Alphace_email / checkpoint-1569 /trainer_state.json
Sourabh2's picture
Upload folder using huggingface_hub
e68b96d verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 393,
"global_step": 1569,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0006373486297004461,
"grad_norm": 0.8087297677993774,
"learning_rate": 1e-05,
"loss": 1.034,
"step": 1
},
{
"epoch": 0.0006373486297004461,
"eval_loss": 1.095216155052185,
"eval_runtime": 6.6487,
"eval_samples_per_second": 9.927,
"eval_steps_per_second": 4.963,
"step": 1
},
{
"epoch": 0.0012746972594008922,
"grad_norm": 1.013185977935791,
"learning_rate": 2e-05,
"loss": 1.4419,
"step": 2
},
{
"epoch": 0.0019120458891013384,
"grad_norm": 0.8052219748497009,
"learning_rate": 3e-05,
"loss": 0.5742,
"step": 3
},
{
"epoch": 0.0025493945188017845,
"grad_norm": 1.2126351594924927,
"learning_rate": 4e-05,
"loss": 1.5246,
"step": 4
},
{
"epoch": 0.0031867431485022306,
"grad_norm": 0.6865807175636292,
"learning_rate": 5e-05,
"loss": 0.5572,
"step": 5
},
{
"epoch": 0.0038240917782026767,
"grad_norm": 0.8945586085319519,
"learning_rate": 6e-05,
"loss": 0.833,
"step": 6
},
{
"epoch": 0.004461440407903123,
"grad_norm": 0.9489598870277405,
"learning_rate": 7e-05,
"loss": 1.3068,
"step": 7
},
{
"epoch": 0.005098789037603569,
"grad_norm": 0.8268119096755981,
"learning_rate": 8e-05,
"loss": 0.9671,
"step": 8
},
{
"epoch": 0.0057361376673040155,
"grad_norm": 0.5656170845031738,
"learning_rate": 9e-05,
"loss": 0.5968,
"step": 9
},
{
"epoch": 0.006373486297004461,
"grad_norm": 0.681001603603363,
"learning_rate": 0.0001,
"loss": 0.757,
"step": 10
},
{
"epoch": 0.007010834926704908,
"grad_norm": 1.0039377212524414,
"learning_rate": 0.00011000000000000002,
"loss": 0.9267,
"step": 11
},
{
"epoch": 0.0076481835564053535,
"grad_norm": 1.0792843103408813,
"learning_rate": 0.00012,
"loss": 1.0037,
"step": 12
},
{
"epoch": 0.0082855321861058,
"grad_norm": 1.0027225017547607,
"learning_rate": 0.00013000000000000002,
"loss": 0.6703,
"step": 13
},
{
"epoch": 0.008922880815806247,
"grad_norm": 0.9296356439590454,
"learning_rate": 0.00014,
"loss": 0.9243,
"step": 14
},
{
"epoch": 0.009560229445506692,
"grad_norm": 0.9406685829162598,
"learning_rate": 0.00015000000000000001,
"loss": 0.6625,
"step": 15
},
{
"epoch": 0.010197578075207138,
"grad_norm": 1.0628087520599365,
"learning_rate": 0.00016,
"loss": 0.5259,
"step": 16
},
{
"epoch": 0.010834926704907584,
"grad_norm": 0.9398377537727356,
"learning_rate": 0.00017,
"loss": 0.51,
"step": 17
},
{
"epoch": 0.011472275334608031,
"grad_norm": 1.2091987133026123,
"learning_rate": 0.00018,
"loss": 0.6041,
"step": 18
},
{
"epoch": 0.012109623964308477,
"grad_norm": 1.3757166862487793,
"learning_rate": 0.00019,
"loss": 0.7462,
"step": 19
},
{
"epoch": 0.012746972594008922,
"grad_norm": 1.3699002265930176,
"learning_rate": 0.0002,
"loss": 0.4536,
"step": 20
},
{
"epoch": 0.01338432122370937,
"grad_norm": 1.3036978244781494,
"learning_rate": 0.00019999979433198083,
"loss": 0.1559,
"step": 21
},
{
"epoch": 0.014021669853409816,
"grad_norm": 1.7420039176940918,
"learning_rate": 0.00019999917732876928,
"loss": 0.4959,
"step": 22
},
{
"epoch": 0.014659018483110261,
"grad_norm": 1.591464877128601,
"learning_rate": 0.00019999814899290325,
"loss": 0.6398,
"step": 23
},
{
"epoch": 0.015296367112810707,
"grad_norm": 2.641218662261963,
"learning_rate": 0.00019999670932861277,
"loss": 0.5111,
"step": 24
},
{
"epoch": 0.015933715742511154,
"grad_norm": 1.4207754135131836,
"learning_rate": 0.0001999948583418196,
"loss": 0.2899,
"step": 25
},
{
"epoch": 0.0165710643722116,
"grad_norm": 1.778639316558838,
"learning_rate": 0.0001999925960401376,
"loss": 0.3126,
"step": 26
},
{
"epoch": 0.017208413001912046,
"grad_norm": 2.876039743423462,
"learning_rate": 0.00019998992243287233,
"loss": 0.4185,
"step": 27
},
{
"epoch": 0.017845761631612493,
"grad_norm": 2.236043930053711,
"learning_rate": 0.00019998683753102138,
"loss": 0.4636,
"step": 28
},
{
"epoch": 0.018483110261312937,
"grad_norm": 2.8687374591827393,
"learning_rate": 0.00019998334134727405,
"loss": 0.6388,
"step": 29
},
{
"epoch": 0.019120458891013385,
"grad_norm": 2.3187761306762695,
"learning_rate": 0.00019997943389601138,
"loss": 0.1663,
"step": 30
},
{
"epoch": 0.019757807520713832,
"grad_norm": 2.0581860542297363,
"learning_rate": 0.00019997511519330617,
"loss": 0.1046,
"step": 31
},
{
"epoch": 0.020395156150414276,
"grad_norm": 1.3978091478347778,
"learning_rate": 0.00019997038525692272,
"loss": 0.238,
"step": 32
},
{
"epoch": 0.021032504780114723,
"grad_norm": 1.8944082260131836,
"learning_rate": 0.00019996524410631703,
"loss": 0.5103,
"step": 33
},
{
"epoch": 0.021669853409815167,
"grad_norm": 1.8764029741287231,
"learning_rate": 0.00019995969176263653,
"loss": 0.1161,
"step": 34
},
{
"epoch": 0.022307202039515615,
"grad_norm": 1.4465802907943726,
"learning_rate": 0.00019995372824871993,
"loss": 0.3947,
"step": 35
},
{
"epoch": 0.022944550669216062,
"grad_norm": 1.5151886940002441,
"learning_rate": 0.0001999473535890973,
"loss": 0.4018,
"step": 36
},
{
"epoch": 0.023581899298916506,
"grad_norm": 1.7926268577575684,
"learning_rate": 0.00019994056780999,
"loss": 0.4217,
"step": 37
},
{
"epoch": 0.024219247928616953,
"grad_norm": 1.238084316253662,
"learning_rate": 0.0001999333709393104,
"loss": 0.5063,
"step": 38
},
{
"epoch": 0.0248565965583174,
"grad_norm": 1.0243855714797974,
"learning_rate": 0.00019992576300666173,
"loss": 0.2177,
"step": 39
},
{
"epoch": 0.025493945188017845,
"grad_norm": 1.3816263675689697,
"learning_rate": 0.00019991774404333818,
"loss": 0.4822,
"step": 40
},
{
"epoch": 0.026131293817718292,
"grad_norm": 0.7292724251747131,
"learning_rate": 0.00019990931408232467,
"loss": 0.1341,
"step": 41
},
{
"epoch": 0.02676864244741874,
"grad_norm": 0.9624613523483276,
"learning_rate": 0.00019990047315829665,
"loss": 0.1826,
"step": 42
},
{
"epoch": 0.027405991077119184,
"grad_norm": 1.473120927810669,
"learning_rate": 0.00019989122130762007,
"loss": 0.3706,
"step": 43
},
{
"epoch": 0.02804333970681963,
"grad_norm": 1.0263862609863281,
"learning_rate": 0.00019988155856835103,
"loss": 0.25,
"step": 44
},
{
"epoch": 0.028680688336520075,
"grad_norm": 1.335708737373352,
"learning_rate": 0.00019987148498023594,
"loss": 0.3509,
"step": 45
},
{
"epoch": 0.029318036966220522,
"grad_norm": 1.3829046487808228,
"learning_rate": 0.00019986100058471103,
"loss": 0.5565,
"step": 46
},
{
"epoch": 0.02995538559592097,
"grad_norm": 1.0559632778167725,
"learning_rate": 0.0001998501054249025,
"loss": 0.3021,
"step": 47
},
{
"epoch": 0.030592734225621414,
"grad_norm": 1.2597861289978027,
"learning_rate": 0.00019983879954562598,
"loss": 0.1621,
"step": 48
},
{
"epoch": 0.03123008285532186,
"grad_norm": 0.7574909925460815,
"learning_rate": 0.00019982708299338663,
"loss": 0.2451,
"step": 49
},
{
"epoch": 0.03186743148502231,
"grad_norm": 1.3010400533676147,
"learning_rate": 0.00019981495581637893,
"loss": 0.5569,
"step": 50
},
{
"epoch": 0.032504780114722756,
"grad_norm": 1.0032603740692139,
"learning_rate": 0.00019980241806448622,
"loss": 0.3329,
"step": 51
},
{
"epoch": 0.0331421287444232,
"grad_norm": 0.5539001226425171,
"learning_rate": 0.00019978946978928088,
"loss": 0.0715,
"step": 52
},
{
"epoch": 0.033779477374123644,
"grad_norm": 1.075184941291809,
"learning_rate": 0.00019977611104402378,
"loss": 0.5566,
"step": 53
},
{
"epoch": 0.03441682600382409,
"grad_norm": 0.799937903881073,
"learning_rate": 0.00019976234188366429,
"loss": 0.1628,
"step": 54
},
{
"epoch": 0.03505417463352454,
"grad_norm": 1.0306012630462646,
"learning_rate": 0.00019974816236483992,
"loss": 0.3826,
"step": 55
},
{
"epoch": 0.035691523263224986,
"grad_norm": 1.4506051540374756,
"learning_rate": 0.00019973357254587611,
"loss": 0.536,
"step": 56
},
{
"epoch": 0.036328871892925434,
"grad_norm": 1.5342416763305664,
"learning_rate": 0.00019971857248678607,
"loss": 0.6018,
"step": 57
},
{
"epoch": 0.036966220522625874,
"grad_norm": 1.2872010469436646,
"learning_rate": 0.00019970316224927044,
"loss": 0.4248,
"step": 58
},
{
"epoch": 0.03760356915232632,
"grad_norm": 0.7495999336242676,
"learning_rate": 0.00019968734189671712,
"loss": 0.3695,
"step": 59
},
{
"epoch": 0.03824091778202677,
"grad_norm": 0.43265894055366516,
"learning_rate": 0.00019967111149420084,
"loss": 0.0358,
"step": 60
},
{
"epoch": 0.038878266411727216,
"grad_norm": 0.6488798260688782,
"learning_rate": 0.00019965447110848315,
"loss": 0.3138,
"step": 61
},
{
"epoch": 0.039515615041427664,
"grad_norm": 0.8329582214355469,
"learning_rate": 0.00019963742080801198,
"loss": 0.3543,
"step": 62
},
{
"epoch": 0.040152963671128104,
"grad_norm": 1.14444899559021,
"learning_rate": 0.0001996199606629213,
"loss": 0.5133,
"step": 63
},
{
"epoch": 0.04079031230082855,
"grad_norm": 1.1341776847839355,
"learning_rate": 0.00019960209074503101,
"loss": 0.2256,
"step": 64
},
{
"epoch": 0.041427660930529,
"grad_norm": 0.8539294600486755,
"learning_rate": 0.00019958381112784653,
"loss": 0.2212,
"step": 65
},
{
"epoch": 0.04206500956022945,
"grad_norm": 0.9306990504264832,
"learning_rate": 0.00019956512188655845,
"loss": 0.3296,
"step": 66
},
{
"epoch": 0.042702358189929894,
"grad_norm": 1.3836973905563354,
"learning_rate": 0.00019954602309804246,
"loss": 0.7184,
"step": 67
},
{
"epoch": 0.043339706819630335,
"grad_norm": 1.0024141073226929,
"learning_rate": 0.00019952651484085867,
"loss": 0.343,
"step": 68
},
{
"epoch": 0.04397705544933078,
"grad_norm": 1.2005387544631958,
"learning_rate": 0.00019950659719525162,
"loss": 0.6287,
"step": 69
},
{
"epoch": 0.04461440407903123,
"grad_norm": 1.4316396713256836,
"learning_rate": 0.00019948627024314974,
"loss": 0.5861,
"step": 70
},
{
"epoch": 0.04525175270873168,
"grad_norm": 1.3682820796966553,
"learning_rate": 0.00019946553406816514,
"loss": 0.5075,
"step": 71
},
{
"epoch": 0.045889101338432124,
"grad_norm": 1.0369032621383667,
"learning_rate": 0.00019944438875559316,
"loss": 0.1549,
"step": 72
},
{
"epoch": 0.04652644996813257,
"grad_norm": 0.9389836192131042,
"learning_rate": 0.00019942283439241209,
"loss": 0.3257,
"step": 73
},
{
"epoch": 0.04716379859783301,
"grad_norm": 1.2492974996566772,
"learning_rate": 0.00019940087106728278,
"loss": 0.3257,
"step": 74
},
{
"epoch": 0.04780114722753346,
"grad_norm": 0.8527800440788269,
"learning_rate": 0.00019937849887054833,
"loss": 0.4185,
"step": 75
},
{
"epoch": 0.04843849585723391,
"grad_norm": 1.4771286249160767,
"learning_rate": 0.00019935571789423363,
"loss": 0.4198,
"step": 76
},
{
"epoch": 0.049075844486934354,
"grad_norm": 1.241165280342102,
"learning_rate": 0.00019933252823204505,
"loss": 0.4651,
"step": 77
},
{
"epoch": 0.0497131931166348,
"grad_norm": 0.8001418113708496,
"learning_rate": 0.00019930892997937005,
"loss": 0.1951,
"step": 78
},
{
"epoch": 0.05035054174633524,
"grad_norm": 0.7575386762619019,
"learning_rate": 0.0001992849232332767,
"loss": 0.2423,
"step": 79
},
{
"epoch": 0.05098789037603569,
"grad_norm": 1.619662880897522,
"learning_rate": 0.00019926050809251344,
"loss": 0.5817,
"step": 80
},
{
"epoch": 0.05162523900573614,
"grad_norm": 1.3025641441345215,
"learning_rate": 0.00019923568465750853,
"loss": 0.2982,
"step": 81
},
{
"epoch": 0.052262587635436585,
"grad_norm": 1.0632684230804443,
"learning_rate": 0.00019921045303036969,
"loss": 0.3157,
"step": 82
},
{
"epoch": 0.05289993626513703,
"grad_norm": 0.910922646522522,
"learning_rate": 0.0001991848133148837,
"loss": 0.2298,
"step": 83
},
{
"epoch": 0.05353728489483748,
"grad_norm": 1.309713363647461,
"learning_rate": 0.000199158765616516,
"loss": 0.5097,
"step": 84
},
{
"epoch": 0.05417463352453792,
"grad_norm": 0.8286350965499878,
"learning_rate": 0.00019913231004241007,
"loss": 0.2403,
"step": 85
},
{
"epoch": 0.05481198215423837,
"grad_norm": 1.2789089679718018,
"learning_rate": 0.00019910544670138726,
"loss": 0.5141,
"step": 86
},
{
"epoch": 0.055449330783938815,
"grad_norm": 0.7211568355560303,
"learning_rate": 0.00019907817570394623,
"loss": 0.0756,
"step": 87
},
{
"epoch": 0.05608667941363926,
"grad_norm": 1.2010267972946167,
"learning_rate": 0.00019905049716226233,
"loss": 0.2571,
"step": 88
},
{
"epoch": 0.05672402804333971,
"grad_norm": 1.2290879487991333,
"learning_rate": 0.00019902241119018746,
"loss": 0.4677,
"step": 89
},
{
"epoch": 0.05736137667304015,
"grad_norm": 1.0563770532608032,
"learning_rate": 0.00019899391790324927,
"loss": 0.2588,
"step": 90
},
{
"epoch": 0.0579987253027406,
"grad_norm": 1.1716020107269287,
"learning_rate": 0.00019896501741865094,
"loss": 0.2625,
"step": 91
},
{
"epoch": 0.058636073932441045,
"grad_norm": 1.0361477136611938,
"learning_rate": 0.0001989357098552706,
"loss": 0.4864,
"step": 92
},
{
"epoch": 0.05927342256214149,
"grad_norm": 1.0492510795593262,
"learning_rate": 0.00019890599533366082,
"loss": 0.4091,
"step": 93
},
{
"epoch": 0.05991077119184194,
"grad_norm": 1.1766313314437866,
"learning_rate": 0.00019887587397604813,
"loss": 0.3389,
"step": 94
},
{
"epoch": 0.06054811982154239,
"grad_norm": 1.1423076391220093,
"learning_rate": 0.0001988453459063325,
"loss": 0.3951,
"step": 95
},
{
"epoch": 0.06118546845124283,
"grad_norm": 1.3329870700836182,
"learning_rate": 0.00019881441125008692,
"loss": 0.5579,
"step": 96
},
{
"epoch": 0.061822817080943275,
"grad_norm": 1.0614429712295532,
"learning_rate": 0.00019878307013455678,
"loss": 0.3509,
"step": 97
},
{
"epoch": 0.06246016571064372,
"grad_norm": 0.8363850712776184,
"learning_rate": 0.00019875132268865934,
"loss": 0.2288,
"step": 98
},
{
"epoch": 0.06309751434034416,
"grad_norm": 0.9025006294250488,
"learning_rate": 0.00019871916904298337,
"loss": 0.313,
"step": 99
},
{
"epoch": 0.06373486297004462,
"grad_norm": 0.9582391977310181,
"learning_rate": 0.0001986866093297883,
"loss": 0.3139,
"step": 100
},
{
"epoch": 0.06437221159974506,
"grad_norm": 0.8045271039009094,
"learning_rate": 0.000198653643683004,
"loss": 0.2165,
"step": 101
},
{
"epoch": 0.06500956022944551,
"grad_norm": 1.4800572395324707,
"learning_rate": 0.0001986202722382301,
"loss": 0.4728,
"step": 102
},
{
"epoch": 0.06564690885914595,
"grad_norm": 1.1153333187103271,
"learning_rate": 0.00019858649513273532,
"loss": 0.3246,
"step": 103
},
{
"epoch": 0.0662842574888464,
"grad_norm": 0.6812078952789307,
"learning_rate": 0.00019855231250545712,
"loss": 0.0905,
"step": 104
},
{
"epoch": 0.06692160611854685,
"grad_norm": 1.549744963645935,
"learning_rate": 0.00019851772449700094,
"loss": 0.3597,
"step": 105
},
{
"epoch": 0.06755895474824729,
"grad_norm": 1.2806156873703003,
"learning_rate": 0.00019848273124963969,
"loss": 0.2911,
"step": 106
},
{
"epoch": 0.06819630337794774,
"grad_norm": 0.5269983410835266,
"learning_rate": 0.00019844733290731327,
"loss": 0.0589,
"step": 107
},
{
"epoch": 0.06883365200764818,
"grad_norm": 1.1872228384017944,
"learning_rate": 0.00019841152961562778,
"loss": 0.4659,
"step": 108
},
{
"epoch": 0.06947100063734862,
"grad_norm": 0.9378387331962585,
"learning_rate": 0.00019837532152185508,
"loss": 0.2319,
"step": 109
},
{
"epoch": 0.07010834926704908,
"grad_norm": 0.7823953628540039,
"learning_rate": 0.00019833870877493208,
"loss": 0.1788,
"step": 110
},
{
"epoch": 0.07074569789674952,
"grad_norm": 1.0174436569213867,
"learning_rate": 0.00019830169152546025,
"loss": 0.171,
"step": 111
},
{
"epoch": 0.07138304652644997,
"grad_norm": 1.21722412109375,
"learning_rate": 0.0001982642699257048,
"loss": 0.3396,
"step": 112
},
{
"epoch": 0.07202039515615041,
"grad_norm": 0.6042284965515137,
"learning_rate": 0.00019822644412959437,
"loss": 0.2543,
"step": 113
},
{
"epoch": 0.07265774378585087,
"grad_norm": 1.0360757112503052,
"learning_rate": 0.00019818821429271998,
"loss": 0.2087,
"step": 114
},
{
"epoch": 0.07329509241555131,
"grad_norm": 0.9517643451690674,
"learning_rate": 0.00019814958057233479,
"loss": 0.3145,
"step": 115
},
{
"epoch": 0.07393244104525175,
"grad_norm": 0.9021937251091003,
"learning_rate": 0.0001981105431273532,
"loss": 0.3069,
"step": 116
},
{
"epoch": 0.0745697896749522,
"grad_norm": 0.8300794959068298,
"learning_rate": 0.00019807110211835028,
"loss": 0.3572,
"step": 117
},
{
"epoch": 0.07520713830465264,
"grad_norm": 0.9179697036743164,
"learning_rate": 0.00019803125770756114,
"loss": 0.3215,
"step": 118
},
{
"epoch": 0.0758444869343531,
"grad_norm": 0.9362103939056396,
"learning_rate": 0.00019799101005888017,
"loss": 0.2391,
"step": 119
},
{
"epoch": 0.07648183556405354,
"grad_norm": 0.9023969173431396,
"learning_rate": 0.00019795035933786045,
"loss": 0.2932,
"step": 120
},
{
"epoch": 0.07711918419375398,
"grad_norm": 1.320633888244629,
"learning_rate": 0.0001979093057117131,
"loss": 0.4263,
"step": 121
},
{
"epoch": 0.07775653282345443,
"grad_norm": 0.7515263557434082,
"learning_rate": 0.0001978678493493064,
"loss": 0.2072,
"step": 122
},
{
"epoch": 0.07839388145315487,
"grad_norm": 1.3262919187545776,
"learning_rate": 0.00019782599042116535,
"loss": 0.3808,
"step": 123
},
{
"epoch": 0.07903123008285533,
"grad_norm": 0.6986315846443176,
"learning_rate": 0.00019778372909947086,
"loss": 0.1227,
"step": 124
},
{
"epoch": 0.07966857871255577,
"grad_norm": 1.0611870288848877,
"learning_rate": 0.00019774106555805885,
"loss": 0.322,
"step": 125
},
{
"epoch": 0.08030592734225621,
"grad_norm": 0.7359690070152283,
"learning_rate": 0.00019769799997241996,
"loss": 0.1202,
"step": 126
},
{
"epoch": 0.08094327597195666,
"grad_norm": 0.8817009925842285,
"learning_rate": 0.0001976545325196984,
"loss": 0.3929,
"step": 127
},
{
"epoch": 0.0815806246016571,
"grad_norm": 0.9690378308296204,
"learning_rate": 0.0001976106633786915,
"loss": 0.2834,
"step": 128
},
{
"epoch": 0.08221797323135756,
"grad_norm": 1.5191131830215454,
"learning_rate": 0.00019756639272984882,
"loss": 0.631,
"step": 129
},
{
"epoch": 0.082855321861058,
"grad_norm": 0.8324887156486511,
"learning_rate": 0.0001975217207552715,
"loss": 0.3293,
"step": 130
},
{
"epoch": 0.08349267049075844,
"grad_norm": 1.0263890027999878,
"learning_rate": 0.0001974766476387115,
"loss": 0.3834,
"step": 131
},
{
"epoch": 0.0841300191204589,
"grad_norm": 1.1872267723083496,
"learning_rate": 0.00019743117356557074,
"loss": 0.4707,
"step": 132
},
{
"epoch": 0.08476736775015933,
"grad_norm": 1.1256309747695923,
"learning_rate": 0.00019738529872290054,
"loss": 0.3615,
"step": 133
},
{
"epoch": 0.08540471637985979,
"grad_norm": 1.1773260831832886,
"learning_rate": 0.00019733902329940057,
"loss": 0.306,
"step": 134
},
{
"epoch": 0.08604206500956023,
"grad_norm": 0.7083126306533813,
"learning_rate": 0.0001972923474854184,
"loss": 0.2916,
"step": 135
},
{
"epoch": 0.08667941363926067,
"grad_norm": 0.8032341003417969,
"learning_rate": 0.00019724527147294845,
"loss": 0.2193,
"step": 136
},
{
"epoch": 0.08731676226896112,
"grad_norm": 0.5595875382423401,
"learning_rate": 0.0001971977954556313,
"loss": 0.0722,
"step": 137
},
{
"epoch": 0.08795411089866156,
"grad_norm": 1.139371633529663,
"learning_rate": 0.00019714991962875296,
"loss": 0.3586,
"step": 138
},
{
"epoch": 0.08859145952836202,
"grad_norm": 1.0842418670654297,
"learning_rate": 0.00019710164418924395,
"loss": 0.3676,
"step": 139
},
{
"epoch": 0.08922880815806246,
"grad_norm": 1.0619008541107178,
"learning_rate": 0.00019705296933567853,
"loss": 0.2884,
"step": 140
},
{
"epoch": 0.08986615678776291,
"grad_norm": 0.9296439290046692,
"learning_rate": 0.00019700389526827391,
"loss": 0.3713,
"step": 141
},
{
"epoch": 0.09050350541746335,
"grad_norm": 0.902638852596283,
"learning_rate": 0.00019695442218888945,
"loss": 0.2775,
"step": 142
},
{
"epoch": 0.0911408540471638,
"grad_norm": 0.685797393321991,
"learning_rate": 0.00019690455030102572,
"loss": 0.0716,
"step": 143
},
{
"epoch": 0.09177820267686425,
"grad_norm": 1.3578416109085083,
"learning_rate": 0.0001968542798098238,
"loss": 0.5199,
"step": 144
},
{
"epoch": 0.09241555130656469,
"grad_norm": 1.3487476110458374,
"learning_rate": 0.0001968036109220643,
"loss": 0.1956,
"step": 145
},
{
"epoch": 0.09305289993626514,
"grad_norm": 0.9793888926506042,
"learning_rate": 0.00019675254384616663,
"loss": 0.3956,
"step": 146
},
{
"epoch": 0.09369024856596558,
"grad_norm": 1.2398440837860107,
"learning_rate": 0.00019670107879218813,
"loss": 0.3269,
"step": 147
},
{
"epoch": 0.09432759719566602,
"grad_norm": 0.6424081325531006,
"learning_rate": 0.00019664921597182302,
"loss": 0.0824,
"step": 148
},
{
"epoch": 0.09496494582536648,
"grad_norm": 1.1197257041931152,
"learning_rate": 0.00019659695559840184,
"loss": 0.3915,
"step": 149
},
{
"epoch": 0.09560229445506692,
"grad_norm": 0.9322795271873474,
"learning_rate": 0.00019654429788689028,
"loss": 0.3921,
"step": 150
},
{
"epoch": 0.09623964308476737,
"grad_norm": 0.5158027410507202,
"learning_rate": 0.00019649124305388854,
"loss": 0.0981,
"step": 151
},
{
"epoch": 0.09687699171446781,
"grad_norm": 0.9111659526824951,
"learning_rate": 0.00019643779131763028,
"loss": 0.3237,
"step": 152
},
{
"epoch": 0.09751434034416825,
"grad_norm": 1.2434680461883545,
"learning_rate": 0.0001963839428979817,
"loss": 0.2862,
"step": 153
},
{
"epoch": 0.09815168897386871,
"grad_norm": 1.0688103437423706,
"learning_rate": 0.00019632969801644076,
"loss": 0.4604,
"step": 154
},
{
"epoch": 0.09878903760356915,
"grad_norm": 0.9399598836898804,
"learning_rate": 0.00019627505689613625,
"loss": 0.4419,
"step": 155
},
{
"epoch": 0.0994263862332696,
"grad_norm": 0.7369008660316467,
"learning_rate": 0.00019622001976182677,
"loss": 0.3467,
"step": 156
},
{
"epoch": 0.10006373486297004,
"grad_norm": 0.5455231070518494,
"learning_rate": 0.0001961645868398999,
"loss": 0.1122,
"step": 157
},
{
"epoch": 0.10070108349267048,
"grad_norm": 0.5688629746437073,
"learning_rate": 0.00019610875835837118,
"loss": 0.1,
"step": 158
},
{
"epoch": 0.10133843212237094,
"grad_norm": 1.2653007507324219,
"learning_rate": 0.0001960525345468833,
"loss": 0.4819,
"step": 159
},
{
"epoch": 0.10197578075207138,
"grad_norm": 1.2717375755310059,
"learning_rate": 0.00019599591563670504,
"loss": 0.2624,
"step": 160
},
{
"epoch": 0.10261312938177183,
"grad_norm": 1.0098074674606323,
"learning_rate": 0.00019593890186073046,
"loss": 0.2277,
"step": 161
},
{
"epoch": 0.10325047801147227,
"grad_norm": 0.9294418096542358,
"learning_rate": 0.00019588149345347767,
"loss": 0.245,
"step": 162
},
{
"epoch": 0.10388782664117271,
"grad_norm": 1.0674294233322144,
"learning_rate": 0.00019582369065108818,
"loss": 0.6072,
"step": 163
},
{
"epoch": 0.10452517527087317,
"grad_norm": 0.8809716105461121,
"learning_rate": 0.00019576549369132578,
"loss": 0.1988,
"step": 164
},
{
"epoch": 0.10516252390057361,
"grad_norm": 1.060187816619873,
"learning_rate": 0.0001957069028135755,
"loss": 0.3857,
"step": 165
},
{
"epoch": 0.10579987253027406,
"grad_norm": 0.985283613204956,
"learning_rate": 0.0001956479182588427,
"loss": 0.4099,
"step": 166
},
{
"epoch": 0.1064372211599745,
"grad_norm": 0.898350179195404,
"learning_rate": 0.0001955885402697522,
"loss": 0.3636,
"step": 167
},
{
"epoch": 0.10707456978967496,
"grad_norm": 0.9302679300308228,
"learning_rate": 0.00019552876909054701,
"loss": 0.3034,
"step": 168
},
{
"epoch": 0.1077119184193754,
"grad_norm": 0.7727479934692383,
"learning_rate": 0.0001954686049670875,
"loss": 0.3398,
"step": 169
},
{
"epoch": 0.10834926704907584,
"grad_norm": 1.0249948501586914,
"learning_rate": 0.00019540804814685046,
"loss": 0.3645,
"step": 170
},
{
"epoch": 0.1089866156787763,
"grad_norm": 0.8279457092285156,
"learning_rate": 0.0001953470988789279,
"loss": 0.339,
"step": 171
},
{
"epoch": 0.10962396430847673,
"grad_norm": 0.7484726309776306,
"learning_rate": 0.0001952857574140261,
"loss": 0.1843,
"step": 172
},
{
"epoch": 0.11026131293817719,
"grad_norm": 0.8363127708435059,
"learning_rate": 0.00019522402400446462,
"loss": 0.2382,
"step": 173
},
{
"epoch": 0.11089866156787763,
"grad_norm": 0.9513233304023743,
"learning_rate": 0.00019516189890417525,
"loss": 0.5213,
"step": 174
},
{
"epoch": 0.11153601019757807,
"grad_norm": 1.2631983757019043,
"learning_rate": 0.00019509938236870084,
"loss": 0.472,
"step": 175
},
{
"epoch": 0.11217335882727852,
"grad_norm": 0.7087992429733276,
"learning_rate": 0.0001950364746551945,
"loss": 0.2074,
"step": 176
},
{
"epoch": 0.11281070745697896,
"grad_norm": 1.0447160005569458,
"learning_rate": 0.00019497317602241833,
"loss": 0.2685,
"step": 177
},
{
"epoch": 0.11344805608667942,
"grad_norm": 1.447125792503357,
"learning_rate": 0.00019490948673074238,
"loss": 0.4409,
"step": 178
},
{
"epoch": 0.11408540471637986,
"grad_norm": 1.2068469524383545,
"learning_rate": 0.0001948454070421437,
"loss": 0.5064,
"step": 179
},
{
"epoch": 0.1147227533460803,
"grad_norm": 1.3105393648147583,
"learning_rate": 0.0001947809372202051,
"loss": 0.4046,
"step": 180
},
{
"epoch": 0.11536010197578075,
"grad_norm": 0.7638639807701111,
"learning_rate": 0.00019471607753011422,
"loss": 0.2528,
"step": 181
},
{
"epoch": 0.1159974506054812,
"grad_norm": 0.2580053508281708,
"learning_rate": 0.0001946508282386623,
"loss": 0.029,
"step": 182
},
{
"epoch": 0.11663479923518165,
"grad_norm": 1.3025866746902466,
"learning_rate": 0.00019458518961424325,
"loss": 0.6058,
"step": 183
},
{
"epoch": 0.11727214786488209,
"grad_norm": 1.274926781654358,
"learning_rate": 0.0001945191619268523,
"loss": 0.5549,
"step": 184
},
{
"epoch": 0.11790949649458253,
"grad_norm": 0.7398247718811035,
"learning_rate": 0.00019445274544808525,
"loss": 0.1499,
"step": 185
},
{
"epoch": 0.11854684512428298,
"grad_norm": 0.993326723575592,
"learning_rate": 0.0001943859404511369,
"loss": 0.3668,
"step": 186
},
{
"epoch": 0.11918419375398343,
"grad_norm": 1.080265998840332,
"learning_rate": 0.00019431874721080033,
"loss": 0.385,
"step": 187
},
{
"epoch": 0.11982154238368388,
"grad_norm": 1.0789588689804077,
"learning_rate": 0.00019425116600346558,
"loss": 0.6067,
"step": 188
},
{
"epoch": 0.12045889101338432,
"grad_norm": 0.7704159021377563,
"learning_rate": 0.00019418319710711845,
"loss": 0.2714,
"step": 189
},
{
"epoch": 0.12109623964308477,
"grad_norm": 0.7008885145187378,
"learning_rate": 0.00019411484080133953,
"loss": 0.1862,
"step": 190
},
{
"epoch": 0.12173358827278521,
"grad_norm": 0.6024148464202881,
"learning_rate": 0.00019404609736730293,
"loss": 0.1756,
"step": 191
},
{
"epoch": 0.12237093690248566,
"grad_norm": 0.23649345338344574,
"learning_rate": 0.00019397696708777518,
"loss": 0.0546,
"step": 192
},
{
"epoch": 0.12300828553218611,
"grad_norm": 0.7000377178192139,
"learning_rate": 0.00019390745024711406,
"loss": 0.2412,
"step": 193
},
{
"epoch": 0.12364563416188655,
"grad_norm": 1.114310622215271,
"learning_rate": 0.00019383754713126733,
"loss": 0.1716,
"step": 194
},
{
"epoch": 0.124282982791587,
"grad_norm": 1.0355890989303589,
"learning_rate": 0.00019376725802777172,
"loss": 0.3847,
"step": 195
},
{
"epoch": 0.12492033142128744,
"grad_norm": 0.938788890838623,
