
      "learning_rate": 3e-05,
      "loss": 2.291,
      "step": 564
    },
    {
      "epoch": 0.17432891082999075,
      "grad_norm": 0.5599997639656067,
      "learning_rate": 3e-05,
      "loss": 2.761,
      "step": 565
    },
    {
      "epoch": 0.17463745757482257,
      "grad_norm": 0.895828127861023,
      "learning_rate": 3e-05,
      "loss": 3.2657,
      "step": 566
    },
    {
      "epoch": 0.17494600431965443,
      "grad_norm": 0.48360684514045715,
      "learning_rate": 3e-05,
      "loss": 3.0389,
      "step": 567
    },
    {
      "epoch": 0.17525455106448626,
      "grad_norm": 0.4734765887260437,
      "learning_rate": 3e-05,
      "loss": 2.7359,
      "step": 568
    },
    {
      "epoch": 0.1755630978093181,
      "grad_norm": 0.5472472310066223,
      "learning_rate": 3e-05,
      "loss": 2.5452,
      "step": 569
    },
    {
      "epoch": 0.17587164455414994,
      "grad_norm": 0.5069577693939209,
      "learning_rate": 3e-05,
      "loss": 3.0763,
      "step": 570
    },
    {
      "epoch": 0.1761801912989818,
      "grad_norm": 0.6229672431945801,
      "learning_rate": 3e-05,
      "loss": 3.2135,
      "step": 571
    },
    {
      "epoch": 0.17648873804381363,
      "grad_norm": 0.5861048102378845,
      "learning_rate": 3e-05,
      "loss": 3.4906,
      "step": 572
    },
    {
      "epoch": 0.17679728478864548,
      "grad_norm": 0.4805934429168701,
      "learning_rate": 3e-05,
      "loss": 2.8969,
      "step": 573
    },
    {
      "epoch": 0.1771058315334773,
      "grad_norm": 0.6094735860824585,
      "learning_rate": 3e-05,
      "loss": 3.0226,
      "step": 574
    },
    {
      "epoch": 0.17741437827830916,
      "grad_norm": 0.44033339619636536,
      "learning_rate": 3e-05,
      "loss": 2.4886,
      "step": 575
    },
    {
      "epoch": 0.177722925023141,
      "grad_norm": 1.1240124702453613,
      "learning_rate": 3e-05,
      "loss": 3.4272,
      "step": 576
    },
    {
      "epoch": 0.17803147176797285,
      "grad_norm": 0.4928964078426361,
      "learning_rate": 3e-05,
      "loss": 2.3991,
      "step": 577
    },
    {
      "epoch": 0.17834001851280468,
      "grad_norm": 0.45760032534599304,
      "learning_rate": 3e-05,
      "loss": 2.5567,
      "step": 578
    },
    {
      "epoch": 0.17864856525763653,
      "grad_norm": 0.46122652292251587,
      "learning_rate": 3e-05,
      "loss": 2.4686,
      "step": 579
    },
    {
      "epoch": 0.17895711200246836,
      "grad_norm": 0.4241897761821747,
      "learning_rate": 3e-05,
      "loss": 2.3896,
      "step": 580
    },
    {
      "epoch": 0.17926565874730022,
      "grad_norm": 0.7329486012458801,
      "learning_rate": 3e-05,
      "loss": 2.8377,
      "step": 581
    },
    {
      "epoch": 0.17957420549213207,
      "grad_norm": 0.5408822894096375,
      "learning_rate": 3e-05,
      "loss": 2.895,
      "step": 582
    },
    {
      "epoch": 0.1798827522369639,
      "grad_norm": 0.4442683756351471,
      "learning_rate": 3e-05,
      "loss": 2.4398,
      "step": 583
    },
    {
      "epoch": 0.18019129898179576,
      "grad_norm": 0.7230685949325562,
      "learning_rate": 3e-05,
      "loss": 3.0208,
      "step": 584
    },
    {
      "epoch": 0.18049984572662758,
      "grad_norm": 0.45991250872612,
      "learning_rate": 3e-05,
      "loss": 2.8429,
      "step": 585
    },
    {
      "epoch": 0.18080839247145944,
      "grad_norm": 0.5567431449890137,
      "learning_rate": 3e-05,
      "loss": 3.0907,
      "step": 586
    },
    {
      "epoch": 0.18111693921629127,
      "grad_norm": 0.39667779207229614,
      "learning_rate": 3e-05,
      "loss": 2.5474,
      "step": 587
    },
    {
      "epoch": 0.18142548596112312,
      "grad_norm": 0.4259324371814728,
      "learning_rate": 3e-05,
      "loss": 2.937,
      "step": 588
    },
    {
      "epoch": 0.18173403270595495,
      "grad_norm": 0.5309762358665466,
      "learning_rate": 3e-05,
      "loss": 2.904,
      "step": 589
    },
    {
      "epoch": 0.1820425794507868,
      "grad_norm": 0.4646052420139313,
      "learning_rate": 3e-05,
      "loss": 2.7825,
      "step": 590
    },
    {
      "epoch": 0.18235112619561863,
      "grad_norm": 0.5502670407295227,
      "learning_rate": 3e-05,
      "loss": 3.3798,
      "step": 591
    },
    {
      "epoch": 0.1826596729404505,
      "grad_norm": 0.4678283929824829,
      "learning_rate": 3e-05,
      "loss": 2.7709,
      "step": 592
    },
    {
      "epoch": 0.18296821968528232,
      "grad_norm": 0.44272318482398987,
      "learning_rate": 3e-05,
      "loss": 2.7197,
      "step": 593
    },
    {
      "epoch": 0.18327676643011417,
      "grad_norm": 0.34605056047439575,
      "learning_rate": 3e-05,
      "loss": 2.6256,
      "step": 594
    },
    {
      "epoch": 0.183585313174946,
      "grad_norm": 0.32907217741012573,
      "learning_rate": 3e-05,
      "loss": 2.183,
      "step": 595
    },
    {
      "epoch": 0.18389385991977786,
      "grad_norm": 0.5004732608795166,
      "learning_rate": 3e-05,
      "loss": 2.4808,
      "step": 596
    },
    {
      "epoch": 0.18420240666460969,
      "grad_norm": 0.4207105338573456,
      "learning_rate": 3e-05,
      "loss": 2.7112,
      "step": 597
    },
    {
      "epoch": 0.18451095340944154,
      "grad_norm": 0.4427211284637451,
      "learning_rate": 3e-05,
      "loss": 2.7198,
      "step": 598
    },
    {
      "epoch": 0.18481950015427337,
      "grad_norm": 0.3510686159133911,
      "learning_rate": 3e-05,
      "loss": 2.6319,
      "step": 599
    },
    {
      "epoch": 0.18512804689910523,
      "grad_norm": 0.3347536623477936,
      "learning_rate": 3e-05,
      "loss": 2.479,
      "step": 600
    },
    {
      "epoch": 0.18543659364393705,
      "grad_norm": 0.4486810266971588,
      "learning_rate": 3e-05,
      "loss": 2.7633,
      "step": 601
    },
    {
      "epoch": 0.1857451403887689,
      "grad_norm": 0.7267583012580872,
      "learning_rate": 3e-05,
      "loss": 3.2443,
      "step": 602
    },
    {
      "epoch": 0.18605368713360074,
      "grad_norm": 0.5485150814056396,
      "learning_rate": 3e-05,
      "loss": 3.0257,
      "step": 603
    },
    {
      "epoch": 0.1863622338784326,
      "grad_norm": 0.6103752255439758,
      "learning_rate": 3e-05,
      "loss": 2.8336,
      "step": 604
    },
    {
      "epoch": 0.18667078062326442,
      "grad_norm": 0.4755418002605438,
      "learning_rate": 3e-05,
      "loss": 2.7703,
      "step": 605
    },
    {
      "epoch": 0.18697932736809628,
      "grad_norm": 0.523270308971405,
      "learning_rate": 3e-05,
      "loss": 2.792,
      "step": 606
    },
    {
      "epoch": 0.1872878741129281,
      "grad_norm": 0.7174893021583557,
      "learning_rate": 3e-05,
      "loss": 3.1485,
      "step": 607
    },
    {
      "epoch": 0.18759642085775996,
      "grad_norm": 0.4338546097278595,
      "learning_rate": 3e-05,
      "loss": 2.6676,
      "step": 608
    },
    {
      "epoch": 0.1879049676025918,
      "grad_norm": 0.5212303996086121,
      "learning_rate": 3e-05,
      "loss": 2.5303,
      "step": 609
    },
    {
      "epoch": 0.18821351434742364,
      "grad_norm": 0.4013713002204895,
      "learning_rate": 3e-05,
      "loss": 2.5814,
      "step": 610
    },
    {
      "epoch": 0.18852206109225547,
      "grad_norm": 0.38565197587013245,
      "learning_rate": 3e-05,
      "loss": 2.5641,
      "step": 611
    },
    {
      "epoch": 0.18883060783708733,
      "grad_norm": 0.5289196372032166,
      "learning_rate": 3e-05,
      "loss": 3.1017,
      "step": 612
    },
    {
      "epoch": 0.18913915458191916,
      "grad_norm": 0.4261009097099304,
      "learning_rate": 3e-05,
      "loss": 2.7536,
      "step": 613
    },
    {
      "epoch": 0.189447701326751,
      "grad_norm": 0.6289636492729187,
      "learning_rate": 3e-05,
      "loss": 2.5995,
      "step": 614
    },
    {
      "epoch": 0.18975624807158284,
      "grad_norm": 0.3841557800769806,
      "learning_rate": 3e-05,
      "loss": 2.9889,
      "step": 615
    },
    {
      "epoch": 0.1900647948164147,
      "grad_norm": 0.37914833426475525,
      "learning_rate": 3e-05,
      "loss": 2.4147,
      "step": 616
    },
    {
      "epoch": 0.19037334156124652,
      "grad_norm": 0.5321159958839417,
      "learning_rate": 3e-05,
      "loss": 2.8679,
      "step": 617
    },
    {
      "epoch": 0.19068188830607838,
      "grad_norm": 0.4842563271522522,
      "learning_rate": 3e-05,
      "loss": 2.7819,
      "step": 618
    },
    {
      "epoch": 0.1909904350509102,
      "grad_norm": 0.35459670424461365,
      "learning_rate": 3e-05,
      "loss": 2.768,
      "step": 619
    },
    {
      "epoch": 0.19129898179574206,
      "grad_norm": 0.6710373163223267,
      "learning_rate": 3e-05,
      "loss": 3.077,
      "step": 620
    },
    {
      "epoch": 0.1916075285405739,
      "grad_norm": 0.4304696023464203,
      "learning_rate": 3e-05,
      "loss": 2.4043,
      "step": 621
    },
    {
      "epoch": 0.19191607528540575,
      "grad_norm": 0.500830352306366,
      "learning_rate": 3e-05,
      "loss": 2.6387,
      "step": 622
    },
    {
      "epoch": 0.19222462203023757,
      "grad_norm": 0.5076797008514404,
      "learning_rate": 3e-05,
      "loss": 2.3244,
      "step": 623
    },
    {
      "epoch": 0.19253316877506943,
      "grad_norm": 0.30540212988853455,
      "learning_rate": 3e-05,
      "loss": 2.0853,
      "step": 624
    },
    {
      "epoch": 0.19284171551990126,
      "grad_norm": 0.6184133887290955,
      "learning_rate": 3e-05,
      "loss": 2.6062,
      "step": 625
    },
    {
      "epoch": 0.1931502622647331,
      "grad_norm": 0.4118046164512634,
      "learning_rate": 3e-05,
      "loss": 2.1961,
      "step": 626
    },
    {
      "epoch": 0.19345880900956494,
      "grad_norm": 0.6084356904029846,
      "learning_rate": 3e-05,
      "loss": 2.8398,
      "step": 627
    },
    {
      "epoch": 0.1937673557543968,
      "grad_norm": 0.44872429966926575,
      "learning_rate": 3e-05,
      "loss": 2.9109,
      "step": 628
    },
    {
      "epoch": 0.19407590249922863,
      "grad_norm": 0.5909293293952942,
      "learning_rate": 3e-05,
      "loss": 2.4278,
      "step": 629
    },
    {
      "epoch": 0.19438444924406048,
      "grad_norm": 0.5620918273925781,
      "learning_rate": 3e-05,
      "loss": 2.8395,
      "step": 630
    },
    {
      "epoch": 0.1946929959888923,
      "grad_norm": 0.5947256088256836,
      "learning_rate": 3e-05,
      "loss": 3.1204,
      "step": 631
    },
    {
      "epoch": 0.19500154273372416,
      "grad_norm": 0.42888760566711426,
      "learning_rate": 3e-05,
      "loss": 2.6668,
      "step": 632
    },
    {
      "epoch": 0.195310089478556,
      "grad_norm": 0.8526038527488708,
      "learning_rate": 3e-05,
      "loss": 2.9337,
      "step": 633
    },
    {
      "epoch": 0.19561863622338785,
      "grad_norm": 0.5663727521896362,
      "learning_rate": 3e-05,
      "loss": 3.103,
      "step": 634
    },
    {
      "epoch": 0.19592718296821968,
      "grad_norm": 0.5431751012802124,
      "learning_rate": 3e-05,
      "loss": 2.6397,
      "step": 635
    },
    {
      "epoch": 0.19623572971305153,
      "grad_norm": 0.5891326069831848,
      "learning_rate": 3e-05,
      "loss": 3.2095,
      "step": 636
    },
    {
      "epoch": 0.19654427645788336,
      "grad_norm": 0.5660468339920044,
      "learning_rate": 3e-05,
      "loss": 2.9853,
      "step": 637
    },
    {
      "epoch": 0.19685282320271522,
      "grad_norm": 0.3460230529308319,
      "learning_rate": 3e-05,
      "loss": 2.4755,
      "step": 638
    },
    {
      "epoch": 0.19716136994754704,
      "grad_norm": 0.30621814727783203,
      "learning_rate": 3e-05,
      "loss": 2.5505,
      "step": 639
    },
    {
      "epoch": 0.1974699166923789,
      "grad_norm": 0.4897528290748596,
      "learning_rate": 3e-05,
      "loss": 2.8266,
      "step": 640
    },
    {
      "epoch": 0.19777846343721073,
      "grad_norm": 0.7390705347061157,
      "learning_rate": 3e-05,
      "loss": 3.0495,
      "step": 641
    },
    {
      "epoch": 0.19808701018204258,
      "grad_norm": 0.403690904378891,
      "learning_rate": 3e-05,
      "loss": 2.6824,
      "step": 642
    },
    {
      "epoch": 0.1983955569268744,
      "grad_norm": 0.6214193105697632,
      "learning_rate": 3e-05,
      "loss": 2.9677,
      "step": 643
    },
    {
      "epoch": 0.19870410367170627,
      "grad_norm": 0.3992424011230469,
      "learning_rate": 3e-05,
      "loss": 2.7133,
      "step": 644
    },
    {
      "epoch": 0.1990126504165381,
      "grad_norm": 0.5738241672515869,
      "learning_rate": 3e-05,
      "loss": 2.7717,
      "step": 645
    },
    {
      "epoch": 0.19932119716136995,
      "grad_norm": 0.5687406659126282,
      "learning_rate": 3e-05,
      "loss": 3.1489,
      "step": 646
    },
    {
      "epoch": 0.19962974390620178,
      "grad_norm": 0.44407641887664795,
      "learning_rate": 3e-05,
      "loss": 3.0794,
      "step": 647
    },
    {
      "epoch": 0.19993829065103363,
      "grad_norm": 0.5712667107582092,
      "learning_rate": 3e-05,
      "loss": 2.7482,
      "step": 648
    },
    {
      "epoch": 0.20024683739586546,
      "grad_norm": 0.3764342963695526,
      "learning_rate": 3e-05,
      "loss": 2.6002,
      "step": 649
    },
    {
      "epoch": 0.20055538414069732,
      "grad_norm": 0.3133735954761505,
      "learning_rate": 3e-05,
      "loss": 2.3715,
      "step": 650
    },
    {
      "epoch": 0.20086393088552915,
      "grad_norm": 0.5320349931716919,
      "learning_rate": 3e-05,
      "loss": 3.0982,
      "step": 651
    },
    {
      "epoch": 0.201172477630361,
      "grad_norm": 0.498043030500412,
      "learning_rate": 3e-05,
      "loss": 3.0611,
      "step": 652
    },
    {
      "epoch": 0.20148102437519283,
      "grad_norm": 0.611578106880188,
      "learning_rate": 3e-05,
      "loss": 3.2704,
      "step": 653
    },
    {
      "epoch": 0.20178957112002469,
      "grad_norm": 0.43669798970222473,
      "learning_rate": 3e-05,
      "loss": 2.5288,
      "step": 654
    },
    {
      "epoch": 0.2020981178648565,
      "grad_norm": 0.4220907688140869,
      "learning_rate": 3e-05,
      "loss": 2.8483,
      "step": 655
    },
    {
      "epoch": 0.20240666460968837,
      "grad_norm": 0.46355968713760376,
      "learning_rate": 3e-05,
      "loss": 3.1131,
      "step": 656
    },
    {
      "epoch": 0.2027152113545202,
      "grad_norm": 0.3843071758747101,
      "learning_rate": 3e-05,
      "loss": 2.6449,
      "step": 657
    },
    {
      "epoch": 0.20302375809935205,
      "grad_norm": 0.3454858660697937,
      "learning_rate": 3e-05,
      "loss": 2.4922,
      "step": 658
    },
    {
      "epoch": 0.20333230484418388,
      "grad_norm": 0.31287965178489685,
      "learning_rate": 3e-05,
      "loss": 2.5445,
      "step": 659
    },
    {
      "epoch": 0.20364085158901574,
      "grad_norm": 0.49061527848243713,
      "learning_rate": 3e-05,
      "loss": 2.4979,
      "step": 660
    },
    {
      "epoch": 0.20394939833384756,
      "grad_norm": 0.5406720638275146,
      "learning_rate": 3e-05,
      "loss": 2.7902,
      "step": 661
    },
    {
      "epoch": 0.20425794507867942,
      "grad_norm": 0.5429649353027344,
      "learning_rate": 3e-05,
      "loss": 2.8121,
      "step": 662
    },
    {
      "epoch": 0.20456649182351125,
      "grad_norm": 0.47672000527381897,
      "learning_rate": 3e-05,
      "loss": 3.1438,
      "step": 663
    },
    {
      "epoch": 0.2048750385683431,
      "grad_norm": 0.3863488435745239,
      "learning_rate": 3e-05,
      "loss": 2.4956,
      "step": 664
    },
    {
      "epoch": 0.20518358531317496,
      "grad_norm": 0.5311136245727539,
      "learning_rate": 3e-05,
      "loss": 2.404,
      "step": 665
    },
    {
      "epoch": 0.2054921320580068,
      "grad_norm": 0.30825114250183105,
      "learning_rate": 3e-05,
      "loss": 2.2225,
      "step": 666
    },
    {
      "epoch": 0.20580067880283864,
      "grad_norm": 0.4122128188610077,
      "learning_rate": 3e-05,
      "loss": 2.4823,
      "step": 667
    },
    {
      "epoch": 0.20610922554767047,
      "grad_norm": 0.45997029542922974,
      "learning_rate": 3e-05,
      "loss": 2.456,
      "step": 668
    },
    {
      "epoch": 0.20641777229250233,
      "grad_norm": 0.38074791431427,
      "learning_rate": 3e-05,
      "loss": 2.4485,
      "step": 669
    },
    {
      "epoch": 0.20672631903733416,
      "grad_norm": 0.7642366886138916,
      "learning_rate": 3e-05,
      "loss": 3.2876,
      "step": 670
    },
    {
      "epoch": 0.207034865782166,
      "grad_norm": 0.6110647320747375,
      "learning_rate": 3e-05,
      "loss": 2.901,
      "step": 671
    },
    {
      "epoch": 0.20734341252699784,
      "grad_norm": 0.36458998918533325,
      "learning_rate": 3e-05,
      "loss": 2.2886,
      "step": 672
    },
    {
      "epoch": 0.2076519592718297,
      "grad_norm": 0.49400243163108826,
      "learning_rate": 3e-05,
      "loss": 2.7685,
      "step": 673
    },
    {
      "epoch": 0.20796050601666152,
      "grad_norm": 0.4072665274143219,
      "learning_rate": 3e-05,
      "loss": 2.6262,
      "step": 674
    },
    {
      "epoch": 0.20826905276149338,
      "grad_norm": 0.3744329512119293,
      "learning_rate": 3e-05,
      "loss": 2.6089,
      "step": 675
    },
    {
      "epoch": 0.2085775995063252,
      "grad_norm": 0.4581242501735687,
      "learning_rate": 3e-05,
      "loss": 2.6608,
      "step": 676
    },
    {
      "epoch": 0.20888614625115706,
      "grad_norm": 0.46983492374420166,
      "learning_rate": 3e-05,
      "loss": 2.4597,
      "step": 677
    },
    {
      "epoch": 0.2091946929959889,
      "grad_norm": 0.5508672595024109,
      "learning_rate": 3e-05,
      "loss": 2.8413,
      "step": 678
    },
    {
      "epoch": 0.20950323974082075,
      "grad_norm": 0.3159502148628235,
      "learning_rate": 3e-05,
      "loss": 2.3252,
      "step": 679
    },
    {
      "epoch": 0.20981178648565257,
      "grad_norm": 0.5377674102783203,
      "learning_rate": 3e-05,
      "loss": 2.2233,
      "step": 680
    },
    {
      "epoch": 0.21012033323048443,
      "grad_norm": 0.40905702114105225,
      "learning_rate": 3e-05,
      "loss": 2.1659,
      "step": 681
    },
    {
      "epoch": 0.21042887997531626,
      "grad_norm": 0.3416579067707062,
      "learning_rate": 3e-05,
      "loss": 2.0977,
      "step": 682
    },
    {
      "epoch": 0.2107374267201481,
      "grad_norm": 0.5408785939216614,
      "learning_rate": 3e-05,
      "loss": 3.0337,
      "step": 683
    },
    {
      "epoch": 0.21104597346497994,
      "grad_norm": 0.6323843002319336,
      "learning_rate": 3e-05,
      "loss": 2.9733,
      "step": 684
    },
    {
      "epoch": 0.2113545202098118,
      "grad_norm": 0.8620406985282898,
      "learning_rate": 3e-05,
      "loss": 3.3109,
      "step": 685
    },
    {
      "epoch": 0.21166306695464362,
      "grad_norm": 0.38318270444869995,
      "learning_rate": 3e-05,
      "loss": 2.786,
      "step": 686
    },
    {
      "epoch": 0.21197161369947548,
      "grad_norm": 0.44390928745269775,
      "learning_rate": 3e-05,
      "loss": 2.9674,
      "step": 687
    },
    {
      "epoch": 0.2122801604443073,
      "grad_norm": 0.6823718547821045,
      "learning_rate": 3e-05,
      "loss": 3.0883,
      "step": 688
    },
    {
      "epoch": 0.21258870718913916,
      "grad_norm": 0.451790988445282,
      "learning_rate": 3e-05,
      "loss": 2.5927,
      "step": 689
    },
    {
      "epoch": 0.212897253933971,
      "grad_norm": 0.4029495120048523,
      "learning_rate": 3e-05,
      "loss": 2.6611,
      "step": 690
    },
    {
      "epoch": 0.21320580067880285,
      "grad_norm": 0.4392080307006836,
      "learning_rate": 3e-05,
      "loss": 2.7574,
      "step": 691
    },
    {
      "epoch": 0.21351434742363468,
      "grad_norm": 0.6579667925834656,
      "learning_rate": 3e-05,
      "loss": 3.1202,
      "step": 692
    },
    {
      "epoch": 0.21382289416846653,
      "grad_norm": 0.5804861783981323,
      "learning_rate": 3e-05,
      "loss": 3.3316,
      "step": 693
    },
    {
      "epoch": 0.21413144091329836,
      "grad_norm": 0.4412527084350586,
      "learning_rate": 3e-05,
      "loss": 3.7257,
      "step": 694
    },
    {
      "epoch": 0.21443998765813022,
      "grad_norm": 0.6037363409996033,
      "learning_rate": 3e-05,
      "loss": 2.5479,
      "step": 695
    },
    {
      "epoch": 0.21474853440296204,
      "grad_norm": 0.6885823011398315,
      "learning_rate": 3e-05,
      "loss": 3.063,
      "step": 696
    },
    {
      "epoch": 0.2150570811477939,
      "grad_norm": 0.5372409224510193,
      "learning_rate": 3e-05,
      "loss": 2.7169,
      "step": 697
    },
    {
      "epoch": 0.21536562789262573,
      "grad_norm": 0.5530219674110413,
      "learning_rate": 3e-05,
      "loss": 3.0568,
      "step": 698
    },
    {
      "epoch": 0.21567417463745758,
      "grad_norm": 0.7457128167152405,
      "learning_rate": 3e-05,
      "loss": 2.5894,
      "step": 699
    },
    {
      "epoch": 0.2159827213822894,
      "grad_norm": 0.6711127161979675,
      "learning_rate": 3e-05,
      "loss": 2.7789,
      "step": 700
    },
    {
      "epoch": 0.21629126812712127,
      "grad_norm": 0.6075369119644165,
      "learning_rate": 3e-05,
      "loss": 2.9943,
      "step": 701
    },
    {
      "epoch": 0.2165998148719531,
      "grad_norm": 0.5779369473457336,
      "learning_rate": 3e-05,
      "loss": 2.5556,
      "step": 702
    },
    {
      "epoch": 0.21690836161678495,
      "grad_norm": 0.9133353233337402,
      "learning_rate": 3e-05,
      "loss": 2.6227,
      "step": 703
    },
    {
      "epoch": 0.21721690836161678,
      "grad_norm": 0.8465149402618408,
      "learning_rate": 3e-05,
      "loss": 3.1094,
      "step": 704
    },
    {
      "epoch": 0.21752545510644863,
      "grad_norm": 0.30936139822006226,
      "learning_rate": 3e-05,
      "loss": 2.7395,
      "step": 705
    },
    {
      "epoch": 0.21783400185128046,
      "grad_norm": 0.7300909757614136,
      "learning_rate": 3e-05,
      "loss": 2.9833,
      "step": 706
    },
    {
      "epoch": 0.21814254859611232,
      "grad_norm": 0.7368529438972473,
      "learning_rate": 3e-05,
      "loss": 2.7301,
      "step": 707
    },
    {
      "epoch": 0.21845109534094415,
      "grad_norm": 0.4861367642879486,
      "learning_rate": 3e-05,
      "loss": 2.5439,
      "step": 708
    },
    {
      "epoch": 0.218759642085776,
      "grad_norm": 0.4448826313018799,
      "learning_rate": 3e-05,
      "loss": 2.7462,
      "step": 709
    },
    {
      "epoch": 0.21906818883060783,
      "grad_norm": 0.43111348152160645,
      "learning_rate": 3e-05,
      "loss": 2.2179,
      "step": 710
    },
    {
      "epoch": 0.21937673557543969,
      "grad_norm": 0.6709856390953064,
      "learning_rate": 3e-05,
      "loss": 2.3648,
      "step": 711
    },
    {
      "epoch": 0.2196852823202715,
      "grad_norm": 0.5811859965324402,
      "learning_rate": 3e-05,
      "loss": 2.2098,
      "step": 712
    },
    {
      "epoch": 0.21999382906510337,
      "grad_norm": 0.44703492522239685,
      "learning_rate": 3e-05,
      "loss": 2.4668,
      "step": 713
    },
    {
      "epoch": 0.2203023758099352,
      "grad_norm": 0.32512566447257996,
      "learning_rate": 3e-05,
      "loss": 2.5741,
      "step": 714
    },
    {
      "epoch": 0.22061092255476705,
      "grad_norm": 0.4460441768169403,
      "learning_rate": 3e-05,
      "loss": 2.5822,
      "step": 715
    },
    {
      "epoch": 0.22091946929959888,
      "grad_norm": 0.46903368830680847,
      "learning_rate": 3e-05,
      "loss": 2.3405,
      "step": 716
    },
    {
      "epoch": 0.22122801604443074,
      "grad_norm": 0.4886974096298218,
      "learning_rate": 3e-05,
      "loss": 2.3086,
      "step": 717
    },
    {
      "epoch": 0.22153656278926256,
      "grad_norm": 0.7623785734176636,
      "learning_rate": 3e-05,
      "loss": 3.0441,
      "step": 718
    },
    {
      "epoch": 0.22184510953409442,
      "grad_norm": 0.39860019087791443,
      "learning_rate": 3e-05,
      "loss": 2.7577,
      "step": 719
    },
    {
      "epoch": 0.22215365627892625,
      "grad_norm": 0.5369051098823547,
      "learning_rate": 3e-05,
      "loss": 2.4445,
      "step": 720
    },
    {
      "epoch": 0.2224622030237581,
      "grad_norm": 0.6205699443817139,
      "learning_rate": 3e-05,
      "loss": 3.322,
      "step": 721
    },
    {
      "epoch": 0.22277074976858993,
      "grad_norm": 0.4737318456172943,
      "learning_rate": 3e-05,
      "loss": 2.5368,
      "step": 722
    },
    {
      "epoch": 0.2230792965134218,
      "grad_norm": 0.4158051311969757,
      "learning_rate": 3e-05,
      "loss": 2.3423,
      "step": 723
    },
    {
      "epoch": 0.22338784325825362,
      "grad_norm": 0.44715094566345215,
      "learning_rate": 3e-05,
      "loss": 2.9117,
      "step": 724
    },
    {
      "epoch": 0.22369639000308547,
      "grad_norm": 0.40355610847473145,
      "learning_rate": 3e-05,
      "loss": 2.8261,
      "step": 725
    },
    {
      "epoch": 0.2240049367479173,
      "grad_norm": 0.40845543146133423,
      "learning_rate": 3e-05,
      "loss": 2.9491,
      "step": 726
    },
    {
      "epoch": 0.22431348349274915,
      "grad_norm": 0.40344637632369995,
      "learning_rate": 3e-05,
      "loss": 2.6911,
      "step": 727
    },
    {
      "epoch": 0.22462203023758098,
      "grad_norm": 0.44544005393981934,
      "learning_rate": 3e-05,
      "loss": 2.5146,
      "step": 728
    },
    {
      "epoch": 0.22493057698241284,
      "grad_norm": 0.33173346519470215,
      "learning_rate": 3e-05,
      "loss": 2.3432,
      "step": 729
    },
    {
      "epoch": 0.22523912372724467,
      "grad_norm": 0.3437408208847046,
      "learning_rate": 3e-05,
      "loss": 2.642,
      "step": 730
    },
    {
      "epoch": 0.22554767047207652,
      "grad_norm": 0.3419845700263977,
      "learning_rate": 3e-05,
      "loss": 2.7049,
      "step": 731
    },
    {
      "epoch": 0.22585621721690835,
      "grad_norm": 0.38495057821273804,
      "learning_rate": 3e-05,
      "loss": 2.7977,
      "step": 732
    },
    {
      "epoch": 0.2261647639617402,
      "grad_norm": 0.46839049458503723,
      "learning_rate": 3e-05,
      "loss": 2.3997,
      "step": 733
    },
    {
      "epoch": 0.22647331070657203,
      "grad_norm": 0.5214048624038696,
      "learning_rate": 3e-05,
      "loss": 3.0801,
      "step": 734
    },
    {
      "epoch": 0.2267818574514039,
      "grad_norm": 0.32730308175086975,
      "learning_rate": 3e-05,
      "loss": 2.6502,
      "step": 735
    },
    {
      "epoch": 0.22709040419623572,
      "grad_norm": 0.459028035402298,
      "learning_rate": 3e-05,
      "loss": 2.8593,
      "step": 736
    },
    {
      "epoch": 0.22739895094106757,
      "grad_norm": 0.5912216901779175,
      "learning_rate": 3e-05,
      "loss": 3.0517,
      "step": 737
    },
    {
      "epoch": 0.2277074976858994,
      "grad_norm": 0.5747165083885193,
      "learning_rate": 3e-05,
      "loss": 3.1073,
      "step": 738
    },
    {
      "epoch": 0.22801604443073126,
      "grad_norm": 0.39477410912513733,
      "learning_rate": 3e-05,
      "loss": 2.531,
      "step": 739
    },
    {
      "epoch": 0.22832459117556309,
      "grad_norm": 0.47743627429008484,
      "learning_rate": 3e-05,
      "loss": 2.7878,
      "step": 740
    },
    {
      "epoch": 0.22863313792039494,
      "grad_norm": 0.6934757232666016,
      "learning_rate": 3e-05,
      "loss": 2.844,
      "step": 741
    },
    {
      "epoch": 0.22894168466522677,
      "grad_norm": 0.39980217814445496,
      "learning_rate": 3e-05,
      "loss": 2.7462,
      "step": 742
    },
    {
      "epoch": 0.22925023141005862,
      "grad_norm": 0.393174409866333,
      "learning_rate": 3e-05,
      "loss": 2.6384,
      "step": 743
    },
    {
      "epoch": 0.22955877815489045,
      "grad_norm": 0.5676349997520447,
      "learning_rate": 3e-05,
      "loss": 3.0358,
      "step": 744
    },
    {
      "epoch": 0.2298673248997223,
      "grad_norm": 0.8145315647125244,
      "learning_rate": 3e-05,
      "loss": 3.4264,
      "step": 745
    },
    {
      "epoch": 0.23017587164455414,
      "grad_norm": 0.3674360513687134,
      "learning_rate": 3e-05,
      "loss": 2.6974,
      "step": 746
    },
    {
      "epoch": 0.230484418389386,
      "grad_norm": 0.39227521419525146,
      "learning_rate": 3e-05,
      "loss": 2.466,
      "step": 747
    },
    {
      "epoch": 0.23079296513421785,
      "grad_norm": 0.6145668625831604,
      "learning_rate": 3e-05,
      "loss": 2.9016,
      "step": 748
    },
    {
      "epoch": 0.23110151187904968,
      "grad_norm": 0.3767462968826294,
      "learning_rate": 3e-05,
      "loss": 2.8179,
      "step": 749
    },
    {
      "epoch": 0.23141005862388153,
      "grad_norm": 0.358978807926178,
      "learning_rate": 3e-05,
      "loss": 2.4825,
      "step": 750
    },
    {
      "epoch": 0.23171860536871336,
      "grad_norm": 0.4695563018321991,
      "learning_rate": 3e-05,
      "loss": 3.0992,
      "step": 751
    },
    {
      "epoch": 0.23202715211354522,
      "grad_norm": 0.4999081790447235,
      "learning_rate": 3e-05,
      "loss": 2.8782,
      "step": 752
    },
    {
      "epoch": 0.23233569885837704,
      "grad_norm": 0.4756012558937073,
      "learning_rate": 3e-05,
      "loss": 3.299,
      "step": 753
    },
    {
      "epoch": 0.2326442456032089,
      "grad_norm": 0.42113542556762695,
      "learning_rate": 3e-05,
      "loss": 2.9299,
      "step": 754
    },
    {
      "epoch": 0.23295279234804073,
      "grad_norm": 0.4862482249736786,
      "learning_rate": 3e-05,
      "loss": 2.6376,
      "step": 755
    },
    {
      "epoch": 0.23326133909287258,
      "grad_norm": 0.44577065110206604,
      "learning_rate": 3e-05,
      "loss": 2.8854,
      "step": 756
    },
    {
      "epoch": 0.2335698858377044,
      "grad_norm": 0.36711904406547546,
      "learning_rate": 3e-05,
      "loss": 2.7238,
      "step": 757
    },
    {
      "epoch": 0.23387843258253627,
      "grad_norm": 0.45583394169807434,
      "learning_rate": 3e-05,
      "loss": 2.9171,
      "step": 758
    },
    {
      "epoch": 0.2341869793273681,
      "grad_norm": 0.4282878041267395,
      "learning_rate": 3e-05,
      "loss": 2.753,
      "step": 759
    },
    {
      "epoch": 0.23449552607219995,
      "grad_norm": 0.3927743136882782,
      "learning_rate": 3e-05,
      "loss": 2.8325,
      "step": 760
    },
    {
      "epoch": 0.23480407281703178,
      "grad_norm": 0.4064248204231262,
      "learning_rate": 3e-05,
      "loss": 2.3949,
      "step": 761
    },
    {
      "epoch": 0.23511261956186363,
      "grad_norm": 0.5085972547531128,
      "learning_rate": 3e-05,
      "loss": 3.3134,
      "step": 762
    },
    {
      "epoch": 0.23542116630669546,
      "grad_norm": 0.43735840916633606,
      "learning_rate": 3e-05,
      "loss": 2.8876,
      "step": 763
    },
    {
      "epoch": 0.23572971305152732,
      "grad_norm": 0.5270010232925415,
      "learning_rate": 3e-05,
      "loss": 3.1755,
      "step": 764
    },
    {
      "epoch": 0.23603825979635915,
      "grad_norm": 0.4824548065662384,
      "learning_rate": 3e-05,
      "loss": 2.8461,
      "step": 765
    },
    {
      "epoch": 0.236346806541191,
      "grad_norm": 0.5891832113265991,
      "learning_rate": 3e-05,
      "loss": 2.864,
      "step": 766
    },
    {
      "epoch": 0.23665535328602283,
      "grad_norm": 0.3647357225418091,
      "learning_rate": 3e-05,
      "loss": 3.0185,
      "step": 767
    },
    {
      "epoch": 0.23696390003085469,
      "grad_norm": 0.3849000930786133,
      "learning_rate": 3e-05,
      "loss": 2.5547,
      "step": 768
    },
    {
      "epoch": 0.2372724467756865,
      "grad_norm": 0.3377460241317749,
      "learning_rate": 3e-05,
      "loss": 2.3713,
      "step": 769
    },
    {
      "epoch": 0.23758099352051837,
      "grad_norm": 0.4434642493724823,
      "learning_rate": 3e-05,
      "loss": 2.5315,
      "step": 770
    },
    {
      "epoch": 0.2378895402653502,
      "grad_norm": 0.3554992079734802,
      "learning_rate": 3e-05,
      "loss": 2.2536,
      "step": 771
    },
    {
      "epoch": 0.23819808701018205,
      "grad_norm": 0.49346596002578735,
      "learning_rate": 3e-05,
      "loss": 2.6857,
      "step": 772
    },
    {
      "epoch": 0.23850663375501388,
      "grad_norm": 0.5243836641311646,
      "learning_rate": 3e-05,
      "loss": 2.9841,
      "step": 773
    },
    {
      "epoch": 0.23881518049984574,
      "grad_norm": 0.47306209802627563,
      "learning_rate": 3e-05,
      "loss": 2.5651,
      "step": 774
    },
    {
      "epoch": 0.23912372724467756,
      "grad_norm": 0.4085429310798645,
      "learning_rate": 3e-05,
      "loss": 2.7709,
      "step": 775
    },
    {
      "epoch": 0.23943227398950942,
      "grad_norm": 0.5058587789535522,
      "learning_rate": 3e-05,
      "loss": 2.8431,
      "step": 776
    },
    {
      "epoch": 0.23974082073434125,
      "grad_norm": 0.423749178647995,
      "learning_rate": 3e-05,
      "loss": 3.1645,
      "step": 777
    },
    {
      "epoch": 0.2400493674791731,
      "grad_norm": 0.44455215334892273,
      "learning_rate": 3e-05,
      "loss": 2.4221,
      "step": 778
    },
    {
      "epoch": 0.24035791422400493,
      "grad_norm": 0.4022437334060669,
      "learning_rate": 3e-05,
      "loss": 2.7517,
      "step": 779
    },
    {
      "epoch": 0.2406664609688368,
      "grad_norm": 0.3913156986236572,
      "learning_rate": 3e-05,
      "loss": 2.8265,
      "step": 780
    },
    {
      "epoch": 0.24097500771366862,
      "grad_norm": 0.5418980717658997,
      "learning_rate": 3e-05,
      "loss": 3.1478,
      "step": 781
    },
    {
      "epoch": 0.24128355445850047,
      "grad_norm": 0.3950507938861847,
      "learning_rate": 3e-05,
      "loss": 2.5474,
      "step": 782
    },
    {
      "epoch": 0.2415921012033323,
      "grad_norm": 0.43344080448150635,
      "learning_rate": 3e-05,
      "loss": 2.5813,
      "step": 783
    },
    {
      "epoch": 0.24190064794816415,
      "grad_norm": 0.32457104325294495,
      "learning_rate": 3e-05,
      "loss": 2.6936,
      "step": 784
    },
    {
      "epoch": 0.24220919469299598,
      "grad_norm": 0.29046568274497986,
      "learning_rate": 3e-05,
      "loss": 2.2027,
      "step": 785
    },
    {
      "epoch": 0.24251774143782784,
      "grad_norm": 0.47084319591522217,
      "learning_rate": 3e-05,
      "loss": 2.7889,
      "step": 786
    },
    {
      "epoch": 0.24282628818265967,
      "grad_norm": 0.4512122571468353,
      "learning_rate": 3e-05,
      "loss": 2.8724,
      "step": 787
    },
    {
      "epoch": 0.24313483492749152,
      "grad_norm": 0.3429430425167084,
      "learning_rate": 3e-05,
      "loss": 2.5882,
      "step": 788
    },
    {
      "epoch": 0.24344338167232335,
      "grad_norm": 0.2793130576610565,
      "learning_rate": 3e-05,
      "loss": 2.3702,
      "step": 789
    },
    {
      "epoch": 0.2437519284171552,
      "grad_norm": 0.6752581000328064,
      "learning_rate": 3e-05,
      "loss": 2.9438,
      "step": 790
    },
    {
      "epoch": 0.24406047516198703,
      "grad_norm": 0.3320951461791992,
      "learning_rate": 3e-05,
      "loss": 2.5836,
      "step": 791
    },
    {
      "epoch": 0.2443690219068189,
      "grad_norm": 0.443316787481308,
      "learning_rate": 3e-05,
      "loss": 2.8967,
      "step": 792
    },
    {
      "epoch": 0.24467756865165072,
      "grad_norm": 0.44384002685546875,
      "learning_rate": 3e-05,
      "loss": 2.1344,
      "step": 793
    },
    {
      "epoch": 0.24498611539648257,
      "grad_norm": 0.5478724241256714,
      "learning_rate": 3e-05,
      "loss": 2.8676,
      "step": 794
    },
    {
      "epoch": 0.2452946621413144,
      "grad_norm": 0.4789648950099945,
      "learning_rate": 3e-05,
      "loss": 2.954,
      "step": 795
    },
    {
      "epoch": 0.24560320888614626,
      "grad_norm": 0.30075979232788086,
      "learning_rate": 3e-05,
      "loss": 2.2147,
      "step": 796
    },
    {
      "epoch": 0.24591175563097808,
      "grad_norm": 0.8388468027114868,
      "learning_rate": 3e-05,
      "loss": 3.0801,
      "step": 797
    },
    {
      "epoch": 0.24622030237580994,
      "grad_norm": 0.46896687150001526,
      "learning_rate": 3e-05,
      "loss": 3.0219,
      "step": 798
    },
    {
      "epoch": 0.24652884912064177,
      "grad_norm": 0.37118545174598694,
      "learning_rate": 3e-05,
      "loss": 2.7809,
      "step": 799
    },
    {
      "epoch": 0.24683739586547362,
      "grad_norm": 0.3290826380252838,
      "learning_rate": 3e-05,
      "loss": 2.4913,
      "step": 800
    },
    {
      "epoch": 0.24714594261030545,
      "grad_norm": 0.4458785951137543,
      "learning_rate": 3e-05,
      "loss": 2.8324,
      "step": 801
    },
    {
      "epoch": 0.2474544893551373,
      "grad_norm": 0.39164820313453674,
      "learning_rate": 3e-05,
      "loss": 2.8643,
      "step": 802
    },
    {
      "epoch": 0.24776303609996914,
      "grad_norm": 0.37570327520370483,
      "learning_rate": 3e-05,
      "loss": 2.6302,
      "step": 803
    },
    {
      "epoch": 0.248071582844801,
      "grad_norm": 0.36291664838790894,
      "learning_rate": 3e-05,
      "loss": 2.3188,
      "step": 804
    },
    {
      "epoch": 0.24838012958963282,
      "grad_norm": 0.7497803568840027,
      "learning_rate": 3e-05,
      "loss": 3.1793,
      "step": 805
    },
    {
      "epoch": 0.24868867633446468,
      "grad_norm": 0.3503245413303375,
      "learning_rate": 3e-05,
      "loss": 2.6478,
      "step": 806
    },
    {
      "epoch": 0.2489972230792965,
      "grad_norm": 0.4229944944381714,
      "learning_rate": 3e-05,
      "loss": 3.166,
      "step": 807
    },
    {
      "epoch": 0.24930576982412836,
      "grad_norm": 0.39964744448661804,
      "learning_rate": 3e-05,
      "loss": 2.5633,
      "step": 808
    },
    {
      "epoch": 0.2496143165689602,
      "grad_norm": 0.4407540559768677,
      "learning_rate": 3e-05,
      "loss": 2.8108,
      "step": 809
    },
    {
      "epoch": 0.24992286331379204,
      "grad_norm": 0.5962763428688049,
      "learning_rate": 3e-05,
      "loss": 2.7762,
      "step": 810
    },
    {
      "epoch": 0.25023141005862387,
      "grad_norm": 0.35750457644462585,
      "learning_rate": 3e-05,
      "loss": 2.5792,
      "step": 811
    },
    {
      "epoch": 0.2505399568034557,
      "grad_norm": 0.6067323088645935,
      "learning_rate": 3e-05,
      "loss": 2.9188,
      "step": 812
    },
    {
      "epoch": 0.2508485035482876,
      "grad_norm": 0.5473812818527222,
      "learning_rate": 3e-05,
      "loss": 2.9941,
      "step": 813
    },
    {
      "epoch": 0.2511570502931194,
      "grad_norm": 0.4247765839099884,
      "learning_rate": 3e-05,
      "loss": 2.9066,
      "step": 814
    },
    {
      "epoch": 0.25146559703795124,
      "grad_norm": 0.36779630184173584,
      "learning_rate": 3e-05,
      "loss": 2.1791,
      "step": 815
    },
    {
      "epoch": 0.25177414378278307,
      "grad_norm": 0.4601776599884033,
      "learning_rate": 3e-05,
      "loss": 2.4513,
      "step": 816
    },
    {
      "epoch": 0.25208269052761495,
      "grad_norm": 0.4455990493297577,
      "learning_rate": 3e-05,
      "loss": 3.0133,
      "step": 817
    },
    {
      "epoch": 0.2523912372724468,
      "grad_norm": 0.469711035490036,
      "learning_rate": 3e-05,
      "loss": 2.7601,
      "step": 818
    },
    {
      "epoch": 0.2526997840172786,
      "grad_norm": 0.35248449444770813,
      "learning_rate": 3e-05,
      "loss": 2.4141,
      "step": 819
    },
    {
      "epoch": 0.25300833076211043,
      "grad_norm": 0.5230732560157776,
      "learning_rate": 3e-05,
      "loss": 2.7922,
      "step": 820
    },
    {
      "epoch": 0.2533168775069423,
      "grad_norm": 0.4009235203266144,
      "learning_rate": 3e-05,
      "loss": 2.8466,
      "step": 821
    },
    {
      "epoch": 0.25362542425177415,
      "grad_norm": 0.4277903139591217,
      "learning_rate": 3e-05,
      "loss": 2.6193,
      "step": 822
    },
    {
      "epoch": 0.253933970996606,
      "grad_norm": 0.35308367013931274,
      "learning_rate": 3e-05,
      "loss": 2.7359,
      "step": 823
    },
    {
      "epoch": 0.2542425177414378,
      "grad_norm": 0.5506107807159424,
      "learning_rate": 3e-05,
      "loss": 3.3045,
      "step": 824
    },
    {
      "epoch": 0.2545510644862697,
      "grad_norm": 0.4381810128688812,
      "learning_rate": 3e-05,
      "loss": 2.5703,
      "step": 825
    },
    {
      "epoch": 0.2548596112311015,
      "grad_norm": 0.6535665392875671,
      "learning_rate": 3e-05,
      "loss": 2.9752,
      "step": 826
    },
    {
      "epoch": 0.25516815797593334,
      "grad_norm": 0.31507301330566406,
      "learning_rate": 3e-05,
      "loss": 2.4288,
      "step": 827
    },
    {
      "epoch": 0.25547670472076517,
      "grad_norm": 0.628995418548584,
      "learning_rate": 3e-05,
      "loss": 2.9277,
      "step": 828
    },
    {
      "epoch": 0.25578525146559705,
      "grad_norm": 0.3475759029388428,
      "learning_rate": 3e-05,
      "loss": 2.6116,
      "step": 829
    },
    {
      "epoch": 0.2560937982104289,
      "grad_norm": 0.35610201954841614,
      "learning_rate": 3e-05,
      "loss": 2.5474,
      "step": 830
    },
    {
      "epoch": 0.2564023449552607,
      "grad_norm": 0.3411518335342407,
      "learning_rate": 3e-05,
      "loss": 2.6045,
      "step": 831
    },
    {
      "epoch": 0.2567108917000926,
      "grad_norm": 0.36465325951576233,
      "learning_rate": 3e-05,
      "loss": 2.3901,
      "step": 832
    },
    {
      "epoch": 0.2570194384449244,
      "grad_norm": 0.3037632703781128,
      "learning_rate": 3e-05,
      "loss": 2.3478,
      "step": 833
    },
    {
      "epoch": 0.25732798518975625,
      "grad_norm": 0.44962364435195923,
      "learning_rate": 3e-05,
      "loss": 2.8154,
      "step": 834
    },
    {
      "epoch": 0.2576365319345881,
      "grad_norm": 0.3245837688446045,
      "learning_rate": 3e-05,
      "loss": 2.4922,
      "step": 835
    },
    {
      "epoch": 0.25794507867941996,
      "grad_norm": 0.37248003482818604,
      "learning_rate": 3e-05,
      "loss": 2.6044,
      "step": 836
    },
    {
      "epoch": 0.2582536254242518,
      "grad_norm": 0.3687119483947754,
      "learning_rate": 3e-05,
      "loss": 2.7271,
      "step": 837
    },
    {
      "epoch": 0.2585621721690836,
      "grad_norm": 0.3978298604488373,
      "learning_rate": 3e-05,
      "loss": 2.9363,
      "step": 838
    },
    {
      "epoch": 0.25887071891391544,
      "grad_norm": 0.43831732869148254,
      "learning_rate": 3e-05,
      "loss": 2.633,
      "step": 839
    },
    {
      "epoch": 0.2591792656587473,
      "grad_norm": 0.39611366391181946,
      "learning_rate": 3e-05,
      "loss": 2.9044,
      "step": 840
    },
    {
      "epoch": 0.25948781240357915,
      "grad_norm": 0.37481510639190674,
      "learning_rate": 3e-05,
      "loss": 2.7067,
      "step": 841
    },
    {
      "epoch": 0.259796359148411,
      "grad_norm": 0.33353596925735474,
      "learning_rate": 3e-05,
      "loss": 2.4132,
      "step": 842
    },
    {
      "epoch": 0.2601049058932428,
      "grad_norm": 0.42422613501548767,
      "learning_rate": 3e-05,
      "loss": 2.4306,
      "step": 843
    },
    {
      "epoch": 0.2604134526380747,
      "grad_norm": 0.7475231289863586,
      "learning_rate": 3e-05,
      "loss": 3.2851,
      "step": 844
    },
    {
      "epoch": 0.2607219993829065,
      "grad_norm": 0.8938561081886292,
      "learning_rate": 3e-05,
      "loss": 3.2137,
      "step": 845
    },
    {
      "epoch": 0.26103054612773835,
      "grad_norm": 0.5334686636924744,
      "learning_rate": 3e-05,
      "loss": 2.3994,
      "step": 846
    },
    {
      "epoch": 0.2613390928725702,
      "grad_norm": 0.6619825959205627,
      "learning_rate": 3e-05,
      "loss": 2.5175,
      "step": 847
    },
    {
      "epoch": 0.26164763961740206,
      "grad_norm": 0.8206683993339539,
      "learning_rate": 3e-05,
      "loss": 2.6581,
      "step": 848
    },
    {
      "epoch": 0.2619561863622339,
      "grad_norm": 0.4362182319164276,
      "learning_rate": 3e-05,
      "loss": 3.1203,
      "step": 849
    },
    {
      "epoch": 0.2622647331070657,
      "grad_norm": 0.5157708525657654,
      "learning_rate": 3e-05,
      "loss": 2.8433,
      "step": 850
    },
    {
      "epoch": 0.26257327985189755,
      "grad_norm": 0.5386294722557068,
      "learning_rate": 3e-05,
      "loss": 2.5983,
      "step": 851
    },
    {
      "epoch": 0.26288182659672943,
      "grad_norm": 0.7352175712585449,
      "learning_rate": 3e-05,
      "loss": 2.8269,
      "step": 852
    },
    {
      "epoch": 0.26319037334156126,
      "grad_norm": 0.4389644265174866,
      "learning_rate": 3e-05,
      "loss": 2.7614,
      "step": 853
    },
    {
      "epoch": 0.2634989200863931,
      "grad_norm": 0.5747153162956238,
      "learning_rate": 3e-05,
      "loss": 2.8439,
      "step": 854
    },
    {
      "epoch": 0.2638074668312249,
      "grad_norm": 0.442501962184906,
      "learning_rate": 3e-05,
      "loss": 2.5673,
      "step": 855
    },
    {
      "epoch": 0.2641160135760568,
      "grad_norm": 0.5041112899780273,
      "learning_rate": 3e-05,
      "loss": 3.4168,
      "step": 856
    },
    {
      "epoch": 0.2644245603208886,
      "grad_norm": 0.4333510994911194,
      "learning_rate": 3e-05,
      "loss": 2.5954,
      "step": 857
    },
    {
      "epoch": 0.26473310706572045,
      "grad_norm": 0.2841757535934448,
      "learning_rate": 3e-05,
      "loss": 2.2267,
      "step": 858
    },
    {
      "epoch": 0.2650416538105523,
      "grad_norm": 0.3240799903869629,
      "learning_rate": 3e-05,
      "loss": 2.4861,
      "step": 859
    },
    {
      "epoch": 0.26535020055538416,
      "grad_norm": 0.8906437754631042,
      "learning_rate": 3e-05,
      "loss": 3.1556,
      "step": 860
    },
    {
      "epoch": 0.265658747300216,
      "grad_norm": 0.503738284111023,
      "learning_rate": 3e-05,
      "loss": 2.8386,
      "step": 861
    },
    {
      "epoch": 0.2659672940450478,
      "grad_norm": 0.3821960985660553,
      "learning_rate": 3e-05,
      "loss": 2.4592,
      "step": 862
    },
    {
      "epoch": 0.26627584078987965,
      "grad_norm": 0.4790137708187103,
      "learning_rate": 3e-05,
      "loss": 2.9267,
      "step": 863
    },
    {
      "epoch": 0.26658438753471153,
      "grad_norm": 0.4529775083065033,
      "learning_rate": 3e-05,
      "loss": 2.7667,
      "step": 864
    },
    {
      "epoch": 0.26689293427954336,
      "grad_norm": 0.38088151812553406,
      "learning_rate": 3e-05,
      "loss": 2.8161,
      "step": 865
    },
    {
      "epoch": 0.2672014810243752,
      "grad_norm": 0.38208597898483276,
      "learning_rate": 3e-05,
      "loss": 2.6606,
      "step": 866
    },
    {
      "epoch": 0.267510027769207,
      "grad_norm": 0.3878786861896515,
      "learning_rate": 3e-05,
      "loss": 2.8936,
      "step": 867
    },
    {
      "epoch": 0.2678185745140389,
      "grad_norm": 0.3223757743835449,
      "learning_rate": 3e-05,
      "loss": 2.6747,
      "step": 868
    },
    {
      "epoch": 0.2681271212588707,
      "grad_norm": 0.380728155374527,
      "learning_rate": 3e-05,
      "loss": 2.9688,
      "step": 869
    },
    {
      "epoch": 0.26843566800370255,
      "grad_norm": 0.3616756796836853,
      "learning_rate": 3e-05,
      "loss": 2.9107,
      "step": 870
    },
    {
      "epoch": 0.2687442147485344,
      "grad_norm": 0.4382908344268799,
      "learning_rate": 3e-05,
      "loss": 2.7253,
      "step": 871
    },
    {
      "epoch": 0.26905276149336627,
      "grad_norm": 0.6541967988014221,
      "learning_rate": 3e-05,
      "loss": 2.9171,
      "step": 872
    },
    {
      "epoch": 0.2693613082381981,
      "grad_norm": 0.5041164755821228,
      "learning_rate": 3e-05,
      "loss": 2.697,
      "step": 873
    },
    {
      "epoch": 0.2696698549830299,
      "grad_norm": 0.40384742617607117,
      "learning_rate": 3e-05,
      "loss": 2.832,
      "step": 874
    },
    {
      "epoch": 0.26997840172786175,
      "grad_norm": 0.4746786952018738,
      "learning_rate": 3e-05,
      "loss": 3.0941,
      "step": 875
    },
    {
      "epoch": 0.27028694847269363,
      "grad_norm": 0.49661922454833984,
      "learning_rate": 3e-05,
      "loss": 2.9883,
      "step": 876
    },
    {
      "epoch": 0.27059549521752546,
      "grad_norm": 0.323425829410553,
      "learning_rate": 3e-05,
      "loss": 2.4742,
      "step": 877
    },
    {
      "epoch": 0.2709040419623573,
      "grad_norm": 0.383983314037323,
      "learning_rate": 3e-05,
      "loss": 2.5993,
      "step": 878
    },
    {
      "epoch": 0.2712125887071891,
      "grad_norm": 0.44553110003471375,
      "learning_rate": 3e-05,
      "loss": 2.6372,
      "step": 879
    },
    {
      "epoch": 0.271521135452021,
      "grad_norm": 0.5241272449493408,
      "learning_rate": 3e-05,
      "loss": 3.2384,
      "step": 880
    },
    {
      "epoch": 0.27182968219685283,
      "grad_norm": 0.6117975115776062,
      "learning_rate": 3e-05,
      "loss": 2.8964,
      "step": 881
    },
    {
      "epoch": 0.27213822894168466,
      "grad_norm": 0.3425690233707428,
      "learning_rate": 3e-05,
      "loss": 2.3628,
      "step": 882
    },
    {
      "epoch": 0.2724467756865165,
      "grad_norm": 0.49667888879776,
      "learning_rate": 3e-05,
      "loss": 3.1708,
      "step": 883
    },
    {
      "epoch": 0.27275532243134837,
      "grad_norm": 0.3282347321510315,
      "learning_rate": 3e-05,
      "loss": 2.5675,
      "step": 884
    },
    {
      "epoch": 0.2730638691761802,
      "grad_norm": 0.371405690908432,
      "learning_rate": 3e-05,
      "loss": 2.3648,
      "step": 885
    },
    {
      "epoch": 0.273372415921012,
      "grad_norm": 0.46397897601127625,
      "learning_rate": 3e-05,
      "loss": 2.4369,
      "step": 886
    },
    {
      "epoch": 0.27368096266584385,
      "grad_norm": 0.45648089051246643,
      "learning_rate": 3e-05,
      "loss": 2.5306,
      "step": 887
    },
    {
      "epoch": 0.27398950941067574,
      "grad_norm": 0.3604811728000641,
      "learning_rate": 3e-05,
      "loss": 2.6633,
      "step": 888
    },
    {
      "epoch": 0.27429805615550756,
      "grad_norm": 0.3597293496131897,
      "learning_rate": 3e-05,
      "loss": 2.5673,
      "step": 889
    },
    {
      "epoch": 0.2746066029003394,
      "grad_norm": 0.4115181565284729,
      "learning_rate": 3e-05,
      "loss": 2.6737,
      "step": 890
    },
    {
      "epoch": 0.2749151496451712,
      "grad_norm": 0.5663295388221741,
      "learning_rate": 3e-05,
      "loss": 3.3189,
      "step": 891
    },
    {
      "epoch": 0.2752236963900031,
      "grad_norm": 0.3877559006214142,
      "learning_rate": 3e-05,
      "loss": 2.3662,
      "step": 892
    },
    {
      "epoch": 0.27553224313483493,
      "grad_norm": 0.3957102596759796,
      "learning_rate": 3e-05,
      "loss": 2.6481,
      "step": 893
    },
    {
      "epoch": 0.27584078987966676,
      "grad_norm": 0.4517107307910919,
      "learning_rate": 3e-05,
      "loss": 2.6274,
      "step": 894
    },
    {
      "epoch": 0.2761493366244986,
      "grad_norm": 0.3920423090457916,
      "learning_rate": 3e-05,
      "loss": 2.3127,
      "step": 895
    },
    {
      "epoch": 0.27645788336933047,
      "grad_norm": 0.4129326641559601,
      "learning_rate": 3e-05,
      "loss": 2.8305,
      "step": 896
    },
    {
      "epoch": 0.2767664301141623,
      "grad_norm": 0.434194415807724,
      "learning_rate": 3e-05,
      "loss": 2.9748,
      "step": 897
    },
    {
      "epoch": 0.2770749768589941,
      "grad_norm": 0.35281941294670105,
      "learning_rate": 3e-05,
      "loss": 2.6046,
      "step": 898
    },
    {
      "epoch": 0.27738352360382595,
      "grad_norm": 0.3993188142776489,
      "learning_rate": 3e-05,
      "loss": 3.0656,
      "step": 899
    },
    {
      "epoch": 0.27769207034865784,
      "grad_norm": 0.38514411449432373,
