Spark2Scale / checkpoint-1000 /trainer_state.json
Dohahemdann's picture
Checkpoint at step 1000
65e923f verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.6916825177243645,
"eval_steps": 500,
"global_step": 1000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0006916825177243646,
"grad_norm": 0.17447508871555328,
"learning_rate": 0.0,
"loss": 2.4721,
"step": 1
},
{
"epoch": 0.0013833650354487291,
"grad_norm": 0.24757890403270721,
"learning_rate": 1.3698630136986302e-06,
"loss": 2.7788,
"step": 2
},
{
"epoch": 0.0020750475531730937,
"grad_norm": 0.20169955492019653,
"learning_rate": 2.7397260273972604e-06,
"loss": 2.817,
"step": 3
},
{
"epoch": 0.0027667300708974583,
"grad_norm": 0.2597520053386688,
"learning_rate": 4.10958904109589e-06,
"loss": 2.8883,
"step": 4
},
{
"epoch": 0.0034584125886218224,
"grad_norm": 0.2476697564125061,
"learning_rate": 5.479452054794521e-06,
"loss": 2.5194,
"step": 5
},
{
"epoch": 0.004150095106346187,
"grad_norm": 0.19448921084403992,
"learning_rate": 6.849315068493151e-06,
"loss": 2.0166,
"step": 6
},
{
"epoch": 0.0048417776240705515,
"grad_norm": 0.30129846930503845,
"learning_rate": 8.21917808219178e-06,
"loss": 2.0576,
"step": 7
},
{
"epoch": 0.0055334601417949165,
"grad_norm": 0.29668128490448,
"learning_rate": 9.589041095890411e-06,
"loss": 2.0307,
"step": 8
},
{
"epoch": 0.006225142659519281,
"grad_norm": 0.20030352473258972,
"learning_rate": 1.0958904109589042e-05,
"loss": 2.6198,
"step": 9
},
{
"epoch": 0.006916825177243645,
"grad_norm": 0.2806167006492615,
"learning_rate": 1.2328767123287671e-05,
"loss": 2.7184,
"step": 10
},
{
"epoch": 0.00760850769496801,
"grad_norm": 0.3155820965766907,
"learning_rate": 1.3698630136986302e-05,
"loss": 2.5599,
"step": 11
},
{
"epoch": 0.008300190212692375,
"grad_norm": 0.37342140078544617,
"learning_rate": 1.5068493150684931e-05,
"loss": 1.6116,
"step": 12
},
{
"epoch": 0.008991872730416739,
"grad_norm": 0.3110700845718384,
"learning_rate": 1.643835616438356e-05,
"loss": 2.03,
"step": 13
},
{
"epoch": 0.009683555248141103,
"grad_norm": 0.22983220219612122,
"learning_rate": 1.780821917808219e-05,
"loss": 3.0022,
"step": 14
},
{
"epoch": 0.010375237765865467,
"grad_norm": 0.20027688145637512,
"learning_rate": 1.9178082191780822e-05,
"loss": 2.3026,
"step": 15
},
{
"epoch": 0.011066920283589833,
"grad_norm": 0.2387463003396988,
"learning_rate": 2.0547945205479453e-05,
"loss": 2.9633,
"step": 16
},
{
"epoch": 0.011758602801314197,
"grad_norm": 0.21235717833042145,
"learning_rate": 2.1917808219178083e-05,
"loss": 2.8526,
"step": 17
},
{
"epoch": 0.012450285319038561,
"grad_norm": 0.29500314593315125,
"learning_rate": 2.328767123287671e-05,
"loss": 2.5277,
"step": 18
},
{
"epoch": 0.013141967836762926,
"grad_norm": 0.20916298031806946,
"learning_rate": 2.4657534246575342e-05,
"loss": 2.3433,
"step": 19
},
{
"epoch": 0.01383365035448729,
"grad_norm": 0.29092636704444885,
"learning_rate": 2.6027397260273973e-05,
"loss": 2.6725,
"step": 20