"learning_rate": 0.00019369658322575167,
"loss": 0.1959,
"step": 196
},
{
"epoch": 0.12555768005098789,
"grad_norm": 0.8382909297943115,
"learning_rate": 0.00019362552301591807,
"loss": 0.3532,
"step": 197
},
{
"epoch": 0.12619502868068833,
"grad_norm": 0.7558854222297668,
"learning_rate": 0.00019355407769056725,
"loss": 0.156,
"step": 198
},
{
"epoch": 0.1268323773103888,
"grad_norm": 1.151506781578064,
"learning_rate": 0.0001934822475435795,
"loss": 0.4139,
"step": 199
},
{
"epoch": 0.12746972594008923,
"grad_norm": 1.3888171911239624,
"learning_rate": 0.0001934100328704181,
"loss": 0.4695,
"step": 200
},
{
"epoch": 0.12810707456978968,
"grad_norm": 0.9342585206031799,
"learning_rate": 0.00019333743396812806,
"loss": 0.3449,
"step": 201
},
{
"epoch": 0.12874442319949012,
"grad_norm": 1.0079127550125122,
"learning_rate": 0.0001932644511353348,
"loss": 0.4085,
"step": 202
},
{
"epoch": 0.12938177182919056,
"grad_norm": 1.2367472648620605,
"learning_rate": 0.00019319108467224307,
"loss": 0.4151,
"step": 203
},
{
"epoch": 0.13001912045889102,
"grad_norm": 1.1219055652618408,
"learning_rate": 0.00019311733488063554,
"loss": 0.4242,
"step": 204
},
{
"epoch": 0.13065646908859146,
"grad_norm": 0.7425576448440552,
"learning_rate": 0.00019304320206387166,
"loss": 0.2183,
"step": 205
},
{
"epoch": 0.1312938177182919,
"grad_norm": 1.2308136224746704,
"learning_rate": 0.00019296868652688643,
"loss": 0.5093,
"step": 206
},
{
"epoch": 0.13193116634799235,
"grad_norm": 0.9649097919464111,
"learning_rate": 0.00019289378857618913,
"loss": 0.2678,
"step": 207
},
{
"epoch": 0.1325685149776928,
"grad_norm": 0.726058304309845,
"learning_rate": 0.00019281850851986203,
"loss": 0.22,
"step": 208
},
{
"epoch": 0.13320586360739325,
"grad_norm": 0.5078527331352234,
"learning_rate": 0.00019274284666755907,
"loss": 0.1266,
"step": 209
},
{
"epoch": 0.1338432122370937,
"grad_norm": 0.8264751434326172,
"learning_rate": 0.00019266680333050477,
"loss": 0.2768,
"step": 210
},
{
"epoch": 0.13448056086679414,
"grad_norm": 0.3217345178127289,
"learning_rate": 0.00019259037882149278,
"loss": 0.0361,
"step": 211
},
{
"epoch": 0.13511790949649458,
"grad_norm": 0.8530462384223938,
"learning_rate": 0.00019251357345488463,
"loss": 0.2718,
"step": 212
},
{
"epoch": 0.13575525812619502,
"grad_norm": 0.9174484014511108,
"learning_rate": 0.00019243638754660848,
"loss": 0.2713,
"step": 213
},
{
"epoch": 0.13639260675589548,
"grad_norm": 0.6848881840705872,
"learning_rate": 0.00019235882141415778,
"loss": 0.141,
"step": 214
},
{
"epoch": 0.13702995538559593,
"grad_norm": 0.9194259643554688,
"learning_rate": 0.00019228087537658998,
"loss": 0.3329,
"step": 215
},
{
"epoch": 0.13766730401529637,
"grad_norm": 0.26792770624160767,
"learning_rate": 0.00019220254975452526,
"loss": 0.0345,
"step": 216
},
{
"epoch": 0.1383046526449968,
"grad_norm": 1.121355414390564,
"learning_rate": 0.00019212384487014508,
"loss": 0.3415,
"step": 217
},
{
"epoch": 0.13894200127469725,
"grad_norm": 1.1536093950271606,
"learning_rate": 0.00019204476104719103,
"loss": 0.5155,
"step": 218
},
{
"epoch": 0.13957934990439771,
"grad_norm": 0.9080384373664856,
"learning_rate": 0.00019196529861096337,
"loss": 0.267,
"step": 219
},
{
"epoch": 0.14021669853409816,
"grad_norm": 0.808695375919342,
"learning_rate": 0.00019188545788831972,
"loss": 0.3074,
"step": 220
},
{
"epoch": 0.1408540471637986,
"grad_norm": 0.9998703002929688,
"learning_rate": 0.0001918052392076737,
"loss": 0.2856,
"step": 221
},
{
"epoch": 0.14149139579349904,
"grad_norm": 0.24331898987293243,
"learning_rate": 0.0001917246428989938,
"loss": 0.0151,
"step": 222
},
{
"epoch": 0.14212874442319948,
"grad_norm": 0.9334642291069031,
"learning_rate": 0.00019164366929380155,
"loss": 0.1379,
"step": 223
},
{
"epoch": 0.14276609305289995,
"grad_norm": 1.1146843433380127,
"learning_rate": 0.00019156231872517063,
"loss": 0.3853,
"step": 224
},
{
"epoch": 0.14340344168260039,
"grad_norm": 1.1395857334136963,
"learning_rate": 0.00019148059152772522,
"loss": 0.3466,
"step": 225
},
{
"epoch": 0.14404079031230083,
"grad_norm": 1.0670769214630127,
"learning_rate": 0.00019139848803763872,
"loss": 0.2507,
"step": 226
},
{
"epoch": 0.14467813894200127,
"grad_norm": 1.2081040143966675,
"learning_rate": 0.00019131600859263238,
"loss": 0.3428,
"step": 227
},
{
"epoch": 0.14531548757170173,
"grad_norm": 1.2847809791564941,
"learning_rate": 0.00019123315353197396,
"loss": 0.4016,
"step": 228
},
{
"epoch": 0.14595283620140218,
"grad_norm": 0.7287536263465881,
"learning_rate": 0.00019114992319647608,
"loss": 0.1856,
"step": 229
},
{
"epoch": 0.14659018483110262,
"grad_norm": 1.0922931432724,
"learning_rate": 0.00019106631792849515,
"loss": 0.3057,
"step": 230
},
{
"epoch": 0.14722753346080306,
"grad_norm": 0.6916879415512085,
"learning_rate": 0.00019098233807192977,
"loss": 0.2311,
"step": 231
},
{
"epoch": 0.1478648820905035,
"grad_norm": 0.8120045065879822,
"learning_rate": 0.00019089798397221936,
"loss": 0.3721,
"step": 232
},
{
"epoch": 0.14850223072020396,
"grad_norm": 1.0548818111419678,
"learning_rate": 0.00019081325597634273,
"loss": 0.2549,
"step": 233
},
{
"epoch": 0.1491395793499044,
"grad_norm": 0.8204748630523682,
"learning_rate": 0.00019072815443281664,
"loss": 0.1589,
"step": 234
},
{
"epoch": 0.14977692797960485,
"grad_norm": 0.6061741709709167,
"learning_rate": 0.00019064267969169443,
"loss": 0.1243,
"step": 235
},
{
"epoch": 0.1504142766093053,
"grad_norm": 0.9242045283317566,
"learning_rate": 0.0001905568321045645,
"loss": 0.4587,
"step": 236
},
{
"epoch": 0.15105162523900573,
"grad_norm": 0.8071498870849609,
"learning_rate": 0.0001904706120245489,
"loss": 0.2391,
"step": 237
},
{
"epoch": 0.1516889738687062,
"grad_norm": 0.9680665135383606,
"learning_rate": 0.00019038401980630193,
"loss": 0.2078,
"step": 238
},
{
"epoch": 0.15232632249840664,
"grad_norm": 0.9554654955863953,
"learning_rate": 0.00019029705580600855,
"loss": 0.5775,
"step": 239
},
{
"epoch": 0.15296367112810708,
"grad_norm": 0.5851937532424927,
"learning_rate": 0.00019020972038138307,
"loss": 0.1093,
"step": 240
},
{
"epoch": 0.15360101975780752,
"grad_norm": 0.6797639727592468,
"learning_rate": 0.00019012201389166755,
"loss": 0.2853,
"step": 241
},
{
"epoch": 0.15423836838750796,
"grad_norm": 1.1129157543182373,
"learning_rate": 0.00019003393669763036,
"loss": 0.3796,
"step": 242
},
{
"epoch": 0.15487571701720843,
"grad_norm": 1.0510376691818237,
"learning_rate": 0.0001899454891615648,
"loss": 0.2435,
"step": 243
},
{
"epoch": 0.15551306564690887,
"grad_norm": 0.3660753071308136,
"learning_rate": 0.00018985667164728742,
"loss": 0.0622,
"step": 244
},
{
"epoch": 0.1561504142766093,
"grad_norm": 0.7310383319854736,
"learning_rate": 0.00018976748452013671,
"loss": 0.3131,
"step": 245
},
{
"epoch": 0.15678776290630975,
"grad_norm": 0.6510698795318604,
"learning_rate": 0.00018967792814697138,
"loss": 0.3482,
"step": 246
},
{
"epoch": 0.1574251115360102,
"grad_norm": 0.5597620606422424,
"learning_rate": 0.00018958800289616915,
"loss": 0.1017,
"step": 247
},
{
"epoch": 0.15806246016571066,
"grad_norm": 0.6275910139083862,
"learning_rate": 0.00018949770913762496,
"loss": 0.237,
"step": 248
},
{
"epoch": 0.1586998087954111,
"grad_norm": 0.6911898255348206,
"learning_rate": 0.00018940704724274956,
"loss": 0.2781,
"step": 249
},
{
"epoch": 0.15933715742511154,
"grad_norm": 0.63053959608078,
"learning_rate": 0.000189316017584468,
"loss": 0.1995,
"step": 250
},
{
"epoch": 0.15997450605481198,
"grad_norm": 0.7132065296173096,
"learning_rate": 0.0001892246205372181,
"loss": 0.2891,
"step": 251
},
{
"epoch": 0.16061185468451242,
"grad_norm": 0.7136152982711792,
"learning_rate": 0.00018913285647694877,
"loss": 0.1669,
"step": 252
},
{
"epoch": 0.16124920331421289,
"grad_norm": 0.7467982172966003,
"learning_rate": 0.0001890407257811188,
"loss": 0.1904,
"step": 253
},
{
"epoch": 0.16188655194391333,
"grad_norm": 1.2307111024856567,
"learning_rate": 0.00018894822882869482,
"loss": 0.6067,
"step": 254
},
{
"epoch": 0.16252390057361377,
"grad_norm": 0.7953675389289856,
"learning_rate": 0.00018885536600015022,
"loss": 0.2748,
"step": 255
},
{
"epoch": 0.1631612492033142,
"grad_norm": 0.8050816059112549,
"learning_rate": 0.0001887621376774632,
"loss": 0.1564,
"step": 256
},
{
"epoch": 0.16379859783301465,
"grad_norm": 0.9376430511474609,
"learning_rate": 0.00018866854424411548,
"loss": 0.3872,
"step": 257
},
{
"epoch": 0.16443594646271512,
"grad_norm": 0.8895918726921082,
"learning_rate": 0.00018857458608509062,
"loss": 0.3349,
"step": 258
},
{
"epoch": 0.16507329509241556,
"grad_norm": 1.0301870107650757,
"learning_rate": 0.00018848026358687234,
"loss": 0.3629,
"step": 259
},
{
"epoch": 0.165710643722116,
"grad_norm": 0.629410445690155,
"learning_rate": 0.0001883855771374431,
"loss": 0.1107,
"step": 260
},
{
"epoch": 0.16634799235181644,
"grad_norm": 0.8364279270172119,
"learning_rate": 0.00018829052712628235,
"loss": 0.2686,
"step": 261
},
{
"epoch": 0.16698534098151688,
"grad_norm": 1.1152483224868774,
"learning_rate": 0.00018819511394436506,
"loss": 0.3722,
"step": 262
},
{
"epoch": 0.16762268961121735,
"grad_norm": 0.8939811587333679,
"learning_rate": 0.00018809933798416006,
"loss": 0.26,
"step": 263
},
{
"epoch": 0.1682600382409178,
"grad_norm": 1.091820240020752,
"learning_rate": 0.00018800319963962835,
"loss": 0.3782,
"step": 264
},
{
"epoch": 0.16889738687061823,
"grad_norm": 0.9514994025230408,
"learning_rate": 0.0001879066993062216,
"loss": 0.1535,
"step": 265
},
{
"epoch": 0.16953473550031867,
"grad_norm": 0.9090454578399658,
"learning_rate": 0.00018780983738088047,
"loss": 0.0907,
"step": 266
},
{
"epoch": 0.1701720841300191,
"grad_norm": 1.1316412687301636,
"learning_rate": 0.00018771261426203298,
"loss": 0.453,
"step": 267
},
{
"epoch": 0.17080943275971958,
"grad_norm": 0.9637196063995361,
"learning_rate": 0.0001876150303495928,
"loss": 0.2135,
"step": 268
},
{
"epoch": 0.17144678138942002,
"grad_norm": 1.0072875022888184,
"learning_rate": 0.00018751708604495778,
"loss": 0.1907,
"step": 269
},
{
"epoch": 0.17208413001912046,
"grad_norm": 0.6125558018684387,
"learning_rate": 0.0001874187817510081,
"loss": 0.1078,
"step": 270
},
{
"epoch": 0.1727214786488209,
"grad_norm": 1.3461196422576904,
"learning_rate": 0.0001873201178721048,
"loss": 0.5376,
"step": 271
},
{
"epoch": 0.17335882727852134,
"grad_norm": 1.156919002532959,
"learning_rate": 0.00018722109481408795,
"loss": 0.5093,
"step": 272
},
{
"epoch": 0.1739961759082218,
"grad_norm": 1.216074824333191,
"learning_rate": 0.0001871217129842751,
"loss": 0.3761,
"step": 273
},
{
"epoch": 0.17463352453792225,
"grad_norm": 1.5678287744522095,
"learning_rate": 0.00018702197279145944,
"loss": 0.4145,
"step": 274
},
{
"epoch": 0.1752708731676227,
"grad_norm": 1.0253586769104004,
"learning_rate": 0.00018692187464590846,
"loss": 0.3711,
"step": 275
},
{
"epoch": 0.17590822179732313,
"grad_norm": 1.0861526727676392,
"learning_rate": 0.00018682141895936177,
"loss": 0.2918,
"step": 276
},
{
"epoch": 0.17654557042702357,
"grad_norm": 1.0223357677459717,
"learning_rate": 0.0001867206061450299,
"loss": 0.2159,
"step": 277
},
{
"epoch": 0.17718291905672404,
"grad_norm": 0.9765048027038574,
"learning_rate": 0.00018661943661759227,
"loss": 0.5828,
"step": 278
},
{
"epoch": 0.17782026768642448,
"grad_norm": 0.8800041079521179,
"learning_rate": 0.0001865179107931956,
"loss": 0.2818,
"step": 279
},
{
"epoch": 0.17845761631612492,
"grad_norm": 0.9124730825424194,
"learning_rate": 0.00018641602908945216,
"loss": 0.2472,
"step": 280
},
{
"epoch": 0.17909496494582536,
"grad_norm": 0.961738109588623,
"learning_rate": 0.00018631379192543817,
"loss": 0.5051,
"step": 281
},
{
"epoch": 0.17973231357552583,
"grad_norm": 0.9458169341087341,
"learning_rate": 0.00018621119972169189,
"loss": 0.3851,
"step": 282
},
{
"epoch": 0.18036966220522627,
"grad_norm": 1.1568273305892944,
"learning_rate": 0.0001861082529002121,
"loss": 0.4793,
"step": 283
},
{
"epoch": 0.1810070108349267,
"grad_norm": 0.49251148104667664,
"learning_rate": 0.00018600495188445608,
"loss": 0.0806,
"step": 284
},
{
"epoch": 0.18164435946462715,
"grad_norm": 0.5726460814476013,
"learning_rate": 0.0001859012970993382,
"loss": 0.1974,
"step": 285
},
{
"epoch": 0.1822817080943276,
"grad_norm": 0.4672144055366516,
"learning_rate": 0.00018579728897122788,
"loss": 0.0525,
"step": 286
},
{
"epoch": 0.18291905672402806,
"grad_norm": 0.4259563088417053,
"learning_rate": 0.00018569292792794813,
"loss": 0.0774,
"step": 287
},
{
"epoch": 0.1835564053537285,
"grad_norm": 1.1044259071350098,
"learning_rate": 0.00018558821439877346,
"loss": 0.4283,
"step": 288
},
{
"epoch": 0.18419375398342894,
"grad_norm": 0.8057593107223511,
"learning_rate": 0.00018548314881442838,
"loss": 0.2405,
"step": 289
},
{
"epoch": 0.18483110261312938,
"grad_norm": 0.6809104681015015,
"learning_rate": 0.00018537773160708546,
"loss": 0.1399,
"step": 290
},
{
"epoch": 0.18546845124282982,
"grad_norm": 0.45731616020202637,
"learning_rate": 0.00018527196321036372,
"loss": 0.0962,
"step": 291
},
{
"epoch": 0.1861057998725303,
"grad_norm": 0.5318043828010559,
"learning_rate": 0.00018516584405932666,
"loss": 0.1472,
"step": 292
},
{
"epoch": 0.18674314850223073,
"grad_norm": 1.0502890348434448,
"learning_rate": 0.00018505937459048063,
"loss": 0.4424,
"step": 293
},
{
"epoch": 0.18738049713193117,
"grad_norm": 0.4580780267715454,
"learning_rate": 0.00018495255524177286,
"loss": 0.067,
"step": 294
},
{
"epoch": 0.1880178457616316,
"grad_norm": 1.0052136182785034,
"learning_rate": 0.0001848453864525899,
"loss": 0.3343,
"step": 295
},
{
"epoch": 0.18865519439133205,
"grad_norm": 1.4164037704467773,
"learning_rate": 0.00018473786866375554,
"loss": 0.2421,
"step": 296
},
{
"epoch": 0.18929254302103252,
"grad_norm": 1.17645263671875,
"learning_rate": 0.00018463000231752922,
"loss": 0.4515,
"step": 297
},
{
"epoch": 0.18992989165073296,
"grad_norm": 0.6446077227592468,
"learning_rate": 0.00018452178785760413,
"loss": 0.1835,
"step": 298
},
{
"epoch": 0.1905672402804334,
"grad_norm": 0.6400480270385742,
"learning_rate": 0.00018441322572910527,
"loss": 0.1489,
"step": 299
},
{
"epoch": 0.19120458891013384,
"grad_norm": 0.8352667689323425,
"learning_rate": 0.00018430431637858784,
"loss": 0.2003,
"step": 300
},
{
"epoch": 0.19184193753983428,
"grad_norm": 0.9608398079872131,
"learning_rate": 0.00018419506025403528,
"loss": 0.2198,
"step": 301
},
{
"epoch": 0.19247928616953475,
"grad_norm": 0.9239013195037842,
"learning_rate": 0.00018408545780485732,
"loss": 0.3025,
"step": 302
},
{
"epoch": 0.1931166347992352,
"grad_norm": 0.9933524131774902,
"learning_rate": 0.0001839755094818884,
"loss": 0.4879,
"step": 303
},
{
"epoch": 0.19375398342893563,
"grad_norm": 1.0836195945739746,
"learning_rate": 0.00018386521573738563,
"loss": 0.2516,
"step": 304
},
{
"epoch": 0.19439133205863607,
"grad_norm": 0.8679178953170776,
"learning_rate": 0.0001837545770250268,
"loss": 0.1686,
"step": 305
},
{
"epoch": 0.1950286806883365,
"grad_norm": 1.0821306705474854,
"learning_rate": 0.00018364359379990895,
"loss": 0.3905,
"step": 306
},
{
"epoch": 0.19566602931803698,
"grad_norm": 1.0355212688446045,
"learning_rate": 0.00018353226651854595,
"loss": 0.3138,
"step": 307
},
{
"epoch": 0.19630337794773742,
"grad_norm": 1.0600056648254395,
"learning_rate": 0.00018342059563886714,
"loss": 0.2368,
"step": 308
},
{
"epoch": 0.19694072657743786,
"grad_norm": 0.8498077392578125,
"learning_rate": 0.000183308581620215,
"loss": 0.4302,
"step": 309
},
{
"epoch": 0.1975780752071383,
"grad_norm": 1.2844269275665283,
"learning_rate": 0.00018319622492334362,
"loss": 0.2915,
"step": 310
},
{
"epoch": 0.19821542383683874,
"grad_norm": 0.9590743184089661,
"learning_rate": 0.00018308352601041657,
"loss": 0.3272,
"step": 311
},
{
"epoch": 0.1988527724665392,
"grad_norm": 0.6332211494445801,
"learning_rate": 0.00018297048534500512,
"loss": 0.11,
"step": 312
},
{
"epoch": 0.19949012109623965,
"grad_norm": 0.7648255825042725,
"learning_rate": 0.00018285710339208618,
"loss": 0.2498,
"step": 313
},
{
"epoch": 0.2001274697259401,
"grad_norm": 0.9084885120391846,
"learning_rate": 0.00018274338061804072,
"loss": 0.3317,
"step": 314
},
{
"epoch": 0.20076481835564053,
"grad_norm": 1.0678030252456665,
"learning_rate": 0.00018262931749065135,
"loss": 0.3189,
"step": 315
},
{
"epoch": 0.20140216698534097,
"grad_norm": 1.0375890731811523,
"learning_rate": 0.0001825149144791009,
"loss": 0.3358,
"step": 316
},
{
"epoch": 0.20203951561504144,
"grad_norm": 0.9634979367256165,
"learning_rate": 0.00018240017205397015,
"loss": 0.3936,
"step": 317
},
{
"epoch": 0.20267686424474188,
"grad_norm": 1.104101538658142,
"learning_rate": 0.00018228509068723606,
"loss": 0.307,
"step": 318
},
{
"epoch": 0.20331421287444232,
"grad_norm": 1.8222627639770508,
"learning_rate": 0.00018216967085226982,
"loss": 0.4134,
"step": 319
},
{
"epoch": 0.20395156150414276,
"grad_norm": 1.027024745941162,
"learning_rate": 0.00018205391302383473,
"loss": 0.2113,
"step": 320
},
{
"epoch": 0.2045889101338432,
"grad_norm": 1.1963961124420166,
"learning_rate": 0.00018193781767808447,
"loss": 0.334,
"step": 321
},
{
"epoch": 0.20522625876354367,
"grad_norm": 0.98575758934021,
"learning_rate": 0.00018182138529256105,
"loss": 0.2368,
"step": 322
},
{
"epoch": 0.2058636073932441,
"grad_norm": 0.5618329048156738,
"learning_rate": 0.00018170461634619284,
"loss": 0.081,
"step": 323
},
{
"epoch": 0.20650095602294455,
"grad_norm": 0.7279055118560791,
"learning_rate": 0.00018158751131929258,
"loss": 0.1631,
"step": 324
},
{
"epoch": 0.207138304652645,
"grad_norm": 0.7400451898574829,
"learning_rate": 0.00018147007069355543,
"loss": 0.2246,
"step": 325
},
{
"epoch": 0.20777565328234543,
"grad_norm": 0.7073190212249756,
"learning_rate": 0.00018135229495205705,
"loss": 0.2475,
"step": 326
},
{
"epoch": 0.2084130019120459,
"grad_norm": 1.0298449993133545,
"learning_rate": 0.00018123418457925152,
"loss": 0.4364,
"step": 327
},
{
"epoch": 0.20905035054174634,
"grad_norm": 1.2232853174209595,
"learning_rate": 0.00018111574006096932,
"loss": 0.5125,
"step": 328
},
{
"epoch": 0.20968769917144678,
"grad_norm": 0.9231348633766174,
"learning_rate": 0.00018099696188441546,
"loss": 0.3662,
"step": 329
},
{
"epoch": 0.21032504780114722,
"grad_norm": 0.8744479417800903,
"learning_rate": 0.00018087785053816737,
"loss": 0.3611,
"step": 330
},
{
"epoch": 0.2109623964308477,
"grad_norm": 0.9583303332328796,
"learning_rate": 0.00018075840651217298,
"loss": 0.3413,
"step": 331
},
{
"epoch": 0.21159974506054813,
"grad_norm": 1.2141737937927246,
"learning_rate": 0.0001806386302977486,
"loss": 0.1618,
"step": 332
},
{
"epoch": 0.21223709369024857,
"grad_norm": 0.5310484766960144,
"learning_rate": 0.00018051852238757694,
"loss": 0.1623,
"step": 333
},
{
"epoch": 0.212874442319949,
"grad_norm": 0.6970574855804443,
"learning_rate": 0.00018039808327570514,
"loss": 0.2001,
"step": 334
},
{
"epoch": 0.21351179094964945,
"grad_norm": 1.144284725189209,
"learning_rate": 0.00018027731345754267,
"loss": 0.5024,
"step": 335
},
{
"epoch": 0.21414913957934992,
"grad_norm": 1.0011065006256104,
"learning_rate": 0.00018015621342985932,
"loss": 0.2895,
"step": 336
},
{
"epoch": 0.21478648820905036,
"grad_norm": 1.0536068677902222,
"learning_rate": 0.00018003478369078312,
"loss": 0.296,
"step": 337
},
{
"epoch": 0.2154238368387508,
"grad_norm": 1.138771414756775,
"learning_rate": 0.0001799130247397984,
"loss": 0.3287,
"step": 338
},
{
"epoch": 0.21606118546845124,
"grad_norm": 0.8639943599700928,
"learning_rate": 0.00017979093707774354,
"loss": 0.278,
"step": 339
},
{
"epoch": 0.21669853409815168,
"grad_norm": 1.0771543979644775,
"learning_rate": 0.00017966852120680917,
"loss": 0.3549,
"step": 340
},
{
"epoch": 0.21733588272785215,
"grad_norm": 0.8185917735099792,
"learning_rate": 0.00017954577763053581,
"loss": 0.2701,
"step": 341
},
{
"epoch": 0.2179732313575526,
"grad_norm": 0.6478885412216187,
"learning_rate": 0.00017942270685381207,
"loss": 0.2148,
"step": 342
},
{
"epoch": 0.21861057998725303,
"grad_norm": 0.9637883305549622,
"learning_rate": 0.00017929930938287235,
"loss": 0.2345,
"step": 343
},
{
"epoch": 0.21924792861695347,
"grad_norm": 0.9976230263710022,
"learning_rate": 0.000179175585725295,
"loss": 0.3691,
"step": 344
},
{
"epoch": 0.2198852772466539,
"grad_norm": 1.0075472593307495,
"learning_rate": 0.00017905153638999997,
"loss": 0.3384,
"step": 345
},
{
"epoch": 0.22052262587635438,
"grad_norm": 0.6782506704330444,
"learning_rate": 0.00017892716188724688,
"loss": 0.1469,
"step": 346
},
{
"epoch": 0.22115997450605482,
"grad_norm": 0.9429759383201599,
"learning_rate": 0.0001788024627286329,
"loss": 0.3664,
"step": 347
},
{
"epoch": 0.22179732313575526,
"grad_norm": 1.3366090059280396,
"learning_rate": 0.00017867743942709058,
"loss": 0.772,
"step": 348
},
{
"epoch": 0.2224346717654557,
"grad_norm": 1.0743153095245361,
"learning_rate": 0.00017855209249688584,
"loss": 0.4907,
"step": 349
},
{
"epoch": 0.22307202039515614,
"grad_norm": 1.136670470237732,
"learning_rate": 0.00017842642245361577,
"loss": 0.2522,
"step": 350
},
{
"epoch": 0.2237093690248566,
"grad_norm": 0.7300241589546204,
"learning_rate": 0.00017830042981420652,
"loss": 0.1462,
"step": 351
},
{
"epoch": 0.22434671765455705,
"grad_norm": 0.8688197731971741,
"learning_rate": 0.00017817411509691127,
"loss": 0.4543,
"step": 352
},
{
"epoch": 0.2249840662842575,
"grad_norm": 0.49793678522109985,
"learning_rate": 0.00017804747882130793,
"loss": 0.0915,
"step": 353
},
{
"epoch": 0.22562141491395793,
"grad_norm": 0.9274761080741882,
"learning_rate": 0.00017792052150829716,
"loss": 0.5446,
"step": 354
},
{
"epoch": 0.22625876354365837,
"grad_norm": 0.6762598156929016,
"learning_rate": 0.00017779324368010014,
"loss": 0.1679,
"step": 355
},
{
"epoch": 0.22689611217335884,
"grad_norm": 1.1085275411605835,
"learning_rate": 0.00017766564586025647,
"loss": 0.5198,
"step": 356
},
{
"epoch": 0.22753346080305928,
"grad_norm": 0.7479355931282043,
"learning_rate": 0.00017753772857362194,
"loss": 0.1265,
"step": 357
},
{
"epoch": 0.22817080943275972,
"grad_norm": 0.5710874795913696,
"learning_rate": 0.00017740949234636646,
"loss": 0.0997,
"step": 358
},
{
"epoch": 0.22880815806246016,
"grad_norm": 0.25429093837738037,
"learning_rate": 0.00017728093770597183,
"loss": 0.0217,
"step": 359
},
{
"epoch": 0.2294455066921606,
"grad_norm": 0.8081142902374268,
"learning_rate": 0.0001771520651812296,
"loss": 0.301,
"step": 360
},
{
"epoch": 0.23008285532186107,
"grad_norm": 1.7022185325622559,
"learning_rate": 0.000177022875302239,
"loss": 0.6673,
"step": 361
},
{
"epoch": 0.2307202039515615,
"grad_norm": 1.0308760404586792,
"learning_rate": 0.00017689336860040445,
"loss": 0.3698,
"step": 362
},
{
"epoch": 0.23135755258126195,
"grad_norm": 1.2259387969970703,
"learning_rate": 0.00017676354560843368,
"loss": 0.3355,
"step": 363
},
{
"epoch": 0.2319949012109624,
"grad_norm": 1.0926016569137573,
"learning_rate": 0.00017663340686033556,
"loss": 0.2998,
"step": 364
},
{
"epoch": 0.23263224984066283,
"grad_norm": 0.646233320236206,
"learning_rate": 0.00017650295289141758,
"loss": 0.2323,
"step": 365
},
{
"epoch": 0.2332695984703633,
"grad_norm": 0.7851653099060059,
"learning_rate": 0.00017637218423828394,
"loss": 0.2409,
"step": 366
},
{
"epoch": 0.23390694710006374,
"grad_norm": 0.8424603939056396,
"learning_rate": 0.00017624110143883327,
"loss": 0.2696,
"step": 367
},
{
"epoch": 0.23454429572976418,
"grad_norm": 0.9404459595680237,
"learning_rate": 0.00017610970503225627,
"loss": 0.2925,
"step": 368
},
{
"epoch": 0.23518164435946462,
"grad_norm": 0.27406007051467896,
"learning_rate": 0.0001759779955590338,
"loss": 0.0533,
"step": 369
},
{
"epoch": 0.23581899298916506,
"grad_norm": 0.8875910639762878,
"learning_rate": 0.0001758459735609344,
"loss": 0.2735,
"step": 370
},
{
"epoch": 0.23645634161886553,
"grad_norm": 0.7697153687477112,
"learning_rate": 0.00017571363958101204,
"loss": 0.235,
"step": 371
},
{
"epoch": 0.23709369024856597,
"grad_norm": 1.0897194147109985,
"learning_rate": 0.00017558099416360414,
"loss": 0.3492,
"step": 372
},
{
"epoch": 0.2377310388782664,
"grad_norm": 0.9798071980476379,
"learning_rate": 0.0001754480378543291,
"loss": 0.2151,
"step": 373
},
{
"epoch": 0.23836838750796685,
"grad_norm": 0.9741231799125671,
"learning_rate": 0.00017531477120008414,
"loss": 0.3095,
"step": 374
},
{
"epoch": 0.2390057361376673,
"grad_norm": 0.9084604978561401,
"learning_rate": 0.00017518119474904298,
"loss": 0.2813,
"step": 375
},
{
"epoch": 0.23964308476736776,
"grad_norm": 0.895915150642395,
"learning_rate": 0.00017504730905065378,
"loss": 0.2367,
"step": 376
},
{
"epoch": 0.2402804333970682,
"grad_norm": 1.1402713060379028,
"learning_rate": 0.0001749131146556366,
"loss": 0.4494,
"step": 377
},
{
"epoch": 0.24091778202676864,
"grad_norm": 0.848270058631897,
"learning_rate": 0.00017477861211598138,
"loss": 0.198,
"step": 378
},
{
"epoch": 0.24155513065646908,
"grad_norm": 1.1824499368667603,
"learning_rate": 0.00017464380198494554,
"loss": 0.4257,
"step": 379
},
{
"epoch": 0.24219247928616955,
"grad_norm": 0.6983247995376587,
"learning_rate": 0.0001745086848170517,
"loss": 0.1436,
"step": 380
},
{
"epoch": 0.24282982791587,
"grad_norm": 0.9400845766067505,
"learning_rate": 0.00017437326116808555,
"loss": 0.5486,
"step": 381
},
{
"epoch": 0.24346717654557043,
"grad_norm": 0.9566679000854492,
"learning_rate": 0.00017423753159509327,
"loss": 0.3754,
"step": 382
},
{
"epoch": 0.24410452517527087,
"grad_norm": 0.7652778029441833,
"learning_rate": 0.00017410149665637953,
"loss": 0.2078,
"step": 383
},
{
"epoch": 0.2447418738049713,
"grad_norm": 0.8942538499832153,
"learning_rate": 0.0001739651569115051,
"loss": 0.1942,
"step": 384
},
{
"epoch": 0.24537922243467178,
"grad_norm": 1.1474512815475464,
"learning_rate": 0.00017382851292128444,
"loss": 0.4145,
"step": 385
},
{
"epoch": 0.24601657106437222,
"grad_norm": 1.0497283935546875,
"learning_rate": 0.00017369156524778357,
"loss": 0.1826,
"step": 386
},
{
"epoch": 0.24665391969407266,
"grad_norm": 1.2420200109481812,
"learning_rate": 0.00017355431445431756,
"loss": 0.3426,
"step": 387
},
{
"epoch": 0.2472912683237731,
"grad_norm": 1.3453154563903809,
"learning_rate": 0.00017341676110544844,
"loss": 0.5391,
"step": 388
},
{
"epoch": 0.24792861695347354,
"grad_norm": 1.3887192010879517,
"learning_rate": 0.00017327890576698267,
"loss": 0.3919,
"step": 389
},
{
"epoch": 0.248565965583174,
"grad_norm": 1.5828278064727783,
"learning_rate": 0.00017314074900596897,
"loss": 0.4103,
"step": 390
},
{
"epoch": 0.24920331421287445,
"grad_norm": 0.3824993073940277,
"learning_rate": 0.0001730022913906959,
"loss": 0.0552,
"step": 391
},
{
"epoch": 0.2498406628425749,
"grad_norm": 1.5478538274765015,
"learning_rate": 0.0001728635334906895,