      "learning_rate": 3e-05,
      "loss": 2.6093,
      "step": 900
    },
    {
      "epoch": 0.27800061709348967,
      "grad_norm": 0.46261972188949585,
      "learning_rate": 3e-05,
      "loss": 2.7449,
      "step": 901
    },
    {
      "epoch": 0.2783091638383215,
      "grad_norm": 0.35552743077278137,
      "learning_rate": 3e-05,
      "loss": 2.4989,
      "step": 902
    },
    {
      "epoch": 0.2786177105831533,
      "grad_norm": 0.3849910795688629,
      "learning_rate": 3e-05,
      "loss": 2.8201,
      "step": 903
    },
    {
      "epoch": 0.2789262573279852,
      "grad_norm": 0.3854105770587921,
      "learning_rate": 3e-05,
      "loss": 2.866,
      "step": 904
    },
    {
      "epoch": 0.27923480407281703,
      "grad_norm": 0.41260039806365967,
      "learning_rate": 3e-05,
      "loss": 2.6602,
      "step": 905
    },
    {
      "epoch": 0.27954335081764886,
      "grad_norm": 0.347160667181015,
      "learning_rate": 3e-05,
      "loss": 2.3522,
      "step": 906
    },
    {
      "epoch": 0.2798518975624807,
      "grad_norm": 0.2681293785572052,
      "learning_rate": 3e-05,
      "loss": 2.2104,
      "step": 907
    },
    {
      "epoch": 0.2801604443073126,
      "grad_norm": 0.5947923064231873,
      "learning_rate": 3e-05,
      "loss": 2.9183,
      "step": 908
    },
    {
      "epoch": 0.2804689910521444,
      "grad_norm": 0.3329774737358093,
      "learning_rate": 3e-05,
      "loss": 2.6277,
      "step": 909
    },
    {
      "epoch": 0.28077753779697623,
      "grad_norm": 0.5521478652954102,
      "learning_rate": 3e-05,
      "loss": 3.0918,
      "step": 910
    },
    {
      "epoch": 0.28108608454180806,
      "grad_norm": 0.4059542417526245,
      "learning_rate": 3e-05,
      "loss": 2.7583,
      "step": 911
    },
    {
      "epoch": 0.28139463128663994,
      "grad_norm": 0.406534343957901,
      "learning_rate": 3e-05,
      "loss": 2.8539,
      "step": 912
    },
    {
      "epoch": 0.28170317803147177,
      "grad_norm": 0.6377788782119751,
      "learning_rate": 3e-05,
      "loss": 2.7076,
      "step": 913
    },
    {
      "epoch": 0.2820117247763036,
      "grad_norm": 0.3285945653915405,
      "learning_rate": 3e-05,
      "loss": 2.233,
      "step": 914
    },
    {
      "epoch": 0.2823202715211355,
      "grad_norm": 0.3589547276496887,
      "learning_rate": 3e-05,
      "loss": 2.6757,
      "step": 915
    },
    {
      "epoch": 0.2826288182659673,
      "grad_norm": 0.5185350775718689,
      "learning_rate": 3e-05,
      "loss": 2.84,
      "step": 916
    },
    {
      "epoch": 0.28293736501079914,
      "grad_norm": 0.5448761582374573,
      "learning_rate": 3e-05,
      "loss": 2.6148,
      "step": 917
    },
    {
      "epoch": 0.28324591175563096,
      "grad_norm": 0.40239185094833374,
      "learning_rate": 3e-05,
      "loss": 2.6744,
      "step": 918
    },
    {
      "epoch": 0.28355445850046285,
      "grad_norm": 0.4547919034957886,
      "learning_rate": 3e-05,
      "loss": 3.0762,
      "step": 919
    },
    {
      "epoch": 0.2838630052452947,
      "grad_norm": 0.6027969717979431,
      "learning_rate": 3e-05,
      "loss": 2.5929,
      "step": 920
    },
    {
      "epoch": 0.2841715519901265,
      "grad_norm": 0.4554506242275238,
      "learning_rate": 3e-05,
      "loss": 2.4028,
      "step": 921
    },
    {
      "epoch": 0.28448009873495833,
      "grad_norm": 0.4088301360607147,
      "learning_rate": 3e-05,
      "loss": 2.2121,
      "step": 922
    },
    {
      "epoch": 0.2847886454797902,
      "grad_norm": 0.38209256529808044,
      "learning_rate": 3e-05,
      "loss": 2.8373,
      "step": 923
    },
    {
      "epoch": 0.28509719222462204,
      "grad_norm": 0.3328472673892975,
      "learning_rate": 3e-05,
      "loss": 2.4444,
      "step": 924
    },
    {
      "epoch": 0.28540573896945387,
      "grad_norm": 0.33774638175964355,
      "learning_rate": 3e-05,
      "loss": 2.3775,
      "step": 925
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 0.5520322322845459,
      "learning_rate": 3e-05,
      "loss": 2.6919,
      "step": 926
    },
    {
      "epoch": 0.2860228324591176,
      "grad_norm": 0.42964938282966614,
      "learning_rate": 3e-05,
      "loss": 2.7063,
      "step": 927
    },
    {
      "epoch": 0.2863313792039494,
      "grad_norm": 0.3997243642807007,
      "learning_rate": 3e-05,
      "loss": 2.4581,
      "step": 928
    },
    {
      "epoch": 0.28663992594878124,
      "grad_norm": 0.34988248348236084,
      "learning_rate": 3e-05,
      "loss": 2.6436,
      "step": 929
    },
    {
      "epoch": 0.28694847269361307,
      "grad_norm": 0.5160298347473145,
      "learning_rate": 3e-05,
      "loss": 2.5583,
      "step": 930
    },
    {
      "epoch": 0.28725701943844495,
      "grad_norm": 0.5120193362236023,
      "learning_rate": 3e-05,
      "loss": 2.879,
      "step": 931
    },
    {
      "epoch": 0.2875655661832768,
      "grad_norm": 0.380231112241745,
      "learning_rate": 3e-05,
      "loss": 2.8559,
      "step": 932
    },
    {
      "epoch": 0.2878741129281086,
      "grad_norm": 0.3843991756439209,
      "learning_rate": 3e-05,
      "loss": 2.547,
      "step": 933
    },
    {
      "epoch": 0.28818265967294043,
      "grad_norm": 0.48810240626335144,
      "learning_rate": 3e-05,
      "loss": 2.7941,
      "step": 934
    },
    {
      "epoch": 0.2884912064177723,
      "grad_norm": 0.4541907012462616,
      "learning_rate": 3e-05,
      "loss": 2.6922,
      "step": 935
    },
    {
      "epoch": 0.28879975316260414,
      "grad_norm": 0.45531538128852844,
      "learning_rate": 3e-05,
      "loss": 2.9196,
      "step": 936
    },
    {
      "epoch": 0.289108299907436,
      "grad_norm": 0.33889535069465637,
      "learning_rate": 3e-05,
      "loss": 2.6043,
      "step": 937
    },
    {
      "epoch": 0.2894168466522678,
      "grad_norm": 0.41679373383522034,
      "learning_rate": 3e-05,
      "loss": 2.3954,
      "step": 938
    },
    {
      "epoch": 0.2897253933970997,
      "grad_norm": 0.5239324569702148,
      "learning_rate": 3e-05,
      "loss": 2.7328,
      "step": 939
    },
    {
      "epoch": 0.2900339401419315,
      "grad_norm": 0.45569634437561035,
      "learning_rate": 3e-05,
      "loss": 2.8944,
      "step": 940
    },
    {
      "epoch": 0.29034248688676334,
      "grad_norm": 0.29057788848876953,
      "learning_rate": 3e-05,
      "loss": 2.1483,
      "step": 941
    },
    {
      "epoch": 0.29065103363159517,
      "grad_norm": 0.3790557384490967,
      "learning_rate": 3e-05,
      "loss": 2.4723,
      "step": 942
    },
    {
      "epoch": 0.29095958037642705,
      "grad_norm": 0.4859466850757599,
      "learning_rate": 3e-05,
      "loss": 2.8131,
      "step": 943
    },
    {
      "epoch": 0.2912681271212589,
      "grad_norm": 0.5878845453262329,
      "learning_rate": 3e-05,
      "loss": 2.8616,
      "step": 944
    },
    {
      "epoch": 0.2915766738660907,
      "grad_norm": 0.3621906638145447,
      "learning_rate": 3e-05,
      "loss": 2.6452,
      "step": 945
    },
    {
      "epoch": 0.29188522061092254,
      "grad_norm": 0.39704960584640503,
      "learning_rate": 3e-05,
      "loss": 2.688,
      "step": 946
    },
    {
      "epoch": 0.2921937673557544,
      "grad_norm": 0.438273549079895,
      "learning_rate": 3e-05,
      "loss": 2.6861,
      "step": 947
    },
    {
      "epoch": 0.29250231410058625,
      "grad_norm": 0.42470529675483704,
      "learning_rate": 3e-05,
      "loss": 2.7117,
      "step": 948
    },
    {
      "epoch": 0.2928108608454181,
      "grad_norm": 0.3004021942615509,
      "learning_rate": 3e-05,
      "loss": 2.2417,
      "step": 949
    },
    {
      "epoch": 0.2931194075902499,
      "grad_norm": 0.3381553590297699,
      "learning_rate": 3e-05,
      "loss": 2.7548,
      "step": 950
    },
    {
      "epoch": 0.2934279543350818,
      "grad_norm": 0.32822176814079285,
      "learning_rate": 3e-05,
      "loss": 2.6632,
      "step": 951
    },
    {
      "epoch": 0.2937365010799136,
      "grad_norm": 0.48720303177833557,
      "learning_rate": 3e-05,
      "loss": 2.9131,
      "step": 952
    },
    {
      "epoch": 0.29404504782474544,
      "grad_norm": 0.5922092795372009,
      "learning_rate": 3e-05,
      "loss": 2.6821,
      "step": 953
    },
    {
      "epoch": 0.29435359456957727,
      "grad_norm": 0.36750638484954834,
      "learning_rate": 3e-05,
      "loss": 2.9744,
      "step": 954
    },
    {
      "epoch": 0.29466214131440915,
      "grad_norm": 0.3950546085834503,
      "learning_rate": 3e-05,
      "loss": 2.8421,
      "step": 955
    },
    {
      "epoch": 0.294970688059241,
      "grad_norm": 0.33338621258735657,
      "learning_rate": 3e-05,
      "loss": 2.3283,
      "step": 956
    },
    {
      "epoch": 0.2952792348040728,
      "grad_norm": 0.32808199524879456,
      "learning_rate": 3e-05,
      "loss": 2.575,
      "step": 957
    },
    {
      "epoch": 0.29558778154890464,
      "grad_norm": 0.40493538975715637,
      "learning_rate": 3e-05,
      "loss": 2.696,
      "step": 958
    },
    {
      "epoch": 0.2958963282937365,
      "grad_norm": 0.5946307182312012,
      "learning_rate": 3e-05,
      "loss": 3.0542,
      "step": 959
    },
    {
      "epoch": 0.29620487503856835,
      "grad_norm": 0.49499088525772095,
      "learning_rate": 3e-05,
      "loss": 2.9535,
      "step": 960
    },
    {
      "epoch": 0.2965134217834002,
      "grad_norm": 0.3947402238845825,
      "learning_rate": 3e-05,
      "loss": 2.2794,
      "step": 961
    },
    {
      "epoch": 0.296821968528232,
      "grad_norm": 0.6068766713142395,
      "learning_rate": 3e-05,
      "loss": 2.7137,
      "step": 962
    },
    {
      "epoch": 0.2971305152730639,
      "grad_norm": 0.3848678469657898,
      "learning_rate": 3e-05,
      "loss": 2.8856,
      "step": 963
    },
    {
      "epoch": 0.2974390620178957,
      "grad_norm": 0.3639110028743744,
      "learning_rate": 3e-05,
      "loss": 2.6542,
      "step": 964
    },
    {
      "epoch": 0.29774760876272754,
      "grad_norm": 0.4631502330303192,
      "learning_rate": 3e-05,
      "loss": 2.801,
      "step": 965
    },
    {
      "epoch": 0.2980561555075594,
      "grad_norm": 0.4895254671573639,
      "learning_rate": 3e-05,
      "loss": 2.5745,
      "step": 966
    },
    {
      "epoch": 0.29836470225239126,
      "grad_norm": 0.32088950276374817,
      "learning_rate": 3e-05,
      "loss": 2.5099,
      "step": 967
    },
    {
      "epoch": 0.2986732489972231,
      "grad_norm": 0.4241865277290344,
      "learning_rate": 3e-05,
      "loss": 2.8371,
      "step": 968
    },
    {
      "epoch": 0.2989817957420549,
      "grad_norm": 0.4249207675457001,
      "learning_rate": 3e-05,
      "loss": 2.786,
      "step": 969
    },
    {
      "epoch": 0.29929034248688674,
      "grad_norm": 0.3566865622997284,
      "learning_rate": 3e-05,
      "loss": 2.4608,
      "step": 970
    },
    {
      "epoch": 0.2995988892317186,
      "grad_norm": 0.4732634127140045,
      "learning_rate": 3e-05,
      "loss": 2.4006,
      "step": 971
    },
    {
      "epoch": 0.29990743597655045,
      "grad_norm": 0.31523409485816956,
      "learning_rate": 3e-05,
      "loss": 2.5125,
      "step": 972
    },
    {
      "epoch": 0.3002159827213823,
      "grad_norm": 0.31227585673332214,
      "learning_rate": 3e-05,
      "loss": 2.5859,
      "step": 973
    },
    {
      "epoch": 0.3005245294662141,
      "grad_norm": 0.5433022379875183,
      "learning_rate": 3e-05,
      "loss": 2.7815,
      "step": 974
    },
    {
      "epoch": 0.300833076211046,
      "grad_norm": 0.32463932037353516,
      "learning_rate": 3e-05,
      "loss": 2.5574,
      "step": 975
    },
    {
      "epoch": 0.3011416229558778,
      "grad_norm": 0.45843788981437683,
      "learning_rate": 3e-05,
      "loss": 3.3021,
      "step": 976
    },
    {
      "epoch": 0.30145016970070965,
      "grad_norm": 0.43129628896713257,
      "learning_rate": 3e-05,
      "loss": 2.7665,
      "step": 977
    },
    {
      "epoch": 0.3017587164455415,
      "grad_norm": 0.4639509618282318,
      "learning_rate": 3e-05,
      "loss": 2.9852,
      "step": 978
    },
    {
      "epoch": 0.30206726319037336,
      "grad_norm": 0.6034327745437622,
      "learning_rate": 3e-05,
      "loss": 2.7809,
      "step": 979
    },
    {
      "epoch": 0.3023758099352052,
      "grad_norm": 0.4655909538269043,
      "learning_rate": 3e-05,
      "loss": 2.788,
      "step": 980
    },
    {
      "epoch": 0.302684356680037,
      "grad_norm": 0.43474000692367554,
      "learning_rate": 3e-05,
      "loss": 2.7155,
      "step": 981
    },
    {
      "epoch": 0.30299290342486884,
      "grad_norm": 0.4621677100658417,
      "learning_rate": 3e-05,
      "loss": 2.649,
      "step": 982
    },
    {
      "epoch": 0.3033014501697007,
      "grad_norm": 0.39728179574012756,
      "learning_rate": 3e-05,
      "loss": 2.7499,
      "step": 983
    },
    {
      "epoch": 0.30360999691453255,
      "grad_norm": 0.41515350341796875,
      "learning_rate": 3e-05,
      "loss": 2.4945,
      "step": 984
    },
    {
      "epoch": 0.3039185436593644,
      "grad_norm": 0.4541874825954437,
      "learning_rate": 3e-05,
      "loss": 2.9019,
      "step": 985
    },
    {
      "epoch": 0.3042270904041962,
      "grad_norm": 0.4044342339038849,
      "learning_rate": 3e-05,
      "loss": 2.5939,
      "step": 986
    },
    {
      "epoch": 0.3045356371490281,
      "grad_norm": 0.5499434471130371,
      "learning_rate": 3e-05,
      "loss": 3.1831,
      "step": 987
    },
    {
      "epoch": 0.3048441838938599,
      "grad_norm": 0.4253259599208832,
      "learning_rate": 3e-05,
      "loss": 2.6482,
      "step": 988
    },
    {
      "epoch": 0.30515273063869175,
      "grad_norm": 0.5634761452674866,
      "learning_rate": 3e-05,
      "loss": 3.0694,
      "step": 989
    },
    {
      "epoch": 0.3054612773835236,
      "grad_norm": 0.3745432496070862,
      "learning_rate": 3e-05,
      "loss": 2.864,
      "step": 990
    },
    {
      "epoch": 0.30576982412835546,
      "grad_norm": 0.31692636013031006,
      "learning_rate": 3e-05,
      "loss": 2.4637,
      "step": 991
    },
    {
      "epoch": 0.3060783708731873,
      "grad_norm": 0.45177918672561646,
      "learning_rate": 3e-05,
      "loss": 2.7045,
      "step": 992
    },
    {
      "epoch": 0.3063869176180191,
      "grad_norm": 0.3548758924007416,
      "learning_rate": 3e-05,
      "loss": 2.4143,
      "step": 993
    },
    {
      "epoch": 0.30669546436285094,
      "grad_norm": 0.4161062240600586,
      "learning_rate": 3e-05,
      "loss": 3.0351,
      "step": 994
    },
    {
      "epoch": 0.30700401110768283,
      "grad_norm": 0.5312502384185791,
      "learning_rate": 3e-05,
      "loss": 2.8126,
      "step": 995
    },
    {
      "epoch": 0.30731255785251466,
      "grad_norm": 0.4165649712085724,
      "learning_rate": 3e-05,
      "loss": 2.311,
      "step": 996
    },
    {
      "epoch": 0.3076211045973465,
      "grad_norm": 0.4763493537902832,
      "learning_rate": 3e-05,
      "loss": 2.5855,
      "step": 997
    },
    {
      "epoch": 0.30792965134217837,
      "grad_norm": 0.3590225279331207,
      "learning_rate": 3e-05,
      "loss": 2.4563,
      "step": 998
    },
    {
      "epoch": 0.3082381980870102,
      "grad_norm": 0.43557247519493103,
      "learning_rate": 3e-05,
      "loss": 2.7756,
      "step": 999
    },
    {
      "epoch": 0.308546744831842,
      "grad_norm": 0.5323253870010376,
      "learning_rate": 3e-05,
      "loss": 2.9261,
      "step": 1000
    },
    {
      "epoch": 0.30885529157667385,
      "grad_norm": 0.3511796295642853,
      "learning_rate": 3e-05,
      "loss": 2.7418,
      "step": 1001
    },
    {
      "epoch": 0.30916383832150574,
      "grad_norm": 0.37199896574020386,
      "learning_rate": 3e-05,
      "loss": 2.6009,
      "step": 1002
    },
    {
      "epoch": 0.30947238506633756,
      "grad_norm": 0.34085142612457275,
      "learning_rate": 3e-05,
      "loss": 2.6959,
      "step": 1003
    },
    {
      "epoch": 0.3097809318111694,
      "grad_norm": 0.4471310079097748,
      "learning_rate": 3e-05,
      "loss": 2.8914,
      "step": 1004
    },
    {
      "epoch": 0.3100894785560012,
      "grad_norm": 0.3000938296318054,
      "learning_rate": 3e-05,
      "loss": 2.3222,
      "step": 1005
    },
    {
      "epoch": 0.3103980253008331,
      "grad_norm": 0.4500029385089874,
      "learning_rate": 3e-05,
      "loss": 2.6863,
      "step": 1006
    },
    {
      "epoch": 0.31070657204566493,
      "grad_norm": 0.36391642689704895,
      "learning_rate": 3e-05,
      "loss": 2.4052,
      "step": 1007
    },
    {
      "epoch": 0.31101511879049676,
      "grad_norm": 0.33557793498039246,
      "learning_rate": 3e-05,
      "loss": 2.4544,
      "step": 1008
    },
    {
      "epoch": 0.3113236655353286,
      "grad_norm": 0.5357667803764343,
      "learning_rate": 3e-05,
      "loss": 2.9365,
      "step": 1009
    },
    {
      "epoch": 0.31163221228016047,
      "grad_norm": 0.40679875016212463,
      "learning_rate": 3e-05,
      "loss": 2.6265,
      "step": 1010
    },
    {
      "epoch": 0.3119407590249923,
      "grad_norm": 0.4948636591434479,
      "learning_rate": 3e-05,
      "loss": 3.1441,
      "step": 1011
    },
    {
      "epoch": 0.3122493057698241,
      "grad_norm": 0.3034040331840515,
      "learning_rate": 3e-05,
      "loss": 2.331,
      "step": 1012
    },
    {
      "epoch": 0.31255785251465595,
      "grad_norm": 0.4402029812335968,
      "learning_rate": 3e-05,
      "loss": 2.7928,
      "step": 1013
    },
    {
      "epoch": 0.31286639925948784,
      "grad_norm": 0.40119442343711853,
      "learning_rate": 3e-05,
      "loss": 2.7245,
      "step": 1014
    },
    {
      "epoch": 0.31317494600431967,
      "grad_norm": 0.4709239900112152,
      "learning_rate": 3e-05,
      "loss": 2.8281,
      "step": 1015
    },
    {
      "epoch": 0.3134834927491515,
      "grad_norm": 0.43850257992744446,
      "learning_rate": 3e-05,
      "loss": 2.4768,
      "step": 1016
    },
    {
      "epoch": 0.3137920394939833,
      "grad_norm": 0.3689449429512024,
      "learning_rate": 3e-05,
      "loss": 2.4916,
      "step": 1017
    },
    {
      "epoch": 0.3141005862388152,
      "grad_norm": 0.3573774993419647,
      "learning_rate": 3e-05,
      "loss": 2.5842,
      "step": 1018
    },
    {
      "epoch": 0.31440913298364703,
      "grad_norm": 0.5338215231895447,
      "learning_rate": 3e-05,
      "loss": 3.1707,
      "step": 1019
    },
    {
      "epoch": 0.31471767972847886,
      "grad_norm": 0.47685402631759644,
      "learning_rate": 3e-05,
      "loss": 2.8287,
      "step": 1020
    },
    {
      "epoch": 0.3150262264733107,
      "grad_norm": 0.29378730058670044,
      "learning_rate": 3e-05,
      "loss": 2.3566,
      "step": 1021
    },
    {
      "epoch": 0.31533477321814257,
      "grad_norm": 0.3345448076725006,
      "learning_rate": 3e-05,
      "loss": 2.8883,
      "step": 1022
    },
    {
      "epoch": 0.3156433199629744,
      "grad_norm": 0.3130493760108948,
      "learning_rate": 3e-05,
      "loss": 2.2636,
      "step": 1023
    },
    {
      "epoch": 0.31595186670780623,
      "grad_norm": 0.4794589877128601,
      "learning_rate": 3e-05,
      "loss": 2.515,
      "step": 1024
    },
    {
      "epoch": 0.31626041345263806,
      "grad_norm": 0.4308234751224518,
      "learning_rate": 3e-05,
      "loss": 3.2071,
      "step": 1025
    },
    {
      "epoch": 0.31656896019746994,
      "grad_norm": 0.33414486050605774,
      "learning_rate": 3e-05,
      "loss": 2.5691,
      "step": 1026
    },
    {
      "epoch": 0.31687750694230177,
      "grad_norm": 0.3229790925979614,
      "learning_rate": 3e-05,
      "loss": 2.3644,
      "step": 1027
    },
    {
      "epoch": 0.3171860536871336,
      "grad_norm": 0.5011124610900879,
      "learning_rate": 3e-05,
      "loss": 2.7893,
      "step": 1028
    },
    {
      "epoch": 0.3174946004319654,
      "grad_norm": 0.37093424797058105,
      "learning_rate": 3e-05,
      "loss": 2.7773,
      "step": 1029
    },
    {
      "epoch": 0.3178031471767973,
      "grad_norm": 0.5820983052253723,
      "learning_rate": 3e-05,
      "loss": 2.826,
      "step": 1030
    },
    {
      "epoch": 0.31811169392162914,
      "grad_norm": 0.3745529353618622,
      "learning_rate": 3e-05,
      "loss": 2.678,
      "step": 1031
    },
    {
      "epoch": 0.31842024066646096,
      "grad_norm": 0.4362945854663849,
      "learning_rate": 3e-05,
      "loss": 2.7813,
      "step": 1032
    },
    {
      "epoch": 0.3187287874112928,
      "grad_norm": 0.45973077416419983,
      "learning_rate": 3e-05,
      "loss": 3.0007,
      "step": 1033
    },
    {
      "epoch": 0.3190373341561247,
      "grad_norm": 0.3230498731136322,
      "learning_rate": 3e-05,
      "loss": 2.1809,
      "step": 1034
    },
    {
      "epoch": 0.3193458809009565,
      "grad_norm": 0.3093605041503906,
      "learning_rate": 3e-05,
      "loss": 2.6337,
      "step": 1035
    },
    {
      "epoch": 0.31965442764578833,
      "grad_norm": 0.33831560611724854,
      "learning_rate": 3e-05,
      "loss": 2.667,
      "step": 1036
    },
    {
      "epoch": 0.31996297439062016,
      "grad_norm": 0.5394222140312195,
      "learning_rate": 3e-05,
      "loss": 3.0874,
      "step": 1037
    },
    {
      "epoch": 0.32027152113545204,
      "grad_norm": 0.46193042397499084,
      "learning_rate": 3e-05,
      "loss": 3.1571,
      "step": 1038
    },
    {
      "epoch": 0.32058006788028387,
      "grad_norm": 0.4071999788284302,
      "learning_rate": 3e-05,
      "loss": 2.7319,
      "step": 1039
    },
    {
      "epoch": 0.3208886146251157,
      "grad_norm": 0.4034838080406189,
      "learning_rate": 3e-05,
      "loss": 2.5469,
      "step": 1040
    },
    {
      "epoch": 0.3211971613699475,
      "grad_norm": 0.4690152406692505,
      "learning_rate": 3e-05,
      "loss": 2.7243,
      "step": 1041
    },
    {
      "epoch": 0.3215057081147794,
      "grad_norm": 0.33456355333328247,
      "learning_rate": 3e-05,
      "loss": 2.2573,
      "step": 1042
    },
    {
      "epoch": 0.32181425485961124,
      "grad_norm": 0.3276689350605011,
      "learning_rate": 3e-05,
      "loss": 2.3228,
      "step": 1043
    },
    {
      "epoch": 0.32212280160444307,
      "grad_norm": 0.3934038579463959,
      "learning_rate": 3e-05,
      "loss": 2.7742,
      "step": 1044
    },
    {
      "epoch": 0.3224313483492749,
      "grad_norm": 0.3162141740322113,
      "learning_rate": 3e-05,
      "loss": 2.3948,
      "step": 1045
    },
    {
      "epoch": 0.3227398950941068,
      "grad_norm": 0.35283178091049194,
      "learning_rate": 3e-05,
      "loss": 2.5482,
      "step": 1046
    },
    {
      "epoch": 0.3230484418389386,
      "grad_norm": 0.30801281332969666,
      "learning_rate": 3e-05,
      "loss": 2.3499,
      "step": 1047
    },
    {
      "epoch": 0.32335698858377043,
      "grad_norm": 0.2911578118801117,
      "learning_rate": 3e-05,
      "loss": 2.5108,
      "step": 1048
    },
    {
      "epoch": 0.32366553532860226,
      "grad_norm": 0.4680328965187073,
      "learning_rate": 3e-05,
      "loss": 2.8629,
      "step": 1049
    },
    {
      "epoch": 0.32397408207343414,
      "grad_norm": 0.27995920181274414,
      "learning_rate": 3e-05,
      "loss": 2.2291,
      "step": 1050
    },
    {
      "epoch": 0.32428262881826597,
      "grad_norm": 0.29879486560821533,
      "learning_rate": 3e-05,
      "loss": 2.4503,
      "step": 1051
    },
    {
      "epoch": 0.3245911755630978,
      "grad_norm": 0.4487646520137787,
      "learning_rate": 3e-05,
      "loss": 3.1378,
      "step": 1052
    },
    {
      "epoch": 0.32489972230792963,
      "grad_norm": 0.4540494680404663,
      "learning_rate": 3e-05,
      "loss": 3.0161,
      "step": 1053
    },
    {
      "epoch": 0.3252082690527615,
      "grad_norm": 0.6287296414375305,
      "learning_rate": 3e-05,
      "loss": 2.7699,
      "step": 1054
    },
    {
      "epoch": 0.32551681579759334,
      "grad_norm": 0.3549972176551819,
      "learning_rate": 3e-05,
      "loss": 2.517,
      "step": 1055
    },
    {
      "epoch": 0.32582536254242517,
      "grad_norm": 0.35076603293418884,
      "learning_rate": 3e-05,
      "loss": 2.5529,
      "step": 1056
    },
    {
      "epoch": 0.326133909287257,
      "grad_norm": 0.5306907296180725,
      "learning_rate": 3e-05,
      "loss": 2.6375,
      "step": 1057
    },
    {
      "epoch": 0.3264424560320889,
      "grad_norm": 0.6471700072288513,
      "learning_rate": 3e-05,
      "loss": 3.0257,
      "step": 1058
    },
    {
      "epoch": 0.3267510027769207,
      "grad_norm": 0.4791683256626129,
      "learning_rate": 3e-05,
      "loss": 2.673,
      "step": 1059
    },
    {
      "epoch": 0.32705954952175254,
      "grad_norm": 0.3479447364807129,
      "learning_rate": 3e-05,
      "loss": 2.4918,
      "step": 1060
    },
    {
      "epoch": 0.32736809626658436,
      "grad_norm": 0.5895714163780212,
      "learning_rate": 3e-05,
      "loss": 2.7425,
      "step": 1061
    },
    {
      "epoch": 0.32767664301141625,
      "grad_norm": 0.600107729434967,
      "learning_rate": 3e-05,
      "loss": 3.1736,
      "step": 1062
    },
    {
      "epoch": 0.3279851897562481,
      "grad_norm": 0.37535980343818665,
      "learning_rate": 3e-05,
      "loss": 2.5499,
      "step": 1063
    },
    {
      "epoch": 0.3282937365010799,
      "grad_norm": 0.5155684351921082,
      "learning_rate": 3e-05,
      "loss": 2.6339,
      "step": 1064
    },
    {
      "epoch": 0.32860228324591173,
      "grad_norm": 0.4304593801498413,
      "learning_rate": 3e-05,
      "loss": 2.9669,
      "step": 1065
    },
    {
      "epoch": 0.3289108299907436,
      "grad_norm": 0.38700851798057556,
      "learning_rate": 3e-05,
      "loss": 2.6592,
      "step": 1066
    },
    {
      "epoch": 0.32921937673557544,
      "grad_norm": 0.42323800921440125,
      "learning_rate": 3e-05,
      "loss": 2.468,
      "step": 1067
    },
    {
      "epoch": 0.32952792348040727,
      "grad_norm": 0.4084995687007904,
      "learning_rate": 3e-05,
      "loss": 2.908,
      "step": 1068
    },
    {
      "epoch": 0.3298364702252391,
      "grad_norm": 0.4313088655471802,
      "learning_rate": 3e-05,
      "loss": 2.6691,
      "step": 1069
    },
    {
      "epoch": 0.330145016970071,
      "grad_norm": 0.7936158776283264,
      "learning_rate": 3e-05,
      "loss": 2.912,
      "step": 1070
    },
    {
      "epoch": 0.3304535637149028,
      "grad_norm": 0.300073504447937,
      "learning_rate": 3e-05,
      "loss": 2.2288,
      "step": 1071
    },
    {
      "epoch": 0.33076211045973464,
      "grad_norm": 0.37451618909835815,
      "learning_rate": 3e-05,
      "loss": 2.9299,
      "step": 1072
    },
    {
      "epoch": 0.33107065720456647,
      "grad_norm": 0.5415565967559814,
      "learning_rate": 3e-05,
      "loss": 2.7914,
      "step": 1073
    },
    {
      "epoch": 0.33137920394939835,
      "grad_norm": 0.35867804288864136,
      "learning_rate": 3e-05,
      "loss": 2.8536,
      "step": 1074
    },
    {
      "epoch": 0.3316877506942302,
      "grad_norm": 0.4286486506462097,
      "learning_rate": 3e-05,
      "loss": 2.7861,
      "step": 1075
    },
    {
      "epoch": 0.331996297439062,
      "grad_norm": 0.31424498558044434,
      "learning_rate": 3e-05,
      "loss": 2.5602,
      "step": 1076
    },
    {
      "epoch": 0.33230484418389383,
      "grad_norm": 0.2903349697589874,
      "learning_rate": 3e-05,
      "loss": 2.2673,
      "step": 1077
    },
    {
      "epoch": 0.3326133909287257,
      "grad_norm": 0.3674512505531311,
      "learning_rate": 3e-05,
      "loss": 2.7277,
      "step": 1078
    },
    {
      "epoch": 0.33292193767355754,
      "grad_norm": 0.5027512311935425,
      "learning_rate": 3e-05,
      "loss": 2.6751,
      "step": 1079
    },
    {
      "epoch": 0.33323048441838937,
      "grad_norm": 0.33197149634361267,
      "learning_rate": 3e-05,
      "loss": 2.2715,
      "step": 1080
    },
    {
      "epoch": 0.33353903116322126,
      "grad_norm": 0.3024391233921051,
      "learning_rate": 3e-05,
      "loss": 2.5931,
      "step": 1081
    },
    {
      "epoch": 0.3338475779080531,
      "grad_norm": 0.5677501559257507,
      "learning_rate": 3e-05,
      "loss": 2.7854,
      "step": 1082
    },
    {
      "epoch": 0.3341561246528849,
      "grad_norm": 0.349448025226593,
      "learning_rate": 3e-05,
      "loss": 2.5758,
      "step": 1083
    },
    {
      "epoch": 0.33446467139771674,
      "grad_norm": 0.4105369746685028,
      "learning_rate": 3e-05,
      "loss": 2.5552,
      "step": 1084
    },
    {
      "epoch": 0.3347732181425486,
      "grad_norm": 0.423350989818573,
      "learning_rate": 3e-05,
      "loss": 2.8894,
      "step": 1085
    },
    {
      "epoch": 0.33508176488738045,
      "grad_norm": 0.45168107748031616,
      "learning_rate": 3e-05,
      "loss": 2.7562,
      "step": 1086
    },
    {
      "epoch": 0.3353903116322123,
      "grad_norm": 0.354754775762558,
      "learning_rate": 3e-05,
      "loss": 2.8504,
      "step": 1087
    },
    {
      "epoch": 0.3356988583770441,
      "grad_norm": 0.31282752752304077,
      "learning_rate": 3e-05,
      "loss": 2.6208,
      "step": 1088
    },
    {
      "epoch": 0.336007405121876,
      "grad_norm": 0.3565126359462738,
      "learning_rate": 3e-05,
      "loss": 2.7854,
      "step": 1089
    },
    {
      "epoch": 0.3363159518667078,
      "grad_norm": 0.4445134401321411,
      "learning_rate": 3e-05,
      "loss": 2.8022,
      "step": 1090
    },
    {
      "epoch": 0.33662449861153965,
      "grad_norm": 0.682829737663269,
      "learning_rate": 3e-05,
      "loss": 2.7264,
      "step": 1091
    },
    {
      "epoch": 0.3369330453563715,
      "grad_norm": 0.4143361747264862,
      "learning_rate": 3e-05,
      "loss": 2.8082,
      "step": 1092
    },
    {
      "epoch": 0.33724159210120336,
      "grad_norm": 0.4276883602142334,
      "learning_rate": 3e-05,
      "loss": 2.5857,
      "step": 1093
    },
    {
      "epoch": 0.3375501388460352,
      "grad_norm": 0.5904972553253174,
      "learning_rate": 3e-05,
      "loss": 2.4174,
      "step": 1094
    },
    {
      "epoch": 0.337858685590867,
      "grad_norm": 0.5379177927970886,
      "learning_rate": 3e-05,
      "loss": 2.9506,
      "step": 1095
    },
    {
      "epoch": 0.33816723233569884,
      "grad_norm": 0.3578889071941376,
      "learning_rate": 3e-05,
      "loss": 2.6252,
      "step": 1096
    },
    {
      "epoch": 0.3384757790805307,
      "grad_norm": 0.459988534450531,
      "learning_rate": 3e-05,
      "loss": 2.7379,
      "step": 1097
    },
    {
      "epoch": 0.33878432582536255,
      "grad_norm": 0.5335902571678162,
      "learning_rate": 3e-05,
      "loss": 2.7689,
      "step": 1098
    },
    {
      "epoch": 0.3390928725701944,
      "grad_norm": 0.427128404378891,
      "learning_rate": 3e-05,
      "loss": 2.5377,
      "step": 1099
    },
    {
      "epoch": 0.3394014193150262,
      "grad_norm": 0.36371105909347534,
      "learning_rate": 3e-05,
      "loss": 2.3228,
      "step": 1100
    },
    {
      "epoch": 0.3397099660598581,
      "grad_norm": 0.6602774262428284,
      "learning_rate": 3e-05,
      "loss": 2.7878,
      "step": 1101
    },
    {
      "epoch": 0.3400185128046899,
      "grad_norm": 0.27885714173316956,
      "learning_rate": 3e-05,
      "loss": 2.2656,
      "step": 1102
    },
    {
      "epoch": 0.34032705954952175,
      "grad_norm": 0.5975661277770996,
      "learning_rate": 3e-05,
      "loss": 2.9421,
      "step": 1103
    },
    {
      "epoch": 0.3406356062943536,
      "grad_norm": 0.6099026203155518,
      "learning_rate": 3e-05,
      "loss": 3.1214,
      "step": 1104
    },
    {
      "epoch": 0.34094415303918546,
      "grad_norm": 0.4762013554573059,
      "learning_rate": 3e-05,
      "loss": 2.3022,
      "step": 1105
    },
    {
      "epoch": 0.3412526997840173,
      "grad_norm": 0.6812524795532227,
      "learning_rate": 3e-05,
      "loss": 2.5754,
      "step": 1106
    },
    {
      "epoch": 0.3415612465288491,
      "grad_norm": 0.41604599356651306,
      "learning_rate": 3e-05,
      "loss": 2.7934,
      "step": 1107
    },
    {
      "epoch": 0.34186979327368094,
      "grad_norm": 0.40861639380455017,
      "learning_rate": 3e-05,
      "loss": 2.7791,
      "step": 1108
    },
    {
      "epoch": 0.3421783400185128,
      "grad_norm": 0.533877968788147,
      "learning_rate": 3e-05,
      "loss": 3.0471,
      "step": 1109
    },
    {
      "epoch": 0.34248688676334466,
      "grad_norm": 0.34040915966033936,
      "learning_rate": 3e-05,
      "loss": 2.1942,
      "step": 1110
    },
    {
      "epoch": 0.3427954335081765,
      "grad_norm": 0.5931209921836853,
      "learning_rate": 3e-05,
      "loss": 2.6718,
      "step": 1111
    },
    {
      "epoch": 0.3431039802530083,
      "grad_norm": 0.4841914772987366,
      "learning_rate": 3e-05,
      "loss": 2.5222,
      "step": 1112
    },
    {
      "epoch": 0.3434125269978402,
      "grad_norm": 0.32610583305358887,
      "learning_rate": 3e-05,
      "loss": 2.5451,
      "step": 1113
    },
    {
      "epoch": 0.343721073742672,
      "grad_norm": 0.42859357595443726,
      "learning_rate": 3e-05,
      "loss": 2.9902,
      "step": 1114
    },
    {
      "epoch": 0.34402962048750385,
      "grad_norm": 0.4420587122440338,
      "learning_rate": 3e-05,
      "loss": 2.6019,
      "step": 1115
    },
    {
      "epoch": 0.3443381672323357,
      "grad_norm": 0.4523164927959442,
      "learning_rate": 3e-05,
      "loss": 2.6256,
      "step": 1116
    },
    {
      "epoch": 0.34464671397716756,
      "grad_norm": 0.4764743447303772,
      "learning_rate": 3e-05,
      "loss": 2.942,
      "step": 1117
    },
    {
      "epoch": 0.3449552607219994,
      "grad_norm": 0.40522849559783936,
      "learning_rate": 3e-05,
      "loss": 2.6806,
      "step": 1118
    },
    {
      "epoch": 0.3452638074668312,
      "grad_norm": 0.331582248210907,
      "learning_rate": 3e-05,
      "loss": 2.3276,
      "step": 1119
    },
    {
      "epoch": 0.34557235421166305,
      "grad_norm": 0.4719706177711487,
      "learning_rate": 3e-05,
      "loss": 3.0291,
      "step": 1120
    },
    {
      "epoch": 0.34588090095649493,
      "grad_norm": 0.44904038310050964,
      "learning_rate": 3e-05,
      "loss": 2.8547,
      "step": 1121
    },
    {
      "epoch": 0.34618944770132676,
      "grad_norm": 0.43688273429870605,
      "learning_rate": 3e-05,
      "loss": 2.4615,
      "step": 1122
    },
    {
      "epoch": 0.3464979944461586,
      "grad_norm": 0.4403668940067291,
      "learning_rate": 3e-05,
      "loss": 3.0372,
      "step": 1123
    },
    {
      "epoch": 0.3468065411909904,
      "grad_norm": 0.3114491105079651,
      "learning_rate": 3e-05,
      "loss": 2.5904,
      "step": 1124
    },
    {
      "epoch": 0.3471150879358223,
      "grad_norm": 0.43030625581741333,
      "learning_rate": 3e-05,
      "loss": 2.5914,
      "step": 1125
    },
    {
      "epoch": 0.3474236346806541,
      "grad_norm": 0.5228371620178223,
      "learning_rate": 3e-05,
      "loss": 3.269,
      "step": 1126
    },
    {
      "epoch": 0.34773218142548595,
      "grad_norm": 0.35842400789260864,
      "learning_rate": 3e-05,
      "loss": 2.6338,
      "step": 1127
    },
    {
      "epoch": 0.3480407281703178,
      "grad_norm": 0.33594179153442383,
      "learning_rate": 3e-05,
      "loss": 2.5405,
      "step": 1128
    },
    {
      "epoch": 0.34834927491514966,
      "grad_norm": 0.34110668301582336,
      "learning_rate": 3e-05,
      "loss": 2.5228,
      "step": 1129
    },
    {
      "epoch": 0.3486578216599815,
      "grad_norm": 0.42432937026023865,
      "learning_rate": 3e-05,
      "loss": 2.8293,
      "step": 1130
    },
    {
      "epoch": 0.3489663684048133,
      "grad_norm": 0.42318588495254517,
      "learning_rate": 3e-05,
      "loss": 2.2771,
      "step": 1131
    },
    {
      "epoch": 0.34927491514964515,
      "grad_norm": 0.34015128016471863,
      "learning_rate": 3e-05,
      "loss": 2.7245,
      "step": 1132
    },
    {
      "epoch": 0.34958346189447703,
      "grad_norm": 0.3692342936992645,
      "learning_rate": 3e-05,
      "loss": 2.5157,
      "step": 1133
    },
    {
      "epoch": 0.34989200863930886,
      "grad_norm": 0.5286168456077576,
      "learning_rate": 3e-05,
      "loss": 3.3151,
      "step": 1134
    },
    {
      "epoch": 0.3502005553841407,
      "grad_norm": 0.4966541528701782,
      "learning_rate": 3e-05,
      "loss": 3.3489,
      "step": 1135
    },
    {
      "epoch": 0.3505091021289725,
      "grad_norm": 0.5056973695755005,
      "learning_rate": 3e-05,
      "loss": 2.8628,
      "step": 1136
    },
    {
      "epoch": 0.3508176488738044,
      "grad_norm": 0.40126872062683105,
      "learning_rate": 3e-05,
      "loss": 3.2355,
      "step": 1137
    },
    {
      "epoch": 0.3511261956186362,
      "grad_norm": 0.4347645938396454,
      "learning_rate": 3e-05,
      "loss": 2.3129,
      "step": 1138
    },
    {
      "epoch": 0.35143474236346806,
      "grad_norm": 0.5079309940338135,
      "learning_rate": 3e-05,
      "loss": 2.6779,
      "step": 1139
    },
    {
      "epoch": 0.3517432891082999,
      "grad_norm": 0.34270283579826355,
      "learning_rate": 3e-05,
      "loss": 2.6367,
      "step": 1140
    },
    {
      "epoch": 0.35205183585313177,
      "grad_norm": 0.3936125636100769,
      "learning_rate": 3e-05,
      "loss": 2.7157,
      "step": 1141
    },
    {
      "epoch": 0.3523603825979636,
      "grad_norm": 0.5401539206504822,
      "learning_rate": 3e-05,
      "loss": 3.1242,
      "step": 1142
    },
    {
      "epoch": 0.3526689293427954,
      "grad_norm": 0.3918668329715729,
      "learning_rate": 3e-05,
      "loss": 2.9379,
      "step": 1143
    },
    {
      "epoch": 0.35297747608762725,
      "grad_norm": 0.3724942207336426,
      "learning_rate": 3e-05,
      "loss": 2.5844,
      "step": 1144
    },
    {
      "epoch": 0.35328602283245913,
      "grad_norm": 0.4116598963737488,
      "learning_rate": 3e-05,
      "loss": 2.6892,
      "step": 1145
    },
    {
      "epoch": 0.35359456957729096,
      "grad_norm": 0.5307642817497253,
      "learning_rate": 3e-05,
      "loss": 2.7121,
      "step": 1146
    },
    {
      "epoch": 0.3539031163221228,
      "grad_norm": 0.34816232323646545,
      "learning_rate": 3e-05,
      "loss": 2.6197,
      "step": 1147
    },
    {
      "epoch": 0.3542116630669546,
      "grad_norm": 0.3448748290538788,
      "learning_rate": 3e-05,
      "loss": 2.5077,
      "step": 1148
    },
    {
      "epoch": 0.3545202098117865,
      "grad_norm": 0.39467447996139526,
      "learning_rate": 3e-05,
      "loss": 2.3641,
      "step": 1149
    },
    {
      "epoch": 0.35482875655661833,
      "grad_norm": 0.39322465658187866,
      "learning_rate": 3e-05,
      "loss": 2.6104,
      "step": 1150
    },
    {
      "epoch": 0.35513730330145016,
      "grad_norm": 0.34313109517097473,
      "learning_rate": 3e-05,
      "loss": 2.7748,
      "step": 1151
    },
    {
      "epoch": 0.355445850046282,
      "grad_norm": 0.32297638058662415,
      "learning_rate": 3e-05,
      "loss": 2.6006,
      "step": 1152
    },
    {
      "epoch": 0.35575439679111387,
      "grad_norm": 0.35026779770851135,
      "learning_rate": 3e-05,
      "loss": 2.3611,
      "step": 1153
    },
    {
      "epoch": 0.3560629435359457,
      "grad_norm": 0.4179763197898865,
      "learning_rate": 3e-05,
      "loss": 3.1878,
      "step": 1154
    },
    {
      "epoch": 0.3563714902807775,
      "grad_norm": 0.3217645287513733,
      "learning_rate": 3e-05,
      "loss": 2.3414,
      "step": 1155
    },
    {
      "epoch": 0.35668003702560935,
      "grad_norm": 0.43228569626808167,
      "learning_rate": 3e-05,
      "loss": 3.0959,
      "step": 1156
    },
    {
      "epoch": 0.35698858377044124,
      "grad_norm": 0.5505173802375793,
      "learning_rate": 3e-05,
      "loss": 3.0136,
      "step": 1157
    },
    {
      "epoch": 0.35729713051527306,
      "grad_norm": 0.31909194588661194,
      "learning_rate": 3e-05,
      "loss": 2.4238,
      "step": 1158
    },
    {
      "epoch": 0.3576056772601049,
      "grad_norm": 0.38182443380355835,
      "learning_rate": 3e-05,
      "loss": 2.8554,
      "step": 1159
    },
    {
      "epoch": 0.3579142240049367,
      "grad_norm": 0.3917055130004883,
      "learning_rate": 3e-05,
      "loss": 2.4918,
      "step": 1160
    },
    {
      "epoch": 0.3582227707497686,
      "grad_norm": 0.30668020248413086,
      "learning_rate": 3e-05,
      "loss": 2.4466,
      "step": 1161
    },
    {
      "epoch": 0.35853131749460043,
      "grad_norm": 0.4843204617500305,
      "learning_rate": 3e-05,
      "loss": 3.0447,
      "step": 1162
    },
    {
      "epoch": 0.35883986423943226,
      "grad_norm": 0.3819452226161957,
      "learning_rate": 3e-05,
      "loss": 2.8378,
      "step": 1163
    },
    {
      "epoch": 0.35914841098426414,
      "grad_norm": 0.41383904218673706,
      "learning_rate": 3e-05,
      "loss": 2.558,
      "step": 1164
    },
    {
      "epoch": 0.35945695772909597,
      "grad_norm": 0.31443139910697937,
      "learning_rate": 3e-05,
      "loss": 2.7383,
      "step": 1165
    },
    {
      "epoch": 0.3597655044739278,
      "grad_norm": 0.39199626445770264,
      "learning_rate": 3e-05,
      "loss": 2.7038,
      "step": 1166
    },
    {
      "epoch": 0.3600740512187596,
      "grad_norm": 0.3552488386631012,
      "learning_rate": 3e-05,
      "loss": 2.3299,
      "step": 1167
    },
    {
      "epoch": 0.3603825979635915,
      "grad_norm": 0.45004716515541077,
      "learning_rate": 3e-05,
      "loss": 3.0991,
      "step": 1168
    },
    {
      "epoch": 0.36069114470842334,
      "grad_norm": 0.3720468282699585,
      "learning_rate": 3e-05,
      "loss": 2.5241,
      "step": 1169
    },
    {
      "epoch": 0.36099969145325517,
      "grad_norm": 0.4281458258628845,
      "learning_rate": 3e-05,
      "loss": 2.9697,
      "step": 1170
    },
    {
      "epoch": 0.361308238198087,
      "grad_norm": 0.4345632791519165,
      "learning_rate": 3e-05,
      "loss": 2.473,
      "step": 1171
    },
    {
      "epoch": 0.3616167849429189,
      "grad_norm": 0.28444620966911316,
      "learning_rate": 3e-05,
      "loss": 2.2297,
      "step": 1172
    },
    {
      "epoch": 0.3619253316877507,
      "grad_norm": 0.33514413237571716,
      "learning_rate": 3e-05,
      "loss": 2.7778,
      "step": 1173
    },
    {
      "epoch": 0.36223387843258253,
      "grad_norm": 0.29754677414894104,
      "learning_rate": 3e-05,
      "loss": 2.5157,
      "step": 1174
    },
    {
      "epoch": 0.36254242517741436,
      "grad_norm": 0.3521910309791565,
      "learning_rate": 3e-05,
      "loss": 2.8631,
      "step": 1175
    },
    {
      "epoch": 0.36285097192224625,
      "grad_norm": 0.35081109404563904,
      "learning_rate": 3e-05,
      "loss": 2.3547,
      "step": 1176
    },
    {
      "epoch": 0.3631595186670781,
      "grad_norm": 0.6115928292274475,
      "learning_rate": 3e-05,
      "loss": 3.0622,
      "step": 1177
    },
    {
      "epoch": 0.3634680654119099,
      "grad_norm": 0.3421701192855835,
      "learning_rate": 3e-05,
      "loss": 2.5053,
      "step": 1178
    },
    {
      "epoch": 0.36377661215674173,
      "grad_norm": 0.42282208800315857,
      "learning_rate": 3e-05,
      "loss": 3.1358,
      "step": 1179
    },
    {
      "epoch": 0.3640851589015736,
      "grad_norm": 0.3527933359146118,
      "learning_rate": 3e-05,
      "loss": 2.7091,
      "step": 1180
    },
    {
      "epoch": 0.36439370564640544,
      "grad_norm": 0.40788233280181885,
      "learning_rate": 3e-05,
      "loss": 2.7765,
      "step": 1181
    },
    {
      "epoch": 0.36470225239123727,
      "grad_norm": 0.39160051941871643,
      "learning_rate": 3e-05,
      "loss": 2.8549,
      "step": 1182
    },
    {
      "epoch": 0.3650107991360691,
      "grad_norm": 0.4997316598892212,
      "learning_rate": 3e-05,
      "loss": 2.7178,
      "step": 1183
    },
    {
      "epoch": 0.365319345880901,
      "grad_norm": 0.376682847738266,
      "learning_rate": 3e-05,
      "loss": 2.7221,
      "step": 1184
    },
    {
      "epoch": 0.3656278926257328,
      "grad_norm": 0.5439577102661133,
      "learning_rate": 3e-05,
      "loss": 2.7612,
      "step": 1185
    },
    {
      "epoch": 0.36593643937056464,
      "grad_norm": 0.375169962644577,
      "learning_rate": 3e-05,
      "loss": 2.8623,
      "step": 1186
    },
    {
      "epoch": 0.36624498611539646,
      "grad_norm": 0.3876233994960785,
      "learning_rate": 3e-05,
      "loss": 2.7473,
      "step": 1187
    },
    {
      "epoch": 0.36655353286022835,
      "grad_norm": 0.36275288462638855,
      "learning_rate": 3e-05,
      "loss": 2.7514,
      "step": 1188
    },
    {
      "epoch": 0.3668620796050602,
      "grad_norm": 0.36678746342658997,
      "learning_rate": 3e-05,
      "loss": 2.8159,
      "step": 1189
    },
    {
      "epoch": 0.367170626349892,
      "grad_norm": 0.4482974410057068,
      "learning_rate": 3e-05,
      "loss": 2.6262,
      "step": 1190
    },
    {
      "epoch": 0.36747917309472383,
      "grad_norm": 0.5268934369087219,
      "learning_rate": 3e-05,
      "loss": 3.0075,
      "step": 1191
    },
    {
      "epoch": 0.3677877198395557,
      "grad_norm": 0.5856001377105713,
      "learning_rate": 3e-05,
      "loss": 3.1768,
      "step": 1192
    },
    {
      "epoch": 0.36809626658438754,
      "grad_norm": 0.36955198645591736,
      "learning_rate": 3e-05,
      "loss": 2.2928,
      "step": 1193
    },
    {
      "epoch": 0.36840481332921937,
      "grad_norm": 0.4243110120296478,
      "learning_rate": 3e-05,
      "loss": 2.8757,
      "step": 1194
    },
    {
      "epoch": 0.3687133600740512,
      "grad_norm": 0.5759037733078003,
      "learning_rate": 3e-05,
      "loss": 3.1536,
      "step": 1195
    },
    {
      "epoch": 0.3690219068188831,
      "grad_norm": 0.39592766761779785,
      "learning_rate": 3e-05,
      "loss": 2.6084,
      "step": 1196
    },
    {
      "epoch": 0.3693304535637149,
      "grad_norm": 0.4080641567707062,
      "learning_rate": 3e-05,
      "loss": 2.5055,
      "step": 1197
    },
    {
      "epoch": 0.36963900030854674,
      "grad_norm": 0.46009865403175354,
      "learning_rate": 3e-05,
      "loss": 2.9701,
      "step": 1198
    },
    {
      "epoch": 0.36994754705337857,
      "grad_norm": 0.2832425832748413,
      "learning_rate": 3e-05,
      "loss": 2.3101,
      "step": 1199
    },
    {
      "epoch": 0.37025609379821045,
      "grad_norm": 0.3997393548488617,
      "learning_rate": 3e-05,
      "loss": 2.5356,
      "step": 1200
    },
    {
      "epoch": 0.3705646405430423,
      "grad_norm": 0.3031362295150757,
      "learning_rate": 3e-05,
      "loss": 2.0162,
      "step": 1201
    },
    {
      "epoch": 0.3708731872878741,
      "grad_norm": 0.29112961888313293,
      "learning_rate": 3e-05,
      "loss": 2.1462,
      "step": 1202
    },
    {
      "epoch": 0.37118173403270593,
      "grad_norm": 0.38319069147109985,
      "learning_rate": 3e-05,
      "loss": 2.0506,
      "step": 1203
    },
    {
      "epoch": 0.3714902807775378,
      "grad_norm": 0.4330032467842102,
      "learning_rate": 3e-05,
      "loss": 2.3374,
      "step": 1204
    },
    {
      "epoch": 0.37179882752236965,
      "grad_norm": 0.2930798828601837,
      "learning_rate": 3e-05,
      "loss": 2.4065,
      "step": 1205
    },
    {
      "epoch": 0.3721073742672015,
      "grad_norm": 0.37502041459083557,
      "learning_rate": 3e-05,
      "loss": 2.5894,
      "step": 1206
    },
    {
      "epoch": 0.3724159210120333,
      "grad_norm": 0.3615609109401703,
      "learning_rate": 3e-05,
      "loss": 2.4882,
      "step": 1207
    },
    {
      "epoch": 0.3727244677568652,
      "grad_norm": 0.4277931749820709,
      "learning_rate": 3e-05,
      "loss": 2.4733,
      "step": 1208
    },
    {
      "epoch": 0.373033014501697,
      "grad_norm": 0.33783355355262756,
      "learning_rate": 3e-05,
      "loss": 2.66,
      "step": 1209
    },
    {
      "epoch": 0.37334156124652884,
      "grad_norm": 0.37479284405708313,
      "learning_rate": 3e-05,
      "loss": 2.575,
      "step": 1210
    },
    {
      "epoch": 0.37365010799136067,
      "grad_norm": 0.4059560298919678,
      "learning_rate": 3e-05,
      "loss": 2.8207,
      "step": 1211
    },
    {
      "epoch": 0.37395865473619255,
      "grad_norm": 0.50154709815979,
      "learning_rate": 3e-05,
      "loss": 2.8768,
      "step": 1212
    },
    {
      "epoch": 0.3742672014810244,
      "grad_norm": 0.3010912239551544,
      "learning_rate": 3e-05,
      "loss": 2.2838,
      "step": 1213
    },
    {
      "epoch": 0.3745757482258562,
      "grad_norm": 0.45400598645210266,
      "learning_rate": 3e-05,
      "loss": 2.504,
      "step": 1214
    },
    {
      "epoch": 0.37488429497068804,
      "grad_norm": 0.3583899140357971,
      "learning_rate": 3e-05,
      "loss": 2.8666,
      "step": 1215
    },
    {
      "epoch": 0.3751928417155199,
      "grad_norm": 0.5383654236793518,
      "learning_rate": 3e-05,
      "loss": 2.9554,
      "step": 1216
    },
    {
      "epoch": 0.37550138846035175,
      "grad_norm": 0.29187503457069397,
      "learning_rate": 3e-05,
      "loss": 2.2723,
      "step": 1217
    },
    {
      "epoch": 0.3758099352051836,
      "grad_norm": 0.32889237999916077,
      "learning_rate": 3e-05,
      "loss": 2.3481,
      "step": 1218
    },
    {
      "epoch": 0.3761184819500154,
      "grad_norm": 0.29874467849731445,
      "learning_rate": 3e-05,
      "loss": 2.1303,
      "step": 1219
    },
    {
      "epoch": 0.3764270286948473,
      "grad_norm": 0.3520304560661316,
      "learning_rate": 3e-05,
      "loss": 2.5011,
      "step": 1220
    },
    {
      "epoch": 0.3767355754396791,
      "grad_norm": 0.31358984112739563,
      "learning_rate": 3e-05,
      "loss": 2.3208,
      "step": 1221
    },
    {
      "epoch": 0.37704412218451094,
      "grad_norm": 0.921454668045044,
      "learning_rate": 3e-05,
      "loss": 3.7345,
      "step": 1222
    },
    {
      "epoch": 0.37735266892934277,
      "grad_norm": 0.600284218788147,
      "learning_rate": 3e-05,
      "loss": 2.6006,
      "step": 1223
    },
    {
      "epoch": 0.37766121567417466,
      "grad_norm": 0.6821380853652954,
      "learning_rate": 3e-05,
      "loss": 2.7102,
      "step": 1224
    },
    {
      "epoch": 0.3779697624190065,
      "grad_norm": 0.46705061197280884,
      "learning_rate": 3e-05,
      "loss": 2.6111,
      "step": 1225
    },
    {
      "epoch": 0.3782783091638383,
      "grad_norm": 0.41461801528930664,
      "learning_rate": 3e-05,
      "loss": 2.5879,
      "step": 1226
    },
    {
      "epoch": 0.37858685590867014,
      "grad_norm": 0.3981480896472931,
      "learning_rate": 3e-05,
      "loss": 2.5165,
      "step": 1227
    },
    {
      "epoch": 0.378895402653502,
      "grad_norm": 0.6518423557281494,
      "learning_rate": 3e-05,
      "loss": 2.8353,
      "step": 1228
    },
    {
      "epoch": 0.37920394939833385,
      "grad_norm": 0.5395017862319946,
      "learning_rate": 3e-05,
      "loss": 2.6184,
      "step": 1229
    },
    {
      "epoch": 0.3795124961431657,
      "grad_norm": 0.4610118865966797,
      "learning_rate": 3e-05,
      "loss": 2.6831,
      "step": 1230
    },
    {
      "epoch": 0.3798210428879975,
      "grad_norm": 0.38779354095458984,
      "learning_rate": 3e-05,
      "loss": 2.6815,
      "step": 1231
    },
    {
      "epoch": 0.3801295896328294,
      "grad_norm": 0.5472552180290222,
      "learning_rate": 3e-05,
      "loss": 2.4982,
      "step": 1232
    },
    {
      "epoch": 0.3804381363776612,
      "grad_norm": 0.5459097623825073,
      "learning_rate": 3e-05,
      "loss": 2.4581,
      "step": 1233
    },
    {
      "epoch": 0.38074668312249305,
      "grad_norm": 0.47404393553733826,
      "learning_rate": 3e-05,
      "loss": 2.7392,
      "step": 1234
    },
    {