},
{
"epoch": 0.014525332872211656,
"grad_norm": 0.2366064488887787,
"learning_rate": 2.7397260273972603e-05,
"loss": 2.7656,
"step": 21
},
{
"epoch": 0.01521701538993602,
"grad_norm": 0.23627543449401855,
"learning_rate": 2.8767123287671234e-05,
"loss": 2.5008,
"step": 22
},
{
"epoch": 0.015908697907660384,
"grad_norm": 0.30645817518234253,
"learning_rate": 3.0136986301369862e-05,
"loss": 1.8266,
"step": 23
},
{
"epoch": 0.01660038042538475,
"grad_norm": 0.5485689043998718,
"learning_rate": 3.1506849315068496e-05,
"loss": 2.7961,
"step": 24
},
{
"epoch": 0.017292062943109112,
"grad_norm": 0.2701663076877594,
"learning_rate": 3.287671232876712e-05,
"loss": 2.7765,
"step": 25
},
{
"epoch": 0.017983745460833478,
"grad_norm": 0.28117069602012634,
"learning_rate": 3.424657534246575e-05,
"loss": 2.7523,
"step": 26
},
{
"epoch": 0.01867542797855784,
"grad_norm": 0.38187164068222046,
"learning_rate": 3.561643835616438e-05,
"loss": 2.2493,
"step": 27
},
{
"epoch": 0.019367110496282206,
"grad_norm": 0.23057831823825836,
"learning_rate": 3.698630136986301e-05,
"loss": 2.8409,
"step": 28
},
{
"epoch": 0.020058793014006572,
"grad_norm": 0.314023494720459,
"learning_rate": 3.8356164383561644e-05,
"loss": 2.7994,
"step": 29
},
{
"epoch": 0.020750475531730934,
"grad_norm": 0.5843497514724731,
"learning_rate": 3.9726027397260274e-05,
"loss": 2.8692,
"step": 30
},
{
"epoch": 0.0214421580494553,
"grad_norm": 0.3582310974597931,
"learning_rate": 4.1095890410958905e-05,
"loss": 2.5466,
"step": 31
},
{
"epoch": 0.022133840567179666,
"grad_norm": 0.2534823715686798,
"learning_rate": 4.2465753424657536e-05,
"loss": 2.5193,
"step": 32
},
{
"epoch": 0.02282552308490403,
"grad_norm": 0.34849318861961365,
"learning_rate": 4.383561643835617e-05,
"loss": 1.7904,
"step": 33
},
{
"epoch": 0.023517205602628394,
"grad_norm": 0.35676291584968567,
"learning_rate": 4.520547945205479e-05,
"loss": 1.6514,
"step": 34
},
{
"epoch": 0.024208888120352757,
"grad_norm": 0.24594038724899292,
"learning_rate": 4.657534246575342e-05,
"loss": 2.3192,
"step": 35
},
{
"epoch": 0.024900570638077123,
"grad_norm": 0.3562544882297516,
"learning_rate": 4.794520547945205e-05,
"loss": 2.8913,
"step": 36
},
{
"epoch": 0.02559225315580149,
"grad_norm": 0.4177575409412384,
"learning_rate": 4.9315068493150684e-05,
"loss": 2.4487,
"step": 37
},
{
"epoch": 0.02628393567352585,
"grad_norm": 0.3126346170902252,
"learning_rate": 5.068493150684932e-05,
"loss": 1.4528,
"step": 38
},
{
"epoch": 0.026975618191250217,
"grad_norm": 0.30482178926467896,
"learning_rate": 5.2054794520547945e-05,
"loss": 2.6085,
"step": 39
},
{
"epoch": 0.02766730070897458,
"grad_norm": 0.31493324041366577,
"learning_rate": 5.342465753424658e-05,
"loss": 1.8078,
"step": 40
},
{
"epoch": 0.028358983226698945,
"grad_norm": 0.4693777859210968,
"learning_rate": 5.479452054794521e-05,
"loss": 1.363,
"step": 41
},
{
"epoch": 0.02905066574442331,
"grad_norm": 0.6538149118423462,
"learning_rate": 5.616438356164384e-05,