"loss": 0.6023,
"step": 392
},
{
"epoch": 0.25047801147227533,
"grad_norm": 0.9932207465171814,
"learning_rate": 0.000172724475876711,
"loss": 0.2894,
"step": 393
},
{
"epoch": 0.25047801147227533,
"eval_loss": 0.3186395764350891,
"eval_runtime": 6.8366,
"eval_samples_per_second": 9.654,
"eval_steps_per_second": 4.827,
"step": 393
},
{
"epoch": 0.25111536010197577,
"grad_norm": 0.9736688137054443,
"learning_rate": 0.0001725851191207545,
"loss": 0.3935,
"step": 394
},
{
"epoch": 0.2517527087316762,
"grad_norm": 1.0824806690216064,
"learning_rate": 0.0001724454637960446,
"loss": 0.3848,
"step": 395
},
{
"epoch": 0.25239005736137665,
"grad_norm": 1.4277781248092651,
"learning_rate": 0.00017230551047703392,
"loss": 0.6109,
"step": 396
},
{
"epoch": 0.2530274059910771,
"grad_norm": 0.6426457166671753,
"learning_rate": 0.00017216525973940092,
"loss": 0.1823,
"step": 397
},
{
"epoch": 0.2536647546207776,
"grad_norm": 0.8359407186508179,
"learning_rate": 0.00017202471216004744,
"loss": 0.2087,
"step": 398
},
{
"epoch": 0.25430210325047803,
"grad_norm": 0.8000611662864685,
"learning_rate": 0.00017188386831709632,
"loss": 0.1584,
"step": 399
},
{
"epoch": 0.25493945188017847,
"grad_norm": 1.2975292205810547,
"learning_rate": 0.000171742728789889,
"loss": 0.4532,
"step": 400
},
{
"epoch": 0.2555768005098789,
"grad_norm": 1.097817063331604,
"learning_rate": 0.0001716012941589833,
"loss": 0.2365,
"step": 401
},
{
"epoch": 0.25621414913957935,
"grad_norm": 0.9788956046104431,
"learning_rate": 0.00017145956500615075,
"loss": 0.2692,
"step": 402
},
{
"epoch": 0.2568514977692798,
"grad_norm": 0.7173243165016174,
"learning_rate": 0.00017131754191437447,
"loss": 0.1889,
"step": 403
},
{
"epoch": 0.25748884639898023,
"grad_norm": 0.7509357929229736,
"learning_rate": 0.00017117522546784665,
"loss": 0.1934,
"step": 404
},
{
"epoch": 0.25812619502868067,
"grad_norm": 1.1611324548721313,
"learning_rate": 0.00017103261625196607,
"loss": 0.3122,
"step": 405
},
{
"epoch": 0.2587635436583811,
"grad_norm": 0.9995123147964478,
"learning_rate": 0.00017088971485333586,
"loss": 0.4476,
"step": 406
},
{
"epoch": 0.2594008922880816,
"grad_norm": 0.8044164776802063,
"learning_rate": 0.00017074652185976097,
"loss": 0.2134,
"step": 407
},
{
"epoch": 0.26003824091778205,
"grad_norm": 0.8801809549331665,
"learning_rate": 0.00017060303786024577,
"loss": 0.4108,
"step": 408
},
{
"epoch": 0.2606755895474825,
"grad_norm": 1.0581964254379272,
"learning_rate": 0.00017045926344499165,
"loss": 0.1347,
"step": 409
},
{
"epoch": 0.26131293817718293,
"grad_norm": 1.0970072746276855,
"learning_rate": 0.00017031519920539462,
"loss": 0.2554,
"step": 410
},
{
"epoch": 0.26195028680688337,
"grad_norm": 0.9846743941307068,
"learning_rate": 0.00017017084573404284,
"loss": 0.2267,
"step": 411
},
{
"epoch": 0.2625876354365838,
"grad_norm": 0.8163756132125854,
"learning_rate": 0.00017002620362471406,
"loss": 0.1773,
"step": 412
},
{
"epoch": 0.26322498406628425,
"grad_norm": 0.9301058650016785,
"learning_rate": 0.00016988127347237353,
"loss": 0.3135,
"step": 413
},
{
"epoch": 0.2638623326959847,
"grad_norm": 0.6529296636581421,
"learning_rate": 0.00016973605587317108,
"loss": 0.1011,
"step": 414
},
{
"epoch": 0.26449968132568513,
"grad_norm": 1.206225037574768,
"learning_rate": 0.00016959055142443913,
"loss": 0.3957,
"step": 415
},
{
"epoch": 0.2651370299553856,
"grad_norm": 0.800182580947876,
"learning_rate": 0.00016944476072468988,
"loss": 0.1754,
"step": 416
},
{
"epoch": 0.26577437858508607,
"grad_norm": 0.896647036075592,
"learning_rate": 0.00016929868437361297,
"loss": 0.433,
"step": 417
},
{
"epoch": 0.2664117272147865,
"grad_norm": 0.6034906506538391,
"learning_rate": 0.00016915232297207315,
"loss": 0.1634,
"step": 418
},
{
"epoch": 0.26704907584448695,
"grad_norm": 1.0596516132354736,
"learning_rate": 0.00016900567712210758,
"loss": 0.2917,
"step": 419
},
{
"epoch": 0.2676864244741874,
"grad_norm": 0.9491992592811584,
"learning_rate": 0.00016885874742692347,
"loss": 0.3213,
"step": 420
},
{
"epoch": 0.26832377310388783,
"grad_norm": 0.97361159324646,
"learning_rate": 0.0001687115344908956,
"loss": 0.1555,
"step": 421
},
{
"epoch": 0.26896112173358827,
"grad_norm": 0.6845320463180542,
"learning_rate": 0.0001685640389195639,
"loss": 0.1863,
"step": 422
},
{
"epoch": 0.2695984703632887,
"grad_norm": 0.9510777592658997,
"learning_rate": 0.00016841626131963074,
"loss": 0.2023,
"step": 423
},
{
"epoch": 0.27023581899298915,
"grad_norm": 0.6615311503410339,
"learning_rate": 0.00016826820229895865,
"loss": 0.2278,
"step": 424
},
{
"epoch": 0.2708731676226896,
"grad_norm": 0.9706951379776001,
"learning_rate": 0.00016811986246656778,
"loss": 0.3091,
"step": 425
},
{
"epoch": 0.27151051625239003,
"grad_norm": 0.6907554268836975,
"learning_rate": 0.0001679712424326333,
"loss": 0.2132,
"step": 426
},
{
"epoch": 0.27214786488209053,
"grad_norm": 0.776425838470459,
"learning_rate": 0.00016782234280848294,
"loss": 0.3146,
"step": 427
},
{
"epoch": 0.27278521351179097,
"grad_norm": 0.8611300587654114,
"learning_rate": 0.00016767316420659458,
"loss": 0.1037,
"step": 428
},
{
"epoch": 0.2734225621414914,
"grad_norm": 0.5898683071136475,
"learning_rate": 0.00016752370724059352,
"loss": 0.0605,
"step": 429
},
{
"epoch": 0.27405991077119185,
"grad_norm": 0.8071754574775696,
"learning_rate": 0.00016737397252525012,
"loss": 0.2362,
"step": 430
},
{
"epoch": 0.2746972594008923,
"grad_norm": 0.7139734029769897,
"learning_rate": 0.00016722396067647726,
"loss": 0.165,
"step": 431
},
{
"epoch": 0.27533460803059273,
"grad_norm": 0.8318780660629272,
"learning_rate": 0.0001670736723113277,
"loss": 0.3067,
"step": 432
},
{
"epoch": 0.27597195666029317,
"grad_norm": 0.8397043943405151,
"learning_rate": 0.0001669231080479917,
"loss": 0.2011,
"step": 433
},
{
"epoch": 0.2766093052899936,
"grad_norm": 0.5502068996429443,
"learning_rate": 0.0001667722685057943,
"loss": 0.1465,
"step": 434
},
{
"epoch": 0.27724665391969405,
"grad_norm": 0.8265907168388367,
"learning_rate": 0.00016662115430519288,
"loss": 0.2035,
"step": 435
},
{
"epoch": 0.2778840025493945,
"grad_norm": 1.110367774963379,
"learning_rate": 0.0001664697660677746,
"loss": 0.4208,
"step": 436
},
{
"epoch": 0.278521351179095,
"grad_norm": 0.6079518795013428,
"learning_rate": 0.0001663181044162539,
"loss": 0.1592,
"step": 437
},
{
"epoch": 0.27915869980879543,
"grad_norm": 0.8618608713150024,
"learning_rate": 0.00016616616997446978,
"loss": 0.3148,
"step": 438
},
{
"epoch": 0.27979604843849587,
"grad_norm": 1.0243940353393555,
"learning_rate": 0.0001660139633673833,
"loss": 0.2974,
"step": 439
},
{
"epoch": 0.2804333970681963,
"grad_norm": 1.6069848537445068,
"learning_rate": 0.00016586148522107518,
"loss": 0.4136,
"step": 440
},
{
"epoch": 0.28107074569789675,
"grad_norm": 0.42425957322120667,
"learning_rate": 0.0001657087361627429,
"loss": 0.0253,
"step": 441
},
{
"epoch": 0.2817080943275972,
"grad_norm": 0.855467677116394,
"learning_rate": 0.00016555571682069848,
"loss": 0.4067,
"step": 442
},
{
"epoch": 0.28234544295729763,
"grad_norm": 1.3716607093811035,
"learning_rate": 0.00016540242782436555,
"loss": 0.5167,
"step": 443
},
{
"epoch": 0.2829827915869981,
"grad_norm": 1.206874132156372,
"learning_rate": 0.00016524886980427697,
"loss": 0.393,
"step": 444
},
{
"epoch": 0.2836201402166985,
"grad_norm": 0.6875660419464111,
"learning_rate": 0.0001650950433920723,
"loss": 0.1238,
"step": 445
},
{
"epoch": 0.28425748884639895,
"grad_norm": 1.154205560684204,
"learning_rate": 0.00016494094922049493,
"loss": 0.3138,
"step": 446
},
{
"epoch": 0.28489483747609945,
"grad_norm": 1.0486526489257812,
"learning_rate": 0.00016478658792338979,
"loss": 0.4165,
"step": 447
},
{
"epoch": 0.2855321861057999,
"grad_norm": 0.4852248728275299,
"learning_rate": 0.00016463196013570048,
"loss": 0.0771,
"step": 448
},
{
"epoch": 0.28616953473550033,
"grad_norm": 0.9326637983322144,
"learning_rate": 0.00016447706649346682,
"loss": 0.3766,
"step": 449
},
{
"epoch": 0.28680688336520077,
"grad_norm": 0.7197770476341248,
"learning_rate": 0.0001643219076338222,
"loss": 0.0751,
"step": 450
},
{
"epoch": 0.2874442319949012,
"grad_norm": 0.8876821994781494,
"learning_rate": 0.0001641664841949909,
"loss": 0.0775,
"step": 451
},
{
"epoch": 0.28808158062460165,
"grad_norm": 0.7619658708572388,
"learning_rate": 0.00016401079681628558,
"loss": 0.2741,
"step": 452
},
{
"epoch": 0.2887189292543021,
"grad_norm": 0.8145762085914612,
"learning_rate": 0.00016385484613810447,
"loss": 0.2084,
"step": 453
},
{
"epoch": 0.28935627788400253,
"grad_norm": 0.6327223777770996,
"learning_rate": 0.00016369863280192898,
"loss": 0.1823,
"step": 454
},
{
"epoch": 0.289993626513703,
"grad_norm": 1.0835278034210205,
"learning_rate": 0.00016354215745032082,
"loss": 0.2354,
"step": 455
},
{
"epoch": 0.29063097514340347,
"grad_norm": 1.0292389392852783,
"learning_rate": 0.00016338542072691953,
"loss": 0.3637,
"step": 456
},
{
"epoch": 0.2912683237731039,
"grad_norm": 0.12314633280038834,
"learning_rate": 0.00016322842327643969,
"loss": 0.0265,
"step": 457
},
{
"epoch": 0.29190567240280435,
"grad_norm": 0.9486640691757202,
"learning_rate": 0.0001630711657446684,
"loss": 0.2199,
"step": 458
},
{
"epoch": 0.2925430210325048,
"grad_norm": 0.7531243562698364,
"learning_rate": 0.0001629136487784626,
"loss": 0.151,
"step": 459
},
{
"epoch": 0.29318036966220523,
"grad_norm": 0.35610267519950867,
"learning_rate": 0.00016275587302574634,
"loss": 0.0462,
"step": 460
},
{
"epoch": 0.29381771829190567,
"grad_norm": 1.1195050477981567,
"learning_rate": 0.00016259783913550812,
"loss": 0.3106,
"step": 461
},
{
"epoch": 0.2944550669216061,
"grad_norm": 1.2721606492996216,
"learning_rate": 0.0001624395477577983,
"loss": 0.439,
"step": 462
},
{
"epoch": 0.29509241555130655,
"grad_norm": 1.0469766855239868,
"learning_rate": 0.00016228099954372635,
"loss": 0.2614,
"step": 463
},
{
"epoch": 0.295729764181007,
"grad_norm": 1.191555380821228,
"learning_rate": 0.00016212219514545823,
"loss": 0.3795,
"step": 464
},
{
"epoch": 0.29636711281070743,
"grad_norm": 0.8967866897583008,
"learning_rate": 0.00016196313521621365,
"loss": 0.313,
"step": 465
},
{
"epoch": 0.29700446144040793,
"grad_norm": 0.8236089944839478,
"learning_rate": 0.00016180382041026343,
"loss": 0.2226,
"step": 466
},
{
"epoch": 0.29764181007010837,
"grad_norm": 0.8141946196556091,
"learning_rate": 0.00016164425138292675,
"loss": 0.1926,
"step": 467
},
{
"epoch": 0.2982791586998088,
"grad_norm": 0.5971007347106934,
"learning_rate": 0.00016148442879056855,
"loss": 0.1583,
"step": 468
},
{
"epoch": 0.29891650732950925,
"grad_norm": 1.1322076320648193,
"learning_rate": 0.00016132435329059677,
"loss": 0.4261,
"step": 469
},
{
"epoch": 0.2995538559592097,
"grad_norm": 1.0180432796478271,
"learning_rate": 0.00016116402554145962,
"loss": 0.4002,
"step": 470
},
{
"epoch": 0.30019120458891013,
"grad_norm": 1.0787919759750366,
"learning_rate": 0.00016100344620264284,
"loss": 0.2316,
"step": 471
},
{
"epoch": 0.3008285532186106,
"grad_norm": 0.9993605017662048,
"learning_rate": 0.00016084261593466725,
"loss": 0.2434,
"step": 472
},
{
"epoch": 0.301465901848311,
"grad_norm": 0.9957782030105591,
"learning_rate": 0.00016068153539908558,
"loss": 0.2615,
"step": 473
},
{
"epoch": 0.30210325047801145,
"grad_norm": 1.2358076572418213,
"learning_rate": 0.0001605202052584802,
"loss": 0.3877,
"step": 474
},
{
"epoch": 0.3027405991077119,
"grad_norm": 1.317139744758606,
"learning_rate": 0.00016035862617646011,
"loss": 0.3078,
"step": 475
},
{
"epoch": 0.3033779477374124,
"grad_norm": 0.9851285815238953,
"learning_rate": 0.0001601967988176583,
"loss": 0.1896,
"step": 476
},
{
"epoch": 0.30401529636711283,
"grad_norm": 1.122254729270935,
"learning_rate": 0.00016003472384772897,
"loss": 0.4244,
"step": 477
},
{
"epoch": 0.30465264499681327,
"grad_norm": 1.017385721206665,
"learning_rate": 0.00015987240193334495,
"loss": 0.346,
"step": 478
},
{
"epoch": 0.3052899936265137,
"grad_norm": 1.242531180381775,
"learning_rate": 0.00015970983374219473,
"loss": 0.426,
"step": 479
},
{
"epoch": 0.30592734225621415,
"grad_norm": 0.6854953169822693,
"learning_rate": 0.00015954701994297987,
"loss": 0.1504,
"step": 480
},
{
"epoch": 0.3065646908859146,
"grad_norm": 0.9415954947471619,
"learning_rate": 0.00015938396120541222,
"loss": 0.2962,
"step": 481
},
{
"epoch": 0.30720203951561503,
"grad_norm": 0.9208685755729675,
"learning_rate": 0.00015922065820021111,
"loss": 0.4935,
"step": 482
},
{
"epoch": 0.3078393881453155,
"grad_norm": 0.5450341701507568,
"learning_rate": 0.0001590571115991007,
"loss": 0.1196,
"step": 483
},
{
"epoch": 0.3084767367750159,
"grad_norm": 0.9460034370422363,
"learning_rate": 0.00015889332207480705,
"loss": 0.3683,
"step": 484
},
{
"epoch": 0.30911408540471635,
"grad_norm": 0.6469952464103699,
"learning_rate": 0.0001587292903010555,
"loss": 0.1868,
"step": 485
},
{
"epoch": 0.30975143403441685,
"grad_norm": 0.9460652470588684,
"learning_rate": 0.00015856501695256785,
"loss": 0.2783,
"step": 486
},
{
"epoch": 0.3103887826641173,
"grad_norm": 1.3356025218963623,
"learning_rate": 0.0001584005027050596,
"loss": 0.0991,
"step": 487
},
{
"epoch": 0.31102613129381773,
"grad_norm": 0.8451143503189087,
"learning_rate": 0.00015823574823523715,
"loss": 0.2775,
"step": 488
},
{
"epoch": 0.31166347992351817,
"grad_norm": 1.184491753578186,
"learning_rate": 0.000158070754220795,
"loss": 0.3025,
"step": 489
},
{
"epoch": 0.3123008285532186,
"grad_norm": 1.445219874382019,
"learning_rate": 0.000157905521340413,
"loss": 0.4519,
"step": 490
},
{
"epoch": 0.31293817718291905,
"grad_norm": 0.803226888179779,
"learning_rate": 0.00015774005027375347,
"loss": 0.163,
"step": 491
},
{
"epoch": 0.3135755258126195,
"grad_norm": 0.6444011926651001,
"learning_rate": 0.0001575743417014586,
"loss": 0.2383,
"step": 492
},
{
"epoch": 0.31421287444231993,
"grad_norm": 0.8938921093940735,
"learning_rate": 0.0001574083963051475,
"loss": 0.2663,
"step": 493
},
{
"epoch": 0.3148502230720204,
"grad_norm": 1.0509055852890015,
"learning_rate": 0.00015724221476741334,
"loss": 0.3239,
"step": 494
},
{
"epoch": 0.3154875717017208,
"grad_norm": 0.757455587387085,
"learning_rate": 0.00015707579777182065,
"loss": 0.2904,
"step": 495
},
{
"epoch": 0.3161249203314213,
"grad_norm": 0.689369797706604,
"learning_rate": 0.00015690914600290258,
"loss": 0.1419,
"step": 496
},
{
"epoch": 0.31676226896112175,
"grad_norm": 0.759578287601471,
"learning_rate": 0.00015674226014615784,
"loss": 0.1816,
"step": 497
},
{
"epoch": 0.3173996175908222,
"grad_norm": 0.46891000866889954,
"learning_rate": 0.00015657514088804813,
"loss": 0.0876,
"step": 498
},
{
"epoch": 0.31803696622052263,
"grad_norm": 0.7397509217262268,
"learning_rate": 0.0001564077889159952,
"loss": 0.2972,
"step": 499
},
{
"epoch": 0.3186743148502231,
"grad_norm": 0.7705743312835693,
"learning_rate": 0.00015624020491837804,
"loss": 0.2357,
"step": 500
},
{
"epoch": 0.3193116634799235,
"grad_norm": 0.79763263463974,
"learning_rate": 0.00015607238958452994,
"loss": 0.2184,
"step": 501
},
{
"epoch": 0.31994901210962395,
"grad_norm": 0.7162529826164246,
"learning_rate": 0.00015590434360473595,
"loss": 0.2179,
"step": 502
},
{
"epoch": 0.3205863607393244,
"grad_norm": 0.9361833333969116,
"learning_rate": 0.00015573606767022968,
"loss": 0.1798,
"step": 503
},
{
"epoch": 0.32122370936902483,
"grad_norm": 1.368133783340454,
"learning_rate": 0.00015556756247319073,
"loss": 0.418,
"step": 504
},
{
"epoch": 0.3218610579987253,
"grad_norm": 1.0074046850204468,
"learning_rate": 0.00015539882870674166,
"loss": 0.3763,
"step": 505
},
{
"epoch": 0.32249840662842577,
"grad_norm": 0.974919319152832,
"learning_rate": 0.0001552298670649453,
"loss": 0.3241,
"step": 506
},
{
"epoch": 0.3231357552581262,
"grad_norm": 0.7152822613716125,
"learning_rate": 0.00015506067824280176,
"loss": 0.1718,
"step": 507
},
{
"epoch": 0.32377310388782665,
"grad_norm": 1.2417292594909668,
"learning_rate": 0.00015489126293624566,
"loss": 0.4152,
"step": 508
},
{
"epoch": 0.3244104525175271,
"grad_norm": 0.3077457845211029,
"learning_rate": 0.00015472162184214315,
"loss": 0.0384,
"step": 509
},
{
"epoch": 0.32504780114722753,
"grad_norm": 0.8533461689949036,
"learning_rate": 0.0001545517556582892,
"loss": 0.2198,
"step": 510
},
{
"epoch": 0.325685149776928,
"grad_norm": 0.9196656346321106,
"learning_rate": 0.0001543816650834047,
"loss": 0.2798,
"step": 511
},
{
"epoch": 0.3263224984066284,
"grad_norm": 0.7563450932502747,
"learning_rate": 0.00015421135081713345,
"loss": 0.1582,
"step": 512
},
{
"epoch": 0.32695984703632885,
"grad_norm": 0.9521018266677856,
"learning_rate": 0.00015404081356003941,
"loss": 0.2324,
"step": 513
},
{
"epoch": 0.3275971956660293,
"grad_norm": 1.535440444946289,
"learning_rate": 0.00015387005401360375,
"loss": 0.5038,
"step": 514
},
{
"epoch": 0.3282345442957298,
"grad_norm": 0.8014768362045288,
"learning_rate": 0.00015369907288022205,
"loss": 0.0883,
"step": 515
},
{
"epoch": 0.32887189292543023,
"grad_norm": 0.7590137124061584,
"learning_rate": 0.00015352787086320136,
"loss": 0.195,
"step": 516
},
{
"epoch": 0.32950924155513067,
"grad_norm": 0.3749714493751526,
"learning_rate": 0.00015335644866675715,
"loss": 0.055,
"step": 517
},
{
"epoch": 0.3301465901848311,
"grad_norm": 1.0182459354400635,
"learning_rate": 0.00015318480699601084,
"loss": 0.302,
"step": 518
},
{
"epoch": 0.33078393881453155,
"grad_norm": 0.7702544927597046,
"learning_rate": 0.00015301294655698642,
"loss": 0.4102,
"step": 519
},
{
"epoch": 0.331421287444232,
"grad_norm": 1.2974505424499512,
"learning_rate": 0.0001528408680566078,
"loss": 0.5051,
"step": 520
},
{
"epoch": 0.33205863607393243,
"grad_norm": 0.7069422006607056,
"learning_rate": 0.00015266857220269588,
"loss": 0.1705,
"step": 521
},
{
"epoch": 0.3326959847036329,
"grad_norm": 1.0596264600753784,
"learning_rate": 0.00015249605970396558,
"loss": 0.3468,
"step": 522
},
{
"epoch": 0.3333333333333333,
"grad_norm": 1.1788617372512817,
"learning_rate": 0.000152323331270023,
"loss": 0.4014,
"step": 523
},
{
"epoch": 0.33397068196303376,
"grad_norm": 0.679743766784668,
"learning_rate": 0.00015215038761136248,
"loss": 0.2643,
"step": 524
},
{
"epoch": 0.33460803059273425,
"grad_norm": 0.6367809772491455,
"learning_rate": 0.00015197722943936348,
"loss": 0.1732,
"step": 525
},
{
"epoch": 0.3352453792224347,
"grad_norm": 0.9102736711502075,
"learning_rate": 0.0001518038574662881,
"loss": 0.2655,
"step": 526
},
{
"epoch": 0.33588272785213513,
"grad_norm": 0.7565219402313232,
"learning_rate": 0.0001516302724052777,
"loss": 0.2432,
"step": 527
},
{
"epoch": 0.3365200764818356,
"grad_norm": 1.1752339601516724,
"learning_rate": 0.00015145647497035017,
"loss": 0.4928,
"step": 528
},
{
"epoch": 0.337157425111536,
"grad_norm": 1.0684925317764282,
"learning_rate": 0.00015128246587639704,
"loss": 0.4376,
"step": 529
},
{
"epoch": 0.33779477374123645,
"grad_norm": 0.874518632888794,
"learning_rate": 0.00015110824583918034,
"loss": 0.2538,
"step": 530
},
{
"epoch": 0.3384321223709369,
"grad_norm": 0.9514549374580383,
"learning_rate": 0.00015093381557532997,
"loss": 0.2469,
"step": 531
},
{
"epoch": 0.33906947100063733,
"grad_norm": 1.0768482685089111,
"learning_rate": 0.00015075917580234042,
"loss": 0.1939,
"step": 532
},
{
"epoch": 0.3397068196303378,
"grad_norm": 1.386254072189331,
"learning_rate": 0.00015058432723856802,
"loss": 0.3986,
"step": 533
},
{
"epoch": 0.3403441682600382,
"grad_norm": 0.921574056148529,
"learning_rate": 0.0001504092706032279,
"loss": 0.2328,
"step": 534
},
{
"epoch": 0.3409815168897387,
"grad_norm": 1.0182220935821533,
"learning_rate": 0.00015023400661639112,
"loss": 0.194,
"step": 535
},
{
"epoch": 0.34161886551943915,
"grad_norm": 0.9648236036300659,
"learning_rate": 0.00015005853599898161,
"loss": 0.3766,
"step": 536
},
{
"epoch": 0.3422562141491396,
"grad_norm": 0.7621607184410095,
"learning_rate": 0.00014988285947277328,
"loss": 0.2251,
"step": 537
},
{
"epoch": 0.34289356277884003,
"grad_norm": 0.7492721676826477,
"learning_rate": 0.00014970697776038692,
"loss": 0.2206,
"step": 538
},
{
"epoch": 0.3435309114085405,
"grad_norm": 0.7960018515586853,
"learning_rate": 0.0001495308915852875,
"loss": 0.1679,
"step": 539
},
{
"epoch": 0.3441682600382409,
"grad_norm": 0.20154525339603424,
"learning_rate": 0.00014935460167178085,
"loss": 0.0283,
"step": 540
},
{
"epoch": 0.34480560866794135,
"grad_norm": 0.5639259219169617,
"learning_rate": 0.0001491781087450109,
"loss": 0.0939,
"step": 541
},
{
"epoch": 0.3454429572976418,
"grad_norm": 0.7699045538902283,
"learning_rate": 0.00014900141353095672,
"loss": 0.2282,
"step": 542
},
{
"epoch": 0.34608030592734224,
"grad_norm": 0.8119546175003052,
"learning_rate": 0.00014882451675642938,
"loss": 0.0554,
"step": 543
},
{
"epoch": 0.3467176545570427,
"grad_norm": 0.9836156964302063,
"learning_rate": 0.000148647419149069,
"loss": 0.2998,
"step": 544
},
{
"epoch": 0.34735500318674317,
"grad_norm": 1.9436732530593872,
"learning_rate": 0.000148470121437342,
"loss": 0.4807,
"step": 545
},
{
"epoch": 0.3479923518164436,
"grad_norm": 1.3825451135635376,
"learning_rate": 0.00014829262435053767,
"loss": 0.3518,
"step": 546
},
{
"epoch": 0.34862970044614405,
"grad_norm": 0.983121395111084,
"learning_rate": 0.0001481149286187655,
"loss": 0.2036,
"step": 547
},
{
"epoch": 0.3492670490758445,
"grad_norm": 0.5847717523574829,
"learning_rate": 0.00014793703497295208,
"loss": 0.1869,
"step": 548
},
{
"epoch": 0.34990439770554493,
"grad_norm": 1.198561429977417,
"learning_rate": 0.0001477589441448381,
"loss": 0.377,
"step": 549
},
{
"epoch": 0.3505417463352454,
"grad_norm": 0.8971907496452332,
"learning_rate": 0.00014758065686697533,
"loss": 0.3066,
"step": 550
},
{
"epoch": 0.3511790949649458,
"grad_norm": 1.1344964504241943,
"learning_rate": 0.00014740217387272355,
"loss": 0.3708,
"step": 551
},
{
"epoch": 0.35181644359464626,
"grad_norm": 1.0182440280914307,
"learning_rate": 0.00014722349589624763,
"loss": 0.2802,
"step": 552
},
{
"epoch": 0.3524537922243467,
"grad_norm": 0.9438650608062744,
"learning_rate": 0.00014704462367251458,
"loss": 0.2689,
"step": 553
},
{
"epoch": 0.35309114085404714,
"grad_norm": 0.9879765510559082,
"learning_rate": 0.0001468655579372902,
"loss": 0.2899,
"step": 554
},
{
"epoch": 0.35372848948374763,
"grad_norm": 0.803093433380127,
"learning_rate": 0.00014668629942713644,
"loss": 0.2963,
"step": 555
},
{
"epoch": 0.3543658381134481,
"grad_norm": 0.7461884021759033,
"learning_rate": 0.00014650684887940817,
"loss": 0.2399,
"step": 556
},
{
"epoch": 0.3550031867431485,
"grad_norm": 0.6757521629333496,
"learning_rate": 0.00014632720703225015,
"loss": 0.1402,
"step": 557
},
{
"epoch": 0.35564053537284895,
"grad_norm": 1.083609938621521,
"learning_rate": 0.00014614737462459405,
"loss": 0.4086,
"step": 558
},
{
"epoch": 0.3562778840025494,
"grad_norm": 0.786292314529419,
"learning_rate": 0.00014596735239615533,
"loss": 0.2077,
"step": 559
},
{
"epoch": 0.35691523263224983,
"grad_norm": 0.7855844497680664,
"learning_rate": 0.00014578714108743035,
"loss": 0.2561,
"step": 560
},
{
"epoch": 0.3575525812619503,
"grad_norm": 0.8310897350311279,
"learning_rate": 0.00014560674143969312,
"loss": 0.2841,
"step": 561
},
{
"epoch": 0.3581899298916507,
"grad_norm": 0.9358211159706116,
"learning_rate": 0.00014542615419499243,
"loss": 0.1825,
"step": 562
},
{
"epoch": 0.35882727852135116,
"grad_norm": 0.7215493321418762,
"learning_rate": 0.00014524538009614868,
"loss": 0.136,
"step": 563
},
{
"epoch": 0.35946462715105165,
"grad_norm": 1.0193787813186646,
"learning_rate": 0.0001450644198867509,
"loss": 0.2766,
"step": 564
},
{
"epoch": 0.3601019757807521,
"grad_norm": 0.9681364893913269,
"learning_rate": 0.00014488327431115365,
"loss": 0.4552,
"step": 565
},
{
"epoch": 0.36073932441045253,
"grad_norm": 0.5649060010910034,
"learning_rate": 0.00014470194411447395,
"loss": 0.0584,
"step": 566
},
{
"epoch": 0.361376673040153,
"grad_norm": 1.0573179721832275,
"learning_rate": 0.00014452043004258827,
"loss": 0.4635,
"step": 567
},
{
"epoch": 0.3620140216698534,
"grad_norm": 0.8015258312225342,
"learning_rate": 0.0001443387328421294,
"loss": 0.2213,
"step": 568
},
{
"epoch": 0.36265137029955385,
"grad_norm": 0.7115965485572815,
"learning_rate": 0.00014415685326048337,
"loss": 0.1824,
"step": 569
},
{
"epoch": 0.3632887189292543,
"grad_norm": 0.7246455550193787,
"learning_rate": 0.00014397479204578643,
"loss": 0.1991,
"step": 570
},
{
"epoch": 0.36392606755895474,
"grad_norm": 0.8708940148353577,
"learning_rate": 0.00014379254994692205,
"loss": 0.1796,
"step": 571
},
{
"epoch": 0.3645634161886552,
"grad_norm": 1.4185974597930908,
"learning_rate": 0.00014361012771351765,
"loss": 0.3133,
"step": 572
},
{
"epoch": 0.3652007648183556,
"grad_norm": 0.9905892610549927,
"learning_rate": 0.00014342752609594153,
"loss": 0.2534,
"step": 573
},
{
"epoch": 0.3658381134480561,
"grad_norm": 1.5535361766815186,
"learning_rate": 0.00014324474584530004,
"loss": 0.2541,
"step": 574
},
{
"epoch": 0.36647546207775655,
"grad_norm": 0.7428743839263916,
"learning_rate": 0.00014306178771343419,
"loss": 0.2141,
"step": 575
},
{
"epoch": 0.367112810707457,
"grad_norm": 0.8181564807891846,
"learning_rate": 0.00014287865245291668,
"loss": 0.2067,
"step": 576
},
{
"epoch": 0.36775015933715743,
"grad_norm": 0.8336141705513,
"learning_rate": 0.0001426953408170489,
"loss": 0.1796,
"step": 577
},
{
"epoch": 0.3683875079668579,
"grad_norm": 0.9294024705886841,
"learning_rate": 0.00014251185355985762,
"loss": 0.4238,
"step": 578
},
{
"epoch": 0.3690248565965583,
"grad_norm": 1.274320125579834,
"learning_rate": 0.00014232819143609205,
"loss": 0.3288,
"step": 579
},
{
"epoch": 0.36966220522625876,
"grad_norm": 0.6250787377357483,
"learning_rate": 0.0001421443552012207,
"loss": 0.1523,
"step": 580
},
{
"epoch": 0.3702995538559592,
"grad_norm": 0.9842786192893982,
"learning_rate": 0.00014196034561142831,
"loss": 0.3179,
"step": 581
},
{
"epoch": 0.37093690248565964,
"grad_norm": 0.9008854031562805,
"learning_rate": 0.0001417761634236126,
"loss": 0.235,
"step": 582
},
{
"epoch": 0.3715742511153601,
"grad_norm": 0.799509584903717,
"learning_rate": 0.00014159180939538128,
"loss": 0.2348,
"step": 583
},
{
"epoch": 0.3722115997450606,
"grad_norm": 1.1855205297470093,
"learning_rate": 0.00014140728428504887,
"loss": 0.4213,
"step": 584
},
{
"epoch": 0.372848948374761,
"grad_norm": 0.7991178631782532,
"learning_rate": 0.0001412225888516337,
"loss": 0.3277,
"step": 585
},
{
"epoch": 0.37348629700446145,
"grad_norm": 0.8852383494377136,
"learning_rate": 0.00014103772385485466,
"loss": 0.2058,
"step": 586
},
{
"epoch": 0.3741236456341619,
"grad_norm": 0.8656224012374878,
"learning_rate": 0.00014085269005512804,
"loss": 0.378,
"step": 587
},
{
"epoch": 0.37476099426386233,