      "epoch": 0.3810552298673249,
      "grad_norm": 0.428065687417984,
      "learning_rate": 3e-05,
      "loss": 2.6488,
      "step": 1235
    },
    {
      "epoch": 0.38136377661215676,
      "grad_norm": 0.6436012983322144,
      "learning_rate": 3e-05,
      "loss": 2.8324,
      "step": 1236
    },
    {
      "epoch": 0.3816723233569886,
      "grad_norm": 0.7773986458778381,
      "learning_rate": 3e-05,
      "loss": 2.7858,
      "step": 1237
    },
    {
      "epoch": 0.3819808701018204,
      "grad_norm": 0.4160424470901489,
      "learning_rate": 3e-05,
      "loss": 2.7753,
      "step": 1238
    },
    {
      "epoch": 0.38228941684665224,
      "grad_norm": 0.3402724266052246,
      "learning_rate": 3e-05,
      "loss": 2.6381,
      "step": 1239
    },
    {
      "epoch": 0.3825979635914841,
      "grad_norm": 0.4877413511276245,
      "learning_rate": 3e-05,
      "loss": 3.0552,
      "step": 1240
    },
    {
      "epoch": 0.38290651033631595,
      "grad_norm": 0.565540611743927,
      "learning_rate": 3e-05,
      "loss": 2.4953,
      "step": 1241
    },
    {
      "epoch": 0.3832150570811478,
      "grad_norm": 0.4378884434700012,
      "learning_rate": 3e-05,
      "loss": 2.3667,
      "step": 1242
    },
    {
      "epoch": 0.3835236038259796,
      "grad_norm": 0.31386178731918335,
      "learning_rate": 3e-05,
      "loss": 2.5664,
      "step": 1243
    },
    {
      "epoch": 0.3838321505708115,
      "grad_norm": 0.3567987382411957,
      "learning_rate": 3e-05,
      "loss": 2.6788,
      "step": 1244
    },
    {
      "epoch": 0.3841406973156433,
      "grad_norm": 0.31117892265319824,
      "learning_rate": 3e-05,
      "loss": 2.3698,
      "step": 1245
    },
    {
      "epoch": 0.38444924406047515,
      "grad_norm": 0.32202041149139404,
      "learning_rate": 3e-05,
      "loss": 2.4283,
      "step": 1246
    },
    {
      "epoch": 0.38475779080530703,
      "grad_norm": 0.3621535897254944,
      "learning_rate": 3e-05,
      "loss": 2.4739,
      "step": 1247
    },
    {
      "epoch": 0.38506633755013886,
      "grad_norm": 0.3923341929912567,
      "learning_rate": 3e-05,
      "loss": 2.7244,
      "step": 1248
    },
    {
      "epoch": 0.3853748842949707,
      "grad_norm": 0.3279580771923065,
      "learning_rate": 3e-05,
      "loss": 2.395,
      "step": 1249
    },
    {
      "epoch": 0.3856834310398025,
      "grad_norm": 0.36575204133987427,
      "learning_rate": 3e-05,
      "loss": 2.5109,
      "step": 1250
    },
    {
      "epoch": 0.3859919777846344,
      "grad_norm": 0.4692917764186859,
      "learning_rate": 3e-05,
      "loss": 3.1119,
      "step": 1251
    },
    {
      "epoch": 0.3863005245294662,
      "grad_norm": 0.4051806330680847,
      "learning_rate": 3e-05,
      "loss": 2.6323,
      "step": 1252
    },
    {
      "epoch": 0.38660907127429806,
      "grad_norm": 0.4611109793186188,
      "learning_rate": 3e-05,
      "loss": 2.4552,
      "step": 1253
    },
    {
      "epoch": 0.3869176180191299,
      "grad_norm": 0.5289100408554077,
      "learning_rate": 3e-05,
      "loss": 3.3514,
      "step": 1254
    },
    {
      "epoch": 0.38722616476396177,
      "grad_norm": 0.3087036907672882,
      "learning_rate": 3e-05,
      "loss": 2.6185,
      "step": 1255
    },
    {
      "epoch": 0.3875347115087936,
      "grad_norm": 0.46260780096054077,
      "learning_rate": 3e-05,
      "loss": 2.8345,
      "step": 1256
    },
    {
      "epoch": 0.3878432582536254,
      "grad_norm": 0.3516508936882019,
      "learning_rate": 3e-05,
      "loss": 2.7413,
      "step": 1257
    },
    {
      "epoch": 0.38815180499845725,
      "grad_norm": 0.3896012008190155,
      "learning_rate": 3e-05,
      "loss": 2.8107,
      "step": 1258
    },
    {
      "epoch": 0.38846035174328913,
      "grad_norm": 0.35049739480018616,
      "learning_rate": 3e-05,
      "loss": 2.3946,
      "step": 1259
    },
    {
      "epoch": 0.38876889848812096,
      "grad_norm": 0.401770681142807,
      "learning_rate": 3e-05,
      "loss": 2.8708,
      "step": 1260
    },
    {
      "epoch": 0.3890774452329528,
      "grad_norm": 0.3014017939567566,
      "learning_rate": 3e-05,
      "loss": 2.7118,
      "step": 1261
    },
    {
      "epoch": 0.3893859919777846,
      "grad_norm": 0.32284262776374817,
      "learning_rate": 3e-05,
      "loss": 2.7672,
      "step": 1262
    },
    {
      "epoch": 0.3896945387226165,
      "grad_norm": 0.38780876994132996,
      "learning_rate": 3e-05,
      "loss": 2.5998,
      "step": 1263
    },
    {
      "epoch": 0.39000308546744833,
      "grad_norm": 0.2658548057079315,
      "learning_rate": 3e-05,
      "loss": 2.3036,
      "step": 1264
    },
    {
      "epoch": 0.39031163221228016,
      "grad_norm": 0.5888837575912476,
      "learning_rate": 3e-05,
      "loss": 3.4526,
      "step": 1265
    },
    {
      "epoch": 0.390620178957112,
      "grad_norm": 0.3705357015132904,
      "learning_rate": 3e-05,
      "loss": 2.9579,
      "step": 1266
    },
    {
      "epoch": 0.39092872570194387,
      "grad_norm": 0.28175267577171326,
      "learning_rate": 3e-05,
      "loss": 2.1632,
      "step": 1267
    },
    {
      "epoch": 0.3912372724467757,
      "grad_norm": 0.33350950479507446,
      "learning_rate": 3e-05,
      "loss": 2.9705,
      "step": 1268
    },
    {
      "epoch": 0.3915458191916075,
      "grad_norm": 0.4463180601596832,
      "learning_rate": 3e-05,
      "loss": 2.6577,
      "step": 1269
    },
    {
      "epoch": 0.39185436593643935,
      "grad_norm": 0.5363352298736572,
      "learning_rate": 3e-05,
      "loss": 3.4181,
      "step": 1270
    },
    {
      "epoch": 0.39216291268127124,
      "grad_norm": 0.32660403847694397,
      "learning_rate": 3e-05,
      "loss": 2.6473,
      "step": 1271
    },
    {
      "epoch": 0.39247145942610306,
      "grad_norm": 0.4326009750366211,
      "learning_rate": 3e-05,
      "loss": 2.6244,
      "step": 1272
    },
    {
      "epoch": 0.3927800061709349,
      "grad_norm": 0.43666842579841614,
      "learning_rate": 3e-05,
      "loss": 2.6443,
      "step": 1273
    },
    {
      "epoch": 0.3930885529157667,
      "grad_norm": 0.3615734279155731,
      "learning_rate": 3e-05,
      "loss": 2.4956,
      "step": 1274
    },
    {
      "epoch": 0.3933970996605986,
      "grad_norm": 0.6071893572807312,
      "learning_rate": 3e-05,
      "loss": 3.1909,
      "step": 1275
    },
    {
      "epoch": 0.39370564640543043,
      "grad_norm": 0.3208840489387512,
      "learning_rate": 3e-05,
      "loss": 2.4802,
      "step": 1276
    },
    {
      "epoch": 0.39401419315026226,
      "grad_norm": 0.3932779133319855,
      "learning_rate": 3e-05,
      "loss": 2.7946,
      "step": 1277
    },
    {
      "epoch": 0.3943227398950941,
      "grad_norm": 0.3697027862071991,
      "learning_rate": 3e-05,
      "loss": 2.7385,
      "step": 1278
    },
    {
      "epoch": 0.39463128663992597,
      "grad_norm": 0.6150335669517517,
      "learning_rate": 3e-05,
      "loss": 2.8217,
      "step": 1279
    },
    {
      "epoch": 0.3949398333847578,
      "grad_norm": 0.3720497190952301,
      "learning_rate": 3e-05,
      "loss": 2.7876,
      "step": 1280
    },
    {
      "epoch": 0.3952483801295896,
      "grad_norm": 0.2867538034915924,
      "learning_rate": 3e-05,
      "loss": 2.6693,
      "step": 1281
    },
    {
      "epoch": 0.39555692687442146,
      "grad_norm": 0.3619423806667328,
      "learning_rate": 3e-05,
      "loss": 2.4328,
      "step": 1282
    },
    {
      "epoch": 0.39586547361925334,
      "grad_norm": 0.37228068709373474,
      "learning_rate": 3e-05,
      "loss": 3.076,
      "step": 1283
    },
    {
      "epoch": 0.39617402036408517,
      "grad_norm": 0.39720866084098816,
      "learning_rate": 3e-05,
      "loss": 2.6513,
      "step": 1284
    },
    {
      "epoch": 0.396482567108917,
      "grad_norm": 0.3063104748725891,
      "learning_rate": 3e-05,
      "loss": 2.3469,
      "step": 1285
    },
    {
      "epoch": 0.3967911138537488,
      "grad_norm": 0.3493167459964752,
      "learning_rate": 3e-05,
      "loss": 2.9156,
      "step": 1286
    },
    {
      "epoch": 0.3970996605985807,
      "grad_norm": 0.4875669479370117,
      "learning_rate": 3e-05,
      "loss": 3.035,
      "step": 1287
    },
    {
      "epoch": 0.39740820734341253,
      "grad_norm": 0.5185248255729675,
      "learning_rate": 3e-05,
      "loss": 3.5732,
      "step": 1288
    },
    {
      "epoch": 0.39771675408824436,
      "grad_norm": 0.5274576544761658,
      "learning_rate": 3e-05,
      "loss": 2.708,
      "step": 1289
    },
    {
      "epoch": 0.3980253008330762,
      "grad_norm": 0.3633483648300171,
      "learning_rate": 3e-05,
      "loss": 2.4895,
      "step": 1290
    },
    {
      "epoch": 0.3983338475779081,
      "grad_norm": 0.283071905374527,
      "learning_rate": 3e-05,
      "loss": 2.1112,
      "step": 1291
    },
    {
      "epoch": 0.3986423943227399,
      "grad_norm": 0.31986409425735474,
      "learning_rate": 3e-05,
      "loss": 2.2668,
      "step": 1292
    },
    {
      "epoch": 0.39895094106757173,
      "grad_norm": 0.7418584227561951,
      "learning_rate": 3e-05,
      "loss": 3.4388,
      "step": 1293
    },
    {
      "epoch": 0.39925948781240356,
      "grad_norm": 0.4027244746685028,
      "learning_rate": 3e-05,
      "loss": 2.592,
      "step": 1294
    },
    {
      "epoch": 0.39956803455723544,
      "grad_norm": 0.29873213171958923,
      "learning_rate": 3e-05,
      "loss": 2.4954,
      "step": 1295
    },
    {
      "epoch": 0.39987658130206727,
      "grad_norm": 0.3049558401107788,
      "learning_rate": 3e-05,
      "loss": 2.4802,
      "step": 1296
    },
    {
      "epoch": 0.4001851280468991,
      "grad_norm": 0.4675412178039551,
      "learning_rate": 3e-05,
      "loss": 2.6097,
      "step": 1297
    },
    {
      "epoch": 0.4004936747917309,
      "grad_norm": 0.5763147473335266,
      "learning_rate": 3e-05,
      "loss": 2.6048,
      "step": 1298
    },
    {
      "epoch": 0.4008022215365628,
      "grad_norm": 0.45807382464408875,
      "learning_rate": 3e-05,
      "loss": 2.7465,
      "step": 1299
    },
    {
      "epoch": 0.40111076828139464,
      "grad_norm": 0.5493050813674927,
      "learning_rate": 3e-05,
      "loss": 2.9135,
      "step": 1300
    },
    {
      "epoch": 0.40141931502622646,
      "grad_norm": 0.610711395740509,
      "learning_rate": 3e-05,
      "loss": 2.7627,
      "step": 1301
    },
    {
      "epoch": 0.4017278617710583,
      "grad_norm": 0.31486013531684875,
      "learning_rate": 3e-05,
      "loss": 2.3012,
      "step": 1302
    },
    {
      "epoch": 0.4020364085158902,
      "grad_norm": 0.37602242827415466,
      "learning_rate": 3e-05,
      "loss": 2.7181,
      "step": 1303
    },
    {
      "epoch": 0.402344955260722,
      "grad_norm": 0.33764341473579407,
      "learning_rate": 3e-05,
      "loss": 2.4903,
      "step": 1304
    },
    {
      "epoch": 0.40265350200555383,
      "grad_norm": 0.35062283277511597,
      "learning_rate": 3e-05,
      "loss": 2.5738,
      "step": 1305
    },
    {
      "epoch": 0.40296204875038566,
      "grad_norm": 0.3441402316093445,
      "learning_rate": 3e-05,
      "loss": 2.3253,
      "step": 1306
    },
    {
      "epoch": 0.40327059549521754,
      "grad_norm": 0.5749024748802185,
      "learning_rate": 3e-05,
      "loss": 2.9657,
      "step": 1307
    },
    {
      "epoch": 0.40357914224004937,
      "grad_norm": 0.30308666825294495,
      "learning_rate": 3e-05,
      "loss": 2.2349,
      "step": 1308
    },
    {
      "epoch": 0.4038876889848812,
      "grad_norm": 0.40580713748931885,
      "learning_rate": 3e-05,
      "loss": 2.5753,
      "step": 1309
    },
    {
      "epoch": 0.404196235729713,
      "grad_norm": 0.419880211353302,
      "learning_rate": 3e-05,
      "loss": 2.5283,
      "step": 1310
    },
    {
      "epoch": 0.4045047824745449,
      "grad_norm": 0.42699503898620605,
      "learning_rate": 3e-05,
      "loss": 2.6258,
      "step": 1311
    },
    {
      "epoch": 0.40481332921937674,
      "grad_norm": 0.3106153607368469,
      "learning_rate": 3e-05,
      "loss": 2.1418,
      "step": 1312
    },
    {
      "epoch": 0.40512187596420857,
      "grad_norm": 0.32699504494667053,
      "learning_rate": 3e-05,
      "loss": 2.6335,
      "step": 1313
    },
    {
      "epoch": 0.4054304227090404,
      "grad_norm": 0.4115197956562042,
      "learning_rate": 3e-05,
      "loss": 2.7467,
      "step": 1314
    },
    {
      "epoch": 0.4057389694538723,
      "grad_norm": 0.4634549021720886,
      "learning_rate": 3e-05,
      "loss": 2.6238,
      "step": 1315
    },
    {
      "epoch": 0.4060475161987041,
      "grad_norm": 0.396408349275589,
      "learning_rate": 3e-05,
      "loss": 2.8943,
      "step": 1316
    },
    {
      "epoch": 0.40635606294353593,
      "grad_norm": 0.253656804561615,
      "learning_rate": 3e-05,
      "loss": 2.0561,
      "step": 1317
    },
    {
      "epoch": 0.40666460968836776,
      "grad_norm": 0.3059282600879669,
      "learning_rate": 3e-05,
      "loss": 2.6393,
      "step": 1318
    },
    {
      "epoch": 0.40697315643319965,
      "grad_norm": 0.3378694951534271,
      "learning_rate": 3e-05,
      "loss": 2.6338,
      "step": 1319
    },
    {
      "epoch": 0.4072817031780315,
      "grad_norm": 0.5034033060073853,
      "learning_rate": 3e-05,
      "loss": 3.1591,
      "step": 1320
    },
    {
      "epoch": 0.4075902499228633,
      "grad_norm": 0.35838401317596436,
      "learning_rate": 3e-05,
      "loss": 2.3609,
      "step": 1321
    },
    {
      "epoch": 0.40789879666769513,
      "grad_norm": 0.4967813789844513,
      "learning_rate": 3e-05,
      "loss": 3.0929,
      "step": 1322
    },
    {
      "epoch": 0.408207343412527,
      "grad_norm": 0.3491065800189972,
      "learning_rate": 3e-05,
      "loss": 2.9074,
      "step": 1323
    },
    {
      "epoch": 0.40851589015735884,
      "grad_norm": 0.37351828813552856,
      "learning_rate": 3e-05,
      "loss": 2.6009,
      "step": 1324
    },
    {
      "epoch": 0.40882443690219067,
      "grad_norm": 0.4349062442779541,
      "learning_rate": 3e-05,
      "loss": 3.0212,
      "step": 1325
    },
    {
      "epoch": 0.4091329836470225,
      "grad_norm": 0.3631637692451477,
      "learning_rate": 3e-05,
      "loss": 2.7591,
      "step": 1326
    },
    {
      "epoch": 0.4094415303918544,
      "grad_norm": 0.392930805683136,
      "learning_rate": 3e-05,
      "loss": 2.7822,
      "step": 1327
    },
    {
      "epoch": 0.4097500771366862,
      "grad_norm": 0.571189820766449,
      "learning_rate": 3e-05,
      "loss": 3.0766,
      "step": 1328
    },
    {
      "epoch": 0.41005862388151804,
      "grad_norm": 0.30686402320861816,
      "learning_rate": 3e-05,
      "loss": 2.684,
      "step": 1329
    },
    {
      "epoch": 0.4103671706263499,
      "grad_norm": 0.3239377737045288,
      "learning_rate": 3e-05,
      "loss": 2.7809,
      "step": 1330
    },
    {
      "epoch": 0.41067571737118175,
      "grad_norm": 0.35130640864372253,
      "learning_rate": 3e-05,
      "loss": 2.452,
      "step": 1331
    },
    {
      "epoch": 0.4109842641160136,
      "grad_norm": 0.3733953535556793,
      "learning_rate": 3e-05,
      "loss": 3.0438,
      "step": 1332
    },
    {
      "epoch": 0.4112928108608454,
      "grad_norm": 0.2904767692089081,
      "learning_rate": 3e-05,
      "loss": 2.5109,
      "step": 1333
    },
    {
      "epoch": 0.4116013576056773,
      "grad_norm": 0.3014267086982727,
      "learning_rate": 3e-05,
      "loss": 2.2298,
      "step": 1334
    },
    {
      "epoch": 0.4119099043505091,
      "grad_norm": 0.40220725536346436,
      "learning_rate": 3e-05,
      "loss": 2.7179,
      "step": 1335
    },
    {
      "epoch": 0.41221845109534094,
      "grad_norm": 0.3351367115974426,
      "learning_rate": 3e-05,
      "loss": 2.2936,
      "step": 1336
    },
    {
      "epoch": 0.41252699784017277,
      "grad_norm": 0.3121192753314972,
      "learning_rate": 3e-05,
      "loss": 2.265,
      "step": 1337
    },
    {
      "epoch": 0.41283554458500465,
      "grad_norm": 0.4137488305568695,
      "learning_rate": 3e-05,
      "loss": 2.7979,
      "step": 1338
    },
    {
      "epoch": 0.4131440913298365,
      "grad_norm": 0.4146516025066376,
      "learning_rate": 3e-05,
      "loss": 3.174,
      "step": 1339
    },
    {
      "epoch": 0.4134526380746683,
      "grad_norm": 0.39826127886772156,
      "learning_rate": 3e-05,
      "loss": 2.5499,
      "step": 1340
    },
    {
      "epoch": 0.41376118481950014,
      "grad_norm": 0.46398836374282837,
      "learning_rate": 3e-05,
      "loss": 2.6571,
      "step": 1341
    },
    {
      "epoch": 0.414069731564332,
      "grad_norm": 0.42536866664886475,
      "learning_rate": 3e-05,
      "loss": 2.6297,
      "step": 1342
    },
    {
      "epoch": 0.41437827830916385,
      "grad_norm": 0.3115003705024719,
      "learning_rate": 3e-05,
      "loss": 2.4774,
      "step": 1343
    },
    {
      "epoch": 0.4146868250539957,
      "grad_norm": 0.3130146563053131,
      "learning_rate": 3e-05,
      "loss": 2.5763,
      "step": 1344
    },
    {
      "epoch": 0.4149953717988275,
      "grad_norm": 0.4441472291946411,
      "learning_rate": 3e-05,
      "loss": 2.9602,
      "step": 1345
    },
    {
      "epoch": 0.4153039185436594,
      "grad_norm": 0.8076886534690857,
      "learning_rate": 3e-05,
      "loss": 2.6896,
      "step": 1346
    },
    {
      "epoch": 0.4156124652884912,
      "grad_norm": 0.7255078554153442,
      "learning_rate": 3e-05,
      "loss": 3.3754,
      "step": 1347
    },
    {
      "epoch": 0.41592101203332305,
      "grad_norm": 0.3382607698440552,
      "learning_rate": 3e-05,
      "loss": 2.5589,
      "step": 1348
    },
    {
      "epoch": 0.4162295587781549,
      "grad_norm": 0.46596959233283997,
      "learning_rate": 3e-05,
      "loss": 2.6977,
      "step": 1349
    },
    {
      "epoch": 0.41653810552298676,
      "grad_norm": 0.6461755037307739,
      "learning_rate": 3e-05,
      "loss": 3.2361,
      "step": 1350
    },
    {
      "epoch": 0.4168466522678186,
      "grad_norm": 0.3971562087535858,
      "learning_rate": 3e-05,
      "loss": 2.4609,
      "step": 1351
    },
    {
      "epoch": 0.4171551990126504,
      "grad_norm": 0.4004113972187042,
      "learning_rate": 3e-05,
      "loss": 2.763,
      "step": 1352
    },
    {
      "epoch": 0.41746374575748224,
      "grad_norm": 0.37319812178611755,
      "learning_rate": 3e-05,
      "loss": 2.8121,
      "step": 1353
    },
    {
      "epoch": 0.4177722925023141,
      "grad_norm": 0.5117070078849792,
      "learning_rate": 3e-05,
      "loss": 2.9268,
      "step": 1354
    },
    {
      "epoch": 0.41808083924714595,
      "grad_norm": 0.5662671327590942,
      "learning_rate": 3e-05,
      "loss": 3.2365,
      "step": 1355
    },
    {
      "epoch": 0.4183893859919778,
      "grad_norm": 0.3150057792663574,
      "learning_rate": 3e-05,
      "loss": 2.7096,
      "step": 1356
    },
    {
      "epoch": 0.4186979327368096,
      "grad_norm": 0.3373428285121918,
      "learning_rate": 3e-05,
      "loss": 2.9863,
      "step": 1357
    },
    {
      "epoch": 0.4190064794816415,
      "grad_norm": 0.3957524001598358,
      "learning_rate": 3e-05,
      "loss": 2.6506,
      "step": 1358
    },
    {
      "epoch": 0.4193150262264733,
      "grad_norm": 0.3644879460334778,
      "learning_rate": 3e-05,
      "loss": 2.4042,
      "step": 1359
    },
    {
      "epoch": 0.41962357297130515,
      "grad_norm": 0.44521406292915344,
      "learning_rate": 3e-05,
      "loss": 2.5451,
      "step": 1360
    },
    {
      "epoch": 0.419932119716137,
      "grad_norm": 0.3452889621257782,
      "learning_rate": 3e-05,
      "loss": 2.4424,
      "step": 1361
    },
    {
      "epoch": 0.42024066646096886,
      "grad_norm": 0.4142851233482361,
      "learning_rate": 3e-05,
      "loss": 2.7783,
      "step": 1362
    },
    {
      "epoch": 0.4205492132058007,
      "grad_norm": 0.5085824728012085,
      "learning_rate": 3e-05,
      "loss": 2.8652,
      "step": 1363
    },
    {
      "epoch": 0.4208577599506325,
      "grad_norm": 0.2973617911338806,
      "learning_rate": 3e-05,
      "loss": 2.4887,
      "step": 1364
    },
    {
      "epoch": 0.42116630669546434,
      "grad_norm": 0.48054924607276917,
      "learning_rate": 3e-05,
      "loss": 3.1104,
      "step": 1365
    },
    {
      "epoch": 0.4214748534402962,
      "grad_norm": 0.3088330030441284,
      "learning_rate": 3e-05,
      "loss": 2.8398,
      "step": 1366
    },
    {
      "epoch": 0.42178340018512805,
      "grad_norm": 0.34353551268577576,
      "learning_rate": 3e-05,
      "loss": 2.6529,
      "step": 1367
    },
    {
      "epoch": 0.4220919469299599,
      "grad_norm": 0.3366641700267792,
      "learning_rate": 3e-05,
      "loss": 2.7679,
      "step": 1368
    },
    {
      "epoch": 0.4224004936747917,
      "grad_norm": 0.3791579604148865,
      "learning_rate": 3e-05,
      "loss": 2.8372,
      "step": 1369
    },
    {
      "epoch": 0.4227090404196236,
      "grad_norm": 0.33375903964042664,
      "learning_rate": 3e-05,
      "loss": 2.7064,
      "step": 1370
    },
    {
      "epoch": 0.4230175871644554,
      "grad_norm": 0.3554198741912842,
      "learning_rate": 3e-05,
      "loss": 2.953,
      "step": 1371
    },
    {
      "epoch": 0.42332613390928725,
      "grad_norm": 0.3274098336696625,
      "learning_rate": 3e-05,
      "loss": 2.4668,
      "step": 1372
    },
    {
      "epoch": 0.4236346806541191,
      "grad_norm": 0.4714450538158417,
      "learning_rate": 3e-05,
      "loss": 3.1614,
      "step": 1373
    },
    {
      "epoch": 0.42394322739895096,
      "grad_norm": 0.4060400724411011,
      "learning_rate": 3e-05,
      "loss": 3.0015,
      "step": 1374
    },
    {
      "epoch": 0.4242517741437828,
      "grad_norm": 0.32625189423561096,
      "learning_rate": 3e-05,
      "loss": 2.4997,
      "step": 1375
    },
    {
      "epoch": 0.4245603208886146,
      "grad_norm": 0.2833174765110016,
      "learning_rate": 3e-05,
      "loss": 2.207,
      "step": 1376
    },
    {
      "epoch": 0.42486886763344645,
      "grad_norm": 0.36846381425857544,
      "learning_rate": 3e-05,
      "loss": 2.5808,
      "step": 1377
    },
    {
      "epoch": 0.42517741437827833,
      "grad_norm": 0.3334082365036011,
      "learning_rate": 3e-05,
      "loss": 2.643,
      "step": 1378
    },
    {
      "epoch": 0.42548596112311016,
      "grad_norm": 0.3491818606853485,
      "learning_rate": 3e-05,
      "loss": 2.8605,
      "step": 1379
    },
    {
      "epoch": 0.425794507867942,
      "grad_norm": 0.4421471953392029,
      "learning_rate": 3e-05,
      "loss": 3.3898,
      "step": 1380
    },
    {
      "epoch": 0.4261030546127738,
      "grad_norm": 0.2874299883842468,
      "learning_rate": 3e-05,
      "loss": 2.4055,
      "step": 1381
    },
    {
      "epoch": 0.4264116013576057,
      "grad_norm": 0.2746589779853821,
      "learning_rate": 3e-05,
      "loss": 2.5162,
      "step": 1382
    },
    {
      "epoch": 0.4267201481024375,
      "grad_norm": 0.33837470412254333,
      "learning_rate": 3e-05,
      "loss": 2.3699,
      "step": 1383
    },
    {
      "epoch": 0.42702869484726935,
      "grad_norm": 0.40471774339675903,
      "learning_rate": 3e-05,
      "loss": 2.9301,
      "step": 1384
    },
    {
      "epoch": 0.4273372415921012,
      "grad_norm": 0.3374662399291992,
      "learning_rate": 3e-05,
      "loss": 2.4965,
      "step": 1385
    },
    {
      "epoch": 0.42764578833693306,
      "grad_norm": 0.42680230736732483,
      "learning_rate": 3e-05,
      "loss": 3.1878,
      "step": 1386
    },
    {
      "epoch": 0.4279543350817649,
      "grad_norm": 0.31313976645469666,
      "learning_rate": 3e-05,
      "loss": 2.0341,
      "step": 1387
    },
    {
      "epoch": 0.4282628818265967,
      "grad_norm": 0.25408491492271423,
      "learning_rate": 3e-05,
      "loss": 2.4309,
      "step": 1388
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 0.31811076402664185,
      "learning_rate": 3e-05,
      "loss": 2.6167,
      "step": 1389
    },
    {
      "epoch": 0.42887997531626043,
      "grad_norm": 0.3181273341178894,
      "learning_rate": 3e-05,
      "loss": 2.4618,
      "step": 1390
    },
    {
      "epoch": 0.42918852206109226,
      "grad_norm": 0.3027171790599823,
      "learning_rate": 3e-05,
      "loss": 2.6972,
      "step": 1391
    },
    {
      "epoch": 0.4294970688059241,
      "grad_norm": 0.3148479759693146,
      "learning_rate": 3e-05,
      "loss": 2.5805,
      "step": 1392
    },
    {
      "epoch": 0.4298056155507559,
      "grad_norm": 0.3551836311817169,
      "learning_rate": 3e-05,
      "loss": 2.5305,
      "step": 1393
    },
    {
      "epoch": 0.4301141622955878,
      "grad_norm": 0.3691128194332123,
      "learning_rate": 3e-05,
      "loss": 2.5768,
      "step": 1394
    },
    {
      "epoch": 0.4304227090404196,
      "grad_norm": 0.30554333329200745,
      "learning_rate": 3e-05,
      "loss": 2.2353,
      "step": 1395
    },
    {
      "epoch": 0.43073125578525145,
      "grad_norm": 0.390206515789032,
      "learning_rate": 3e-05,
      "loss": 2.8924,
      "step": 1396
    },
    {
      "epoch": 0.4310398025300833,
      "grad_norm": 0.3919602036476135,
      "learning_rate": 3e-05,
      "loss": 2.7581,
      "step": 1397
    },
    {
      "epoch": 0.43134834927491517,
      "grad_norm": 0.48594582080841064,
      "learning_rate": 3e-05,
      "loss": 2.9195,
      "step": 1398
    },
    {
      "epoch": 0.431656896019747,
      "grad_norm": 0.33339229226112366,
      "learning_rate": 3e-05,
      "loss": 2.703,
      "step": 1399
    },
    {
      "epoch": 0.4319654427645788,
      "grad_norm": 0.3424081802368164,
      "learning_rate": 3e-05,
      "loss": 2.3623,
      "step": 1400
    },
    {
      "epoch": 0.43227398950941065,
      "grad_norm": 0.37659206986427307,
      "learning_rate": 3e-05,
      "loss": 3.0833,
      "step": 1401
    },
    {
      "epoch": 0.43258253625424253,
      "grad_norm": 0.4342521131038666,
      "learning_rate": 3e-05,
      "loss": 2.5336,
      "step": 1402
    },
    {
      "epoch": 0.43289108299907436,
      "grad_norm": 0.3215969204902649,
      "learning_rate": 3e-05,
      "loss": 2.6314,
      "step": 1403
    },
    {
      "epoch": 0.4331996297439062,
      "grad_norm": 0.3317095637321472,
      "learning_rate": 3e-05,
      "loss": 2.2092,
      "step": 1404
    },
    {
      "epoch": 0.433508176488738,
      "grad_norm": 0.4227984845638275,
      "learning_rate": 3e-05,
      "loss": 2.9534,
      "step": 1405
    },
    {
      "epoch": 0.4338167232335699,
      "grad_norm": 0.38567742705345154,
      "learning_rate": 3e-05,
      "loss": 2.7021,
      "step": 1406
    },
    {
      "epoch": 0.43412526997840173,
      "grad_norm": 0.3301149606704712,
      "learning_rate": 3e-05,
      "loss": 2.6226,
      "step": 1407
    },
    {
      "epoch": 0.43443381672323356,
      "grad_norm": 0.3673669099807739,
      "learning_rate": 3e-05,
      "loss": 2.6008,
      "step": 1408
    },
    {
      "epoch": 0.4347423634680654,
      "grad_norm": 0.3407856523990631,
      "learning_rate": 3e-05,
      "loss": 2.3941,
      "step": 1409
    },
    {
      "epoch": 0.43505091021289727,
      "grad_norm": 0.43250688910484314,
      "learning_rate": 3e-05,
      "loss": 3.001,
      "step": 1410
    },
    {
      "epoch": 0.4353594569577291,
      "grad_norm": 0.3385581374168396,
      "learning_rate": 3e-05,
      "loss": 2.9602,
      "step": 1411
    },
    {
      "epoch": 0.4356680037025609,
      "grad_norm": 0.3739365041255951,
      "learning_rate": 3e-05,
      "loss": 2.8393,
      "step": 1412
    },
    {
      "epoch": 0.4359765504473928,
      "grad_norm": 0.5004026889801025,
      "learning_rate": 3e-05,
      "loss": 2.8947,
      "step": 1413
    },
    {
      "epoch": 0.43628509719222464,
      "grad_norm": 0.3384886085987091,
      "learning_rate": 3e-05,
      "loss": 2.4085,
      "step": 1414
    },
    {
      "epoch": 0.43659364393705646,
      "grad_norm": 0.511809229850769,
      "learning_rate": 3e-05,
      "loss": 2.864,
      "step": 1415
    },
    {
      "epoch": 0.4369021906818883,
      "grad_norm": 0.37064921855926514,
      "learning_rate": 3e-05,
      "loss": 2.8768,
      "step": 1416
    },
    {
      "epoch": 0.4372107374267202,
      "grad_norm": 0.3196764290332794,
      "learning_rate": 3e-05,
      "loss": 2.4019,
      "step": 1417
    },
    {
      "epoch": 0.437519284171552,
      "grad_norm": 0.3673863708972931,
      "learning_rate": 3e-05,
      "loss": 2.6811,
      "step": 1418
    },
    {
      "epoch": 0.43782783091638383,
      "grad_norm": 0.3803464472293854,
      "learning_rate": 3e-05,
      "loss": 2.3291,
      "step": 1419
    },
    {
      "epoch": 0.43813637766121566,
      "grad_norm": 0.5707358717918396,
      "learning_rate": 3e-05,
      "loss": 3.048,
      "step": 1420
    },
    {
      "epoch": 0.43844492440604754,
      "grad_norm": 0.3424266278743744,
      "learning_rate": 3e-05,
      "loss": 2.3497,
      "step": 1421
    },
    {
      "epoch": 0.43875347115087937,
      "grad_norm": 0.5494610071182251,
      "learning_rate": 3e-05,
      "loss": 3.0777,
      "step": 1422
    },
    {
      "epoch": 0.4390620178957112,
      "grad_norm": 0.5954529643058777,
      "learning_rate": 3e-05,
      "loss": 2.7765,
      "step": 1423
    },
    {
      "epoch": 0.439370564640543,
      "grad_norm": 0.33620685338974,
      "learning_rate": 3e-05,
      "loss": 2.3952,
      "step": 1424
    },
    {
      "epoch": 0.4396791113853749,
      "grad_norm": 0.3967253565788269,
      "learning_rate": 3e-05,
      "loss": 2.8079,
      "step": 1425
    },
    {
      "epoch": 0.43998765813020674,
      "grad_norm": 0.2961389124393463,
      "learning_rate": 3e-05,
      "loss": 2.3434,
      "step": 1426
    },
    {
      "epoch": 0.44029620487503857,
      "grad_norm": 0.41013649106025696,
      "learning_rate": 3e-05,
      "loss": 2.79,
      "step": 1427
    },
    {
      "epoch": 0.4406047516198704,
      "grad_norm": 0.3336520791053772,
      "learning_rate": 3e-05,
      "loss": 2.6722,
      "step": 1428
    },
    {
      "epoch": 0.4409132983647023,
      "grad_norm": 0.4669612646102905,
      "learning_rate": 3e-05,
      "loss": 2.6199,
      "step": 1429
    },
    {
      "epoch": 0.4412218451095341,
      "grad_norm": 0.3168468475341797,
      "learning_rate": 3e-05,
      "loss": 2.499,
      "step": 1430
    },
    {
      "epoch": 0.44153039185436593,
      "grad_norm": 0.3851400911808014,
      "learning_rate": 3e-05,
      "loss": 2.5927,
      "step": 1431
    },
    {
      "epoch": 0.44183893859919776,
      "grad_norm": 0.31471362709999084,
      "learning_rate": 3e-05,
      "loss": 2.5986,
      "step": 1432
    },
    {
      "epoch": 0.44214748534402964,
      "grad_norm": 0.405691534280777,
      "learning_rate": 3e-05,
      "loss": 2.8578,
      "step": 1433
    },
    {
      "epoch": 0.4424560320888615,
      "grad_norm": 0.3536216914653778,
      "learning_rate": 3e-05,
      "loss": 2.6194,
      "step": 1434
    },
    {
      "epoch": 0.4427645788336933,
      "grad_norm": 0.3257675766944885,
      "learning_rate": 3e-05,
      "loss": 2.5817,
      "step": 1435
    },
    {
      "epoch": 0.44307312557852513,
      "grad_norm": 0.3345182538032532,
      "learning_rate": 3e-05,
      "loss": 2.423,
      "step": 1436
    },
    {
      "epoch": 0.443381672323357,
      "grad_norm": 0.47484976053237915,
      "learning_rate": 3e-05,
      "loss": 2.8201,
      "step": 1437
    },
    {
      "epoch": 0.44369021906818884,
      "grad_norm": 0.4050092399120331,
      "learning_rate": 3e-05,
      "loss": 2.4457,
      "step": 1438
    },
    {
      "epoch": 0.44399876581302067,
      "grad_norm": 0.3550967574119568,
      "learning_rate": 3e-05,
      "loss": 2.5056,
      "step": 1439
    },
    {
      "epoch": 0.4443073125578525,
      "grad_norm": 0.34624183177948,
      "learning_rate": 3e-05,
      "loss": 2.5355,
      "step": 1440
    },
    {
      "epoch": 0.4446158593026844,
      "grad_norm": 0.4038742482662201,
      "learning_rate": 3e-05,
      "loss": 2.7499,
      "step": 1441
    },
    {
      "epoch": 0.4449244060475162,
      "grad_norm": 0.3393418788909912,
      "learning_rate": 3e-05,
      "loss": 2.7613,
      "step": 1442
    },
    {
      "epoch": 0.44523295279234804,
      "grad_norm": 0.4069559872150421,
      "learning_rate": 3e-05,
      "loss": 2.7251,
      "step": 1443
    },
    {
      "epoch": 0.44554149953717986,
      "grad_norm": 0.42860448360443115,
      "learning_rate": 3e-05,
      "loss": 2.7757,
      "step": 1444
    },
    {
      "epoch": 0.44585004628201175,
      "grad_norm": 0.3287833034992218,
      "learning_rate": 3e-05,
      "loss": 2.3999,
      "step": 1445
    },
    {
      "epoch": 0.4461585930268436,
      "grad_norm": 0.359164297580719,
      "learning_rate": 3e-05,
      "loss": 2.7029,
      "step": 1446
    },
    {
      "epoch": 0.4464671397716754,
      "grad_norm": 0.4843924641609192,
      "learning_rate": 3e-05,
      "loss": 3.0231,
      "step": 1447
    },
    {
      "epoch": 0.44677568651650723,
      "grad_norm": 0.4894542396068573,
      "learning_rate": 3e-05,
      "loss": 3.0614,
      "step": 1448
    },
    {
      "epoch": 0.4470842332613391,
      "grad_norm": 0.33690640330314636,
      "learning_rate": 3e-05,
      "loss": 2.6946,
      "step": 1449
    },
    {
      "epoch": 0.44739278000617094,
      "grad_norm": 0.2866448163986206,
      "learning_rate": 3e-05,
      "loss": 2.3611,
      "step": 1450
    },
    {
      "epoch": 0.44770132675100277,
      "grad_norm": 0.7917083501815796,
      "learning_rate": 3e-05,
      "loss": 3.5579,
      "step": 1451
    },
    {
      "epoch": 0.4480098734958346,
      "grad_norm": 0.2805081903934479,
      "learning_rate": 3e-05,
      "loss": 2.1076,
      "step": 1452
    },
    {
      "epoch": 0.4483184202406665,
      "grad_norm": 0.3295937180519104,
      "learning_rate": 3e-05,
      "loss": 2.8668,
      "step": 1453
    },
    {
      "epoch": 0.4486269669854983,
      "grad_norm": 0.30099961161613464,
      "learning_rate": 3e-05,
      "loss": 2.3013,
      "step": 1454
    },
    {
      "epoch": 0.44893551373033014,
      "grad_norm": 0.543934166431427,
      "learning_rate": 3e-05,
      "loss": 2.9462,
      "step": 1455
    },
    {
      "epoch": 0.44924406047516197,
      "grad_norm": 0.36943310499191284,
      "learning_rate": 3e-05,
      "loss": 2.93,
      "step": 1456
    },
    {
      "epoch": 0.44955260721999385,
      "grad_norm": 0.5213918685913086,
      "learning_rate": 3e-05,
      "loss": 2.5094,
      "step": 1457
    },
    {
      "epoch": 0.4498611539648257,
      "grad_norm": 0.2893174886703491,
      "learning_rate": 3e-05,
      "loss": 2.4578,
      "step": 1458
    },
    {
      "epoch": 0.4501697007096575,
      "grad_norm": 0.36389556527137756,
      "learning_rate": 3e-05,
      "loss": 2.8613,
      "step": 1459
    },
    {
      "epoch": 0.45047824745448933,
      "grad_norm": 0.4273573160171509,
      "learning_rate": 3e-05,
      "loss": 2.7277,
      "step": 1460
    },
    {
      "epoch": 0.4507867941993212,
      "grad_norm": 0.5575438141822815,
      "learning_rate": 3e-05,
      "loss": 3.4139,
      "step": 1461
    },
    {
      "epoch": 0.45109534094415304,
      "grad_norm": 0.4661589562892914,
      "learning_rate": 3e-05,
      "loss": 2.9162,
      "step": 1462
    },
    {
      "epoch": 0.4514038876889849,
      "grad_norm": 0.350808709859848,
      "learning_rate": 3e-05,
      "loss": 2.8438,
      "step": 1463
    },
    {
      "epoch": 0.4517124344338167,
      "grad_norm": 0.26493504643440247,
      "learning_rate": 3e-05,
      "loss": 2.3362,
      "step": 1464
    },
    {
      "epoch": 0.4520209811786486,
      "grad_norm": 0.30495163798332214,
      "learning_rate": 3e-05,
      "loss": 2.6332,
      "step": 1465
    },
    {
      "epoch": 0.4523295279234804,
      "grad_norm": 0.32672178745269775,
      "learning_rate": 3e-05,
      "loss": 2.4577,
      "step": 1466
    },
    {
      "epoch": 0.45263807466831224,
      "grad_norm": 0.29373109340667725,
      "learning_rate": 3e-05,
      "loss": 2.3206,
      "step": 1467
    },
    {
      "epoch": 0.45294662141314407,
      "grad_norm": 0.47060441970825195,
      "learning_rate": 3e-05,
      "loss": 2.6481,
      "step": 1468
    },
    {
      "epoch": 0.45325516815797595,
      "grad_norm": 0.4044017195701599,
      "learning_rate": 3e-05,
      "loss": 2.888,
      "step": 1469
    },
    {
      "epoch": 0.4535637149028078,
      "grad_norm": 0.5061796307563782,
      "learning_rate": 3e-05,
      "loss": 2.5166,
      "step": 1470
    },
    {
      "epoch": 0.4538722616476396,
      "grad_norm": 0.35933852195739746,
      "learning_rate": 3e-05,
      "loss": 2.6528,
      "step": 1471
    },
    {
      "epoch": 0.45418080839247144,
      "grad_norm": 0.36869826912879944,
      "learning_rate": 3e-05,
      "loss": 2.8693,
      "step": 1472
    },
    {
      "epoch": 0.4544893551373033,
      "grad_norm": 0.3967495560646057,
      "learning_rate": 3e-05,
      "loss": 2.6779,
      "step": 1473
    },
    {
      "epoch": 0.45479790188213515,
      "grad_norm": 0.4149707853794098,
      "learning_rate": 3e-05,
      "loss": 2.8755,
      "step": 1474
    },
    {
      "epoch": 0.455106448626967,
      "grad_norm": 0.39458996057510376,
      "learning_rate": 3e-05,
      "loss": 2.6415,
      "step": 1475
    },
    {
      "epoch": 0.4554149953717988,
      "grad_norm": 0.3556062877178192,
      "learning_rate": 3e-05,
      "loss": 2.7043,
      "step": 1476
    },
    {
      "epoch": 0.4557235421166307,
      "grad_norm": 0.5347737073898315,
      "learning_rate": 3e-05,
      "loss": 3.1037,
      "step": 1477
    },
    {
      "epoch": 0.4560320888614625,
      "grad_norm": 0.3868623673915863,
      "learning_rate": 3e-05,
      "loss": 2.9386,
      "step": 1478
    },
    {
      "epoch": 0.45634063560629434,
      "grad_norm": 0.3561474680900574,
      "learning_rate": 3e-05,
      "loss": 2.9683,
      "step": 1479
    },
    {
      "epoch": 0.45664918235112617,
      "grad_norm": 0.4054335951805115,
      "learning_rate": 3e-05,
      "loss": 3.0608,
      "step": 1480
    },
    {
      "epoch": 0.45695772909595805,
      "grad_norm": 0.3604312241077423,
      "learning_rate": 3e-05,
      "loss": 2.858,
      "step": 1481
    },
    {
      "epoch": 0.4572662758407899,
      "grad_norm": 0.41026541590690613,
      "learning_rate": 3e-05,
      "loss": 2.8464,
      "step": 1482
    },
    {
      "epoch": 0.4575748225856217,
      "grad_norm": 0.39170393347740173,
      "learning_rate": 3e-05,
      "loss": 2.5438,
      "step": 1483
    },
    {
      "epoch": 0.45788336933045354,
      "grad_norm": 0.3100852072238922,
      "learning_rate": 3e-05,
      "loss": 2.4244,
      "step": 1484
    },
    {
      "epoch": 0.4581919160752854,
      "grad_norm": 0.3728077709674835,
      "learning_rate": 3e-05,
      "loss": 2.9126,
      "step": 1485
    },
    {
      "epoch": 0.45850046282011725,
      "grad_norm": 0.32681816816329956,
      "learning_rate": 3e-05,
      "loss": 2.7418,
      "step": 1486
    },
    {
      "epoch": 0.4588090095649491,
      "grad_norm": 0.31139644980430603,
      "learning_rate": 3e-05,
      "loss": 2.587,
      "step": 1487
    },
    {
      "epoch": 0.4591175563097809,
      "grad_norm": 0.3643485903739929,
      "learning_rate": 3e-05,
      "loss": 2.3148,
      "step": 1488
    },
    {
      "epoch": 0.4594261030546128,
      "grad_norm": 0.37991657853126526,
      "learning_rate": 3e-05,
      "loss": 2.6316,
      "step": 1489
    },
    {
      "epoch": 0.4597346497994446,
      "grad_norm": 0.33139222860336304,
      "learning_rate": 3e-05,
      "loss": 2.577,
      "step": 1490
    },
    {
      "epoch": 0.46004319654427644,
      "grad_norm": 0.39415374398231506,
      "learning_rate": 3e-05,
      "loss": 2.6474,
      "step": 1491
    },
    {
      "epoch": 0.4603517432891083,
      "grad_norm": 0.43900495767593384,
      "learning_rate": 3e-05,
      "loss": 2.9312,
      "step": 1492
    },
    {
      "epoch": 0.46066029003394016,
      "grad_norm": 0.4236634075641632,
      "learning_rate": 3e-05,
      "loss": 2.4871,
      "step": 1493
    },
    {
      "epoch": 0.460968836778772,
      "grad_norm": 0.37256962060928345,
      "learning_rate": 3e-05,
      "loss": 2.6436,
      "step": 1494
    },
    {
      "epoch": 0.4612773835236038,
      "grad_norm": 0.3243587613105774,
      "learning_rate": 3e-05,
      "loss": 2.6336,
      "step": 1495
    },
    {
      "epoch": 0.4615859302684357,
      "grad_norm": 0.40223172307014465,
      "learning_rate": 3e-05,
      "loss": 2.4542,
      "step": 1496
    },
    {
      "epoch": 0.4618944770132675,
      "grad_norm": 0.24286819994449615,
      "learning_rate": 3e-05,
      "loss": 1.953,
      "step": 1497
    },
    {
      "epoch": 0.46220302375809935,
      "grad_norm": 0.3773837685585022,
      "learning_rate": 3e-05,
      "loss": 2.7311,
      "step": 1498
    },
    {
      "epoch": 0.4625115705029312,
      "grad_norm": 0.43442192673683167,
      "learning_rate": 3e-05,
      "loss": 2.3662,
      "step": 1499
    },
    {
      "epoch": 0.46282011724776306,
      "grad_norm": 0.35620713233947754,
      "learning_rate": 3e-05,
      "loss": 2.562,
      "step": 1500
    },
    {
      "epoch": 0.4631286639925949,
      "grad_norm": 0.3047196567058563,
      "learning_rate": 3e-05,
      "loss": 2.4955,
      "step": 1501
    },
    {
      "epoch": 0.4634372107374267,
      "grad_norm": 0.31453654170036316,
      "learning_rate": 3e-05,
      "loss": 2.3444,
      "step": 1502
    },
    {
      "epoch": 0.46374575748225855,
      "grad_norm": 0.39251789450645447,
      "learning_rate": 3e-05,
      "loss": 2.405,
      "step": 1503
    },
    {
      "epoch": 0.46405430422709043,
      "grad_norm": 0.3584946393966675,
      "learning_rate": 3e-05,
      "loss": 2.5729,
      "step": 1504
    },
    {
      "epoch": 0.46436285097192226,
      "grad_norm": 0.3041662573814392,
      "learning_rate": 3e-05,
      "loss": 2.3803,
      "step": 1505
    },
    {
      "epoch": 0.4646713977167541,
      "grad_norm": 0.44750258326530457,
      "learning_rate": 3e-05,
      "loss": 3.1487,
      "step": 1506
    },
    {
      "epoch": 0.4649799444615859,
      "grad_norm": 0.5141081809997559,
      "learning_rate": 3e-05,
      "loss": 2.8217,
      "step": 1507
    },
    {
      "epoch": 0.4652884912064178,
      "grad_norm": 0.5313743352890015,
      "learning_rate": 3e-05,
      "loss": 3.2112,
      "step": 1508
    },
    {
      "epoch": 0.4655970379512496,
      "grad_norm": 0.2799358367919922,
      "learning_rate": 3e-05,
      "loss": 2.3401,
      "step": 1509
    },
    {
      "epoch": 0.46590558469608145,
      "grad_norm": 0.485971063375473,
      "learning_rate": 3e-05,
      "loss": 3.0464,
      "step": 1510
    },
    {
      "epoch": 0.4662141314409133,
      "grad_norm": 0.2679968774318695,
      "learning_rate": 3e-05,
      "loss": 2.3394,
      "step": 1511
    },
    {
      "epoch": 0.46652267818574517,
      "grad_norm": 0.37413671612739563,
      "learning_rate": 3e-05,
      "loss": 2.8974,
      "step": 1512
    },
    {
      "epoch": 0.466831224930577,
      "grad_norm": 0.39536064863204956,
      "learning_rate": 3e-05,
      "loss": 3.2332,
      "step": 1513
    },
    {
      "epoch": 0.4671397716754088,
      "grad_norm": 0.38000598549842834,
      "learning_rate": 3e-05,
      "loss": 2.7148,
      "step": 1514
    },
    {
      "epoch": 0.46744831842024065,
      "grad_norm": 0.48554322123527527,
      "learning_rate": 3e-05,
      "loss": 2.6587,
      "step": 1515
    },
    {
      "epoch": 0.46775686516507253,
      "grad_norm": 0.4193074405193329,
      "learning_rate": 3e-05,
      "loss": 2.9054,
      "step": 1516
    },
    {
      "epoch": 0.46806541190990436,
      "grad_norm": 0.29958027601242065,
      "learning_rate": 3e-05,
      "loss": 2.3647,
      "step": 1517
    },
    {
      "epoch": 0.4683739586547362,
      "grad_norm": 0.2912478744983673,
      "learning_rate": 3e-05,
      "loss": 2.3545,
      "step": 1518
    },
    {
      "epoch": 0.468682505399568,
      "grad_norm": 0.268606036901474,
      "learning_rate": 3e-05,
      "loss": 2.1819,
      "step": 1519
    },
    {
      "epoch": 0.4689910521443999,
      "grad_norm": 0.38793525099754333,
      "learning_rate": 3e-05,
      "loss": 2.8047,
      "step": 1520
    },
    {
      "epoch": 0.46929959888923173,
      "grad_norm": 0.33550551533699036,
      "learning_rate": 3e-05,
      "loss": 3.0669,
      "step": 1521
    },
    {
      "epoch": 0.46960814563406356,
      "grad_norm": 0.36648938059806824,
      "learning_rate": 3e-05,
      "loss": 2.513,
      "step": 1522
    },
    {
      "epoch": 0.4699166923788954,
      "grad_norm": 0.4586170017719269,
      "learning_rate": 3e-05,
      "loss": 2.8867,
      "step": 1523
    },
    {
      "epoch": 0.47022523912372727,
      "grad_norm": 0.3348916471004486,
      "learning_rate": 3e-05,
      "loss": 2.6242,
      "step": 1524
    },
    {
      "epoch": 0.4705337858685591,
      "grad_norm": 0.29232484102249146,
      "learning_rate": 3e-05,
      "loss": 2.3149,
      "step": 1525
    },
    {
      "epoch": 0.4708423326133909,
      "grad_norm": 0.3628198504447937,
      "learning_rate": 3e-05,
      "loss": 2.3215,
      "step": 1526
    },
    {
      "epoch": 0.47115087935822275,
      "grad_norm": 0.46447861194610596,
      "learning_rate": 3e-05,
      "loss": 2.948,
      "step": 1527
    },
    {
      "epoch": 0.47145942610305464,
      "grad_norm": 0.3290894627571106,
      "learning_rate": 3e-05,
      "loss": 2.8251,
      "step": 1528
    },
    {
      "epoch": 0.47176797284788646,
      "grad_norm": 0.36014366149902344,
      "learning_rate": 3e-05,
      "loss": 2.499,
      "step": 1529
    },
    {
      "epoch": 0.4720765195927183,
      "grad_norm": 0.29595625400543213,
      "learning_rate": 3e-05,
      "loss": 2.2951,
      "step": 1530
    },
    {
      "epoch": 0.4723850663375501,
      "grad_norm": 0.4257338345050812,
      "learning_rate": 3e-05,
      "loss": 3.1167,
      "step": 1531
    },
    {
      "epoch": 0.472693613082382,
      "grad_norm": 0.36370155215263367,
      "learning_rate": 3e-05,
      "loss": 2.7506,
      "step": 1532
    },
    {
      "epoch": 0.47300215982721383,
      "grad_norm": 0.2574162185192108,
      "learning_rate": 3e-05,
      "loss": 2.2529,
      "step": 1533
    },
    {
      "epoch": 0.47331070657204566,
      "grad_norm": 0.3576168417930603,
      "learning_rate": 3e-05,
      "loss": 2.6559,
      "step": 1534
    },
    {
      "epoch": 0.4736192533168775,
      "grad_norm": 0.5774997472763062,
      "learning_rate": 3e-05,
      "loss": 2.8159,
      "step": 1535
    },
    {
      "epoch": 0.47392780006170937,
      "grad_norm": 0.37460872530937195,
      "learning_rate": 3e-05,
      "loss": 3.3666,
      "step": 1536
    },
    {
      "epoch": 0.4742363468065412,
      "grad_norm": 0.36277443170547485,
      "learning_rate": 3e-05,
      "loss": 2.6648,
      "step": 1537
    },
    {
      "epoch": 0.474544893551373,
      "grad_norm": 0.41957375407218933,
      "learning_rate": 3e-05,
      "loss": 2.6658,
      "step": 1538
    },
    {
      "epoch": 0.47485344029620485,
      "grad_norm": 0.4152233898639679,
      "learning_rate": 3e-05,
      "loss": 2.6572,
      "step": 1539
    },
    {
      "epoch": 0.47516198704103674,
      "grad_norm": 0.2616024911403656,
      "learning_rate": 3e-05,
      "loss": 2.228,
      "step": 1540
    },
    {
      "epoch": 0.47547053378586857,
      "grad_norm": 0.31821849942207336,
      "learning_rate": 3e-05,
      "loss": 2.5554,
      "step": 1541
    },
    {
      "epoch": 0.4757790805307004,
      "grad_norm": 0.4284205734729767,
      "learning_rate": 3e-05,
      "loss": 2.8339,
      "step": 1542
    },
    {
      "epoch": 0.4760876272755322,
      "grad_norm": 0.4077942669391632,
      "learning_rate": 3e-05,
      "loss": 2.7667,
      "step": 1543
    },
    {
      "epoch": 0.4763961740203641,
      "grad_norm": 0.46257877349853516,
      "learning_rate": 3e-05,
      "loss": 2.9476,
      "step": 1544
    },
    {
      "epoch": 0.47670472076519593,
      "grad_norm": 0.37353515625,
      "learning_rate": 3e-05,
      "loss": 2.6627,
      "step": 1545
    },
    {
      "epoch": 0.47701326751002776,
      "grad_norm": 0.29298368096351624,
      "learning_rate": 3e-05,
      "loss": 2.3048,
      "step": 1546
    },
    {
      "epoch": 0.4773218142548596,
      "grad_norm": 0.46679583191871643,
      "learning_rate": 3e-05,
      "loss": 2.4399,
      "step": 1547
    },
    {
      "epoch": 0.4776303609996915,
      "grad_norm": 0.38306209444999695,
      "learning_rate": 3e-05,
      "loss": 2.4255,
      "step": 1548
    },
    {
      "epoch": 0.4779389077445233,
      "grad_norm": 0.2915458679199219,
      "learning_rate": 3e-05,
      "loss": 2.4377,
      "step": 1549
    },
    {
      "epoch": 0.47824745448935513,
      "grad_norm": 0.3321347236633301,
      "learning_rate": 3e-05,
      "loss": 2.1925,
      "step": 1550
    },
    {
      "epoch": 0.47855600123418696,
      "grad_norm": 0.5388374924659729,
      "learning_rate": 3e-05,
      "loss": 2.6837,
      "step": 1551
    },
    {
      "epoch": 0.47886454797901884,
      "grad_norm": 0.5425001978874207,
      "learning_rate": 3e-05,
      "loss": 2.747,
      "step": 1552
    },
    {
      "epoch": 0.47917309472385067,
      "grad_norm": 0.3192477524280548,
      "learning_rate": 3e-05,
      "loss": 2.6687,
      "step": 1553
    },
    {
      "epoch": 0.4794816414686825,
      "grad_norm": 0.542382001876831,
      "learning_rate": 3e-05,
      "loss": 3.0074,
      "step": 1554
    },
    {
      "epoch": 0.4797901882135143,
      "grad_norm": 0.5276587605476379,
      "learning_rate": 3e-05,
      "loss": 2.5688,
      "step": 1555
    },
    {
      "epoch": 0.4800987349583462,
      "grad_norm": 0.35362979769706726,
      "learning_rate": 3e-05,
      "loss": 2.2445,
      "step": 1556
    },
    {
      "epoch": 0.48040728170317804,
      "grad_norm": 0.309165894985199,
      "learning_rate": 3e-05,
      "loss": 2.7821,
      "step": 1557
    },
    {
      "epoch": 0.48071582844800986,
      "grad_norm": 0.335504412651062,
      "learning_rate": 3e-05,
      "loss": 2.4358,
      "step": 1558
    },
    {
      "epoch": 0.4810243751928417,
      "grad_norm": 0.6521844863891602,
      "learning_rate": 3e-05,
      "loss": 3.2967,
      "step": 1559
    },
    {
      "epoch": 0.4813329219376736,
      "grad_norm": 0.32530689239501953,
      "learning_rate": 3e-05,
      "loss": 2.4366,
      "step": 1560
    },
    {
      "epoch": 0.4816414686825054,
      "grad_norm": 0.27962008118629456,
      "learning_rate": 3e-05,
      "loss": 2.1322,
      "step": 1561
    },
    {
      "epoch": 0.48195001542733723,
      "grad_norm": 0.328016072511673,
      "learning_rate": 3e-05,
      "loss": 2.6175,
      "step": 1562
    },
    {
      "epoch": 0.48225856217216906,
      "grad_norm": 0.45414265990257263,
      "learning_rate": 3e-05,
      "loss": 2.5736,
      "step": 1563
    },
    {
      "epoch": 0.48256710891700094,
      "grad_norm": 0.4768297076225281,
      "learning_rate": 3e-05,
      "loss": 2.4566,
      "step": 1564
    },
    {
      "epoch": 0.48287565566183277,
      "grad_norm": 0.31878459453582764,
      "learning_rate": 3e-05,
      "loss": 2.3206,
      "step": 1565
    },
    {
      "epoch": 0.4831842024066646,
      "grad_norm": 0.34920403361320496,
      "learning_rate": 3e-05,
      "loss": 2.5046,
      "step": 1566
    },
    {
      "epoch": 0.4834927491514964,
      "grad_norm": 0.3578214943408966,
      "learning_rate": 3e-05,
      "loss": 2.5467,
      "step": 1567
    },
    {
      "epoch": 0.4838012958963283,
      "grad_norm": 0.31990334391593933,
      "learning_rate": 3e-05,
      "loss": 2.2028,
      "step": 1568
    },
    {
      "epoch": 0.48410984264116014,