"loss": 1.2077,
"step": 42
},
{
"epoch": 0.029742348262147673,
"grad_norm": 0.32594457268714905,
"learning_rate": 5.753424657534247e-05,
"loss": 2.2301,
"step": 43
},
{
"epoch": 0.03043403077987204,
"grad_norm": 0.35883522033691406,
"learning_rate": 5.89041095890411e-05,
"loss": 2.5199,
"step": 44
},
{
"epoch": 0.0311257132975964,
"grad_norm": 0.3142642080783844,
"learning_rate": 6.0273972602739724e-05,
"loss": 2.0425,
"step": 45
},
{
"epoch": 0.03181739581532077,
"grad_norm": 0.31059974431991577,
"learning_rate": 6.164383561643835e-05,
"loss": 2.6123,
"step": 46
},
{
"epoch": 0.03250907833304513,
"grad_norm": 0.24734480679035187,
"learning_rate": 6.301369863013699e-05,
"loss": 2.5415,
"step": 47
},
{
"epoch": 0.0332007608507695,
"grad_norm": 0.3828027546405792,
"learning_rate": 6.438356164383562e-05,
"loss": 2.3858,
"step": 48
},
{
"epoch": 0.03389244336849386,
"grad_norm": 0.32175979018211365,
"learning_rate": 6.575342465753424e-05,
"loss": 2.135,
"step": 49
},
{
"epoch": 0.034584125886218224,
"grad_norm": 0.5965486764907837,
"learning_rate": 6.712328767123288e-05,
"loss": 2.2111,
"step": 50
},
{
"epoch": 0.03527580840394259,
"grad_norm": 0.5561261177062988,
"learning_rate": 6.84931506849315e-05,
"loss": 1.725,
"step": 51
},
{
"epoch": 0.035967490921666956,
"grad_norm": 0.46931296586990356,
"learning_rate": 6.986301369863014e-05,
"loss": 2.5344,
"step": 52
},
{
"epoch": 0.03665917343939132,
"grad_norm": 0.4662770926952362,
"learning_rate": 7.123287671232876e-05,
"loss": 2.4031,
"step": 53
},
{
"epoch": 0.03735085595711568,
"grad_norm": 0.3315083980560303,
"learning_rate": 7.26027397260274e-05,
"loss": 2.0916,
"step": 54
},
{
"epoch": 0.03804253847484005,
"grad_norm": 0.36718690395355225,
"learning_rate": 7.397260273972603e-05,
"loss": 2.6542,
"step": 55
},
{
"epoch": 0.03873422099256441,
"grad_norm": 0.43732333183288574,
"learning_rate": 7.534246575342466e-05,
"loss": 2.3609,
"step": 56
},
{
"epoch": 0.039425903510288775,
"grad_norm": 0.5333709716796875,
"learning_rate": 7.671232876712329e-05,
"loss": 2.0736,
"step": 57
},
{
"epoch": 0.040117586028013144,
"grad_norm": 0.609667181968689,
"learning_rate": 7.808219178082192e-05,
"loss": 1.9863,
"step": 58
},
{
"epoch": 0.040809268545737507,
"grad_norm": 0.38461530208587646,
"learning_rate": 7.945205479452055e-05,
"loss": 2.0244,
"step": 59
},
{
"epoch": 0.04150095106346187,
"grad_norm": 0.37384161353111267,
"learning_rate": 8.082191780821919e-05,
"loss": 2.5084,
"step": 60
},
{
"epoch": 0.04219263358118624,
"grad_norm": 0.575130820274353,
"learning_rate": 8.219178082191781e-05,
"loss": 1.5943,
"step": 61
},
{
"epoch": 0.0428843160989106,
"grad_norm": 0.47650450468063354,
"learning_rate": 8.356164383561645e-05,
"loss": 2.4752,
"step": 62
},
{
"epoch": 0.04357599861663496,
"grad_norm": 0.36863836646080017,
"learning_rate": 8.493150684931507e-05,
"loss": 2.5023,
"step": 63
},
{
"epoch": 0.04426768113435933,
"grad_norm": 0.3590114116668701,
"learning_rate": 8.630136986301371e-05,