"grad_norm": 0.7828904390335083,
"learning_rate": 0.0001406674882135646,
"loss": 0.29,
"step": 588
},
{
"epoch": 0.3753983428935628,
"grad_norm": 0.5957149863243103,
"learning_rate": 0.00014048211909196626,
"loss": 0.1579,
"step": 589
},
{
"epoch": 0.3760356915232632,
"grad_norm": 0.7790771126747131,
"learning_rate": 0.00014029658345282296,
"loss": 0.2006,
"step": 590
},
{
"epoch": 0.37667304015296366,
"grad_norm": 0.3924875557422638,
"learning_rate": 0.00014011088205930972,
"loss": 0.0415,
"step": 591
},
{
"epoch": 0.3773103887826641,
"grad_norm": 0.7297988533973694,
"learning_rate": 0.00013992501567528326,
"loss": 0.2259,
"step": 592
},
{
"epoch": 0.37794773741236454,
"grad_norm": 0.8644670248031616,
"learning_rate": 0.00013973898506527895,
"loss": 0.1879,
"step": 593
},
{
"epoch": 0.37858508604206503,
"grad_norm": 0.6984097957611084,
"learning_rate": 0.00013955279099450786,
"loss": 0.213,
"step": 594
},
{
"epoch": 0.3792224346717655,
"grad_norm": 0.8078986406326294,
"learning_rate": 0.00013936643422885316,
"loss": 0.2318,
"step": 595
},
{
"epoch": 0.3798597833014659,
"grad_norm": 0.8908408880233765,
"learning_rate": 0.0001391799155348675,
"loss": 0.2093,
"step": 596
},
{
"epoch": 0.38049713193116635,
"grad_norm": 0.9252097606658936,
"learning_rate": 0.00013899323567976942,
"loss": 0.2628,
"step": 597
},
{
"epoch": 0.3811344805608668,
"grad_norm": 0.7688030004501343,
"learning_rate": 0.0001388063954314405,
"loss": 0.2409,
"step": 598
},
{
"epoch": 0.38177182919056724,
"grad_norm": 0.6707478761672974,
"learning_rate": 0.00013861939555842194,
"loss": 0.2762,
"step": 599
},
{
"epoch": 0.3824091778202677,
"grad_norm": 0.9521300792694092,
"learning_rate": 0.00013843223682991166,
"loss": 0.2672,
"step": 600
},
{
"epoch": 0.3830465264499681,
"grad_norm": 0.6128231883049011,
"learning_rate": 0.00013824492001576093,
"loss": 0.0728,
"step": 601
},
{
"epoch": 0.38368387507966856,
"grad_norm": 0.7471403479576111,
"learning_rate": 0.00013805744588647138,
"loss": 0.1727,
"step": 602
},
{
"epoch": 0.384321223709369,
"grad_norm": 0.8444771766662598,
"learning_rate": 0.0001378698152131916,
"loss": 0.1536,
"step": 603
},
{
"epoch": 0.3849585723390695,
"grad_norm": 0.19250012934207916,
"learning_rate": 0.00013768202876771418,
"loss": 0.0251,
"step": 604
},
{
"epoch": 0.38559592096876993,
"grad_norm": 1.0509403944015503,
"learning_rate": 0.0001374940873224724,
"loss": 0.3439,
"step": 605
},
{
"epoch": 0.3862332695984704,
"grad_norm": 1.0547863245010376,
"learning_rate": 0.00013730599165053727,
"loss": 0.4169,
"step": 606
},
{
"epoch": 0.3868706182281708,
"grad_norm": 1.0675448179244995,
"learning_rate": 0.000137117742525614,
"loss": 0.3226,
"step": 607
},
{
"epoch": 0.38750796685787126,
"grad_norm": 1.3179535865783691,
"learning_rate": 0.00013692934072203903,
"loss": 0.5797,
"step": 608
},
{
"epoch": 0.3881453154875717,
"grad_norm": 1.1440210342407227,
"learning_rate": 0.00013674078701477702,
"loss": 0.2382,
"step": 609
},
{
"epoch": 0.38878266411727214,
"grad_norm": 0.8661467432975769,
"learning_rate": 0.0001365520821794172,
"loss": 0.2992,
"step": 610
},
{
"epoch": 0.3894200127469726,
"grad_norm": 0.7292525172233582,
"learning_rate": 0.00013636322699217058,
"loss": 0.1423,
"step": 611
},
{
"epoch": 0.390057361376673,
"grad_norm": 0.7456052899360657,
"learning_rate": 0.00013617422222986667,
"loss": 0.1798,
"step": 612
},
{
"epoch": 0.3906947100063735,
"grad_norm": 0.9199575781822205,
"learning_rate": 0.0001359850686699501,
"loss": 0.3076,
"step": 613
},
{
"epoch": 0.39133205863607395,
"grad_norm": 0.6483279466629028,
"learning_rate": 0.00013579576709047764,
"loss": 0.2881,
"step": 614
},
{
"epoch": 0.3919694072657744,
"grad_norm": 0.38735392689704895,
"learning_rate": 0.00013560631827011502,
"loss": 0.0693,
"step": 615
},
{
"epoch": 0.39260675589547483,
"grad_norm": 1.5048894882202148,
"learning_rate": 0.0001354167229881333,
"loss": 0.3857,
"step": 616
},
{
"epoch": 0.3932441045251753,
"grad_norm": 0.9635381102561951,
"learning_rate": 0.0001352269820244064,
"loss": 0.2325,
"step": 617
},
{
"epoch": 0.3938814531548757,
"grad_norm": 0.37867435812950134,
"learning_rate": 0.0001350370961594072,
"loss": 0.0515,
"step": 618
},
{
"epoch": 0.39451880178457616,
"grad_norm": 0.5940608382225037,
"learning_rate": 0.00013484706617420466,
"loss": 0.1663,
"step": 619
},
{
"epoch": 0.3951561504142766,
"grad_norm": 0.42715132236480713,
"learning_rate": 0.00013465689285046065,
"loss": 0.0571,
"step": 620
},
{
"epoch": 0.39579349904397704,
"grad_norm": 0.41966956853866577,
"learning_rate": 0.00013446657697042656,
"loss": 0.0397,
"step": 621
},
{
"epoch": 0.3964308476736775,
"grad_norm": 0.8110173940658569,
"learning_rate": 0.0001342761193169402,
"loss": 0.2813,
"step": 622
},
{
"epoch": 0.397068196303378,
"grad_norm": 0.9587869048118591,
"learning_rate": 0.00013408552067342255,
"loss": 0.3634,
"step": 623
},
{
"epoch": 0.3977055449330784,
"grad_norm": 0.5752767324447632,
"learning_rate": 0.00013389478182387448,
"loss": 0.073,
"step": 624
},
{
"epoch": 0.39834289356277885,
"grad_norm": 1.452166199684143,
"learning_rate": 0.00013370390355287368,
"loss": 0.5348,
"step": 625
},
{
"epoch": 0.3989802421924793,
"grad_norm": 1.0857925415039062,
"learning_rate": 0.00013351288664557122,
"loss": 0.3476,
"step": 626
},
{
"epoch": 0.39961759082217974,
"grad_norm": 0.8397477269172668,
"learning_rate": 0.00013332173188768844,
"loss": 0.2725,
"step": 627
},
{
"epoch": 0.4002549394518802,
"grad_norm": 0.8080865740776062,
"learning_rate": 0.00013313044006551386,
"loss": 0.2189,
"step": 628
},
{
"epoch": 0.4008922880815806,
"grad_norm": 1.233829379081726,
"learning_rate": 0.00013293901196589957,
"loss": 0.3334,
"step": 629
},
{
"epoch": 0.40152963671128106,
"grad_norm": 1.0986390113830566,
"learning_rate": 0.0001327474483762584,
"loss": 0.5244,
"step": 630
},
{
"epoch": 0.4021669853409815,
"grad_norm": 0.9726091027259827,
"learning_rate": 0.00013255575008456043,
"loss": 0.2878,
"step": 631
},
{
"epoch": 0.40280433397068194,
"grad_norm": 0.6543111801147461,
"learning_rate": 0.0001323639178793298,
"loss": 0.1106,
"step": 632
},
{
"epoch": 0.40344168260038243,
"grad_norm": 0.8783746361732483,
"learning_rate": 0.00013217195254964148,
"loss": 0.2465,
"step": 633
},
{
"epoch": 0.4040790312300829,
"grad_norm": 1.016592264175415,
"learning_rate": 0.0001319798548851181,
"loss": 0.215,
"step": 634
},
{
"epoch": 0.4047163798597833,
"grad_norm": 0.13829180598258972,
"learning_rate": 0.0001317876256759265,
"loss": 0.0262,
"step": 635
},
{
"epoch": 0.40535372848948376,
"grad_norm": 1.0304619073867798,
"learning_rate": 0.00013159526571277484,
"loss": 0.3301,
"step": 636
},
{
"epoch": 0.4059910771191842,
"grad_norm": 0.9620900750160217,
"learning_rate": 0.00013140277578690883,
"loss": 0.1977,
"step": 637
},
{
"epoch": 0.40662842574888464,
"grad_norm": 0.9158233404159546,
"learning_rate": 0.00013121015669010898,
"loss": 0.1949,
"step": 638
},
{
"epoch": 0.4072657743785851,
"grad_norm": 0.6509270071983337,
"learning_rate": 0.00013101740921468702,
"loss": 0.1415,
"step": 639
},
{
"epoch": 0.4079031230082855,
"grad_norm": 0.8859730958938599,
"learning_rate": 0.0001308245341534828,
"loss": 0.3595,
"step": 640
},
{
"epoch": 0.40854047163798596,
"grad_norm": 0.8790940642356873,
"learning_rate": 0.00013063153229986096,
"loss": 0.1857,
"step": 641
},
{
"epoch": 0.4091778202676864,
"grad_norm": 0.7152032852172852,
"learning_rate": 0.00013043840444770764,
"loss": 0.1413,
"step": 642
},
{
"epoch": 0.4098151688973869,
"grad_norm": 0.6586755514144897,
"learning_rate": 0.00013024515139142733,
"loss": 0.2221,
"step": 643
},
{
"epoch": 0.41045251752708733,
"grad_norm": 1.3934643268585205,
"learning_rate": 0.0001300517739259395,
"loss": 0.3246,
"step": 644
},
{
"epoch": 0.4110898661567878,
"grad_norm": 1.247506856918335,
"learning_rate": 0.00012985827284667535,
"loss": 0.4791,
"step": 645
},
{
"epoch": 0.4117272147864882,
"grad_norm": 0.8062496781349182,
"learning_rate": 0.00012966464894957454,
"loss": 0.2348,
"step": 646
},
{
"epoch": 0.41236456341618866,
"grad_norm": 0.7484445571899414,
"learning_rate": 0.00012947090303108192,
"loss": 0.1685,
"step": 647
},
{
"epoch": 0.4130019120458891,
"grad_norm": 0.4937773048877716,
"learning_rate": 0.0001292770358881443,
"loss": 0.0591,
"step": 648
},
{
"epoch": 0.41363926067558954,
"grad_norm": 0.8465187549591064,
"learning_rate": 0.00012908304831820715,
"loss": 0.311,
"step": 649
},
{
"epoch": 0.41427660930529,
"grad_norm": 0.822917103767395,
"learning_rate": 0.00012888894111921117,
"loss": 0.2862,
"step": 650
},
{
"epoch": 0.4149139579349904,
"grad_norm": 1.4749966859817505,
"learning_rate": 0.00012869471508958927,
"loss": 0.3425,
"step": 651
},
{
"epoch": 0.41555130656469086,
"grad_norm": 0.5220008492469788,
"learning_rate": 0.0001285003710282631,
"loss": 0.0361,
"step": 652
},
{
"epoch": 0.41618865519439135,
"grad_norm": 1.2537230253219604,
"learning_rate": 0.00012830590973463986,
"loss": 0.4289,
"step": 653
},
{
"epoch": 0.4168260038240918,
"grad_norm": 0.7489885687828064,
"learning_rate": 0.00012811133200860885,
"loss": 0.2206,
"step": 654
},
{
"epoch": 0.41746335245379224,
"grad_norm": 0.5201739072799683,
"learning_rate": 0.00012791663865053846,
"loss": 0.0773,
"step": 655
},
{
"epoch": 0.4181007010834927,
"grad_norm": 0.8189405798912048,
"learning_rate": 0.00012772183046127258,
"loss": 0.1746,
"step": 656
},
{
"epoch": 0.4187380497131931,
"grad_norm": 1.1056138277053833,
"learning_rate": 0.0001275269082421275,
"loss": 0.2854,
"step": 657
},
{
"epoch": 0.41937539834289356,
"grad_norm": 1.0581618547439575,
"learning_rate": 0.00012733187279488859,
"loss": 0.3187,
"step": 658
},
{
"epoch": 0.420012746972594,
"grad_norm": 1.0861239433288574,
"learning_rate": 0.00012713672492180688,
"loss": 0.3675,
"step": 659
},
{
"epoch": 0.42065009560229444,
"grad_norm": 1.0314252376556396,
"learning_rate": 0.000126941465425596,
"loss": 0.2726,
"step": 660
},
{
"epoch": 0.4212874442319949,
"grad_norm": 0.8676585555076599,
"learning_rate": 0.0001267460951094285,
"loss": 0.3376,
"step": 661
},
{
"epoch": 0.4219247928616954,
"grad_norm": 0.7947667837142944,
"learning_rate": 0.000126550614776933,
"loss": 0.2963,
"step": 662
},
{
"epoch": 0.4225621414913958,
"grad_norm": 0.4266347587108612,
"learning_rate": 0.00012635502523219047,
"loss": 0.0582,
"step": 663
},
{
"epoch": 0.42319949012109626,
"grad_norm": 0.7492831945419312,
"learning_rate": 0.00012615932727973127,
"loss": 0.1781,
"step": 664
},
{
"epoch": 0.4238368387507967,
"grad_norm": 0.8613919019699097,
"learning_rate": 0.00012596352172453154,
"loss": 0.2941,
"step": 665
},
{
"epoch": 0.42447418738049714,
"grad_norm": 0.9851914644241333,
"learning_rate": 0.00012576760937201017,
"loss": 0.1585,
"step": 666
},
{
"epoch": 0.4251115360101976,
"grad_norm": 1.1824510097503662,
"learning_rate": 0.0001255715910280252,
"loss": 0.2881,
"step": 667
},
{
"epoch": 0.425748884639898,
"grad_norm": 0.7090943455696106,
"learning_rate": 0.00012537546749887078,
"loss": 0.2857,
"step": 668
},
{
"epoch": 0.42638623326959846,
"grad_norm": 1.1374778747558594,
"learning_rate": 0.0001251792395912736,
"loss": 0.4139,
"step": 669
},
{
"epoch": 0.4270235818992989,
"grad_norm": 0.7184798717498779,
"learning_rate": 0.0001249829081123898,
"loss": 0.1267,
"step": 670
},
{
"epoch": 0.42766093052899934,
"grad_norm": 0.3961735665798187,
"learning_rate": 0.0001247864738698015,
"loss": 0.05,
"step": 671
},
{
"epoch": 0.42829827915869984,
"grad_norm": 1.0486527681350708,
"learning_rate": 0.00012458993767151355,
"loss": 0.3901,
"step": 672
},
{
"epoch": 0.4289356277884003,
"grad_norm": 1.1963367462158203,
"learning_rate": 0.0001243933003259501,
"loss": 0.3241,
"step": 673
},
{
"epoch": 0.4295729764181007,
"grad_norm": 1.0635004043579102,
"learning_rate": 0.00012419656264195153,
"loss": 0.3195,
"step": 674
},
{
"epoch": 0.43021032504780116,
"grad_norm": 0.2920896112918854,
"learning_rate": 0.00012399972542877073,
"loss": 0.0203,
"step": 675
},
{
"epoch": 0.4308476736775016,
"grad_norm": 1.0419822931289673,
"learning_rate": 0.00012380278949607011,
"loss": 0.2975,
"step": 676
},
{
"epoch": 0.43148502230720204,
"grad_norm": 1.0537307262420654,
"learning_rate": 0.00012360575565391818,
"loss": 0.3471,
"step": 677
},
{
"epoch": 0.4321223709369025,
"grad_norm": 1.255276083946228,
"learning_rate": 0.00012340862471278612,
"loss": 0.194,
"step": 678
},
{
"epoch": 0.4327597195666029,
"grad_norm": 0.7135950326919556,
"learning_rate": 0.00012321139748354453,
"loss": 0.0743,
"step": 679
},
{
"epoch": 0.43339706819630336,
"grad_norm": 1.033130168914795,
"learning_rate": 0.0001230140747774601,
"loss": 0.3348,
"step": 680
},
{
"epoch": 0.4340344168260038,
"grad_norm": 1.0107977390289307,
"learning_rate": 0.00012281665740619216,
"loss": 0.2979,
"step": 681
},
{
"epoch": 0.4346717654557043,
"grad_norm": 0.8340908288955688,
"learning_rate": 0.00012261914618178957,
"loss": 0.3267,
"step": 682
},
{
"epoch": 0.43530911408540474,
"grad_norm": 0.5547581315040588,
"learning_rate": 0.0001224215419166872,
"loss": 0.1254,
"step": 683
},
{
"epoch": 0.4359464627151052,
"grad_norm": 0.8941380381584167,
"learning_rate": 0.00012222384542370253,
"loss": 0.197,
"step": 684
},
{
"epoch": 0.4365838113448056,
"grad_norm": 1.0526996850967407,
"learning_rate": 0.00012202605751603256,
"loss": 0.3715,
"step": 685
},
{
"epoch": 0.43722115997450606,
"grad_norm": 0.2321912944316864,
"learning_rate": 0.00012182817900725014,
"loss": 0.0142,
"step": 686
},
{
"epoch": 0.4378585086042065,
"grad_norm": 0.9421070218086243,
"learning_rate": 0.00012163021071130101,
"loss": 0.3526,
"step": 687
},
{
"epoch": 0.43849585723390694,
"grad_norm": 1.404322624206543,
"learning_rate": 0.00012143215344250001,
"loss": 0.262,
"step": 688
},
{
"epoch": 0.4391332058636074,
"grad_norm": 1.3302491903305054,
"learning_rate": 0.00012123400801552809,
"loss": 0.2989,
"step": 689
},
{
"epoch": 0.4397705544933078,
"grad_norm": 0.7755722999572754,
"learning_rate": 0.00012103577524542884,
"loss": 0.1756,
"step": 690
},
{
"epoch": 0.44040790312300826,
"grad_norm": 1.0167707204818726,
"learning_rate": 0.00012083745594760502,
"loss": 0.4299,
"step": 691
},
{
"epoch": 0.44104525175270876,
"grad_norm": 1.37527596950531,
"learning_rate": 0.00012063905093781544,
"loss": 0.3424,
"step": 692
},
{
"epoch": 0.4416826003824092,
"grad_norm": 0.8296977281570435,
"learning_rate": 0.00012044056103217138,
"loss": 0.2459,
"step": 693
},
{
"epoch": 0.44231994901210964,
"grad_norm": 0.887431263923645,
"learning_rate": 0.00012024198704713335,
"loss": 0.1254,
"step": 694
},
{
"epoch": 0.4429572976418101,
"grad_norm": 0.4647270739078522,
"learning_rate": 0.00012004332979950769,
"loss": 0.0226,
"step": 695
},
{
"epoch": 0.4435946462715105,
"grad_norm": 1.0587822198867798,
"learning_rate": 0.00011984459010644328,
"loss": 0.3479,
"step": 696
},
{
"epoch": 0.44423199490121096,
"grad_norm": 0.5554285645484924,
"learning_rate": 0.0001196457687854281,
"loss": 0.1074,
"step": 697
},
{
"epoch": 0.4448693435309114,
"grad_norm": 1.135880708694458,
"learning_rate": 0.00011944686665428593,
"loss": 0.348,
"step": 698
},
{
"epoch": 0.44550669216061184,
"grad_norm": 0.7486273646354675,
"learning_rate": 0.00011924788453117281,
"loss": 0.3408,
"step": 699
},
{
"epoch": 0.4461440407903123,
"grad_norm": 1.0744110345840454,
"learning_rate": 0.00011904882323457402,
"loss": 0.1309,
"step": 700
},
{
"epoch": 0.4467813894200127,
"grad_norm": 1.2191270589828491,
"learning_rate": 0.0001188496835833004,
"loss": 0.411,
"step": 701
},
{
"epoch": 0.4474187380497132,
"grad_norm": 0.9536232948303223,
"learning_rate": 0.00011865046639648506,
"loss": 0.2967,
"step": 702
},
{
"epoch": 0.44805608667941366,
"grad_norm": 0.612731397151947,
"learning_rate": 0.00011845117249358008,
"loss": 0.0502,
"step": 703
},
{
"epoch": 0.4486934353091141,
"grad_norm": 0.7568250894546509,
"learning_rate": 0.00011825180269435321,
"loss": 0.2956,
"step": 704
},
{
"epoch": 0.44933078393881454,
"grad_norm": 0.6128911375999451,
"learning_rate": 0.00011805235781888413,
"loss": 0.14,
"step": 705
},
{
"epoch": 0.449968132568515,
"grad_norm": 0.5269573926925659,
"learning_rate": 0.0001178528386875616,
"loss": 0.0322,
"step": 706
},
{
"epoch": 0.4506054811982154,
"grad_norm": 0.8908300399780273,
"learning_rate": 0.00011765324612107969,
"loss": 0.3312,
"step": 707
},
{
"epoch": 0.45124282982791586,
"grad_norm": 0.7696492671966553,
"learning_rate": 0.0001174535809404345,
"loss": 0.2417,
"step": 708
},
{
"epoch": 0.4518801784576163,
"grad_norm": 0.7858138084411621,
"learning_rate": 0.00011725384396692094,
"loss": 0.2774,
"step": 709
},
{
"epoch": 0.45251752708731674,
"grad_norm": 0.8619891405105591,
"learning_rate": 0.00011705403602212915,
"loss": 0.249,
"step": 710
},
{
"epoch": 0.45315487571701724,
"grad_norm": 0.7546841502189636,
"learning_rate": 0.0001168541579279412,
"loss": 0.116,
"step": 711
},
{
"epoch": 0.4537922243467177,
"grad_norm": 0.6840965747833252,
"learning_rate": 0.00011665421050652773,
"loss": 0.0325,
"step": 712
},
{
"epoch": 0.4544295729764181,
"grad_norm": 0.6608375906944275,
"learning_rate": 0.00011645419458034452,
"loss": 0.1757,
"step": 713
},
{
"epoch": 0.45506692160611856,
"grad_norm": 1.4142905473709106,
"learning_rate": 0.00011625411097212919,
"loss": 0.4896,
"step": 714
},
{
"epoch": 0.455704270235819,
"grad_norm": 0.8743811249732971,
"learning_rate": 0.00011605396050489771,
"loss": 0.2809,
"step": 715
},
{
"epoch": 0.45634161886551944,
"grad_norm": 1.1313313245773315,
"learning_rate": 0.00011585374400194111,
"loss": 0.1011,
"step": 716
},
{
"epoch": 0.4569789674952199,
"grad_norm": 1.1805555820465088,
"learning_rate": 0.000115653462286822,
"loss": 0.3602,
"step": 717
},
{
"epoch": 0.4576163161249203,
"grad_norm": 0.679884672164917,
"learning_rate": 0.0001154531161833712,
"loss": 0.1479,
"step": 718
},
{
"epoch": 0.45825366475462076,
"grad_norm": 0.5877489447593689,
"learning_rate": 0.00011525270651568457,
"loss": 0.1843,
"step": 719
},
{
"epoch": 0.4588910133843212,
"grad_norm": 0.9065275192260742,
"learning_rate": 0.00011505223410811914,
"loss": 0.2348,
"step": 720
},
{
"epoch": 0.4595283620140217,
"grad_norm": 1.0555546283721924,
"learning_rate": 0.00011485169978529026,
"loss": 0.1238,
"step": 721
},
{
"epoch": 0.46016571064372214,
"grad_norm": 0.8118847012519836,
"learning_rate": 0.00011465110437206787,
"loss": 0.3498,
"step": 722
},
{
"epoch": 0.4608030592734226,
"grad_norm": 0.8980100154876709,
"learning_rate": 0.0001144504486935732,
"loss": 0.1469,
"step": 723
},
{
"epoch": 0.461440407903123,
"grad_norm": 1.0857048034667969,
"learning_rate": 0.0001142497335751753,
"loss": 0.346,
"step": 724
},
{
"epoch": 0.46207775653282346,
"grad_norm": 0.6706693768501282,
"learning_rate": 0.00011404895984248787,
"loss": 0.1239,
"step": 725
},
{
"epoch": 0.4627151051625239,
"grad_norm": 1.0879980325698853,
"learning_rate": 0.00011384812832136554,
"loss": 0.3405,
"step": 726
},
{
"epoch": 0.46335245379222434,
"grad_norm": 0.24894720315933228,
"learning_rate": 0.00011364723983790085,
"loss": 0.0306,
"step": 727
},
{
"epoch": 0.4639898024219248,
"grad_norm": 1.1150575876235962,
"learning_rate": 0.00011344629521842043,
"loss": 0.3916,
"step": 728
},
{
"epoch": 0.4646271510516252,
"grad_norm": 1.1501226425170898,
"learning_rate": 0.00011324529528948193,
"loss": 0.2166,
"step": 729
},
{
"epoch": 0.46526449968132566,
"grad_norm": 0.47070667147636414,
"learning_rate": 0.00011304424087787056,
"loss": 0.1058,
"step": 730
},
{
"epoch": 0.46590184831102616,
"grad_norm": 0.7818524241447449,
"learning_rate": 0.00011284313281059553,
"loss": 0.2251,
"step": 731
},
{
"epoch": 0.4665391969407266,
"grad_norm": 1.16464364528656,
"learning_rate": 0.00011264197191488677,
"loss": 0.4106,
"step": 732
},
{
"epoch": 0.46717654557042704,
"grad_norm": 1.4244799613952637,
"learning_rate": 0.00011244075901819156,
"loss": 0.3916,
"step": 733
},
{
"epoch": 0.4678138942001275,
"grad_norm": 0.31768059730529785,
"learning_rate": 0.00011223949494817105,
"loss": 0.032,
"step": 734
},
{
"epoch": 0.4684512428298279,
"grad_norm": 0.9916287660598755,
"learning_rate": 0.00011203818053269692,
"loss": 0.3095,
"step": 735
},
{
"epoch": 0.46908859145952836,
"grad_norm": 1.0889006853103638,
"learning_rate": 0.00011183681659984793,
"loss": 0.1475,
"step": 736
},
{
"epoch": 0.4697259400892288,
"grad_norm": 0.9469870328903198,
"learning_rate": 0.0001116354039779064,
"loss": 0.2689,
"step": 737
},
{
"epoch": 0.47036328871892924,
"grad_norm": 0.8823866844177246,
"learning_rate": 0.00011143394349535517,
"loss": 0.1573,
"step": 738
},
{
"epoch": 0.4710006373486297,
"grad_norm": 0.9090831279754639,
"learning_rate": 0.0001112324359808737,
"loss": 0.235,
"step": 739
},
{
"epoch": 0.4716379859783301,
"grad_norm": 0.5955209136009216,
"learning_rate": 0.00011103088226333505,
"loss": 0.1461,
"step": 740
},
{
"epoch": 0.4722753346080306,
"grad_norm": 0.9746752381324768,
"learning_rate": 0.0001108292831718023,
"loss": 0.3041,
"step": 741
},
{
"epoch": 0.47291268323773106,
"grad_norm": 1.1742018461227417,
"learning_rate": 0.0001106276395355252,
"loss": 0.2913,
"step": 742
},
{
"epoch": 0.4735500318674315,
"grad_norm": 0.7824221253395081,
"learning_rate": 0.00011042595218393661,
"loss": 0.2339,
"step": 743
},
{
"epoch": 0.47418738049713194,
"grad_norm": 0.8328511118888855,
"learning_rate": 0.00011022422194664938,
"loss": 0.0809,
"step": 744
},
{
"epoch": 0.4748247291268324,
"grad_norm": 0.8351703882217407,
"learning_rate": 0.00011002244965345261,
"loss": 0.2407,
"step": 745
},
{
"epoch": 0.4754620777565328,
"grad_norm": 0.8491295576095581,
"learning_rate": 0.00010982063613430853,
"loss": 0.114,
"step": 746
},
{
"epoch": 0.47609942638623326,
"grad_norm": 1.4426089525222778,
"learning_rate": 0.0001096187822193488,
"loss": 0.5476,
"step": 747
},
{
"epoch": 0.4767367750159337,
"grad_norm": 0.7609091997146606,
"learning_rate": 0.00010941688873887137,
"loss": 0.2276,
"step": 748
},
{
"epoch": 0.47737412364563414,
"grad_norm": 1.0052447319030762,
"learning_rate": 0.00010921495652333684,
"loss": 0.2333,
"step": 749
},
{
"epoch": 0.4780114722753346,
"grad_norm": 0.5735613703727722,
"learning_rate": 0.00010901298640336526,
"loss": 0.0968,
"step": 750
},
{
"epoch": 0.4786488209050351,
"grad_norm": 1.0442891120910645,
"learning_rate": 0.00010881097920973242,
"loss": 0.3025,
"step": 751
},
{
"epoch": 0.4792861695347355,
"grad_norm": 1.3371227979660034,
"learning_rate": 0.00010860893577336677,
"loss": 0.4465,
"step": 752
},
{
"epoch": 0.47992351816443596,
"grad_norm": 0.9134352207183838,
"learning_rate": 0.00010840685692534575,
"loss": 0.2264,
"step": 753
},
{
"epoch": 0.4805608667941364,
"grad_norm": 1.2226171493530273,
"learning_rate": 0.00010820474349689253,
"loss": 0.4715,
"step": 754
},
{
"epoch": 0.48119821542383684,
"grad_norm": 1.1333131790161133,
"learning_rate": 0.00010800259631937246,
"loss": 0.4335,
"step": 755
},
{
"epoch": 0.4818355640535373,
"grad_norm": 1.3913917541503906,
"learning_rate": 0.00010780041622428967,
"loss": 0.4192,
"step": 756
},
{
"epoch": 0.4824729126832377,
"grad_norm": 0.9874146580696106,
"learning_rate": 0.00010759820404328386,
"loss": 0.1385,
"step": 757
},
{
"epoch": 0.48311026131293816,
"grad_norm": 0.8050394058227539,
"learning_rate": 0.00010739596060812653,
"loss": 0.3024,
"step": 758
},
{
"epoch": 0.4837476099426386,
"grad_norm": 1.3845244646072388,
"learning_rate": 0.00010719368675071783,
"loss": 0.3909,
"step": 759
},
{
"epoch": 0.4843849585723391,
"grad_norm": 0.6297693252563477,
"learning_rate": 0.00010699138330308305,
"loss": 0.1826,
"step": 760
},
{
"epoch": 0.48502230720203954,
"grad_norm": 0.8287470936775208,
"learning_rate": 0.00010678905109736914,
"loss": 0.1334,
"step": 761
},
{
"epoch": 0.48565965583174,
"grad_norm": 1.2835328578948975,
"learning_rate": 0.00010658669096584141,
"loss": 0.5079,
"step": 762
},
{
"epoch": 0.4862970044614404,
"grad_norm": 1.0681860446929932,
"learning_rate": 0.00010638430374088004,
"loss": 0.3174,
"step": 763
},
{
"epoch": 0.48693435309114086,
"grad_norm": 1.0765239000320435,
"learning_rate": 0.00010618189025497654,
"loss": 0.2913,
"step": 764
},
{
"epoch": 0.4875717017208413,
"grad_norm": 1.0276625156402588,
"learning_rate": 0.0001059794513407306,
"loss": 0.2761,
"step": 765
},
{
"epoch": 0.48820905035054174,
"grad_norm": 1.0593966245651245,
"learning_rate": 0.00010577698783084638,
"loss": 0.2988,
"step": 766
},
{
"epoch": 0.4888463989802422,
"grad_norm": 1.0131773948669434,
"learning_rate": 0.00010557450055812928,
"loss": 0.3252,
"step": 767
},
{
"epoch": 0.4894837476099426,
"grad_norm": 0.7672984600067139,
"learning_rate": 0.00010537199035548244,
"loss": 0.2886,
"step": 768
},
{
"epoch": 0.49012109623964306,
"grad_norm": 1.368405818939209,
"learning_rate": 0.00010516945805590326,
"loss": 0.4342,
"step": 769
},
{
"epoch": 0.49075844486934356,
"grad_norm": 0.912278950214386,
"learning_rate": 0.00010496690449248014,
"loss": 0.2477,
"step": 770
},
{
"epoch": 0.491395793499044,
"grad_norm": 0.22493362426757812,
"learning_rate": 0.00010476433049838886,
"loss": 0.027,
"step": 771
},
{
"epoch": 0.49203314212874444,
"grad_norm": 0.5740759372711182,
"learning_rate": 0.00010456173690688922,
"loss": 0.1142,
"step": 772
},
{
"epoch": 0.4926704907584449,
"grad_norm": 0.36011382937431335,
"learning_rate": 0.00010435912455132171,
"loss": 0.0727,
"step": 773
},
{
"epoch": 0.4933078393881453,
"grad_norm": 0.8948391079902649,
"learning_rate": 0.000104156494265104,
"loss": 0.3636,
"step": 774
},
{
"epoch": 0.49394518801784576,
"grad_norm": 0.9722794890403748,
"learning_rate": 0.00010395384688172738,
"loss": 0.3425,
"step": 775
},
{
"epoch": 0.4945825366475462,
"grad_norm": 0.6774589419364929,
"learning_rate": 0.00010375118323475368,
"loss": 0.1503,
"step": 776
},
{
"epoch": 0.49521988527724664,
"grad_norm": 1.1837798357009888,
"learning_rate": 0.00010354850415781145,
"loss": 0.2366,
"step": 777
},
{
"epoch": 0.4958572339069471,
"grad_norm": 0.37189412117004395,
"learning_rate": 0.00010334581048459282,
"loss": 0.0572,
"step": 778
},
{
"epoch": 0.4964945825366475,
"grad_norm": 0.7028501033782959,
"learning_rate": 0.00010314310304884986,
"loss": 0.168,
"step": 779
},
{
"epoch": 0.497131931166348,
"grad_norm": 0.49016547203063965,
"learning_rate": 0.00010294038268439136,
"loss": 0.09,
"step": 780
},
{
"epoch": 0.49776927979604846,
"grad_norm": 0.7516559958457947,
"learning_rate": 0.0001027376502250792,
"loss": 0.2503,
"step": 781
},
{
"epoch": 0.4984066284257489,
"grad_norm": 1.2607611417770386,
"learning_rate": 0.0001025349065048251,
"loss": 0.3195,
"step": 782
},
{
"epoch": 0.49904397705544934,
"grad_norm": 1.0620837211608887,
"learning_rate": 0.00010233215235758692,
"loss": 0.3199,
"step": 783
},
{
"epoch": 0.4996813256851498,
"grad_norm": 0.8697445392608643,
"learning_rate": 0.00010212938861736572,
"loss": 0.248,
"step": 784
},
{