      "grad_norm": 0.5464645624160767,
      "learning_rate": 3e-05,
      "loss": 2.9154,
      "step": 1569
    },
    {
      "epoch": 0.48441838938599197,
      "grad_norm": 0.3837225139141083,
      "learning_rate": 3e-05,
      "loss": 3.1693,
      "step": 1570
    },
    {
      "epoch": 0.4847269361308238,
      "grad_norm": 0.4334166646003723,
      "learning_rate": 3e-05,
      "loss": 2.6183,
      "step": 1571
    },
    {
      "epoch": 0.4850354828756557,
      "grad_norm": 0.6002141237258911,
      "learning_rate": 3e-05,
      "loss": 2.7773,
      "step": 1572
    },
    {
      "epoch": 0.4853440296204875,
      "grad_norm": 0.4231824278831482,
      "learning_rate": 3e-05,
      "loss": 3.0285,
      "step": 1573
    },
    {
      "epoch": 0.48565257636531933,
      "grad_norm": 0.5055903196334839,
      "learning_rate": 3e-05,
      "loss": 2.8379,
      "step": 1574
    },
    {
      "epoch": 0.48596112311015116,
      "grad_norm": 0.4704802632331848,
      "learning_rate": 3e-05,
      "loss": 2.9247,
      "step": 1575
    },
    {
      "epoch": 0.48626966985498304,
      "grad_norm": 0.34470391273498535,
      "learning_rate": 3e-05,
      "loss": 2.595,
      "step": 1576
    },
    {
      "epoch": 0.4865782165998149,
      "grad_norm": 0.3747197091579437,
      "learning_rate": 3e-05,
      "loss": 2.7041,
      "step": 1577
    },
    {
      "epoch": 0.4868867633446467,
      "grad_norm": 0.3777099549770355,
      "learning_rate": 3e-05,
      "loss": 2.8153,
      "step": 1578
    },
    {
      "epoch": 0.4871953100894786,
      "grad_norm": 0.3668653666973114,
      "learning_rate": 3e-05,
      "loss": 2.8264,
      "step": 1579
    },
    {
      "epoch": 0.4875038568343104,
      "grad_norm": 0.36916372179985046,
      "learning_rate": 3e-05,
      "loss": 2.6379,
      "step": 1580
    },
    {
      "epoch": 0.48781240357914224,
      "grad_norm": 0.4489986300468445,
      "learning_rate": 3e-05,
      "loss": 2.743,
      "step": 1581
    },
    {
      "epoch": 0.48812095032397407,
      "grad_norm": 0.4028119146823883,
      "learning_rate": 3e-05,
      "loss": 2.7651,
      "step": 1582
    },
    {
      "epoch": 0.48842949706880595,
      "grad_norm": 0.5546548366546631,
      "learning_rate": 3e-05,
      "loss": 2.9192,
      "step": 1583
    },
    {
      "epoch": 0.4887380438136378,
      "grad_norm": 0.30430009961128235,
      "learning_rate": 3e-05,
      "loss": 2.5499,
      "step": 1584
    },
    {
      "epoch": 0.4890465905584696,
      "grad_norm": 0.27419066429138184,
      "learning_rate": 3e-05,
      "loss": 2.1188,
      "step": 1585
    },
    {
      "epoch": 0.48935513730330144,
      "grad_norm": 0.4444429576396942,
      "learning_rate": 3e-05,
      "loss": 2.5329,
      "step": 1586
    },
    {
      "epoch": 0.4896636840481333,
      "grad_norm": 0.5690097808837891,
      "learning_rate": 3e-05,
      "loss": 3.0373,
      "step": 1587
    },
    {
      "epoch": 0.48997223079296515,
      "grad_norm": 0.48694705963134766,
      "learning_rate": 3e-05,
      "loss": 3.1918,
      "step": 1588
    },
    {
      "epoch": 0.490280777537797,
      "grad_norm": 0.34520235657691956,
      "learning_rate": 3e-05,
      "loss": 2.7452,
      "step": 1589
    },
    {
      "epoch": 0.4905893242826288,
      "grad_norm": 0.3020474910736084,
      "learning_rate": 3e-05,
      "loss": 2.078,
      "step": 1590
    },
    {
      "epoch": 0.4908978710274607,
      "grad_norm": 0.6240926384925842,
      "learning_rate": 3e-05,
      "loss": 2.9503,
      "step": 1591
    },
    {
      "epoch": 0.4912064177722925,
      "grad_norm": 0.45031264424324036,
      "learning_rate": 3e-05,
      "loss": 2.7756,
      "step": 1592
    },
    {
      "epoch": 0.49151496451712434,
      "grad_norm": 0.394246369600296,
      "learning_rate": 3e-05,
      "loss": 2.353,
      "step": 1593
    },
    {
      "epoch": 0.49182351126195617,
      "grad_norm": 0.3401414155960083,
      "learning_rate": 3e-05,
      "loss": 2.9227,
      "step": 1594
    },
    {
      "epoch": 0.49213205800678805,
      "grad_norm": 0.37477779388427734,
      "learning_rate": 3e-05,
      "loss": 2.2483,
      "step": 1595
    },
    {
      "epoch": 0.4924406047516199,
      "grad_norm": 0.451190710067749,
      "learning_rate": 3e-05,
      "loss": 2.6409,
      "step": 1596
    },
    {
      "epoch": 0.4927491514964517,
      "grad_norm": 0.5501927733421326,
      "learning_rate": 3e-05,
      "loss": 3.0029,
      "step": 1597
    },
    {
      "epoch": 0.49305769824128354,
      "grad_norm": 0.4859730303287506,
      "learning_rate": 3e-05,
      "loss": 2.7112,
      "step": 1598
    },
    {
      "epoch": 0.4933662449861154,
      "grad_norm": 0.6070705056190491,
      "learning_rate": 3e-05,
      "loss": 2.9249,
      "step": 1599
    },
    {
      "epoch": 0.49367479173094725,
      "grad_norm": 0.5782085061073303,
      "learning_rate": 3e-05,
      "loss": 2.7502,
      "step": 1600
    },
    {
      "epoch": 0.4939833384757791,
      "grad_norm": 0.43459030985832214,
      "learning_rate": 3e-05,
      "loss": 2.6673,
      "step": 1601
    },
    {
      "epoch": 0.4942918852206109,
      "grad_norm": 0.3972328305244446,
      "learning_rate": 3e-05,
      "loss": 2.1902,
      "step": 1602
    },
    {
      "epoch": 0.4946004319654428,
      "grad_norm": 0.4112645089626312,
      "learning_rate": 3e-05,
      "loss": 2.5025,
      "step": 1603
    },
    {
      "epoch": 0.4949089787102746,
      "grad_norm": 0.31184130907058716,
      "learning_rate": 3e-05,
      "loss": 2.4411,
      "step": 1604
    },
    {
      "epoch": 0.49521752545510644,
      "grad_norm": 0.396083265542984,
      "learning_rate": 3e-05,
      "loss": 2.5261,
      "step": 1605
    },
    {
      "epoch": 0.4955260721999383,
      "grad_norm": 0.4644416272640228,
      "learning_rate": 3e-05,
      "loss": 2.9669,
      "step": 1606
    },
    {
      "epoch": 0.49583461894477016,
      "grad_norm": 0.3891617953777313,
      "learning_rate": 3e-05,
      "loss": 2.6902,
      "step": 1607
    },
    {
      "epoch": 0.496143165689602,
      "grad_norm": 0.5242297649383545,
      "learning_rate": 3e-05,
      "loss": 3.0694,
      "step": 1608
    },
    {
      "epoch": 0.4964517124344338,
      "grad_norm": 0.4301431477069855,
      "learning_rate": 3e-05,
      "loss": 2.5702,
      "step": 1609
    },
    {
      "epoch": 0.49676025917926564,
      "grad_norm": 0.40124115347862244,
      "learning_rate": 3e-05,
      "loss": 2.6013,
      "step": 1610
    },
    {
      "epoch": 0.4970688059240975,
      "grad_norm": 0.6594104170799255,
      "learning_rate": 3e-05,
      "loss": 2.7237,
      "step": 1611
    },
    {
      "epoch": 0.49737735266892935,
      "grad_norm": 0.533812940120697,
      "learning_rate": 3e-05,
      "loss": 3.1564,
      "step": 1612
    },
    {
      "epoch": 0.4976858994137612,
      "grad_norm": 0.29430219531059265,
      "learning_rate": 3e-05,
      "loss": 2.6577,
      "step": 1613
    },
    {
      "epoch": 0.497994446158593,
      "grad_norm": 0.40662962198257446,
      "learning_rate": 3e-05,
      "loss": 2.8116,
      "step": 1614
    },
    {
      "epoch": 0.4983029929034249,
      "grad_norm": 0.3725249767303467,
      "learning_rate": 3e-05,
      "loss": 2.3387,
      "step": 1615
    },
    {
      "epoch": 0.4986115396482567,
      "grad_norm": 0.37109196186065674,
      "learning_rate": 3e-05,
      "loss": 2.4788,
      "step": 1616
    },
    {
      "epoch": 0.49892008639308855,
      "grad_norm": 0.3833397626876831,
      "learning_rate": 3e-05,
      "loss": 2.4994,
      "step": 1617
    },
    {
      "epoch": 0.4992286331379204,
      "grad_norm": 0.6564522981643677,
      "learning_rate": 3e-05,
      "loss": 3.3482,
      "step": 1618
    },
    {
      "epoch": 0.49953717988275226,
      "grad_norm": 0.3033236265182495,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 1619
    },
    {
      "epoch": 0.4998457266275841,
      "grad_norm": 0.3454381823539734,
      "learning_rate": 3e-05,
      "loss": 2.4953,
      "step": 1620
    },
    {
      "epoch": 0.5001542733724159,
      "grad_norm": 0.42523840069770813,
      "learning_rate": 3e-05,
      "loss": 2.7933,
      "step": 1621
    },
    {
      "epoch": 0.5004628201172477,
      "grad_norm": 0.4610409140586853,
      "learning_rate": 3e-05,
      "loss": 2.6401,
      "step": 1622
    },
    {
      "epoch": 0.5007713668620796,
      "grad_norm": 0.3483074903488159,
      "learning_rate": 3e-05,
      "loss": 2.6013,
      "step": 1623
    },
    {
      "epoch": 0.5010799136069114,
      "grad_norm": 0.36740410327911377,
      "learning_rate": 3e-05,
      "loss": 2.7199,
      "step": 1624
    },
    {
      "epoch": 0.5013884603517433,
      "grad_norm": 0.3452313542366028,
      "learning_rate": 3e-05,
      "loss": 2.6963,
      "step": 1625
    },
    {
      "epoch": 0.5016970070965752,
      "grad_norm": 0.5462652444839478,
      "learning_rate": 3e-05,
      "loss": 2.8419,
      "step": 1626
    },
    {
      "epoch": 0.502005553841407,
      "grad_norm": 0.31306546926498413,
      "learning_rate": 3e-05,
      "loss": 2.3641,
      "step": 1627
    },
    {
      "epoch": 0.5023141005862388,
      "grad_norm": 0.3925100266933441,
      "learning_rate": 3e-05,
      "loss": 2.6604,
      "step": 1628
    },
    {
      "epoch": 0.5026226473310706,
      "grad_norm": 0.3354152739048004,
      "learning_rate": 3e-05,
      "loss": 2.4355,
      "step": 1629
    },
    {
      "epoch": 0.5029311940759025,
      "grad_norm": 0.38791340589523315,
      "learning_rate": 3e-05,
      "loss": 2.6835,
      "step": 1630
    },
    {
      "epoch": 0.5032397408207343,
      "grad_norm": 0.405977725982666,
      "learning_rate": 3e-05,
      "loss": 3.1869,
      "step": 1631
    },
    {
      "epoch": 0.5035482875655661,
      "grad_norm": 0.34836137294769287,
      "learning_rate": 3e-05,
      "loss": 2.661,
      "step": 1632
    },
    {
      "epoch": 0.5038568343103981,
      "grad_norm": 0.3889111280441284,
      "learning_rate": 3e-05,
      "loss": 2.7348,
      "step": 1633
    },
    {
      "epoch": 0.5041653810552299,
      "grad_norm": 0.30361315608024597,
      "learning_rate": 3e-05,
      "loss": 2.4406,
      "step": 1634
    },
    {
      "epoch": 0.5044739278000617,
      "grad_norm": 0.3601548373699188,
      "learning_rate": 3e-05,
      "loss": 2.8035,
      "step": 1635
    },
    {
      "epoch": 0.5047824745448936,
      "grad_norm": 0.4318731725215912,
      "learning_rate": 3e-05,
      "loss": 2.6381,
      "step": 1636
    },
    {
      "epoch": 0.5050910212897254,
      "grad_norm": 0.30395030975341797,
      "learning_rate": 3e-05,
      "loss": 2.3321,
      "step": 1637
    },
    {
      "epoch": 0.5053995680345572,
      "grad_norm": 0.4935227036476135,
      "learning_rate": 3e-05,
      "loss": 3.006,
      "step": 1638
    },
    {
      "epoch": 0.505708114779389,
      "grad_norm": 0.32266750931739807,
      "learning_rate": 3e-05,
      "loss": 2.7027,
      "step": 1639
    },
    {
      "epoch": 0.5060166615242209,
      "grad_norm": 0.4440366327762604,
      "learning_rate": 3e-05,
      "loss": 3.0615,
      "step": 1640
    },
    {
      "epoch": 0.5063252082690528,
      "grad_norm": 0.3273499608039856,
      "learning_rate": 3e-05,
      "loss": 2.489,
      "step": 1641
    },
    {
      "epoch": 0.5066337550138846,
      "grad_norm": 0.33660799264907837,
      "learning_rate": 3e-05,
      "loss": 2.4475,
      "step": 1642
    },
    {
      "epoch": 0.5069423017587165,
      "grad_norm": 0.3202866017818451,
      "learning_rate": 3e-05,
      "loss": 2.4899,
      "step": 1643
    },
    {
      "epoch": 0.5072508485035483,
      "grad_norm": 0.292321115732193,
      "learning_rate": 3e-05,
      "loss": 2.3956,
      "step": 1644
    },
    {
      "epoch": 0.5075593952483801,
      "grad_norm": 0.35103198885917664,
      "learning_rate": 3e-05,
      "loss": 2.6357,
      "step": 1645
    },
    {
      "epoch": 0.507867941993212,
      "grad_norm": 0.4939737319946289,
      "learning_rate": 3e-05,
      "loss": 2.8856,
      "step": 1646
    },
    {
      "epoch": 0.5081764887380438,
      "grad_norm": 0.2993530333042145,
      "learning_rate": 3e-05,
      "loss": 2.3868,
      "step": 1647
    },
    {
      "epoch": 0.5084850354828756,
      "grad_norm": 0.3006252944469452,
      "learning_rate": 3e-05,
      "loss": 2.5518,
      "step": 1648
    },
    {
      "epoch": 0.5087935822277075,
      "grad_norm": 0.3943158984184265,
      "learning_rate": 3e-05,
      "loss": 2.7638,
      "step": 1649
    },
    {
      "epoch": 0.5091021289725394,
      "grad_norm": 0.5232028961181641,
      "learning_rate": 3e-05,
      "loss": 3.0938,
      "step": 1650
    },
    {
      "epoch": 0.5094106757173712,
      "grad_norm": 0.2864915132522583,
      "learning_rate": 3e-05,
      "loss": 2.5598,
      "step": 1651
    },
    {
      "epoch": 0.509719222462203,
      "grad_norm": 0.3905394673347473,
      "learning_rate": 3e-05,
      "loss": 2.8282,
      "step": 1652
    },
    {
      "epoch": 0.5100277692070349,
      "grad_norm": 0.3154192566871643,
      "learning_rate": 3e-05,
      "loss": 2.4099,
      "step": 1653
    },
    {
      "epoch": 0.5103363159518667,
      "grad_norm": 0.4238516092300415,
      "learning_rate": 3e-05,
      "loss": 2.7145,
      "step": 1654
    },
    {
      "epoch": 0.5106448626966985,
      "grad_norm": 0.37606295943260193,
      "learning_rate": 3e-05,
      "loss": 2.4632,
      "step": 1655
    },
    {
      "epoch": 0.5109534094415303,
      "grad_norm": 0.3119703531265259,
      "learning_rate": 3e-05,
      "loss": 2.7921,
      "step": 1656
    },
    {
      "epoch": 0.5112619561863623,
      "grad_norm": 0.4015417993068695,
      "learning_rate": 3e-05,
      "loss": 2.7777,
      "step": 1657
    },
    {
      "epoch": 0.5115705029311941,
      "grad_norm": 0.314378947019577,
      "learning_rate": 3e-05,
      "loss": 2.7296,
      "step": 1658
    },
    {
      "epoch": 0.5118790496760259,
      "grad_norm": 0.3702321946620941,
      "learning_rate": 3e-05,
      "loss": 2.8218,
      "step": 1659
    },
    {
      "epoch": 0.5121875964208578,
      "grad_norm": 0.6027920842170715,
      "learning_rate": 3e-05,
      "loss": 3.028,
      "step": 1660
    },
    {
      "epoch": 0.5124961431656896,
      "grad_norm": 0.2812676727771759,
      "learning_rate": 3e-05,
      "loss": 2.4171,
      "step": 1661
    },
    {
      "epoch": 0.5128046899105214,
      "grad_norm": 0.442354679107666,
      "learning_rate": 3e-05,
      "loss": 2.8097,
      "step": 1662
    },
    {
      "epoch": 0.5131132366553532,
      "grad_norm": 0.3620881736278534,
      "learning_rate": 3e-05,
      "loss": 2.2628,
      "step": 1663
    },
    {
      "epoch": 0.5134217834001852,
      "grad_norm": 0.530297040939331,
      "learning_rate": 3e-05,
      "loss": 2.602,
      "step": 1664
    },
    {
      "epoch": 0.513730330145017,
      "grad_norm": 0.31801244616508484,
      "learning_rate": 3e-05,
      "loss": 2.471,
      "step": 1665
    },
    {
      "epoch": 0.5140388768898488,
      "grad_norm": 0.32291916012763977,
      "learning_rate": 3e-05,
      "loss": 2.7074,
      "step": 1666
    },
    {
      "epoch": 0.5143474236346807,
      "grad_norm": 0.49042975902557373,
      "learning_rate": 3e-05,
      "loss": 2.6864,
      "step": 1667
    },
    {
      "epoch": 0.5146559703795125,
      "grad_norm": 0.3325033187866211,
      "learning_rate": 3e-05,
      "loss": 2.5405,
      "step": 1668
    },
    {
      "epoch": 0.5149645171243443,
      "grad_norm": 0.3791443109512329,
      "learning_rate": 3e-05,
      "loss": 2.7489,
      "step": 1669
    },
    {
      "epoch": 0.5152730638691762,
      "grad_norm": 0.37896838784217834,
      "learning_rate": 3e-05,
      "loss": 3.0854,
      "step": 1670
    },
    {
      "epoch": 0.515581610614008,
      "grad_norm": 0.4317682087421417,
      "learning_rate": 3e-05,
      "loss": 3.3075,
      "step": 1671
    },
    {
      "epoch": 0.5158901573588399,
      "grad_norm": 0.4479047954082489,
      "learning_rate": 3e-05,
      "loss": 2.5176,
      "step": 1672
    },
    {
      "epoch": 0.5161987041036717,
      "grad_norm": 0.37080270051956177,
      "learning_rate": 3e-05,
      "loss": 2.942,
      "step": 1673
    },
    {
      "epoch": 0.5165072508485036,
      "grad_norm": 0.28573471307754517,
      "learning_rate": 3e-05,
      "loss": 2.2821,
      "step": 1674
    },
    {
      "epoch": 0.5168157975933354,
      "grad_norm": 0.46885693073272705,
      "learning_rate": 3e-05,
      "loss": 3.0661,
      "step": 1675
    },
    {
      "epoch": 0.5171243443381672,
      "grad_norm": 0.2819637060165405,
      "learning_rate": 3e-05,
      "loss": 2.4839,
      "step": 1676
    },
    {
      "epoch": 0.5174328910829991,
      "grad_norm": 0.41865482926368713,
      "learning_rate": 3e-05,
      "loss": 2.9538,
      "step": 1677
    },
    {
      "epoch": 0.5177414378278309,
      "grad_norm": 0.43568989634513855,
      "learning_rate": 3e-05,
      "loss": 2.9169,
      "step": 1678
    },
    {
      "epoch": 0.5180499845726627,
      "grad_norm": 0.30851465463638306,
      "learning_rate": 3e-05,
      "loss": 2.8745,
      "step": 1679
    },
    {
      "epoch": 0.5183585313174947,
      "grad_norm": 0.3105289340019226,
      "learning_rate": 3e-05,
      "loss": 2.4255,
      "step": 1680
    },
    {
      "epoch": 0.5186670780623265,
      "grad_norm": 0.32589539885520935,
      "learning_rate": 3e-05,
      "loss": 2.727,
      "step": 1681
    },
    {
      "epoch": 0.5189756248071583,
      "grad_norm": 0.29433050751686096,
      "learning_rate": 3e-05,
      "loss": 2.4772,
      "step": 1682
    },
    {
      "epoch": 0.5192841715519901,
      "grad_norm": 0.3651641607284546,
      "learning_rate": 3e-05,
      "loss": 2.5789,
      "step": 1683
    },
    {
      "epoch": 0.519592718296822,
      "grad_norm": 0.2723684310913086,
      "learning_rate": 3e-05,
      "loss": 2.5751,
      "step": 1684
    },
    {
      "epoch": 0.5199012650416538,
      "grad_norm": 0.3247137665748596,
      "learning_rate": 3e-05,
      "loss": 2.4918,
      "step": 1685
    },
    {
      "epoch": 0.5202098117864856,
      "grad_norm": 0.32286760210990906,
      "learning_rate": 3e-05,
      "loss": 2.3303,
      "step": 1686
    },
    {
      "epoch": 0.5205183585313174,
      "grad_norm": 0.25823745131492615,
      "learning_rate": 3e-05,
      "loss": 2.3304,
      "step": 1687
    },
    {
      "epoch": 0.5208269052761494,
      "grad_norm": 0.36339297890663147,
      "learning_rate": 3e-05,
      "loss": 2.6724,
      "step": 1688
    },
    {
      "epoch": 0.5211354520209812,
      "grad_norm": 0.29126620292663574,
      "learning_rate": 3e-05,
      "loss": 2.6284,
      "step": 1689
    },
    {
      "epoch": 0.521443998765813,
      "grad_norm": 0.3497116267681122,
      "learning_rate": 3e-05,
      "loss": 2.6851,
      "step": 1690
    },
    {
      "epoch": 0.5217525455106449,
      "grad_norm": 0.324299156665802,
      "learning_rate": 3e-05,
      "loss": 2.5865,
      "step": 1691
    },
    {
      "epoch": 0.5220610922554767,
      "grad_norm": 0.3677506148815155,
      "learning_rate": 3e-05,
      "loss": 2.4579,
      "step": 1692
    },
    {
      "epoch": 0.5223696390003085,
      "grad_norm": 0.307575523853302,
      "learning_rate": 3e-05,
      "loss": 2.4014,
      "step": 1693
    },
    {
      "epoch": 0.5226781857451404,
      "grad_norm": 0.32490015029907227,
      "learning_rate": 3e-05,
      "loss": 2.5131,
      "step": 1694
    },
    {
      "epoch": 0.5229867324899722,
      "grad_norm": 0.2470085620880127,
      "learning_rate": 3e-05,
      "loss": 2.326,
      "step": 1695
    },
    {
      "epoch": 0.5232952792348041,
      "grad_norm": 0.4868578314781189,
      "learning_rate": 3e-05,
      "loss": 3.2556,
      "step": 1696
    },
    {
      "epoch": 0.523603825979636,
      "grad_norm": 0.362221896648407,
      "learning_rate": 3e-05,
      "loss": 2.7828,
      "step": 1697
    },
    {
      "epoch": 0.5239123727244678,
      "grad_norm": 0.40775471925735474,
      "learning_rate": 3e-05,
      "loss": 3.2336,
      "step": 1698
    },
    {
      "epoch": 0.5242209194692996,
      "grad_norm": 0.3011956810951233,
      "learning_rate": 3e-05,
      "loss": 2.2592,
      "step": 1699
    },
    {
      "epoch": 0.5245294662141314,
      "grad_norm": 0.3967321515083313,
      "learning_rate": 3e-05,
      "loss": 2.81,
      "step": 1700
    },
    {
      "epoch": 0.5248380129589633,
      "grad_norm": 0.37166154384613037,
      "learning_rate": 3e-05,
      "loss": 2.5431,
      "step": 1701
    },
    {
      "epoch": 0.5251465597037951,
      "grad_norm": 0.4062604010105133,
      "learning_rate": 3e-05,
      "loss": 2.6376,
      "step": 1702
    },
    {
      "epoch": 0.5254551064486269,
      "grad_norm": 0.313347727060318,
      "learning_rate": 3e-05,
      "loss": 2.3459,
      "step": 1703
    },
    {
      "epoch": 0.5257636531934589,
      "grad_norm": 0.8226602077484131,
      "learning_rate": 3e-05,
      "loss": 3.1876,
      "step": 1704
    },
    {
      "epoch": 0.5260721999382907,
      "grad_norm": 0.3581107556819916,
      "learning_rate": 3e-05,
      "loss": 2.6067,
      "step": 1705
    },
    {
      "epoch": 0.5263807466831225,
      "grad_norm": 0.410627543926239,
      "learning_rate": 3e-05,
      "loss": 2.6455,
      "step": 1706
    },
    {
      "epoch": 0.5266892934279543,
      "grad_norm": 0.4492070972919464,
      "learning_rate": 3e-05,
      "loss": 2.634,
      "step": 1707
    },
    {
      "epoch": 0.5269978401727862,
      "grad_norm": 0.40184956789016724,
      "learning_rate": 3e-05,
      "loss": 2.8047,
      "step": 1708
    },
    {
      "epoch": 0.527306386917618,
      "grad_norm": 0.2819567024707794,
      "learning_rate": 3e-05,
      "loss": 1.965,
      "step": 1709
    },
    {
      "epoch": 0.5276149336624498,
      "grad_norm": 0.3046351671218872,
      "learning_rate": 3e-05,
      "loss": 2.6007,
      "step": 1710
    },
    {
      "epoch": 0.5279234804072817,
      "grad_norm": 0.338138222694397,
      "learning_rate": 3e-05,
      "loss": 3.0538,
      "step": 1711
    },
    {
      "epoch": 0.5282320271521136,
      "grad_norm": 0.38633912801742554,
      "learning_rate": 3e-05,
      "loss": 2.5772,
      "step": 1712
    },
    {
      "epoch": 0.5285405738969454,
      "grad_norm": 0.3535228669643402,
      "learning_rate": 3e-05,
      "loss": 2.6554,
      "step": 1713
    },
    {
      "epoch": 0.5288491206417772,
      "grad_norm": 0.6468623876571655,
      "learning_rate": 3e-05,
      "loss": 3.2737,
      "step": 1714
    },
    {
      "epoch": 0.5291576673866091,
      "grad_norm": 0.31310710310935974,
      "learning_rate": 3e-05,
      "loss": 2.7591,
      "step": 1715
    },
    {
      "epoch": 0.5294662141314409,
      "grad_norm": 0.3655259609222412,
      "learning_rate": 3e-05,
      "loss": 2.7045,
      "step": 1716
    },
    {
      "epoch": 0.5297747608762727,
      "grad_norm": 0.5266317129135132,
      "learning_rate": 3e-05,
      "loss": 3.0872,
      "step": 1717
    },
    {
      "epoch": 0.5300833076211046,
      "grad_norm": 0.3490053713321686,
      "learning_rate": 3e-05,
      "loss": 2.7577,
      "step": 1718
    },
    {
      "epoch": 0.5303918543659364,
      "grad_norm": 0.3306678235530853,
      "learning_rate": 3e-05,
      "loss": 2.4778,
      "step": 1719
    },
    {
      "epoch": 0.5307004011107683,
      "grad_norm": 0.376544713973999,
      "learning_rate": 3e-05,
      "loss": 2.6798,
      "step": 1720
    },
    {
      "epoch": 0.5310089478556002,
      "grad_norm": 0.3733018636703491,
      "learning_rate": 3e-05,
      "loss": 2.7735,
      "step": 1721
    },
    {
      "epoch": 0.531317494600432,
      "grad_norm": 0.3042423725128174,
      "learning_rate": 3e-05,
      "loss": 2.6467,
      "step": 1722
    },
    {
      "epoch": 0.5316260413452638,
      "grad_norm": 0.39278408885002136,
      "learning_rate": 3e-05,
      "loss": 2.9289,
      "step": 1723
    },
    {
      "epoch": 0.5319345880900956,
      "grad_norm": 0.38252609968185425,
      "learning_rate": 3e-05,
      "loss": 2.5449,
      "step": 1724
    },
    {
      "epoch": 0.5322431348349275,
      "grad_norm": 0.355350524187088,
      "learning_rate": 3e-05,
      "loss": 2.4288,
      "step": 1725
    },
    {
      "epoch": 0.5325516815797593,
      "grad_norm": 0.34510883688926697,
      "learning_rate": 3e-05,
      "loss": 2.811,
      "step": 1726
    },
    {
      "epoch": 0.5328602283245911,
      "grad_norm": 0.29719817638397217,
      "learning_rate": 3e-05,
      "loss": 2.6064,
      "step": 1727
    },
    {
      "epoch": 0.5331687750694231,
      "grad_norm": 0.297100305557251,
      "learning_rate": 3e-05,
      "loss": 2.7171,
      "step": 1728
    },
    {
      "epoch": 0.5334773218142549,
      "grad_norm": 0.42260539531707764,
      "learning_rate": 3e-05,
      "loss": 2.88,
      "step": 1729
    },
    {
      "epoch": 0.5337858685590867,
      "grad_norm": 0.3438571095466614,
      "learning_rate": 3e-05,
      "loss": 2.7997,
      "step": 1730
    },
    {
      "epoch": 0.5340944153039185,
      "grad_norm": 0.38454294204711914,
      "learning_rate": 3e-05,
      "loss": 2.8103,
      "step": 1731
    },
    {
      "epoch": 0.5344029620487504,
      "grad_norm": 0.3376411199569702,
      "learning_rate": 3e-05,
      "loss": 2.7868,
      "step": 1732
    },
    {
      "epoch": 0.5347115087935822,
      "grad_norm": 0.3197532892227173,
      "learning_rate": 3e-05,
      "loss": 2.9247,
      "step": 1733
    },
    {
      "epoch": 0.535020055538414,
      "grad_norm": 0.380571573972702,
      "learning_rate": 3e-05,
      "loss": 2.3086,
      "step": 1734
    },
    {
      "epoch": 0.5353286022832459,
      "grad_norm": 0.42782899737358093,
      "learning_rate": 3e-05,
      "loss": 2.9169,
      "step": 1735
    },
    {
      "epoch": 0.5356371490280778,
      "grad_norm": 0.3073023557662964,
      "learning_rate": 3e-05,
      "loss": 2.5997,
      "step": 1736
    },
    {
      "epoch": 0.5359456957729096,
      "grad_norm": 0.33757370710372925,
      "learning_rate": 3e-05,
      "loss": 2.4294,
      "step": 1737
    },
    {
      "epoch": 0.5362542425177415,
      "grad_norm": 0.3303975462913513,
      "learning_rate": 3e-05,
      "loss": 2.308,
      "step": 1738
    },
    {
      "epoch": 0.5365627892625733,
      "grad_norm": 0.35160350799560547,
      "learning_rate": 3e-05,
      "loss": 2.6371,
      "step": 1739
    },
    {
      "epoch": 0.5368713360074051,
      "grad_norm": 0.39013320207595825,
      "learning_rate": 3e-05,
      "loss": 2.9465,
      "step": 1740
    },
    {
      "epoch": 0.5371798827522369,
      "grad_norm": 0.3293541371822357,
      "learning_rate": 3e-05,
      "loss": 2.2803,
      "step": 1741
    },
    {
      "epoch": 0.5374884294970688,
      "grad_norm": 0.34581053256988525,
      "learning_rate": 3e-05,
      "loss": 2.5024,
      "step": 1742
    },
    {
      "epoch": 0.5377969762419006,
      "grad_norm": 0.35894790291786194,
      "learning_rate": 3e-05,
      "loss": 2.6478,
      "step": 1743
    },
    {
      "epoch": 0.5381055229867325,
      "grad_norm": 0.33879491686820984,
      "learning_rate": 3e-05,
      "loss": 2.6175,
      "step": 1744
    },
    {
      "epoch": 0.5384140697315644,
      "grad_norm": 0.3612264394760132,
      "learning_rate": 3e-05,
      "loss": 2.565,
      "step": 1745
    },
    {
      "epoch": 0.5387226164763962,
      "grad_norm": 0.49857962131500244,
      "learning_rate": 3e-05,
      "loss": 3.2948,
      "step": 1746
    },
    {
      "epoch": 0.539031163221228,
      "grad_norm": 0.3405458629131317,
      "learning_rate": 3e-05,
      "loss": 2.2905,
      "step": 1747
    },
    {
      "epoch": 0.5393397099660598,
      "grad_norm": 0.44803398847579956,
      "learning_rate": 3e-05,
      "loss": 3.7067,
      "step": 1748
    },
    {
      "epoch": 0.5396482567108917,
      "grad_norm": 0.5877761244773865,
      "learning_rate": 3e-05,
      "loss": 3.3724,
      "step": 1749
    },
    {
      "epoch": 0.5399568034557235,
      "grad_norm": 0.34262049198150635,
      "learning_rate": 3e-05,
      "loss": 3.0155,
      "step": 1750
    },
    {
      "epoch": 0.5402653502005554,
      "grad_norm": 0.3923507332801819,
      "learning_rate": 3e-05,
      "loss": 3.2323,
      "step": 1751
    },
    {
      "epoch": 0.5405738969453873,
      "grad_norm": 0.3871021568775177,
      "learning_rate": 3e-05,
      "loss": 2.7126,
      "step": 1752
    },
    {
      "epoch": 0.5408824436902191,
      "grad_norm": 0.31630492210388184,
      "learning_rate": 3e-05,
      "loss": 2.4423,
      "step": 1753
    },
    {
      "epoch": 0.5411909904350509,
      "grad_norm": 0.3699815571308136,
      "learning_rate": 3e-05,
      "loss": 2.8711,
      "step": 1754
    },
    {
      "epoch": 0.5414995371798828,
      "grad_norm": 0.3360150456428528,
      "learning_rate": 3e-05,
      "loss": 2.6574,
      "step": 1755
    },
    {
      "epoch": 0.5418080839247146,
      "grad_norm": 0.5751543641090393,
      "learning_rate": 3e-05,
      "loss": 3.2725,
      "step": 1756
    },
    {
      "epoch": 0.5421166306695464,
      "grad_norm": 0.2963869571685791,
      "learning_rate": 3e-05,
      "loss": 2.5501,
      "step": 1757
    },
    {
      "epoch": 0.5424251774143782,
      "grad_norm": 0.35280632972717285,
      "learning_rate": 3e-05,
      "loss": 2.6562,
      "step": 1758
    },
    {
      "epoch": 0.5427337241592102,
      "grad_norm": 0.42578262090682983,
      "learning_rate": 3e-05,
      "loss": 2.9552,
      "step": 1759
    },
    {
      "epoch": 0.543042270904042,
      "grad_norm": 0.26292774081230164,
      "learning_rate": 3e-05,
      "loss": 2.3602,
      "step": 1760
    },
    {
      "epoch": 0.5433508176488738,
      "grad_norm": 0.4971487522125244,
      "learning_rate": 3e-05,
      "loss": 3.02,
      "step": 1761
    },
    {
      "epoch": 0.5436593643937057,
      "grad_norm": 0.42923474311828613,
      "learning_rate": 3e-05,
      "loss": 2.9607,
      "step": 1762
    },
    {
      "epoch": 0.5439679111385375,
      "grad_norm": 0.3830646872520447,
      "learning_rate": 3e-05,
      "loss": 2.438,
      "step": 1763
    },
    {
      "epoch": 0.5442764578833693,
      "grad_norm": 0.2715272903442383,
      "learning_rate": 3e-05,
      "loss": 2.5599,
      "step": 1764
    },
    {
      "epoch": 0.5445850046282011,
      "grad_norm": 0.3658992052078247,
      "learning_rate": 3e-05,
      "loss": 2.6659,
      "step": 1765
    },
    {
      "epoch": 0.544893551373033,
      "grad_norm": 0.4321707785129547,
      "learning_rate": 3e-05,
      "loss": 2.4854,
      "step": 1766
    },
    {
      "epoch": 0.5452020981178649,
      "grad_norm": 0.31105655431747437,
      "learning_rate": 3e-05,
      "loss": 2.6108,
      "step": 1767
    },
    {
      "epoch": 0.5455106448626967,
      "grad_norm": 0.3162234425544739,
      "learning_rate": 3e-05,
      "loss": 2.4305,
      "step": 1768
    },
    {
      "epoch": 0.5458191916075286,
      "grad_norm": 0.2657429277896881,
      "learning_rate": 3e-05,
      "loss": 2.6477,
      "step": 1769
    },
    {
      "epoch": 0.5461277383523604,
      "grad_norm": 0.29710811376571655,
      "learning_rate": 3e-05,
      "loss": 2.446,
      "step": 1770
    },
    {
      "epoch": 0.5464362850971922,
      "grad_norm": 0.3215124309062958,
      "learning_rate": 3e-05,
      "loss": 2.2926,
      "step": 1771
    },
    {
      "epoch": 0.546744831842024,
      "grad_norm": 0.39421600103378296,
      "learning_rate": 3e-05,
      "loss": 2.9613,
      "step": 1772
    },
    {
      "epoch": 0.5470533785868559,
      "grad_norm": 0.34313124418258667,
      "learning_rate": 3e-05,
      "loss": 2.802,
      "step": 1773
    },
    {
      "epoch": 0.5473619253316877,
      "grad_norm": 0.398807555437088,
      "learning_rate": 3e-05,
      "loss": 3.447,
      "step": 1774
    },
    {
      "epoch": 0.5476704720765196,
      "grad_norm": 0.3040825128555298,
      "learning_rate": 3e-05,
      "loss": 2.3763,
      "step": 1775
    },
    {
      "epoch": 0.5479790188213515,
      "grad_norm": 0.29558873176574707,
      "learning_rate": 3e-05,
      "loss": 2.5285,
      "step": 1776
    },
    {
      "epoch": 0.5482875655661833,
      "grad_norm": 0.35539618134498596,
      "learning_rate": 3e-05,
      "loss": 2.7637,
      "step": 1777
    },
    {
      "epoch": 0.5485961123110151,
      "grad_norm": 0.34798794984817505,
      "learning_rate": 3e-05,
      "loss": 2.5633,
      "step": 1778
    },
    {
      "epoch": 0.548904659055847,
      "grad_norm": 0.2828461825847626,
      "learning_rate": 3e-05,
      "loss": 2.6647,
      "step": 1779
    },
    {
      "epoch": 0.5492132058006788,
      "grad_norm": 0.48053836822509766,
      "learning_rate": 3e-05,
      "loss": 2.5918,
      "step": 1780
    },
    {
      "epoch": 0.5495217525455106,
      "grad_norm": 0.31791162490844727,
      "learning_rate": 3e-05,
      "loss": 2.4877,
      "step": 1781
    },
    {
      "epoch": 0.5498302992903424,
      "grad_norm": 0.2918544411659241,
      "learning_rate": 3e-05,
      "loss": 2.3345,
      "step": 1782
    },
    {
      "epoch": 0.5501388460351744,
      "grad_norm": 0.3297959566116333,
      "learning_rate": 3e-05,
      "loss": 2.6386,
      "step": 1783
    },
    {
      "epoch": 0.5504473927800062,
      "grad_norm": 0.3603450357913971,
      "learning_rate": 3e-05,
      "loss": 2.3877,
      "step": 1784
    },
    {
      "epoch": 0.550755939524838,
      "grad_norm": 0.4325063228607178,
      "learning_rate": 3e-05,
      "loss": 2.9395,
      "step": 1785
    },
    {
      "epoch": 0.5510644862696699,
      "grad_norm": 0.5895464420318604,
      "learning_rate": 3e-05,
      "loss": 3.4669,
      "step": 1786
    },
    {
      "epoch": 0.5513730330145017,
      "grad_norm": 0.2964521646499634,
      "learning_rate": 3e-05,
      "loss": 2.728,
      "step": 1787
    },
    {
      "epoch": 0.5516815797593335,
      "grad_norm": 0.30058997869491577,
      "learning_rate": 3e-05,
      "loss": 2.6638,
      "step": 1788
    },
    {
      "epoch": 0.5519901265041653,
      "grad_norm": 0.3132179081439972,
      "learning_rate": 3e-05,
      "loss": 2.7093,
      "step": 1789
    },
    {
      "epoch": 0.5522986732489972,
      "grad_norm": 0.3626979887485504,
      "learning_rate": 3e-05,
      "loss": 2.4565,
      "step": 1790
    },
    {
      "epoch": 0.5526072199938291,
      "grad_norm": 0.2594972252845764,
      "learning_rate": 3e-05,
      "loss": 2.1917,
      "step": 1791
    },
    {
      "epoch": 0.5529157667386609,
      "grad_norm": 0.2946871817111969,
      "learning_rate": 3e-05,
      "loss": 2.4383,
      "step": 1792
    },
    {
      "epoch": 0.5532243134834928,
      "grad_norm": 0.3073224127292633,
      "learning_rate": 3e-05,
      "loss": 2.4735,
      "step": 1793
    },
    {
      "epoch": 0.5535328602283246,
      "grad_norm": 0.33085721731185913,
      "learning_rate": 3e-05,
      "loss": 2.9155,
      "step": 1794
    },
    {
      "epoch": 0.5538414069731564,
      "grad_norm": 0.3167915344238281,
      "learning_rate": 3e-05,
      "loss": 2.5648,
      "step": 1795
    },
    {
      "epoch": 0.5541499537179883,
      "grad_norm": 0.3947449326515198,
      "learning_rate": 3e-05,
      "loss": 2.7195,
      "step": 1796
    },
    {
      "epoch": 0.5544585004628201,
      "grad_norm": 0.29865971207618713,
      "learning_rate": 3e-05,
      "loss": 2.8072,
      "step": 1797
    },
    {
      "epoch": 0.5547670472076519,
      "grad_norm": 0.32342618703842163,
      "learning_rate": 3e-05,
      "loss": 2.7242,
      "step": 1798
    },
    {
      "epoch": 0.5550755939524838,
      "grad_norm": 0.35544127225875854,
      "learning_rate": 3e-05,
      "loss": 2.4616,
      "step": 1799
    },
    {
      "epoch": 0.5553841406973157,
      "grad_norm": 0.30906030535697937,
      "learning_rate": 3e-05,
      "loss": 2.4555,
      "step": 1800
    },
    {
      "epoch": 0.5556926874421475,
      "grad_norm": 0.2638324499130249,
      "learning_rate": 3e-05,
      "loss": 2.5501,
      "step": 1801
    },
    {
      "epoch": 0.5560012341869793,
      "grad_norm": 0.5023085474967957,
      "learning_rate": 3e-05,
      "loss": 3.0855,
      "step": 1802
    },
    {
      "epoch": 0.5563097809318112,
      "grad_norm": 0.35941094160079956,
      "learning_rate": 3e-05,
      "loss": 2.4908,
      "step": 1803
    },
    {
      "epoch": 0.556618327676643,
      "grad_norm": 0.4491455554962158,
      "learning_rate": 3e-05,
      "loss": 2.9572,
      "step": 1804
    },
    {
      "epoch": 0.5569268744214748,
      "grad_norm": 0.44301337003707886,
      "learning_rate": 3e-05,
      "loss": 2.8863,
      "step": 1805
    },
    {
      "epoch": 0.5572354211663066,
      "grad_norm": 0.3584831655025482,
      "learning_rate": 3e-05,
      "loss": 2.6321,
      "step": 1806
    },
    {
      "epoch": 0.5575439679111386,
      "grad_norm": 0.3179319202899933,
      "learning_rate": 3e-05,
      "loss": 2.5629,
      "step": 1807
    },
    {
      "epoch": 0.5578525146559704,
      "grad_norm": 0.39769089221954346,
      "learning_rate": 3e-05,
      "loss": 2.3598,
      "step": 1808
    },
    {
      "epoch": 0.5581610614008022,
      "grad_norm": 0.45672088861465454,
      "learning_rate": 3e-05,
      "loss": 2.7827,
      "step": 1809
    },
    {
      "epoch": 0.5584696081456341,
      "grad_norm": 0.40564146637916565,
      "learning_rate": 3e-05,
      "loss": 2.6394,
      "step": 1810
    },
    {
      "epoch": 0.5587781548904659,
      "grad_norm": 0.3717137575149536,
      "learning_rate": 3e-05,
      "loss": 2.4997,
      "step": 1811
    },
    {
      "epoch": 0.5590867016352977,
      "grad_norm": 0.4004184901714325,
      "learning_rate": 3e-05,
      "loss": 2.4249,
      "step": 1812
    },
    {
      "epoch": 0.5593952483801296,
      "grad_norm": 0.3864268362522125,
      "learning_rate": 3e-05,
      "loss": 2.7476,
      "step": 1813
    },
    {
      "epoch": 0.5597037951249614,
      "grad_norm": 0.2754631042480469,
      "learning_rate": 3e-05,
      "loss": 2.3015,
      "step": 1814
    },
    {
      "epoch": 0.5600123418697933,
      "grad_norm": 0.317277193069458,
      "learning_rate": 3e-05,
      "loss": 2.56,
      "step": 1815
    },
    {
      "epoch": 0.5603208886146251,
      "grad_norm": 0.4582974314689636,
      "learning_rate": 3e-05,
      "loss": 3.0188,
      "step": 1816
    },
    {
      "epoch": 0.560629435359457,
      "grad_norm": 0.3826389014720917,
      "learning_rate": 3e-05,
      "loss": 2.8016,
      "step": 1817
    },
    {
      "epoch": 0.5609379821042888,
      "grad_norm": 0.3235574960708618,
      "learning_rate": 3e-05,
      "loss": 2.5016,
      "step": 1818
    },
    {
      "epoch": 0.5612465288491206,
      "grad_norm": 0.3079756498336792,
      "learning_rate": 3e-05,
      "loss": 2.6037,
      "step": 1819
    },
    {
      "epoch": 0.5615550755939525,
      "grad_norm": 0.3370635509490967,
      "learning_rate": 3e-05,
      "loss": 2.6834,
      "step": 1820
    },
    {
      "epoch": 0.5618636223387843,
      "grad_norm": 0.3856896162033081,
      "learning_rate": 3e-05,
      "loss": 2.5547,
      "step": 1821
    },
    {
      "epoch": 0.5621721690836161,
      "grad_norm": 0.29247716069221497,
      "learning_rate": 3e-05,
      "loss": 2.3593,
      "step": 1822
    },
    {
      "epoch": 0.562480715828448,
      "grad_norm": 0.2804856300354004,
      "learning_rate": 3e-05,
      "loss": 2.2592,
      "step": 1823
    },
    {
      "epoch": 0.5627892625732799,
      "grad_norm": 0.38057059049606323,
      "learning_rate": 3e-05,
      "loss": 2.7212,
      "step": 1824
    },
    {
      "epoch": 0.5630978093181117,
      "grad_norm": 0.31085407733917236,
      "learning_rate": 3e-05,
      "loss": 2.3131,
      "step": 1825
    },
    {
      "epoch": 0.5634063560629435,
      "grad_norm": 0.30950838327407837,
      "learning_rate": 3e-05,
      "loss": 2.3015,
      "step": 1826
    },
    {
      "epoch": 0.5637149028077754,
      "grad_norm": 0.2703428864479065,
      "learning_rate": 3e-05,
      "loss": 2.4096,
      "step": 1827
    },
    {
      "epoch": 0.5640234495526072,
      "grad_norm": 0.2752431035041809,
      "learning_rate": 3e-05,
      "loss": 2.2907,
      "step": 1828
    },
    {
      "epoch": 0.564331996297439,
      "grad_norm": 0.3768213391304016,
      "learning_rate": 3e-05,
      "loss": 2.706,
      "step": 1829
    },
    {
      "epoch": 0.564640543042271,
      "grad_norm": 0.3653735816478729,
      "learning_rate": 3e-05,
      "loss": 2.6885,
      "step": 1830
    },
    {
      "epoch": 0.5649490897871028,
      "grad_norm": 0.300034761428833,
      "learning_rate": 3e-05,
      "loss": 2.6125,
      "step": 1831
    },
    {
      "epoch": 0.5652576365319346,
      "grad_norm": 0.32249727845191956,
      "learning_rate": 3e-05,
      "loss": 2.6107,
      "step": 1832
    },
    {
      "epoch": 0.5655661832767664,
      "grad_norm": 0.2547013759613037,
      "learning_rate": 3e-05,
      "loss": 2.1702,
      "step": 1833
    },
    {
      "epoch": 0.5658747300215983,
      "grad_norm": 0.3587866723537445,
      "learning_rate": 3e-05,
      "loss": 2.4859,
      "step": 1834
    },
    {
      "epoch": 0.5661832767664301,
      "grad_norm": 0.34305340051651,
      "learning_rate": 3e-05,
      "loss": 2.51,
      "step": 1835
    },
    {
      "epoch": 0.5664918235112619,
      "grad_norm": 0.3700675368309021,
      "learning_rate": 3e-05,
      "loss": 3.0394,
      "step": 1836
    },
    {
      "epoch": 0.5668003702560938,
      "grad_norm": 0.3234868347644806,
      "learning_rate": 3e-05,
      "loss": 2.5686,
      "step": 1837
    },
    {
      "epoch": 0.5671089170009257,
      "grad_norm": 0.39365333318710327,
      "learning_rate": 3e-05,
      "loss": 3.0999,
      "step": 1838
    },
    {
      "epoch": 0.5674174637457575,
      "grad_norm": 0.41990581154823303,
      "learning_rate": 3e-05,
      "loss": 2.8201,
      "step": 1839
    },
    {
      "epoch": 0.5677260104905894,
      "grad_norm": 0.3234783411026001,
      "learning_rate": 3e-05,
      "loss": 2.9017,
      "step": 1840
    },
    {
      "epoch": 0.5680345572354212,
      "grad_norm": 0.4942884147167206,
      "learning_rate": 3e-05,
      "loss": 2.6666,
      "step": 1841
    },
    {
      "epoch": 0.568343103980253,
      "grad_norm": 0.32762080430984497,
      "learning_rate": 3e-05,
      "loss": 2.8604,
      "step": 1842
    },
    {
      "epoch": 0.5686516507250848,
      "grad_norm": 0.32969486713409424,
      "learning_rate": 3e-05,
      "loss": 2.8232,
      "step": 1843
    },
    {
      "epoch": 0.5689601974699167,
      "grad_norm": 0.41623249650001526,
      "learning_rate": 3e-05,
      "loss": 2.7924,
      "step": 1844
    },
    {
      "epoch": 0.5692687442147485,
      "grad_norm": 0.47765466570854187,
      "learning_rate": 3e-05,
      "loss": 3.2767,
      "step": 1845
    },
    {
      "epoch": 0.5695772909595804,
      "grad_norm": 0.3534036874771118,
      "learning_rate": 3e-05,
      "loss": 2.6968,
      "step": 1846
    },
    {
      "epoch": 0.5698858377044123,
      "grad_norm": 0.36876803636550903,
      "learning_rate": 3e-05,
      "loss": 2.7692,
      "step": 1847
    },
    {
      "epoch": 0.5701943844492441,
      "grad_norm": 0.30010735988616943,
      "learning_rate": 3e-05,
      "loss": 2.7339,
      "step": 1848
    },
    {
      "epoch": 0.5705029311940759,
      "grad_norm": 0.41695329546928406,
      "learning_rate": 3e-05,
      "loss": 2.6282,
      "step": 1849
    },
    {
      "epoch": 0.5708114779389077,
      "grad_norm": 0.43772029876708984,
      "learning_rate": 3e-05,
      "loss": 2.8142,
      "step": 1850
    },
    {
      "epoch": 0.5711200246837396,
      "grad_norm": 0.37401148676872253,
      "learning_rate": 3e-05,
      "loss": 2.6245,
      "step": 1851
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.3227090835571289,
      "learning_rate": 3e-05,
      "loss": 2.7928,
      "step": 1852
    },
    {
      "epoch": 0.5717371181734032,
      "grad_norm": 0.319191575050354,
      "learning_rate": 3e-05,
      "loss": 2.478,
      "step": 1853
    },
    {
      "epoch": 0.5720456649182352,
      "grad_norm": 0.3070712983608246,
      "learning_rate": 3e-05,
      "loss": 2.5363,
      "step": 1854
    },
    {
      "epoch": 0.572354211663067,
      "grad_norm": 0.31813663244247437,
      "learning_rate": 3e-05,
      "loss": 2.9719,
      "step": 1855
    },
    {
      "epoch": 0.5726627584078988,
      "grad_norm": 0.3057642877101898,
      "learning_rate": 3e-05,
      "loss": 2.4343,
      "step": 1856
    },
    {
      "epoch": 0.5729713051527306,
      "grad_norm": 0.3192763328552246,
      "learning_rate": 3e-05,
      "loss": 2.6212,
      "step": 1857
    },
    {
      "epoch": 0.5732798518975625,
      "grad_norm": 0.42249295115470886,
      "learning_rate": 3e-05,
      "loss": 2.5999,
      "step": 1858
    },
    {
      "epoch": 0.5735883986423943,
      "grad_norm": 0.29535210132598877,
      "learning_rate": 3e-05,
      "loss": 2.4362,
      "step": 1859
    },
    {
      "epoch": 0.5738969453872261,
      "grad_norm": 0.4919167757034302,
      "learning_rate": 3e-05,
      "loss": 2.7813,
      "step": 1860
    },
    {
      "epoch": 0.574205492132058,
      "grad_norm": 0.320404976606369,
      "learning_rate": 3e-05,
      "loss": 2.6745,
      "step": 1861
    },
    {
      "epoch": 0.5745140388768899,
      "grad_norm": 0.31205329298973083,
      "learning_rate": 3e-05,
      "loss": 2.496,
      "step": 1862
    },
    {
      "epoch": 0.5748225856217217,
      "grad_norm": 0.2652618885040283,
      "learning_rate": 3e-05,
      "loss": 2.2061,
      "step": 1863
    },
    {
      "epoch": 0.5751311323665536,
      "grad_norm": 0.3447934687137604,
      "learning_rate": 3e-05,
      "loss": 2.8457,
      "step": 1864
    },
    {
      "epoch": 0.5754396791113854,
      "grad_norm": 0.323257714509964,
      "learning_rate": 3e-05,
      "loss": 2.7075,
      "step": 1865
    },
    {
      "epoch": 0.5757482258562172,
      "grad_norm": 0.2977539300918579,
      "learning_rate": 3e-05,
      "loss": 2.576,
      "step": 1866
    },
    {
      "epoch": 0.576056772601049,
      "grad_norm": 0.3683079183101654,
      "learning_rate": 3e-05,
      "loss": 2.7179,
      "step": 1867
    },
    {
      "epoch": 0.5763653193458809,
      "grad_norm": 0.4104117155075073,
      "learning_rate": 3e-05,
      "loss": 2.6387,
      "step": 1868
    },
    {
      "epoch": 0.5766738660907127,
      "grad_norm": 0.4164542257785797,
      "learning_rate": 3e-05,
      "loss": 3.1218,
      "step": 1869
    },
    {
      "epoch": 0.5769824128355446,
      "grad_norm": 0.35790860652923584,
      "learning_rate": 3e-05,
      "loss": 3.0051,
      "step": 1870
    },
    {
      "epoch": 0.5772909595803765,
      "grad_norm": 0.38595104217529297,
      "learning_rate": 3e-05,
      "loss": 2.5418,
      "step": 1871
    },
    {
      "epoch": 0.5775995063252083,
      "grad_norm": 0.30152931809425354,
      "learning_rate": 3e-05,
      "loss": 2.6598,
      "step": 1872
    },
    {
      "epoch": 0.5779080530700401,
      "grad_norm": 0.45305702090263367,
      "learning_rate": 3e-05,
      "loss": 3.059,
      "step": 1873
    },
    {
      "epoch": 0.578216599814872,
      "grad_norm": 0.3736101984977722,
      "learning_rate": 3e-05,
      "loss": 2.6267,
      "step": 1874
    },
    {
      "epoch": 0.5785251465597038,
      "grad_norm": 0.3496648371219635,
      "learning_rate": 3e-05,
      "loss": 2.5717,
      "step": 1875
    },
    {
      "epoch": 0.5788336933045356,
      "grad_norm": 0.4360155463218689,
      "learning_rate": 3e-05,
      "loss": 3.1713,
      "step": 1876
    },
    {
      "epoch": 0.5791422400493674,
      "grad_norm": 0.3444565236568451,
      "learning_rate": 3e-05,
      "loss": 2.5872,
      "step": 1877
    },
    {
      "epoch": 0.5794507867941994,
      "grad_norm": 0.3689393103122711,
      "learning_rate": 3e-05,
      "loss": 2.8699,
      "step": 1878
    },
    {
      "epoch": 0.5797593335390312,
      "grad_norm": 0.3451610803604126,
      "learning_rate": 3e-05,
      "loss": 2.7843,
      "step": 1879
    },
    {
      "epoch": 0.580067880283863,
      "grad_norm": 0.32051753997802734,
      "learning_rate": 3e-05,
      "loss": 2.6073,
      "step": 1880
    },
    {
      "epoch": 0.5803764270286949,
      "grad_norm": 0.49073344469070435,
      "learning_rate": 3e-05,
      "loss": 2.6237,
      "step": 1881
    },
    {
      "epoch": 0.5806849737735267,
      "grad_norm": 0.3311246633529663,
      "learning_rate": 3e-05,
      "loss": 2.5781,
      "step": 1882
    },
    {
      "epoch": 0.5809935205183585,
      "grad_norm": 0.31692609190940857,
      "learning_rate": 3e-05,
      "loss": 2.4897,
      "step": 1883
    },
    {
      "epoch": 0.5813020672631903,
      "grad_norm": 0.5018044114112854,
      "learning_rate": 3e-05,
      "loss": 2.7427,
      "step": 1884
    },
    {
      "epoch": 0.5816106140080222,
      "grad_norm": 0.30075499415397644,
      "learning_rate": 3e-05,
      "loss": 2.4889,
      "step": 1885
    },
    {
      "epoch": 0.5819191607528541,
      "grad_norm": 0.36535558104515076,
      "learning_rate": 3e-05,
      "loss": 2.696,
      "step": 1886
    },
    {
      "epoch": 0.5822277074976859,
      "grad_norm": 0.3899480104446411,
      "learning_rate": 3e-05,
      "loss": 2.7502,
      "step": 1887
    },
    {
      "epoch": 0.5825362542425178,
      "grad_norm": 0.3767227232456207,
      "learning_rate": 3e-05,
      "loss": 2.8132,
      "step": 1888
    },
    {
      "epoch": 0.5828448009873496,
      "grad_norm": 0.6538413763046265,
      "learning_rate": 3e-05,
      "loss": 3.192,
      "step": 1889
    },
    {
      "epoch": 0.5831533477321814,
      "grad_norm": 0.4631282389163971,
      "learning_rate": 3e-05,
      "loss": 2.9992,
      "step": 1890
    },
    {
      "epoch": 0.5834618944770132,
      "grad_norm": 0.3080557584762573,
      "learning_rate": 3e-05,
      "loss": 2.6843,
      "step": 1891
    },
    {
      "epoch": 0.5837704412218451,
      "grad_norm": 0.5268321633338928,
      "learning_rate": 3e-05,
      "loss": 2.722,
      "step": 1892
    },
    {
      "epoch": 0.5840789879666769,
      "grad_norm": 0.38069623708724976,
      "learning_rate": 3e-05,
      "loss": 2.6325,
      "step": 1893
    },
    {
      "epoch": 0.5843875347115088,
      "grad_norm": 0.3551689386367798,
      "learning_rate": 3e-05,
      "loss": 2.734,
      "step": 1894
    },
    {
      "epoch": 0.5846960814563407,
      "grad_norm": 0.49280086159706116,
      "learning_rate": 3e-05,
      "loss": 2.7902,
      "step": 1895
    },
    {
      "epoch": 0.5850046282011725,
      "grad_norm": 0.334772527217865,
      "learning_rate": 3e-05,
      "loss": 2.5278,
      "step": 1896
    },
    {
      "epoch": 0.5853131749460043,
      "grad_norm": 0.4307243227958679,
      "learning_rate": 3e-05,
      "loss": 3.0175,
      "step": 1897
    },
    {
      "epoch": 0.5856217216908362,
      "grad_norm": 0.4769163429737091,
      "learning_rate": 3e-05,
      "loss": 2.7155,
      "step": 1898
    },
    {
      "epoch": 0.585930268435668,
      "grad_norm": 0.4567604660987854,
      "learning_rate": 3e-05,
      "loss": 2.6868,
      "step": 1899
    },
    {
      "epoch": 0.5862388151804998,
      "grad_norm": 0.2927672564983368,
      "learning_rate": 3e-05,
      "loss": 2.4188,
      "step": 1900
    },
    {
      "epoch": 0.5865473619253316,
      "grad_norm": 0.35626882314682007,
      "learning_rate": 3e-05,
      "loss": 2.4569,
      "step": 1901
    },
    {
      "epoch": 0.5868559086701636,
      "grad_norm": 0.3931029736995697,
      "learning_rate": 3e-05,
      "loss": 2.704,
      "step": 1902
    },
    {
      "epoch": 0.5871644554149954,
      "grad_norm": 0.3271540105342865,
      "learning_rate": 3e-05,
      "loss": 2.8947,
      "step": 1903
    },
    {
      "epoch": 0.5874730021598272,
      "grad_norm": 0.4138778746128082,