"loss": 2.1982,
"step": 64
},
{
"epoch": 0.044959363652083695,
"grad_norm": 0.4362975060939789,
"learning_rate": 8.767123287671233e-05,
"loss": 2.0463,
"step": 65
},
{
"epoch": 0.04565104616980806,
"grad_norm": 0.39568379521369934,
"learning_rate": 8.904109589041096e-05,
"loss": 2.5595,
"step": 66
},
{
"epoch": 0.04634272868753242,
"grad_norm": 0.41195258498191833,
"learning_rate": 9.041095890410958e-05,
"loss": 2.348,
"step": 67
},
{
"epoch": 0.04703441120525679,
"grad_norm": 0.5394885540008545,
"learning_rate": 9.178082191780822e-05,
"loss": 1.2385,
"step": 68
},
{
"epoch": 0.04772609372298115,
"grad_norm": 0.41482311487197876,
"learning_rate": 9.315068493150684e-05,
"loss": 2.7036,
"step": 69
},
{
"epoch": 0.048417776240705514,
"grad_norm": 0.4505786597728729,
"learning_rate": 9.452054794520548e-05,
"loss": 1.7636,
"step": 70
},
{
"epoch": 0.04910945875842988,
"grad_norm": 0.26118066906929016,
"learning_rate": 9.58904109589041e-05,
"loss": 1.7056,
"step": 71
},
{
"epoch": 0.049801141276154245,
"grad_norm": 0.7240022420883179,
"learning_rate": 9.726027397260274e-05,
"loss": 2.3171,
"step": 72
},
{
"epoch": 0.05049282379387861,
"grad_norm": 0.5030315518379211,
"learning_rate": 9.863013698630137e-05,
"loss": 2.1896,
"step": 73
},
{
"epoch": 0.05118450631160298,
"grad_norm": 0.7940630316734314,
"learning_rate": 0.0001,
"loss": 1.4996,
"step": 74
},
{
"epoch": 0.05187618882932734,
"grad_norm": 0.45342186093330383,
"learning_rate": 9.992716678805537e-05,
"loss": 2.1487,
"step": 75
},
{
"epoch": 0.0525678713470517,
"grad_norm": 0.4451451301574707,
"learning_rate": 9.985433357611071e-05,
"loss": 2.1561,
"step": 76
},
{
"epoch": 0.053259553864776064,
"grad_norm": 0.39434245228767395,
"learning_rate": 9.978150036416607e-05,
"loss": 1.8114,
"step": 77
},
{
"epoch": 0.053951236382500434,
"grad_norm": 0.5808930397033691,
"learning_rate": 9.970866715222141e-05,
"loss": 2.0528,
"step": 78
},
{
"epoch": 0.054642918900224796,
"grad_norm": 0.4741731882095337,
"learning_rate": 9.963583394027677e-05,
"loss": 1.988,
"step": 79
},
{
"epoch": 0.05533460141794916,
"grad_norm": 0.38925132155418396,
"learning_rate": 9.956300072833212e-05,
"loss": 1.1629,
"step": 80
},
{
"epoch": 0.05602628393567353,
"grad_norm": 0.5677796602249146,
"learning_rate": 9.949016751638748e-05,
"loss": 1.7325,
"step": 81
},
{
"epoch": 0.05671796645339789,
"grad_norm": 0.5454609990119934,
"learning_rate": 9.941733430444284e-05,
"loss": 2.0386,
"step": 82
},
{
"epoch": 0.05740964897112225,
"grad_norm": 0.42700880765914917,
"learning_rate": 9.934450109249819e-05,
"loss": 2.1646,
"step": 83
},
{
"epoch": 0.05810133148884662,
"grad_norm": 0.4880882203578949,
"learning_rate": 9.927166788055353e-05,
"loss": 1.8193,
"step": 84
},
{
"epoch": 0.058793014006570984,
"grad_norm": 0.4928286373615265,
"learning_rate": 9.919883466860888e-05,
"loss": 1.6243,
"step": 85
},
{
"epoch": 0.05948469652429535,
"grad_norm": 0.5179658532142639,
"learning_rate": 9.912600145666424e-05,