"epoch": 0.5003186743148502,
"grad_norm": 0.7616758942604065,
"learning_rate": 0.00010192661611820169,
"loss": 0.191,
"step": 785
},
{
"epoch": 0.5009560229445507,
"grad_norm": 0.6131820678710938,
"learning_rate": 0.00010172383569417123,
"loss": 0.2138,
"step": 786
},
{
"epoch": 0.5009560229445507,
"eval_loss": 0.2808364927768707,
"eval_runtime": 6.6134,
"eval_samples_per_second": 9.98,
"eval_steps_per_second": 4.99,
"step": 786
},
{
"epoch": 0.5015933715742511,
"grad_norm": 0.4716302752494812,
"learning_rate": 0.00010152104817938335,
"loss": 0.0558,
"step": 787
},
{
"epoch": 0.5022307202039515,
"grad_norm": 0.5269849300384521,
"learning_rate": 0.00010131825440797613,
"loss": 0.1079,
"step": 788
},
{
"epoch": 0.502868068833652,
"grad_norm": 0.9003429412841797,
"learning_rate": 0.00010111545521411347,
"loss": 0.4007,
"step": 789
},
{
"epoch": 0.5035054174633524,
"grad_norm": 1.0929003953933716,
"learning_rate": 0.00010091265143198152,
"loss": 0.3282,
"step": 790
},
{
"epoch": 0.5041427660930529,
"grad_norm": 0.9409645199775696,
"learning_rate": 0.00010070984389578531,
"loss": 0.3488,
"step": 791
},
{
"epoch": 0.5047801147227533,
"grad_norm": 0.960472822189331,
"learning_rate": 0.00010050703343974535,
"loss": 0.2188,
"step": 792
},
{
"epoch": 0.5054174633524537,
"grad_norm": 0.27374258637428284,
"learning_rate": 0.00010030422089809415,
"loss": 0.0194,
"step": 793
},
{
"epoch": 0.5060548119821542,
"grad_norm": 0.8933336734771729,
"learning_rate": 0.0001001014071050727,
"loss": 0.3983,
"step": 794
},
{
"epoch": 0.5066921606118547,
"grad_norm": 0.7880145311355591,
"learning_rate": 9.989859289492732e-05,
"loss": 0.2839,
"step": 795
},
{
"epoch": 0.5073295092415552,
"grad_norm": 1.562007188796997,
"learning_rate": 9.96957791019059e-05,
"loss": 0.2828,
"step": 796
},
{
"epoch": 0.5079668578712556,
"grad_norm": 0.5563234090805054,
"learning_rate": 9.94929665602547e-05,
"loss": 0.0784,
"step": 797
},
{
"epoch": 0.5086042065009561,
"grad_norm": 0.889250636100769,
"learning_rate": 9.92901561042147e-05,
"loss": 0.255,
"step": 798
},
{
"epoch": 0.5092415551306565,
"grad_norm": 0.8856499195098877,
"learning_rate": 9.90873485680185e-05,
"loss": 0.2629,
"step": 799
},
{
"epoch": 0.5098789037603569,
"grad_norm": 0.7302645444869995,
"learning_rate": 9.888454478588655e-05,
"loss": 0.1107,
"step": 800
},
{
"epoch": 0.5105162523900574,
"grad_norm": 1.236111044883728,
"learning_rate": 9.868174559202388e-05,
"loss": 0.3373,
"step": 801
},
{
"epoch": 0.5111536010197578,
"grad_norm": 0.9791870713233948,
"learning_rate": 9.847895182061667e-05,
"loss": 0.2283,
"step": 802
},
{
"epoch": 0.5117909496494583,
"grad_norm": 1.0384395122528076,
"learning_rate": 9.827616430582875e-05,
"loss": 0.4065,
"step": 803
},
{
"epoch": 0.5124282982791587,
"grad_norm": 0.6283885836601257,
"learning_rate": 9.807338388179835e-05,
"loss": 0.1143,
"step": 804
},
{
"epoch": 0.5130656469088591,
"grad_norm": 0.9107471704483032,
"learning_rate": 9.787061138263431e-05,
"loss": 0.3071,
"step": 805
},
{
"epoch": 0.5137029955385596,
"grad_norm": 0.5751721858978271,
"learning_rate": 9.766784764241309e-05,
"loss": 0.0446,
"step": 806
},
{
"epoch": 0.51434034416826,
"grad_norm": 0.7414814829826355,
"learning_rate": 9.746509349517496e-05,
"loss": 0.2143,
"step": 807
},
{
"epoch": 0.5149776927979605,
"grad_norm": 0.6147415041923523,
"learning_rate": 9.72623497749208e-05,
"loss": 0.0822,
"step": 808
},
{
"epoch": 0.5156150414276609,
"grad_norm": 1.3089789152145386,
"learning_rate": 9.705961731560864e-05,
"loss": 0.434,
"step": 809
},
{
"epoch": 0.5162523900573613,
"grad_norm": 0.5168224573135376,
"learning_rate": 9.685689695115015e-05,
"loss": 0.105,
"step": 810
},
{
"epoch": 0.5168897386870618,
"grad_norm": 0.3691941797733307,
"learning_rate": 9.66541895154072e-05,
"loss": 0.0463,
"step": 811
},
{
"epoch": 0.5175270873167622,
"grad_norm": 0.8521266579627991,
"learning_rate": 9.645149584218857e-05,
"loss": 0.3283,
"step": 812
},
{
"epoch": 0.5181644359464627,
"grad_norm": 0.9181352853775024,
"learning_rate": 9.624881676524636e-05,
"loss": 0.2919,
"step": 813
},
{
"epoch": 0.5188017845761632,
"grad_norm": 0.9536579251289368,
"learning_rate": 9.604615311827261e-05,
"loss": 0.2009,
"step": 814
},
{
"epoch": 0.5194391332058637,
"grad_norm": 1.092387318611145,
"learning_rate": 9.584350573489603e-05,
"loss": 0.2536,
"step": 815
},
{
"epoch": 0.5200764818355641,
"grad_norm": 0.6718846559524536,
"learning_rate": 9.564087544867831e-05,
"loss": 0.1731,
"step": 816
},
{
"epoch": 0.5207138304652645,
"grad_norm": 1.2321211099624634,
"learning_rate": 9.54382630931108e-05,
"loss": 0.2836,
"step": 817
},
{
"epoch": 0.521351179094965,
"grad_norm": 1.0353859663009644,
"learning_rate": 9.523566950161119e-05,
"loss": 0.4172,
"step": 818
},
{
"epoch": 0.5219885277246654,
"grad_norm": 0.8059009313583374,
"learning_rate": 9.503309550751987e-05,
"loss": 0.1833,
"step": 819
},
{
"epoch": 0.5226258763543659,
"grad_norm": 0.8674271702766418,
"learning_rate": 9.483054194409673e-05,
"loss": 0.3013,
"step": 820
},
{
"epoch": 0.5232632249840663,
"grad_norm": 0.7569763660430908,
"learning_rate": 9.46280096445176e-05,
"loss": 0.1148,
"step": 821
},
{
"epoch": 0.5239005736137667,
"grad_norm": 1.4185737371444702,
"learning_rate": 9.442549944187073e-05,
"loss": 0.4481,
"step": 822
},
{
"epoch": 0.5245379222434672,
"grad_norm": 0.5064479112625122,
"learning_rate": 9.422301216915363e-05,
"loss": 0.0859,
"step": 823
},
{
"epoch": 0.5251752708731676,
"grad_norm": 0.8803994059562683,
"learning_rate": 9.402054865926944e-05,
"loss": 0.2393,
"step": 824
},
{
"epoch": 0.5258126195028681,
"grad_norm": 1.4073066711425781,
"learning_rate": 9.381810974502347e-05,
"loss": 0.4285,
"step": 825
},
{
"epoch": 0.5264499681325685,
"grad_norm": 1.291663646697998,
"learning_rate": 9.361569625911998e-05,
"loss": 0.1984,
"step": 826
},
{
"epoch": 0.5270873167622689,
"grad_norm": 1.0464991331100464,
"learning_rate": 9.34133090341586e-05,
"loss": 0.3178,
"step": 827
},
{
"epoch": 0.5277246653919694,
"grad_norm": 0.4726356565952301,
"learning_rate": 9.321094890263087e-05,
"loss": 0.0564,
"step": 828
},
{
"epoch": 0.5283620140216698,
"grad_norm": 0.781101644039154,
"learning_rate": 9.300861669691699e-05,
"loss": 0.1341,
"step": 829
},
{
"epoch": 0.5289993626513703,
"grad_norm": 0.665503203868866,
"learning_rate": 9.280631324928216e-05,
"loss": 0.1025,
"step": 830
},
{
"epoch": 0.5296367112810707,
"grad_norm": 1.0607640743255615,
"learning_rate": 9.260403939187348e-05,
"loss": 0.3632,
"step": 831
},
{
"epoch": 0.5302740599107711,
"grad_norm": 1.196071982383728,
"learning_rate": 9.240179595671616e-05,
"loss": 0.3609,
"step": 832
},
{
"epoch": 0.5309114085404716,
"grad_norm": 0.9817525148391724,
"learning_rate": 9.219958377571034e-05,
"loss": 0.2427,
"step": 833
},
{
"epoch": 0.5315487571701721,
"grad_norm": 0.8917176127433777,
"learning_rate": 9.199740368062757e-05,
"loss": 0.1727,
"step": 834
},
{
"epoch": 0.5321861057998726,
"grad_norm": 0.38516029715538025,
"learning_rate": 9.179525650310748e-05,
"loss": 0.0249,
"step": 835
},
{
"epoch": 0.532823454429573,
"grad_norm": 1.2819154262542725,
"learning_rate": 9.159314307465426e-05,
"loss": 0.4615,
"step": 836
},
{
"epoch": 0.5334608030592735,
"grad_norm": 0.6406480073928833,
"learning_rate": 9.139106422663325e-05,
"loss": 0.1394,
"step": 837
},
{
"epoch": 0.5340981516889739,
"grad_norm": 1.3165202140808105,
"learning_rate": 9.11890207902676e-05,
"loss": 0.4627,
"step": 838
},
{
"epoch": 0.5347355003186743,
"grad_norm": 0.7313929796218872,
"learning_rate": 9.098701359663479e-05,
"loss": 0.1994,
"step": 839
},
{
"epoch": 0.5353728489483748,
"grad_norm": 0.7527788877487183,
"learning_rate": 9.078504347666317e-05,
"loss": 0.1868,
"step": 840
},
{
"epoch": 0.5360101975780752,
"grad_norm": 0.908225953578949,
"learning_rate": 9.058311126112864e-05,
"loss": 0.2091,
"step": 841
},
{
"epoch": 0.5366475462077757,
"grad_norm": 0.677261233329773,
"learning_rate": 9.038121778065122e-05,
"loss": 0.1745,
"step": 842
},
{
"epoch": 0.5372848948374761,
"grad_norm": 0.8489937782287598,
"learning_rate": 9.017936386569149e-05,
"loss": 0.2289,
"step": 843
},
{
"epoch": 0.5379222434671765,
"grad_norm": 1.0118935108184814,
"learning_rate": 8.997755034654741e-05,
"loss": 0.1614,
"step": 844
},
{
"epoch": 0.538559592096877,
"grad_norm": 1.2849200963974,
"learning_rate": 8.977577805335067e-05,
"loss": 0.3285,
"step": 845
},
{
"epoch": 0.5391969407265774,
"grad_norm": 1.1920063495635986,
"learning_rate": 8.95740478160634e-05,
"loss": 0.4803,
"step": 846
},
{
"epoch": 0.5398342893562779,
"grad_norm": 0.8660503029823303,
"learning_rate": 8.937236046447484e-05,
"loss": 0.2482,
"step": 847
},
{
"epoch": 0.5404716379859783,
"grad_norm": 1.1800625324249268,
"learning_rate": 8.917071682819773e-05,
"loss": 0.1665,
"step": 848
},
{
"epoch": 0.5411089866156787,
"grad_norm": 1.0726137161254883,
"learning_rate": 8.896911773666498e-05,
"loss": 0.2917,
"step": 849
},
{
"epoch": 0.5417463352453792,
"grad_norm": 0.9762551188468933,
"learning_rate": 8.876756401912636e-05,
"loss": 0.2669,
"step": 850
},
{
"epoch": 0.5423836838750796,
"grad_norm": 0.7531813979148865,
"learning_rate": 8.856605650464487e-05,
"loss": 0.2345,
"step": 851
},
{
"epoch": 0.5430210325047801,
"grad_norm": 0.9927303194999695,
"learning_rate": 8.836459602209359e-05,
"loss": 0.2542,
"step": 852
},
{
"epoch": 0.5436583811344805,
"grad_norm": 1.3853241205215454,
"learning_rate": 8.81631834001521e-05,
"loss": 0.2715,
"step": 853
},
{
"epoch": 0.5442957297641811,
"grad_norm": 1.5022135972976685,
"learning_rate": 8.796181946730309e-05,
"loss": 0.5391,
"step": 854
},
{
"epoch": 0.5449330783938815,
"grad_norm": 0.5603840947151184,
"learning_rate": 8.776050505182896e-05,
"loss": 0.0779,
"step": 855
},
{
"epoch": 0.5455704270235819,
"grad_norm": 0.9402315616607666,
"learning_rate": 8.755924098180849e-05,
"loss": 0.3259,
"step": 856
},
{
"epoch": 0.5462077756532824,
"grad_norm": 0.9817129969596863,
"learning_rate": 8.735802808511327e-05,
"loss": 0.3746,
"step": 857
},
{
"epoch": 0.5468451242829828,
"grad_norm": 0.7778319120407104,
"learning_rate": 8.71568671894045e-05,
"loss": 0.2461,
"step": 858
},
{
"epoch": 0.5474824729126833,
"grad_norm": 0.876461386680603,
"learning_rate": 8.695575912212947e-05,
"loss": 0.2553,
"step": 859
},
{
"epoch": 0.5481198215423837,
"grad_norm": 0.7117061018943787,
"learning_rate": 8.675470471051808e-05,
"loss": 0.1585,
"step": 860
},
{
"epoch": 0.5487571701720841,
"grad_norm": 0.9419434070587158,
"learning_rate": 8.655370478157961e-05,
"loss": 0.2627,
"step": 861
},
{
"epoch": 0.5493945188017846,
"grad_norm": 0.9806506037712097,
"learning_rate": 8.635276016209915e-05,
"loss": 0.4395,
"step": 862
},
{
"epoch": 0.550031867431485,
"grad_norm": 0.6108279228210449,
"learning_rate": 8.615187167863446e-05,
"loss": 0.1546,
"step": 863
},
{
"epoch": 0.5506692160611855,
"grad_norm": 0.8049284815788269,
"learning_rate": 8.595104015751217e-05,
"loss": 0.1954,
"step": 864
},
{
"epoch": 0.5513065646908859,
"grad_norm": 0.7539802193641663,
"learning_rate": 8.575026642482473e-05,
"loss": 0.1443,
"step": 865
},
{
"epoch": 0.5519439133205863,
"grad_norm": 0.8883464336395264,
"learning_rate": 8.554955130642685e-05,
"loss": 0.1995,
"step": 866
},
{
"epoch": 0.5525812619502868,
"grad_norm": 1.1132943630218506,
"learning_rate": 8.534889562793217e-05,
"loss": 0.3204,
"step": 867
},
{
"epoch": 0.5532186105799872,
"grad_norm": 1.3352077007293701,
"learning_rate": 8.514830021470972e-05,
"loss": 0.3106,
"step": 868
},
{
"epoch": 0.5538559592096877,
"grad_norm": 0.6655024886131287,
"learning_rate": 8.494776589188087e-05,
"loss": 0.1069,
"step": 869
},
{
"epoch": 0.5544933078393881,
"grad_norm": 1.2918314933776855,
"learning_rate": 8.474729348431547e-05,
"loss": 0.3116,
"step": 870
},
{
"epoch": 0.5551306564690885,
"grad_norm": 0.35464590787887573,
"learning_rate": 8.454688381662881e-05,
"loss": 0.0492,
"step": 871
},
{
"epoch": 0.555768005098789,
"grad_norm": 1.0541599988937378,
"learning_rate": 8.434653771317804e-05,
"loss": 0.3152,
"step": 872
},
{
"epoch": 0.5564053537284895,
"grad_norm": 0.9562996029853821,
"learning_rate": 8.41462559980589e-05,
"loss": 0.3294,
"step": 873
},
{
"epoch": 0.55704270235819,
"grad_norm": 0.5530235171318054,
"learning_rate": 8.39460394951023e-05,
"loss": 0.0721,
"step": 874
},
{
"epoch": 0.5576800509878904,
"grad_norm": 0.5161815285682678,
"learning_rate": 8.374588902787082e-05,
"loss": 0.0665,
"step": 875
},
{
"epoch": 0.5583173996175909,
"grad_norm": 1.1355836391448975,
"learning_rate": 8.354580541965549e-05,
"loss": 0.2279,
"step": 876
},
{
"epoch": 0.5589547482472913,
"grad_norm": 0.9556757807731628,
"learning_rate": 8.334578949347232e-05,
"loss": 0.2751,
"step": 877
},
{
"epoch": 0.5595920968769917,
"grad_norm": 0.6779599189758301,
"learning_rate": 8.314584207205881e-05,
"loss": 0.1614,
"step": 878
},
{
"epoch": 0.5602294455066922,
"grad_norm": 0.6609731912612915,
"learning_rate": 8.294596397787086e-05,
"loss": 0.1137,
"step": 879
},
{
"epoch": 0.5608667941363926,
"grad_norm": 1.3658030033111572,
"learning_rate": 8.274615603307907e-05,
"loss": 0.374,
"step": 880
},
{
"epoch": 0.5615041427660931,
"grad_norm": 1.063551664352417,
"learning_rate": 8.254641905956551e-05,
"loss": 0.1769,
"step": 881
},
{
"epoch": 0.5621414913957935,
"grad_norm": 0.8690844178199768,
"learning_rate": 8.234675387892036e-05,
"loss": 0.1612,
"step": 882
},
{
"epoch": 0.5627788400254939,
"grad_norm": 0.4268827438354492,
"learning_rate": 8.214716131243842e-05,
"loss": 0.0395,
"step": 883
},
{
"epoch": 0.5634161886551944,
"grad_norm": 0.9573777318000793,
"learning_rate": 8.194764218111587e-05,
"loss": 0.1268,
"step": 884
},
{
"epoch": 0.5640535372848948,
"grad_norm": 0.6962541341781616,
"learning_rate": 8.174819730564683e-05,
"loss": 0.1325,
"step": 885
},
{
"epoch": 0.5646908859145953,
"grad_norm": 1.0532737970352173,
"learning_rate": 8.154882750641993e-05,
"loss": 0.3171,
"step": 886
},
{
"epoch": 0.5653282345442957,
"grad_norm": 1.311773657798767,
"learning_rate": 8.134953360351497e-05,
"loss": 0.1225,
"step": 887
},
{
"epoch": 0.5659655831739961,
"grad_norm": 0.5788532495498657,
"learning_rate": 8.115031641669965e-05,
"loss": 0.0666,
"step": 888
},
{
"epoch": 0.5666029318036966,
"grad_norm": 1.2444723844528198,
"learning_rate": 8.095117676542599e-05,
"loss": 0.3604,
"step": 889
},
{
"epoch": 0.567240280433397,
"grad_norm": 1.0503877401351929,
"learning_rate": 8.07521154688272e-05,
"loss": 0.2381,
"step": 890
},
{
"epoch": 0.5678776290630975,
"grad_norm": 0.9579775333404541,
"learning_rate": 8.055313334571411e-05,
"loss": 0.4501,
"step": 891
},
{
"epoch": 0.5685149776927979,
"grad_norm": 1.0460937023162842,
"learning_rate": 8.035423121457191e-05,
"loss": 0.2139,
"step": 892
},
{
"epoch": 0.5691523263224985,
"grad_norm": 1.4753398895263672,
"learning_rate": 8.015540989355674e-05,
"loss": 0.454,
"step": 893
},
{
"epoch": 0.5697896749521989,
"grad_norm": 0.6733602285385132,
"learning_rate": 7.995667020049236e-05,
"loss": 0.1095,
"step": 894
},
{
"epoch": 0.5704270235818993,
"grad_norm": 0.7593318819999695,
"learning_rate": 7.975801295286668e-05,
"loss": 0.1321,
"step": 895
},
{
"epoch": 0.5710643722115998,
"grad_norm": 1.0922800302505493,
"learning_rate": 7.955943896782864e-05,
"loss": 0.2866,
"step": 896
},
{
"epoch": 0.5717017208413002,
"grad_norm": 0.6763241291046143,
"learning_rate": 7.936094906218458e-05,
"loss": 0.1905,
"step": 897
},
{
"epoch": 0.5723390694710007,
"grad_norm": 0.6475235223770142,
"learning_rate": 7.9162544052395e-05,
"loss": 0.1879,
"step": 898
},
{
"epoch": 0.5729764181007011,
"grad_norm": 0.7937578558921814,
"learning_rate": 7.89642247545712e-05,
"loss": 0.2387,
"step": 899
},
{
"epoch": 0.5736137667304015,
"grad_norm": 0.885387122631073,
"learning_rate": 7.876599198447191e-05,
"loss": 0.2103,
"step": 900
},
{
"epoch": 0.574251115360102,
"grad_norm": 0.7330776453018188,
"learning_rate": 7.856784655750003e-05,
"loss": 0.1775,
"step": 901
},
{
"epoch": 0.5748884639898024,
"grad_norm": 0.5146239995956421,
"learning_rate": 7.836978928869902e-05,
"loss": 0.0607,
"step": 902
},
{
"epoch": 0.5755258126195029,
"grad_norm": 0.4523480236530304,
"learning_rate": 7.817182099274988e-05,
"loss": 0.0458,
"step": 903
},
{
"epoch": 0.5761631612492033,
"grad_norm": 0.8467149138450623,
"learning_rate": 7.797394248396749e-05,
"loss": 0.0978,
"step": 904
},
{
"epoch": 0.5768005098789037,
"grad_norm": 0.8119767904281616,
"learning_rate": 7.777615457629746e-05,
"loss": 0.2048,
"step": 905
},
{
"epoch": 0.5774378585086042,
"grad_norm": 1.4051626920700073,
"learning_rate": 7.757845808331281e-05,
"loss": 0.4502,
"step": 906
},
{
"epoch": 0.5780752071383046,
"grad_norm": 0.9835460782051086,
"learning_rate": 7.738085381821044e-05,
"loss": 0.1733,
"step": 907
},
{
"epoch": 0.5787125557680051,
"grad_norm": 1.0887149572372437,
"learning_rate": 7.718334259380787e-05,
"loss": 0.2749,
"step": 908
},
{
"epoch": 0.5793499043977055,
"grad_norm": 0.9593770503997803,
"learning_rate": 7.698592522253996e-05,
"loss": 0.2576,
"step": 909
},
{
"epoch": 0.579987253027406,
"grad_norm": 0.8294894099235535,
"learning_rate": 7.678860251645551e-05,
"loss": 0.2141,
"step": 910
},
{
"epoch": 0.5806246016571064,
"grad_norm": 0.9683666229248047,
"learning_rate": 7.659137528721389e-05,
"loss": 0.2942,
"step": 911
},
{
"epoch": 0.5812619502868069,
"grad_norm": 0.9343656301498413,
"learning_rate": 7.639424434608185e-05,
"loss": 0.2191,
"step": 912
},
{
"epoch": 0.5818992989165074,
"grad_norm": 0.924299955368042,
"learning_rate": 7.619721050392991e-05,
"loss": 0.1861,
"step": 913
},
{
"epoch": 0.5825366475462078,
"grad_norm": 0.5737289190292358,
"learning_rate": 7.60002745712293e-05,
"loss": 0.0935,
"step": 914
},
{
"epoch": 0.5831739961759083,
"grad_norm": 1.5057976245880127,
"learning_rate": 7.580343735804852e-05,
"loss": 0.5175,
"step": 915
},
{
"epoch": 0.5838113448056087,
"grad_norm": 1.3809232711791992,
"learning_rate": 7.560669967404988e-05,
"loss": 0.444,
"step": 916
},
{
"epoch": 0.5844486934353091,
"grad_norm": 1.0442137718200684,
"learning_rate": 7.541006232848646e-05,
"loss": 0.3629,
"step": 917
},
{
"epoch": 0.5850860420650096,
"grad_norm": 1.2537975311279297,
"learning_rate": 7.521352613019852e-05,
"loss": 0.4307,
"step": 918
},
{
"epoch": 0.58572339069471,
"grad_norm": 1.245901107788086,
"learning_rate": 7.501709188761022e-05,
"loss": 0.2934,
"step": 919
},
{
"epoch": 0.5863607393244105,
"grad_norm": 0.7782462239265442,
"learning_rate": 7.482076040872646e-05,
"loss": 0.1935,
"step": 920
},
{
"epoch": 0.5869980879541109,
"grad_norm": 0.8486935496330261,
"learning_rate": 7.462453250112923e-05,
"loss": 0.1295,
"step": 921
},
{
"epoch": 0.5876354365838113,
"grad_norm": 1.010790467262268,
"learning_rate": 7.44284089719748e-05,
"loss": 0.149,
"step": 922
},
{
"epoch": 0.5882727852135118,
"grad_norm": 1.242264747619629,
"learning_rate": 7.423239062798984e-05,
"loss": 0.2523,
"step": 923
},
{
"epoch": 0.5889101338432122,
"grad_norm": 1.256464958190918,
"learning_rate": 7.403647827546847e-05,
"loss": 0.3664,
"step": 924
},
{
"epoch": 0.5895474824729127,
"grad_norm": 1.3096874952316284,
"learning_rate": 7.384067272026876e-05,
"loss": 0.4482,
"step": 925
},
{
"epoch": 0.5901848311026131,
"grad_norm": 0.6922174096107483,
"learning_rate": 7.364497476780958e-05,
"loss": 0.205,
"step": 926
},
{
"epoch": 0.5908221797323135,
"grad_norm": 0.8708852529525757,
"learning_rate": 7.344938522306704e-05,
"loss": 0.2199,
"step": 927
},
{
"epoch": 0.591459528362014,
"grad_norm": 0.8645421862602234,
"learning_rate": 7.325390489057152e-05,
"loss": 0.2152,
"step": 928
},
{
"epoch": 0.5920968769917144,
"grad_norm": 0.8442782163619995,
"learning_rate": 7.305853457440403e-05,
"loss": 0.2195,
"step": 929
},
{
"epoch": 0.5927342256214149,
"grad_norm": 0.7548922300338745,
"learning_rate": 7.286327507819313e-05,
"loss": 0.147,
"step": 930
},
{
"epoch": 0.5933715742511153,
"grad_norm": 1.1399472951889038,
"learning_rate": 7.266812720511145e-05,
"loss": 0.3507,
"step": 931
},
{
"epoch": 0.5940089228808159,
"grad_norm": 0.7924764752388,
"learning_rate": 7.247309175787251e-05,
"loss": 0.3002,
"step": 932
},
{
"epoch": 0.5946462715105163,
"grad_norm": 1.5446228981018066,
"learning_rate": 7.227816953872746e-05,
"loss": 0.412,
"step": 933
},
{
"epoch": 0.5952836201402167,
"grad_norm": 0.7069314122200012,
"learning_rate": 7.208336134946156e-05,
"loss": 0.1812,
"step": 934
},
{
"epoch": 0.5959209687699172,
"grad_norm": 0.4605756103992462,
"learning_rate": 7.188866799139118e-05,
"loss": 0.0866,
"step": 935
},
{
"epoch": 0.5965583173996176,
"grad_norm": 0.9317154884338379,
"learning_rate": 7.169409026536016e-05,
"loss": 0.2398,
"step": 936
},
{
"epoch": 0.5971956660293181,
"grad_norm": 0.7285643815994263,
"learning_rate": 7.14996289717369e-05,
"loss": 0.2552,
"step": 937
},
{
"epoch": 0.5978330146590185,
"grad_norm": 0.7581962943077087,
"learning_rate": 7.13052849104107e-05,
"loss": 0.1526,
"step": 938
},
{
"epoch": 0.5984703632887189,
"grad_norm": 1.4109922647476196,
"learning_rate": 7.111105888078884e-05,
"loss": 0.4126,
"step": 939
},
{
"epoch": 0.5991077119184194,
"grad_norm": 0.9118402004241943,
"learning_rate": 7.091695168179287e-05,
"loss": 0.2461,
"step": 940
},
{
"epoch": 0.5997450605481198,
"grad_norm": 0.9958475232124329,
"learning_rate": 7.07229641118557e-05,
"loss": 0.3221,
"step": 941
},
{
"epoch": 0.6003824091778203,
"grad_norm": 0.9631773233413696,
"learning_rate": 7.052909696891811e-05,
"loss": 0.2996,
"step": 942
},
{
"epoch": 0.6010197578075207,
"grad_norm": 0.9885851740837097,
"learning_rate": 7.033535105042547e-05,
"loss": 0.3568,
"step": 943
},
{
"epoch": 0.6016571064372211,
"grad_norm": 0.521555483341217,
"learning_rate": 7.014172715332466e-05,
"loss": 0.0605,
"step": 944
},
{
"epoch": 0.6022944550669216,
"grad_norm": 0.11074019968509674,
"learning_rate": 6.99482260740605e-05,
"loss": 0.0171,
"step": 945
},
{
"epoch": 0.602931803696622,
"grad_norm": 1.1028550863265991,
"learning_rate": 6.975484860857268e-05,
"loss": 0.3904,
"step": 946
},
{
"epoch": 0.6035691523263225,
"grad_norm": 1.007546067237854,
"learning_rate": 6.956159555229238e-05,
"loss": 0.3293,
"step": 947
},
{
"epoch": 0.6042065009560229,
"grad_norm": 0.6614052057266235,
"learning_rate": 6.936846770013905e-05,
"loss": 0.2016,
"step": 948
},
{
"epoch": 0.6048438495857233,
"grad_norm": 0.7158271074295044,
"learning_rate": 6.917546584651719e-05,
"loss": 0.1147,
"step": 949
},
{
"epoch": 0.6054811982154238,
"grad_norm": 0.6712331771850586,
"learning_rate": 6.898259078531299e-05,
"loss": 0.135,
"step": 950
},
{
"epoch": 0.6061185468451242,
"grad_norm": 1.1118035316467285,
"learning_rate": 6.878984330989104e-05,
"loss": 0.3839,
"step": 951
},
{
"epoch": 0.6067558954748248,
"grad_norm": 1.0008307695388794,
"learning_rate": 6.859722421309119e-05,
"loss": 0.2777,
"step": 952
},
{
"epoch": 0.6073932441045252,
"grad_norm": 0.2503453493118286,
"learning_rate": 6.84047342872252e-05,
"loss": 0.0227,
"step": 953
},
{
"epoch": 0.6080305927342257,
"grad_norm": 1.1753374338150024,
"learning_rate": 6.821237432407348e-05,
"loss": 0.2496,
"step": 954
},
{
"epoch": 0.6086679413639261,
"grad_norm": 0.8431404829025269,
"learning_rate": 6.802014511488193e-05,
"loss": 0.2659,
"step": 955
},
{
"epoch": 0.6093052899936265,
"grad_norm": 0.8169496655464172,
"learning_rate": 6.782804745035856e-05,
"loss": 0.1755,
"step": 956
},
{
"epoch": 0.609942638623327,
"grad_norm": 0.9566336274147034,
"learning_rate": 6.763608212067024e-05,
"loss": 0.1891,
"step": 957
},
{
"epoch": 0.6105799872530274,
"grad_norm": 1.0177518129348755,
"learning_rate": 6.744424991543962e-05,
"loss": 0.0882,
"step": 958
},
{
"epoch": 0.6112173358827279,
"grad_norm": 0.8488861322402954,
"learning_rate": 6.725255162374158e-05,
"loss": 0.3185,
"step": 959
},
{
"epoch": 0.6118546845124283,
"grad_norm": 0.6682339310646057,
"learning_rate": 6.706098803410043e-05,
"loss": 0.1464,
"step": 960
},
{
"epoch": 0.6124920331421287,
"grad_norm": 1.0835683345794678,
"learning_rate": 6.686955993448617e-05,
"loss": 0.2024,
"step": 961
},
{
"epoch": 0.6131293817718292,
"grad_norm": 1.1579644680023193,
"learning_rate": 6.667826811231157e-05,
"loss": 0.2215,
"step": 962
},
{
"epoch": 0.6137667304015296,
"grad_norm": 0.514375627040863,
"learning_rate": 6.648711335442883e-05,
"loss": 0.0758,
"step": 963
},
{
"epoch": 0.6144040790312301,
"grad_norm": 1.0089221000671387,
"learning_rate": 6.629609644712633e-05,
"loss": 0.4259,
"step": 964
},
{
"epoch": 0.6150414276609305,
"grad_norm": 0.8962211012840271,
"learning_rate": 6.610521817612553e-05,
"loss": 0.1434,
"step": 965
},
{
"epoch": 0.615678776290631,
"grad_norm": 1.0341756343841553,
"learning_rate": 6.591447932657746e-05,
"loss": 0.2564,
"step": 966
},
{
"epoch": 0.6163161249203314,
"grad_norm": 0.8328891396522522,
"learning_rate": 6.57238806830598e-05,
"loss": 0.273,
"step": 967
},
{
"epoch": 0.6169534735500318,
"grad_norm": 0.5026578903198242,
"learning_rate": 6.553342302957346e-05,
"loss": 0.0481,
"step": 968
},
{
"epoch": 0.6175908221797323,
"grad_norm": 0.9261746406555176,
"learning_rate": 6.534310714953938e-05,
"loss": 0.4245,
"step": 969
},
{
"epoch": 0.6182281708094327,
"grad_norm": 0.8884578347206116,
"learning_rate": 6.515293382579535e-05,
"loss": 0.1722,
"step": 970
},
{
"epoch": 0.6188655194391333,
"grad_norm": 0.7685832977294922,
"learning_rate": 6.496290384059284e-05,
"loss": 0.1776,
"step": 971
},
{
"epoch": 0.6195028680688337,
"grad_norm": 1.0669718980789185,
"learning_rate": 6.477301797559361e-05,
"loss": 0.3195,
"step": 972
},
{
"epoch": 0.6201402166985341,
"grad_norm": 1.0431551933288574,
"learning_rate": 6.458327701186671e-05,
"loss": 0.3094,
"step": 973
},
{
"epoch": 0.6207775653282346,
"grad_norm": 0.6476925611495972,
"learning_rate": 6.439368172988505e-05,
"loss": 0.2227,
"step": 974
},
{
"epoch": 0.621414913957935,
"grad_norm": 0.6935895681381226,
"learning_rate": 6.420423290952233e-05,
"loss": 0.2056,
"step": 975
},
{
"epoch": 0.6220522625876355,
"grad_norm": 0.8205288648605347,
"learning_rate": 6.401493133004989e-05,
"loss": 0.2589,
"step": 976
},
{
"epoch": 0.6226896112173359,
"grad_norm": 0.8635762333869934,
"learning_rate": 6.382577777013337e-05,
"loss": 0.3142,
"step": 977
},
{
"epoch": 0.6233269598470363,
"grad_norm": 1.1652648448944092,
"learning_rate": 6.363677300782943e-05,
"loss": 0.4899,
"step": 978
},
{
"epoch": 0.6239643084767368,
"grad_norm": 0.6844851970672607,
"learning_rate": 6.344791782058286e-05,
"loss": 0.1927,
"step": 979
},
{
"epoch": 0.6246016571064372,
"grad_norm": 0.8243976831436157,
"learning_rate": 6.325921298522304e-05,
"loss": 0.2268,
"step": 980
},
{
"epoch": 0.6252390057361377,
"grad_norm": 0.9082722663879395,
"learning_rate": 6.307065927796096e-05,
"loss": 0.0929,
"step": 981
},
{
"epoch": 0.6258763543658381,
"grad_norm": 1.1699886322021484,