      "learning_rate": 3e-05,
      "loss": 2.9432,
      "step": 1904
    },
    {
      "epoch": 0.5877815489046591,
      "grad_norm": 0.305519163608551,
      "learning_rate": 3e-05,
      "loss": 2.6207,
      "step": 1905
    },
    {
      "epoch": 0.5880900956494909,
      "grad_norm": 0.33586934208869934,
      "learning_rate": 3e-05,
      "loss": 2.5194,
      "step": 1906
    },
    {
      "epoch": 0.5883986423943227,
      "grad_norm": 0.3901278078556061,
      "learning_rate": 3e-05,
      "loss": 2.4353,
      "step": 1907
    },
    {
      "epoch": 0.5887071891391545,
      "grad_norm": 0.28809359669685364,
      "learning_rate": 3e-05,
      "loss": 2.6155,
      "step": 1908
    },
    {
      "epoch": 0.5890157358839864,
      "grad_norm": 0.3274940252304077,
      "learning_rate": 3e-05,
      "loss": 2.2618,
      "step": 1909
    },
    {
      "epoch": 0.5893242826288183,
      "grad_norm": 0.9131669402122498,
      "learning_rate": 3e-05,
      "loss": 2.8335,
      "step": 1910
    },
    {
      "epoch": 0.5896328293736501,
      "grad_norm": 0.3802897036075592,
      "learning_rate": 3e-05,
      "loss": 2.5784,
      "step": 1911
    },
    {
      "epoch": 0.589941376118482,
      "grad_norm": 0.4873533546924591,
      "learning_rate": 3e-05,
      "loss": 2.6349,
      "step": 1912
    },
    {
      "epoch": 0.5902499228633138,
      "grad_norm": 0.33420997858047485,
      "learning_rate": 3e-05,
      "loss": 2.4609,
      "step": 1913
    },
    {
      "epoch": 0.5905584696081456,
      "grad_norm": 0.4681174159049988,
      "learning_rate": 3e-05,
      "loss": 2.4985,
      "step": 1914
    },
    {
      "epoch": 0.5908670163529774,
      "grad_norm": 0.3231395184993744,
      "learning_rate": 3e-05,
      "loss": 2.4141,
      "step": 1915
    },
    {
      "epoch": 0.5911755630978093,
      "grad_norm": 0.37578973174095154,
      "learning_rate": 3e-05,
      "loss": 2.6424,
      "step": 1916
    },
    {
      "epoch": 0.5914841098426412,
      "grad_norm": 0.3429258167743683,
      "learning_rate": 3e-05,
      "loss": 2.3949,
      "step": 1917
    },
    {
      "epoch": 0.591792656587473,
      "grad_norm": 0.42337566614151,
      "learning_rate": 3e-05,
      "loss": 2.5829,
      "step": 1918
    },
    {
      "epoch": 0.5921012033323049,
      "grad_norm": 0.3989293873310089,
      "learning_rate": 3e-05,
      "loss": 2.9044,
      "step": 1919
    },
    {
      "epoch": 0.5924097500771367,
      "grad_norm": 0.351583331823349,
      "learning_rate": 3e-05,
      "loss": 2.5472,
      "step": 1920
    },
    {
      "epoch": 0.5927182968219685,
      "grad_norm": 0.28526926040649414,
      "learning_rate": 3e-05,
      "loss": 2.336,
      "step": 1921
    },
    {
      "epoch": 0.5930268435668004,
      "grad_norm": 0.292166143655777,
      "learning_rate": 3e-05,
      "loss": 2.3399,
      "step": 1922
    },
    {
      "epoch": 0.5933353903116322,
      "grad_norm": 0.3061083257198334,
      "learning_rate": 3e-05,
      "loss": 2.5137,
      "step": 1923
    },
    {
      "epoch": 0.593643937056464,
      "grad_norm": 0.3518201410770416,
      "learning_rate": 3e-05,
      "loss": 2.6903,
      "step": 1924
    },
    {
      "epoch": 0.593952483801296,
      "grad_norm": 0.34315404295921326,
      "learning_rate": 3e-05,
      "loss": 2.7476,
      "step": 1925
    },
    {
      "epoch": 0.5942610305461278,
      "grad_norm": 0.3130031228065491,
      "learning_rate": 3e-05,
      "loss": 2.303,
      "step": 1926
    },
    {
      "epoch": 0.5945695772909596,
      "grad_norm": 0.2985696792602539,
      "learning_rate": 3e-05,
      "loss": 2.1554,
      "step": 1927
    },
    {
      "epoch": 0.5948781240357914,
      "grad_norm": 0.28823089599609375,
      "learning_rate": 3e-05,
      "loss": 2.4544,
      "step": 1928
    },
    {
      "epoch": 0.5951866707806233,
      "grad_norm": 0.2691349387168884,
      "learning_rate": 3e-05,
      "loss": 2.3395,
      "step": 1929
    },
    {
      "epoch": 0.5954952175254551,
      "grad_norm": 0.42032185196876526,
      "learning_rate": 3e-05,
      "loss": 3.0267,
      "step": 1930
    },
    {
      "epoch": 0.5958037642702869,
      "grad_norm": 0.3613561987876892,
      "learning_rate": 3e-05,
      "loss": 2.6168,
      "step": 1931
    },
    {
      "epoch": 0.5961123110151187,
      "grad_norm": 0.34824803471565247,
      "learning_rate": 3e-05,
      "loss": 2.8337,
      "step": 1932
    },
    {
      "epoch": 0.5964208577599507,
      "grad_norm": 0.3117191195487976,
      "learning_rate": 3e-05,
      "loss": 2.7659,
      "step": 1933
    },
    {
      "epoch": 0.5967294045047825,
      "grad_norm": 0.42382311820983887,
      "learning_rate": 3e-05,
      "loss": 3.3237,
      "step": 1934
    },
    {
      "epoch": 0.5970379512496143,
      "grad_norm": 0.3390701711177826,
      "learning_rate": 3e-05,
      "loss": 2.7139,
      "step": 1935
    },
    {
      "epoch": 0.5973464979944462,
      "grad_norm": 0.3503396213054657,
      "learning_rate": 3e-05,
      "loss": 2.9288,
      "step": 1936
    },
    {
      "epoch": 0.597655044739278,
      "grad_norm": 0.3063940405845642,
      "learning_rate": 3e-05,
      "loss": 2.1386,
      "step": 1937
    },
    {
      "epoch": 0.5979635914841098,
      "grad_norm": 0.3633251190185547,
      "learning_rate": 3e-05,
      "loss": 2.7778,
      "step": 1938
    },
    {
      "epoch": 0.5982721382289417,
      "grad_norm": 0.337150514125824,
      "learning_rate": 3e-05,
      "loss": 2.7948,
      "step": 1939
    },
    {
      "epoch": 0.5985806849737735,
      "grad_norm": 0.31596246361732483,
      "learning_rate": 3e-05,
      "loss": 2.5579,
      "step": 1940
    },
    {
      "epoch": 0.5988892317186054,
      "grad_norm": 0.3002220094203949,
      "learning_rate": 3e-05,
      "loss": 2.5221,
      "step": 1941
    },
    {
      "epoch": 0.5991977784634372,
      "grad_norm": 0.3110668957233429,
      "learning_rate": 3e-05,
      "loss": 2.6875,
      "step": 1942
    },
    {
      "epoch": 0.5995063252082691,
      "grad_norm": 0.43362733721733093,
      "learning_rate": 3e-05,
      "loss": 2.8798,
      "step": 1943
    },
    {
      "epoch": 0.5998148719531009,
      "grad_norm": 0.3057228624820709,
      "learning_rate": 3e-05,
      "loss": 2.6287,
      "step": 1944
    },
    {
      "epoch": 0.6001234186979327,
      "grad_norm": 0.42439699172973633,
      "learning_rate": 3e-05,
      "loss": 3.182,
      "step": 1945
    },
    {
      "epoch": 0.6004319654427646,
      "grad_norm": 0.3645000159740448,
      "learning_rate": 3e-05,
      "loss": 3.1542,
      "step": 1946
    },
    {
      "epoch": 0.6007405121875964,
      "grad_norm": 0.32982340455055237,
      "learning_rate": 3e-05,
      "loss": 2.5876,
      "step": 1947
    },
    {
      "epoch": 0.6010490589324282,
      "grad_norm": 0.3214912414550781,
      "learning_rate": 3e-05,
      "loss": 2.392,
      "step": 1948
    },
    {
      "epoch": 0.6013576056772602,
      "grad_norm": 0.39832058548927307,
      "learning_rate": 3e-05,
      "loss": 2.6137,
      "step": 1949
    },
    {
      "epoch": 0.601666152422092,
      "grad_norm": 0.31585508584976196,
      "learning_rate": 3e-05,
      "loss": 2.687,
      "step": 1950
    },
    {
      "epoch": 0.6019746991669238,
      "grad_norm": 0.5885326862335205,
      "learning_rate": 3e-05,
      "loss": 3.2389,
      "step": 1951
    },
    {
      "epoch": 0.6022832459117556,
      "grad_norm": 0.3510866165161133,
      "learning_rate": 3e-05,
      "loss": 2.7795,
      "step": 1952
    },
    {
      "epoch": 0.6025917926565875,
      "grad_norm": 0.29248693585395813,
      "learning_rate": 3e-05,
      "loss": 2.4253,
      "step": 1953
    },
    {
      "epoch": 0.6029003394014193,
      "grad_norm": 0.3324500322341919,
      "learning_rate": 3e-05,
      "loss": 2.4282,
      "step": 1954
    },
    {
      "epoch": 0.6032088861462511,
      "grad_norm": 0.4674556851387024,
      "learning_rate": 3e-05,
      "loss": 3.3254,
      "step": 1955
    },
    {
      "epoch": 0.603517432891083,
      "grad_norm": 0.30010566115379333,
      "learning_rate": 3e-05,
      "loss": 2.3326,
      "step": 1956
    },
    {
      "epoch": 0.6038259796359149,
      "grad_norm": 0.8941420912742615,
      "learning_rate": 3e-05,
      "loss": 2.8991,
      "step": 1957
    },
    {
      "epoch": 0.6041345263807467,
      "grad_norm": 0.3649481534957886,
      "learning_rate": 3e-05,
      "loss": 2.9411,
      "step": 1958
    },
    {
      "epoch": 0.6044430731255785,
      "grad_norm": 0.40504541993141174,
      "learning_rate": 3e-05,
      "loss": 3.1132,
      "step": 1959
    },
    {
      "epoch": 0.6047516198704104,
      "grad_norm": 0.36313414573669434,
      "learning_rate": 3e-05,
      "loss": 2.8904,
      "step": 1960
    },
    {
      "epoch": 0.6050601666152422,
      "grad_norm": 0.5115112662315369,
      "learning_rate": 3e-05,
      "loss": 3.2337,
      "step": 1961
    },
    {
      "epoch": 0.605368713360074,
      "grad_norm": 0.33497315645217896,
      "learning_rate": 3e-05,
      "loss": 2.5356,
      "step": 1962
    },
    {
      "epoch": 0.6056772601049059,
      "grad_norm": 0.38686394691467285,
      "learning_rate": 3e-05,
      "loss": 2.5937,
      "step": 1963
    },
    {
      "epoch": 0.6059858068497377,
      "grad_norm": 0.6908053159713745,
      "learning_rate": 3e-05,
      "loss": 2.828,
      "step": 1964
    },
    {
      "epoch": 0.6062943535945696,
      "grad_norm": 0.31312212347984314,
      "learning_rate": 3e-05,
      "loss": 2.4178,
      "step": 1965
    },
    {
      "epoch": 0.6066029003394015,
      "grad_norm": 0.29186975955963135,
      "learning_rate": 3e-05,
      "loss": 2.1984,
      "step": 1966
    },
    {
      "epoch": 0.6069114470842333,
      "grad_norm": 0.37484097480773926,
      "learning_rate": 3e-05,
      "loss": 2.4074,
      "step": 1967
    },
    {
      "epoch": 0.6072199938290651,
      "grad_norm": 0.32743147015571594,
      "learning_rate": 3e-05,
      "loss": 2.4768,
      "step": 1968
    },
    {
      "epoch": 0.6075285405738969,
      "grad_norm": 0.4841730296611786,
      "learning_rate": 3e-05,
      "loss": 2.5,
      "step": 1969
    },
    {
      "epoch": 0.6078370873187288,
      "grad_norm": 0.2945593297481537,
      "learning_rate": 3e-05,
      "loss": 2.478,
      "step": 1970
    },
    {
      "epoch": 0.6081456340635606,
      "grad_norm": 0.4916343688964844,
      "learning_rate": 3e-05,
      "loss": 2.4608,
      "step": 1971
    },
    {
      "epoch": 0.6084541808083924,
      "grad_norm": 0.4401153028011322,
      "learning_rate": 3e-05,
      "loss": 2.4569,
      "step": 1972
    },
    {
      "epoch": 0.6087627275532244,
      "grad_norm": 0.2683543264865875,
      "learning_rate": 3e-05,
      "loss": 2.4189,
      "step": 1973
    },
    {
      "epoch": 0.6090712742980562,
      "grad_norm": 0.35089823603630066,
      "learning_rate": 3e-05,
      "loss": 2.4496,
      "step": 1974
    },
    {
      "epoch": 0.609379821042888,
      "grad_norm": 0.3995623290538788,
      "learning_rate": 3e-05,
      "loss": 2.6005,
      "step": 1975
    },
    {
      "epoch": 0.6096883677877198,
      "grad_norm": 0.37342095375061035,
      "learning_rate": 3e-05,
      "loss": 2.5541,
      "step": 1976
    },
    {
      "epoch": 0.6099969145325517,
      "grad_norm": 0.3966485857963562,
      "learning_rate": 3e-05,
      "loss": 2.7672,
      "step": 1977
    },
    {
      "epoch": 0.6103054612773835,
      "grad_norm": 0.33932629227638245,
      "learning_rate": 3e-05,
      "loss": 2.3939,
      "step": 1978
    },
    {
      "epoch": 0.6106140080222153,
      "grad_norm": 0.3646182119846344,
      "learning_rate": 3e-05,
      "loss": 2.8896,
      "step": 1979
    },
    {
      "epoch": 0.6109225547670472,
      "grad_norm": 0.3388557434082031,
      "learning_rate": 3e-05,
      "loss": 2.5835,
      "step": 1980
    },
    {
      "epoch": 0.6112311015118791,
      "grad_norm": 0.3514586091041565,
      "learning_rate": 3e-05,
      "loss": 2.8596,
      "step": 1981
    },
    {
      "epoch": 0.6115396482567109,
      "grad_norm": 0.3328869044780731,
      "learning_rate": 3e-05,
      "loss": 2.6065,
      "step": 1982
    },
    {
      "epoch": 0.6118481950015427,
      "grad_norm": 0.3014579713344574,
      "learning_rate": 3e-05,
      "loss": 2.3762,
      "step": 1983
    },
    {
      "epoch": 0.6121567417463746,
      "grad_norm": 0.3286787271499634,
      "learning_rate": 3e-05,
      "loss": 2.819,
      "step": 1984
    },
    {
      "epoch": 0.6124652884912064,
      "grad_norm": 0.30594438314437866,
      "learning_rate": 3e-05,
      "loss": 2.2396,
      "step": 1985
    },
    {
      "epoch": 0.6127738352360382,
      "grad_norm": 0.34099289774894714,
      "learning_rate": 3e-05,
      "loss": 2.9636,
      "step": 1986
    },
    {
      "epoch": 0.6130823819808701,
      "grad_norm": 0.3839113414287567,
      "learning_rate": 3e-05,
      "loss": 2.8976,
      "step": 1987
    },
    {
      "epoch": 0.6133909287257019,
      "grad_norm": 0.3939029574394226,
      "learning_rate": 3e-05,
      "loss": 2.964,
      "step": 1988
    },
    {
      "epoch": 0.6136994754705338,
      "grad_norm": 0.31318363547325134,
      "learning_rate": 3e-05,
      "loss": 2.5215,
      "step": 1989
    },
    {
      "epoch": 0.6140080222153657,
      "grad_norm": 0.44314906001091003,
      "learning_rate": 3e-05,
      "loss": 2.7397,
      "step": 1990
    },
    {
      "epoch": 0.6143165689601975,
      "grad_norm": 0.2839794158935547,
      "learning_rate": 3e-05,
      "loss": 2.6783,
      "step": 1991
    },
    {
      "epoch": 0.6146251157050293,
      "grad_norm": 0.4167096018791199,
      "learning_rate": 3e-05,
      "loss": 2.8253,
      "step": 1992
    },
    {
      "epoch": 0.6149336624498611,
      "grad_norm": 0.48042523860931396,
      "learning_rate": 3e-05,
      "loss": 2.6572,
      "step": 1993
    },
    {
      "epoch": 0.615242209194693,
      "grad_norm": 0.3192354142665863,
      "learning_rate": 3e-05,
      "loss": 2.5442,
      "step": 1994
    },
    {
      "epoch": 0.6155507559395248,
      "grad_norm": 0.3579767048358917,
      "learning_rate": 3e-05,
      "loss": 2.8193,
      "step": 1995
    },
    {
      "epoch": 0.6158593026843567,
      "grad_norm": 0.4392094314098358,
      "learning_rate": 3e-05,
      "loss": 2.7971,
      "step": 1996
    },
    {
      "epoch": 0.6161678494291886,
      "grad_norm": 0.383989155292511,
      "learning_rate": 3e-05,
      "loss": 2.8751,
      "step": 1997
    },
    {
      "epoch": 0.6164763961740204,
      "grad_norm": 0.36968907713890076,
      "learning_rate": 3e-05,
      "loss": 3.19,
      "step": 1998
    },
    {
      "epoch": 0.6167849429188522,
      "grad_norm": 0.4232364594936371,
      "learning_rate": 3e-05,
      "loss": 3.0788,
      "step": 1999
    },
    {
      "epoch": 0.617093489663684,
      "grad_norm": 0.41558605432510376,
      "learning_rate": 3e-05,
      "loss": 3.3741,
      "step": 2000
    },
    {
      "epoch": 0.6174020364085159,
      "grad_norm": 0.3285547196865082,
      "learning_rate": 3e-05,
      "loss": 2.6336,
      "step": 2001
    },
    {
      "epoch": 0.6177105831533477,
      "grad_norm": 0.311625212430954,
      "learning_rate": 3e-05,
      "loss": 2.6533,
      "step": 2002
    },
    {
      "epoch": 0.6180191298981795,
      "grad_norm": 0.4083147644996643,
      "learning_rate": 3e-05,
      "loss": 3.1001,
      "step": 2003
    },
    {
      "epoch": 0.6183276766430115,
      "grad_norm": 0.44464874267578125,
      "learning_rate": 3e-05,
      "loss": 2.9055,
      "step": 2004
    },
    {
      "epoch": 0.6186362233878433,
      "grad_norm": 0.2967284619808197,
      "learning_rate": 3e-05,
      "loss": 2.8229,
      "step": 2005
    },
    {
      "epoch": 0.6189447701326751,
      "grad_norm": 0.3772972822189331,
      "learning_rate": 3e-05,
      "loss": 2.5243,
      "step": 2006
    },
    {
      "epoch": 0.619253316877507,
      "grad_norm": 0.3279479146003723,
      "learning_rate": 3e-05,
      "loss": 2.9745,
      "step": 2007
    },
    {
      "epoch": 0.6195618636223388,
      "grad_norm": 0.3854341208934784,
      "learning_rate": 3e-05,
      "loss": 2.936,
      "step": 2008
    },
    {
      "epoch": 0.6198704103671706,
      "grad_norm": 0.355197548866272,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 2009
    },
    {
      "epoch": 0.6201789571120024,
      "grad_norm": 0.3269404172897339,
      "learning_rate": 3e-05,
      "loss": 2.3624,
      "step": 2010
    },
    {
      "epoch": 0.6204875038568343,
      "grad_norm": 0.40607500076293945,
      "learning_rate": 3e-05,
      "loss": 2.9721,
      "step": 2011
    },
    {
      "epoch": 0.6207960506016662,
      "grad_norm": 0.3014424741268158,
      "learning_rate": 3e-05,
      "loss": 2.641,
      "step": 2012
    },
    {
      "epoch": 0.621104597346498,
      "grad_norm": 0.3271016478538513,
      "learning_rate": 3e-05,
      "loss": 2.8161,
      "step": 2013
    },
    {
      "epoch": 0.6214131440913299,
      "grad_norm": 0.3622637689113617,
      "learning_rate": 3e-05,
      "loss": 3.0245,
      "step": 2014
    },
    {
      "epoch": 0.6217216908361617,
      "grad_norm": 0.33080458641052246,
      "learning_rate": 3e-05,
      "loss": 2.3558,
      "step": 2015
    },
    {
      "epoch": 0.6220302375809935,
      "grad_norm": 0.2808302938938141,
      "learning_rate": 3e-05,
      "loss": 2.3971,
      "step": 2016
    },
    {
      "epoch": 0.6223387843258253,
      "grad_norm": 0.3635484576225281,
      "learning_rate": 3e-05,
      "loss": 2.6886,
      "step": 2017
    },
    {
      "epoch": 0.6226473310706572,
      "grad_norm": 0.33641016483306885,
      "learning_rate": 3e-05,
      "loss": 2.6517,
      "step": 2018
    },
    {
      "epoch": 0.622955877815489,
      "grad_norm": 0.3255454897880554,
      "learning_rate": 3e-05,
      "loss": 2.7133,
      "step": 2019
    },
    {
      "epoch": 0.6232644245603209,
      "grad_norm": 0.46022775769233704,
      "learning_rate": 3e-05,
      "loss": 2.9981,
      "step": 2020
    },
    {
      "epoch": 0.6235729713051528,
      "grad_norm": 0.26282113790512085,
      "learning_rate": 3e-05,
      "loss": 2.0888,
      "step": 2021
    },
    {
      "epoch": 0.6238815180499846,
      "grad_norm": 0.3235068619251251,
      "learning_rate": 3e-05,
      "loss": 2.5061,
      "step": 2022
    },
    {
      "epoch": 0.6241900647948164,
      "grad_norm": 0.41372352838516235,
      "learning_rate": 3e-05,
      "loss": 2.9735,
      "step": 2023
    },
    {
      "epoch": 0.6244986115396483,
      "grad_norm": 0.3517864942550659,
      "learning_rate": 3e-05,
      "loss": 2.5726,
      "step": 2024
    },
    {
      "epoch": 0.6248071582844801,
      "grad_norm": 0.38335707783699036,
      "learning_rate": 3e-05,
      "loss": 2.7474,
      "step": 2025
    },
    {
      "epoch": 0.6251157050293119,
      "grad_norm": 0.3006962537765503,
      "learning_rate": 3e-05,
      "loss": 2.5282,
      "step": 2026
    },
    {
      "epoch": 0.6254242517741437,
      "grad_norm": 0.47257086634635925,
      "learning_rate": 3e-05,
      "loss": 3.1522,
      "step": 2027
    },
    {
      "epoch": 0.6257327985189757,
      "grad_norm": 0.3759831488132477,
      "learning_rate": 3e-05,
      "loss": 2.8287,
      "step": 2028
    },
    {
      "epoch": 0.6260413452638075,
      "grad_norm": 0.27641305327415466,
      "learning_rate": 3e-05,
      "loss": 2.0199,
      "step": 2029
    },
    {
      "epoch": 0.6263498920086393,
      "grad_norm": 0.3770776391029358,
      "learning_rate": 3e-05,
      "loss": 3.1056,
      "step": 2030
    },
    {
      "epoch": 0.6266584387534712,
      "grad_norm": 0.32263755798339844,
      "learning_rate": 3e-05,
      "loss": 2.6827,
      "step": 2031
    },
    {
      "epoch": 0.626966985498303,
      "grad_norm": 0.29439646005630493,
      "learning_rate": 3e-05,
      "loss": 2.7194,
      "step": 2032
    },
    {
      "epoch": 0.6272755322431348,
      "grad_norm": 0.35637763142585754,
      "learning_rate": 3e-05,
      "loss": 2.4663,
      "step": 2033
    },
    {
      "epoch": 0.6275840789879666,
      "grad_norm": 0.30794280767440796,
      "learning_rate": 3e-05,
      "loss": 2.5098,
      "step": 2034
    },
    {
      "epoch": 0.6278926257327985,
      "grad_norm": 0.4070144593715668,
      "learning_rate": 3e-05,
      "loss": 2.6898,
      "step": 2035
    },
    {
      "epoch": 0.6282011724776304,
      "grad_norm": 0.3939863443374634,
      "learning_rate": 3e-05,
      "loss": 2.9802,
      "step": 2036
    },
    {
      "epoch": 0.6285097192224622,
      "grad_norm": 0.338907390832901,
      "learning_rate": 3e-05,
      "loss": 2.8044,
      "step": 2037
    },
    {
      "epoch": 0.6288182659672941,
      "grad_norm": 0.30836790800094604,
      "learning_rate": 3e-05,
      "loss": 2.4929,
      "step": 2038
    },
    {
      "epoch": 0.6291268127121259,
      "grad_norm": 0.3368026614189148,
      "learning_rate": 3e-05,
      "loss": 2.291,
      "step": 2039
    },
    {
      "epoch": 0.6294353594569577,
      "grad_norm": 0.290962278842926,
      "learning_rate": 3e-05,
      "loss": 2.4745,
      "step": 2040
    },
    {
      "epoch": 0.6297439062017895,
      "grad_norm": 0.41740748286247253,
      "learning_rate": 3e-05,
      "loss": 2.7949,
      "step": 2041
    },
    {
      "epoch": 0.6300524529466214,
      "grad_norm": 0.338375449180603,
      "learning_rate": 3e-05,
      "loss": 2.8744,
      "step": 2042
    },
    {
      "epoch": 0.6303609996914532,
      "grad_norm": 0.35139888525009155,
      "learning_rate": 3e-05,
      "loss": 2.7917,
      "step": 2043
    },
    {
      "epoch": 0.6306695464362851,
      "grad_norm": 0.2601359486579895,
      "learning_rate": 3e-05,
      "loss": 2.1908,
      "step": 2044
    },
    {
      "epoch": 0.630978093181117,
      "grad_norm": 0.46779265999794006,
      "learning_rate": 3e-05,
      "loss": 3.4857,
      "step": 2045
    },
    {
      "epoch": 0.6312866399259488,
      "grad_norm": 0.3154003322124481,
      "learning_rate": 3e-05,
      "loss": 2.4742,
      "step": 2046
    },
    {
      "epoch": 0.6315951866707806,
      "grad_norm": 0.3528044521808624,
      "learning_rate": 3e-05,
      "loss": 2.511,
      "step": 2047
    },
    {
      "epoch": 0.6319037334156125,
      "grad_norm": 0.33536121249198914,
      "learning_rate": 3e-05,
      "loss": 2.5747,
      "step": 2048
    },
    {
      "epoch": 0.6322122801604443,
      "grad_norm": 0.2813878357410431,
      "learning_rate": 3e-05,
      "loss": 2.7885,
      "step": 2049
    },
    {
      "epoch": 0.6325208269052761,
      "grad_norm": 0.37101298570632935,
      "learning_rate": 3e-05,
      "loss": 2.4285,
      "step": 2050
    },
    {
      "epoch": 0.6328293736501079,
      "grad_norm": 0.36067718267440796,
      "learning_rate": 3e-05,
      "loss": 2.4265,
      "step": 2051
    },
    {
      "epoch": 0.6331379203949399,
      "grad_norm": 0.2749241292476654,
      "learning_rate": 3e-05,
      "loss": 2.348,
      "step": 2052
    },
    {
      "epoch": 0.6334464671397717,
      "grad_norm": 0.37541571259498596,
      "learning_rate": 3e-05,
      "loss": 2.9153,
      "step": 2053
    },
    {
      "epoch": 0.6337550138846035,
      "grad_norm": 0.3796222507953644,
      "learning_rate": 3e-05,
      "loss": 2.5818,
      "step": 2054
    },
    {
      "epoch": 0.6340635606294354,
      "grad_norm": 0.36146241426467896,
      "learning_rate": 3e-05,
      "loss": 2.6657,
      "step": 2055
    },
    {
      "epoch": 0.6343721073742672,
      "grad_norm": 0.37858709692955017,
      "learning_rate": 3e-05,
      "loss": 2.8254,
      "step": 2056
    },
    {
      "epoch": 0.634680654119099,
      "grad_norm": 0.38670918345451355,
      "learning_rate": 3e-05,
      "loss": 2.6653,
      "step": 2057
    },
    {
      "epoch": 0.6349892008639308,
      "grad_norm": 0.3363918960094452,
      "learning_rate": 3e-05,
      "loss": 2.6632,
      "step": 2058
    },
    {
      "epoch": 0.6352977476087627,
      "grad_norm": 0.23550312221050262,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 2059
    },
    {
      "epoch": 0.6356062943535946,
      "grad_norm": 0.332854688167572,
      "learning_rate": 3e-05,
      "loss": 2.6517,
      "step": 2060
    },
    {
      "epoch": 0.6359148410984264,
      "grad_norm": 0.3064941167831421,
      "learning_rate": 3e-05,
      "loss": 2.746,
      "step": 2061
    },
    {
      "epoch": 0.6362233878432583,
      "grad_norm": 0.40314847230911255,
      "learning_rate": 3e-05,
      "loss": 2.5667,
      "step": 2062
    },
    {
      "epoch": 0.6365319345880901,
      "grad_norm": 0.36766573786735535,
      "learning_rate": 3e-05,
      "loss": 2.8295,
      "step": 2063
    },
    {
      "epoch": 0.6368404813329219,
      "grad_norm": 0.4863782227039337,
      "learning_rate": 3e-05,
      "loss": 3.1202,
      "step": 2064
    },
    {
      "epoch": 0.6371490280777538,
      "grad_norm": 0.35342922806739807,
      "learning_rate": 3e-05,
      "loss": 2.8253,
      "step": 2065
    },
    {
      "epoch": 0.6374575748225856,
      "grad_norm": 0.2922891676425934,
      "learning_rate": 3e-05,
      "loss": 2.1615,
      "step": 2066
    },
    {
      "epoch": 0.6377661215674174,
      "grad_norm": 0.4159747362136841,
      "learning_rate": 3e-05,
      "loss": 2.7532,
      "step": 2067
    },
    {
      "epoch": 0.6380746683122493,
      "grad_norm": 0.24633346498012543,
      "learning_rate": 3e-05,
      "loss": 2.3447,
      "step": 2068
    },
    {
      "epoch": 0.6383832150570812,
      "grad_norm": 0.48890185356140137,
      "learning_rate": 3e-05,
      "loss": 2.9833,
      "step": 2069
    },
    {
      "epoch": 0.638691761801913,
      "grad_norm": 0.49807408452033997,
      "learning_rate": 3e-05,
      "loss": 3.0582,
      "step": 2070
    },
    {
      "epoch": 0.6390003085467448,
      "grad_norm": 0.3025662899017334,
      "learning_rate": 3e-05,
      "loss": 2.3173,
      "step": 2071
    },
    {
      "epoch": 0.6393088552915767,
      "grad_norm": 0.2955687642097473,
      "learning_rate": 3e-05,
      "loss": 2.4217,
      "step": 2072
    },
    {
      "epoch": 0.6396174020364085,
      "grad_norm": 0.44148918986320496,
      "learning_rate": 3e-05,
      "loss": 2.7124,
      "step": 2073
    },
    {
      "epoch": 0.6399259487812403,
      "grad_norm": 0.39978307485580444,
      "learning_rate": 3e-05,
      "loss": 2.7428,
      "step": 2074
    },
    {
      "epoch": 0.6402344955260721,
      "grad_norm": 0.2851976454257965,
      "learning_rate": 3e-05,
      "loss": 2.4972,
      "step": 2075
    },
    {
      "epoch": 0.6405430422709041,
      "grad_norm": 0.33834779262542725,
      "learning_rate": 3e-05,
      "loss": 2.503,
      "step": 2076
    },
    {
      "epoch": 0.6408515890157359,
      "grad_norm": 0.4237276017665863,
      "learning_rate": 3e-05,
      "loss": 2.973,
      "step": 2077
    },
    {
      "epoch": 0.6411601357605677,
      "grad_norm": 0.4513942301273346,
      "learning_rate": 3e-05,
      "loss": 3.0904,
      "step": 2078
    },
    {
      "epoch": 0.6414686825053996,
      "grad_norm": 0.35915786027908325,
      "learning_rate": 3e-05,
      "loss": 2.4066,
      "step": 2079
    },
    {
      "epoch": 0.6417772292502314,
      "grad_norm": 0.3897479772567749,
      "learning_rate": 3e-05,
      "loss": 2.8671,
      "step": 2080
    },
    {
      "epoch": 0.6420857759950632,
      "grad_norm": 0.3450873792171478,
      "learning_rate": 3e-05,
      "loss": 2.9807,
      "step": 2081
    },
    {
      "epoch": 0.642394322739895,
      "grad_norm": 0.3396153151988983,
      "learning_rate": 3e-05,
      "loss": 2.7134,
      "step": 2082
    },
    {
      "epoch": 0.642702869484727,
      "grad_norm": 0.41863974928855896,
      "learning_rate": 3e-05,
      "loss": 2.6853,
      "step": 2083
    },
    {
      "epoch": 0.6430114162295588,
      "grad_norm": 0.5577157735824585,
      "learning_rate": 3e-05,
      "loss": 2.989,
      "step": 2084
    },
    {
      "epoch": 0.6433199629743906,
      "grad_norm": 0.4542025327682495,
      "learning_rate": 3e-05,
      "loss": 3.2736,
      "step": 2085
    },
    {
      "epoch": 0.6436285097192225,
      "grad_norm": 0.3049847185611725,
      "learning_rate": 3e-05,
      "loss": 2.6011,
      "step": 2086
    },
    {
      "epoch": 0.6439370564640543,
      "grad_norm": 0.5583184957504272,
      "learning_rate": 3e-05,
      "loss": 2.9329,
      "step": 2087
    },
    {
      "epoch": 0.6442456032088861,
      "grad_norm": 0.31898394227027893,
      "learning_rate": 3e-05,
      "loss": 2.7457,
      "step": 2088
    },
    {
      "epoch": 0.644554149953718,
      "grad_norm": 0.3565310537815094,
      "learning_rate": 3e-05,
      "loss": 2.7388,
      "step": 2089
    },
    {
      "epoch": 0.6448626966985498,
      "grad_norm": 0.2690911889076233,
      "learning_rate": 3e-05,
      "loss": 2.061,
      "step": 2090
    },
    {
      "epoch": 0.6451712434433817,
      "grad_norm": 0.339018851518631,
      "learning_rate": 3e-05,
      "loss": 2.5711,
      "step": 2091
    },
    {
      "epoch": 0.6454797901882136,
      "grad_norm": 0.314058780670166,
      "learning_rate": 3e-05,
      "loss": 2.5641,
      "step": 2092
    },
    {
      "epoch": 0.6457883369330454,
      "grad_norm": 0.28353381156921387,
      "learning_rate": 3e-05,
      "loss": 2.3022,
      "step": 2093
    },
    {
      "epoch": 0.6460968836778772,
      "grad_norm": 0.4895442724227905,
      "learning_rate": 3e-05,
      "loss": 2.861,
      "step": 2094
    },
    {
      "epoch": 0.646405430422709,
      "grad_norm": 0.3713444769382477,
      "learning_rate": 3e-05,
      "loss": 2.5318,
      "step": 2095
    },
    {
      "epoch": 0.6467139771675409,
      "grad_norm": 0.2743374705314636,
      "learning_rate": 3e-05,
      "loss": 2.2817,
      "step": 2096
    },
    {
      "epoch": 0.6470225239123727,
      "grad_norm": 0.3508983254432678,
      "learning_rate": 3e-05,
      "loss": 2.4427,
      "step": 2097
    },
    {
      "epoch": 0.6473310706572045,
      "grad_norm": 0.2961727976799011,
      "learning_rate": 3e-05,
      "loss": 2.2046,
      "step": 2098
    },
    {
      "epoch": 0.6476396174020365,
      "grad_norm": 0.32885807752609253,
      "learning_rate": 3e-05,
      "loss": 2.6128,
      "step": 2099
    },
    {
      "epoch": 0.6479481641468683,
      "grad_norm": 0.34111347794532776,
      "learning_rate": 3e-05,
      "loss": 2.4245,
      "step": 2100
    },
    {
      "epoch": 0.6482567108917001,
      "grad_norm": 0.3031350076198578,
      "learning_rate": 3e-05,
      "loss": 2.7567,
      "step": 2101
    },
    {
      "epoch": 0.6485652576365319,
      "grad_norm": 0.3701673746109009,
      "learning_rate": 3e-05,
      "loss": 3.1212,
      "step": 2102
    },
    {
      "epoch": 0.6488738043813638,
      "grad_norm": 0.3395228683948517,
      "learning_rate": 3e-05,
      "loss": 2.7044,
      "step": 2103
    },
    {
      "epoch": 0.6491823511261956,
      "grad_norm": 0.40879321098327637,
      "learning_rate": 3e-05,
      "loss": 2.6475,
      "step": 2104
    },
    {
      "epoch": 0.6494908978710274,
      "grad_norm": 0.41927212476730347,
      "learning_rate": 3e-05,
      "loss": 2.9938,
      "step": 2105
    },
    {
      "epoch": 0.6497994446158593,
      "grad_norm": 0.2867232859134674,
      "learning_rate": 3e-05,
      "loss": 2.229,
      "step": 2106
    },
    {
      "epoch": 0.6501079913606912,
      "grad_norm": 0.41298848390579224,
      "learning_rate": 3e-05,
      "loss": 2.9525,
      "step": 2107
    },
    {
      "epoch": 0.650416538105523,
      "grad_norm": 0.38462120294570923,
      "learning_rate": 3e-05,
      "loss": 2.833,
      "step": 2108
    },
    {
      "epoch": 0.6507250848503549,
      "grad_norm": 0.2860693633556366,
      "learning_rate": 3e-05,
      "loss": 2.6447,
      "step": 2109
    },
    {
      "epoch": 0.6510336315951867,
      "grad_norm": 0.3113842308521271,
      "learning_rate": 3e-05,
      "loss": 2.4201,
      "step": 2110
    },
    {
      "epoch": 0.6513421783400185,
      "grad_norm": 0.27946069836616516,
      "learning_rate": 3e-05,
      "loss": 2.4673,
      "step": 2111
    },
    {
      "epoch": 0.6516507250848503,
      "grad_norm": 0.2512831687927246,
      "learning_rate": 3e-05,
      "loss": 2.2427,
      "step": 2112
    },
    {
      "epoch": 0.6519592718296822,
      "grad_norm": 0.5824667811393738,
      "learning_rate": 3e-05,
      "loss": 2.8793,
      "step": 2113
    },
    {
      "epoch": 0.652267818574514,
      "grad_norm": 0.6139059066772461,
      "learning_rate": 3e-05,
      "loss": 3.2765,
      "step": 2114
    },
    {
      "epoch": 0.6525763653193459,
      "grad_norm": 0.3947453200817108,
      "learning_rate": 3e-05,
      "loss": 2.6216,
      "step": 2115
    },
    {
      "epoch": 0.6528849120641778,
      "grad_norm": 0.2993795871734619,
      "learning_rate": 3e-05,
      "loss": 2.2209,
      "step": 2116
    },
    {
      "epoch": 0.6531934588090096,
      "grad_norm": 0.33383437991142273,
      "learning_rate": 3e-05,
      "loss": 2.427,
      "step": 2117
    },
    {
      "epoch": 0.6535020055538414,
      "grad_norm": 0.46814224123954773,
      "learning_rate": 3e-05,
      "loss": 2.6804,
      "step": 2118
    },
    {
      "epoch": 0.6538105522986732,
      "grad_norm": 0.35672876238822937,
      "learning_rate": 3e-05,
      "loss": 2.8386,
      "step": 2119
    },
    {
      "epoch": 0.6541190990435051,
      "grad_norm": 0.44050756096839905,
      "learning_rate": 3e-05,
      "loss": 2.8095,
      "step": 2120
    },
    {
      "epoch": 0.6544276457883369,
      "grad_norm": 0.3837873339653015,
      "learning_rate": 3e-05,
      "loss": 2.6712,
      "step": 2121
    },
    {
      "epoch": 0.6547361925331687,
      "grad_norm": 0.44696587324142456,
      "learning_rate": 3e-05,
      "loss": 2.8583,
      "step": 2122
    },
    {
      "epoch": 0.6550447392780007,
      "grad_norm": 0.3543272912502289,
      "learning_rate": 3e-05,
      "loss": 2.6896,
      "step": 2123
    },
    {
      "epoch": 0.6553532860228325,
      "grad_norm": 0.3472782373428345,
      "learning_rate": 3e-05,
      "loss": 2.7716,
      "step": 2124
    },
    {
      "epoch": 0.6556618327676643,
      "grad_norm": 0.3943750858306885,
      "learning_rate": 3e-05,
      "loss": 2.6629,
      "step": 2125
    },
    {
      "epoch": 0.6559703795124961,
      "grad_norm": 0.37474364042282104,
      "learning_rate": 3e-05,
      "loss": 2.7249,
      "step": 2126
    },
    {
      "epoch": 0.656278926257328,
      "grad_norm": 0.32506734132766724,
      "learning_rate": 3e-05,
      "loss": 2.7404,
      "step": 2127
    },
    {
      "epoch": 0.6565874730021598,
      "grad_norm": 0.30232688784599304,
      "learning_rate": 3e-05,
      "loss": 2.6386,
      "step": 2128
    },
    {
      "epoch": 0.6568960197469916,
      "grad_norm": 0.3200061321258545,
      "learning_rate": 3e-05,
      "loss": 3.0038,
      "step": 2129
    },
    {
      "epoch": 0.6572045664918235,
      "grad_norm": 0.43252453207969666,
      "learning_rate": 3e-05,
      "loss": 2.8757,
      "step": 2130
    },
    {
      "epoch": 0.6575131132366554,
      "grad_norm": 0.3240208923816681,
      "learning_rate": 3e-05,
      "loss": 2.5438,
      "step": 2131
    },
    {
      "epoch": 0.6578216599814872,
      "grad_norm": 0.38468319177627563,
      "learning_rate": 3e-05,
      "loss": 3.0328,
      "step": 2132
    },
    {
      "epoch": 0.6581302067263191,
      "grad_norm": 0.28264886140823364,
      "learning_rate": 3e-05,
      "loss": 2.3216,
      "step": 2133
    },
    {
      "epoch": 0.6584387534711509,
      "grad_norm": 0.3538890480995178,
      "learning_rate": 3e-05,
      "loss": 3.0705,
      "step": 2134
    },
    {
      "epoch": 0.6587473002159827,
      "grad_norm": 0.29800945520401,
      "learning_rate": 3e-05,
      "loss": 2.5841,
      "step": 2135
    },
    {
      "epoch": 0.6590558469608145,
      "grad_norm": 0.39261680841445923,
      "learning_rate": 3e-05,
      "loss": 2.4659,
      "step": 2136
    },
    {
      "epoch": 0.6593643937056464,
      "grad_norm": 0.25242921710014343,
      "learning_rate": 3e-05,
      "loss": 2.4418,
      "step": 2137
    },
    {
      "epoch": 0.6596729404504782,
      "grad_norm": 0.35038331151008606,
      "learning_rate": 3e-05,
      "loss": 2.399,
      "step": 2138
    },
    {
      "epoch": 0.6599814871953101,
      "grad_norm": 0.36697402596473694,
      "learning_rate": 3e-05,
      "loss": 2.377,
      "step": 2139
    },
    {
      "epoch": 0.660290033940142,
      "grad_norm": 0.3055066466331482,
      "learning_rate": 3e-05,
      "loss": 2.4264,
      "step": 2140
    },
    {
      "epoch": 0.6605985806849738,
      "grad_norm": 0.3446613848209381,
      "learning_rate": 3e-05,
      "loss": 2.5709,
      "step": 2141
    },
    {
      "epoch": 0.6609071274298056,
      "grad_norm": 0.49223238229751587,
      "learning_rate": 3e-05,
      "loss": 2.8673,
      "step": 2142
    },
    {
      "epoch": 0.6612156741746374,
      "grad_norm": 0.29385796189308167,
      "learning_rate": 3e-05,
      "loss": 2.4047,
      "step": 2143
    },
    {
      "epoch": 0.6615242209194693,
      "grad_norm": 0.3069056570529938,
      "learning_rate": 3e-05,
      "loss": 2.6242,
      "step": 2144
    },
    {
      "epoch": 0.6618327676643011,
      "grad_norm": 0.3820006251335144,
      "learning_rate": 3e-05,
      "loss": 2.9872,
      "step": 2145
    },
    {
      "epoch": 0.6621413144091329,
      "grad_norm": 0.27881932258605957,
      "learning_rate": 3e-05,
      "loss": 2.3606,
      "step": 2146
    },
    {
      "epoch": 0.6624498611539649,
      "grad_norm": 0.4184323847293854,
      "learning_rate": 3e-05,
      "loss": 3.138,
      "step": 2147
    },
    {
      "epoch": 0.6627584078987967,
      "grad_norm": 0.2955660820007324,
      "learning_rate": 3e-05,
      "loss": 2.3884,
      "step": 2148
    },
    {
      "epoch": 0.6630669546436285,
      "grad_norm": 0.31548479199409485,
      "learning_rate": 3e-05,
      "loss": 2.3839,
      "step": 2149
    },
    {
      "epoch": 0.6633755013884604,
      "grad_norm": 0.3798123896121979,
      "learning_rate": 3e-05,
      "loss": 2.5527,
      "step": 2150
    },
    {
      "epoch": 0.6636840481332922,
      "grad_norm": 0.31726929545402527,
      "learning_rate": 3e-05,
      "loss": 2.4075,
      "step": 2151
    },
    {
      "epoch": 0.663992594878124,
      "grad_norm": 0.25782397389411926,
      "learning_rate": 3e-05,
      "loss": 1.9784,
      "step": 2152
    },
    {
      "epoch": 0.6643011416229558,
      "grad_norm": 0.23787148296833038,
      "learning_rate": 3e-05,
      "loss": 2.0268,
      "step": 2153
    },
    {
      "epoch": 0.6646096883677877,
      "grad_norm": 0.43553170561790466,
      "learning_rate": 3e-05,
      "loss": 2.7882,
      "step": 2154
    },
    {
      "epoch": 0.6649182351126196,
      "grad_norm": 0.3706422448158264,
      "learning_rate": 3e-05,
      "loss": 2.8122,
      "step": 2155
    },
    {
      "epoch": 0.6652267818574514,
      "grad_norm": 0.3282051682472229,
      "learning_rate": 3e-05,
      "loss": 2.22,
      "step": 2156
    },
    {
      "epoch": 0.6655353286022833,
      "grad_norm": 0.3240806460380554,
      "learning_rate": 3e-05,
      "loss": 2.84,
      "step": 2157
    },
    {
      "epoch": 0.6658438753471151,
      "grad_norm": 0.3654410243034363,
      "learning_rate": 3e-05,
      "loss": 2.6916,
      "step": 2158
    },
    {
      "epoch": 0.6661524220919469,
      "grad_norm": 0.36625534296035767,
      "learning_rate": 3e-05,
      "loss": 2.4161,
      "step": 2159
    },
    {
      "epoch": 0.6664609688367787,
      "grad_norm": 0.30736368894577026,
      "learning_rate": 3e-05,
      "loss": 2.566,
      "step": 2160
    },
    {
      "epoch": 0.6667695155816106,
      "grad_norm": 0.3654194176197052,
      "learning_rate": 3e-05,
      "loss": 2.8405,
      "step": 2161
    },
    {
      "epoch": 0.6670780623264425,
      "grad_norm": 0.319302499294281,
      "learning_rate": 3e-05,
      "loss": 2.8201,
      "step": 2162
    },
    {
      "epoch": 0.6673866090712743,
      "grad_norm": 0.41080933809280396,
      "learning_rate": 3e-05,
      "loss": 2.881,
      "step": 2163
    },
    {
      "epoch": 0.6676951558161062,
      "grad_norm": 0.2875690162181854,
      "learning_rate": 3e-05,
      "loss": 2.4054,
      "step": 2164
    },
    {
      "epoch": 0.668003702560938,
      "grad_norm": 0.4758496582508087,
      "learning_rate": 3e-05,
      "loss": 2.9073,
      "step": 2165
    },
    {
      "epoch": 0.6683122493057698,
      "grad_norm": 0.40784206986427307,
      "learning_rate": 3e-05,
      "loss": 3.0187,
      "step": 2166
    },
    {
      "epoch": 0.6686207960506017,
      "grad_norm": 0.3624708652496338,
      "learning_rate": 3e-05,
      "loss": 2.4545,
      "step": 2167
    },
    {
      "epoch": 0.6689293427954335,
      "grad_norm": 0.2465389370918274,
      "learning_rate": 3e-05,
      "loss": 2.3118,
      "step": 2168
    },
    {
      "epoch": 0.6692378895402653,
      "grad_norm": 0.45873787999153137,
      "learning_rate": 3e-05,
      "loss": 2.9189,
      "step": 2169
    },
    {
      "epoch": 0.6695464362850972,
      "grad_norm": 0.42442765831947327,
      "learning_rate": 3e-05,
      "loss": 2.6501,
      "step": 2170
    },
    {
      "epoch": 0.6698549830299291,
      "grad_norm": 0.31840279698371887,
      "learning_rate": 3e-05,
      "loss": 2.7538,
      "step": 2171
    },
    {
      "epoch": 0.6701635297747609,
      "grad_norm": 0.5180163383483887,
      "learning_rate": 3e-05,
      "loss": 3.147,
      "step": 2172
    },
    {
      "epoch": 0.6704720765195927,
      "grad_norm": 0.32313767075538635,
      "learning_rate": 3e-05,
      "loss": 2.4871,
      "step": 2173
    },
    {
      "epoch": 0.6707806232644246,
      "grad_norm": 0.46035996079444885,
      "learning_rate": 3e-05,
      "loss": 2.6938,
      "step": 2174
    },
    {
      "epoch": 0.6710891700092564,
      "grad_norm": 0.36801087856292725,
      "learning_rate": 3e-05,
      "loss": 2.5027,
      "step": 2175
    },
    {
      "epoch": 0.6713977167540882,
      "grad_norm": 0.32743167877197266,
      "learning_rate": 3e-05,
      "loss": 2.8545,
      "step": 2176
    },
    {
      "epoch": 0.67170626349892,
      "grad_norm": 0.29395878314971924,
      "learning_rate": 3e-05,
      "loss": 2.5953,
      "step": 2177
    },
    {
      "epoch": 0.672014810243752,
      "grad_norm": 0.289230078458786,
      "learning_rate": 3e-05,
      "loss": 2.4384,
      "step": 2178
    },
    {
      "epoch": 0.6723233569885838,
      "grad_norm": 0.45950746536254883,
      "learning_rate": 3e-05,
      "loss": 2.8386,
      "step": 2179
    },
    {
      "epoch": 0.6726319037334156,
      "grad_norm": 0.3325382471084595,
      "learning_rate": 3e-05,
      "loss": 2.7935,
      "step": 2180
    },
    {
      "epoch": 0.6729404504782475,
      "grad_norm": 0.26703134179115295,
      "learning_rate": 3e-05,
      "loss": 2.4701,
      "step": 2181
    },
    {
      "epoch": 0.6732489972230793,
      "grad_norm": 0.5108687877655029,
      "learning_rate": 3e-05,
      "loss": 3.3905,
      "step": 2182
    },
    {
      "epoch": 0.6735575439679111,
      "grad_norm": 0.3172953128814697,
      "learning_rate": 3e-05,
      "loss": 2.1606,
      "step": 2183
    },
    {
      "epoch": 0.673866090712743,
      "grad_norm": 0.35311633348464966,
      "learning_rate": 3e-05,
      "loss": 2.3739,
      "step": 2184
    },
    {
      "epoch": 0.6741746374575748,
      "grad_norm": 0.46544140577316284,
      "learning_rate": 3e-05,
      "loss": 2.7557,
      "step": 2185
    },
    {
      "epoch": 0.6744831842024067,
      "grad_norm": 0.2531960904598236,
      "learning_rate": 3e-05,
      "loss": 2.2056,
      "step": 2186
    },
    {
      "epoch": 0.6747917309472385,
      "grad_norm": 0.33246132731437683,
      "learning_rate": 3e-05,
      "loss": 2.6602,
      "step": 2187
    },
    {
      "epoch": 0.6751002776920704,
      "grad_norm": 0.3067862391471863,
      "learning_rate": 3e-05,
      "loss": 2.1033,
      "step": 2188
    },
    {
      "epoch": 0.6754088244369022,
      "grad_norm": 0.749748170375824,
      "learning_rate": 3e-05,
      "loss": 2.6375,
      "step": 2189
    },
    {
      "epoch": 0.675717371181734,
      "grad_norm": 0.48797351121902466,
      "learning_rate": 3e-05,
      "loss": 3.048,
      "step": 2190
    },
    {
      "epoch": 0.6760259179265659,
      "grad_norm": 0.7949218153953552,
      "learning_rate": 3e-05,
      "loss": 2.4808,
      "step": 2191
    },
    {
      "epoch": 0.6763344646713977,
      "grad_norm": 0.5216763019561768,
      "learning_rate": 3e-05,
      "loss": 3.0119,
      "step": 2192
    },
    {
      "epoch": 0.6766430114162295,
      "grad_norm": 0.48725980520248413,
      "learning_rate": 3e-05,
      "loss": 2.2706,
      "step": 2193
    },
    {
      "epoch": 0.6769515581610615,
      "grad_norm": 0.34293922781944275,
      "learning_rate": 3e-05,
      "loss": 2.3799,
      "step": 2194
    },
    {
      "epoch": 0.6772601049058933,
      "grad_norm": 0.286062628030777,
      "learning_rate": 3e-05,
      "loss": 2.4532,
      "step": 2195
    },
    {
      "epoch": 0.6775686516507251,
      "grad_norm": 0.3767567276954651,
      "learning_rate": 3e-05,
      "loss": 2.7406,
      "step": 2196
    },
    {
      "epoch": 0.6778771983955569,
      "grad_norm": 0.31832119822502136,
      "learning_rate": 3e-05,
      "loss": 2.4595,
      "step": 2197
    },
    {
      "epoch": 0.6781857451403888,
      "grad_norm": 0.5705392360687256,
      "learning_rate": 3e-05,
      "loss": 3.2181,
      "step": 2198
    },
    {
      "epoch": 0.6784942918852206,
      "grad_norm": 0.3068188726902008,
      "learning_rate": 3e-05,
      "loss": 2.3873,
      "step": 2199
    },
    {
      "epoch": 0.6788028386300524,
      "grad_norm": 0.3332638144493103,
      "learning_rate": 3e-05,
      "loss": 2.7779,
      "step": 2200
    },
    {
      "epoch": 0.6791113853748842,
      "grad_norm": 0.30800509452819824,
      "learning_rate": 3e-05,
      "loss": 2.1011,
      "step": 2201
    },
    {
      "epoch": 0.6794199321197162,
      "grad_norm": 0.5739889144897461,
      "learning_rate": 3e-05,
      "loss": 2.9388,
      "step": 2202
    },
    {
      "epoch": 0.679728478864548,
      "grad_norm": 0.29751482605934143,
      "learning_rate": 3e-05,
      "loss": 2.6044,
      "step": 2203
    },
    {
      "epoch": 0.6800370256093798,
      "grad_norm": 0.29991525411605835,
      "learning_rate": 3e-05,
      "loss": 2.3541,
      "step": 2204
    },
    {
      "epoch": 0.6803455723542117,
      "grad_norm": 0.526593029499054,
      "learning_rate": 3e-05,
      "loss": 3.2972,
      "step": 2205
    },
    {
      "epoch": 0.6806541190990435,
      "grad_norm": 0.32622289657592773,
      "learning_rate": 3e-05,
      "loss": 2.6286,
      "step": 2206
    },
    {
      "epoch": 0.6809626658438753,
      "grad_norm": 0.45844268798828125,
      "learning_rate": 3e-05,
      "loss": 3.0493,
      "step": 2207
    },
    {
      "epoch": 0.6812712125887072,
      "grad_norm": 0.3432754576206207,
      "learning_rate": 3e-05,
      "loss": 2.6928,
      "step": 2208
    },
    {
      "epoch": 0.681579759333539,
      "grad_norm": 0.413926362991333,
      "learning_rate": 3e-05,
      "loss": 3.2148,
      "step": 2209
    },
    {
      "epoch": 0.6818883060783709,
      "grad_norm": 0.3242056369781494,
      "learning_rate": 3e-05,
      "loss": 2.8108,
      "step": 2210
    },
    {
      "epoch": 0.6821968528232027,
      "grad_norm": 0.3689955770969391,
      "learning_rate": 3e-05,
      "loss": 2.2757,
      "step": 2211
    },
    {
      "epoch": 0.6825053995680346,
      "grad_norm": 0.4456981122493744,
      "learning_rate": 3e-05,
      "loss": 2.6915,
      "step": 2212
    },
    {
      "epoch": 0.6828139463128664,
      "grad_norm": 0.3460436761379242,
      "learning_rate": 3e-05,
      "loss": 2.4635,
      "step": 2213
    },
    {
      "epoch": 0.6831224930576982,
      "grad_norm": 0.32900822162628174,
      "learning_rate": 3e-05,
      "loss": 2.423,
      "step": 2214
    },
    {
      "epoch": 0.6834310398025301,
      "grad_norm": 0.37845373153686523,
      "learning_rate": 3e-05,
      "loss": 2.34,
      "step": 2215
    },
    {
      "epoch": 0.6837395865473619,
      "grad_norm": 0.432632178068161,
      "learning_rate": 3e-05,
      "loss": 2.6306,
      "step": 2216
    },
    {
      "epoch": 0.6840481332921937,
      "grad_norm": 0.4410339891910553,
      "learning_rate": 3e-05,
      "loss": 2.7681,
      "step": 2217
    },
    {
      "epoch": 0.6843566800370257,
      "grad_norm": 0.3067202866077423,
      "learning_rate": 3e-05,
      "loss": 2.4562,
      "step": 2218
    },
    {
      "epoch": 0.6846652267818575,
      "grad_norm": 0.39880362153053284,
      "learning_rate": 3e-05,
      "loss": 2.6653,
      "step": 2219
    },
    {
      "epoch": 0.6849737735266893,
      "grad_norm": 0.3374853730201721,
      "learning_rate": 3e-05,
      "loss": 2.5644,
      "step": 2220
    },
    {
      "epoch": 0.6852823202715211,
      "grad_norm": 0.32741785049438477,
      "learning_rate": 3e-05,
      "loss": 2.5256,
      "step": 2221
    },
    {
      "epoch": 0.685590867016353,
      "grad_norm": 0.3135084807872772,
      "learning_rate": 3e-05,
      "loss": 2.8652,
      "step": 2222
    },
    {
      "epoch": 0.6858994137611848,
      "grad_norm": 0.2992698848247528,
      "learning_rate": 3e-05,
      "loss": 2.5192,
      "step": 2223
    },
    {
      "epoch": 0.6862079605060166,
      "grad_norm": 0.2987714409828186,
      "learning_rate": 3e-05,
      "loss": 2.3355,
      "step": 2224
    },
    {
      "epoch": 0.6865165072508485,
      "grad_norm": 0.28761017322540283,
      "learning_rate": 3e-05,
      "loss": 2.6733,
      "step": 2225
    },
    {
      "epoch": 0.6868250539956804,
      "grad_norm": 0.45268577337265015,
      "learning_rate": 3e-05,
      "loss": 3.0758,
      "step": 2226
    },
    {
      "epoch": 0.6871336007405122,
      "grad_norm": 0.44058606028556824,
      "learning_rate": 3e-05,
      "loss": 3.1584,
      "step": 2227
    },
    {
      "epoch": 0.687442147485344,
      "grad_norm": 0.3867708146572113,
      "learning_rate": 3e-05,
      "loss": 3.1498,
      "step": 2228
    },
    {
      "epoch": 0.6877506942301759,
      "grad_norm": 0.3580076992511749,
      "learning_rate": 3e-05,
      "loss": 2.8775,
      "step": 2229
    },
    {
      "epoch": 0.6880592409750077,
      "grad_norm": 0.4155197739601135,
      "learning_rate": 3e-05,
      "loss": 2.7409,
      "step": 2230
    },
    {
      "epoch": 0.6883677877198395,
      "grad_norm": 0.4667437672615051,
      "learning_rate": 3e-05,
      "loss": 2.9636,
      "step": 2231
    },
    {
      "epoch": 0.6886763344646714,
      "grad_norm": 0.27438244223594666,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 2232
    },
    {
      "epoch": 0.6889848812095032,
      "grad_norm": 0.32291701436042786,
      "learning_rate": 3e-05,
      "loss": 2.4505,
      "step": 2233
    },
    {
      "epoch": 0.6892934279543351,
      "grad_norm": 0.33502185344696045,
      "learning_rate": 3e-05,
      "loss": 2.3098,
      "step": 2234
    },
    {
      "epoch": 0.689601974699167,
      "grad_norm": 0.37546586990356445,
      "learning_rate": 3e-05,
      "loss": 2.105,
      "step": 2235
    },
    {
      "epoch": 0.6899105214439988,
      "grad_norm": 0.7218531370162964,
      "learning_rate": 3e-05,
      "loss": 3.3048,
      "step": 2236
    },
    {
      "epoch": 0.6902190681888306,
      "grad_norm": 0.29030007123947144,
      "learning_rate": 3e-05,
      "loss": 2.2994,
      "step": 2237
    },
    {
      "epoch": 0.6905276149336624,
      "grad_norm": 0.35308146476745605,
      "learning_rate": 3e-05,
      "loss": 3.0852,
      "step": 2238
    },
    {
      "epoch": 0.6908361616784943,
      "grad_norm": 0.38699471950531006,
      "learning_rate": 3e-05,