"loss": 1.6846,
"step": 86
},
{
"epoch": 0.060176379042019716,
"grad_norm": 0.6409149765968323,
"learning_rate": 9.905316824471959e-05,
"loss": 0.8419,
"step": 87
},
{
"epoch": 0.06086806155974408,
"grad_norm": 0.7547211050987244,
"learning_rate": 9.898033503277495e-05,
"loss": 2.399,
"step": 88
},
{
"epoch": 0.06155974407746844,
"grad_norm": 0.412672758102417,
"learning_rate": 9.890750182083031e-05,
"loss": 2.2212,
"step": 89
},
{
"epoch": 0.0622514265951928,
"grad_norm": 0.5434851050376892,
"learning_rate": 9.883466860888566e-05,
"loss": 2.3083,
"step": 90
},
{
"epoch": 0.06294310911291717,
"grad_norm": 0.4428962767124176,
"learning_rate": 9.876183539694101e-05,
"loss": 2.3586,
"step": 91
},
{
"epoch": 0.06363479163064154,
"grad_norm": 0.643883466720581,
"learning_rate": 9.868900218499635e-05,
"loss": 2.4555,
"step": 92
},
{
"epoch": 0.0643264741483659,
"grad_norm": 1.8360555171966553,
"learning_rate": 9.861616897305172e-05,
"loss": 2.3463,
"step": 93
},
{
"epoch": 0.06501815666609026,
"grad_norm": 0.5649489164352417,
"learning_rate": 9.854333576110706e-05,
"loss": 1.9054,
"step": 94
},
{
"epoch": 0.06570983918381462,
"grad_norm": 0.5043310523033142,
"learning_rate": 9.847050254916242e-05,
"loss": 2.0604,
"step": 95
},
{
"epoch": 0.066401521701539,
"grad_norm": 0.49205994606018066,
"learning_rate": 9.839766933721779e-05,
"loss": 2.5182,
"step": 96
},
{
"epoch": 0.06709320421926336,
"grad_norm": 0.9012492299079895,
"learning_rate": 9.832483612527313e-05,
"loss": 1.9895,
"step": 97
},
{
"epoch": 0.06778488673698772,
"grad_norm": 0.576653778553009,
"learning_rate": 9.825200291332848e-05,
"loss": 2.5201,
"step": 98
},
{
"epoch": 0.06847656925471209,
"grad_norm": 0.4745285212993622,
"learning_rate": 9.817916970138383e-05,
"loss": 1.9348,
"step": 99
},
{
"epoch": 0.06916825177243645,
"grad_norm": 0.5546420216560364,
"learning_rate": 9.810633648943919e-05,
"loss": 1.6732,
"step": 100
},
{
"epoch": 0.06985993429016081,
"grad_norm": 0.8806080222129822,
"learning_rate": 9.803350327749454e-05,
"loss": 1.713,
"step": 101
},
{
"epoch": 0.07055161680788519,
"grad_norm": 0.6029040217399597,
"learning_rate": 9.79606700655499e-05,
"loss": 1.5178,
"step": 102
},
{
"epoch": 0.07124329932560955,
"grad_norm": 0.56731778383255,
"learning_rate": 9.788783685360526e-05,
"loss": 1.6848,
"step": 103
},
{
"epoch": 0.07193498184333391,
"grad_norm": 0.5833808183670044,
"learning_rate": 9.781500364166059e-05,
"loss": 1.703,
"step": 104
},
{
"epoch": 0.07262666436105827,
"grad_norm": 0.724471390247345,
"learning_rate": 9.774217042971595e-05,
"loss": 2.227,
"step": 105
},
{
"epoch": 0.07331834687878264,
"grad_norm": 0.722097635269165,
"learning_rate": 9.76693372177713e-05,
"loss": 1.2173,
"step": 106
},
{
"epoch": 0.074010029396507,
"grad_norm": 0.7538579702377319,
"learning_rate": 9.759650400582666e-05,
"loss": 1.8546,
"step": 107
},
{
"epoch": 0.07470171191423136,
"grad_norm": 0.7746326327323914,
"learning_rate": 9.752367079388201e-05,