"learning_rate": 6.288225747438604e-05,
"loss": 0.1539,
"step": 982
},
{
"epoch": 0.6265137029955385,
"grad_norm": 0.8078843355178833,
"learning_rate": 6.269400834946277e-05,
"loss": 0.105,
"step": 983
},
{
"epoch": 0.627151051625239,
"grad_norm": 0.9834449887275696,
"learning_rate": 6.250591267752761e-05,
"loss": 0.3612,
"step": 984
},
{
"epoch": 0.6277884002549394,
"grad_norm": 0.7381107211112976,
"learning_rate": 6.231797123228589e-05,
"loss": 0.2275,
"step": 985
},
{
"epoch": 0.6284257488846399,
"grad_norm": 0.6575921773910522,
"learning_rate": 6.213018478680843e-05,
"loss": 0.1963,
"step": 986
},
{
"epoch": 0.6290630975143403,
"grad_norm": 0.9638115763664246,
"learning_rate": 6.194255411352863e-05,
"loss": 0.3496,
"step": 987
},
{
"epoch": 0.6297004461440407,
"grad_norm": 1.0635380744934082,
"learning_rate": 6.175507998423907e-05,
"loss": 0.2985,
"step": 988
},
{
"epoch": 0.6303377947737412,
"grad_norm": 0.4694770574569702,
"learning_rate": 6.156776317008835e-05,
"loss": 0.052,
"step": 989
},
{
"epoch": 0.6309751434034416,
"grad_norm": 0.8005937933921814,
"learning_rate": 6.138060444157807e-05,
"loss": 0.2596,
"step": 990
},
{
"epoch": 0.6316124920331422,
"grad_norm": 0.7719281911849976,
"learning_rate": 6.11936045685595e-05,
"loss": 0.2659,
"step": 991
},
{
"epoch": 0.6322498406628426,
"grad_norm": 1.0329854488372803,
"learning_rate": 6.100676432023058e-05,
"loss": 0.2494,
"step": 992
},
{
"epoch": 0.6328871892925431,
"grad_norm": 0.6563783288002014,
"learning_rate": 6.082008446513251e-05,
"loss": 0.1409,
"step": 993
},
{
"epoch": 0.6335245379222435,
"grad_norm": 0.8198005557060242,
"learning_rate": 6.063356577114685e-05,
"loss": 0.2563,
"step": 994
},
{
"epoch": 0.6341618865519439,
"grad_norm": 1.0802313089370728,
"learning_rate": 6.0447209005492186e-05,
"loss": 0.4181,
"step": 995
},
{
"epoch": 0.6347992351816444,
"grad_norm": 1.1442335844039917,
"learning_rate": 6.0261014934721074e-05,
"loss": 0.1336,
"step": 996
},
{
"epoch": 0.6354365838113448,
"grad_norm": 0.7847735285758972,
"learning_rate": 6.007498432471677e-05,
"loss": 0.2257,
"step": 997
},
{
"epoch": 0.6360739324410453,
"grad_norm": 0.652402400970459,
"learning_rate": 5.988911794069031e-05,
"loss": 0.1231,
"step": 998
},
{
"epoch": 0.6367112810707457,
"grad_norm": 0.5656065344810486,
"learning_rate": 5.970341654717705e-05,
"loss": 0.1101,
"step": 999
},
{
"epoch": 0.6373486297004461,
"grad_norm": 1.004534363746643,
"learning_rate": 5.951788090803379e-05,
"loss": 0.3001,
"step": 1000
},
{
"epoch": 0.6379859783301466,
"grad_norm": 1.1844885349273682,
"learning_rate": 5.9332511786435416e-05,
"loss": 0.2872,
"step": 1001
},
{
"epoch": 0.638623326959847,
"grad_norm": 1.362561821937561,
"learning_rate": 5.9147309944871954e-05,
"loss": 0.4581,
"step": 1002
},
{
"epoch": 0.6392606755895475,
"grad_norm": 0.8407707214355469,
"learning_rate": 5.896227614514538e-05,
"loss": 0.2224,
"step": 1003
},
{
"epoch": 0.6398980242192479,
"grad_norm": 0.7078567147254944,
"learning_rate": 5.877741114836631e-05,
"loss": 0.1334,
"step": 1004
},
{
"epoch": 0.6405353728489483,
"grad_norm": 0.4722069799900055,
"learning_rate": 5.859271571495115e-05,
"loss": 0.0606,
"step": 1005
},
{
"epoch": 0.6411727214786488,
"grad_norm": 0.9850430488586426,
"learning_rate": 5.840819060461877e-05,
"loss": 0.2342,
"step": 1006
},
{
"epoch": 0.6418100701083492,
"grad_norm": 1.0159498453140259,
"learning_rate": 5.8223836576387416e-05,
"loss": 0.405,
"step": 1007
},
{
"epoch": 0.6424474187380497,
"grad_norm": 0.5655425786972046,
"learning_rate": 5.803965438857171e-05,
"loss": 0.1927,
"step": 1008
},
{
"epoch": 0.6430847673677501,
"grad_norm": 0.6650108098983765,
"learning_rate": 5.785564479877927e-05,
"loss": 0.1383,
"step": 1009
},
{
"epoch": 0.6437221159974506,
"grad_norm": 0.23765377700328827,
"learning_rate": 5.767180856390797e-05,
"loss": 0.0372,
"step": 1010
},
{
"epoch": 0.6443594646271511,
"grad_norm": 0.8595127463340759,
"learning_rate": 5.7488146440142445e-05,
"loss": 0.2417,
"step": 1011
},
{
"epoch": 0.6449968132568515,
"grad_norm": 0.942446231842041,
"learning_rate": 5.730465918295117e-05,
"loss": 0.2268,
"step": 1012
},
{
"epoch": 0.645634161886552,
"grad_norm": 0.5906773805618286,
"learning_rate": 5.712134754708331e-05,
"loss": 0.1145,
"step": 1013
},
{
"epoch": 0.6462715105162524,
"grad_norm": 0.6804829239845276,
"learning_rate": 5.693821228656583e-05,
"loss": 0.1795,
"step": 1014
},
{
"epoch": 0.6469088591459529,
"grad_norm": 1.2095837593078613,
"learning_rate": 5.675525415469998e-05,
"loss": 0.4063,
"step": 1015
},
{
"epoch": 0.6475462077756533,
"grad_norm": 0.6919000148773193,
"learning_rate": 5.65724739040585e-05,
"loss": 0.1653,
"step": 1016
},
{
"epoch": 0.6481835564053537,
"grad_norm": 0.8155308961868286,
"learning_rate": 5.638987228648238e-05,
"loss": 0.28,
"step": 1017
},
{
"epoch": 0.6488209050350542,
"grad_norm": 1.0539478063583374,
"learning_rate": 5.620745005307793e-05,
"loss": 0.3918,
"step": 1018
},
{
"epoch": 0.6494582536647546,
"grad_norm": 0.5815137028694153,
"learning_rate": 5.602520795421357e-05,
"loss": 0.1478,
"step": 1019
},
{
"epoch": 0.6500956022944551,
"grad_norm": 1.2935404777526855,
"learning_rate": 5.5843146739516683e-05,
"loss": 0.3363,
"step": 1020
},
{
"epoch": 0.6507329509241555,
"grad_norm": 1.2574087381362915,
"learning_rate": 5.5661267157870636e-05,
"loss": 0.3255,
"step": 1021
},
{
"epoch": 0.651370299553856,
"grad_norm": 0.601564347743988,
"learning_rate": 5.547956995741176e-05,
"loss": 0.1497,
"step": 1022
},
{
"epoch": 0.6520076481835564,
"grad_norm": 0.789741039276123,
"learning_rate": 5.529805588552609e-05,
"loss": 0.3208,
"step": 1023
},
{
"epoch": 0.6526449968132568,
"grad_norm": 0.9937177896499634,
"learning_rate": 5.511672568884637e-05,
"loss": 0.3073,
"step": 1024
},
{
"epoch": 0.6532823454429573,
"grad_norm": 0.7651997208595276,
"learning_rate": 5.493558011324913e-05,
"loss": 0.1496,
"step": 1025
},
{
"epoch": 0.6539196940726577,
"grad_norm": 0.9328089356422424,
"learning_rate": 5.475461990385133e-05,
"loss": 0.1427,
"step": 1026
},
{
"epoch": 0.6545570427023581,
"grad_norm": 0.9245437979698181,
"learning_rate": 5.4573845805007617e-05,
"loss": 0.2799,
"step": 1027
},
{
"epoch": 0.6551943913320586,
"grad_norm": 0.9113796353340149,
"learning_rate": 5.4393258560306926e-05,
"loss": 0.1567,
"step": 1028
},
{
"epoch": 0.655831739961759,
"grad_norm": 0.8145269751548767,
"learning_rate": 5.4212858912569685e-05,
"loss": 0.2705,
"step": 1029
},
{
"epoch": 0.6564690885914596,
"grad_norm": 0.6442999243736267,
"learning_rate": 5.4032647603844655e-05,
"loss": 0.1417,
"step": 1030
},
{
"epoch": 0.65710643722116,
"grad_norm": 0.6885161399841309,
"learning_rate": 5.385262537540597e-05,
"loss": 0.1749,
"step": 1031
},
{
"epoch": 0.6577437858508605,
"grad_norm": 0.6260651350021362,
"learning_rate": 5.367279296774988e-05,
"loss": 0.0769,
"step": 1032
},
{
"epoch": 0.6583811344805609,
"grad_norm": 0.7134547829627991,
"learning_rate": 5.3493151120591875e-05,
"loss": 0.1942,
"step": 1033
},
{
"epoch": 0.6590184831102613,
"grad_norm": 0.8658597469329834,
"learning_rate": 5.331370057286358e-05,
"loss": 0.2577,
"step": 1034
},
{
"epoch": 0.6596558317399618,
"grad_norm": 0.7364838123321533,
"learning_rate": 5.313444206270981e-05,
"loss": 0.0842,
"step": 1035
},
{
"epoch": 0.6602931803696622,
"grad_norm": 0.5216448307037354,
"learning_rate": 5.295537632748546e-05,
"loss": 0.1065,
"step": 1036
},
{
"epoch": 0.6609305289993627,
"grad_norm": 1.4036343097686768,
"learning_rate": 5.277650410375239e-05,
"loss": 0.3339,
"step": 1037
},
{
"epoch": 0.6615678776290631,
"grad_norm": 0.96217280626297,
"learning_rate": 5.259782612727648e-05,
"loss": 0.3326,
"step": 1038
},
{
"epoch": 0.6622052262587635,
"grad_norm": 1.4202228784561157,
"learning_rate": 5.241934313302472e-05,
"loss": 0.3551,
"step": 1039
},
{
"epoch": 0.662842574888464,
"grad_norm": 0.9574637413024902,
"learning_rate": 5.22410558551619e-05,
"loss": 0.2012,
"step": 1040
},
{
"epoch": 0.6634799235181644,
"grad_norm": 0.876212477684021,
"learning_rate": 5.206296502704794e-05,
"loss": 0.0826,
"step": 1041
},
{
"epoch": 0.6641172721478649,
"grad_norm": 1.614162802696228,
"learning_rate": 5.188507138123455e-05,
"loss": 0.4925,
"step": 1042
},
{
"epoch": 0.6647546207775653,
"grad_norm": 0.8218491673469543,
"learning_rate": 5.1707375649462354e-05,
"loss": 0.1169,
"step": 1043
},
{
"epoch": 0.6653919694072657,
"grad_norm": 0.5850120186805725,
"learning_rate": 5.152987856265803e-05,
"loss": 0.0584,
"step": 1044
},
{
"epoch": 0.6660293180369662,
"grad_norm": 0.47425946593284607,
"learning_rate": 5.1352580850930985e-05,
"loss": 0.0503,
"step": 1045
},
{
"epoch": 0.6666666666666666,
"grad_norm": 1.0167031288146973,
"learning_rate": 5.1175483243570665e-05,
"loss": 0.4912,
"step": 1046
},
{
"epoch": 0.6673040152963671,
"grad_norm": 0.36353054642677307,
"learning_rate": 5.09985864690433e-05,
"loss": 0.0543,
"step": 1047
},
{
"epoch": 0.6679413639260675,
"grad_norm": 0.9460597634315491,
"learning_rate": 5.082189125498913e-05,
"loss": 0.2532,
"step": 1048
},
{
"epoch": 0.668578712555768,
"grad_norm": 1.4120869636535645,
"learning_rate": 5.064539832821922e-05,
"loss": 0.4935,
"step": 1049
},
{
"epoch": 0.6692160611854685,
"grad_norm": 1.2351733446121216,
"learning_rate": 5.046910841471253e-05,
"loss": 0.4287,
"step": 1050
},
{
"epoch": 0.6698534098151689,
"grad_norm": 0.8982768058776855,
"learning_rate": 5.0293022239613066e-05,
"loss": 0.2379,
"step": 1051
},
{
"epoch": 0.6704907584448694,
"grad_norm": 0.8030502796173096,
"learning_rate": 5.0117140527226745e-05,
"loss": 0.237,
"step": 1052
},
{
"epoch": 0.6711281070745698,
"grad_norm": 0.8648181557655334,
"learning_rate": 4.994146400101841e-05,
"loss": 0.1707,
"step": 1053
},
{
"epoch": 0.6717654557042703,
"grad_norm": 1.600630283355713,
"learning_rate": 4.976599338360892e-05,
"loss": 0.182,
"step": 1054
},
{
"epoch": 0.6724028043339707,
"grad_norm": 1.1526744365692139,
"learning_rate": 4.959072939677212e-05,
"loss": 0.2043,
"step": 1055
},
{
"epoch": 0.6730401529636711,
"grad_norm": 0.9903702735900879,
"learning_rate": 4.9415672761431986e-05,
"loss": 0.3084,
"step": 1056
},
{
"epoch": 0.6736775015933716,
"grad_norm": 1.160250186920166,
"learning_rate": 4.924082419765959e-05,
"loss": 0.3857,
"step": 1057
},
{
"epoch": 0.674314850223072,
"grad_norm": 1.0104570388793945,
"learning_rate": 4.906618442467005e-05,
"loss": 0.2836,
"step": 1058
},
{
"epoch": 0.6749521988527725,
"grad_norm": 0.5635807514190674,
"learning_rate": 4.8891754160819646e-05,
"loss": 0.1643,
"step": 1059
},
{
"epoch": 0.6755895474824729,
"grad_norm": 1.1220759153366089,
"learning_rate": 4.8717534123603e-05,
"loss": 0.3154,
"step": 1060
},
{
"epoch": 0.6762268961121733,
"grad_norm": 0.6994805335998535,
"learning_rate": 4.854352502964983e-05,
"loss": 0.1236,
"step": 1061
},
{
"epoch": 0.6768642447418738,
"grad_norm": 0.8309299349784851,
"learning_rate": 4.836972759472232e-05,
"loss": 0.1669,
"step": 1062
},
{
"epoch": 0.6775015933715742,
"grad_norm": 1.1031311750411987,
"learning_rate": 4.819614253371193e-05,
"loss": 0.3242,
"step": 1063
},
{
"epoch": 0.6781389420012747,
"grad_norm": 0.8850547671318054,
"learning_rate": 4.802277056063652e-05,
"loss": 0.172,
"step": 1064
},
{
"epoch": 0.6787762906309751,
"grad_norm": 1.1537615060806274,
"learning_rate": 4.784961238863758e-05,
"loss": 0.1838,
"step": 1065
},
{
"epoch": 0.6794136392606756,
"grad_norm": 0.8512182235717773,
"learning_rate": 4.7676668729977036e-05,
"loss": 0.2223,
"step": 1066
},
{
"epoch": 0.680050987890376,
"grad_norm": 0.9696006774902344,
"learning_rate": 4.750394029603443e-05,
"loss": 0.3536,
"step": 1067
},
{
"epoch": 0.6806883365200764,
"grad_norm": 0.4729258716106415,
"learning_rate": 4.733142779730413e-05,
"loss": 0.0984,
"step": 1068
},
{
"epoch": 0.681325685149777,
"grad_norm": 0.7963859438896179,
"learning_rate": 4.715913194339222e-05,
"loss": 0.219,
"step": 1069
},
{
"epoch": 0.6819630337794774,
"grad_norm": 0.3708473742008209,
"learning_rate": 4.6987053443013614e-05,
"loss": 0.0208,
"step": 1070
},
{
"epoch": 0.6826003824091779,
"grad_norm": 1.0191763639450073,
"learning_rate": 4.68151930039892e-05,
"loss": 0.1902,
"step": 1071
},
{
"epoch": 0.6832377310388783,
"grad_norm": 1.0406640768051147,
"learning_rate": 4.664355133324286e-05,
"loss": 0.2446,
"step": 1072
},
{
"epoch": 0.6838750796685787,
"grad_norm": 1.0915453433990479,
"learning_rate": 4.647212913679868e-05,
"loss": 0.3546,
"step": 1073
},
{
"epoch": 0.6845124282982792,
"grad_norm": 1.5243350267410278,
"learning_rate": 4.630092711977797e-05,
"loss": 0.4329,
"step": 1074
},
{
"epoch": 0.6851497769279796,
"grad_norm": 0.6093116402626038,
"learning_rate": 4.6129945986396284e-05,
"loss": 0.1466,
"step": 1075
},
{
"epoch": 0.6857871255576801,
"grad_norm": 1.0749497413635254,
"learning_rate": 4.595918643996061e-05,
"loss": 0.3498,
"step": 1076
},
{
"epoch": 0.6864244741873805,
"grad_norm": 0.8853207230567932,
"learning_rate": 4.578864918286654e-05,
"loss": 0.3243,
"step": 1077
},
{
"epoch": 0.687061822817081,
"grad_norm": 0.11305562406778336,
"learning_rate": 4.5618334916595294e-05,
"loss": 0.0241,
"step": 1078
},
{
"epoch": 0.6876991714467814,
"grad_norm": 0.6961015462875366,
"learning_rate": 4.544824434171081e-05,
"loss": 0.2231,
"step": 1079
},
{
"epoch": 0.6883365200764818,
"grad_norm": 0.923985481262207,
"learning_rate": 4.5278378157856905e-05,
"loss": 0.4028,
"step": 1080
},
{
"epoch": 0.6889738687061823,
"grad_norm": 0.7815952897071838,
"learning_rate": 4.510873706375439e-05,
"loss": 0.2175,
"step": 1081
},
{
"epoch": 0.6896112173358827,
"grad_norm": 0.38949084281921387,
"learning_rate": 4.493932175719827e-05,
"loss": 0.0579,
"step": 1082
},
{
"epoch": 0.6902485659655831,
"grad_norm": 0.3501385748386383,
"learning_rate": 4.477013293505471e-05,
"loss": 0.0195,
"step": 1083
},
{
"epoch": 0.6908859145952836,
"grad_norm": 1.1248000860214233,
"learning_rate": 4.460117129325836e-05,
"loss": 0.3429,
"step": 1084
},
{
"epoch": 0.691523263224984,
"grad_norm": 1.1931782960891724,
"learning_rate": 4.443243752680929e-05,
"loss": 0.2969,
"step": 1085
},
{
"epoch": 0.6921606118546845,
"grad_norm": 0.6619778275489807,
"learning_rate": 4.426393232977034e-05,
"loss": 0.0847,
"step": 1086
},
{
"epoch": 0.6927979604843849,
"grad_norm": 0.9904046058654785,
"learning_rate": 4.409565639526409e-05,
"loss": 0.333,
"step": 1087
},
{
"epoch": 0.6934353091140854,
"grad_norm": 0.10180588066577911,
"learning_rate": 4.3927610415470064e-05,
"loss": 0.024,
"step": 1088
},
{
"epoch": 0.6940726577437859,
"grad_norm": 0.7957285046577454,
"learning_rate": 4.3759795081621965e-05,
"loss": 0.1791,
"step": 1089
},
{
"epoch": 0.6947100063734863,
"grad_norm": 0.37493160367012024,
"learning_rate": 4.35922110840048e-05,
"loss": 0.0705,
"step": 1090
},
{
"epoch": 0.6953473550031868,
"grad_norm": 1.064832329750061,
"learning_rate": 4.342485911195189e-05,
"loss": 0.3587,
"step": 1091
},
{
"epoch": 0.6959847036328872,
"grad_norm": 0.7557972073554993,
"learning_rate": 4.32577398538422e-05,
"loss": 0.0983,
"step": 1092
},
{
"epoch": 0.6966220522625877,
"grad_norm": 0.3659322261810303,
"learning_rate": 4.309085399709745e-05,
"loss": 0.0511,
"step": 1093
},
{
"epoch": 0.6972594008922881,
"grad_norm": 0.9731507301330566,
"learning_rate": 4.2924202228179336e-05,
"loss": 0.2037,
"step": 1094
},
{
"epoch": 0.6978967495219885,
"grad_norm": 0.7988548278808594,
"learning_rate": 4.275778523258668e-05,
"loss": 0.2168,
"step": 1095
},
{
"epoch": 0.698534098151689,
"grad_norm": 1.153964638710022,
"learning_rate": 4.259160369485252e-05,
"loss": 0.2931,
"step": 1096
},
{
"epoch": 0.6991714467813894,
"grad_norm": 0.9738572835922241,
"learning_rate": 4.2425658298541385e-05,
"loss": 0.3106,
"step": 1097
},
{
"epoch": 0.6998087954110899,
"grad_norm": 1.024255633354187,
"learning_rate": 4.225994972624655e-05,
"loss": 0.2967,
"step": 1098
},
{
"epoch": 0.7004461440407903,
"grad_norm": 0.42764341831207275,
"learning_rate": 4.2094478659587024e-05,
"loss": 0.0411,
"step": 1099
},
{
"epoch": 0.7010834926704907,
"grad_norm": 1.0172414779663086,
"learning_rate": 4.192924577920502e-05,
"loss": 0.3063,
"step": 1100
},
{
"epoch": 0.7017208413001912,
"grad_norm": 0.7158370018005371,
"learning_rate": 4.1764251764762875e-05,
"loss": 0.111,
"step": 1101
},
{
"epoch": 0.7023581899298916,
"grad_norm": 0.9178307056427002,
"learning_rate": 4.15994972949404e-05,
"loss": 0.1868,
"step": 1102
},
{
"epoch": 0.7029955385595921,
"grad_norm": 1.4824291467666626,
"learning_rate": 4.143498304743218e-05,
"loss": 0.6061,
"step": 1103
},
{
"epoch": 0.7036328871892925,
"grad_norm": 1.0024702548980713,
"learning_rate": 4.127070969894452e-05,
"loss": 0.364,
"step": 1104
},
{
"epoch": 0.704270235818993,
"grad_norm": 0.8822917938232422,
"learning_rate": 4.110667792519298e-05,
"loss": 0.2657,
"step": 1105
},
{
"epoch": 0.7049075844486934,
"grad_norm": 0.8142839074134827,
"learning_rate": 4.09428884008993e-05,
"loss": 0.2811,
"step": 1106
},
{
"epoch": 0.7055449330783938,
"grad_norm": 1.2901910543441772,
"learning_rate": 4.077934179978888e-05,
"loss": 0.2791,
"step": 1107
},
{
"epoch": 0.7061822817080943,
"grad_norm": 1.0088354349136353,
"learning_rate": 4.0616038794587794e-05,
"loss": 0.248,
"step": 1108
},
{
"epoch": 0.7068196303377948,
"grad_norm": 0.7809550166130066,
"learning_rate": 4.045298005702016e-05,
"loss": 0.2062,
"step": 1109
},
{
"epoch": 0.7074569789674953,
"grad_norm": 0.5823033452033997,
"learning_rate": 4.0290166257805295e-05,
"loss": 0.0761,
"step": 1110
},
{
"epoch": 0.7080943275971957,
"grad_norm": 0.7609939575195312,
"learning_rate": 4.012759806665505e-05,
"loss": 0.2447,
"step": 1111
},
{
"epoch": 0.7087316762268961,
"grad_norm": 1.5352245569229126,
"learning_rate": 3.9965276152271034e-05,
"loss": 0.6704,
"step": 1112
},
{
"epoch": 0.7093690248565966,
"grad_norm": 0.5743124485015869,
"learning_rate": 3.980320118234174e-05,
"loss": 0.1252,
"step": 1113
},
{
"epoch": 0.710006373486297,
"grad_norm": 1.141689419746399,
"learning_rate": 3.9641373823539895e-05,
"loss": 0.3259,
"step": 1114
},
{
"epoch": 0.7106437221159975,
"grad_norm": 0.8912404179573059,
"learning_rate": 3.947979474151977e-05,
"loss": 0.2098,
"step": 1115
},
{
"epoch": 0.7112810707456979,
"grad_norm": 0.7555819153785706,
"learning_rate": 3.931846460091442e-05,
"loss": 0.2027,
"step": 1116
},
{
"epoch": 0.7119184193753983,
"grad_norm": 0.9930626749992371,
"learning_rate": 3.915738406533278e-05,
"loss": 0.1773,
"step": 1117
},
{
"epoch": 0.7125557680050988,
"grad_norm": 1.0219807624816895,
"learning_rate": 3.8996553797357186e-05,
"loss": 0.2019,
"step": 1118
},
{
"epoch": 0.7131931166347992,
"grad_norm": 1.0251866579055786,
"learning_rate": 3.8835974458540426e-05,
"loss": 0.27,
"step": 1119
},
{
"epoch": 0.7138304652644997,
"grad_norm": 1.012486219406128,
"learning_rate": 3.867564670940324e-05,
"loss": 0.3643,
"step": 1120
},
{
"epoch": 0.7144678138942001,
"grad_norm": 1.1464511156082153,
"learning_rate": 3.851557120943145e-05,
"loss": 0.3727,
"step": 1121
},
{
"epoch": 0.7151051625239006,
"grad_norm": 0.5951558947563171,
"learning_rate": 3.835574861707328e-05,
"loss": 0.1905,
"step": 1122
},
{
"epoch": 0.715742511153601,
"grad_norm": 0.36077195405960083,
"learning_rate": 3.8196179589736594e-05,
"loss": 0.0442,
"step": 1123
},
{
"epoch": 0.7163798597833014,
"grad_norm": 0.9777058362960815,
"learning_rate": 3.8036864783786376e-05,
"loss": 0.2665,
"step": 1124
},
{
"epoch": 0.7170172084130019,
"grad_norm": 0.1312219500541687,
"learning_rate": 3.787780485454181e-05,
"loss": 0.0127,
"step": 1125
},
{
"epoch": 0.7176545570427023,
"grad_norm": 0.49427250027656555,
"learning_rate": 3.7719000456273666e-05,
"loss": 0.0867,
"step": 1126
},
{
"epoch": 0.7182919056724028,
"grad_norm": 1.0151724815368652,
"learning_rate": 3.7560452242201695e-05,
"loss": 0.2289,
"step": 1127
},
{
"epoch": 0.7189292543021033,
"grad_norm": 1.1875512599945068,
"learning_rate": 3.7402160864491885e-05,
"loss": 0.3492,
"step": 1128
},
{
"epoch": 0.7195666029318037,
"grad_norm": 0.23903441429138184,
"learning_rate": 3.7244126974253694e-05,
"loss": 0.0416,
"step": 1129
},
{
"epoch": 0.7202039515615042,
"grad_norm": 0.45936426520347595,
"learning_rate": 3.708635122153743e-05,
"loss": 0.0738,
"step": 1130
},
{
"epoch": 0.7208413001912046,
"grad_norm": 0.41278624534606934,
"learning_rate": 3.692883425533162e-05,
"loss": 0.0704,
"step": 1131
},
{
"epoch": 0.7214786488209051,
"grad_norm": 0.40299734473228455,
"learning_rate": 3.6771576723560344e-05,
"loss": 0.0526,
"step": 1132
},
{
"epoch": 0.7221159974506055,
"grad_norm": 1.3829057216644287,
"learning_rate": 3.661457927308052e-05,
"loss": 0.4581,
"step": 1133
},
{
"epoch": 0.722753346080306,
"grad_norm": 1.3514536619186401,
"learning_rate": 3.645784254967921e-05,
"loss": 0.3453,
"step": 1134
},
{
"epoch": 0.7233906947100064,
"grad_norm": 0.6790758371353149,
"learning_rate": 3.630136719807103e-05,
"loss": 0.1359,
"step": 1135
},
{
"epoch": 0.7240280433397068,
"grad_norm": 0.9672769904136658,
"learning_rate": 3.614515386189554e-05,
"loss": 0.4651,
"step": 1136
},
{
"epoch": 0.7246653919694073,
"grad_norm": 0.48652079701423645,
"learning_rate": 3.598920318371444e-05,
"loss": 0.0738,
"step": 1137
},
{
"epoch": 0.7253027405991077,
"grad_norm": 1.1108102798461914,
"learning_rate": 3.583351580500911e-05,
"loss": 0.2818,
"step": 1138
},
{
"epoch": 0.7259400892288081,
"grad_norm": 1.0056939125061035,
"learning_rate": 3.5678092366177837e-05,
"loss": 0.3547,
"step": 1139
},
{
"epoch": 0.7265774378585086,
"grad_norm": 1.0355867147445679,
"learning_rate": 3.55229335065332e-05,
"loss": 0.3093,
"step": 1140
},
{
"epoch": 0.727214786488209,
"grad_norm": 0.7016434669494629,
"learning_rate": 3.536803986429955e-05,
"loss": 0.1727,
"step": 1141
},
{
"epoch": 0.7278521351179095,
"grad_norm": 1.6175440549850464,
"learning_rate": 3.521341207661022e-05,
"loss": 0.2091,
"step": 1142
},
{
"epoch": 0.7284894837476099,
"grad_norm": 0.7671099305152893,
"learning_rate": 3.505905077950509e-05,
"loss": 0.2335,
"step": 1143
},
{
"epoch": 0.7291268323773104,
"grad_norm": 1.0727221965789795,
"learning_rate": 3.490495660792771e-05,
"loss": 0.2663,
"step": 1144
},
{
"epoch": 0.7297641810070108,
"grad_norm": 1.1853899955749512,
"learning_rate": 3.475113019572304e-05,
"loss": 0.3304,
"step": 1145
},
{
"epoch": 0.7304015296367112,
"grad_norm": 0.6341899633407593,
"learning_rate": 3.4597572175634496e-05,
"loss": 0.1595,
"step": 1146
},
{
"epoch": 0.7310388782664117,
"grad_norm": 0.3361879587173462,
"learning_rate": 3.444428317930154e-05,
"loss": 0.0573,
"step": 1147
},
{
"epoch": 0.7316762268961122,
"grad_norm": 0.7132753729820251,
"learning_rate": 3.42912638372571e-05,
"loss": 0.1388,
"step": 1148
},
{
"epoch": 0.7323135755258127,
"grad_norm": 1.1019666194915771,
"learning_rate": 3.413851477892483e-05,
"loss": 0.2709,
"step": 1149
},
{
"epoch": 0.7329509241555131,
"grad_norm": 1.1032804250717163,
"learning_rate": 3.39860366326167e-05,
"loss": 0.2131,
"step": 1150
},
{
"epoch": 0.7335882727852135,
"grad_norm": 0.5099560022354126,
"learning_rate": 3.383383002553027e-05,
"loss": 0.1082,
"step": 1151
},
{
"epoch": 0.734225621414914,
"grad_norm": 1.0154937505722046,
"learning_rate": 3.368189558374611e-05,
"loss": 0.2699,
"step": 1152
},
{
"epoch": 0.7348629700446144,
"grad_norm": 0.9836935997009277,
"learning_rate": 3.353023393222538e-05,
"loss": 0.2327,
"step": 1153
},
{
"epoch": 0.7355003186743149,
"grad_norm": 1.0049062967300415,
"learning_rate": 3.337884569480715e-05,
"loss": 0.1805,
"step": 1154
},
{
"epoch": 0.7361376673040153,
"grad_norm": 1.2056697607040405,
"learning_rate": 3.322773149420575e-05,
"loss": 0.4456,
"step": 1155
},
{
"epoch": 0.7367750159337157,
"grad_norm": 1.254693627357483,
"learning_rate": 3.307689195200834e-05,
"loss": 0.3486,
"step": 1156
},
{
"epoch": 0.7374123645634162,
"grad_norm": 0.8442389369010925,
"learning_rate": 3.2926327688672295e-05,
"loss": 0.2581,
"step": 1157
},
{
"epoch": 0.7380497131931166,
"grad_norm": 0.8513213992118835,
"learning_rate": 3.2776039323522736e-05,
"loss": 0.3164,
"step": 1158
},
{
"epoch": 0.7386870618228171,
"grad_norm": 1.337807059288025,
"learning_rate": 3.262602747474989e-05,
"loss": 0.2845,
"step": 1159
},
{
"epoch": 0.7393244104525175,
"grad_norm": 1.0269842147827148,
"learning_rate": 3.247629275940651e-05,
"loss": 0.1763,
"step": 1160
},
{
"epoch": 0.739961759082218,
"grad_norm": 1.0552716255187988,
"learning_rate": 3.232683579340543e-05,
"loss": 0.3236,
"step": 1161
},
{
"epoch": 0.7405991077119184,
"grad_norm": 0.9615634083747864,
"learning_rate": 3.217765719151706e-05,
"loss": 0.3243,
"step": 1162
},
{
"epoch": 0.7412364563416188,
"grad_norm": 1.0001686811447144,
"learning_rate": 3.2028757567366707e-05,
"loss": 0.181,
"step": 1163
},
{
"epoch": 0.7418738049713193,
"grad_norm": 1.150784969329834,
"learning_rate": 3.188013753343224e-05,
"loss": 0.195,
"step": 1164
},
{
"epoch": 0.7425111536010197,
"grad_norm": 0.8778766393661499,
"learning_rate": 3.173179770104134e-05,
"loss": 0.1776,
"step": 1165
},
{
"epoch": 0.7431485022307202,
"grad_norm": 1.0586961507797241,
"learning_rate": 3.158373868036928e-05,
"loss": 0.2934,
"step": 1166
},
{
"epoch": 0.7437858508604207,
"grad_norm": 0.08438076078891754,
"learning_rate": 3.143596108043614e-05,
"loss": 0.0222,
"step": 1167
},
{
"epoch": 0.7444231994901211,
"grad_norm": 0.9016368985176086,
"learning_rate": 3.1288465509104424e-05,
"loss": 0.1791,
"step": 1168
},
{
"epoch": 0.7450605481198216,
"grad_norm": 1.1317733526229858,
"learning_rate": 3.114125257307656e-05,
"loss": 0.3478,
"step": 1169
},
{
"epoch": 0.745697896749522,
"grad_norm": 0.7872707843780518,
"learning_rate": 3.0994322877892434e-05,
"loss": 0.2126,
"step": 1170
},
{
"epoch": 0.7463352453792225,
"grad_norm": 1.1280791759490967,
"learning_rate": 3.084767702792687e-05,
"loss": 0.3023,
"step": 1171
},
{
"epoch": 0.7469725940089229,
"grad_norm": 1.445211410522461,
"learning_rate": 3.070131562638705e-05,
"loss": 0.2478,
"step": 1172
},
{
"epoch": 0.7476099426386233,
"grad_norm": 0.7021994590759277,
"learning_rate": 3.0555239275310145e-05,
"loss": 0.1644,
"step": 1173
},
{
"epoch": 0.7482472912683238,
"grad_norm": 0.9670385122299194,
"learning_rate": 3.040944857556085e-05,
"loss": 0.2388,
"step": 1174
},
{
"epoch": 0.7488846398980242,
"grad_norm": 1.1045423746109009,
"learning_rate": 3.0263944126828903e-05,
"loss": 0.4179,
"step": 1175
},
{
"epoch": 0.7495219885277247,
"grad_norm": 0.876715362071991,
"learning_rate": 3.0118726527626507e-05,
"loss": 0.2002,
"step": 1176
},
{
"epoch": 0.7501593371574251,
"grad_norm": 0.40951767563819885,
"learning_rate": 2.9973796375285967e-05,
"loss": 0.0675,
"step": 1177
},
{
"epoch": 0.7507966857871256,
"grad_norm": 0.5713475942611694,
"learning_rate": 2.9829154265957205e-05,
"loss": 0.0742,
"step": 1178
},
{
"epoch": 0.751434034416826,