      "loss": 2.7178,
      "step": 2239
    },
    {
      "epoch": 0.6911447084233261,
      "grad_norm": 0.39251628518104553,
      "learning_rate": 3e-05,
      "loss": 2.2669,
      "step": 2240
    },
    {
      "epoch": 0.6914532551681579,
      "grad_norm": 0.48583975434303284,
      "learning_rate": 3e-05,
      "loss": 3.0221,
      "step": 2241
    },
    {
      "epoch": 0.6917618019129899,
      "grad_norm": 0.310846209526062,
      "learning_rate": 3e-05,
      "loss": 2.5904,
      "step": 2242
    },
    {
      "epoch": 0.6920703486578217,
      "grad_norm": 0.30022484064102173,
      "learning_rate": 3e-05,
      "loss": 2.4378,
      "step": 2243
    },
    {
      "epoch": 0.6923788954026535,
      "grad_norm": 0.48664936423301697,
      "learning_rate": 3e-05,
      "loss": 2.8075,
      "step": 2244
    },
    {
      "epoch": 0.6926874421474853,
      "grad_norm": 0.3220667839050293,
      "learning_rate": 3e-05,
      "loss": 2.5881,
      "step": 2245
    },
    {
      "epoch": 0.6929959888923172,
      "grad_norm": 0.28465014696121216,
      "learning_rate": 3e-05,
      "loss": 2.2752,
      "step": 2246
    },
    {
      "epoch": 0.693304535637149,
      "grad_norm": 0.2912628650665283,
      "learning_rate": 3e-05,
      "loss": 2.4114,
      "step": 2247
    },
    {
      "epoch": 0.6936130823819808,
      "grad_norm": 0.348395973443985,
      "learning_rate": 3e-05,
      "loss": 2.812,
      "step": 2248
    },
    {
      "epoch": 0.6939216291268128,
      "grad_norm": 0.3242470622062683,
      "learning_rate": 3e-05,
      "loss": 2.7246,
      "step": 2249
    },
    {
      "epoch": 0.6942301758716446,
      "grad_norm": 0.3317355215549469,
      "learning_rate": 3e-05,
      "loss": 2.4041,
      "step": 2250
    },
    {
      "epoch": 0.6945387226164764,
      "grad_norm": 0.4242478609085083,
      "learning_rate": 3e-05,
      "loss": 2.5728,
      "step": 2251
    },
    {
      "epoch": 0.6948472693613083,
      "grad_norm": 0.35139262676239014,
      "learning_rate": 3e-05,
      "loss": 2.47,
      "step": 2252
    },
    {
      "epoch": 0.6951558161061401,
      "grad_norm": 0.3759293854236603,
      "learning_rate": 3e-05,
      "loss": 2.8471,
      "step": 2253
    },
    {
      "epoch": 0.6954643628509719,
      "grad_norm": 0.42364293336868286,
      "learning_rate": 3e-05,
      "loss": 2.7891,
      "step": 2254
    },
    {
      "epoch": 0.6957729095958037,
      "grad_norm": 0.37337446212768555,
      "learning_rate": 3e-05,
      "loss": 2.5027,
      "step": 2255
    },
    {
      "epoch": 0.6960814563406356,
      "grad_norm": 0.5470480918884277,
      "learning_rate": 3e-05,
      "loss": 3.0105,
      "step": 2256
    },
    {
      "epoch": 0.6963900030854675,
      "grad_norm": 0.46059450507164,
      "learning_rate": 3e-05,
      "loss": 2.7711,
      "step": 2257
    },
    {
      "epoch": 0.6966985498302993,
      "grad_norm": 0.259276807308197,
      "learning_rate": 3e-05,
      "loss": 2.2483,
      "step": 2258
    },
    {
      "epoch": 0.6970070965751312,
      "grad_norm": 0.5910064578056335,
      "learning_rate": 3e-05,
      "loss": 2.8551,
      "step": 2259
    },
    {
      "epoch": 0.697315643319963,
      "grad_norm": 0.4100971221923828,
      "learning_rate": 3e-05,
      "loss": 2.7009,
      "step": 2260
    },
    {
      "epoch": 0.6976241900647948,
      "grad_norm": 0.4783870279788971,
      "learning_rate": 3e-05,
      "loss": 3.0588,
      "step": 2261
    },
    {
      "epoch": 0.6979327368096266,
      "grad_norm": 0.2990737855434418,
      "learning_rate": 3e-05,
      "loss": 2.6226,
      "step": 2262
    },
    {
      "epoch": 0.6982412835544585,
      "grad_norm": 0.3180517554283142,
      "learning_rate": 3e-05,
      "loss": 2.7246,
      "step": 2263
    },
    {
      "epoch": 0.6985498302992903,
      "grad_norm": 0.47897204756736755,
      "learning_rate": 3e-05,
      "loss": 2.9415,
      "step": 2264
    },
    {
      "epoch": 0.6988583770441222,
      "grad_norm": 0.36933448910713196,
      "learning_rate": 3e-05,
      "loss": 2.5065,
      "step": 2265
    },
    {
      "epoch": 0.6991669237889541,
      "grad_norm": 0.3363001346588135,
      "learning_rate": 3e-05,
      "loss": 2.4655,
      "step": 2266
    },
    {
      "epoch": 0.6994754705337859,
      "grad_norm": 0.35361260175704956,
      "learning_rate": 3e-05,
      "loss": 2.7079,
      "step": 2267
    },
    {
      "epoch": 0.6997840172786177,
      "grad_norm": 0.6832797527313232,
      "learning_rate": 3e-05,
      "loss": 3.281,
      "step": 2268
    },
    {
      "epoch": 0.7000925640234495,
      "grad_norm": 0.4083409309387207,
      "learning_rate": 3e-05,
      "loss": 3.1841,
      "step": 2269
    },
    {
      "epoch": 0.7004011107682814,
      "grad_norm": 0.36310717463493347,
      "learning_rate": 3e-05,
      "loss": 2.4988,
      "step": 2270
    },
    {
      "epoch": 0.7007096575131132,
      "grad_norm": 0.3296610713005066,
      "learning_rate": 3e-05,
      "loss": 2.5112,
      "step": 2271
    },
    {
      "epoch": 0.701018204257945,
      "grad_norm": 0.3596273958683014,
      "learning_rate": 3e-05,
      "loss": 3.1719,
      "step": 2272
    },
    {
      "epoch": 0.701326751002777,
      "grad_norm": 0.30044686794281006,
      "learning_rate": 3e-05,
      "loss": 2.525,
      "step": 2273
    },
    {
      "epoch": 0.7016352977476088,
      "grad_norm": 0.3514106869697571,
      "learning_rate": 3e-05,
      "loss": 2.8474,
      "step": 2274
    },
    {
      "epoch": 0.7019438444924406,
      "grad_norm": 0.3375399708747864,
      "learning_rate": 3e-05,
      "loss": 2.6595,
      "step": 2275
    },
    {
      "epoch": 0.7022523912372725,
      "grad_norm": 0.3137775659561157,
      "learning_rate": 3e-05,
      "loss": 2.7717,
      "step": 2276
    },
    {
      "epoch": 0.7025609379821043,
      "grad_norm": 0.440255343914032,
      "learning_rate": 3e-05,
      "loss": 2.9353,
      "step": 2277
    },
    {
      "epoch": 0.7028694847269361,
      "grad_norm": 0.3127903342247009,
      "learning_rate": 3e-05,
      "loss": 2.4178,
      "step": 2278
    },
    {
      "epoch": 0.7031780314717679,
      "grad_norm": 0.3616493046283722,
      "learning_rate": 3e-05,
      "loss": 2.8495,
      "step": 2279
    },
    {
      "epoch": 0.7034865782165998,
      "grad_norm": 0.3875301480293274,
      "learning_rate": 3e-05,
      "loss": 2.7257,
      "step": 2280
    },
    {
      "epoch": 0.7037951249614317,
      "grad_norm": 0.44572484493255615,
      "learning_rate": 3e-05,
      "loss": 2.7335,
      "step": 2281
    },
    {
      "epoch": 0.7041036717062635,
      "grad_norm": 0.35633429884910583,
      "learning_rate": 3e-05,
      "loss": 2.919,
      "step": 2282
    },
    {
      "epoch": 0.7044122184510954,
      "grad_norm": 0.3577132821083069,
      "learning_rate": 3e-05,
      "loss": 2.7033,
      "step": 2283
    },
    {
      "epoch": 0.7047207651959272,
      "grad_norm": 0.37120798230171204,
      "learning_rate": 3e-05,
      "loss": 2.8152,
      "step": 2284
    },
    {
      "epoch": 0.705029311940759,
      "grad_norm": 0.2999494969844818,
      "learning_rate": 3e-05,
      "loss": 2.2868,
      "step": 2285
    },
    {
      "epoch": 0.7053378586855908,
      "grad_norm": 0.2723415493965149,
      "learning_rate": 3e-05,
      "loss": 2.2569,
      "step": 2286
    },
    {
      "epoch": 0.7056464054304227,
      "grad_norm": 0.2613217532634735,
      "learning_rate": 3e-05,
      "loss": 2.1881,
      "step": 2287
    },
    {
      "epoch": 0.7059549521752545,
      "grad_norm": 0.30193865299224854,
      "learning_rate": 3e-05,
      "loss": 2.6832,
      "step": 2288
    },
    {
      "epoch": 0.7062634989200864,
      "grad_norm": 0.35747671127319336,
      "learning_rate": 3e-05,
      "loss": 2.6943,
      "step": 2289
    },
    {
      "epoch": 0.7065720456649183,
      "grad_norm": 0.33091968297958374,
      "learning_rate": 3e-05,
      "loss": 2.4817,
      "step": 2290
    },
    {
      "epoch": 0.7068805924097501,
      "grad_norm": 0.28639718890190125,
      "learning_rate": 3e-05,
      "loss": 2.3967,
      "step": 2291
    },
    {
      "epoch": 0.7071891391545819,
      "grad_norm": 0.33338162302970886,
      "learning_rate": 3e-05,
      "loss": 2.5865,
      "step": 2292
    },
    {
      "epoch": 0.7074976858994138,
      "grad_norm": 0.33903318643569946,
      "learning_rate": 3e-05,
      "loss": 2.795,
      "step": 2293
    },
    {
      "epoch": 0.7078062326442456,
      "grad_norm": 0.38681912422180176,
      "learning_rate": 3e-05,
      "loss": 2.6127,
      "step": 2294
    },
    {
      "epoch": 0.7081147793890774,
      "grad_norm": 0.3772861659526825,
      "learning_rate": 3e-05,
      "loss": 2.6537,
      "step": 2295
    },
    {
      "epoch": 0.7084233261339092,
      "grad_norm": 0.31994327902793884,
      "learning_rate": 3e-05,
      "loss": 2.4539,
      "step": 2296
    },
    {
      "epoch": 0.7087318728787412,
      "grad_norm": 0.4272918403148651,
      "learning_rate": 3e-05,
      "loss": 2.7252,
      "step": 2297
    },
    {
      "epoch": 0.709040419623573,
      "grad_norm": 0.3606460392475128,
      "learning_rate": 3e-05,
      "loss": 2.3777,
      "step": 2298
    },
    {
      "epoch": 0.7093489663684048,
      "grad_norm": 0.3049137592315674,
      "learning_rate": 3e-05,
      "loss": 2.5834,
      "step": 2299
    },
    {
      "epoch": 0.7096575131132367,
      "grad_norm": 0.30312472581863403,
      "learning_rate": 3e-05,
      "loss": 2.4526,
      "step": 2300
    },
    {
      "epoch": 0.7099660598580685,
      "grad_norm": 0.39378124475479126,
      "learning_rate": 3e-05,
      "loss": 2.5959,
      "step": 2301
    },
    {
      "epoch": 0.7102746066029003,
      "grad_norm": 0.3223033547401428,
      "learning_rate": 3e-05,
      "loss": 2.5562,
      "step": 2302
    },
    {
      "epoch": 0.7105831533477321,
      "grad_norm": 0.30526670813560486,
      "learning_rate": 3e-05,
      "loss": 2.6841,
      "step": 2303
    },
    {
      "epoch": 0.710891700092564,
      "grad_norm": 0.29967185854911804,
      "learning_rate": 3e-05,
      "loss": 2.8474,
      "step": 2304
    },
    {
      "epoch": 0.7112002468373959,
      "grad_norm": 0.44183349609375,
      "learning_rate": 3e-05,
      "loss": 2.7405,
      "step": 2305
    },
    {
      "epoch": 0.7115087935822277,
      "grad_norm": 0.3366570770740509,
      "learning_rate": 3e-05,
      "loss": 2.5091,
      "step": 2306
    },
    {
      "epoch": 0.7118173403270596,
      "grad_norm": 0.3268311023712158,
      "learning_rate": 3e-05,
      "loss": 2.7072,
      "step": 2307
    },
    {
      "epoch": 0.7121258870718914,
      "grad_norm": 0.32370471954345703,
      "learning_rate": 3e-05,
      "loss": 2.7471,
      "step": 2308
    },
    {
      "epoch": 0.7124344338167232,
      "grad_norm": 0.24951313436031342,
      "learning_rate": 3e-05,
      "loss": 2.3114,
      "step": 2309
    },
    {
      "epoch": 0.712742980561555,
      "grad_norm": 0.31242477893829346,
      "learning_rate": 3e-05,
      "loss": 2.6775,
      "step": 2310
    },
    {
      "epoch": 0.7130515273063869,
      "grad_norm": 0.42314693331718445,
      "learning_rate": 3e-05,
      "loss": 2.7215,
      "step": 2311
    },
    {
      "epoch": 0.7133600740512187,
      "grad_norm": 0.41124439239501953,
      "learning_rate": 3e-05,
      "loss": 3.0069,
      "step": 2312
    },
    {
      "epoch": 0.7136686207960506,
      "grad_norm": 0.34377771615982056,
      "learning_rate": 3e-05,
      "loss": 2.6546,
      "step": 2313
    },
    {
      "epoch": 0.7139771675408825,
      "grad_norm": 0.3865838348865509,
      "learning_rate": 3e-05,
      "loss": 3.2176,
      "step": 2314
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.3056287169456482,
      "learning_rate": 3e-05,
      "loss": 2.6893,
      "step": 2315
    },
    {
      "epoch": 0.7145942610305461,
      "grad_norm": 0.3255837857723236,
      "learning_rate": 3e-05,
      "loss": 2.6466,
      "step": 2316
    },
    {
      "epoch": 0.714902807775378,
      "grad_norm": 0.3461015522480011,
      "learning_rate": 3e-05,
      "loss": 2.2741,
      "step": 2317
    },
    {
      "epoch": 0.7152113545202098,
      "grad_norm": 0.3450034558773041,
      "learning_rate": 3e-05,
      "loss": 2.343,
      "step": 2318
    },
    {
      "epoch": 0.7155199012650416,
      "grad_norm": 0.31079721450805664,
      "learning_rate": 3e-05,
      "loss": 2.9378,
      "step": 2319
    },
    {
      "epoch": 0.7158284480098734,
      "grad_norm": 0.39686399698257446,
      "learning_rate": 3e-05,
      "loss": 2.8702,
      "step": 2320
    },
    {
      "epoch": 0.7161369947547054,
      "grad_norm": 0.44539371132850647,
      "learning_rate": 3e-05,
      "loss": 2.6015,
      "step": 2321
    },
    {
      "epoch": 0.7164455414995372,
      "grad_norm": 0.2945401668548584,
      "learning_rate": 3e-05,
      "loss": 2.2512,
      "step": 2322
    },
    {
      "epoch": 0.716754088244369,
      "grad_norm": 0.46408987045288086,
      "learning_rate": 3e-05,
      "loss": 3.037,
      "step": 2323
    },
    {
      "epoch": 0.7170626349892009,
      "grad_norm": 0.36283478140830994,
      "learning_rate": 3e-05,
      "loss": 2.5426,
      "step": 2324
    },
    {
      "epoch": 0.7173711817340327,
      "grad_norm": 0.37654486298561096,
      "learning_rate": 3e-05,
      "loss": 2.5525,
      "step": 2325
    },
    {
      "epoch": 0.7176797284788645,
      "grad_norm": 0.32116761803627014,
      "learning_rate": 3e-05,
      "loss": 2.357,
      "step": 2326
    },
    {
      "epoch": 0.7179882752236963,
      "grad_norm": 0.3310379683971405,
      "learning_rate": 3e-05,
      "loss": 2.7374,
      "step": 2327
    },
    {
      "epoch": 0.7182968219685283,
      "grad_norm": 0.33591607213020325,
      "learning_rate": 3e-05,
      "loss": 2.8397,
      "step": 2328
    },
    {
      "epoch": 0.7186053687133601,
      "grad_norm": 0.2981739640235901,
      "learning_rate": 3e-05,
      "loss": 2.5991,
      "step": 2329
    },
    {
      "epoch": 0.7189139154581919,
      "grad_norm": 0.2821868360042572,
      "learning_rate": 3e-05,
      "loss": 2.3263,
      "step": 2330
    },
    {
      "epoch": 0.7192224622030238,
      "grad_norm": 0.31829720735549927,
      "learning_rate": 3e-05,
      "loss": 2.9571,
      "step": 2331
    },
    {
      "epoch": 0.7195310089478556,
      "grad_norm": 0.2890639901161194,
      "learning_rate": 3e-05,
      "loss": 2.5357,
      "step": 2332
    },
    {
      "epoch": 0.7198395556926874,
      "grad_norm": 0.3317052721977234,
      "learning_rate": 3e-05,
      "loss": 2.4097,
      "step": 2333
    },
    {
      "epoch": 0.7201481024375193,
      "grad_norm": 0.3682270050048828,
      "learning_rate": 3e-05,
      "loss": 2.7066,
      "step": 2334
    },
    {
      "epoch": 0.7204566491823511,
      "grad_norm": 0.35335469245910645,
      "learning_rate": 3e-05,
      "loss": 2.5117,
      "step": 2335
    },
    {
      "epoch": 0.720765195927183,
      "grad_norm": 0.26136720180511475,
      "learning_rate": 3e-05,
      "loss": 2.3658,
      "step": 2336
    },
    {
      "epoch": 0.7210737426720149,
      "grad_norm": 0.2938997447490692,
      "learning_rate": 3e-05,
      "loss": 2.8205,
      "step": 2337
    },
    {
      "epoch": 0.7213822894168467,
      "grad_norm": 0.37984588742256165,
      "learning_rate": 3e-05,
      "loss": 2.796,
      "step": 2338
    },
    {
      "epoch": 0.7216908361616785,
      "grad_norm": 0.34478721022605896,
      "learning_rate": 3e-05,
      "loss": 2.4592,
      "step": 2339
    },
    {
      "epoch": 0.7219993829065103,
      "grad_norm": 0.2819859981536865,
      "learning_rate": 3e-05,
      "loss": 2.3211,
      "step": 2340
    },
    {
      "epoch": 0.7223079296513422,
      "grad_norm": 0.3126926124095917,
      "learning_rate": 3e-05,
      "loss": 2.5203,
      "step": 2341
    },
    {
      "epoch": 0.722616476396174,
      "grad_norm": 0.28043824434280396,
      "learning_rate": 3e-05,
      "loss": 2.7156,
      "step": 2342
    },
    {
      "epoch": 0.7229250231410058,
      "grad_norm": 0.48454147577285767,
      "learning_rate": 3e-05,
      "loss": 3.0008,
      "step": 2343
    },
    {
      "epoch": 0.7232335698858378,
      "grad_norm": 0.4411979615688324,
      "learning_rate": 3e-05,
      "loss": 3.1171,
      "step": 2344
    },
    {
      "epoch": 0.7235421166306696,
      "grad_norm": 0.3608051836490631,
      "learning_rate": 3e-05,
      "loss": 2.8618,
      "step": 2345
    },
    {
      "epoch": 0.7238506633755014,
      "grad_norm": 0.3951874077320099,
      "learning_rate": 3e-05,
      "loss": 2.7568,
      "step": 2346
    },
    {
      "epoch": 0.7241592101203332,
      "grad_norm": 0.4376699924468994,
      "learning_rate": 3e-05,
      "loss": 2.5694,
      "step": 2347
    },
    {
      "epoch": 0.7244677568651651,
      "grad_norm": 0.3012159466743469,
      "learning_rate": 3e-05,
      "loss": 2.6437,
      "step": 2348
    },
    {
      "epoch": 0.7247763036099969,
      "grad_norm": 0.2726997435092926,
      "learning_rate": 3e-05,
      "loss": 2.3439,
      "step": 2349
    },
    {
      "epoch": 0.7250848503548287,
      "grad_norm": 0.4913809895515442,
      "learning_rate": 3e-05,
      "loss": 2.9323,
      "step": 2350
    },
    {
      "epoch": 0.7253933970996606,
      "grad_norm": 0.34712734818458557,
      "learning_rate": 3e-05,
      "loss": 2.352,
      "step": 2351
    },
    {
      "epoch": 0.7257019438444925,
      "grad_norm": 0.2753252387046814,
      "learning_rate": 3e-05,
      "loss": 2.6161,
      "step": 2352
    },
    {
      "epoch": 0.7260104905893243,
      "grad_norm": 0.290897935628891,
      "learning_rate": 3e-05,
      "loss": 2.5956,
      "step": 2353
    },
    {
      "epoch": 0.7263190373341561,
      "grad_norm": 0.3504074215888977,
      "learning_rate": 3e-05,
      "loss": 2.9418,
      "step": 2354
    },
    {
      "epoch": 0.726627584078988,
      "grad_norm": 0.4185696840286255,
      "learning_rate": 3e-05,
      "loss": 2.6918,
      "step": 2355
    },
    {
      "epoch": 0.7269361308238198,
      "grad_norm": 0.36609721183776855,
      "learning_rate": 3e-05,
      "loss": 2.7385,
      "step": 2356
    },
    {
      "epoch": 0.7272446775686516,
      "grad_norm": 0.35895973443984985,
      "learning_rate": 3e-05,
      "loss": 2.9608,
      "step": 2357
    },
    {
      "epoch": 0.7275532243134835,
      "grad_norm": 0.39654770493507385,
      "learning_rate": 3e-05,
      "loss": 3.4329,
      "step": 2358
    },
    {
      "epoch": 0.7278617710583153,
      "grad_norm": 0.33351320028305054,
      "learning_rate": 3e-05,
      "loss": 2.3718,
      "step": 2359
    },
    {
      "epoch": 0.7281703178031472,
      "grad_norm": 0.4044846296310425,
      "learning_rate": 3e-05,
      "loss": 2.6055,
      "step": 2360
    },
    {
      "epoch": 0.728478864547979,
      "grad_norm": 0.3423720598220825,
      "learning_rate": 3e-05,
      "loss": 2.7944,
      "step": 2361
    },
    {
      "epoch": 0.7287874112928109,
      "grad_norm": 0.3391677439212799,
      "learning_rate": 3e-05,
      "loss": 3.0567,
      "step": 2362
    },
    {
      "epoch": 0.7290959580376427,
      "grad_norm": 0.2994338572025299,
      "learning_rate": 3e-05,
      "loss": 2.5462,
      "step": 2363
    },
    {
      "epoch": 0.7294045047824745,
      "grad_norm": 0.5571287870407104,
      "learning_rate": 3e-05,
      "loss": 3.2259,
      "step": 2364
    },
    {
      "epoch": 0.7297130515273064,
      "grad_norm": 0.36647090315818787,
      "learning_rate": 3e-05,
      "loss": 3.1436,
      "step": 2365
    },
    {
      "epoch": 0.7300215982721382,
      "grad_norm": 0.3304094970226288,
      "learning_rate": 3e-05,
      "loss": 2.7925,
      "step": 2366
    },
    {
      "epoch": 0.73033014501697,
      "grad_norm": 0.3421422243118286,
      "learning_rate": 3e-05,
      "loss": 2.757,
      "step": 2367
    },
    {
      "epoch": 0.730638691761802,
      "grad_norm": 0.5458205342292786,
      "learning_rate": 3e-05,
      "loss": 2.8702,
      "step": 2368
    },
    {
      "epoch": 0.7309472385066338,
      "grad_norm": 0.32926011085510254,
      "learning_rate": 3e-05,
      "loss": 2.3621,
      "step": 2369
    },
    {
      "epoch": 0.7312557852514656,
      "grad_norm": 0.3372330963611603,
      "learning_rate": 3e-05,
      "loss": 2.8953,
      "step": 2370
    },
    {
      "epoch": 0.7315643319962974,
      "grad_norm": 0.3673820197582245,
      "learning_rate": 3e-05,
      "loss": 2.7747,
      "step": 2371
    },
    {
      "epoch": 0.7318728787411293,
      "grad_norm": 0.5362467765808105,
      "learning_rate": 3e-05,
      "loss": 3.1949,
      "step": 2372
    },
    {
      "epoch": 0.7321814254859611,
      "grad_norm": 0.37843358516693115,
      "learning_rate": 3e-05,
      "loss": 2.982,
      "step": 2373
    },
    {
      "epoch": 0.7324899722307929,
      "grad_norm": 0.31489112973213196,
      "learning_rate": 3e-05,
      "loss": 2.5436,
      "step": 2374
    },
    {
      "epoch": 0.7327985189756248,
      "grad_norm": 0.3280992805957794,
      "learning_rate": 3e-05,
      "loss": 2.4243,
      "step": 2375
    },
    {
      "epoch": 0.7331070657204567,
      "grad_norm": 0.32287856936454773,
      "learning_rate": 3e-05,
      "loss": 2.4056,
      "step": 2376
    },
    {
      "epoch": 0.7334156124652885,
      "grad_norm": 0.3963980972766876,
      "learning_rate": 3e-05,
      "loss": 2.8322,
      "step": 2377
    },
    {
      "epoch": 0.7337241592101204,
      "grad_norm": 0.3684110641479492,
      "learning_rate": 3e-05,
      "loss": 2.9733,
      "step": 2378
    },
    {
      "epoch": 0.7340327059549522,
      "grad_norm": 0.2812758684158325,
      "learning_rate": 3e-05,
      "loss": 2.3254,
      "step": 2379
    },
    {
      "epoch": 0.734341252699784,
      "grad_norm": 0.2867478132247925,
      "learning_rate": 3e-05,
      "loss": 2.3585,
      "step": 2380
    },
    {
      "epoch": 0.7346497994446158,
      "grad_norm": 0.34234970808029175,
      "learning_rate": 3e-05,
      "loss": 2.4571,
      "step": 2381
    },
    {
      "epoch": 0.7349583461894477,
      "grad_norm": 0.29766973853111267,
      "learning_rate": 3e-05,
      "loss": 2.1928,
      "step": 2382
    },
    {
      "epoch": 0.7352668929342795,
      "grad_norm": 0.41757407784461975,
      "learning_rate": 3e-05,
      "loss": 2.9724,
      "step": 2383
    },
    {
      "epoch": 0.7355754396791114,
      "grad_norm": 0.2655417323112488,
      "learning_rate": 3e-05,
      "loss": 2.3478,
      "step": 2384
    },
    {
      "epoch": 0.7358839864239433,
      "grad_norm": 0.4270665645599365,
      "learning_rate": 3e-05,
      "loss": 3.1655,
      "step": 2385
    },
    {
      "epoch": 0.7361925331687751,
      "grad_norm": 0.35242992639541626,
      "learning_rate": 3e-05,
      "loss": 2.8984,
      "step": 2386
    },
    {
      "epoch": 0.7365010799136069,
      "grad_norm": 0.32121220231056213,
      "learning_rate": 3e-05,
      "loss": 2.8393,
      "step": 2387
    },
    {
      "epoch": 0.7368096266584387,
      "grad_norm": 0.31955909729003906,
      "learning_rate": 3e-05,
      "loss": 2.5745,
      "step": 2388
    },
    {
      "epoch": 0.7371181734032706,
      "grad_norm": 0.362232506275177,
      "learning_rate": 3e-05,
      "loss": 2.6013,
      "step": 2389
    },
    {
      "epoch": 0.7374267201481024,
      "grad_norm": 0.321138471364975,
      "learning_rate": 3e-05,
      "loss": 2.6018,
      "step": 2390
    },
    {
      "epoch": 0.7377352668929342,
      "grad_norm": 0.33530327677726746,
      "learning_rate": 3e-05,
      "loss": 2.4887,
      "step": 2391
    },
    {
      "epoch": 0.7380438136377662,
      "grad_norm": 0.2641347348690033,
      "learning_rate": 3e-05,
      "loss": 2.4245,
      "step": 2392
    },
    {
      "epoch": 0.738352360382598,
      "grad_norm": 0.3557533025741577,
      "learning_rate": 3e-05,
      "loss": 2.9062,
      "step": 2393
    },
    {
      "epoch": 0.7386609071274298,
      "grad_norm": 0.3297819495201111,
      "learning_rate": 3e-05,
      "loss": 2.3035,
      "step": 2394
    },
    {
      "epoch": 0.7389694538722617,
      "grad_norm": 0.3426196277141571,
      "learning_rate": 3e-05,
      "loss": 2.7256,
      "step": 2395
    },
    {
      "epoch": 0.7392780006170935,
      "grad_norm": 0.305744469165802,
      "learning_rate": 3e-05,
      "loss": 2.3195,
      "step": 2396
    },
    {
      "epoch": 0.7395865473619253,
      "grad_norm": 0.31769564747810364,
      "learning_rate": 3e-05,
      "loss": 2.4878,
      "step": 2397
    },
    {
      "epoch": 0.7398950941067571,
      "grad_norm": 0.25210025906562805,
      "learning_rate": 3e-05,
      "loss": 2.2466,
      "step": 2398
    },
    {
      "epoch": 0.740203640851589,
      "grad_norm": 0.34738588333129883,
      "learning_rate": 3e-05,
      "loss": 2.6832,
      "step": 2399
    },
    {
      "epoch": 0.7405121875964209,
      "grad_norm": 0.32754284143447876,
      "learning_rate": 3e-05,
      "loss": 2.5207,
      "step": 2400
    },
    {
      "epoch": 0.7408207343412527,
      "grad_norm": 0.5114676356315613,
      "learning_rate": 3e-05,
      "loss": 2.9858,
      "step": 2401
    },
    {
      "epoch": 0.7411292810860846,
      "grad_norm": 0.2886255085468292,
      "learning_rate": 3e-05,
      "loss": 2.4498,
      "step": 2402
    },
    {
      "epoch": 0.7414378278309164,
      "grad_norm": 0.2532082796096802,
      "learning_rate": 3e-05,
      "loss": 2.3472,
      "step": 2403
    },
    {
      "epoch": 0.7417463745757482,
      "grad_norm": 0.30410200357437134,
      "learning_rate": 3e-05,
      "loss": 2.3324,
      "step": 2404
    },
    {
      "epoch": 0.74205492132058,
      "grad_norm": 0.3447681665420532,
      "learning_rate": 3e-05,
      "loss": 2.6529,
      "step": 2405
    },
    {
      "epoch": 0.7423634680654119,
      "grad_norm": 0.24459058046340942,
      "learning_rate": 3e-05,
      "loss": 2.2374,
      "step": 2406
    },
    {
      "epoch": 0.7426720148102437,
      "grad_norm": 0.35860294103622437,
      "learning_rate": 3e-05,
      "loss": 2.8066,
      "step": 2407
    },
    {
      "epoch": 0.7429805615550756,
      "grad_norm": 0.3267250657081604,
      "learning_rate": 3e-05,
      "loss": 2.5574,
      "step": 2408
    },
    {
      "epoch": 0.7432891082999075,
      "grad_norm": 0.44709357619285583,
      "learning_rate": 3e-05,
      "loss": 2.9453,
      "step": 2409
    },
    {
      "epoch": 0.7435976550447393,
      "grad_norm": 0.3135888874530792,
      "learning_rate": 3e-05,
      "loss": 2.5602,
      "step": 2410
    },
    {
      "epoch": 0.7439062017895711,
      "grad_norm": 0.33394119143486023,
      "learning_rate": 3e-05,
      "loss": 2.9956,
      "step": 2411
    },
    {
      "epoch": 0.744214748534403,
      "grad_norm": 0.33460256457328796,
      "learning_rate": 3e-05,
      "loss": 2.7939,
      "step": 2412
    },
    {
      "epoch": 0.7445232952792348,
      "grad_norm": 0.4133302569389343,
      "learning_rate": 3e-05,
      "loss": 2.9322,
      "step": 2413
    },
    {
      "epoch": 0.7448318420240666,
      "grad_norm": 0.27985629439353943,
      "learning_rate": 3e-05,
      "loss": 2.4283,
      "step": 2414
    },
    {
      "epoch": 0.7451403887688985,
      "grad_norm": 0.30831363797187805,
      "learning_rate": 3e-05,
      "loss": 2.8271,
      "step": 2415
    },
    {
      "epoch": 0.7454489355137304,
      "grad_norm": 0.3428017795085907,
      "learning_rate": 3e-05,
      "loss": 2.4585,
      "step": 2416
    },
    {
      "epoch": 0.7457574822585622,
      "grad_norm": 0.26562222838401794,
      "learning_rate": 3e-05,
      "loss": 2.4727,
      "step": 2417
    },
    {
      "epoch": 0.746066029003394,
      "grad_norm": 0.333884060382843,
      "learning_rate": 3e-05,
      "loss": 2.7364,
      "step": 2418
    },
    {
      "epoch": 0.7463745757482259,
      "grad_norm": 0.28853392601013184,
      "learning_rate": 3e-05,
      "loss": 2.1919,
      "step": 2419
    },
    {
      "epoch": 0.7466831224930577,
      "grad_norm": 0.37203434109687805,
      "learning_rate": 3e-05,
      "loss": 2.8172,
      "step": 2420
    },
    {
      "epoch": 0.7469916692378895,
      "grad_norm": 0.3368702530860901,
      "learning_rate": 3e-05,
      "loss": 2.8392,
      "step": 2421
    },
    {
      "epoch": 0.7473002159827213,
      "grad_norm": 0.4434995651245117,
      "learning_rate": 3e-05,
      "loss": 3.0815,
      "step": 2422
    },
    {
      "epoch": 0.7476087627275533,
      "grad_norm": 0.3538492023944855,
      "learning_rate": 3e-05,
      "loss": 3.1175,
      "step": 2423
    },
    {
      "epoch": 0.7479173094723851,
      "grad_norm": 0.3190656304359436,
      "learning_rate": 3e-05,
      "loss": 2.9098,
      "step": 2424
    },
    {
      "epoch": 0.7482258562172169,
      "grad_norm": 0.5804755687713623,
      "learning_rate": 3e-05,
      "loss": 2.8014,
      "step": 2425
    },
    {
      "epoch": 0.7485344029620488,
      "grad_norm": 0.3076252043247223,
      "learning_rate": 3e-05,
      "loss": 2.4758,
      "step": 2426
    },
    {
      "epoch": 0.7488429497068806,
      "grad_norm": 0.32138678431510925,
      "learning_rate": 3e-05,
      "loss": 2.3818,
      "step": 2427
    },
    {
      "epoch": 0.7491514964517124,
      "grad_norm": 0.31957435607910156,
      "learning_rate": 3e-05,
      "loss": 2.6408,
      "step": 2428
    },
    {
      "epoch": 0.7494600431965442,
      "grad_norm": 0.4527863562107086,
      "learning_rate": 3e-05,
      "loss": 2.5108,
      "step": 2429
    },
    {
      "epoch": 0.7497685899413761,
      "grad_norm": 0.3417857587337494,
      "learning_rate": 3e-05,
      "loss": 2.3055,
      "step": 2430
    },
    {
      "epoch": 0.750077136686208,
      "grad_norm": 0.3859625458717346,
      "learning_rate": 3e-05,
      "loss": 2.7654,
      "step": 2431
    },
    {
      "epoch": 0.7503856834310398,
      "grad_norm": 0.3281620740890503,
      "learning_rate": 3e-05,
      "loss": 2.9784,
      "step": 2432
    },
    {
      "epoch": 0.7506942301758717,
      "grad_norm": 0.3883509635925293,
      "learning_rate": 3e-05,
      "loss": 2.731,
      "step": 2433
    },
    {
      "epoch": 0.7510027769207035,
      "grad_norm": 0.4876944422721863,
      "learning_rate": 3e-05,
      "loss": 3.2308,
      "step": 2434
    },
    {
      "epoch": 0.7513113236655353,
      "grad_norm": 0.31618931889533997,
      "learning_rate": 3e-05,
      "loss": 2.4108,
      "step": 2435
    },
    {
      "epoch": 0.7516198704103672,
      "grad_norm": 0.3203408420085907,
      "learning_rate": 3e-05,
      "loss": 3.0112,
      "step": 2436
    },
    {
      "epoch": 0.751928417155199,
      "grad_norm": 0.2931792438030243,
      "learning_rate": 3e-05,
      "loss": 2.475,
      "step": 2437
    },
    {
      "epoch": 0.7522369639000308,
      "grad_norm": 0.41018813848495483,
      "learning_rate": 3e-05,
      "loss": 2.6797,
      "step": 2438
    },
    {
      "epoch": 0.7525455106448627,
      "grad_norm": 0.3470573127269745,
      "learning_rate": 3e-05,
      "loss": 2.7325,
      "step": 2439
    },
    {
      "epoch": 0.7528540573896946,
      "grad_norm": 0.40421903133392334,
      "learning_rate": 3e-05,
      "loss": 3.2362,
      "step": 2440
    },
    {
      "epoch": 0.7531626041345264,
      "grad_norm": 0.3297579288482666,
      "learning_rate": 3e-05,
      "loss": 2.7426,
      "step": 2441
    },
    {
      "epoch": 0.7534711508793582,
      "grad_norm": 0.34871771931648254,
      "learning_rate": 3e-05,
      "loss": 2.6116,
      "step": 2442
    },
    {
      "epoch": 0.7537796976241901,
      "grad_norm": 0.3342503011226654,
      "learning_rate": 3e-05,
      "loss": 2.5352,
      "step": 2443
    },
    {
      "epoch": 0.7540882443690219,
      "grad_norm": 0.3271142542362213,
      "learning_rate": 3e-05,
      "loss": 2.4884,
      "step": 2444
    },
    {
      "epoch": 0.7543967911138537,
      "grad_norm": 0.28418102860450745,
      "learning_rate": 3e-05,
      "loss": 2.2455,
      "step": 2445
    },
    {
      "epoch": 0.7547053378586855,
      "grad_norm": 0.3343852758407593,
      "learning_rate": 3e-05,
      "loss": 2.6427,
      "step": 2446
    },
    {
      "epoch": 0.7550138846035175,
      "grad_norm": 0.33747485280036926,
      "learning_rate": 3e-05,
      "loss": 2.789,
      "step": 2447
    },
    {
      "epoch": 0.7553224313483493,
      "grad_norm": 0.3218127191066742,
      "learning_rate": 3e-05,
      "loss": 2.5194,
      "step": 2448
    },
    {
      "epoch": 0.7556309780931811,
      "grad_norm": 0.26363837718963623,
      "learning_rate": 3e-05,
      "loss": 2.407,
      "step": 2449
    },
    {
      "epoch": 0.755939524838013,
      "grad_norm": 0.31267601251602173,
      "learning_rate": 3e-05,
      "loss": 2.7804,
      "step": 2450
    },
    {
      "epoch": 0.7562480715828448,
      "grad_norm": 0.3237861394882202,
      "learning_rate": 3e-05,
      "loss": 2.7229,
      "step": 2451
    },
    {
      "epoch": 0.7565566183276766,
      "grad_norm": 0.26739397644996643,
      "learning_rate": 3e-05,
      "loss": 2.2267,
      "step": 2452
    },
    {
      "epoch": 0.7568651650725085,
      "grad_norm": 0.3156261742115021,
      "learning_rate": 3e-05,
      "loss": 2.7353,
      "step": 2453
    },
    {
      "epoch": 0.7571737118173403,
      "grad_norm": 0.3504122495651245,
      "learning_rate": 3e-05,
      "loss": 2.583,
      "step": 2454
    },
    {
      "epoch": 0.7574822585621722,
      "grad_norm": 0.2620641589164734,
      "learning_rate": 3e-05,
      "loss": 2.4883,
      "step": 2455
    },
    {
      "epoch": 0.757790805307004,
      "grad_norm": 0.283641517162323,
      "learning_rate": 3e-05,
      "loss": 2.6536,
      "step": 2456
    },
    {
      "epoch": 0.7580993520518359,
      "grad_norm": 0.26506707072257996,
      "learning_rate": 3e-05,
      "loss": 2.183,
      "step": 2457
    },
    {
      "epoch": 0.7584078987966677,
      "grad_norm": 0.5000653862953186,
      "learning_rate": 3e-05,
      "loss": 2.5543,
      "step": 2458
    },
    {
      "epoch": 0.7587164455414995,
      "grad_norm": 0.3939869701862335,
      "learning_rate": 3e-05,
      "loss": 2.7877,
      "step": 2459
    },
    {
      "epoch": 0.7590249922863314,
      "grad_norm": 0.3873560428619385,
      "learning_rate": 3e-05,
      "loss": 2.6711,
      "step": 2460
    },
    {
      "epoch": 0.7593335390311632,
      "grad_norm": 0.26135027408599854,
      "learning_rate": 3e-05,
      "loss": 2.3347,
      "step": 2461
    },
    {
      "epoch": 0.759642085775995,
      "grad_norm": 0.3263419270515442,
      "learning_rate": 3e-05,
      "loss": 2.5798,
      "step": 2462
    },
    {
      "epoch": 0.759950632520827,
      "grad_norm": 0.30152130126953125,
      "learning_rate": 3e-05,
      "loss": 2.722,
      "step": 2463
    },
    {
      "epoch": 0.7602591792656588,
      "grad_norm": 0.31374019384384155,
      "learning_rate": 3e-05,
      "loss": 2.6143,
      "step": 2464
    },
    {
      "epoch": 0.7605677260104906,
      "grad_norm": 0.3111976385116577,
      "learning_rate": 3e-05,
      "loss": 2.4367,
      "step": 2465
    },
    {
      "epoch": 0.7608762727553224,
      "grad_norm": 0.27732405066490173,
      "learning_rate": 3e-05,
      "loss": 2.4787,
      "step": 2466
    },
    {
      "epoch": 0.7611848195001543,
      "grad_norm": 0.36006230115890503,
      "learning_rate": 3e-05,
      "loss": 3.1052,
      "step": 2467
    },
    {
      "epoch": 0.7614933662449861,
      "grad_norm": 0.41372624039649963,
      "learning_rate": 3e-05,
      "loss": 2.6298,
      "step": 2468
    },
    {
      "epoch": 0.7618019129898179,
      "grad_norm": 0.2705595791339874,
      "learning_rate": 3e-05,
      "loss": 2.3408,
      "step": 2469
    },
    {
      "epoch": 0.7621104597346497,
      "grad_norm": 0.4729202091693878,
      "learning_rate": 3e-05,
      "loss": 2.77,
      "step": 2470
    },
    {
      "epoch": 0.7624190064794817,
      "grad_norm": 0.3792128264904022,
      "learning_rate": 3e-05,
      "loss": 2.4871,
      "step": 2471
    },
    {
      "epoch": 0.7627275532243135,
      "grad_norm": 0.36836135387420654,
      "learning_rate": 3e-05,
      "loss": 3.0535,
      "step": 2472
    },
    {
      "epoch": 0.7630360999691453,
      "grad_norm": 0.29513564705848694,
      "learning_rate": 3e-05,
      "loss": 2.3313,
      "step": 2473
    },
    {
      "epoch": 0.7633446467139772,
      "grad_norm": 0.2971630394458771,
      "learning_rate": 3e-05,
      "loss": 2.5533,
      "step": 2474
    },
    {
      "epoch": 0.763653193458809,
      "grad_norm": 0.4797176718711853,
      "learning_rate": 3e-05,
      "loss": 3.1834,
      "step": 2475
    },
    {
      "epoch": 0.7639617402036408,
      "grad_norm": 0.4530516266822815,
      "learning_rate": 3e-05,
      "loss": 3.1081,
      "step": 2476
    },
    {
      "epoch": 0.7642702869484727,
      "grad_norm": 0.312676340341568,
      "learning_rate": 3e-05,
      "loss": 2.3509,
      "step": 2477
    },
    {
      "epoch": 0.7645788336933045,
      "grad_norm": 0.3664294481277466,
      "learning_rate": 3e-05,
      "loss": 2.8283,
      "step": 2478
    },
    {
      "epoch": 0.7648873804381364,
      "grad_norm": 0.3851703107357025,
      "learning_rate": 3e-05,
      "loss": 2.6576,
      "step": 2479
    },
    {
      "epoch": 0.7651959271829682,
      "grad_norm": 0.28264495730400085,
      "learning_rate": 3e-05,
      "loss": 2.2982,
      "step": 2480
    },
    {
      "epoch": 0.7655044739278001,
      "grad_norm": 0.3776361048221588,
      "learning_rate": 3e-05,
      "loss": 2.6804,
      "step": 2481
    },
    {
      "epoch": 0.7658130206726319,
      "grad_norm": 0.4029763340950012,
      "learning_rate": 3e-05,
      "loss": 2.7296,
      "step": 2482
    },
    {
      "epoch": 0.7661215674174637,
      "grad_norm": 0.2946641743183136,
      "learning_rate": 3e-05,
      "loss": 2.6946,
      "step": 2483
    },
    {
      "epoch": 0.7664301141622956,
      "grad_norm": 0.3321347236633301,
      "learning_rate": 3e-05,
      "loss": 2.6765,
      "step": 2484
    },
    {
      "epoch": 0.7667386609071274,
      "grad_norm": 0.2816894054412842,
      "learning_rate": 3e-05,
      "loss": 2.2034,
      "step": 2485
    },
    {
      "epoch": 0.7670472076519592,
      "grad_norm": 0.50068199634552,
      "learning_rate": 3e-05,
      "loss": 3.1855,
      "step": 2486
    },
    {
      "epoch": 0.7673557543967912,
      "grad_norm": 0.3912111222743988,
      "learning_rate": 3e-05,
      "loss": 2.7591,
      "step": 2487
    },
    {
      "epoch": 0.767664301141623,
      "grad_norm": 0.48269519209861755,
      "learning_rate": 3e-05,
      "loss": 2.8649,
      "step": 2488
    },
    {
      "epoch": 0.7679728478864548,
      "grad_norm": 0.2754214406013489,
      "learning_rate": 3e-05,
      "loss": 2.3922,
      "step": 2489
    },
    {
      "epoch": 0.7682813946312866,
      "grad_norm": 0.3330560326576233,
      "learning_rate": 3e-05,
      "loss": 2.4132,
      "step": 2490
    },
    {
      "epoch": 0.7685899413761185,
      "grad_norm": 0.3277741074562073,
      "learning_rate": 3e-05,
      "loss": 2.7918,
      "step": 2491
    },
    {
      "epoch": 0.7688984881209503,
      "grad_norm": 0.2994289696216583,
      "learning_rate": 3e-05,
      "loss": 2.3698,
      "step": 2492
    },
    {
      "epoch": 0.7692070348657821,
      "grad_norm": 0.4166587293148041,
      "learning_rate": 3e-05,
      "loss": 2.8805,
      "step": 2493
    },
    {
      "epoch": 0.7695155816106141,
      "grad_norm": 0.31089070439338684,
      "learning_rate": 3e-05,
      "loss": 2.6931,
      "step": 2494
    },
    {
      "epoch": 0.7698241283554459,
      "grad_norm": 0.5258704423904419,
      "learning_rate": 3e-05,
      "loss": 3.2245,
      "step": 2495
    },
    {
      "epoch": 0.7701326751002777,
      "grad_norm": 0.3315356373786926,
      "learning_rate": 3e-05,
      "loss": 2.9022,
      "step": 2496
    },
    {
      "epoch": 0.7704412218451095,
      "grad_norm": 0.6737450957298279,
      "learning_rate": 3e-05,
      "loss": 2.8508,
      "step": 2497
    },
    {
      "epoch": 0.7707497685899414,
      "grad_norm": 0.4406956732273102,
      "learning_rate": 3e-05,
      "loss": 2.5105,
      "step": 2498
    },
    {
      "epoch": 0.7710583153347732,
      "grad_norm": 0.30623260140419006,
      "learning_rate": 3e-05,
      "loss": 2.4407,
      "step": 2499
    },
    {
      "epoch": 0.771366862079605,
      "grad_norm": 0.4266282916069031,
      "learning_rate": 3e-05,
      "loss": 2.4441,
      "step": 2500
    },
    {
      "epoch": 0.7716754088244369,
      "grad_norm": 0.4838431477546692,
      "learning_rate": 3e-05,
      "loss": 3.0301,
      "step": 2501
    },
    {
      "epoch": 0.7719839555692688,
      "grad_norm": 0.31204402446746826,
      "learning_rate": 3e-05,
      "loss": 2.5655,
      "step": 2502
    },
    {
      "epoch": 0.7722925023141006,
      "grad_norm": 0.49737173318862915,
      "learning_rate": 3e-05,
      "loss": 3.3862,
      "step": 2503
    },
    {
      "epoch": 0.7726010490589325,
      "grad_norm": 0.4349178075790405,
      "learning_rate": 3e-05,
      "loss": 3.0957,
      "step": 2504
    },
    {
      "epoch": 0.7729095958037643,
      "grad_norm": 0.305585652589798,
      "learning_rate": 3e-05,
      "loss": 2.2918,
      "step": 2505
    },
    {
      "epoch": 0.7732181425485961,
      "grad_norm": 0.43251293897628784,
      "learning_rate": 3e-05,
      "loss": 2.7933,
      "step": 2506
    },
    {
      "epoch": 0.7735266892934279,
      "grad_norm": 0.3308151960372925,
      "learning_rate": 3e-05,
      "loss": 2.5302,
      "step": 2507
    },
    {
      "epoch": 0.7738352360382598,
      "grad_norm": 0.46171101927757263,
      "learning_rate": 3e-05,
      "loss": 2.8737,
      "step": 2508
    },
    {
      "epoch": 0.7741437827830916,
      "grad_norm": 0.2574649751186371,
      "learning_rate": 3e-05,
      "loss": 2.3048,
      "step": 2509
    },
    {
      "epoch": 0.7744523295279235,
      "grad_norm": 0.3715420365333557,
      "learning_rate": 3e-05,
      "loss": 2.6229,
      "step": 2510
    },
    {
      "epoch": 0.7747608762727554,
      "grad_norm": 0.27585408091545105,
      "learning_rate": 3e-05,
      "loss": 2.3107,
      "step": 2511
    },
    {
      "epoch": 0.7750694230175872,
      "grad_norm": 0.34748390316963196,
      "learning_rate": 3e-05,
      "loss": 2.4872,
      "step": 2512
    },
    {
      "epoch": 0.775377969762419,
      "grad_norm": 0.49627915024757385,
      "learning_rate": 3e-05,
      "loss": 3.1234,
      "step": 2513
    },
    {
      "epoch": 0.7756865165072508,
      "grad_norm": 0.2920449376106262,
      "learning_rate": 3e-05,
      "loss": 2.3504,
      "step": 2514
    },
    {
      "epoch": 0.7759950632520827,
      "grad_norm": 0.4166420102119446,
      "learning_rate": 3e-05,
      "loss": 2.2337,
      "step": 2515
    },
    {
      "epoch": 0.7763036099969145,
      "grad_norm": 0.4382103383541107,
      "learning_rate": 3e-05,
      "loss": 2.5172,
      "step": 2516
    },
    {
      "epoch": 0.7766121567417463,
      "grad_norm": 0.44609513878822327,
      "learning_rate": 3e-05,
      "loss": 2.5689,
      "step": 2517
    },
    {
      "epoch": 0.7769207034865783,
      "grad_norm": 0.2668542265892029,
      "learning_rate": 3e-05,
      "loss": 2.3561,
      "step": 2518
    },
    {
      "epoch": 0.7772292502314101,
      "grad_norm": 0.5886580944061279,
      "learning_rate": 3e-05,
      "loss": 2.9458,
      "step": 2519
    },
    {
      "epoch": 0.7775377969762419,
      "grad_norm": 0.3188645839691162,
      "learning_rate": 3e-05,
      "loss": 2.6815,
      "step": 2520
    },
    {
      "epoch": 0.7778463437210738,
      "grad_norm": 0.2909128963947296,
      "learning_rate": 3e-05,
      "loss": 2.3713,
      "step": 2521
    },
    {
      "epoch": 0.7781548904659056,
      "grad_norm": 0.40471532940864563,
      "learning_rate": 3e-05,
      "loss": 2.8582,
      "step": 2522
    },
    {
      "epoch": 0.7784634372107374,
      "grad_norm": 0.32558944821357727,
      "learning_rate": 3e-05,
      "loss": 3.0118,
      "step": 2523
    },
    {
      "epoch": 0.7787719839555692,
      "grad_norm": 0.44683676958084106,
      "learning_rate": 3e-05,
      "loss": 2.9271,
      "step": 2524
    },
    {
      "epoch": 0.7790805307004011,
      "grad_norm": 0.4091757833957672,
      "learning_rate": 3e-05,
      "loss": 3.0104,
      "step": 2525
    },
    {
      "epoch": 0.779389077445233,
      "grad_norm": 0.32832053303718567,
      "learning_rate": 3e-05,
      "loss": 2.6282,
      "step": 2526
    },
    {
      "epoch": 0.7796976241900648,
      "grad_norm": 0.3501112461090088,
      "learning_rate": 3e-05,
      "loss": 2.8599,
      "step": 2527
    },
    {
      "epoch": 0.7800061709348967,
      "grad_norm": 0.5252228379249573,
      "learning_rate": 3e-05,
      "loss": 3.5915,
      "step": 2528
    },
    {
      "epoch": 0.7803147176797285,
      "grad_norm": 0.3500424921512604,
      "learning_rate": 3e-05,
      "loss": 2.6835,
      "step": 2529
    },
    {
      "epoch": 0.7806232644245603,
      "grad_norm": 0.27792876958847046,
      "learning_rate": 3e-05,
      "loss": 2.3665,
      "step": 2530
    },
    {
      "epoch": 0.7809318111693921,
      "grad_norm": 0.2472742646932602,
      "learning_rate": 3e-05,
      "loss": 2.5189,
      "step": 2531
    },
    {
      "epoch": 0.781240357914224,
      "grad_norm": 0.2625131607055664,
      "learning_rate": 3e-05,
      "loss": 2.2754,
      "step": 2532
    },
    {
      "epoch": 0.7815489046590558,
      "grad_norm": 0.31832006573677063,
      "learning_rate": 3e-05,
      "loss": 2.5445,
      "step": 2533
    },
    {
      "epoch": 0.7818574514038877,
      "grad_norm": 0.438546746969223,
      "learning_rate": 3e-05,
      "loss": 2.5174,
      "step": 2534
    },
    {
      "epoch": 0.7821659981487196,
      "grad_norm": 0.2980039715766907,
      "learning_rate": 3e-05,
      "loss": 2.6125,
      "step": 2535
    },
    {
      "epoch": 0.7824745448935514,
      "grad_norm": 0.47645822167396545,
      "learning_rate": 3e-05,
      "loss": 2.7383,
      "step": 2536
    },
    {
      "epoch": 0.7827830916383832,
      "grad_norm": 0.2914513647556305,
      "learning_rate": 3e-05,
      "loss": 2.329,
      "step": 2537
    },
    {
      "epoch": 0.783091638383215,
      "grad_norm": 0.27299952507019043,
      "learning_rate": 3e-05,
      "loss": 2.1149,
      "step": 2538
    },
    {
      "epoch": 0.7834001851280469,
      "grad_norm": 0.4196101427078247,
      "learning_rate": 3e-05,
      "loss": 2.6852,
      "step": 2539
    },
    {
      "epoch": 0.7837087318728787,
      "grad_norm": 0.36974844336509705,
      "learning_rate": 3e-05,
      "loss": 2.9551,
      "step": 2540
    },
    {
      "epoch": 0.7840172786177105,
      "grad_norm": 0.301588773727417,
      "learning_rate": 3e-05,
      "loss": 2.6668,
      "step": 2541
    },
    {
      "epoch": 0.7843258253625425,
      "grad_norm": 0.3235589861869812,
      "learning_rate": 3e-05,
      "loss": 2.4232,
      "step": 2542
    },
    {
      "epoch": 0.7846343721073743,
      "grad_norm": 0.29920855164527893,
      "learning_rate": 3e-05,
      "loss": 2.71,
      "step": 2543
    },
    {
      "epoch": 0.7849429188522061,
      "grad_norm": 0.2784794867038727,
      "learning_rate": 3e-05,
      "loss": 2.3513,
      "step": 2544
    },
    {
      "epoch": 0.785251465597038,
      "grad_norm": 0.3809240758419037,
      "learning_rate": 3e-05,
      "loss": 2.7195,
      "step": 2545
    },
    {
      "epoch": 0.7855600123418698,
      "grad_norm": 0.6936203837394714,
      "learning_rate": 3e-05,
      "loss": 2.7638,
      "step": 2546
    },
    {
      "epoch": 0.7858685590867016,
      "grad_norm": 0.26078084111213684,
      "learning_rate": 3e-05,
      "loss": 2.3575,
      "step": 2547
    },
    {
      "epoch": 0.7861771058315334,
      "grad_norm": 0.29285866022109985,
      "learning_rate": 3e-05,
      "loss": 2.617,
      "step": 2548
    },
    {
      "epoch": 0.7864856525763653,
      "grad_norm": 0.29151779413223267,
      "learning_rate": 3e-05,
      "loss": 2.3894,
      "step": 2549
    },
    {
      "epoch": 0.7867941993211972,
      "grad_norm": 0.35012853145599365,
      "learning_rate": 3e-05,
      "loss": 2.6173,
      "step": 2550
    },
    {
      "epoch": 0.787102746066029,
      "grad_norm": 0.32691946625709534,
      "learning_rate": 3e-05,
      "loss": 2.519,
      "step": 2551
    },
    {
      "epoch": 0.7874112928108609,
      "grad_norm": 0.4858316481113434,
      "learning_rate": 3e-05,
      "loss": 3.0,
      "step": 2552
    },
    {
      "epoch": 0.7877198395556927,
      "grad_norm": 0.3307975232601166,
      "learning_rate": 3e-05,
      "loss": 2.8458,
      "step": 2553
    },
    {
      "epoch": 0.7880283863005245,
      "grad_norm": 0.2922747731208801,
      "learning_rate": 3e-05,
      "loss": 2.3746,
      "step": 2554
    },
    {
      "epoch": 0.7883369330453563,
      "grad_norm": 0.294562429189682,
      "learning_rate": 3e-05,
      "loss": 2.5383,
      "step": 2555
    },
    {
      "epoch": 0.7886454797901882,
      "grad_norm": 0.33721625804901123,
      "learning_rate": 3e-05,
      "loss": 2.5842,
      "step": 2556
    },
    {
      "epoch": 0.78895402653502,
      "grad_norm": 0.33727625012397766,
      "learning_rate": 3e-05,
      "loss": 2.3681,
      "step": 2557
    },
    {
      "epoch": 0.7892625732798519,
      "grad_norm": 0.3231096565723419,
      "learning_rate": 3e-05,
      "loss": 2.4362,
      "step": 2558
    },
    {
      "epoch": 0.7895711200246838,
      "grad_norm": 0.48459893465042114,
      "learning_rate": 3e-05,
      "loss": 2.8433,
      "step": 2559
    },
    {
      "epoch": 0.7898796667695156,
      "grad_norm": 0.36656931042671204,
      "learning_rate": 3e-05,
      "loss": 2.8116,
      "step": 2560
    },
    {
      "epoch": 0.7901882135143474,
      "grad_norm": 0.3987710475921631,
      "learning_rate": 3e-05,
      "loss": 2.7642,
      "step": 2561
    },
    {
      "epoch": 0.7904967602591793,
      "grad_norm": 0.3305315673351288,
      "learning_rate": 3e-05,
      "loss": 2.4185,
      "step": 2562
    },
    {
      "epoch": 0.7908053070040111,
      "grad_norm": 0.3420153856277466,
      "learning_rate": 3e-05,
      "loss": 2.2899,
      "step": 2563
    },
    {
      "epoch": 0.7911138537488429,
      "grad_norm": 0.2909233272075653,
      "learning_rate": 3e-05,
      "loss": 2.2986,
      "step": 2564
    },
    {
      "epoch": 0.7914224004936747,
      "grad_norm": 0.2866182327270508,
      "learning_rate": 3e-05,
      "loss": 2.2152,
      "step": 2565
    },
    {
      "epoch": 0.7917309472385067,
      "grad_norm": 0.31850704550743103,
      "learning_rate": 3e-05,
      "loss": 2.7282,
      "step": 2566
    },
    {
      "epoch": 0.7920394939833385,
      "grad_norm": 0.32133668661117554,
      "learning_rate": 3e-05,
      "loss": 2.7608,
      "step": 2567
    },
    {
      "epoch": 0.7923480407281703,
      "grad_norm": 0.35847535729408264,
      "learning_rate": 3e-05,
      "loss": 2.774,
      "step": 2568
    },
    {
      "epoch": 0.7926565874730022,
      "grad_norm": 0.23132698237895966,
      "learning_rate": 3e-05,
      "loss": 1.8797,
      "step": 2569
    },
    {
      "epoch": 0.792965134217834,
      "grad_norm": 0.39120107889175415,
      "learning_rate": 3e-05,
      "loss": 2.8025,
      "step": 2570
    },
    {
      "epoch": 0.7932736809626658,
      "grad_norm": 0.3051452934741974,
      "learning_rate": 3e-05,
      "loss": 2.4258,
      "step": 2571
    },
    {
      "epoch": 0.7935822277074976,
      "grad_norm": 0.5285351276397705,
      "learning_rate": 3e-05,
      "loss": 2.6135,
      "step": 2572
    },
    {
      "epoch": 0.7938907744523295,
      "grad_norm": 0.29339203238487244,
      "learning_rate": 3e-05,
      "loss": 2.5413,
      "step": 2573
    },
    {
      "epoch": 0.7941993211971614,
      "grad_norm": 0.37700745463371277,
      "learning_rate": 3e-05,
      "loss": 2.687,