"loss": 2.4067,
"step": 108
},
{
"epoch": 0.07539339443195574,
"grad_norm": 0.5197188258171082,
"learning_rate": 9.745083758193737e-05,
"loss": 1.9193,
"step": 109
},
{
"epoch": 0.0760850769496801,
"grad_norm": 0.4536001682281494,
"learning_rate": 9.737800436999273e-05,
"loss": 2.2905,
"step": 110
},
{
"epoch": 0.07677675946740446,
"grad_norm": 0.6520904302597046,
"learning_rate": 9.730517115804807e-05,
"loss": 2.012,
"step": 111
},
{
"epoch": 0.07746844198512882,
"grad_norm": 0.5670090913772583,
"learning_rate": 9.723233794610343e-05,
"loss": 2.0289,
"step": 112
},
{
"epoch": 0.07816012450285319,
"grad_norm": 0.8609398603439331,
"learning_rate": 9.715950473415877e-05,
"loss": 1.332,
"step": 113
},
{
"epoch": 0.07885180702057755,
"grad_norm": 0.5880535244941711,
"learning_rate": 9.708667152221414e-05,
"loss": 1.7267,
"step": 114
},
{
"epoch": 0.07954348953830193,
"grad_norm": 0.6090431213378906,
"learning_rate": 9.701383831026948e-05,
"loss": 1.3968,
"step": 115
},
{
"epoch": 0.08023517205602629,
"grad_norm": 0.5701532363891602,
"learning_rate": 9.694100509832484e-05,
"loss": 2.1829,
"step": 116
},
{
"epoch": 0.08092685457375065,
"grad_norm": 0.6294519305229187,
"learning_rate": 9.68681718863802e-05,
"loss": 1.7385,
"step": 117
},
{
"epoch": 0.08161853709147501,
"grad_norm": 0.5141008496284485,
"learning_rate": 9.679533867443554e-05,
"loss": 1.8417,
"step": 118
},
{
"epoch": 0.08231021960919938,
"grad_norm": 1.0640853643417358,
"learning_rate": 9.67225054624909e-05,
"loss": 2.1031,
"step": 119
},
{
"epoch": 0.08300190212692374,
"grad_norm": 0.6235936284065247,
"learning_rate": 9.664967225054625e-05,
"loss": 1.9911,
"step": 120
},
{
"epoch": 0.0836935846446481,
"grad_norm": 0.8041340112686157,
"learning_rate": 9.657683903860161e-05,
"loss": 1.9172,
"step": 121
},
{
"epoch": 0.08438526716237248,
"grad_norm": 0.6447221040725708,
"learning_rate": 9.650400582665696e-05,
"loss": 2.0157,
"step": 122
},
{
"epoch": 0.08507694968009684,
"grad_norm": 0.7299740314483643,
"learning_rate": 9.643117261471232e-05,
"loss": 1.9157,
"step": 123
},
{
"epoch": 0.0857686321978212,
"grad_norm": 0.8950564861297607,
"learning_rate": 9.635833940276767e-05,
"loss": 1.4313,
"step": 124
},
{
"epoch": 0.08646031471554556,
"grad_norm": 0.5062771439552307,
"learning_rate": 9.628550619082301e-05,
"loss": 1.7742,
"step": 125
},
{
"epoch": 0.08715199723326993,
"grad_norm": 0.7482561469078064,
"learning_rate": 9.621267297887837e-05,
"loss": 1.6122,
"step": 126
},
{
"epoch": 0.08784367975099429,
"grad_norm": 0.4244556725025177,
"learning_rate": 9.613983976693372e-05,
"loss": 1.3825,
"step": 127
},
{
"epoch": 0.08853536226871866,
"grad_norm": 0.5349308252334595,
"learning_rate": 9.606700655498908e-05,
"loss": 1.9771,
"step": 128
},
{
"epoch": 0.08922704478644303,
"grad_norm": 0.5429610013961792,
"learning_rate": 9.599417334304443e-05,
"loss": 2.5704,
"step": 129
},
{
"epoch": 0.08991872730416739,
"grad_norm": 0.4875735938549042,