"grad_norm": 0.9962145686149597,
"learning_rate": 2.9684800794605407e-05,
"loss": 0.2722,
"step": 1179
},
{
"epoch": 0.751434034416826,
"eval_loss": 0.2601059079170227,
"eval_runtime": 6.7843,
"eval_samples_per_second": 9.728,
"eval_steps_per_second": 4.864,
"step": 1179
},
{
"epoch": 0.7520713830465264,
"grad_norm": 0.9761635661125183,
"learning_rate": 2.9540736555008362e-05,
"loss": 0.1734,
"step": 1180
},
{
"epoch": 0.7527087316762269,
"grad_norm": 0.7004249095916748,
"learning_rate": 2.939696213975427e-05,
"loss": 0.1173,
"step": 1181
},
{
"epoch": 0.7533460803059273,
"grad_norm": 0.5264585614204407,
"learning_rate": 2.9253478140239044e-05,
"loss": 0.0914,
"step": 1182
},
{
"epoch": 0.7539834289356278,
"grad_norm": 0.8496471643447876,
"learning_rate": 2.911028514666415e-05,
"loss": 0.1976,
"step": 1183
},
{
"epoch": 0.7546207775653282,
"grad_norm": 0.9048736691474915,
"learning_rate": 2.896738374803395e-05,
"loss": 0.0954,
"step": 1184
},
{
"epoch": 0.7552581261950286,
"grad_norm": 0.832561731338501,
"learning_rate": 2.8824774532153353e-05,
"loss": 0.1765,
"step": 1185
},
{
"epoch": 0.7558954748247291,
"grad_norm": 0.9162015914916992,
"learning_rate": 2.868245808562553e-05,
"loss": 0.2553,
"step": 1186
},
{
"epoch": 0.7565328234544296,
"grad_norm": 1.2445670366287231,
"learning_rate": 2.854043499384925e-05,
"loss": 0.4392,
"step": 1187
},
{
"epoch": 0.7571701720841301,
"grad_norm": 1.3423705101013184,
"learning_rate": 2.8398705841016726e-05,
"loss": 0.2308,
"step": 1188
},
{
"epoch": 0.7578075207138305,
"grad_norm": 0.7995614409446716,
"learning_rate": 2.8257271210111024e-05,
"loss": 0.2373,
"step": 1189
},
{
"epoch": 0.758444869343531,
"grad_norm": 1.1977143287658691,
"learning_rate": 2.8116131682903713e-05,
"loss": 0.2843,
"step": 1190
},
{
"epoch": 0.7590822179732314,
"grad_norm": 0.7222425937652588,
"learning_rate": 2.797528783995257e-05,
"loss": 0.1413,
"step": 1191
},
{
"epoch": 0.7597195666029318,
"grad_norm": 0.9210628271102905,
"learning_rate": 2.783474026059909e-05,
"loss": 0.2228,
"step": 1192
},
{
"epoch": 0.7603569152326323,
"grad_norm": 1.032714605331421,
"learning_rate": 2.7694489522966104e-05,
"loss": 0.4799,
"step": 1193
},
{
"epoch": 0.7609942638623327,
"grad_norm": 1.0173118114471436,
"learning_rate": 2.7554536203955437e-05,
"loss": 0.282,
"step": 1194
},
{
"epoch": 0.7616316124920331,
"grad_norm": 0.8013007640838623,
"learning_rate": 2.7414880879245508e-05,
"loss": 0.1599,
"step": 1195
},
{
"epoch": 0.7622689611217336,
"grad_norm": 0.8158217072486877,
"learning_rate": 2.7275524123289e-05,
"loss": 0.16,
"step": 1196
},
{
"epoch": 0.762906309751434,
"grad_norm": 0.8877891898155212,
"learning_rate": 2.713646650931053e-05,
"loss": 0.2061,
"step": 1197
},
{
"epoch": 0.7635436583811345,
"grad_norm": 0.9385794401168823,
"learning_rate": 2.6997708609304127e-05,
"loss": 0.25,
"step": 1198
},
{
"epoch": 0.7641810070108349,
"grad_norm": 0.8377261757850647,
"learning_rate": 2.6859250994031028e-05,
"loss": 0.2053,
"step": 1199
},
{
"epoch": 0.7648183556405354,
"grad_norm": 1.0591347217559814,
"learning_rate": 2.6721094233017342e-05,
"loss": 0.2043,
"step": 1200
},
{
"epoch": 0.7654557042702358,
"grad_norm": 0.8733403086662292,
"learning_rate": 2.6583238894551577e-05,
"loss": 0.1734,
"step": 1201
},
{
"epoch": 0.7660930528999362,
"grad_norm": 0.9723019599914551,
"learning_rate": 2.6445685545682462e-05,
"loss": 0.2577,
"step": 1202
},
{
"epoch": 0.7667304015296367,
"grad_norm": 0.9710585474967957,
"learning_rate": 2.630843475221646e-05,
"loss": 0.2403,
"step": 1203
},
{
"epoch": 0.7673677501593371,
"grad_norm": 0.6627768278121948,
"learning_rate": 2.6171487078715574e-05,
"loss": 0.0852,
"step": 1204
},
{
"epoch": 0.7680050987890376,
"grad_norm": 1.220848560333252,
"learning_rate": 2.603484308849493e-05,
"loss": 0.2762,
"step": 1205
},
{
"epoch": 0.768642447418738,
"grad_norm": 1.0246915817260742,
"learning_rate": 2.589850334362047e-05,
"loss": 0.2402,
"step": 1206
},
{
"epoch": 0.7692797960484385,
"grad_norm": 0.4610997140407562,
"learning_rate": 2.576246840490677e-05,
"loss": 0.0778,
"step": 1207
},
{
"epoch": 0.769917144678139,
"grad_norm": 0.7311042547225952,
"learning_rate": 2.5626738831914467e-05,
"loss": 0.2068,
"step": 1208
},
{
"epoch": 0.7705544933078394,
"grad_norm": 1.2306151390075684,
"learning_rate": 2.5491315182948304e-05,
"loss": 0.3927,
"step": 1209
},
{
"epoch": 0.7711918419375399,
"grad_norm": 0.21768520772457123,
"learning_rate": 2.53561980150545e-05,
"loss": 0.0362,
"step": 1210
},
{
"epoch": 0.7718291905672403,
"grad_norm": 0.5796980857849121,
"learning_rate": 2.5221387884018634e-05,
"loss": 0.0969,
"step": 1211
},
{
"epoch": 0.7724665391969407,
"grad_norm": 0.9904603958129883,
"learning_rate": 2.5086885344363407e-05,
"loss": 0.2924,
"step": 1212
},
{
"epoch": 0.7731038878266412,
"grad_norm": 0.8572244644165039,
"learning_rate": 2.495269094934625e-05,
"loss": 0.1887,
"step": 1213
},
{
"epoch": 0.7737412364563416,
"grad_norm": 0.5620453357696533,
"learning_rate": 2.4818805250957035e-05,
"loss": 0.0847,
"step": 1214
},
{
"epoch": 0.7743785850860421,
"grad_norm": 0.6120081543922424,
"learning_rate": 2.4685228799915905e-05,
"loss": 0.0659,
"step": 1215
},
{
"epoch": 0.7750159337157425,
"grad_norm": 0.5261247754096985,
"learning_rate": 2.4551962145670904e-05,
"loss": 0.0788,
"step": 1216
},
{
"epoch": 0.775653282345443,
"grad_norm": 0.7822028398513794,
"learning_rate": 2.4419005836395838e-05,
"loss": 0.1444,
"step": 1217
},
{
"epoch": 0.7762906309751434,
"grad_norm": 0.7344506978988647,
"learning_rate": 2.4286360418987963e-05,
"loss": 0.1334,
"step": 1218
},
{
"epoch": 0.7769279796048438,
"grad_norm": 1.027187705039978,
"learning_rate": 2.4154026439065634e-05,
"loss": 0.2788,
"step": 1219
},
{
"epoch": 0.7775653282345443,
"grad_norm": 0.7066121697425842,
"learning_rate": 2.4022004440966195e-05,
"loss": 0.175,
"step": 1220
},
{
"epoch": 0.7782026768642447,
"grad_norm": 0.8603143692016602,
"learning_rate": 2.3890294967743742e-05,
"loss": 0.1679,
"step": 1221
},
{
"epoch": 0.7788400254939452,
"grad_norm": 0.624467670917511,
"learning_rate": 2.3758898561166787e-05,
"loss": 0.1073,
"step": 1222
},
{
"epoch": 0.7794773741236456,
"grad_norm": 1.2390384674072266,
"learning_rate": 2.3627815761716064e-05,
"loss": 0.2942,
"step": 1223
},
{
"epoch": 0.780114722753346,
"grad_norm": 0.7339411973953247,
"learning_rate": 2.349704710858244e-05,
"loss": 0.1179,
"step": 1224
},
{
"epoch": 0.7807520713830465,
"grad_norm": 0.8258393406867981,
"learning_rate": 2.336659313966444e-05,
"loss": 0.1117,
"step": 1225
},
{
"epoch": 0.781389420012747,
"grad_norm": 0.9481708407402039,
"learning_rate": 2.323645439156632e-05,
"loss": 0.2201,
"step": 1226
},
{
"epoch": 0.7820267686424475,
"grad_norm": 1.064927339553833,
"learning_rate": 2.3106631399595614e-05,
"loss": 0.2947,
"step": 1227
},
{
"epoch": 0.7826641172721479,
"grad_norm": 1.10151207447052,
"learning_rate": 2.2977124697761056e-05,
"loss": 0.3264,
"step": 1228
},
{
"epoch": 0.7833014659018483,
"grad_norm": 1.6724836826324463,
"learning_rate": 2.284793481877039e-05,
"loss": 0.4314,
"step": 1229
},
{
"epoch": 0.7839388145315488,
"grad_norm": 0.6590045094490051,
"learning_rate": 2.2719062294028202e-05,
"loss": 0.1281,
"step": 1230
},
{
"epoch": 0.7845761631612492,
"grad_norm": 1.1867492198944092,
"learning_rate": 2.2590507653633576e-05,
"loss": 0.3377,
"step": 1231
},
{
"epoch": 0.7852135117909497,
"grad_norm": 1.1134761571884155,
"learning_rate": 2.24622714263781e-05,
"loss": 0.2753,
"step": 1232
},
{
"epoch": 0.7858508604206501,
"grad_norm": 0.5853508114814758,
"learning_rate": 2.2334354139743518e-05,
"loss": 0.1181,
"step": 1233
},
{
"epoch": 0.7864882090503506,
"grad_norm": 0.5680104494094849,
"learning_rate": 2.2206756319899846e-05,
"loss": 0.1468,
"step": 1234
},
{
"epoch": 0.787125557680051,
"grad_norm": 0.6641390323638916,
"learning_rate": 2.2079478491702854e-05,
"loss": 0.1407,
"step": 1235
},
{
"epoch": 0.7877629063097514,
"grad_norm": 1.085426926612854,
"learning_rate": 2.1952521178692108e-05,
"loss": 0.2798,
"step": 1236
},
{
"epoch": 0.7884002549394519,
"grad_norm": 0.8215023875236511,
"learning_rate": 2.1825884903088754e-05,
"loss": 0.2216,
"step": 1237
},
{
"epoch": 0.7890376035691523,
"grad_norm": 1.0317955017089844,
"learning_rate": 2.1699570185793515e-05,
"loss": 0.2723,
"step": 1238
},
{
"epoch": 0.7896749521988528,
"grad_norm": 0.8241130113601685,
"learning_rate": 2.157357754638426e-05,
"loss": 0.1662,
"step": 1239
},
{
"epoch": 0.7903123008285532,
"grad_norm": 0.7739633321762085,
"learning_rate": 2.1447907503114193e-05,
"loss": 0.2274,
"step": 1240
},
{
"epoch": 0.7909496494582536,
"grad_norm": 0.8911176919937134,
"learning_rate": 2.132256057290942e-05,
"loss": 0.2596,
"step": 1241
},
{
"epoch": 0.7915869980879541,
"grad_norm": 0.7646307945251465,
"learning_rate": 2.119753727136712e-05,
"loss": 0.1994,
"step": 1242
},
{
"epoch": 0.7922243467176545,
"grad_norm": 0.7631604671478271,
"learning_rate": 2.1072838112753136e-05,
"loss": 0.1738,
"step": 1243
},
{
"epoch": 0.792861695347355,
"grad_norm": 0.17072726786136627,
"learning_rate": 2.0948463610000037e-05,
"loss": 0.02,
"step": 1244
},
{
"epoch": 0.7934990439770554,
"grad_norm": 0.7780157327651978,
"learning_rate": 2.0824414274705007e-05,
"loss": 0.1636,
"step": 1245
},
{
"epoch": 0.794136392606756,
"grad_norm": 0.7698034644126892,
"learning_rate": 2.070069061712764e-05,
"loss": 0.1583,
"step": 1246
},
{
"epoch": 0.7947737412364564,
"grad_norm": 1.2707273960113525,
"learning_rate": 2.057729314618796e-05,
"loss": 0.3502,
"step": 1247
},
{
"epoch": 0.7954110898661568,
"grad_norm": 0.7315055727958679,
"learning_rate": 2.0454222369464226e-05,
"loss": 0.1561,
"step": 1248
},
{
"epoch": 0.7960484384958573,
"grad_norm": 0.4676603078842163,
"learning_rate": 2.033147879319086e-05,
"loss": 0.0313,
"step": 1249
},
{
"epoch": 0.7966857871255577,
"grad_norm": 0.5674156546592712,
"learning_rate": 2.020906292225645e-05,
"loss": 0.0747,
"step": 1250
},
{
"epoch": 0.7973231357552581,
"grad_norm": 0.644936740398407,
"learning_rate": 2.0086975260201624e-05,
"loss": 0.1034,
"step": 1251
},
{
"epoch": 0.7979604843849586,
"grad_norm": 0.6660456657409668,
"learning_rate": 1.9965216309216893e-05,
"loss": 0.1319,
"step": 1252
},
{
"epoch": 0.798597833014659,
"grad_norm": 0.6221739649772644,
"learning_rate": 1.9843786570140723e-05,
"loss": 0.1643,
"step": 1253
},
{
"epoch": 0.7992351816443595,
"grad_norm": 0.7158281207084656,
"learning_rate": 1.9722686542457357e-05,
"loss": 0.2178,
"step": 1254
},
{
"epoch": 0.7998725302740599,
"grad_norm": 0.8408859372138977,
"learning_rate": 1.9601916724294868e-05,
"loss": 0.1846,
"step": 1255
},
{
"epoch": 0.8005098789037604,
"grad_norm": 0.7760481834411621,
"learning_rate": 1.9481477612423083e-05,
"loss": 0.1884,
"step": 1256
},
{
"epoch": 0.8011472275334608,
"grad_norm": 0.7887476086616516,
"learning_rate": 1.936136970225143e-05,
"loss": 0.3213,
"step": 1257
},
{
"epoch": 0.8017845761631612,
"grad_norm": 0.948664665222168,
"learning_rate": 1.9241593487827026e-05,
"loss": 0.1821,
"step": 1258
},
{
"epoch": 0.8024219247928617,
"grad_norm": 0.7216320037841797,
"learning_rate": 1.9122149461832638e-05,
"loss": 0.2067,
"step": 1259
},
{
"epoch": 0.8030592734225621,
"grad_norm": 0.9751262068748474,
"learning_rate": 1.900303811558455e-05,
"loss": 0.1861,
"step": 1260
},
{
"epoch": 0.8036966220522626,
"grad_norm": 1.2144855260849,
"learning_rate": 1.8884259939030703e-05,
"loss": 0.3434,
"step": 1261
},
{
"epoch": 0.804333970681963,
"grad_norm": 0.8199194073677063,
"learning_rate": 1.8765815420748512e-05,
"loss": 0.2145,
"step": 1262
},
{
"epoch": 0.8049713193116634,
"grad_norm": 0.8331034183502197,
"learning_rate": 1.8647705047942944e-05,
"loss": 0.1831,
"step": 1263
},
{
"epoch": 0.8056086679413639,
"grad_norm": 1.0294809341430664,
"learning_rate": 1.8529929306444583e-05,
"loss": 0.2417,
"step": 1264
},
{
"epoch": 0.8062460165710643,
"grad_norm": 1.087731122970581,
"learning_rate": 1.841248868070746e-05,
"loss": 0.3354,
"step": 1265
},
{
"epoch": 0.8068833652007649,
"grad_norm": 0.6705929040908813,
"learning_rate": 1.8295383653807186e-05,
"loss": 0.0872,
"step": 1266
},
{
"epoch": 0.8075207138304653,
"grad_norm": 1.2238281965255737,
"learning_rate": 1.8178614707438957e-05,
"loss": 0.3695,
"step": 1267
},
{
"epoch": 0.8081580624601657,
"grad_norm": 0.5290761590003967,
"learning_rate": 1.8062182321915545e-05,
"loss": 0.0669,
"step": 1268
},
{
"epoch": 0.8087954110898662,
"grad_norm": 0.8032166361808777,
"learning_rate": 1.7946086976165298e-05,
"loss": 0.1655,
"step": 1269
},
{
"epoch": 0.8094327597195666,
"grad_norm": 0.5958200097084045,
"learning_rate": 1.7830329147730217e-05,
"loss": 0.0985,
"step": 1270
},
{
"epoch": 0.8100701083492671,
"grad_norm": 0.7417314648628235,
"learning_rate": 1.7714909312763904e-05,
"loss": 0.1734,
"step": 1271
},
{
"epoch": 0.8107074569789675,
"grad_norm": 0.6442215442657471,
"learning_rate": 1.7599827946029846e-05,
"loss": 0.1247,
"step": 1272
},
{
"epoch": 0.811344805608668,
"grad_norm": 0.7321109175682068,
"learning_rate": 1.7485085520899125e-05,
"loss": 0.1393,
"step": 1273
},
{
"epoch": 0.8119821542383684,
"grad_norm": 1.2598332166671753,
"learning_rate": 1.737068250934868e-05,
"loss": 0.2359,
"step": 1274
},
{
"epoch": 0.8126195028680688,
"grad_norm": 1.3078793287277222,
"learning_rate": 1.7256619381959315e-05,
"loss": 0.4153,
"step": 1275
},
{
"epoch": 0.8132568514977693,
"grad_norm": 0.18744127452373505,
"learning_rate": 1.7142896607913795e-05,
"loss": 0.0292,
"step": 1276
},
{
"epoch": 0.8138942001274697,
"grad_norm": 0.8837572932243347,
"learning_rate": 1.7029514654994893e-05,
"loss": 0.3866,
"step": 1277
},
{
"epoch": 0.8145315487571702,
"grad_norm": 0.9059514999389648,
"learning_rate": 1.6916473989583425e-05,
"loss": 0.2338,
"step": 1278
},
{
"epoch": 0.8151688973868706,
"grad_norm": 0.6876276135444641,
"learning_rate": 1.6803775076656392e-05,
"loss": 0.0777,
"step": 1279
},
{
"epoch": 0.815806246016571,
"grad_norm": 0.22968345880508423,
"learning_rate": 1.6691418379785006e-05,
"loss": 0.0282,
"step": 1280
},
{
"epoch": 0.8164435946462715,
"grad_norm": 0.7356005907058716,
"learning_rate": 1.65794043611329e-05,
"loss": 0.1504,
"step": 1281
},
{
"epoch": 0.8170809432759719,
"grad_norm": 0.7357573509216309,
"learning_rate": 1.646773348145406e-05,
"loss": 0.2722,
"step": 1282
},
{
"epoch": 0.8177182919056724,
"grad_norm": 0.83384108543396,
"learning_rate": 1.63564062000911e-05,
"loss": 0.139,
"step": 1283
},
{
"epoch": 0.8183556405353728,
"grad_norm": 0.9476302266120911,
"learning_rate": 1.624542297497319e-05,
"loss": 0.322,
"step": 1284
},
{
"epoch": 0.8189929891650733,
"grad_norm": 0.6607417464256287,
"learning_rate": 1.61347842626144e-05,
"loss": 0.1329,
"step": 1285
},
{
"epoch": 0.8196303377947738,
"grad_norm": 0.7398813366889954,
"learning_rate": 1.60244905181116e-05,
"loss": 0.1184,
"step": 1286
},
{
"epoch": 0.8202676864244742,
"grad_norm": 0.9953367710113525,
"learning_rate": 1.591454219514267e-05,
"loss": 0.2581,
"step": 1287
},
{
"epoch": 0.8209050350541747,
"grad_norm": 1.211151361465454,
"learning_rate": 1.580493974596472e-05,
"loss": 0.2758,
"step": 1288
},
{
"epoch": 0.8215423836838751,
"grad_norm": 1.1378930807113647,
"learning_rate": 1.569568362141215e-05,
"loss": 0.3088,
"step": 1289
},
{
"epoch": 0.8221797323135756,
"grad_norm": 1.3171985149383545,
"learning_rate": 1.5586774270894743e-05,
"loss": 0.4319,
"step": 1290
},
{
"epoch": 0.822817080943276,
"grad_norm": 0.3057982325553894,
"learning_rate": 1.5478212142395908e-05,
"loss": 0.0384,
"step": 1291
},
{
"epoch": 0.8234544295729764,
"grad_norm": 1.0368626117706299,
"learning_rate": 1.5369997682470782e-05,
"loss": 0.2804,
"step": 1292
},
{
"epoch": 0.8240917782026769,
"grad_norm": 0.871778130531311,
"learning_rate": 1.5262131336244466e-05,
"loss": 0.3481,
"step": 1293
},
{
"epoch": 0.8247291268323773,
"grad_norm": 0.5947046875953674,
"learning_rate": 1.5154613547410124e-05,
"loss": 0.0708,
"step": 1294
},
{
"epoch": 0.8253664754620778,
"grad_norm": 0.8667436838150024,
"learning_rate": 1.5047444758227158e-05,
"loss": 0.1743,
"step": 1295
},
{
"epoch": 0.8260038240917782,
"grad_norm": 1.1830843687057495,
"learning_rate": 1.4940625409519394e-05,
"loss": 0.2759,
"step": 1296
},
{
"epoch": 0.8266411727214786,
"grad_norm": 0.6308046579360962,
"learning_rate": 1.4834155940673356e-05,
"loss": 0.1001,
"step": 1297
},
{
"epoch": 0.8272785213511791,
"grad_norm": 1.1325982809066772,
"learning_rate": 1.4728036789636301e-05,
"loss": 0.3568,
"step": 1298
},
{
"epoch": 0.8279158699808795,
"grad_norm": 0.8669734597206116,
"learning_rate": 1.4622268392914562e-05,
"loss": 0.2937,
"step": 1299
},
{
"epoch": 0.82855321861058,
"grad_norm": 1.186834692955017,
"learning_rate": 1.4516851185571679e-05,
"loss": 0.3148,
"step": 1300
},
{
"epoch": 0.8291905672402804,
"grad_norm": 0.9354113340377808,
"learning_rate": 1.4411785601226568e-05,
"loss": 0.2381,
"step": 1301
},
{
"epoch": 0.8298279158699808,
"grad_norm": 1.5794496536254883,
"learning_rate": 1.4307072072051896e-05,
"loss": 0.5385,
"step": 1302
},
{
"epoch": 0.8304652644996813,
"grad_norm": 0.20612755417823792,
"learning_rate": 1.420271102877212e-05,
"loss": 0.0191,
"step": 1303
},
{
"epoch": 0.8311026131293817,
"grad_norm": 1.3294650316238403,
"learning_rate": 1.409870290066183e-05,
"loss": 0.335,
"step": 1304
},
{
"epoch": 0.8317399617590823,
"grad_norm": 1.126065969467163,
"learning_rate": 1.3995048115543929e-05,
"loss": 0.253,
"step": 1305
},
{
"epoch": 0.8323773103887827,
"grad_norm": 1.0217036008834839,
"learning_rate": 1.3891747099787921e-05,
"loss": 0.2585,
"step": 1306
},
{
"epoch": 0.8330146590184832,
"grad_norm": 0.7787379622459412,
"learning_rate": 1.3788800278308112e-05,
"loss": 0.1751,
"step": 1307
},
{
"epoch": 0.8336520076481836,
"grad_norm": 1.5774219036102295,
"learning_rate": 1.3686208074561857e-05,
"loss": 0.3862,
"step": 1308
},
{
"epoch": 0.834289356277884,
"grad_norm": 0.8790549635887146,
"learning_rate": 1.358397091054786e-05,
"loss": 0.1654,
"step": 1309
},
{
"epoch": 0.8349267049075845,
"grad_norm": 0.8315568566322327,
"learning_rate": 1.3482089206804426e-05,
"loss": 0.2008,
"step": 1310
},
{
"epoch": 0.8355640535372849,
"grad_norm": 0.6818469166755676,
"learning_rate": 1.3380563382407763e-05,
"loss": 0.1506,
"step": 1311
},
{
"epoch": 0.8362014021669854,
"grad_norm": 0.7880542278289795,
"learning_rate": 1.3279393854970135e-05,
"loss": 0.1556,
"step": 1312
},
{
"epoch": 0.8368387507966858,
"grad_norm": 0.8030346035957336,
"learning_rate": 1.317858104063825e-05,
"loss": 0.1284,
"step": 1313
},
{
"epoch": 0.8374760994263862,
"grad_norm": 1.3306667804718018,
"learning_rate": 1.3078125354091564e-05,
"loss": 0.3406,
"step": 1314
},
{
"epoch": 0.8381134480560867,
"grad_norm": 0.5547246932983398,
"learning_rate": 1.2978027208540556e-05,
"loss": 0.0688,
"step": 1315
},
{
"epoch": 0.8387507966857871,
"grad_norm": 1.3354671001434326,
"learning_rate": 1.2878287015724932e-05,
"loss": 0.2598,
"step": 1316
},
{
"epoch": 0.8393881453154876,
"grad_norm": 0.8202941417694092,
"learning_rate": 1.277890518591206e-05,
"loss": 0.3611,
"step": 1317
},
{
"epoch": 0.840025493945188,
"grad_norm": 1.0215110778808594,
"learning_rate": 1.2679882127895193e-05,
"loss": 0.3167,
"step": 1318
},
{
"epoch": 0.8406628425748884,
"grad_norm": 0.5844862461090088,
"learning_rate": 1.2581218248991877e-05,
"loss": 0.0839,
"step": 1319
},
{
"epoch": 0.8413001912045889,
"grad_norm": 0.689862072467804,
"learning_rate": 1.248291395504223e-05,
"loss": 0.1641,
"step": 1320
},
{
"epoch": 0.8419375398342893,
"grad_norm": 0.7083310484886169,
"learning_rate": 1.238496965040722e-05,
"loss": 0.1797,
"step": 1321
},
{
"epoch": 0.8425748884639898,
"grad_norm": 0.7543584108352661,
"learning_rate": 1.228738573796704e-05,
"loss": 0.1806,
"step": 1322
},
{
"epoch": 0.8432122370936902,
"grad_norm": 1.0917845964431763,
"learning_rate": 1.2190162619119539e-05,
"loss": 0.5464,
"step": 1323
},
{
"epoch": 0.8438495857233907,
"grad_norm": 0.6594381928443909,
"learning_rate": 1.209330069377842e-05,
"loss": 0.1569,
"step": 1324
},
{
"epoch": 0.8444869343530912,
"grad_norm": 1.0132554769515991,
"learning_rate": 1.1996800360371673e-05,
"loss": 0.3943,
"step": 1325
},
{
"epoch": 0.8451242829827916,
"grad_norm": 0.6865078806877136,
"learning_rate": 1.1900662015839947e-05,
"loss": 0.1895,
"step": 1326
},
{
"epoch": 0.8457616316124921,
"grad_norm": 1.0825668573379517,
"learning_rate": 1.1804886055634945e-05,
"loss": 0.5547,
"step": 1327
},
{
"epoch": 0.8463989802421925,
"grad_norm": 0.6795901656150818,
"learning_rate": 1.1709472873717675e-05,
"loss": 0.1052,
"step": 1328
},
{
"epoch": 0.847036328871893,
"grad_norm": 1.2898623943328857,
"learning_rate": 1.1614422862556939e-05,
"loss": 0.2324,
"step": 1329
},
{
"epoch": 0.8476736775015934,
"grad_norm": 0.9949779510498047,
"learning_rate": 1.1519736413127668e-05,
"loss": 0.4049,
"step": 1330
},
{
"epoch": 0.8483110261312938,
"grad_norm": 0.5181422233581543,
"learning_rate": 1.1425413914909378e-05,
"loss": 0.0619,
"step": 1331
},
{
"epoch": 0.8489483747609943,
"grad_norm": 0.9850489497184753,
"learning_rate": 1.1331455755884513e-05,
"loss": 0.2784,
"step": 1332
},
{
"epoch": 0.8495857233906947,
"grad_norm": 0.7785831093788147,
"learning_rate": 1.1237862322536818e-05,
"loss": 0.2654,
"step": 1333
},
{
"epoch": 0.8502230720203952,
"grad_norm": 0.990868091583252,
"learning_rate": 1.1144633999849797e-05,
"loss": 0.1563,
"step": 1334
},
{
"epoch": 0.8508604206500956,
"grad_norm": 0.9114882946014404,
"learning_rate": 1.105177117130518e-05,
"loss": 0.204,
"step": 1335
},
{
"epoch": 0.851497769279796,
"grad_norm": 1.6235325336456299,
"learning_rate": 1.095927421888121e-05,
"loss": 0.5495,
"step": 1336
},
{
"epoch": 0.8521351179094965,
"grad_norm": 0.8826497197151184,
"learning_rate": 1.0867143523051216e-05,
"loss": 0.2237,
"step": 1337
},
{
"epoch": 0.8527724665391969,
"grad_norm": 1.0198767185211182,
"learning_rate": 1.0775379462781942e-05,
"loss": 0.2115,
"step": 1338
},
{
"epoch": 0.8534098151688974,
"grad_norm": 0.962053120136261,
"learning_rate": 1.0683982415532023e-05,
"loss": 0.1139,
"step": 1339
},
{
"epoch": 0.8540471637985978,
"grad_norm": 0.9281373620033264,
"learning_rate": 1.0592952757250474e-05,
"loss": 0.3495,
"step": 1340
},
{
"epoch": 0.8546845124282982,
"grad_norm": 1.4295964241027832,
"learning_rate": 1.0502290862375064e-05,
"loss": 0.3274,
"step": 1341
},
{
"epoch": 0.8553218610579987,
"grad_norm": 1.4286195039749146,
"learning_rate": 1.0411997103830862e-05,
"loss": 0.1792,
"step": 1342
},
{
"epoch": 0.8559592096876991,
"grad_norm": 1.1891881227493286,
"learning_rate": 1.0322071853028614e-05,
"loss": 0.3867,
"step": 1343
},
{
"epoch": 0.8565965583173997,
"grad_norm": 0.9817533493041992,
"learning_rate": 1.0232515479863325e-05,
"loss": 0.2731,
"step": 1344
},
{
"epoch": 0.8572339069471001,
"grad_norm": 0.9481297731399536,
"learning_rate": 1.014332835271259e-05,
"loss": 0.2812,
"step": 1345
},
{
"epoch": 0.8578712555768006,
"grad_norm": 0.8336905241012573,
"learning_rate": 1.0054510838435204e-05,
"loss": 0.2135,
"step": 1346
},
{
"epoch": 0.858508604206501,
"grad_norm": 0.49516230821609497,
"learning_rate": 9.966063302369656e-06,
"loss": 0.1312,
"step": 1347
},
{
"epoch": 0.8591459528362014,
"grad_norm": 1.3473557233810425,
"learning_rate": 9.87798610833247e-06,
"loss": 0.2556,
"step": 1348
},
{
"epoch": 0.8597833014659019,
"grad_norm": 0.9342024326324463,
"learning_rate": 9.79027961861696e-06,
"loss": 0.1845,
"step": 1349
},
{
"epoch": 0.8604206500956023,
"grad_norm": 0.6106568574905396,
"learning_rate": 9.70294419399148e-06,
"loss": 0.1444,
"step": 1350
},
{
"epoch": 0.8610579987253028,
"grad_norm": 0.9477304220199585,
"learning_rate": 9.615980193698105e-06,
"loss": 0.2953,
"step": 1351
},
{
"epoch": 0.8616953473550032,
"grad_norm": 1.5898422002792358,
"learning_rate": 9.52938797545111e-06,
"loss": 0.6002,
"step": 1352
},
{
"epoch": 0.8623326959847036,
"grad_norm": 0.8010792136192322,
"learning_rate": 9.443167895435535e-06,
"loss": 0.2779,
"step": 1353
},
{
"epoch": 0.8629700446144041,
"grad_norm": 0.7801386117935181,
"learning_rate": 9.357320308305606e-06,
"loss": 0.1777,
"step": 1354
},
{
"epoch": 0.8636073932441045,
"grad_norm": 0.8895090818405151,
"learning_rate": 9.27184556718339e-06,
"loss": 0.2623,
"step": 1355
},
{
"epoch": 0.864244741873805,
"grad_norm": 1.5301015377044678,
"learning_rate": 9.18674402365729e-06,
"loss": 0.3295,
"step": 1356
},
{
"epoch": 0.8648820905035054,
"grad_norm": 0.8783523440361023,
"learning_rate": 9.102016027780635e-06,
"loss": 0.3335,
"step": 1357
},
{
"epoch": 0.8655194391332058,
"grad_norm": 0.7379791140556335,
"learning_rate": 9.01766192807022e-06,
"loss": 0.0994,
"step": 1358
},
{
"epoch": 0.8661567877629063,
"grad_norm": 1.0360722541809082,
"learning_rate": 8.933682071504856e-06,
"loss": 0.3078,
"step": 1359
},
{
"epoch": 0.8667941363926067,
"grad_norm": 0.8669149279594421,
"learning_rate": 8.85007680352392e-06,
"loss": 0.2435,
"step": 1360
},
{
"epoch": 0.8674314850223072,
"grad_norm": 0.429298996925354,
"learning_rate": 8.766846468026057e-06,
"loss": 0.0793,
"step": 1361
},
{
"epoch": 0.8680688336520076,
"grad_norm": 0.5863288640975952,
"learning_rate": 8.683991407367598e-06,
"loss": 0.0839,
"step": 1362
},
{
"epoch": 0.868706182281708,
"grad_norm": 1.179316520690918,
"learning_rate": 8.601511962361297e-06,
"loss": 0.3041,
"step": 1363
},
{
"epoch": 0.8693435309114086,
"grad_norm": 1.1625165939331055,
"learning_rate": 8.519408472274804e-06,
"loss": 0.2579,
"step": 1364
},
{
"epoch": 0.869980879541109,
"grad_norm": 1.415980577468872,
"learning_rate": 8.43768127482939e-06,
"loss": 0.2206,
"step": 1365
},
{
"epoch": 0.8706182281708095,
"grad_norm": 0.8789679408073425,
"learning_rate": 8.356330706198467e-06,
"loss": 0.3066,
"step": 1366
},
{
"epoch": 0.8712555768005099,
"grad_norm": 0.9820072054862976,
"learning_rate": 8.27535710100622e-06,
"loss": 0.3191,
"step": 1367
},
{
"epoch": 0.8718929254302104,
"grad_norm": 0.9121773838996887,
"learning_rate": 8.194760792326273e-06,
"loss": 0.2574,
"step": 1368
},
{
"epoch": 0.8725302740599108,
"grad_norm": 0.8674314618110657,
"learning_rate": 8.114542111680301e-06,
"loss": 0.2408,
"step": 1369
},
{
"epoch": 0.8731676226896112,
"grad_norm": 0.9724387526512146,
"learning_rate": 8.034701389036658e-06,
"loss": 0.2501,
"step": 1370
},
{
"epoch": 0.8738049713193117,
"grad_norm": 1.1526381969451904,
"learning_rate": 7.955238952808997e-06,
"loss": 0.2982,
"step": 1371
},
{
"epoch": 0.8744423199490121,
"grad_norm": 0.9064891338348389,
"learning_rate": 7.876155129854934e-06,
"loss": 0.216,
"step": 1372
},
{
"epoch": 0.8750796685787126,
"grad_norm": 0.7723839282989502,
"learning_rate": 7.797450245474747e-06,
"loss": 0.1157,
"step": 1373
},
{
"epoch": 0.875717017208413,