      "step": 2574
    },
    {
      "epoch": 0.7945078679419932,
      "grad_norm": 0.37781018018722534,
      "learning_rate": 3e-05,
      "loss": 3.1863,
      "step": 2575
    },
    {
      "epoch": 0.7948164146868251,
      "grad_norm": 0.30474135279655457,
      "learning_rate": 3e-05,
      "loss": 2.6545,
      "step": 2576
    },
    {
      "epoch": 0.7951249614316569,
      "grad_norm": 0.31511345505714417,
      "learning_rate": 3e-05,
      "loss": 2.4357,
      "step": 2577
    },
    {
      "epoch": 0.7954335081764887,
      "grad_norm": 0.28479355573654175,
      "learning_rate": 3e-05,
      "loss": 2.2369,
      "step": 2578
    },
    {
      "epoch": 0.7957420549213206,
      "grad_norm": 0.5550089478492737,
      "learning_rate": 3e-05,
      "loss": 3.1391,
      "step": 2579
    },
    {
      "epoch": 0.7960506016661524,
      "grad_norm": 0.29672971367836,
      "learning_rate": 3e-05,
      "loss": 2.751,
      "step": 2580
    },
    {
      "epoch": 0.7963591484109843,
      "grad_norm": 0.3240298330783844,
      "learning_rate": 3e-05,
      "loss": 2.7024,
      "step": 2581
    },
    {
      "epoch": 0.7966676951558161,
      "grad_norm": 0.2964293956756592,
      "learning_rate": 3e-05,
      "loss": 2.7657,
      "step": 2582
    },
    {
      "epoch": 0.796976241900648,
      "grad_norm": 0.3782241940498352,
      "learning_rate": 3e-05,
      "loss": 2.7153,
      "step": 2583
    },
    {
      "epoch": 0.7972847886454798,
      "grad_norm": 0.3773477077484131,
      "learning_rate": 3e-05,
      "loss": 2.7207,
      "step": 2584
    },
    {
      "epoch": 0.7975933353903116,
      "grad_norm": 0.2937791049480438,
      "learning_rate": 3e-05,
      "loss": 2.5473,
      "step": 2585
    },
    {
      "epoch": 0.7979018821351435,
      "grad_norm": 0.33833467960357666,
      "learning_rate": 3e-05,
      "loss": 2.7048,
      "step": 2586
    },
    {
      "epoch": 0.7982104288799753,
      "grad_norm": 0.5418023467063904,
      "learning_rate": 3e-05,
      "loss": 3.0567,
      "step": 2587
    },
    {
      "epoch": 0.7985189756248071,
      "grad_norm": 0.32741719484329224,
      "learning_rate": 3e-05,
      "loss": 2.475,
      "step": 2588
    },
    {
      "epoch": 0.798827522369639,
      "grad_norm": 0.27377209067344666,
      "learning_rate": 3e-05,
      "loss": 2.3738,
      "step": 2589
    },
    {
      "epoch": 0.7991360691144709,
      "grad_norm": 0.35510802268981934,
      "learning_rate": 3e-05,
      "loss": 3.0255,
      "step": 2590
    },
    {
      "epoch": 0.7994446158593027,
      "grad_norm": 0.3609134256839752,
      "learning_rate": 3e-05,
      "loss": 2.6703,
      "step": 2591
    },
    {
      "epoch": 0.7997531626041345,
      "grad_norm": 0.2867547571659088,
      "learning_rate": 3e-05,
      "loss": 2.6547,
      "step": 2592
    },
    {
      "epoch": 0.8000617093489664,
      "grad_norm": 0.36119839549064636,
      "learning_rate": 3e-05,
      "loss": 3.0336,
      "step": 2593
    },
    {
      "epoch": 0.8003702560937982,
      "grad_norm": 0.38016822934150696,
      "learning_rate": 3e-05,
      "loss": 2.7407,
      "step": 2594
    },
    {
      "epoch": 0.80067880283863,
      "grad_norm": 0.2825953960418701,
      "learning_rate": 3e-05,
      "loss": 2.6102,
      "step": 2595
    },
    {
      "epoch": 0.8009873495834618,
      "grad_norm": 0.31216439604759216,
      "learning_rate": 3e-05,
      "loss": 2.7951,
      "step": 2596
    },
    {
      "epoch": 0.8012958963282938,
      "grad_norm": 0.2898874282836914,
      "learning_rate": 3e-05,
      "loss": 2.7176,
      "step": 2597
    },
    {
      "epoch": 0.8016044430731256,
      "grad_norm": 0.3177148103713989,
      "learning_rate": 3e-05,
      "loss": 2.3081,
      "step": 2598
    },
    {
      "epoch": 0.8019129898179574,
      "grad_norm": 0.3015001714229584,
      "learning_rate": 3e-05,
      "loss": 2.6391,
      "step": 2599
    },
    {
      "epoch": 0.8022215365627893,
      "grad_norm": 0.3405991196632385,
      "learning_rate": 3e-05,
      "loss": 2.5829,
      "step": 2600
    },
    {
      "epoch": 0.8025300833076211,
      "grad_norm": 0.46747663617134094,
      "learning_rate": 3e-05,
      "loss": 2.8445,
      "step": 2601
    },
    {
      "epoch": 0.8028386300524529,
      "grad_norm": 0.3762139081954956,
      "learning_rate": 3e-05,
      "loss": 2.6351,
      "step": 2602
    },
    {
      "epoch": 0.8031471767972848,
      "grad_norm": 0.2907053232192993,
      "learning_rate": 3e-05,
      "loss": 2.4496,
      "step": 2603
    },
    {
      "epoch": 0.8034557235421166,
      "grad_norm": 0.37862899899482727,
      "learning_rate": 3e-05,
      "loss": 2.8259,
      "step": 2604
    },
    {
      "epoch": 0.8037642702869485,
      "grad_norm": 0.28124406933784485,
      "learning_rate": 3e-05,
      "loss": 2.2761,
      "step": 2605
    },
    {
      "epoch": 0.8040728170317804,
      "grad_norm": 0.38084229826927185,
      "learning_rate": 3e-05,
      "loss": 2.8345,
      "step": 2606
    },
    {
      "epoch": 0.8043813637766122,
      "grad_norm": 0.29694682359695435,
      "learning_rate": 3e-05,
      "loss": 2.7248,
      "step": 2607
    },
    {
      "epoch": 0.804689910521444,
      "grad_norm": 0.3554665744304657,
      "learning_rate": 3e-05,
      "loss": 2.6771,
      "step": 2608
    },
    {
      "epoch": 0.8049984572662758,
      "grad_norm": 0.4623112678527832,
      "learning_rate": 3e-05,
      "loss": 2.8049,
      "step": 2609
    },
    {
      "epoch": 0.8053070040111077,
      "grad_norm": 0.3477843701839447,
      "learning_rate": 3e-05,
      "loss": 2.7834,
      "step": 2610
    },
    {
      "epoch": 0.8056155507559395,
      "grad_norm": 0.325731486082077,
      "learning_rate": 3e-05,
      "loss": 2.9074,
      "step": 2611
    },
    {
      "epoch": 0.8059240975007713,
      "grad_norm": 0.32134032249450684,
      "learning_rate": 3e-05,
      "loss": 2.3314,
      "step": 2612
    },
    {
      "epoch": 0.8062326442456033,
      "grad_norm": 0.26371195912361145,
      "learning_rate": 3e-05,
      "loss": 2.0947,
      "step": 2613
    },
    {
      "epoch": 0.8065411909904351,
      "grad_norm": 0.32706505060195923,
      "learning_rate": 3e-05,
      "loss": 2.4039,
      "step": 2614
    },
    {
      "epoch": 0.8068497377352669,
      "grad_norm": 0.42348629236221313,
      "learning_rate": 3e-05,
      "loss": 3.1099,
      "step": 2615
    },
    {
      "epoch": 0.8071582844800987,
      "grad_norm": 0.41025492548942566,
      "learning_rate": 3e-05,
      "loss": 2.8385,
      "step": 2616
    },
    {
      "epoch": 0.8074668312249306,
      "grad_norm": 0.43497204780578613,
      "learning_rate": 3e-05,
      "loss": 2.989,
      "step": 2617
    },
    {
      "epoch": 0.8077753779697624,
      "grad_norm": 0.36154311895370483,
      "learning_rate": 3e-05,
      "loss": 2.5573,
      "step": 2618
    },
    {
      "epoch": 0.8080839247145942,
      "grad_norm": 0.32343530654907227,
      "learning_rate": 3e-05,
      "loss": 2.8221,
      "step": 2619
    },
    {
      "epoch": 0.808392471459426,
      "grad_norm": 0.3395596146583557,
      "learning_rate": 3e-05,
      "loss": 2.7506,
      "step": 2620
    },
    {
      "epoch": 0.808701018204258,
      "grad_norm": 0.2722172439098358,
      "learning_rate": 3e-05,
      "loss": 2.3515,
      "step": 2621
    },
    {
      "epoch": 0.8090095649490898,
      "grad_norm": 0.3083324432373047,
      "learning_rate": 3e-05,
      "loss": 2.7371,
      "step": 2622
    },
    {
      "epoch": 0.8093181116939216,
      "grad_norm": 0.3601662516593933,
      "learning_rate": 3e-05,
      "loss": 2.9247,
      "step": 2623
    },
    {
      "epoch": 0.8096266584387535,
      "grad_norm": 0.32455846667289734,
      "learning_rate": 3e-05,
      "loss": 2.722,
      "step": 2624
    },
    {
      "epoch": 0.8099352051835853,
      "grad_norm": 0.41457393765449524,
      "learning_rate": 3e-05,
      "loss": 3.0136,
      "step": 2625
    },
    {
      "epoch": 0.8102437519284171,
      "grad_norm": 0.41762587428092957,
      "learning_rate": 3e-05,
      "loss": 2.8492,
      "step": 2626
    },
    {
      "epoch": 0.810552298673249,
      "grad_norm": 0.42402178049087524,
      "learning_rate": 3e-05,
      "loss": 3.0007,
      "step": 2627
    },
    {
      "epoch": 0.8108608454180808,
      "grad_norm": 0.3585397005081177,
      "learning_rate": 3e-05,
      "loss": 2.8061,
      "step": 2628
    },
    {
      "epoch": 0.8111693921629127,
      "grad_norm": 0.3151569366455078,
      "learning_rate": 3e-05,
      "loss": 2.4244,
      "step": 2629
    },
    {
      "epoch": 0.8114779389077446,
      "grad_norm": 0.30198997259140015,
      "learning_rate": 3e-05,
      "loss": 2.4461,
      "step": 2630
    },
    {
      "epoch": 0.8117864856525764,
      "grad_norm": 0.33992353081703186,
      "learning_rate": 3e-05,
      "loss": 2.7244,
      "step": 2631
    },
    {
      "epoch": 0.8120950323974082,
      "grad_norm": 0.41628336906433105,
      "learning_rate": 3e-05,
      "loss": 2.6324,
      "step": 2632
    },
    {
      "epoch": 0.81240357914224,
      "grad_norm": 0.3838220238685608,
      "learning_rate": 3e-05,
      "loss": 2.8536,
      "step": 2633
    },
    {
      "epoch": 0.8127121258870719,
      "grad_norm": 0.2986370921134949,
      "learning_rate": 3e-05,
      "loss": 2.6393,
      "step": 2634
    },
    {
      "epoch": 0.8130206726319037,
      "grad_norm": 0.4952439069747925,
      "learning_rate": 3e-05,
      "loss": 2.938,
      "step": 2635
    },
    {
      "epoch": 0.8133292193767355,
      "grad_norm": 0.3465867340564728,
      "learning_rate": 3e-05,
      "loss": 3.242,
      "step": 2636
    },
    {
      "epoch": 0.8136377661215675,
      "grad_norm": 0.4376803934574127,
      "learning_rate": 3e-05,
      "loss": 2.6619,
      "step": 2637
    },
    {
      "epoch": 0.8139463128663993,
      "grad_norm": 0.3181321918964386,
      "learning_rate": 3e-05,
      "loss": 2.4334,
      "step": 2638
    },
    {
      "epoch": 0.8142548596112311,
      "grad_norm": 0.29201793670654297,
      "learning_rate": 3e-05,
      "loss": 2.5562,
      "step": 2639
    },
    {
      "epoch": 0.814563406356063,
      "grad_norm": 0.3365500569343567,
      "learning_rate": 3e-05,
      "loss": 2.501,
      "step": 2640
    },
    {
      "epoch": 0.8148719531008948,
      "grad_norm": 0.3156362771987915,
      "learning_rate": 3e-05,
      "loss": 2.5262,
      "step": 2641
    },
    {
      "epoch": 0.8151804998457266,
      "grad_norm": 0.35805749893188477,
      "learning_rate": 3e-05,
      "loss": 2.7805,
      "step": 2642
    },
    {
      "epoch": 0.8154890465905584,
      "grad_norm": 0.4785265326499939,
      "learning_rate": 3e-05,
      "loss": 3.0171,
      "step": 2643
    },
    {
      "epoch": 0.8157975933353903,
      "grad_norm": 0.35679998993873596,
      "learning_rate": 3e-05,
      "loss": 2.8537,
      "step": 2644
    },
    {
      "epoch": 0.8161061400802222,
      "grad_norm": 0.3198235332965851,
      "learning_rate": 3e-05,
      "loss": 2.6577,
      "step": 2645
    },
    {
      "epoch": 0.816414686825054,
      "grad_norm": 0.35438936948776245,
      "learning_rate": 3e-05,
      "loss": 2.6866,
      "step": 2646
    },
    {
      "epoch": 0.8167232335698859,
      "grad_norm": 0.36192646622657776,
      "learning_rate": 3e-05,
      "loss": 3.2355,
      "step": 2647
    },
    {
      "epoch": 0.8170317803147177,
      "grad_norm": 0.5714770555496216,
      "learning_rate": 3e-05,
      "loss": 2.6698,
      "step": 2648
    },
    {
      "epoch": 0.8173403270595495,
      "grad_norm": 0.31469258666038513,
      "learning_rate": 3e-05,
      "loss": 2.3305,
      "step": 2649
    },
    {
      "epoch": 0.8176488738043813,
      "grad_norm": 0.2749940752983093,
      "learning_rate": 3e-05,
      "loss": 2.53,
      "step": 2650
    },
    {
      "epoch": 0.8179574205492132,
      "grad_norm": 0.45836618542671204,
      "learning_rate": 3e-05,
      "loss": 3.169,
      "step": 2651
    },
    {
      "epoch": 0.818265967294045,
      "grad_norm": 0.34690895676612854,
      "learning_rate": 3e-05,
      "loss": 2.5242,
      "step": 2652
    },
    {
      "epoch": 0.8185745140388769,
      "grad_norm": 0.31061866879463196,
      "learning_rate": 3e-05,
      "loss": 2.6201,
      "step": 2653
    },
    {
      "epoch": 0.8188830607837088,
      "grad_norm": 0.3154580891132355,
      "learning_rate": 3e-05,
      "loss": 2.5049,
      "step": 2654
    },
    {
      "epoch": 0.8191916075285406,
      "grad_norm": 0.3114047348499298,
      "learning_rate": 3e-05,
      "loss": 2.7704,
      "step": 2655
    },
    {
      "epoch": 0.8195001542733724,
      "grad_norm": 0.3438331186771393,
      "learning_rate": 3e-05,
      "loss": 2.5986,
      "step": 2656
    },
    {
      "epoch": 0.8198087010182042,
      "grad_norm": 0.3506864309310913,
      "learning_rate": 3e-05,
      "loss": 2.4758,
      "step": 2657
    },
    {
      "epoch": 0.8201172477630361,
      "grad_norm": 0.36063385009765625,
      "learning_rate": 3e-05,
      "loss": 2.4194,
      "step": 2658
    },
    {
      "epoch": 0.8204257945078679,
      "grad_norm": 0.3021495044231415,
      "learning_rate": 3e-05,
      "loss": 2.3148,
      "step": 2659
    },
    {
      "epoch": 0.8207343412526998,
      "grad_norm": 0.26607340574264526,
      "learning_rate": 3e-05,
      "loss": 2.1565,
      "step": 2660
    },
    {
      "epoch": 0.8210428879975317,
      "grad_norm": 0.2785079777240753,
      "learning_rate": 3e-05,
      "loss": 2.4643,
      "step": 2661
    },
    {
      "epoch": 0.8213514347423635,
      "grad_norm": 0.42992788553237915,
      "learning_rate": 3e-05,
      "loss": 2.6826,
      "step": 2662
    },
    {
      "epoch": 0.8216599814871953,
      "grad_norm": 0.40426865220069885,
      "learning_rate": 3e-05,
      "loss": 3.0313,
      "step": 2663
    },
    {
      "epoch": 0.8219685282320272,
      "grad_norm": 0.2787652313709259,
      "learning_rate": 3e-05,
      "loss": 2.8305,
      "step": 2664
    },
    {
      "epoch": 0.822277074976859,
      "grad_norm": 0.2849484980106354,
      "learning_rate": 3e-05,
      "loss": 2.6048,
      "step": 2665
    },
    {
      "epoch": 0.8225856217216908,
      "grad_norm": 0.40019720792770386,
      "learning_rate": 3e-05,
      "loss": 3.027,
      "step": 2666
    },
    {
      "epoch": 0.8228941684665226,
      "grad_norm": 0.3828701376914978,
      "learning_rate": 3e-05,
      "loss": 2.7492,
      "step": 2667
    },
    {
      "epoch": 0.8232027152113546,
      "grad_norm": 0.47117698192596436,
      "learning_rate": 3e-05,
      "loss": 2.9442,
      "step": 2668
    },
    {
      "epoch": 0.8235112619561864,
      "grad_norm": 0.4787018895149231,
      "learning_rate": 3e-05,
      "loss": 2.9662,
      "step": 2669
    },
    {
      "epoch": 0.8238198087010182,
      "grad_norm": 0.34115880727767944,
      "learning_rate": 3e-05,
      "loss": 2.5396,
      "step": 2670
    },
    {
      "epoch": 0.8241283554458501,
      "grad_norm": 0.351630836725235,
      "learning_rate": 3e-05,
      "loss": 3.0178,
      "step": 2671
    },
    {
      "epoch": 0.8244369021906819,
      "grad_norm": 0.3947547674179077,
      "learning_rate": 3e-05,
      "loss": 2.9531,
      "step": 2672
    },
    {
      "epoch": 0.8247454489355137,
      "grad_norm": 0.38128045201301575,
      "learning_rate": 3e-05,
      "loss": 2.7299,
      "step": 2673
    },
    {
      "epoch": 0.8250539956803455,
      "grad_norm": 0.3259409964084625,
      "learning_rate": 3e-05,
      "loss": 2.2616,
      "step": 2674
    },
    {
      "epoch": 0.8253625424251774,
      "grad_norm": 0.36533522605895996,
      "learning_rate": 3e-05,
      "loss": 2.8248,
      "step": 2675
    },
    {
      "epoch": 0.8256710891700093,
      "grad_norm": 0.2664625644683838,
      "learning_rate": 3e-05,
      "loss": 2.1005,
      "step": 2676
    },
    {
      "epoch": 0.8259796359148411,
      "grad_norm": 0.3426620066165924,
      "learning_rate": 3e-05,
      "loss": 2.7192,
      "step": 2677
    },
    {
      "epoch": 0.826288182659673,
      "grad_norm": 0.5124875903129578,
      "learning_rate": 3e-05,
      "loss": 2.6017,
      "step": 2678
    },
    {
      "epoch": 0.8265967294045048,
      "grad_norm": 0.3299970328807831,
      "learning_rate": 3e-05,
      "loss": 2.6383,
      "step": 2679
    },
    {
      "epoch": 0.8269052761493366,
      "grad_norm": 0.2651534378528595,
      "learning_rate": 3e-05,
      "loss": 2.1519,
      "step": 2680
    },
    {
      "epoch": 0.8272138228941684,
      "grad_norm": 0.30380764603614807,
      "learning_rate": 3e-05,
      "loss": 2.2517,
      "step": 2681
    },
    {
      "epoch": 0.8275223696390003,
      "grad_norm": 0.35554978251457214,
      "learning_rate": 3e-05,
      "loss": 2.8563,
      "step": 2682
    },
    {
      "epoch": 0.8278309163838321,
      "grad_norm": 0.2989911735057831,
      "learning_rate": 3e-05,
      "loss": 2.572,
      "step": 2683
    },
    {
      "epoch": 0.828139463128664,
      "grad_norm": 0.3779597878456116,
      "learning_rate": 3e-05,
      "loss": 2.6493,
      "step": 2684
    },
    {
      "epoch": 0.8284480098734959,
      "grad_norm": 0.2796619236469269,
      "learning_rate": 3e-05,
      "loss": 2.3776,
      "step": 2685
    },
    {
      "epoch": 0.8287565566183277,
      "grad_norm": 0.31584689021110535,
      "learning_rate": 3e-05,
      "loss": 2.5329,
      "step": 2686
    },
    {
      "epoch": 0.8290651033631595,
      "grad_norm": 0.3388359844684601,
      "learning_rate": 3e-05,
      "loss": 2.7156,
      "step": 2687
    },
    {
      "epoch": 0.8293736501079914,
      "grad_norm": 0.28829944133758545,
      "learning_rate": 3e-05,
      "loss": 2.4409,
      "step": 2688
    },
    {
      "epoch": 0.8296821968528232,
      "grad_norm": 0.27991899847984314,
      "learning_rate": 3e-05,
      "loss": 2.5625,
      "step": 2689
    },
    {
      "epoch": 0.829990743597655,
      "grad_norm": 0.37595227360725403,
      "learning_rate": 3e-05,
      "loss": 3.0767,
      "step": 2690
    },
    {
      "epoch": 0.8302992903424868,
      "grad_norm": 0.3056681752204895,
      "learning_rate": 3e-05,
      "loss": 2.7068,
      "step": 2691
    },
    {
      "epoch": 0.8306078370873188,
      "grad_norm": 0.3019465506076813,
      "learning_rate": 3e-05,
      "loss": 2.4894,
      "step": 2692
    },
    {
      "epoch": 0.8309163838321506,
      "grad_norm": 0.31366103887557983,
      "learning_rate": 3e-05,
      "loss": 2.7181,
      "step": 2693
    },
    {
      "epoch": 0.8312249305769824,
      "grad_norm": 0.4328292906284332,
      "learning_rate": 3e-05,
      "loss": 2.8061,
      "step": 2694
    },
    {
      "epoch": 0.8315334773218143,
      "grad_norm": 0.2865118682384491,
      "learning_rate": 3e-05,
      "loss": 2.6221,
      "step": 2695
    },
    {
      "epoch": 0.8318420240666461,
      "grad_norm": 0.3648976981639862,
      "learning_rate": 3e-05,
      "loss": 2.5833,
      "step": 2696
    },
    {
      "epoch": 0.8321505708114779,
      "grad_norm": 0.3463291525840759,
      "learning_rate": 3e-05,
      "loss": 2.8114,
      "step": 2697
    },
    {
      "epoch": 0.8324591175563097,
      "grad_norm": 0.28873756527900696,
      "learning_rate": 3e-05,
      "loss": 2.6665,
      "step": 2698
    },
    {
      "epoch": 0.8327676643011416,
      "grad_norm": 0.370527446269989,
      "learning_rate": 3e-05,
      "loss": 3.2628,
      "step": 2699
    },
    {
      "epoch": 0.8330762110459735,
      "grad_norm": 0.24568140506744385,
      "learning_rate": 3e-05,
      "loss": 2.2701,
      "step": 2700
    },
    {
      "epoch": 0.8333847577908053,
      "grad_norm": 0.4189949631690979,
      "learning_rate": 3e-05,
      "loss": 3.2355,
      "step": 2701
    },
    {
      "epoch": 0.8336933045356372,
      "grad_norm": 0.4199345111846924,
      "learning_rate": 3e-05,
      "loss": 3.0962,
      "step": 2702
    },
    {
      "epoch": 0.834001851280469,
      "grad_norm": 0.3380403518676758,
      "learning_rate": 3e-05,
      "loss": 2.8414,
      "step": 2703
    },
    {
      "epoch": 0.8343103980253008,
      "grad_norm": 0.35557419061660767,
      "learning_rate": 3e-05,
      "loss": 3.1396,
      "step": 2704
    },
    {
      "epoch": 0.8346189447701327,
      "grad_norm": 0.3350517153739929,
      "learning_rate": 3e-05,
      "loss": 2.4304,
      "step": 2705
    },
    {
      "epoch": 0.8349274915149645,
      "grad_norm": 0.35240399837493896,
      "learning_rate": 3e-05,
      "loss": 2.4992,
      "step": 2706
    },
    {
      "epoch": 0.8352360382597963,
      "grad_norm": 0.3389948904514313,
      "learning_rate": 3e-05,
      "loss": 2.7317,
      "step": 2707
    },
    {
      "epoch": 0.8355445850046282,
      "grad_norm": 0.24769270420074463,
      "learning_rate": 3e-05,
      "loss": 2.3968,
      "step": 2708
    },
    {
      "epoch": 0.8358531317494601,
      "grad_norm": 0.23251210153102875,
      "learning_rate": 3e-05,
      "loss": 2.1971,
      "step": 2709
    },
    {
      "epoch": 0.8361616784942919,
      "grad_norm": 0.26604709029197693,
      "learning_rate": 3e-05,
      "loss": 2.2819,
      "step": 2710
    },
    {
      "epoch": 0.8364702252391237,
      "grad_norm": 0.46939799189567566,
      "learning_rate": 3e-05,
      "loss": 2.6994,
      "step": 2711
    },
    {
      "epoch": 0.8367787719839556,
      "grad_norm": 0.28590619564056396,
      "learning_rate": 3e-05,
      "loss": 2.5709,
      "step": 2712
    },
    {
      "epoch": 0.8370873187287874,
      "grad_norm": 0.4010971486568451,
      "learning_rate": 3e-05,
      "loss": 3.0553,
      "step": 2713
    },
    {
      "epoch": 0.8373958654736192,
      "grad_norm": 0.4239964485168457,
      "learning_rate": 3e-05,
      "loss": 2.8816,
      "step": 2714
    },
    {
      "epoch": 0.837704412218451,
      "grad_norm": 0.3587467074394226,
      "learning_rate": 3e-05,
      "loss": 3.1263,
      "step": 2715
    },
    {
      "epoch": 0.838012958963283,
      "grad_norm": 0.4097730815410614,
      "learning_rate": 3e-05,
      "loss": 2.8905,
      "step": 2716
    },
    {
      "epoch": 0.8383215057081148,
      "grad_norm": 0.37514448165893555,
      "learning_rate": 3e-05,
      "loss": 3.0341,
      "step": 2717
    },
    {
      "epoch": 0.8386300524529466,
      "grad_norm": 0.4946966767311096,
      "learning_rate": 3e-05,
      "loss": 2.8306,
      "step": 2718
    },
    {
      "epoch": 0.8389385991977785,
      "grad_norm": 0.3877689838409424,
      "learning_rate": 3e-05,
      "loss": 3.1758,
      "step": 2719
    },
    {
      "epoch": 0.8392471459426103,
      "grad_norm": 0.3463048040866852,
      "learning_rate": 3e-05,
      "loss": 2.4793,
      "step": 2720
    },
    {
      "epoch": 0.8395556926874421,
      "grad_norm": 0.4264662265777588,
      "learning_rate": 3e-05,
      "loss": 2.7259,
      "step": 2721
    },
    {
      "epoch": 0.839864239432274,
      "grad_norm": 0.35136568546295166,
      "learning_rate": 3e-05,
      "loss": 2.8039,
      "step": 2722
    },
    {
      "epoch": 0.8401727861771058,
      "grad_norm": 0.3099788427352905,
      "learning_rate": 3e-05,
      "loss": 2.5442,
      "step": 2723
    },
    {
      "epoch": 0.8404813329219377,
      "grad_norm": 0.3133460581302643,
      "learning_rate": 3e-05,
      "loss": 2.3903,
      "step": 2724
    },
    {
      "epoch": 0.8407898796667695,
      "grad_norm": 0.30640101432800293,
      "learning_rate": 3e-05,
      "loss": 2.4556,
      "step": 2725
    },
    {
      "epoch": 0.8410984264116014,
      "grad_norm": 0.34547948837280273,
      "learning_rate": 3e-05,
      "loss": 2.7361,
      "step": 2726
    },
    {
      "epoch": 0.8414069731564332,
      "grad_norm": 0.3581864833831787,
      "learning_rate": 3e-05,
      "loss": 2.5407,
      "step": 2727
    },
    {
      "epoch": 0.841715519901265,
      "grad_norm": 0.32495856285095215,
      "learning_rate": 3e-05,
      "loss": 2.9167,
      "step": 2728
    },
    {
      "epoch": 0.8420240666460969,
      "grad_norm": 0.37547948956489563,
      "learning_rate": 3e-05,
      "loss": 2.5143,
      "step": 2729
    },
    {
      "epoch": 0.8423326133909287,
      "grad_norm": 0.2961491048336029,
      "learning_rate": 3e-05,
      "loss": 2.5742,
      "step": 2730
    },
    {
      "epoch": 0.8426411601357605,
      "grad_norm": 0.4586155116558075,
      "learning_rate": 3e-05,
      "loss": 3.3644,
      "step": 2731
    },
    {
      "epoch": 0.8429497068805925,
      "grad_norm": 0.43078380823135376,
      "learning_rate": 3e-05,
      "loss": 2.9546,
      "step": 2732
    },
    {
      "epoch": 0.8432582536254243,
      "grad_norm": 0.5761059522628784,
      "learning_rate": 3e-05,
      "loss": 2.8645,
      "step": 2733
    },
    {
      "epoch": 0.8435668003702561,
      "grad_norm": 0.3569721281528473,
      "learning_rate": 3e-05,
      "loss": 2.4675,
      "step": 2734
    },
    {
      "epoch": 0.8438753471150879,
      "grad_norm": 0.274406373500824,
      "learning_rate": 3e-05,
      "loss": 2.3304,
      "step": 2735
    },
    {
      "epoch": 0.8441838938599198,
      "grad_norm": 0.5510557889938354,
      "learning_rate": 3e-05,
      "loss": 3.5121,
      "step": 2736
    },
    {
      "epoch": 0.8444924406047516,
      "grad_norm": 0.5229282379150391,
      "learning_rate": 3e-05,
      "loss": 3.2188,
      "step": 2737
    },
    {
      "epoch": 0.8448009873495834,
      "grad_norm": 0.3136669993400574,
      "learning_rate": 3e-05,
      "loss": 2.337,
      "step": 2738
    },
    {
      "epoch": 0.8451095340944152,
      "grad_norm": 0.35098856687545776,
      "learning_rate": 3e-05,
      "loss": 2.7652,
      "step": 2739
    },
    {
      "epoch": 0.8454180808392472,
      "grad_norm": 0.29193150997161865,
      "learning_rate": 3e-05,
      "loss": 2.1956,
      "step": 2740
    },
    {
      "epoch": 0.845726627584079,
      "grad_norm": 0.28857648372650146,
      "learning_rate": 3e-05,
      "loss": 2.3248,
      "step": 2741
    },
    {
      "epoch": 0.8460351743289108,
      "grad_norm": 0.31692349910736084,
      "learning_rate": 3e-05,
      "loss": 2.745,
      "step": 2742
    },
    {
      "epoch": 0.8463437210737427,
      "grad_norm": 0.3117423951625824,
      "learning_rate": 3e-05,
      "loss": 2.7153,
      "step": 2743
    },
    {
      "epoch": 0.8466522678185745,
      "grad_norm": 0.25578048825263977,
      "learning_rate": 3e-05,
      "loss": 2.062,
      "step": 2744
    },
    {
      "epoch": 0.8469608145634063,
      "grad_norm": 0.351787269115448,
      "learning_rate": 3e-05,
      "loss": 2.753,
      "step": 2745
    },
    {
      "epoch": 0.8472693613082382,
      "grad_norm": 0.3413632810115814,
      "learning_rate": 3e-05,
      "loss": 2.3042,
      "step": 2746
    },
    {
      "epoch": 0.8475779080530701,
      "grad_norm": 0.3948926031589508,
      "learning_rate": 3e-05,
      "loss": 3.0031,
      "step": 2747
    },
    {
      "epoch": 0.8478864547979019,
      "grad_norm": 0.2530933916568756,
      "learning_rate": 3e-05,
      "loss": 2.2207,
      "step": 2748
    },
    {
      "epoch": 0.8481950015427338,
      "grad_norm": 0.3708816468715668,
      "learning_rate": 3e-05,
      "loss": 2.9789,
      "step": 2749
    },
    {
      "epoch": 0.8485035482875656,
      "grad_norm": 0.28466925024986267,
      "learning_rate": 3e-05,
      "loss": 2.4821,
      "step": 2750
    },
    {
      "epoch": 0.8488120950323974,
      "grad_norm": 0.38464730978012085,
      "learning_rate": 3e-05,
      "loss": 2.6309,
      "step": 2751
    },
    {
      "epoch": 0.8491206417772292,
      "grad_norm": 0.248133584856987,
      "learning_rate": 3e-05,
      "loss": 2.3621,
      "step": 2752
    },
    {
      "epoch": 0.8494291885220611,
      "grad_norm": 0.407237708568573,
      "learning_rate": 3e-05,
      "loss": 3.3179,
      "step": 2753
    },
    {
      "epoch": 0.8497377352668929,
      "grad_norm": 0.3646071255207062,
      "learning_rate": 3e-05,
      "loss": 2.9833,
      "step": 2754
    },
    {
      "epoch": 0.8500462820117248,
      "grad_norm": 0.27028679847717285,
      "learning_rate": 3e-05,
      "loss": 2.4697,
      "step": 2755
    },
    {
      "epoch": 0.8503548287565567,
      "grad_norm": 0.28380709886550903,
      "learning_rate": 3e-05,
      "loss": 2.7756,
      "step": 2756
    },
    {
      "epoch": 0.8506633755013885,
      "grad_norm": 0.4484999477863312,
      "learning_rate": 3e-05,
      "loss": 2.9987,
      "step": 2757
    },
    {
      "epoch": 0.8509719222462203,
      "grad_norm": 0.27781498432159424,
      "learning_rate": 3e-05,
      "loss": 2.3449,
      "step": 2758
    },
    {
      "epoch": 0.8512804689910521,
      "grad_norm": 0.3431036174297333,
      "learning_rate": 3e-05,
      "loss": 2.499,
      "step": 2759
    },
    {
      "epoch": 0.851589015735884,
      "grad_norm": 0.3230559229850769,
      "learning_rate": 3e-05,
      "loss": 2.8905,
      "step": 2760
    },
    {
      "epoch": 0.8518975624807158,
      "grad_norm": 0.33799389004707336,
      "learning_rate": 3e-05,
      "loss": 2.8643,
      "step": 2761
    },
    {
      "epoch": 0.8522061092255476,
      "grad_norm": 0.26970624923706055,
      "learning_rate": 3e-05,
      "loss": 2.52,
      "step": 2762
    },
    {
      "epoch": 0.8525146559703796,
      "grad_norm": 0.3634363114833832,
      "learning_rate": 3e-05,
      "loss": 2.716,
      "step": 2763
    },
    {
      "epoch": 0.8528232027152114,
      "grad_norm": 0.32960546016693115,
      "learning_rate": 3e-05,
      "loss": 2.6191,
      "step": 2764
    },
    {
      "epoch": 0.8531317494600432,
      "grad_norm": 0.35056281089782715,
      "learning_rate": 3e-05,
      "loss": 2.8962,
      "step": 2765
    },
    {
      "epoch": 0.853440296204875,
      "grad_norm": 0.3391978442668915,
      "learning_rate": 3e-05,
      "loss": 2.4946,
      "step": 2766
    },
    {
      "epoch": 0.8537488429497069,
      "grad_norm": 0.5727997422218323,
      "learning_rate": 3e-05,
      "loss": 2.5106,
      "step": 2767
    },
    {
      "epoch": 0.8540573896945387,
      "grad_norm": 0.3022432327270508,
      "learning_rate": 3e-05,
      "loss": 2.2175,
      "step": 2768
    },
    {
      "epoch": 0.8543659364393705,
      "grad_norm": 0.32417649030685425,
      "learning_rate": 3e-05,
      "loss": 2.6882,
      "step": 2769
    },
    {
      "epoch": 0.8546744831842024,
      "grad_norm": 0.3196602463722229,
      "learning_rate": 3e-05,
      "loss": 2.5699,
      "step": 2770
    },
    {
      "epoch": 0.8549830299290343,
      "grad_norm": 0.3259548246860504,
      "learning_rate": 3e-05,
      "loss": 2.7953,
      "step": 2771
    },
    {
      "epoch": 0.8552915766738661,
      "grad_norm": 0.2917036712169647,
      "learning_rate": 3e-05,
      "loss": 2.67,
      "step": 2772
    },
    {
      "epoch": 0.855600123418698,
      "grad_norm": 0.24836978316307068,
      "learning_rate": 3e-05,
      "loss": 2.4831,
      "step": 2773
    },
    {
      "epoch": 0.8559086701635298,
      "grad_norm": 0.3184865117073059,
      "learning_rate": 3e-05,
      "loss": 2.4109,
      "step": 2774
    },
    {
      "epoch": 0.8562172169083616,
      "grad_norm": 0.26458442211151123,
      "learning_rate": 3e-05,
      "loss": 2.1854,
      "step": 2775
    },
    {
      "epoch": 0.8565257636531934,
      "grad_norm": 0.3420705199241638,
      "learning_rate": 3e-05,
      "loss": 2.36,
      "step": 2776
    },
    {
      "epoch": 0.8568343103980253,
      "grad_norm": 0.2892613112926483,
      "learning_rate": 3e-05,
      "loss": 2.6378,
      "step": 2777
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.3661330044269562,
      "learning_rate": 3e-05,
      "loss": 2.6805,
      "step": 2778
    },
    {
      "epoch": 0.857451403887689,
      "grad_norm": 0.35357293486595154,
      "learning_rate": 3e-05,
      "loss": 2.8288,
      "step": 2779
    },
    {
      "epoch": 0.8577599506325209,
      "grad_norm": 0.3824053406715393,
      "learning_rate": 3e-05,
      "loss": 2.9341,
      "step": 2780
    },
    {
      "epoch": 0.8580684973773527,
      "grad_norm": 0.5655234456062317,
      "learning_rate": 3e-05,
      "loss": 2.3296,
      "step": 2781
    },
    {
      "epoch": 0.8583770441221845,
      "grad_norm": 0.32333457469940186,
      "learning_rate": 3e-05,
      "loss": 2.9389,
      "step": 2782
    },
    {
      "epoch": 0.8586855908670163,
      "grad_norm": 0.3431369662284851,
      "learning_rate": 3e-05,
      "loss": 2.5488,
      "step": 2783
    },
    {
      "epoch": 0.8589941376118482,
      "grad_norm": 0.4565815329551697,
      "learning_rate": 3e-05,
      "loss": 2.5384,
      "step": 2784
    },
    {
      "epoch": 0.85930268435668,
      "grad_norm": 0.4187653660774231,
      "learning_rate": 3e-05,
      "loss": 2.9172,
      "step": 2785
    },
    {
      "epoch": 0.8596112311015118,
      "grad_norm": 0.28715503215789795,
      "learning_rate": 3e-05,
      "loss": 2.5196,
      "step": 2786
    },
    {
      "epoch": 0.8599197778463438,
      "grad_norm": 0.28365233540534973,
      "learning_rate": 3e-05,
      "loss": 2.2112,
      "step": 2787
    },
    {
      "epoch": 0.8602283245911756,
      "grad_norm": 0.3386557400226593,
      "learning_rate": 3e-05,
      "loss": 2.6215,
      "step": 2788
    },
    {
      "epoch": 0.8605368713360074,
      "grad_norm": 0.4633696675300598,
      "learning_rate": 3e-05,
      "loss": 2.8126,
      "step": 2789
    },
    {
      "epoch": 0.8608454180808393,
      "grad_norm": 0.4659884572029114,
      "learning_rate": 3e-05,
      "loss": 2.7331,
      "step": 2790
    },
    {
      "epoch": 0.8611539648256711,
      "grad_norm": 0.2589796483516693,
      "learning_rate": 3e-05,
      "loss": 2.2866,
      "step": 2791
    },
    {
      "epoch": 0.8614625115705029,
      "grad_norm": 0.2756637930870056,
      "learning_rate": 3e-05,
      "loss": 2.4272,
      "step": 2792
    },
    {
      "epoch": 0.8617710583153347,
      "grad_norm": 0.46968644857406616,
      "learning_rate": 3e-05,
      "loss": 2.8686,
      "step": 2793
    },
    {
      "epoch": 0.8620796050601666,
      "grad_norm": 0.32571494579315186,
      "learning_rate": 3e-05,
      "loss": 2.2118,
      "step": 2794
    },
    {
      "epoch": 0.8623881518049985,
      "grad_norm": 0.26195642352104187,
      "learning_rate": 3e-05,
      "loss": 2.5495,
      "step": 2795
    },
    {
      "epoch": 0.8626966985498303,
      "grad_norm": 0.30574971437454224,
      "learning_rate": 3e-05,
      "loss": 2.4025,
      "step": 2796
    },
    {
      "epoch": 0.8630052452946622,
      "grad_norm": 0.36342573165893555,
      "learning_rate": 3e-05,
      "loss": 3.0031,
      "step": 2797
    },
    {
      "epoch": 0.863313792039494,
      "grad_norm": 0.3585895895957947,
      "learning_rate": 3e-05,
      "loss": 2.3478,
      "step": 2798
    },
    {
      "epoch": 0.8636223387843258,
      "grad_norm": 0.308462530374527,
      "learning_rate": 3e-05,
      "loss": 2.5347,
      "step": 2799
    },
    {
      "epoch": 0.8639308855291576,
      "grad_norm": 0.345546156167984,
      "learning_rate": 3e-05,
      "loss": 2.6358,
      "step": 2800
    },
    {
      "epoch": 0.8642394322739895,
      "grad_norm": 0.40210092067718506,
      "learning_rate": 3e-05,
      "loss": 2.6162,
      "step": 2801
    },
    {
      "epoch": 0.8645479790188213,
      "grad_norm": 0.318162202835083,
      "learning_rate": 3e-05,
      "loss": 2.2974,
      "step": 2802
    },
    {
      "epoch": 0.8648565257636532,
      "grad_norm": 0.26936545968055725,
      "learning_rate": 3e-05,
      "loss": 2.3031,
      "step": 2803
    },
    {
      "epoch": 0.8651650725084851,
      "grad_norm": 0.34567931294441223,
      "learning_rate": 3e-05,
      "loss": 2.4837,
      "step": 2804
    },
    {
      "epoch": 0.8654736192533169,
      "grad_norm": 0.3457011878490448,
      "learning_rate": 3e-05,
      "loss": 3.1204,
      "step": 2805
    },
    {
      "epoch": 0.8657821659981487,
      "grad_norm": 0.33957639336586,
      "learning_rate": 3e-05,
      "loss": 2.7243,
      "step": 2806
    },
    {
      "epoch": 0.8660907127429806,
      "grad_norm": 0.2553959786891937,
      "learning_rate": 3e-05,
      "loss": 2.1489,
      "step": 2807
    },
    {
      "epoch": 0.8663992594878124,
      "grad_norm": 0.2855786681175232,
      "learning_rate": 3e-05,
      "loss": 2.5898,
      "step": 2808
    },
    {
      "epoch": 0.8667078062326442,
      "grad_norm": 0.3008762001991272,
      "learning_rate": 3e-05,
      "loss": 2.6829,
      "step": 2809
    },
    {
      "epoch": 0.867016352977476,
      "grad_norm": 0.24915944039821625,
      "learning_rate": 3e-05,
      "loss": 2.1304,
      "step": 2810
    },
    {
      "epoch": 0.867324899722308,
      "grad_norm": 0.2780129909515381,
      "learning_rate": 3e-05,
      "loss": 2.4334,
      "step": 2811
    },
    {
      "epoch": 0.8676334464671398,
      "grad_norm": 0.29497021436691284,
      "learning_rate": 3e-05,
      "loss": 2.5898,
      "step": 2812
    },
    {
      "epoch": 0.8679419932119716,
      "grad_norm": 0.27366968989372253,
      "learning_rate": 3e-05,
      "loss": 2.3398,
      "step": 2813
    },
    {
      "epoch": 0.8682505399568035,
      "grad_norm": 0.34261178970336914,
      "learning_rate": 3e-05,
      "loss": 2.5341,
      "step": 2814
    },
    {
      "epoch": 0.8685590867016353,
      "grad_norm": 0.35015833377838135,
      "learning_rate": 3e-05,
      "loss": 2.9025,
      "step": 2815
    },
    {
      "epoch": 0.8688676334464671,
      "grad_norm": 0.23909170925617218,
      "learning_rate": 3e-05,
      "loss": 2.0717,
      "step": 2816
    },
    {
      "epoch": 0.8691761801912989,
      "grad_norm": 0.3002399206161499,
      "learning_rate": 3e-05,
      "loss": 2.5967,
      "step": 2817
    },
    {
      "epoch": 0.8694847269361308,
      "grad_norm": 0.3539586365222931,
      "learning_rate": 3e-05,
      "loss": 2.8151,
      "step": 2818
    },
    {
      "epoch": 0.8697932736809627,
      "grad_norm": 0.2728310823440552,
      "learning_rate": 3e-05,
      "loss": 2.3305,
      "step": 2819
    },
    {
      "epoch": 0.8701018204257945,
      "grad_norm": 0.26273563504219055,
      "learning_rate": 3e-05,
      "loss": 2.3888,
      "step": 2820
    },
    {
      "epoch": 0.8704103671706264,
      "grad_norm": 0.387154221534729,
      "learning_rate": 3e-05,
      "loss": 2.6284,
      "step": 2821
    },
    {
      "epoch": 0.8707189139154582,
      "grad_norm": 0.3012195825576782,
      "learning_rate": 3e-05,
      "loss": 2.3913,
      "step": 2822
    },
    {
      "epoch": 0.87102746066029,
      "grad_norm": 0.3411792516708374,
      "learning_rate": 3e-05,
      "loss": 2.3001,
      "step": 2823
    },
    {
      "epoch": 0.8713360074051218,
      "grad_norm": 0.365607887506485,
      "learning_rate": 3e-05,
      "loss": 2.6243,
      "step": 2824
    },
    {
      "epoch": 0.8716445541499537,
      "grad_norm": 0.27793505787849426,
      "learning_rate": 3e-05,
      "loss": 2.641,
      "step": 2825
    },
    {
      "epoch": 0.8719531008947856,
      "grad_norm": 0.42073091864585876,
      "learning_rate": 3e-05,
      "loss": 3.053,
      "step": 2826
    },
    {
      "epoch": 0.8722616476396174,
      "grad_norm": 0.3174329400062561,
      "learning_rate": 3e-05,
      "loss": 2.5301,
      "step": 2827
    },
    {
      "epoch": 0.8725701943844493,
      "grad_norm": 0.3603368401527405,
      "learning_rate": 3e-05,
      "loss": 2.41,
      "step": 2828
    },
    {
      "epoch": 0.8728787411292811,
      "grad_norm": 0.4805625379085541,
      "learning_rate": 3e-05,
      "loss": 2.5923,
      "step": 2829
    },
    {
      "epoch": 0.8731872878741129,
      "grad_norm": 0.3275454342365265,
      "learning_rate": 3e-05,
      "loss": 3.159,
      "step": 2830
    },
    {
      "epoch": 0.8734958346189448,
      "grad_norm": 0.2651139795780182,
      "learning_rate": 3e-05,
      "loss": 2.4072,
      "step": 2831
    },
    {
      "epoch": 0.8738043813637766,
      "grad_norm": 0.365321546792984,
      "learning_rate": 3e-05,
      "loss": 3.0095,
      "step": 2832
    },
    {
      "epoch": 0.8741129281086084,
      "grad_norm": 0.33985111117362976,
      "learning_rate": 3e-05,
      "loss": 2.6292,
      "step": 2833
    },
    {
      "epoch": 0.8744214748534404,
      "grad_norm": 0.26055023074150085,
      "learning_rate": 3e-05,
      "loss": 2.333,
      "step": 2834
    },
    {
      "epoch": 0.8747300215982722,
      "grad_norm": 0.2642119228839874,
      "learning_rate": 3e-05,
      "loss": 2.3454,
      "step": 2835
    },
    {
      "epoch": 0.875038568343104,
      "grad_norm": 0.258099764585495,
      "learning_rate": 3e-05,
      "loss": 2.3525,
      "step": 2836
    },
    {
      "epoch": 0.8753471150879358,
      "grad_norm": 0.4069601893424988,
      "learning_rate": 3e-05,
      "loss": 3.2616,
      "step": 2837
    },
    {
      "epoch": 0.8756556618327677,
      "grad_norm": 0.27788326144218445,
      "learning_rate": 3e-05,
      "loss": 2.2265,
      "step": 2838
    },
    {
      "epoch": 0.8759642085775995,
      "grad_norm": 0.27163106203079224,
      "learning_rate": 3e-05,
      "loss": 2.7881,
      "step": 2839
    },
    {
      "epoch": 0.8762727553224313,
      "grad_norm": 0.358774334192276,
      "learning_rate": 3e-05,
      "loss": 3.2398,
      "step": 2840
    },
    {
      "epoch": 0.8765813020672631,
      "grad_norm": 0.27252623438835144,
      "learning_rate": 3e-05,
      "loss": 2.316,
      "step": 2841
    },
    {
      "epoch": 0.8768898488120951,
      "grad_norm": 0.254801869392395,
      "learning_rate": 3e-05,
      "loss": 2.456,
      "step": 2842
    },
    {
      "epoch": 0.8771983955569269,
      "grad_norm": 0.3303963541984558,
      "learning_rate": 3e-05,
      "loss": 2.4147,
      "step": 2843
    },
    {
      "epoch": 0.8775069423017587,
      "grad_norm": 0.3355199992656708,
      "learning_rate": 3e-05,
      "loss": 2.4005,
      "step": 2844
    },
    {
      "epoch": 0.8778154890465906,
      "grad_norm": 0.31255996227264404,
      "learning_rate": 3e-05,
      "loss": 2.8942,
      "step": 2845
    },
    {
      "epoch": 0.8781240357914224,
      "grad_norm": 0.3798519968986511,
      "learning_rate": 3e-05,
      "loss": 2.8834,
      "step": 2846
    },
    {
      "epoch": 0.8784325825362542,
      "grad_norm": 0.41321948170661926,
      "learning_rate": 3e-05,
      "loss": 2.9411,
      "step": 2847
    },
    {
      "epoch": 0.878741129281086,
      "grad_norm": 0.31465044617652893,
      "learning_rate": 3e-05,
      "loss": 3.1416,
      "step": 2848
    },
    {
      "epoch": 0.8790496760259179,
      "grad_norm": 0.3947771191596985,
      "learning_rate": 3e-05,
      "loss": 2.4846,
      "step": 2849
    },
    {
      "epoch": 0.8793582227707498,
      "grad_norm": 0.264632910490036,
      "learning_rate": 3e-05,
      "loss": 2.4919,
      "step": 2850
    },
    {
      "epoch": 0.8796667695155816,
      "grad_norm": 0.27938830852508545,
      "learning_rate": 3e-05,
      "loss": 2.2086,
      "step": 2851
    },
    {
      "epoch": 0.8799753162604135,
      "grad_norm": 0.4009644389152527,
      "learning_rate": 3e-05,
      "loss": 2.4718,
      "step": 2852
    },
    {
      "epoch": 0.8802838630052453,
      "grad_norm": 0.3178180456161499,
      "learning_rate": 3e-05,
      "loss": 2.656,
      "step": 2853
    },
    {
      "epoch": 0.8805924097500771,
      "grad_norm": 0.46366631984710693,
      "learning_rate": 3e-05,
      "loss": 2.9195,
      "step": 2854
    },
    {
      "epoch": 0.880900956494909,
      "grad_norm": 0.27060753107070923,
      "learning_rate": 3e-05,
      "loss": 2.3284,
      "step": 2855
    },
    {
      "epoch": 0.8812095032397408,
      "grad_norm": 0.2576352059841156,
      "learning_rate": 3e-05,
      "loss": 2.5008,
      "step": 2856
    },
    {
      "epoch": 0.8815180499845726,
      "grad_norm": 0.3131764531135559,
      "learning_rate": 3e-05,
      "loss": 2.8175,
      "step": 2857
    },
    {
      "epoch": 0.8818265967294046,
      "grad_norm": 0.2975521385669708,
      "learning_rate": 3e-05,
      "loss": 2.6477,
      "step": 2858
    },
    {
      "epoch": 0.8821351434742364,
      "grad_norm": 0.34019675850868225,
      "learning_rate": 3e-05,
      "loss": 2.3781,
      "step": 2859
    },
    {
      "epoch": 0.8824436902190682,
      "grad_norm": 0.287097692489624,
      "learning_rate": 3e-05,
      "loss": 2.3506,
      "step": 2860
    },
    {
      "epoch": 0.8827522369639,
      "grad_norm": 0.35531291365623474,
      "learning_rate": 3e-05,
      "loss": 2.792,
      "step": 2861
    },
    {
      "epoch": 0.8830607837087319,
      "grad_norm": 0.3444139361381531,
      "learning_rate": 3e-05,
      "loss": 2.518,
      "step": 2862
    },
    {
      "epoch": 0.8833693304535637,
      "grad_norm": 0.27676403522491455,
      "learning_rate": 3e-05,
      "loss": 2.2698,
      "step": 2863
    },
    {
      "epoch": 0.8836778771983955,
      "grad_norm": 0.6448089480400085,
      "learning_rate": 3e-05,
      "loss": 3.2054,
      "step": 2864
    },
    {
      "epoch": 0.8839864239432274,
      "grad_norm": 0.4776020050048828,
      "learning_rate": 3e-05,
      "loss": 3.3464,
      "step": 2865
    },
    {
      "epoch": 0.8842949706880593,
      "grad_norm": 0.49332940578460693,
      "learning_rate": 3e-05,
      "loss": 3.3828,
      "step": 2866
    },
    {
      "epoch": 0.8846035174328911,
      "grad_norm": 0.42629098892211914,
      "learning_rate": 3e-05,
      "loss": 2.1809,
      "step": 2867
    },
    {
      "epoch": 0.884912064177723,
      "grad_norm": 0.5584227442741394,
      "learning_rate": 3e-05,
      "loss": 2.9333,
      "step": 2868
    },
    {
      "epoch": 0.8852206109225548,
      "grad_norm": 0.26259252429008484,
      "learning_rate": 3e-05,
      "loss": 2.2268,
      "step": 2869
    },
    {
      "epoch": 0.8855291576673866,
      "grad_norm": 0.28166791796684265,
      "learning_rate": 3e-05,
      "loss": 2.5446,
      "step": 2870
    },
    {
      "epoch": 0.8858377044122184,
      "grad_norm": 0.30466386675834656,
      "learning_rate": 3e-05,
      "loss": 2.1517,
      "step": 2871
    },
    {
      "epoch": 0.8861462511570503,
      "grad_norm": 0.4199237823486328,
      "learning_rate": 3e-05,
      "loss": 2.5337,
      "step": 2872
    },
    {
      "epoch": 0.8864547979018821,
      "grad_norm": 0.4641150236129761,
      "learning_rate": 3e-05,
      "loss": 2.1922,
      "step": 2873
    },
    {
      "epoch": 0.886763344646714,
      "grad_norm": 0.463898628950119,
      "learning_rate": 3e-05,
      "loss": 3.3455,
      "step": 2874
    },
    {
      "epoch": 0.8870718913915459,
      "grad_norm": 0.4407634139060974,
      "learning_rate": 3e-05,
      "loss": 3.0445,
      "step": 2875
    },
    {
      "epoch": 0.8873804381363777,
      "grad_norm": 0.3474924564361572,
      "learning_rate": 3e-05,
      "loss": 2.6418,
      "step": 2876
    },
    {
      "epoch": 0.8876889848812095,
      "grad_norm": 0.42064225673675537,
      "learning_rate": 3e-05,
      "loss": 2.3295,
      "step": 2877
    },
    {
      "epoch": 0.8879975316260413,
      "grad_norm": 0.7077708840370178,
      "learning_rate": 3e-05,
      "loss": 3.1318,
      "step": 2878
    },
    {
      "epoch": 0.8883060783708732,
      "grad_norm": 0.39090168476104736,
      "learning_rate": 3e-05,
      "loss": 2.9096,
      "step": 2879
    },
    {
      "epoch": 0.888614625115705,
      "grad_norm": 0.4708465039730072,
      "learning_rate": 3e-05,
      "loss": 2.8185,
      "step": 2880
    },
    {
      "epoch": 0.8889231718605368,
      "grad_norm": 0.5754609107971191,
      "learning_rate": 3e-05,
      "loss": 3.0664,
      "step": 2881
    },
    {
      "epoch": 0.8892317186053688,
      "grad_norm": 0.4405844807624817,
      "learning_rate": 3e-05,
      "loss": 2.8285,
      "step": 2882
    },
    {
      "epoch": 0.8895402653502006,
      "grad_norm": 0.4159712493419647,
      "learning_rate": 3e-05,
      "loss": 3.0555,
      "step": 2883
    },
    {
      "epoch": 0.8898488120950324,
      "grad_norm": 0.2793191075325012,
      "learning_rate": 3e-05,
      "loss": 2.3012,
      "step": 2884
    },
    {
      "epoch": 0.8901573588398642,
      "grad_norm": 0.49622389674186707,
      "learning_rate": 3e-05,
      "loss": 2.7565,
      "step": 2885
    },
    {
      "epoch": 0.8904659055846961,
      "grad_norm": 0.43158969283103943,
      "learning_rate": 3e-05,
      "loss": 3.1146,
      "step": 2886
    },
    {
      "epoch": 0.8907744523295279,
      "grad_norm": 0.38310056924819946,
      "learning_rate": 3e-05,
      "loss": 2.8941,
      "step": 2887
    },
    {
      "epoch": 0.8910829990743597,
      "grad_norm": 0.370707243680954,
      "learning_rate": 3e-05,
      "loss": 2.2146,
      "step": 2888
    },
    {
      "epoch": 0.8913915458191916,
      "grad_norm": 0.26652663946151733,
      "learning_rate": 3e-05,
      "loss": 2.6479,
      "step": 2889
    },
    {
      "epoch": 0.8917000925640235,
      "grad_norm": 0.4532812833786011,
      "learning_rate": 3e-05,
      "loss": 2.9307,
      "step": 2890
    },
    {
      "epoch": 0.8920086393088553,
      "grad_norm": 0.2940596640110016,
      "learning_rate": 3e-05,
      "loss": 2.5485,
      "step": 2891
    },
    {
      "epoch": 0.8923171860536872,
      "grad_norm": 0.3772636950016022,
      "learning_rate": 3e-05,
      "loss": 2.5425,
      "step": 2892
    },
    {
      "epoch": 0.892625732798519,
      "grad_norm": 0.35960617661476135,
      "learning_rate": 3e-05,
      "loss": 2.721,
      "step": 2893
    },
    {
      "epoch": 0.8929342795433508,
      "grad_norm": 0.2909625470638275,
      "learning_rate": 3e-05,
      "loss": 2.6044,
      "step": 2894
    },
    {
      "epoch": 0.8932428262881826,
      "grad_norm": 0.4376247227191925,
      "learning_rate": 3e-05,
      "loss": 2.6146,
      "step": 2895
    },
    {
      "epoch": 0.8935513730330145,
      "grad_norm": 0.34912368655204773,
      "learning_rate": 3e-05,
      "loss": 2.854,
      "step": 2896
    },
    {
      "epoch": 0.8938599197778463,
      "grad_norm": 0.30256229639053345,
      "learning_rate": 3e-05,
      "loss": 3.0707,
      "step": 2897
    },
    {
      "epoch": 0.8941684665226782,
      "grad_norm": 0.3484569787979126,
      "learning_rate": 3e-05,
      "loss": 2.9541,
      "step": 2898
    },
    {
      "epoch": 0.8944770132675101,
      "grad_norm": 0.33366307616233826,
      "learning_rate": 3e-05,
      "loss": 2.6823,
      "step": 2899
    },
    {
      "epoch": 0.8947855600123419,
      "grad_norm": 0.30587753653526306,
      "learning_rate": 3e-05,
      "loss": 2.3358,
      "step": 2900
    },
    {
      "epoch": 0.8950941067571737,
      "grad_norm": 0.43226373195648193,
      "learning_rate": 3e-05,
      "loss": 3.1975,
      "step": 2901
    },
    {
      "epoch": 0.8954026535020055,
      "grad_norm": 0.5207734704017639,
      "learning_rate": 3e-05,
      "loss": 3.0812,
      "step": 2902
    },
    {
      "epoch": 0.8957112002468374,
      "grad_norm": 0.314447820186615,
      "learning_rate": 3e-05,
      "loss": 2.3953,
      "step": 2903
    },
    {
      "epoch": 0.8960197469916692,
      "grad_norm": 0.30643942952156067,
      "learning_rate": 3e-05,
      "loss": 2.7002,
      "step": 2904
    },
    {
      "epoch": 0.896328293736501,
      "grad_norm": 0.30888503789901733,
      "learning_rate": 3e-05,
      "loss": 2.4829,
      "step": 2905
    },
    {
      "epoch": 0.896636840481333,
      "grad_norm": 0.3054732084274292,
      "learning_rate": 3e-05,
      "loss": 2.5466,
      "step": 2906
    },
    {
      "epoch": 0.8969453872261648,
      "grad_norm": 0.39422494173049927,
      "learning_rate": 3e-05,
      "loss": 3.0326,
      "step": 2907
    },
    {
      "epoch": 0.8972539339709966,
      "grad_norm": 0.27239397168159485,
      "learning_rate": 3e-05,
      "loss": 2.5485,
      "step": 2908
    },
    {
      "epoch": 0.8975624807158284,
      "grad_norm": 0.34003734588623047,
      "learning_rate": 3e-05,
      "loss": 2.7929,
      "step": 2909