"learning_rate": 9.592134013109979e-05,
"loss": 2.0568,
"step": 130
},
{
"epoch": 0.09061040982189175,
"grad_norm": 0.5956497192382812,
"learning_rate": 9.584850691915514e-05,
"loss": 2.3604,
"step": 131
},
{
"epoch": 0.09130209233961611,
"grad_norm": 1.0286632776260376,
"learning_rate": 9.577567370721049e-05,
"loss": 1.278,
"step": 132
},
{
"epoch": 0.09199377485734048,
"grad_norm": 0.5342454314231873,
"learning_rate": 9.570284049526585e-05,
"loss": 1.7806,
"step": 133
},
{
"epoch": 0.09268545737506484,
"grad_norm": 0.5517177581787109,
"learning_rate": 9.56300072833212e-05,
"loss": 2.0241,
"step": 134
},
{
"epoch": 0.09337713989278922,
"grad_norm": 0.6514694094657898,
"learning_rate": 9.555717407137656e-05,
"loss": 2.002,
"step": 135
},
{
"epoch": 0.09406882241051358,
"grad_norm": 0.6356580257415771,
"learning_rate": 9.54843408594319e-05,
"loss": 2.2756,
"step": 136
},
{
"epoch": 0.09476050492823794,
"grad_norm": 0.6672503352165222,
"learning_rate": 9.541150764748726e-05,
"loss": 1.4608,
"step": 137
},
{
"epoch": 0.0954521874459623,
"grad_norm": 0.8465030193328857,
"learning_rate": 9.533867443554261e-05,
"loss": 2.0107,
"step": 138
},
{
"epoch": 0.09614386996368667,
"grad_norm": 0.6064625978469849,
"learning_rate": 9.526584122359796e-05,
"loss": 2.1404,
"step": 139
},
{
"epoch": 0.09683555248141103,
"grad_norm": 0.6728728413581848,
"learning_rate": 9.519300801165332e-05,
"loss": 1.7548,
"step": 140
},
{
"epoch": 0.0975272349991354,
"grad_norm": 0.7992367744445801,
"learning_rate": 9.512017479970867e-05,
"loss": 1.5071,
"step": 141
},
{
"epoch": 0.09821891751685977,
"grad_norm": 0.466144859790802,
"learning_rate": 9.504734158776403e-05,
"loss": 1.5908,
"step": 142
},
{
"epoch": 0.09891060003458413,
"grad_norm": 0.694807767868042,
"learning_rate": 9.497450837581938e-05,
"loss": 1.6261,
"step": 143
},
{
"epoch": 0.09960228255230849,
"grad_norm": 0.5561632513999939,
"learning_rate": 9.490167516387472e-05,
"loss": 1.2933,
"step": 144
},
{
"epoch": 0.10029396507003285,
"grad_norm": 0.6609554886817932,
"learning_rate": 9.482884195193008e-05,
"loss": 1.7889,
"step": 145
},
{
"epoch": 0.10098564758775722,
"grad_norm": 0.9272559285163879,
"learning_rate": 9.475600873998543e-05,
"loss": 1.2951,
"step": 146
},
{
"epoch": 0.10167733010548158,
"grad_norm": 1.5346007347106934,
"learning_rate": 9.46831755280408e-05,
"loss": 1.5395,
"step": 147
},
{
"epoch": 0.10236901262320595,
"grad_norm": 0.8031420111656189,
"learning_rate": 9.461034231609614e-05,
"loss": 2.2035,
"step": 148
},
{
"epoch": 0.10306069514093032,
"grad_norm": 0.5305848121643066,
"learning_rate": 9.45375091041515e-05,
"loss": 1.784,
"step": 149
},
{
"epoch": 0.10375237765865468,
"grad_norm": 0.6742457151412964,
"learning_rate": 9.446467589220685e-05,
"loss": 2.0042,
"step": 150
},
{
"epoch": 0.10444406017637904,
"grad_norm": 0.7282371520996094,
"learning_rate": 9.43918426802622e-05,
"loss": 2.2441,
"step": 151
},
{
"epoch": 0.1051357426941034,
"grad_norm": 0.5512478351593018,