"grad_norm": 0.8762597441673279,
"learning_rate": 7.719124623410023e-06,
"loss": 0.1752,
"step": 1374
},
{
"epoch": 0.8763543658381134,
"grad_norm": 1.0921746492385864,
"learning_rate": 7.641178585842234e-06,
"loss": 0.3775,
"step": 1375
},
{
"epoch": 0.8769917144678139,
"grad_norm": 1.1873197555541992,
"learning_rate": 7.5636124533915444e-06,
"loss": 0.3235,
"step": 1376
},
{
"epoch": 0.8776290630975143,
"grad_norm": 1.226360559463501,
"learning_rate": 7.486426545115377e-06,
"loss": 0.4406,
"step": 1377
},
{
"epoch": 0.8782664117272148,
"grad_norm": 0.8150399327278137,
"learning_rate": 7.409621178507242e-06,
"loss": 0.1928,
"step": 1378
},
{
"epoch": 0.8789037603569152,
"grad_norm": 0.8779265880584717,
"learning_rate": 7.333196669495235e-06,
"loss": 0.1795,
"step": 1379
},
{
"epoch": 0.8795411089866156,
"grad_norm": 0.8137993812561035,
"learning_rate": 7.257153332440947e-06,
"loss": 0.1545,
"step": 1380
},
{
"epoch": 0.8801784576163161,
"grad_norm": 0.8821423649787903,
"learning_rate": 7.181491480137992e-06,
"loss": 0.1401,
"step": 1381
},
{
"epoch": 0.8808158062460165,
"grad_norm": 0.6837453246116638,
"learning_rate": 7.106211423810883e-06,
"loss": 0.1533,
"step": 1382
},
{
"epoch": 0.8814531548757171,
"grad_norm": 0.5277004837989807,
"learning_rate": 7.031313473113588e-06,
"loss": 0.109,
"step": 1383
},
{
"epoch": 0.8820905035054175,
"grad_norm": 0.885387659072876,
"learning_rate": 6.956797936128356e-06,
"loss": 0.2267,
"step": 1384
},
{
"epoch": 0.882727852135118,
"grad_norm": 0.4295061230659485,
"learning_rate": 6.882665119364485e-06,
"loss": 0.045,
"step": 1385
},
{
"epoch": 0.8833652007648184,
"grad_norm": 1.128505825996399,
"learning_rate": 6.808915327756937e-06,
"loss": 0.4093,
"step": 1386
},
{
"epoch": 0.8840025493945188,
"grad_norm": 0.6751217246055603,
"learning_rate": 6.735548864665197e-06,
"loss": 0.1345,
"step": 1387
},
{
"epoch": 0.8846398980242193,
"grad_norm": 1.0774973630905151,
"learning_rate": 6.662566031871975e-06,
"loss": 0.268,
"step": 1388
},
{
"epoch": 0.8852772466539197,
"grad_norm": 0.3923548460006714,
"learning_rate": 6.589967129581931e-06,
"loss": 0.0765,
"step": 1389
},
{
"epoch": 0.8859145952836202,
"grad_norm": 0.7430629134178162,
"learning_rate": 6.517752456420534e-06,
"loss": 0.1581,
"step": 1390
},
{
"epoch": 0.8865519439133206,
"grad_norm": 0.3530411124229431,
"learning_rate": 6.445922309432784e-06,
"loss": 0.0437,
"step": 1391
},
{
"epoch": 0.887189292543021,
"grad_norm": 1.0438884496688843,
"learning_rate": 6.3744769840819226e-06,
"loss": 0.203,
"step": 1392
},
{
"epoch": 0.8878266411727215,
"grad_norm": 0.9265220165252686,
"learning_rate": 6.30341677424835e-06,
"loss": 0.181,
"step": 1393
},
{
"epoch": 0.8884639898024219,
"grad_norm": 1.4675178527832031,
"learning_rate": 6.2327419722282956e-06,
"loss": 0.3492,
"step": 1394
},
{
"epoch": 0.8891013384321224,
"grad_norm": 0.9361581802368164,
"learning_rate": 6.162452868732693e-06,
"loss": 0.2527,
"step": 1395
},
{
"epoch": 0.8897386870618228,
"grad_norm": 1.113041877746582,
"learning_rate": 6.092549752885978e-06,
"loss": 0.3245,
"step": 1396
},
{
"epoch": 0.8903760356915232,
"grad_norm": 0.26121023297309875,
"learning_rate": 6.023032912224835e-06,
"loss": 0.0433,
"step": 1397
},
{
"epoch": 0.8910133843212237,
"grad_norm": 1.0306106805801392,
"learning_rate": 5.953902632697084e-06,
"loss": 0.3897,
"step": 1398
},
{
"epoch": 0.8916507329509241,
"grad_norm": 1.0321463346481323,
"learning_rate": 5.885159198660495e-06,
"loss": 0.2508,
"step": 1399
},
{
"epoch": 0.8922880815806246,
"grad_norm": 0.10497007519006729,
"learning_rate": 5.816802892881568e-06,
"loss": 0.0123,
"step": 1400
},
{
"epoch": 0.892925430210325,
"grad_norm": 1.3268046379089355,
"learning_rate": 5.748833996534441e-06,
"loss": 0.535,
"step": 1401
},
{
"epoch": 0.8935627788400254,
"grad_norm": 1.1545087099075317,
"learning_rate": 5.681252789199654e-06,
"loss": 0.3069,
"step": 1402
},
{
"epoch": 0.894200127469726,
"grad_norm": 1.0337591171264648,
"learning_rate": 5.614059548863093e-06,
"loss": 0.4279,
"step": 1403
},
{
"epoch": 0.8948374760994264,
"grad_norm": 0.6125165224075317,
"learning_rate": 5.547254551914771e-06,
"loss": 0.13,
"step": 1404
},
{
"epoch": 0.8954748247291269,
"grad_norm": 0.7726792097091675,
"learning_rate": 5.4808380731476874e-06,
"loss": 0.145,
"step": 1405
},
{
"epoch": 0.8961121733588273,
"grad_norm": 0.8495877981185913,
"learning_rate": 5.414810385756786e-06,
"loss": 0.2936,
"step": 1406
},
{
"epoch": 0.8967495219885278,
"grad_norm": 1.4949146509170532,
"learning_rate": 5.349171761337712e-06,
"loss": 0.3886,
"step": 1407
},
{
"epoch": 0.8973868706182282,
"grad_norm": 0.9550914764404297,
"learning_rate": 5.283922469885816e-06,
"loss": 0.3338,
"step": 1408
},
{
"epoch": 0.8980242192479286,
"grad_norm": 0.7390797734260559,
"learning_rate": 5.2190627797949255e-06,
"loss": 0.1196,
"step": 1409
},
{
"epoch": 0.8986615678776291,
"grad_norm": 0.8610605001449585,
"learning_rate": 5.154592957856319e-06,
"loss": 0.2241,
"step": 1410
},
{
"epoch": 0.8992989165073295,
"grad_norm": 0.8468586206436157,
"learning_rate": 5.090513269257613e-06,
"loss": 0.1022,
"step": 1411
},
{
"epoch": 0.89993626513703,
"grad_norm": 0.5716304779052734,
"learning_rate": 5.0268239775816735e-06,
"loss": 0.087,
"step": 1412
},
{
"epoch": 0.9005736137667304,
"grad_norm": 1.0419034957885742,
"learning_rate": 4.9635253448054974e-06,
"loss": 0.3649,
"step": 1413
},
{
"epoch": 0.9012109623964308,
"grad_norm": 0.9777461886405945,
"learning_rate": 4.900617631299165e-06,
"loss": 0.2934,
"step": 1414
},
{
"epoch": 0.9018483110261313,
"grad_norm": 1.1807531118392944,
"learning_rate": 4.838101095824776e-06,
"loss": 0.3537,
"step": 1415
},
{
"epoch": 0.9024856596558317,
"grad_norm": 0.7663743495941162,
"learning_rate": 4.775975995535387e-06,
"loss": 0.256,
"step": 1416
},
{
"epoch": 0.9031230082855322,
"grad_norm": 1.2599050998687744,
"learning_rate": 4.714242585973916e-06,
"loss": 0.2649,
"step": 1417
},
{
"epoch": 0.9037603569152326,
"grad_norm": 0.6449419260025024,
"learning_rate": 4.652901121072117e-06,
"loss": 0.0656,
"step": 1418
},
{
"epoch": 0.904397705544933,
"grad_norm": 1.3134218454360962,
"learning_rate": 4.591951853149534e-06,
"loss": 0.2475,
"step": 1419
},
{
"epoch": 0.9050350541746335,
"grad_norm": 0.7308253049850464,
"learning_rate": 4.5313950329125e-06,
"loss": 0.1268,
"step": 1420
},
{
"epoch": 0.9056724028043339,
"grad_norm": 0.6597181558609009,
"learning_rate": 4.471230909453028e-06,
"loss": 0.1786,
"step": 1421
},
{
"epoch": 0.9063097514340345,
"grad_norm": 1.1094797849655151,
"learning_rate": 4.411459730247824e-06,
"loss": 0.3357,
"step": 1422
},
{
"epoch": 0.9069471000637349,
"grad_norm": 0.7872187495231628,
"learning_rate": 4.35208174115731e-06,
"loss": 0.1597,
"step": 1423
},
{
"epoch": 0.9075844486934354,
"grad_norm": 1.3347207307815552,
"learning_rate": 4.293097186424522e-06,
"loss": 0.3862,
"step": 1424
},
{
"epoch": 0.9082217973231358,
"grad_norm": 1.0710327625274658,
"learning_rate": 4.234506308674235e-06,
"loss": 0.2792,
"step": 1425
},
{
"epoch": 0.9088591459528362,
"grad_norm": 1.4633266925811768,
"learning_rate": 4.176309348911822e-06,
"loss": 0.1935,
"step": 1426
},
{
"epoch": 0.9094964945825367,
"grad_norm": 0.5617287158966064,
"learning_rate": 4.118506546522338e-06,
"loss": 0.1447,
"step": 1427
},
{
"epoch": 0.9101338432122371,
"grad_norm": 0.9955701231956482,
"learning_rate": 4.061098139269559e-06,
"loss": 0.2185,
"step": 1428
},
{
"epoch": 0.9107711918419376,
"grad_norm": 0.6899940371513367,
"learning_rate": 4.004084363294958e-06,
"loss": 0.1181,
"step": 1429
},
{
"epoch": 0.911408540471638,
"grad_norm": 1.1276353597640991,
"learning_rate": 3.947465453116728e-06,
"loss": 0.3357,
"step": 1430
},
{
"epoch": 0.9120458891013384,
"grad_norm": 0.8343312740325928,
"learning_rate": 3.8912416416288625e-06,
"loss": 0.2647,
"step": 1431
},
{
"epoch": 0.9126832377310389,
"grad_norm": 1.138927936553955,
"learning_rate": 3.835413160100132e-06,
"loss": 0.2489,
"step": 1432
},
{
"epoch": 0.9133205863607393,
"grad_norm": 1.0664654970169067,
"learning_rate": 3.779980238173242e-06,
"loss": 0.2225,
"step": 1433
},
{
"epoch": 0.9139579349904398,
"grad_norm": 1.459571361541748,
"learning_rate": 3.724943103863754e-06,
"loss": 0.326,
"step": 1434
},
{
"epoch": 0.9145952836201402,
"grad_norm": 1.3498846292495728,
"learning_rate": 3.6703019835592433e-06,
"loss": 0.2712,
"step": 1435
},
{
"epoch": 0.9152326322498406,
"grad_norm": 1.1487019062042236,
"learning_rate": 3.6160571020183108e-06,
"loss": 0.2666,
"step": 1436
},
{
"epoch": 0.9158699808795411,
"grad_norm": 1.0268685817718506,
"learning_rate": 3.5622086823697387e-06,
"loss": 0.3231,
"step": 1437
},
{
"epoch": 0.9165073295092415,
"grad_norm": 1.0036664009094238,
"learning_rate": 3.5087569461114467e-06,
"loss": 0.3426,
"step": 1438
},
{
"epoch": 0.917144678138942,
"grad_norm": 1.1599229574203491,
"learning_rate": 3.4557021131097155e-06,
"loss": 0.3579,
"step": 1439
},
{
"epoch": 0.9177820267686424,
"grad_norm": 0.8227235674858093,
"learning_rate": 3.4030444015981767e-06,
"loss": 0.1595,
"step": 1440
},
{
"epoch": 0.9184193753983428,
"grad_norm": 0.7516672611236572,
"learning_rate": 3.35078402817699e-06,
"loss": 0.2132,
"step": 1441
},
{
"epoch": 0.9190567240280434,
"grad_norm": 0.7729995250701904,
"learning_rate": 3.2989212078119004e-06,
"loss": 0.1438,
"step": 1442
},
{
"epoch": 0.9196940726577438,
"grad_norm": 0.09384467452764511,
"learning_rate": 3.2474561538333613e-06,
"loss": 0.0096,
"step": 1443
},
{
"epoch": 0.9203314212874443,
"grad_norm": 1.3122292757034302,
"learning_rate": 3.196389077935713e-06,
"loss": 0.2874,
"step": 1444
},
{
"epoch": 0.9209687699171447,
"grad_norm": 0.6385897994041443,
"learning_rate": 3.145720190176216e-06,
"loss": 0.103,
"step": 1445
},
{
"epoch": 0.9216061185468452,
"grad_norm": 1.1529611349105835,
"learning_rate": 3.095449698974284e-06,
"loss": 0.3297,
"step": 1446
},
{
"epoch": 0.9222434671765456,
"grad_norm": 0.7778424620628357,
"learning_rate": 3.0455778111105672e-06,
"loss": 0.1372,
"step": 1447
},
{
"epoch": 0.922880815806246,
"grad_norm": 1.0398529767990112,
"learning_rate": 2.9961047317260905e-06,
"loss": 0.4004,
"step": 1448
},
{
"epoch": 0.9235181644359465,
"grad_norm": 0.5870305895805359,
"learning_rate": 2.947030664321482e-06,
"loss": 0.1392,
"step": 1449
},
{
"epoch": 0.9241555130656469,
"grad_norm": 0.8366639614105225,
"learning_rate": 2.898355810756059e-06,
"loss": 0.3177,
"step": 1450
},
{
"epoch": 0.9247928616953474,
"grad_norm": 0.5963795781135559,
"learning_rate": 2.8500803712470436e-06,
"loss": 0.1692,
"step": 1451
},
{
"epoch": 0.9254302103250478,
"grad_norm": 0.9313536882400513,
"learning_rate": 2.802204544368703e-06,
"loss": 0.2865,
"step": 1452
},
{
"epoch": 0.9260675589547482,
"grad_norm": 0.32095101475715637,
"learning_rate": 2.7547285270515643e-06,
"loss": 0.0345,
"step": 1453
},
{
"epoch": 0.9267049075844487,
"grad_norm": 0.6625826954841614,
"learning_rate": 2.7076525145816044e-06,
"loss": 0.1175,
"step": 1454
},
{
"epoch": 0.9273422562141491,
"grad_norm": 0.30104562640190125,
"learning_rate": 2.6609767005994268e-06,
"loss": 0.024,
"step": 1455
},
{
"epoch": 0.9279796048438496,
"grad_norm": 0.9450817108154297,
"learning_rate": 2.614701277099485e-06,
"loss": 0.2995,
"step": 1456
},
{
"epoch": 0.92861695347355,
"grad_norm": 1.044792652130127,
"learning_rate": 2.5688264344292613e-06,
"loss": 0.3716,
"step": 1457
},
{
"epoch": 0.9292543021032504,
"grad_norm": 1.2882393598556519,
"learning_rate": 2.5233523612885114e-06,
"loss": 0.2694,
"step": 1458
},
{
"epoch": 0.9298916507329509,
"grad_norm": 0.9697766900062561,
"learning_rate": 2.4782792447284985e-06,
"loss": 0.2698,
"step": 1459
},
{
"epoch": 0.9305289993626513,
"grad_norm": 0.8437670469284058,
"learning_rate": 2.4336072701511947e-06,
"loss": 0.2507,
"step": 1460
},
{
"epoch": 0.9311663479923518,
"grad_norm": 1.0217374563217163,
"learning_rate": 2.3893366213085244e-06,
"loss": 0.2893,
"step": 1461
},
{
"epoch": 0.9318036966220523,
"grad_norm": 0.8927786946296692,
"learning_rate": 2.3454674803016107e-06,
"loss": 0.2851,
"step": 1462
},
{
"epoch": 0.9324410452517528,
"grad_norm": 1.1168006658554077,
"learning_rate": 2.3020000275800536e-06,
"loss": 0.3148,
"step": 1463
},
{
"epoch": 0.9330783938814532,
"grad_norm": 0.7894974946975708,
"learning_rate": 2.258934441941163e-06,
"loss": 0.3083,
"step": 1464
},
{
"epoch": 0.9337157425111536,
"grad_norm": 0.7732872366905212,
"learning_rate": 2.2162709005291827e-06,
"loss": 0.2499,
"step": 1465
},
{
"epoch": 0.9343530911408541,
"grad_norm": 1.1175874471664429,
"learning_rate": 2.174009578834646e-06,
"loss": 0.2457,
"step": 1466
},
{
"epoch": 0.9349904397705545,
"grad_norm": 0.205730140209198,
"learning_rate": 2.1321506506936205e-06,
"loss": 0.0322,
"step": 1467
},
{
"epoch": 0.935627788400255,
"grad_norm": 1.1772785186767578,
"learning_rate": 2.0906942882869318e-06,
"loss": 0.5899,
"step": 1468
},
{
"epoch": 0.9362651370299554,
"grad_norm": 1.3749407529830933,
"learning_rate": 2.0496406621395624e-06,
"loss": 0.339,
"step": 1469
},
{
"epoch": 0.9369024856596558,
"grad_norm": 1.1026709079742432,
"learning_rate": 2.0089899411198433e-06,
"loss": 0.3676,
"step": 1470
},
{
"epoch": 0.9375398342893563,
"grad_norm": 0.9951978325843811,
"learning_rate": 1.9687422924388744e-06,
"loss": 0.2842,
"step": 1471
},
{
"epoch": 0.9381771829190567,
"grad_norm": 1.0075371265411377,
"learning_rate": 1.9288978816497273e-06,
"loss": 0.3074,
"step": 1472
},
{
"epoch": 0.9388145315487572,
"grad_norm": 0.9561700224876404,
"learning_rate": 1.889456872646811e-06,
"loss": 0.17,
"step": 1473
},
{
"epoch": 0.9394518801784576,
"grad_norm": 1.2524032592773438,
"learning_rate": 1.8504194276652175e-06,
"loss": 0.3405,
"step": 1474
},
{
"epoch": 0.940089228808158,
"grad_norm": 1.1642934083938599,
"learning_rate": 1.8117857072800116e-06,
"loss": 0.248,
"step": 1475
},
{
"epoch": 0.9407265774378585,
"grad_norm": 1.4066911935806274,
"learning_rate": 1.773555870405641e-06,
"loss": 0.3247,
"step": 1476
},
{
"epoch": 0.9413639260675589,
"grad_norm": 1.023539423942566,
"learning_rate": 1.7357300742951831e-06,
"loss": 0.2828,
"step": 1477
},
{
"epoch": 0.9420012746972594,
"grad_norm": 1.0929762125015259,
"learning_rate": 1.698308474539767e-06,
"loss": 0.3995,
"step": 1478
},
{
"epoch": 0.9426386233269598,
"grad_norm": 0.8823283910751343,
"learning_rate": 1.661291225067918e-06,
"loss": 0.1882,
"step": 1479
},
{
"epoch": 0.9432759719566602,
"grad_norm": 0.2362847924232483,
"learning_rate": 1.6246784781449364e-06,
"loss": 0.0272,
"step": 1480
},
{
"epoch": 0.9439133205863608,
"grad_norm": 0.7518084645271301,
"learning_rate": 1.5884703843722204e-06,
"loss": 0.212,
"step": 1481
},
{
"epoch": 0.9445506692160612,
"grad_norm": 0.7696990370750427,
"learning_rate": 1.5526670926867436e-06,
"loss": 0.2016,
"step": 1482
},
{
"epoch": 0.9451880178457617,
"grad_norm": 0.8403407335281372,
"learning_rate": 1.5172687503603011e-06,
"loss": 0.1035,
"step": 1483
},
{
"epoch": 0.9458253664754621,
"grad_norm": 1.0381487607955933,
"learning_rate": 1.4822755029990864e-06,
"loss": 0.2041,
"step": 1484
},
{
"epoch": 0.9464627151051626,
"grad_norm": 0.7745662331581116,
"learning_rate": 1.4476874945428931e-06,
"loss": 0.0891,
"step": 1485
},
{
"epoch": 0.947100063734863,
"grad_norm": 0.7510108351707458,
"learning_rate": 1.4135048672646811e-06,
"loss": 0.162,
"step": 1486
},
{
"epoch": 0.9477374123645634,
"grad_norm": 0.8483333587646484,
"learning_rate": 1.3797277617699111e-06,
"loss": 0.2336,
"step": 1487
},
{
"epoch": 0.9483747609942639,
"grad_norm": 0.766934871673584,
"learning_rate": 1.3463563169960003e-06,
"loss": 0.1634,
"step": 1488
},
{
"epoch": 0.9490121096239643,
"grad_norm": 1.3017100095748901,
"learning_rate": 1.3133906702117337e-06,
"loss": 0.2572,
"step": 1489
},
{
"epoch": 0.9496494582536648,
"grad_norm": 1.175657033920288,
"learning_rate": 1.2808309570166766e-06,
"loss": 0.5013,
"step": 1490
},
{
"epoch": 0.9502868068833652,
"grad_norm": 0.9338657259941101,
"learning_rate": 1.2486773113406624e-06,
"loss": 0.2389,
"step": 1491
},
{
"epoch": 0.9509241555130656,
"grad_norm": 1.0877935886383057,
"learning_rate": 1.2169298654432282e-06,
"loss": 0.3559,
"step": 1492
},
{
"epoch": 0.9515615041427661,
"grad_norm": 0.8545557856559753,
"learning_rate": 1.1855887499130802e-06,
"loss": 0.259,
"step": 1493
},
{
"epoch": 0.9521988527724665,
"grad_norm": 1.2284152507781982,
"learning_rate": 1.1546540936675066e-06,
"loss": 0.2525,
"step": 1494
},
{
"epoch": 0.952836201402167,
"grad_norm": 1.2084978818893433,
"learning_rate": 1.124126023951888e-06,
"loss": 0.3555,
"step": 1495
},
{
"epoch": 0.9534735500318674,
"grad_norm": 1.1898581981658936,
"learning_rate": 1.0940046663391878e-06,
"loss": 0.2715,
"step": 1496
},
{
"epoch": 0.9541108986615678,
"grad_norm": 0.8432650566101074,
"learning_rate": 1.0642901447293963e-06,
"loss": 0.156,
"step": 1497
},
{
"epoch": 0.9547482472912683,
"grad_norm": 0.7501076459884644,
"learning_rate": 1.0349825813490642e-06,
"loss": 0.1201,
"step": 1498
},
{
"epoch": 0.9553855959209687,
"grad_norm": 1.186217188835144,
"learning_rate": 1.00608209675076e-06,
"loss": 0.33,
"step": 1499
},
{
"epoch": 0.9560229445506692,
"grad_norm": 1.5549489259719849,
"learning_rate": 9.775888098125686e-07,
"loss": 0.2407,
"step": 1500
},
{
"epoch": 0.9566602931803697,
"grad_norm": 0.9755271077156067,
"learning_rate": 9.495028377376702e-07,
"loss": 0.1678,
"step": 1501
},
{
"epoch": 0.9572976418100702,
"grad_norm": 0.9037211537361145,
"learning_rate": 9.218242960537859e-07,
"loss": 0.3462,
"step": 1502
},
{
"epoch": 0.9579349904397706,
"grad_norm": 0.7009819746017456,
"learning_rate": 8.945532986127325e-07,
"loss": 0.0776,
"step": 1503
},
{
"epoch": 0.958572339069471,
"grad_norm": 0.8708709478378296,
"learning_rate": 8.676899575899455e-07,
"loss": 0.1623,
"step": 1504
},
{
"epoch": 0.9592096876991715,
"grad_norm": 0.9103490710258484,
"learning_rate": 8.412343834840242e-07,
"loss": 0.2678,
"step": 1505
},
{
"epoch": 0.9598470363288719,
"grad_norm": 1.5367730855941772,
"learning_rate": 8.151866851162982e-07,
"loss": 0.3975,
"step": 1506
},
{
"epoch": 0.9604843849585724,
"grad_norm": 1.1457476615905762,
"learning_rate": 7.895469696303171e-07,
"loss": 0.3594,
"step": 1507
},
{
"epoch": 0.9611217335882728,
"grad_norm": 0.7368289232254028,
"learning_rate": 7.643153424914729e-07,
"loss": 0.1727,
"step": 1508
},
{
"epoch": 0.9617590822179732,
"grad_norm": 1.6382460594177246,
"learning_rate": 7.394919074865669e-07,
"loss": 0.4847,
"step": 1509
},
{
"epoch": 0.9623964308476737,
"grad_norm": 0.7683085203170776,
"learning_rate": 7.150767667232994e-07,
"loss": 0.1362,
"step": 1510
},
{
"epoch": 0.9630337794773741,
"grad_norm": 1.4587337970733643,
"learning_rate": 6.910700206299581e-07,
"loss": 0.3706,
"step": 1511
},
{
"epoch": 0.9636711281070746,
"grad_norm": 1.2709547281265259,
"learning_rate": 6.674717679549414e-07,
"loss": 0.2793,
"step": 1512
},
{
"epoch": 0.964308476736775,
"grad_norm": 0.8205375075340271,
"learning_rate": 6.442821057663695e-07,
"loss": 0.2286,
"step": 1513
},
{
"epoch": 0.9649458253664754,
"grad_norm": 0.7279136776924133,
"learning_rate": 6.215011294516737e-07,
"loss": 0.1447,
"step": 1514
},
{
"epoch": 0.9655831739961759,
"grad_norm": 0.6513869166374207,
"learning_rate": 5.991289327172301e-07,
"loss": 0.1233,
"step": 1515
},
{
"epoch": 0.9662205226258763,
"grad_norm": 1.113147258758545,
"learning_rate": 5.771656075879261e-07,
"loss": 0.2089,
"step": 1516
},
{
"epoch": 0.9668578712555768,
"grad_norm": 0.6860173940658569,
"learning_rate": 5.556112444068506e-07,
"loss": 0.1151,
"step": 1517
},
{
"epoch": 0.9674952198852772,
"grad_norm": 1.0006530284881592,
"learning_rate": 5.344659318348599e-07,
"loss": 0.2394,
"step": 1518
},
{
"epoch": 0.9681325685149776,
"grad_norm": 0.9128064513206482,
"learning_rate": 5.137297568502563e-07,
"loss": 0.2196,
"step": 1519
},
{
"epoch": 0.9687699171446782,
"grad_norm": 1.1200000047683716,
"learning_rate": 4.934028047483886e-07,
"loss": 0.266,
"step": 1520
},
{
"epoch": 0.9694072657743786,
"grad_norm": 1.420006513595581,
"learning_rate": 4.7348515914134026e-07,
"loss": 0.306,
"step": 1521
},
{
"epoch": 0.9700446144040791,
"grad_norm": 0.4903380274772644,
"learning_rate": 4.539769019575646e-07,
"loss": 0.0407,
"step": 1522
},
{
"epoch": 0.9706819630337795,
"grad_norm": 0.7415797710418701,
"learning_rate": 4.3487811344155025e-07,
"loss": 0.063,
"step": 1523
},
{
"epoch": 0.97131931166348,
"grad_norm": 0.920664370059967,
"learning_rate": 4.161888721535001e-07,
"loss": 0.362,
"step": 1524
},
{
"epoch": 0.9719566602931804,
"grad_norm": 1.2313122749328613,
"learning_rate": 3.9790925496900886e-07,
"loss": 0.2827,
"step": 1525
},
{
"epoch": 0.9725940089228808,
"grad_norm": 0.6865007877349854,
"learning_rate": 3.800393370787081e-07,
"loss": 0.2067,
"step": 1526
},
{
"epoch": 0.9732313575525813,
"grad_norm": 1.1980940103530884,
"learning_rate": 3.625791919880328e-07,
"loss": 0.3535,
"step": 1527
},
{
"epoch": 0.9738687061822817,
"grad_norm": 1.1220029592514038,
"learning_rate": 3.455288915168442e-07,
"loss": 0.1572,
"step": 1528
},
{
"epoch": 0.9745060548119822,
"grad_norm": 0.74290931224823,
"learning_rate": 3.288885057991631e-07,
"loss": 0.1444,
"step": 1529
},
{
"epoch": 0.9751434034416826,
"grad_norm": 0.6867620944976807,
"learning_rate": 3.1265810328290344e-07,
"loss": 0.1439,
"step": 1530
},
{
"epoch": 0.975780752071383,
"grad_norm": 0.8704963326454163,
"learning_rate": 2.9683775072956165e-07,
"loss": 0.2038,
"step": 1531
},
{
"epoch": 0.9764181007010835,
"grad_norm": 0.4038764536380768,
"learning_rate": 2.814275132139388e-07,
"loss": 0.0261,
"step": 1532
},
{
"epoch": 0.9770554493307839,
"grad_norm": 0.7241180539131165,
"learning_rate": 2.664274541239076e-07,
"loss": 0.1688,
"step": 1533
},
{
"epoch": 0.9776927979604844,
"grad_norm": 1.3348180055618286,
"learning_rate": 2.518376351601015e-07,
"loss": 0.3095,
"step": 1534
},
{
"epoch": 0.9783301465901848,
"grad_norm": 0.7177256941795349,
"learning_rate": 2.37658116335715e-07,
"loss": 0.1417,
"step": 1535
},
{
"epoch": 0.9789674952198852,
"grad_norm": 0.9075137972831726,
"learning_rate": 2.2388895597622582e-07,
"loss": 0.2384,
"step": 1536
},
{
"epoch": 0.9796048438495857,
"grad_norm": 1.026518702507019,
"learning_rate": 2.105302107191398e-07,
"loss": 0.3419,
"step": 1537
},
{
"epoch": 0.9802421924792861,
"grad_norm": 0.664299726486206,
"learning_rate": 1.975819355137909e-07,
"loss": 0.0825,
"step": 1538
},
{
"epoch": 0.9808795411089866,
"grad_norm": 0.5427790880203247,
"learning_rate": 1.85044183621097e-07,
"loss": 0.0743,
"step": 1539
},
{
"epoch": 0.9815168897386871,
"grad_norm": 1.3215463161468506,
"learning_rate": 1.729170066133712e-07,
"loss": 0.2176,
"step": 1540
},
{
"epoch": 0.9821542383683876,
"grad_norm": 0.8941333293914795,
"learning_rate": 1.6120045437404418e-07,
"loss": 0.2315,
"step": 1541
},
{
"epoch": 0.982791586998088,
"grad_norm": 0.5719680786132812,
"learning_rate": 1.4989457509752004e-07,
"loss": 0.1479,
"step": 1542
},
{
"epoch": 0.9834289356277884,
"grad_norm": 0.8282480239868164,
"learning_rate": 1.389994152889651e-07,
"loss": 0.2575,
"step": 1543
},
{
"epoch": 0.9840662842574889,
"grad_norm": 1.1166064739227295,
"learning_rate": 1.2851501976408608e-07,
"loss": 0.2876,
"step": 1544
},
{
"epoch": 0.9847036328871893,
"grad_norm": 1.2853261232376099,
"learning_rate": 1.184414316489857e-07,
"loss": 0.2839,
"step": 1545
},
{
"epoch": 0.9853409815168898,
"grad_norm": 0.8436833620071411,
"learning_rate": 1.0877869237996275e-07,
"loss": 0.3746,
"step": 1546
},
{
"epoch": 0.9859783301465902,
"grad_norm": 0.8336836695671082,
"learning_rate": 9.95268417033457e-08,
"loss": 0.1337,
"step": 1547
},
{
"epoch": 0.9866156787762906,
"grad_norm": 0.6359622478485107,
"learning_rate": 9.068591767533718e-08,
"loss": 0.0673,
"step": 1548
},
{
"epoch": 0.9872530274059911,
"grad_norm": 1.2503803968429565,
"learning_rate": 8.225595666182528e-08,
"loss": 0.3541,
"step": 1549
},
{
"epoch": 0.9878903760356915,
"grad_norm": 0.9076665639877319,
"learning_rate": 7.423699333829471e-08,
"loss": 0.3187,
"step": 1550
},
{
"epoch": 0.988527724665392,
"grad_norm": 0.7137715816497803,
"learning_rate": 6.662906068961583e-08,
"loss": 0.1767,
"step": 1551
},
{
"epoch": 0.9891650732950924,
"grad_norm": 0.895836353302002,
"learning_rate": 5.943219000997813e-08,
"loss": 0.1525,
"step": 1552
},
{
"epoch": 0.9898024219247928,
"grad_norm": 1.2331132888793945,
"learning_rate": 5.264641090269029e-08,
"loss": 0.3631,
"step": 1553
},
{
"epoch": 0.9904397705544933,
"grad_norm": 1.351482629776001,
"learning_rate": 4.627175128010253e-08,
"loss": 0.3099,
"step": 1554
},
{
"epoch": 0.9910771191841937,
"grad_norm": 0.3359313905239105,
"learning_rate": 4.030823736350664e-08,
"loss": 0.0411,
"step": 1555
},
{
"epoch": 0.9917144678138942,
"grad_norm": 0.6981030106544495,
"learning_rate": 3.47558936829695e-08,
"loss": 0.1061,
"step": 1556
},
{
"epoch": 0.9923518164435946,
"grad_norm": 0.891771137714386,
"learning_rate": 2.961474307728862e-08,
"loss": 0.2435,
"step": 1557
},
{
"epoch": 0.992989165073295,
"grad_norm": 1.1760879755020142,
"learning_rate": 2.4884806693858952e-08,
"loss": 0.3979,
"step": 1558
},
{
"epoch": 0.9936265137029955,
"grad_norm": 1.1163259744644165,
"learning_rate": 2.0566103988617357e-08,
"loss": 0.3315,
"step": 1559
},
{
"epoch": 0.994263862332696,
"grad_norm": 0.8212835788726807,
"learning_rate": 1.6658652725953795e-08,
"loss": 0.2559,
"step": 1560
},
{
"epoch": 0.9949012109623965,
"grad_norm": 1.4443345069885254,
"learning_rate": 1.31624689786225e-08,
"loss": 0.3307,
"step": 1561
},
{
"epoch": 0.9955385595920969,
"grad_norm": 0.800864040851593,
"learning_rate": 1.007756712767538e-08,
"loss": 0.1315,
"step": 1562
},
{
"epoch": 0.9961759082217974,
"grad_norm": 0.9934842586517334,
"learning_rate": 7.4039598624287e-09,
"loss": 0.2182,
"step": 1563
},
{
"epoch": 0.9968132568514978,
"grad_norm": 0.7818678021430969,
"learning_rate": 5.141658180396469e-09,
"loss": 0.1754,
"step": 1564
},
{
"epoch": 0.9974506054811982,
"grad_norm": 1.0667723417282104,
"learning_rate": 3.290671387246036e-09,
"loss": 0.096,
"step": 1565
},
{
"epoch": 0.9980879541108987,
"grad_norm": 1.079973816871643,
"learning_rate": 1.851007096742574e-09,
"loss": 0.1026,
"step": 1566
},
{
"epoch": 0.9987253027405991,
"grad_norm": 0.47429341077804565,
"learning_rate": 8.226712307490836e-10,
"loss": 0.1051,
"step": 1567
},
{
"epoch": 0.9993626513702996,
"grad_norm": 1.3972581624984741,
"learning_rate": 2.0566801918198152e-10,
"loss": 0.5891,
"step": 1568
},
{
"epoch": 1.0,
"grad_norm": 0.9605799317359924,
"learning_rate": 0.0,
"loss": 0.3068,
"step": 1569
}
],
"logging_steps": 1,
"max_steps": 1569,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 5.607210879442944e+16,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}