    },
    {
      "epoch": 0.8978710274606603,
      "grad_norm": 0.278083860874176,
      "learning_rate": 3e-05,
      "loss": 2.4814,
      "step": 2910
    },
    {
      "epoch": 0.8981795742054921,
      "grad_norm": 0.30099397897720337,
      "learning_rate": 3e-05,
      "loss": 2.5195,
      "step": 2911
    },
    {
      "epoch": 0.8984881209503239,
      "grad_norm": 0.2887558341026306,
      "learning_rate": 3e-05,
      "loss": 2.7448,
      "step": 2912
    },
    {
      "epoch": 0.8987966676951559,
      "grad_norm": 0.33276084065437317,
      "learning_rate": 3e-05,
      "loss": 2.6402,
      "step": 2913
    },
    {
      "epoch": 0.8991052144399877,
      "grad_norm": 0.3490481972694397,
      "learning_rate": 3e-05,
      "loss": 2.7461,
      "step": 2914
    },
    {
      "epoch": 0.8994137611848195,
      "grad_norm": 0.3352538049221039,
      "learning_rate": 3e-05,
      "loss": 2.9501,
      "step": 2915
    },
    {
      "epoch": 0.8997223079296514,
      "grad_norm": 0.32356688380241394,
      "learning_rate": 3e-05,
      "loss": 2.6937,
      "step": 2916
    },
    {
      "epoch": 0.9000308546744832,
      "grad_norm": 0.3816601634025574,
      "learning_rate": 3e-05,
      "loss": 2.4262,
      "step": 2917
    },
    {
      "epoch": 0.900339401419315,
      "grad_norm": 0.3136836290359497,
      "learning_rate": 3e-05,
      "loss": 2.6908,
      "step": 2918
    },
    {
      "epoch": 0.9006479481641468,
      "grad_norm": 0.5210847854614258,
      "learning_rate": 3e-05,
      "loss": 3.1556,
      "step": 2919
    },
    {
      "epoch": 0.9009564949089787,
      "grad_norm": 0.33751219511032104,
      "learning_rate": 3e-05,
      "loss": 2.2649,
      "step": 2920
    },
    {
      "epoch": 0.9012650416538106,
      "grad_norm": 0.27721014618873596,
      "learning_rate": 3e-05,
      "loss": 2.3146,
      "step": 2921
    },
    {
      "epoch": 0.9015735883986424,
      "grad_norm": 0.46372905373573303,
      "learning_rate": 3e-05,
      "loss": 2.34,
      "step": 2922
    },
    {
      "epoch": 0.9018821351434743,
      "grad_norm": 0.39576542377471924,
      "learning_rate": 3e-05,
      "loss": 2.6124,
      "step": 2923
    },
    {
      "epoch": 0.9021906818883061,
      "grad_norm": 0.2982642352581024,
      "learning_rate": 3e-05,
      "loss": 2.2038,
      "step": 2924
    },
    {
      "epoch": 0.9024992286331379,
      "grad_norm": 0.35037878155708313,
      "learning_rate": 3e-05,
      "loss": 2.4019,
      "step": 2925
    },
    {
      "epoch": 0.9028077753779697,
      "grad_norm": 0.36664894223213196,
      "learning_rate": 3e-05,
      "loss": 2.368,
      "step": 2926
    },
    {
      "epoch": 0.9031163221228016,
      "grad_norm": 0.33474940061569214,
      "learning_rate": 3e-05,
      "loss": 2.3994,
      "step": 2927
    },
    {
      "epoch": 0.9034248688676334,
      "grad_norm": 0.29408782720565796,
      "learning_rate": 3e-05,
      "loss": 2.4927,
      "step": 2928
    },
    {
      "epoch": 0.9037334156124653,
      "grad_norm": 0.25650712847709656,
      "learning_rate": 3e-05,
      "loss": 2.1845,
      "step": 2929
    },
    {
      "epoch": 0.9040419623572972,
      "grad_norm": 0.4002336263656616,
      "learning_rate": 3e-05,
      "loss": 2.874,
      "step": 2930
    },
    {
      "epoch": 0.904350509102129,
      "grad_norm": 0.31816190481185913,
      "learning_rate": 3e-05,
      "loss": 2.7755,
      "step": 2931
    },
    {
      "epoch": 0.9046590558469608,
      "grad_norm": 0.5233032703399658,
      "learning_rate": 3e-05,
      "loss": 3.3622,
      "step": 2932
    },
    {
      "epoch": 0.9049676025917927,
      "grad_norm": 0.3366897702217102,
      "learning_rate": 3e-05,
      "loss": 2.285,
      "step": 2933
    },
    {
      "epoch": 0.9052761493366245,
      "grad_norm": 0.30700457096099854,
      "learning_rate": 3e-05,
      "loss": 2.8356,
      "step": 2934
    },
    {
      "epoch": 0.9055846960814563,
      "grad_norm": 0.2688470780849457,
      "learning_rate": 3e-05,
      "loss": 2.6035,
      "step": 2935
    },
    {
      "epoch": 0.9058932428262881,
      "grad_norm": 0.24110789597034454,
      "learning_rate": 3e-05,
      "loss": 2.1513,
      "step": 2936
    },
    {
      "epoch": 0.9062017895711201,
      "grad_norm": 0.310677170753479,
      "learning_rate": 3e-05,
      "loss": 2.4114,
      "step": 2937
    },
    {
      "epoch": 0.9065103363159519,
      "grad_norm": 0.33301934599876404,
      "learning_rate": 3e-05,
      "loss": 2.6869,
      "step": 2938
    },
    {
      "epoch": 0.9068188830607837,
      "grad_norm": 0.30799493193626404,
      "learning_rate": 3e-05,
      "loss": 3.1069,
      "step": 2939
    },
    {
      "epoch": 0.9071274298056156,
      "grad_norm": 0.3512706458568573,
      "learning_rate": 3e-05,
      "loss": 2.5975,
      "step": 2940
    },
    {
      "epoch": 0.9074359765504474,
      "grad_norm": 0.3067226707935333,
      "learning_rate": 3e-05,
      "loss": 2.9516,
      "step": 2941
    },
    {
      "epoch": 0.9077445232952792,
      "grad_norm": 0.2766515910625458,
      "learning_rate": 3e-05,
      "loss": 1.9597,
      "step": 2942
    },
    {
      "epoch": 0.908053070040111,
      "grad_norm": 0.28473570942878723,
      "learning_rate": 3e-05,
      "loss": 2.5181,
      "step": 2943
    },
    {
      "epoch": 0.9083616167849429,
      "grad_norm": 0.38429906964302063,
      "learning_rate": 3e-05,
      "loss": 3.0435,
      "step": 2944
    },
    {
      "epoch": 0.9086701635297748,
      "grad_norm": 0.31083181500434875,
      "learning_rate": 3e-05,
      "loss": 2.2965,
      "step": 2945
    },
    {
      "epoch": 0.9089787102746066,
      "grad_norm": 0.270593523979187,
      "learning_rate": 3e-05,
      "loss": 2.2807,
      "step": 2946
    },
    {
      "epoch": 0.9092872570194385,
      "grad_norm": 0.2840012311935425,
      "learning_rate": 3e-05,
      "loss": 2.343,
      "step": 2947
    },
    {
      "epoch": 0.9095958037642703,
      "grad_norm": 0.3537600040435791,
      "learning_rate": 3e-05,
      "loss": 2.7333,
      "step": 2948
    },
    {
      "epoch": 0.9099043505091021,
      "grad_norm": 0.32812702655792236,
      "learning_rate": 3e-05,
      "loss": 2.6442,
      "step": 2949
    },
    {
      "epoch": 0.910212897253934,
      "grad_norm": 0.27162882685661316,
      "learning_rate": 3e-05,
      "loss": 2.4263,
      "step": 2950
    },
    {
      "epoch": 0.9105214439987658,
      "grad_norm": 0.37277480959892273,
      "learning_rate": 3e-05,
      "loss": 2.9832,
      "step": 2951
    },
    {
      "epoch": 0.9108299907435976,
      "grad_norm": 0.28416138887405396,
      "learning_rate": 3e-05,
      "loss": 2.7082,
      "step": 2952
    },
    {
      "epoch": 0.9111385374884295,
      "grad_norm": 0.3010144531726837,
      "learning_rate": 3e-05,
      "loss": 2.463,
      "step": 2953
    },
    {
      "epoch": 0.9114470842332614,
      "grad_norm": 0.3462345600128174,
      "learning_rate": 3e-05,
      "loss": 2.603,
      "step": 2954
    },
    {
      "epoch": 0.9117556309780932,
      "grad_norm": 0.3923610746860504,
      "learning_rate": 3e-05,
      "loss": 3.1361,
      "step": 2955
    },
    {
      "epoch": 0.912064177722925,
      "grad_norm": 0.3037774860858917,
      "learning_rate": 3e-05,
      "loss": 2.6023,
      "step": 2956
    },
    {
      "epoch": 0.9123727244677569,
      "grad_norm": 0.4105221927165985,
      "learning_rate": 3e-05,
      "loss": 3.4412,
      "step": 2957
    },
    {
      "epoch": 0.9126812712125887,
      "grad_norm": 0.32107794284820557,
      "learning_rate": 3e-05,
      "loss": 2.5473,
      "step": 2958
    },
    {
      "epoch": 0.9129898179574205,
      "grad_norm": 0.3359169065952301,
      "learning_rate": 3e-05,
      "loss": 2.2961,
      "step": 2959
    },
    {
      "epoch": 0.9132983647022523,
      "grad_norm": 0.3050844073295593,
      "learning_rate": 3e-05,
      "loss": 2.7848,
      "step": 2960
    },
    {
      "epoch": 0.9136069114470843,
      "grad_norm": 0.3049246668815613,
      "learning_rate": 3e-05,
      "loss": 2.6135,
      "step": 2961
    },
    {
      "epoch": 0.9139154581919161,
      "grad_norm": 0.3283320665359497,
      "learning_rate": 3e-05,
      "loss": 2.7752,
      "step": 2962
    },
    {
      "epoch": 0.9142240049367479,
      "grad_norm": 0.3238690495491028,
      "learning_rate": 3e-05,
      "loss": 2.8239,
      "step": 2963
    },
    {
      "epoch": 0.9145325516815798,
      "grad_norm": 0.36948534846305847,
      "learning_rate": 3e-05,
      "loss": 2.6686,
      "step": 2964
    },
    {
      "epoch": 0.9148410984264116,
      "grad_norm": 0.3469662070274353,
      "learning_rate": 3e-05,
      "loss": 2.983,
      "step": 2965
    },
    {
      "epoch": 0.9151496451712434,
      "grad_norm": 0.3363710641860962,
      "learning_rate": 3e-05,
      "loss": 2.9553,
      "step": 2966
    },
    {
      "epoch": 0.9154581919160752,
      "grad_norm": 0.2983749508857727,
      "learning_rate": 3e-05,
      "loss": 2.2588,
      "step": 2967
    },
    {
      "epoch": 0.9157667386609071,
      "grad_norm": 0.43369337916374207,
      "learning_rate": 3e-05,
      "loss": 3.0581,
      "step": 2968
    },
    {
      "epoch": 0.916075285405739,
      "grad_norm": 0.26847320795059204,
      "learning_rate": 3e-05,
      "loss": 2.5063,
      "step": 2969
    },
    {
      "epoch": 0.9163838321505708,
      "grad_norm": 0.29353561997413635,
      "learning_rate": 3e-05,
      "loss": 2.3036,
      "step": 2970
    },
    {
      "epoch": 0.9166923788954027,
      "grad_norm": 0.2675473690032959,
      "learning_rate": 3e-05,
      "loss": 2.763,
      "step": 2971
    },
    {
      "epoch": 0.9170009256402345,
      "grad_norm": 0.2732151746749878,
      "learning_rate": 3e-05,
      "loss": 2.333,
      "step": 2972
    },
    {
      "epoch": 0.9173094723850663,
      "grad_norm": 0.3161793053150177,
      "learning_rate": 3e-05,
      "loss": 2.9189,
      "step": 2973
    },
    {
      "epoch": 0.9176180191298982,
      "grad_norm": 0.2945577800273895,
      "learning_rate": 3e-05,
      "loss": 2.553,
      "step": 2974
    },
    {
      "epoch": 0.91792656587473,
      "grad_norm": 0.3421754837036133,
      "learning_rate": 3e-05,
      "loss": 2.5823,
      "step": 2975
    },
    {
      "epoch": 0.9182351126195618,
      "grad_norm": 0.26880040764808655,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 2976
    },
    {
      "epoch": 0.9185436593643937,
      "grad_norm": 0.2973429262638092,
      "learning_rate": 3e-05,
      "loss": 2.2899,
      "step": 2977
    },
    {
      "epoch": 0.9188522061092256,
      "grad_norm": 0.25606873631477356,
      "learning_rate": 3e-05,
      "loss": 2.2988,
      "step": 2978
    },
    {
      "epoch": 0.9191607528540574,
      "grad_norm": 0.26625877618789673,
      "learning_rate": 3e-05,
      "loss": 2.3976,
      "step": 2979
    },
    {
      "epoch": 0.9194692995988892,
      "grad_norm": 0.3849131464958191,
      "learning_rate": 3e-05,
      "loss": 3.2044,
      "step": 2980
    },
    {
      "epoch": 0.9197778463437211,
      "grad_norm": 0.31118044257164,
      "learning_rate": 3e-05,
      "loss": 2.7884,
      "step": 2981
    },
    {
      "epoch": 0.9200863930885529,
      "grad_norm": 0.3495997190475464,
      "learning_rate": 3e-05,
      "loss": 2.8973,
      "step": 2982
    },
    {
      "epoch": 0.9203949398333847,
      "grad_norm": 0.41311630606651306,
      "learning_rate": 3e-05,
      "loss": 3.1077,
      "step": 2983
    },
    {
      "epoch": 0.9207034865782165,
      "grad_norm": 0.49640941619873047,
      "learning_rate": 3e-05,
      "loss": 2.4385,
      "step": 2984
    },
    {
      "epoch": 0.9210120333230485,
      "grad_norm": 0.44323229789733887,
      "learning_rate": 3e-05,
      "loss": 2.9318,
      "step": 2985
    },
    {
      "epoch": 0.9213205800678803,
      "grad_norm": 0.31517311930656433,
      "learning_rate": 3e-05,
      "loss": 2.6252,
      "step": 2986
    },
    {
      "epoch": 0.9216291268127121,
      "grad_norm": 0.27888211607933044,
      "learning_rate": 3e-05,
      "loss": 2.3119,
      "step": 2987
    },
    {
      "epoch": 0.921937673557544,
      "grad_norm": 0.5321169495582581,
      "learning_rate": 3e-05,
      "loss": 2.8749,
      "step": 2988
    },
    {
      "epoch": 0.9222462203023758,
      "grad_norm": 0.43574631214141846,
      "learning_rate": 3e-05,
      "loss": 3.212,
      "step": 2989
    },
    {
      "epoch": 0.9225547670472076,
      "grad_norm": 0.29864299297332764,
      "learning_rate": 3e-05,
      "loss": 2.5774,
      "step": 2990
    },
    {
      "epoch": 0.9228633137920395,
      "grad_norm": 0.4059937596321106,
      "learning_rate": 3e-05,
      "loss": 2.632,
      "step": 2991
    },
    {
      "epoch": 0.9231718605368714,
      "grad_norm": 0.3200888931751251,
      "learning_rate": 3e-05,
      "loss": 2.7047,
      "step": 2992
    },
    {
      "epoch": 0.9234804072817032,
      "grad_norm": 0.30579492449760437,
      "learning_rate": 3e-05,
      "loss": 2.1994,
      "step": 2993
    },
    {
      "epoch": 0.923788954026535,
      "grad_norm": 0.29702386260032654,
      "learning_rate": 3e-05,
      "loss": 2.4776,
      "step": 2994
    },
    {
      "epoch": 0.9240975007713669,
      "grad_norm": 0.29933762550354004,
      "learning_rate": 3e-05,
      "loss": 2.5334,
      "step": 2995
    },
    {
      "epoch": 0.9244060475161987,
      "grad_norm": 0.6028221845626831,
      "learning_rate": 3e-05,
      "loss": 3.0717,
      "step": 2996
    },
    {
      "epoch": 0.9247145942610305,
      "grad_norm": 0.3761559724807739,
      "learning_rate": 3e-05,
      "loss": 2.8945,
      "step": 2997
    },
    {
      "epoch": 0.9250231410058624,
      "grad_norm": 0.37582680583000183,
      "learning_rate": 3e-05,
      "loss": 3.0179,
      "step": 2998
    },
    {
      "epoch": 0.9253316877506942,
      "grad_norm": 0.2689196765422821,
      "learning_rate": 3e-05,
      "loss": 2.1541,
      "step": 2999
    },
    {
      "epoch": 0.9256402344955261,
      "grad_norm": 0.26969143748283386,
      "learning_rate": 3e-05,
      "loss": 2.5284,
      "step": 3000
    },
    {
      "epoch": 0.925948781240358,
      "grad_norm": 0.30431580543518066,
      "learning_rate": 3e-05,
      "loss": 2.7802,
      "step": 3001
    },
    {
      "epoch": 0.9262573279851898,
      "grad_norm": 0.3005150556564331,
      "learning_rate": 3e-05,
      "loss": 2.6082,
      "step": 3002
    },
    {
      "epoch": 0.9265658747300216,
      "grad_norm": 0.30338209867477417,
      "learning_rate": 3e-05,
      "loss": 2.7605,
      "step": 3003
    },
    {
      "epoch": 0.9268744214748534,
      "grad_norm": 0.38839176297187805,
      "learning_rate": 3e-05,
      "loss": 2.5129,
      "step": 3004
    },
    {
      "epoch": 0.9271829682196853,
      "grad_norm": 0.2709753215312958,
      "learning_rate": 3e-05,
      "loss": 2.3787,
      "step": 3005
    },
    {
      "epoch": 0.9274915149645171,
      "grad_norm": 0.2967621684074402,
      "learning_rate": 3e-05,
      "loss": 2.6576,
      "step": 3006
    },
    {
      "epoch": 0.9278000617093489,
      "grad_norm": 0.2907581925392151,
      "learning_rate": 3e-05,
      "loss": 2.4585,
      "step": 3007
    },
    {
      "epoch": 0.9281086084541809,
      "grad_norm": 0.317719966173172,
      "learning_rate": 3e-05,
      "loss": 2.5241,
      "step": 3008
    },
    {
      "epoch": 0.9284171551990127,
      "grad_norm": 0.2762477993965149,
      "learning_rate": 3e-05,
      "loss": 2.4264,
      "step": 3009
    },
    {
      "epoch": 0.9287257019438445,
      "grad_norm": 0.39438310265541077,
      "learning_rate": 3e-05,
      "loss": 2.7802,
      "step": 3010
    },
    {
      "epoch": 0.9290342486886763,
      "grad_norm": 0.40340957045555115,
      "learning_rate": 3e-05,
      "loss": 2.8903,
      "step": 3011
    },
    {
      "epoch": 0.9293427954335082,
      "grad_norm": 0.31931769847869873,
      "learning_rate": 3e-05,
      "loss": 2.5336,
      "step": 3012
    },
    {
      "epoch": 0.92965134217834,
      "grad_norm": 0.37353551387786865,
      "learning_rate": 3e-05,
      "loss": 3.1989,
      "step": 3013
    },
    {
      "epoch": 0.9299598889231718,
      "grad_norm": 0.3961566090583801,
      "learning_rate": 3e-05,
      "loss": 2.7865,
      "step": 3014
    },
    {
      "epoch": 0.9302684356680037,
      "grad_norm": 0.3376099765300751,
      "learning_rate": 3e-05,
      "loss": 2.8772,
      "step": 3015
    },
    {
      "epoch": 0.9305769824128356,
      "grad_norm": 0.3120309114456177,
      "learning_rate": 3e-05,
      "loss": 2.4902,
      "step": 3016
    },
    {
      "epoch": 0.9308855291576674,
      "grad_norm": 0.3560793101787567,
      "learning_rate": 3e-05,
      "loss": 3.0333,
      "step": 3017
    },
    {
      "epoch": 0.9311940759024993,
      "grad_norm": 0.31683385372161865,
      "learning_rate": 3e-05,
      "loss": 2.4281,
      "step": 3018
    },
    {
      "epoch": 0.9315026226473311,
      "grad_norm": 0.2864643931388855,
      "learning_rate": 3e-05,
      "loss": 2.4216,
      "step": 3019
    },
    {
      "epoch": 0.9318111693921629,
      "grad_norm": 0.37182846665382385,
      "learning_rate": 3e-05,
      "loss": 2.8671,
      "step": 3020
    },
    {
      "epoch": 0.9321197161369947,
      "grad_norm": 0.2641955316066742,
      "learning_rate": 3e-05,
      "loss": 2.4916,
      "step": 3021
    },
    {
      "epoch": 0.9324282628818266,
      "grad_norm": 0.5213820934295654,
      "learning_rate": 3e-05,
      "loss": 3.2508,
      "step": 3022
    },
    {
      "epoch": 0.9327368096266584,
      "grad_norm": 0.3530103266239166,
      "learning_rate": 3e-05,
      "loss": 2.1517,
      "step": 3023
    },
    {
      "epoch": 0.9330453563714903,
      "grad_norm": 0.4100661873817444,
      "learning_rate": 3e-05,
      "loss": 2.4566,
      "step": 3024
    },
    {
      "epoch": 0.9333539031163222,
      "grad_norm": 0.35602623224258423,
      "learning_rate": 3e-05,
      "loss": 2.2966,
      "step": 3025
    },
    {
      "epoch": 0.933662449861154,
      "grad_norm": 0.2843538522720337,
      "learning_rate": 3e-05,
      "loss": 2.2989,
      "step": 3026
    },
    {
      "epoch": 0.9339709966059858,
      "grad_norm": 0.2758252024650574,
      "learning_rate": 3e-05,
      "loss": 2.5311,
      "step": 3027
    },
    {
      "epoch": 0.9342795433508176,
      "grad_norm": 0.3072024881839752,
      "learning_rate": 3e-05,
      "loss": 2.4936,
      "step": 3028
    },
    {
      "epoch": 0.9345880900956495,
      "grad_norm": 0.26311805844306946,
      "learning_rate": 3e-05,
      "loss": 2.3032,
      "step": 3029
    },
    {
      "epoch": 0.9348966368404813,
      "grad_norm": 0.3296268582344055,
      "learning_rate": 3e-05,
      "loss": 2.3564,
      "step": 3030
    },
    {
      "epoch": 0.9352051835853131,
      "grad_norm": 0.5471877455711365,
      "learning_rate": 3e-05,
      "loss": 3.2863,
      "step": 3031
    },
    {
      "epoch": 0.9355137303301451,
      "grad_norm": 0.33055728673934937,
      "learning_rate": 3e-05,
      "loss": 2.7968,
      "step": 3032
    },
    {
      "epoch": 0.9358222770749769,
      "grad_norm": 0.43918079137802124,
      "learning_rate": 3e-05,
      "loss": 3.0845,
      "step": 3033
    },
    {
      "epoch": 0.9361308238198087,
      "grad_norm": 0.4258155822753906,
      "learning_rate": 3e-05,
      "loss": 2.7634,
      "step": 3034
    },
    {
      "epoch": 0.9364393705646405,
      "grad_norm": 0.4386035203933716,
      "learning_rate": 3e-05,
      "loss": 2.7661,
      "step": 3035
    },
    {
      "epoch": 0.9367479173094724,
      "grad_norm": 0.3190198242664337,
      "learning_rate": 3e-05,
      "loss": 2.875,
      "step": 3036
    },
    {
      "epoch": 0.9370564640543042,
      "grad_norm": 0.303249329328537,
      "learning_rate": 3e-05,
      "loss": 2.4214,
      "step": 3037
    },
    {
      "epoch": 0.937365010799136,
      "grad_norm": 0.3624988794326782,
      "learning_rate": 3e-05,
      "loss": 2.5387,
      "step": 3038
    },
    {
      "epoch": 0.9376735575439679,
      "grad_norm": 0.3785219192504883,
      "learning_rate": 3e-05,
      "loss": 2.6126,
      "step": 3039
    },
    {
      "epoch": 0.9379821042887998,
      "grad_norm": 0.3676900565624237,
      "learning_rate": 3e-05,
      "loss": 2.7926,
      "step": 3040
    },
    {
      "epoch": 0.9382906510336316,
      "grad_norm": 0.42263004183769226,
      "learning_rate": 3e-05,
      "loss": 3.0789,
      "step": 3041
    },
    {
      "epoch": 0.9385991977784635,
      "grad_norm": 0.24622249603271484,
      "learning_rate": 3e-05,
      "loss": 2.2995,
      "step": 3042
    },
    {
      "epoch": 0.9389077445232953,
      "grad_norm": 0.2926445007324219,
      "learning_rate": 3e-05,
      "loss": 2.4527,
      "step": 3043
    },
    {
      "epoch": 0.9392162912681271,
      "grad_norm": 0.3879694640636444,
      "learning_rate": 3e-05,
      "loss": 2.8512,
      "step": 3044
    },
    {
      "epoch": 0.9395248380129589,
      "grad_norm": 0.298482209444046,
      "learning_rate": 3e-05,
      "loss": 2.4877,
      "step": 3045
    },
    {
      "epoch": 0.9398333847577908,
      "grad_norm": 0.3013835847377777,
      "learning_rate": 3e-05,
      "loss": 2.2311,
      "step": 3046
    },
    {
      "epoch": 0.9401419315026226,
      "grad_norm": 0.4243490695953369,
      "learning_rate": 3e-05,
      "loss": 2.4454,
      "step": 3047
    },
    {
      "epoch": 0.9404504782474545,
      "grad_norm": 0.3296166658401489,
      "learning_rate": 3e-05,
      "loss": 2.523,
      "step": 3048
    },
    {
      "epoch": 0.9407590249922864,
      "grad_norm": 0.2700132727622986,
      "learning_rate": 3e-05,
      "loss": 2.3791,
      "step": 3049
    },
    {
      "epoch": 0.9410675717371182,
      "grad_norm": 0.29857659339904785,
      "learning_rate": 3e-05,
      "loss": 2.4644,
      "step": 3050
    },
    {
      "epoch": 0.94137611848195,
      "grad_norm": 0.3733389973640442,
      "learning_rate": 3e-05,
      "loss": 2.7213,
      "step": 3051
    },
    {
      "epoch": 0.9416846652267818,
      "grad_norm": 0.26306989789009094,
      "learning_rate": 3e-05,
      "loss": 2.4874,
      "step": 3052
    },
    {
      "epoch": 0.9419932119716137,
      "grad_norm": 0.290010541677475,
      "learning_rate": 3e-05,
      "loss": 2.5925,
      "step": 3053
    },
    {
      "epoch": 0.9423017587164455,
      "grad_norm": 0.2788572907447815,
      "learning_rate": 3e-05,
      "loss": 2.5187,
      "step": 3054
    },
    {
      "epoch": 0.9426103054612773,
      "grad_norm": 0.29622915387153625,
      "learning_rate": 3e-05,
      "loss": 2.9425,
      "step": 3055
    },
    {
      "epoch": 0.9429188522061093,
      "grad_norm": 0.31434181332588196,
      "learning_rate": 3e-05,
      "loss": 2.345,
      "step": 3056
    },
    {
      "epoch": 0.9432273989509411,
      "grad_norm": 0.30648863315582275,
      "learning_rate": 3e-05,
      "loss": 2.4803,
      "step": 3057
    },
    {
      "epoch": 0.9435359456957729,
      "grad_norm": 0.34504854679107666,
      "learning_rate": 3e-05,
      "loss": 2.3747,
      "step": 3058
    },
    {
      "epoch": 0.9438444924406048,
      "grad_norm": 0.2548132836818695,
      "learning_rate": 3e-05,
      "loss": 2.3305,
      "step": 3059
    },
    {
      "epoch": 0.9441530391854366,
      "grad_norm": 0.2902987003326416,
      "learning_rate": 3e-05,
      "loss": 2.4311,
      "step": 3060
    },
    {
      "epoch": 0.9444615859302684,
      "grad_norm": 0.4710998833179474,
      "learning_rate": 3e-05,
      "loss": 2.9562,
      "step": 3061
    },
    {
      "epoch": 0.9447701326751002,
      "grad_norm": 0.25020942091941833,
      "learning_rate": 3e-05,
      "loss": 2.0714,
      "step": 3062
    },
    {
      "epoch": 0.9450786794199321,
      "grad_norm": 0.2964976131916046,
      "learning_rate": 3e-05,
      "loss": 2.4769,
      "step": 3063
    },
    {
      "epoch": 0.945387226164764,
      "grad_norm": 0.36232706904411316,
      "learning_rate": 3e-05,
      "loss": 3.2043,
      "step": 3064
    },
    {
      "epoch": 0.9456957729095958,
      "grad_norm": 0.4200202226638794,
      "learning_rate": 3e-05,
      "loss": 2.8737,
      "step": 3065
    },
    {
      "epoch": 0.9460043196544277,
      "grad_norm": 0.31868401169776917,
      "learning_rate": 3e-05,
      "loss": 2.8268,
      "step": 3066
    },
    {
      "epoch": 0.9463128663992595,
      "grad_norm": 0.30646395683288574,
      "learning_rate": 3e-05,
      "loss": 2.6141,
      "step": 3067
    },
    {
      "epoch": 0.9466214131440913,
      "grad_norm": 0.34161990880966187,
      "learning_rate": 3e-05,
      "loss": 2.5159,
      "step": 3068
    },
    {
      "epoch": 0.9469299598889231,
      "grad_norm": 0.38572776317596436,
      "learning_rate": 3e-05,
      "loss": 2.9139,
      "step": 3069
    },
    {
      "epoch": 0.947238506633755,
      "grad_norm": 0.31582701206207275,
      "learning_rate": 3e-05,
      "loss": 2.8355,
      "step": 3070
    },
    {
      "epoch": 0.9475470533785868,
      "grad_norm": 0.3260790705680847,
      "learning_rate": 3e-05,
      "loss": 2.3406,
      "step": 3071
    },
    {
      "epoch": 0.9478556001234187,
      "grad_norm": 0.2831176221370697,
      "learning_rate": 3e-05,
      "loss": 2.5964,
      "step": 3072
    },
    {
      "epoch": 0.9481641468682506,
      "grad_norm": 0.2874281108379364,
      "learning_rate": 3e-05,
      "loss": 2.6755,
      "step": 3073
    },
    {
      "epoch": 0.9484726936130824,
      "grad_norm": 0.298016756772995,
      "learning_rate": 3e-05,
      "loss": 2.4791,
      "step": 3074
    },
    {
      "epoch": 0.9487812403579142,
      "grad_norm": 0.27686169743537903,
      "learning_rate": 3e-05,
      "loss": 2.5687,
      "step": 3075
    },
    {
      "epoch": 0.949089787102746,
      "grad_norm": 0.3424147665500641,
      "learning_rate": 3e-05,
      "loss": 2.8673,
      "step": 3076
    },
    {
      "epoch": 0.9493983338475779,
      "grad_norm": 0.311608225107193,
      "learning_rate": 3e-05,
      "loss": 2.5727,
      "step": 3077
    },
    {
      "epoch": 0.9497068805924097,
      "grad_norm": 0.4081628918647766,
      "learning_rate": 3e-05,
      "loss": 2.4882,
      "step": 3078
    },
    {
      "epoch": 0.9500154273372416,
      "grad_norm": 0.34730857610702515,
      "learning_rate": 3e-05,
      "loss": 2.7376,
      "step": 3079
    },
    {
      "epoch": 0.9503239740820735,
      "grad_norm": 0.39056992530822754,
      "learning_rate": 3e-05,
      "loss": 2.9685,
      "step": 3080
    },
    {
      "epoch": 0.9506325208269053,
      "grad_norm": 0.3361407220363617,
      "learning_rate": 3e-05,
      "loss": 2.6582,
      "step": 3081
    },
    {
      "epoch": 0.9509410675717371,
      "grad_norm": 0.34739845991134644,
      "learning_rate": 3e-05,
      "loss": 3.0305,
      "step": 3082
    },
    {
      "epoch": 0.951249614316569,
      "grad_norm": 0.3131712079048157,
      "learning_rate": 3e-05,
      "loss": 2.81,
      "step": 3083
    },
    {
      "epoch": 0.9515581610614008,
      "grad_norm": 0.2859879732131958,
      "learning_rate": 3e-05,
      "loss": 2.1778,
      "step": 3084
    },
    {
      "epoch": 0.9518667078062326,
      "grad_norm": 0.3703686594963074,
      "learning_rate": 3e-05,
      "loss": 2.7727,
      "step": 3085
    },
    {
      "epoch": 0.9521752545510644,
      "grad_norm": 0.25854915380477905,
      "learning_rate": 3e-05,
      "loss": 2.3804,
      "step": 3086
    },
    {
      "epoch": 0.9524838012958964,
      "grad_norm": 0.3016381561756134,
      "learning_rate": 3e-05,
      "loss": 2.2427,
      "step": 3087
    },
    {
      "epoch": 0.9527923480407282,
      "grad_norm": 0.3392946720123291,
      "learning_rate": 3e-05,
      "loss": 2.8238,
      "step": 3088
    },
    {
      "epoch": 0.95310089478556,
      "grad_norm": 0.3164759874343872,
      "learning_rate": 3e-05,
      "loss": 3.1186,
      "step": 3089
    },
    {
      "epoch": 0.9534094415303919,
      "grad_norm": 0.3679589331150055,
      "learning_rate": 3e-05,
      "loss": 2.8094,
      "step": 3090
    },
    {
      "epoch": 0.9537179882752237,
      "grad_norm": 0.3372891843318939,
      "learning_rate": 3e-05,
      "loss": 2.9655,
      "step": 3091
    },
    {
      "epoch": 0.9540265350200555,
      "grad_norm": 0.3268563449382782,
      "learning_rate": 3e-05,
      "loss": 2.5682,
      "step": 3092
    },
    {
      "epoch": 0.9543350817648873,
      "grad_norm": 0.31734415888786316,
      "learning_rate": 3e-05,
      "loss": 2.3077,
      "step": 3093
    },
    {
      "epoch": 0.9546436285097192,
      "grad_norm": 0.2737109959125519,
      "learning_rate": 3e-05,
      "loss": 2.4163,
      "step": 3094
    },
    {
      "epoch": 0.9549521752545511,
      "grad_norm": 0.2765163779258728,
      "learning_rate": 3e-05,
      "loss": 2.4372,
      "step": 3095
    },
    {
      "epoch": 0.955260721999383,
      "grad_norm": 0.3204086720943451,
      "learning_rate": 3e-05,
      "loss": 2.7632,
      "step": 3096
    },
    {
      "epoch": 0.9555692687442148,
      "grad_norm": 0.33691713213920593,
      "learning_rate": 3e-05,
      "loss": 2.7507,
      "step": 3097
    },
    {
      "epoch": 0.9558778154890466,
      "grad_norm": 0.4027382731437683,
      "learning_rate": 3e-05,
      "loss": 2.8537,
      "step": 3098
    },
    {
      "epoch": 0.9561863622338784,
      "grad_norm": 0.35876747965812683,
      "learning_rate": 3e-05,
      "loss": 2.7271,
      "step": 3099
    },
    {
      "epoch": 0.9564949089787103,
      "grad_norm": 0.5603698492050171,
      "learning_rate": 3e-05,
      "loss": 3.3758,
      "step": 3100
    },
    {
      "epoch": 0.9568034557235421,
      "grad_norm": 0.3716014325618744,
      "learning_rate": 3e-05,
      "loss": 3.2096,
      "step": 3101
    },
    {
      "epoch": 0.9571120024683739,
      "grad_norm": 0.32186856865882874,
      "learning_rate": 3e-05,
      "loss": 2.6839,
      "step": 3102
    },
    {
      "epoch": 0.9574205492132059,
      "grad_norm": 0.38869190216064453,
      "learning_rate": 3e-05,
      "loss": 2.9946,
      "step": 3103
    },
    {
      "epoch": 0.9577290959580377,
      "grad_norm": 0.3592016100883484,
      "learning_rate": 3e-05,
      "loss": 2.2624,
      "step": 3104
    },
    {
      "epoch": 0.9580376427028695,
      "grad_norm": 0.2958182096481323,
      "learning_rate": 3e-05,
      "loss": 2.2418,
      "step": 3105
    },
    {
      "epoch": 0.9583461894477013,
      "grad_norm": 0.41905367374420166,
      "learning_rate": 3e-05,
      "loss": 3.3198,
      "step": 3106
    },
    {
      "epoch": 0.9586547361925332,
      "grad_norm": 0.2440594583749771,
      "learning_rate": 3e-05,
      "loss": 2.1861,
      "step": 3107
    },
    {
      "epoch": 0.958963282937365,
      "grad_norm": 0.321948379278183,
      "learning_rate": 3e-05,
      "loss": 2.4788,
      "step": 3108
    },
    {
      "epoch": 0.9592718296821968,
      "grad_norm": 0.3694811761379242,
      "learning_rate": 3e-05,
      "loss": 2.4157,
      "step": 3109
    },
    {
      "epoch": 0.9595803764270286,
      "grad_norm": 0.2569068968296051,
      "learning_rate": 3e-05,
      "loss": 2.4516,
      "step": 3110
    },
    {
      "epoch": 0.9598889231718606,
      "grad_norm": 0.3426108956336975,
      "learning_rate": 3e-05,
      "loss": 3.172,
      "step": 3111
    },
    {
      "epoch": 0.9601974699166924,
      "grad_norm": 0.25434866547584534,
      "learning_rate": 3e-05,
      "loss": 2.3473,
      "step": 3112
    },
    {
      "epoch": 0.9605060166615242,
      "grad_norm": 0.29196760058403015,
      "learning_rate": 3e-05,
      "loss": 2.7718,
      "step": 3113
    },
    {
      "epoch": 0.9608145634063561,
      "grad_norm": 0.2587181329727173,
      "learning_rate": 3e-05,
      "loss": 2.6479,
      "step": 3114
    },
    {
      "epoch": 0.9611231101511879,
      "grad_norm": 0.3188501298427582,
      "learning_rate": 3e-05,
      "loss": 2.7803,
      "step": 3115
    },
    {
      "epoch": 0.9614316568960197,
      "grad_norm": 0.32521408796310425,
      "learning_rate": 3e-05,
      "loss": 2.298,
      "step": 3116
    },
    {
      "epoch": 0.9617402036408516,
      "grad_norm": 0.25548383593559265,
      "learning_rate": 3e-05,
      "loss": 2.3261,
      "step": 3117
    },
    {
      "epoch": 0.9620487503856834,
      "grad_norm": 0.35770636796951294,
      "learning_rate": 3e-05,
      "loss": 3.0867,
      "step": 3118
    },
    {
      "epoch": 0.9623572971305153,
      "grad_norm": 0.30941280722618103,
      "learning_rate": 3e-05,
      "loss": 2.6328,
      "step": 3119
    },
    {
      "epoch": 0.9626658438753471,
      "grad_norm": 0.29276618361473083,
      "learning_rate": 3e-05,
      "loss": 2.6992,
      "step": 3120
    },
    {
      "epoch": 0.962974390620179,
      "grad_norm": 0.3974364101886749,
      "learning_rate": 3e-05,
      "loss": 2.4356,
      "step": 3121
    },
    {
      "epoch": 0.9632829373650108,
      "grad_norm": 0.29954880475997925,
      "learning_rate": 3e-05,
      "loss": 2.595,
      "step": 3122
    },
    {
      "epoch": 0.9635914841098426,
      "grad_norm": 0.27460622787475586,
      "learning_rate": 3e-05,
      "loss": 2.6094,
      "step": 3123
    },
    {
      "epoch": 0.9639000308546745,
      "grad_norm": 0.3183707892894745,
      "learning_rate": 3e-05,
      "loss": 2.8685,
      "step": 3124
    },
    {
      "epoch": 0.9642085775995063,
      "grad_norm": 0.3339724540710449,
      "learning_rate": 3e-05,
      "loss": 3.0597,
      "step": 3125
    },
    {
      "epoch": 0.9645171243443381,
      "grad_norm": 0.36868351697921753,
      "learning_rate": 3e-05,
      "loss": 2.4267,
      "step": 3126
    },
    {
      "epoch": 0.9648256710891701,
      "grad_norm": 0.2693444490432739,
      "learning_rate": 3e-05,
      "loss": 2.1895,
      "step": 3127
    },
    {
      "epoch": 0.9651342178340019,
      "grad_norm": 0.2520667314529419,
      "learning_rate": 3e-05,
      "loss": 2.439,
      "step": 3128
    },
    {
      "epoch": 0.9654427645788337,
      "grad_norm": 0.36487773060798645,
      "learning_rate": 3e-05,
      "loss": 2.8229,
      "step": 3129
    },
    {
      "epoch": 0.9657513113236655,
      "grad_norm": 0.38236555457115173,
      "learning_rate": 3e-05,
      "loss": 2.9513,
      "step": 3130
    },
    {
      "epoch": 0.9660598580684974,
      "grad_norm": 0.27006927132606506,
      "learning_rate": 3e-05,
      "loss": 2.2393,
      "step": 3131
    },
    {
      "epoch": 0.9663684048133292,
      "grad_norm": 0.2978242039680481,
      "learning_rate": 3e-05,
      "loss": 2.6186,
      "step": 3132
    },
    {
      "epoch": 0.966676951558161,
      "grad_norm": 0.3456973135471344,
      "learning_rate": 3e-05,
      "loss": 2.758,
      "step": 3133
    },
    {
      "epoch": 0.9669854983029929,
      "grad_norm": 0.3259716331958771,
      "learning_rate": 3e-05,
      "loss": 2.7783,
      "step": 3134
    },
    {
      "epoch": 0.9672940450478248,
      "grad_norm": 0.42309021949768066,
      "learning_rate": 3e-05,
      "loss": 2.8363,
      "step": 3135
    },
    {
      "epoch": 0.9676025917926566,
      "grad_norm": 0.3220120370388031,
      "learning_rate": 3e-05,
      "loss": 2.828,
      "step": 3136
    },
    {
      "epoch": 0.9679111385374884,
      "grad_norm": 0.34470850229263306,
      "learning_rate": 3e-05,
      "loss": 2.3379,
      "step": 3137
    },
    {
      "epoch": 0.9682196852823203,
      "grad_norm": 0.33230939507484436,
      "learning_rate": 3e-05,
      "loss": 2.9445,
      "step": 3138
    },
    {
      "epoch": 0.9685282320271521,
      "grad_norm": 0.2927345037460327,
      "learning_rate": 3e-05,
      "loss": 2.3517,
      "step": 3139
    },
    {
      "epoch": 0.9688367787719839,
      "grad_norm": 0.29498234391212463,
      "learning_rate": 3e-05,
      "loss": 2.8351,
      "step": 3140
    },
    {
      "epoch": 0.9691453255168158,
      "grad_norm": 0.24827489256858826,
      "learning_rate": 3e-05,
      "loss": 2.2122,
      "step": 3141
    },
    {
      "epoch": 0.9694538722616476,
      "grad_norm": 0.2774203419685364,
      "learning_rate": 3e-05,
      "loss": 2.3667,
      "step": 3142
    },
    {
      "epoch": 0.9697624190064795,
      "grad_norm": 0.2662864327430725,
      "learning_rate": 3e-05,
      "loss": 2.559,
      "step": 3143
    },
    {
      "epoch": 0.9700709657513114,
      "grad_norm": 0.2957327961921692,
      "learning_rate": 3e-05,
      "loss": 2.4991,
      "step": 3144
    },
    {
      "epoch": 0.9703795124961432,
      "grad_norm": 0.3074386715888977,
      "learning_rate": 3e-05,
      "loss": 2.6943,
      "step": 3145
    },
    {
      "epoch": 0.970688059240975,
      "grad_norm": 0.40129292011260986,
      "learning_rate": 3e-05,
      "loss": 2.7115,
      "step": 3146
    },
    {
      "epoch": 0.9709966059858068,
      "grad_norm": 0.3249188959598541,
      "learning_rate": 3e-05,
      "loss": 2.3602,
      "step": 3147
    },
    {
      "epoch": 0.9713051527306387,
      "grad_norm": 0.2535168528556824,
      "learning_rate": 3e-05,
      "loss": 2.3057,
      "step": 3148
    },
    {
      "epoch": 0.9716136994754705,
      "grad_norm": 0.31744763255119324,
      "learning_rate": 3e-05,
      "loss": 2.8897,
      "step": 3149
    },
    {
      "epoch": 0.9719222462203023,
      "grad_norm": 0.2633349299430847,
      "learning_rate": 3e-05,
      "loss": 2.623,
      "step": 3150
    },
    {
      "epoch": 0.9722307929651343,
      "grad_norm": 0.30141910910606384,
      "learning_rate": 3e-05,
      "loss": 2.4993,
      "step": 3151
    },
    {
      "epoch": 0.9725393397099661,
      "grad_norm": 0.3821033835411072,
      "learning_rate": 3e-05,
      "loss": 3.1187,
      "step": 3152
    },
    {
      "epoch": 0.9728478864547979,
      "grad_norm": 0.2582380771636963,
      "learning_rate": 3e-05,
      "loss": 2.3634,
      "step": 3153
    },
    {
      "epoch": 0.9731564331996297,
      "grad_norm": 0.24997937679290771,
      "learning_rate": 3e-05,
      "loss": 2.2268,
      "step": 3154
    },
    {
      "epoch": 0.9734649799444616,
      "grad_norm": 0.28298869729042053,
      "learning_rate": 3e-05,
      "loss": 2.6604,
      "step": 3155
    },
    {
      "epoch": 0.9737735266892934,
      "grad_norm": 0.27607831358909607,
      "learning_rate": 3e-05,
      "loss": 2.4942,
      "step": 3156
    },
    {
      "epoch": 0.9740820734341252,
      "grad_norm": 0.3434098958969116,
      "learning_rate": 3e-05,
      "loss": 3.1165,
      "step": 3157
    },
    {
      "epoch": 0.9743906201789572,
      "grad_norm": 0.4422709047794342,
      "learning_rate": 3e-05,
      "loss": 2.959,
      "step": 3158
    },
    {
      "epoch": 0.974699166923789,
      "grad_norm": 0.26411694288253784,
      "learning_rate": 3e-05,
      "loss": 2.2174,
      "step": 3159
    },
    {
      "epoch": 0.9750077136686208,
      "grad_norm": 0.4978486895561218,
      "learning_rate": 3e-05,
      "loss": 3.2263,
      "step": 3160
    },
    {
      "epoch": 0.9753162604134527,
      "grad_norm": 0.4098454713821411,
      "learning_rate": 3e-05,
      "loss": 2.8507,
      "step": 3161
    },
    {
      "epoch": 0.9756248071582845,
      "grad_norm": 0.27312561869621277,
      "learning_rate": 3e-05,
      "loss": 2.1998,
      "step": 3162
    },
    {
      "epoch": 0.9759333539031163,
      "grad_norm": 0.3341294825077057,
      "learning_rate": 3e-05,
      "loss": 2.3678,
      "step": 3163
    },
    {
      "epoch": 0.9762419006479481,
      "grad_norm": 0.3833390474319458,
      "learning_rate": 3e-05,
      "loss": 2.5112,
      "step": 3164
    },
    {
      "epoch": 0.97655044739278,
      "grad_norm": 0.3587063252925873,
      "learning_rate": 3e-05,
      "loss": 2.5817,
      "step": 3165
    },
    {
      "epoch": 0.9768589941376119,
      "grad_norm": 0.34508004784584045,
      "learning_rate": 3e-05,
      "loss": 2.853,
      "step": 3166
    },
    {
      "epoch": 0.9771675408824437,
      "grad_norm": 0.29305341839790344,
      "learning_rate": 3e-05,
      "loss": 2.4678,
      "step": 3167
    },
    {
      "epoch": 0.9774760876272756,
      "grad_norm": 0.28883805871009827,
      "learning_rate": 3e-05,
      "loss": 2.438,
      "step": 3168
    },
    {
      "epoch": 0.9777846343721074,
      "grad_norm": 0.328971266746521,
      "learning_rate": 3e-05,
      "loss": 2.6635,
      "step": 3169
    },
    {
      "epoch": 0.9780931811169392,
      "grad_norm": 0.268512099981308,
      "learning_rate": 3e-05,
      "loss": 2.2994,
      "step": 3170
    },
    {
      "epoch": 0.978401727861771,
      "grad_norm": 0.3973873257637024,
      "learning_rate": 3e-05,
      "loss": 2.5279,
      "step": 3171
    },
    {
      "epoch": 0.9787102746066029,
      "grad_norm": 0.37665578722953796,
      "learning_rate": 3e-05,
      "loss": 2.9962,
      "step": 3172
    },
    {
      "epoch": 0.9790188213514347,
      "grad_norm": 0.314129114151001,
      "learning_rate": 3e-05,
      "loss": 2.2576,
      "step": 3173
    },
    {
      "epoch": 0.9793273680962666,
      "grad_norm": 0.3951795995235443,
      "learning_rate": 3e-05,
      "loss": 3.1135,
      "step": 3174
    },
    {
      "epoch": 0.9796359148410985,
      "grad_norm": 0.3149474263191223,
      "learning_rate": 3e-05,
      "loss": 2.7143,
      "step": 3175
    },
    {
      "epoch": 0.9799444615859303,
      "grad_norm": 0.38547009229660034,
      "learning_rate": 3e-05,
      "loss": 2.5525,
      "step": 3176
    },
    {
      "epoch": 0.9802530083307621,
      "grad_norm": 0.4109960198402405,
      "learning_rate": 3e-05,
      "loss": 3.2388,
      "step": 3177
    },
    {
      "epoch": 0.980561555075594,
      "grad_norm": 0.36622384190559387,
      "learning_rate": 3e-05,
      "loss": 2.4511,
      "step": 3178
    },
    {
      "epoch": 0.9808701018204258,
      "grad_norm": 0.30771172046661377,
      "learning_rate": 3e-05,
      "loss": 2.2813,
      "step": 3179
    },
    {
      "epoch": 0.9811786485652576,
      "grad_norm": 0.30412936210632324,
      "learning_rate": 3e-05,
      "loss": 2.5285,
      "step": 3180
    },
    {
      "epoch": 0.9814871953100894,
      "grad_norm": 0.3460204005241394,
      "learning_rate": 3e-05,
      "loss": 2.6651,
      "step": 3181
    },
    {
      "epoch": 0.9817957420549214,
      "grad_norm": 0.31598231196403503,
      "learning_rate": 3e-05,
      "loss": 2.4141,
      "step": 3182
    },
    {
      "epoch": 0.9821042887997532,
      "grad_norm": 0.3302285075187683,
      "learning_rate": 3e-05,
      "loss": 2.9236,
      "step": 3183
    },
    {
      "epoch": 0.982412835544585,
      "grad_norm": 0.342619389295578,
      "learning_rate": 3e-05,
      "loss": 2.8579,
      "step": 3184
    },
    {
      "epoch": 0.9827213822894169,
      "grad_norm": 0.3031438887119293,
      "learning_rate": 3e-05,
      "loss": 2.4702,
      "step": 3185
    },
    {
      "epoch": 0.9830299290342487,
      "grad_norm": 0.3054526746273041,
      "learning_rate": 3e-05,
      "loss": 2.7021,
      "step": 3186
    },
    {
      "epoch": 0.9833384757790805,
      "grad_norm": 0.32133397459983826,
      "learning_rate": 3e-05,
      "loss": 2.8558,
      "step": 3187
    },
    {
      "epoch": 0.9836470225239123,
      "grad_norm": 0.2902354300022125,
      "learning_rate": 3e-05,
      "loss": 2.5618,
      "step": 3188
    },
    {
      "epoch": 0.9839555692687442,
      "grad_norm": 0.31177830696105957,
      "learning_rate": 3e-05,
      "loss": 2.6112,
      "step": 3189
    },
    {
      "epoch": 0.9842641160135761,
      "grad_norm": 0.336733877658844,
      "learning_rate": 3e-05,
      "loss": 2.9405,
      "step": 3190
    },
    {
      "epoch": 0.9845726627584079,
      "grad_norm": 0.34188321232795715,
      "learning_rate": 3e-05,
      "loss": 3.1409,
      "step": 3191
    },
    {
      "epoch": 0.9848812095032398,
      "grad_norm": 0.33917924761772156,
      "learning_rate": 3e-05,
      "loss": 2.4822,
      "step": 3192
    },
    {
      "epoch": 0.9851897562480716,
      "grad_norm": 0.3156522214412689,
      "learning_rate": 3e-05,
      "loss": 2.8875,
      "step": 3193
    },
    {
      "epoch": 0.9854983029929034,
      "grad_norm": 0.3125807046890259,
      "learning_rate": 3e-05,
      "loss": 2.868,
      "step": 3194
    },
    {
      "epoch": 0.9858068497377352,
      "grad_norm": 0.2903933823108673,
      "learning_rate": 3e-05,
      "loss": 2.5265,
      "step": 3195
    },
    {
      "epoch": 0.9861153964825671,
      "grad_norm": 0.3696379065513611,
      "learning_rate": 3e-05,
      "loss": 2.8984,
      "step": 3196
    },
    {
      "epoch": 0.9864239432273989,
      "grad_norm": 0.33664941787719727,
      "learning_rate": 3e-05,
      "loss": 2.9207,
      "step": 3197
    },
    {
      "epoch": 0.9867324899722308,
      "grad_norm": 0.3234494626522064,
      "learning_rate": 3e-05,
      "loss": 2.8782,
      "step": 3198
    },
    {
      "epoch": 0.9870410367170627,
      "grad_norm": 0.33355656266212463,
      "learning_rate": 3e-05,
      "loss": 2.6745,
      "step": 3199
    },
    {
      "epoch": 0.9873495834618945,
      "grad_norm": 0.30930057168006897,
      "learning_rate": 3e-05,
      "loss": 2.3082,
      "step": 3200
    },
    {
      "epoch": 0.9876581302067263,
      "grad_norm": 0.28265678882598877,
      "learning_rate": 3e-05,
      "loss": 2.2819,
      "step": 3201
    },
    {
      "epoch": 0.9879666769515582,
      "grad_norm": 0.4328087568283081,
      "learning_rate": 3e-05,
      "loss": 2.9072,
      "step": 3202
    },
    {
      "epoch": 0.98827522369639,
      "grad_norm": 0.2890687882900238,
      "learning_rate": 3e-05,
      "loss": 2.3122,
      "step": 3203
    },
    {
      "epoch": 0.9885837704412218,
      "grad_norm": 0.26304498314857483,
      "learning_rate": 3e-05,
      "loss": 2.1147,
      "step": 3204
    },
    {
      "epoch": 0.9888923171860536,
      "grad_norm": 0.35160088539123535,
      "learning_rate": 3e-05,
      "loss": 2.8558,
      "step": 3205
    },
    {
      "epoch": 0.9892008639308856,
      "grad_norm": 0.2529274821281433,
      "learning_rate": 3e-05,
      "loss": 2.3776,
      "step": 3206
    },
    {
      "epoch": 0.9895094106757174,
      "grad_norm": 0.32802489399909973,
      "learning_rate": 3e-05,
      "loss": 2.4024,
      "step": 3207
    },
    {
      "epoch": 0.9898179574205492,
      "grad_norm": 0.30960068106651306,
      "learning_rate": 3e-05,
      "loss": 2.624,
      "step": 3208
    },
    {
      "epoch": 0.9901265041653811,
      "grad_norm": 0.2918947637081146,
      "learning_rate": 3e-05,
      "loss": 2.4104,
      "step": 3209
    },
    {
      "epoch": 0.9904350509102129,
      "grad_norm": 0.3591970205307007,
      "learning_rate": 3e-05,
      "loss": 3.1794,
      "step": 3210
    },
    {
      "epoch": 0.9907435976550447,
      "grad_norm": 0.2975943982601166,
      "learning_rate": 3e-05,
      "loss": 2.6902,
      "step": 3211
    },
    {
      "epoch": 0.9910521443998765,
      "grad_norm": 0.27048832178115845,
      "learning_rate": 3e-05,
      "loss": 2.7187,
      "step": 3212
    },
    {
      "epoch": 0.9913606911447084,
      "grad_norm": 0.26648789644241333,
      "learning_rate": 3e-05,
      "loss": 2.521,
      "step": 3213
    },
    {
      "epoch": 0.9916692378895403,
      "grad_norm": 0.3873830735683441,
      "learning_rate": 3e-05,
      "loss": 2.711,
      "step": 3214
    },
    {
      "epoch": 0.9919777846343721,
      "grad_norm": 0.40551069378852844,
      "learning_rate": 3e-05,
      "loss": 2.6142,
      "step": 3215
    },
    {
      "epoch": 0.992286331379204,
      "grad_norm": 0.37170445919036865,
      "learning_rate": 3e-05,
      "loss": 2.7867,
      "step": 3216
    },
    {
      "epoch": 0.9925948781240358,
      "grad_norm": 0.48047205805778503,
      "learning_rate": 3e-05,
      "loss": 2.2022,
      "step": 3217
    },
    {
      "epoch": 0.9929034248688676,
      "grad_norm": 0.28528162837028503,
      "learning_rate": 3e-05,
      "loss": 2.352,
      "step": 3218
    },
    {
      "epoch": 0.9932119716136995,
      "grad_norm": 0.28340965509414673,
      "learning_rate": 3e-05,
      "loss": 2.5986,
      "step": 3219
    },
    {
      "epoch": 0.9935205183585313,
      "grad_norm": 0.3123044967651367,
      "learning_rate": 3e-05,
      "loss": 2.4165,
      "step": 3220
    },
    {
      "epoch": 0.9938290651033631,
      "grad_norm": 0.3788071274757385,
      "learning_rate": 3e-05,
      "loss": 2.6978,
      "step": 3221
    },
    {
      "epoch": 0.994137611848195,
      "grad_norm": 0.30537885427474976,
      "learning_rate": 3e-05,
      "loss": 2.4919,
      "step": 3222
    },
    {
      "epoch": 0.9944461585930269,
      "grad_norm": 0.2902875542640686,
      "learning_rate": 3e-05,
      "loss": 2.3386,
      "step": 3223
    },
    {
      "epoch": 0.9947547053378587,
      "grad_norm": 0.34292489290237427,
      "learning_rate": 3e-05,
      "loss": 2.3101,
      "step": 3224
    },
    {
      "epoch": 0.9950632520826905,
      "grad_norm": 0.33788037300109863,
      "learning_rate": 3e-05,
      "loss": 2.4423,
      "step": 3225
    },
    {
      "epoch": 0.9953717988275224,
      "grad_norm": 0.31140759587287903,
      "learning_rate": 3e-05,
      "loss": 2.8311,
      "step": 3226
    },
    {
      "epoch": 0.9956803455723542,
      "grad_norm": 0.25094249844551086,
      "learning_rate": 3e-05,
      "loss": 2.1127,
      "step": 3227
    },
    {
      "epoch": 0.995988892317186,
      "grad_norm": 0.3538408577442169,
      "learning_rate": 3e-05,
      "loss": 2.6853,
      "step": 3228
    },
    {
      "epoch": 0.9962974390620178,
      "grad_norm": 0.26211321353912354,
      "learning_rate": 3e-05,
      "loss": 2.3652,
      "step": 3229
    },
    {
      "epoch": 0.9966059858068498,
      "grad_norm": 0.43647173047065735,
      "learning_rate": 3e-05,
      "loss": 2.9222,
      "step": 3230
    },
    {
      "epoch": 0.9969145325516816,
      "grad_norm": 0.44634518027305603,
      "learning_rate": 3e-05,
      "loss": 2.5942,
      "step": 3231
    },
    {
      "epoch": 0.9972230792965134,
      "grad_norm": 0.3625410497188568,
      "learning_rate": 3e-05,
      "loss": 2.581,
      "step": 3232
    },
    {
      "epoch": 0.9975316260413453,
      "grad_norm": 0.37610533833503723,
      "learning_rate": 3e-05,
      "loss": 2.5643,
      "step": 3233
    },
    {
      "epoch": 0.9978401727861771,
      "grad_norm": 0.45702043175697327,
      "learning_rate": 3e-05,
      "loss": 2.9001,
      "step": 3234
    },
    {
      "epoch": 0.9981487195310089,
      "grad_norm": 0.4190535247325897,
      "learning_rate": 3e-05,
      "loss": 2.8825,
      "step": 3235
    },
    {
      "epoch": 0.9984572662758407,
      "grad_norm": 0.2578660845756531,
      "learning_rate": 3e-05,
      "loss": 2.3857,
      "step": 3236
    },
    {
      "epoch": 0.9987658130206726,
      "grad_norm": 0.3452734351158142,
      "learning_rate": 3e-05,
      "loss": 2.5024,
      "step": 3237
    },
    {
      "epoch": 0.9990743597655045,
      "grad_norm": 0.3411957323551178,
      "learning_rate": 3e-05,
      "loss": 2.6994,
      "step": 3238
    },
    {
      "epoch": 0.9993829065103363,
      "grad_norm": 0.32112446427345276,
      "learning_rate": 3e-05,
      "loss": 2.6412,
      "step": 3239
    },
    {
      "epoch": 0.9996914532551682,
      "grad_norm": 0.296281635761261,
      "learning_rate": 3e-05,
      "loss": 2.5655,
      "step": 3240
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.3063308894634247,
      "learning_rate": 3e-05,
      "loss": 2.7056,
      "step": 3241
    },
    {
      "epoch": 1.0,
      "step": 3241,
      "total_flos": 2.49799901214199e+18,
      "train_loss": 2.7246624804396427,
      "train_runtime": 56048.741,
      "train_samples_per_second": 0.463,
      "train_steps_per_second": 0.058
    }
  ],
  "logging_steps": 1,
  "max_steps": 3241,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 300,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.49799901214199e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}