"learning_rate": 9.431900946831756e-05,
"loss": 2.0307,
"step": 152
},
{
"epoch": 0.10582742521182777,
"grad_norm": 0.7224892377853394,
"learning_rate": 9.42461762563729e-05,
"loss": 2.4692,
"step": 153
},
{
"epoch": 0.10651910772955213,
"grad_norm": 0.7150198817253113,
"learning_rate": 9.417334304442827e-05,
"loss": 2.0194,
"step": 154
},
{
"epoch": 0.1072107902472765,
"grad_norm": 0.5024279356002808,
"learning_rate": 9.410050983248361e-05,
"loss": 2.0461,
"step": 155
},
{
"epoch": 0.10790247276500087,
"grad_norm": 1.0661766529083252,
"learning_rate": 9.402767662053898e-05,
"loss": 1.3458,
"step": 156
},
{
"epoch": 0.10859415528272523,
"grad_norm": 0.577153205871582,
"learning_rate": 9.395484340859432e-05,
"loss": 1.517,
"step": 157
},
{
"epoch": 0.10928583780044959,
"grad_norm": 0.6876221895217896,
"learning_rate": 9.388201019664967e-05,
"loss": 1.9622,
"step": 158
},
{
"epoch": 0.10997752031817395,
"grad_norm": 0.5576820373535156,
"learning_rate": 9.380917698470503e-05,
"loss": 1.6063,
"step": 159
},
{
"epoch": 0.11066920283589832,
"grad_norm": 0.8604760766029358,
"learning_rate": 9.373634377276038e-05,
"loss": 1.8693,
"step": 160
},
{
"epoch": 0.1113608853536227,
"grad_norm": 0.6998944282531738,
"learning_rate": 9.366351056081574e-05,
"loss": 1.8942,
"step": 161
},
{
"epoch": 0.11205256787134706,
"grad_norm": 0.7106878757476807,
"learning_rate": 9.359067734887109e-05,
"loss": 1.9999,
"step": 162
},
{
"epoch": 0.11274425038907142,
"grad_norm": 1.167450189590454,
"learning_rate": 9.351784413692645e-05,
"loss": 1.26,
"step": 163
},
{
"epoch": 0.11343593290679578,
"grad_norm": 0.6208682060241699,
"learning_rate": 9.34450109249818e-05,
"loss": 1.7881,
"step": 164
},
{
"epoch": 0.11412761542452014,
"grad_norm": 0.7046381235122681,
"learning_rate": 9.337217771303714e-05,
"loss": 1.7403,
"step": 165
},
{
"epoch": 0.1148192979422445,
"grad_norm": 0.6957246661186218,
"learning_rate": 9.32993445010925e-05,
"loss": 2.1148,
"step": 166
},
{
"epoch": 0.11551098045996887,
"grad_norm": 0.6169411540031433,
"learning_rate": 9.322651128914785e-05,
"loss": 1.9182,
"step": 167
},
{
"epoch": 0.11620266297769324,
"grad_norm": 0.8539403080940247,
"learning_rate": 9.315367807720321e-05,
"loss": 1.7015,
"step": 168
},
{
"epoch": 0.1168943454954176,
"grad_norm": 1.037312626838684,
"learning_rate": 9.308084486525856e-05,
"loss": 1.3205,
"step": 169
},
{
"epoch": 0.11758602801314197,
"grad_norm": 0.5971755981445312,
"learning_rate": 9.300801165331392e-05,
"loss": 1.1845,
"step": 170
},
{
"epoch": 0.11827771053086633,
"grad_norm": 0.5809158086776733,
"learning_rate": 9.293517844136927e-05,
"loss": 1.5852,
"step": 171
},
{
"epoch": 0.1189693930485907,
"grad_norm": 0.6503626704216003,
"learning_rate": 9.286234522942462e-05,
"loss": 1.9045,
"step": 172
},
{
"epoch": 0.11966107556631506,
"grad_norm": 0.5208730101585388,
"learning_rate": 9.278951201747998e-05,
"loss": 1.7599,
"step": 173
},
{
"epoch": 0.12035275808403943,
"grad_norm": 0.76451176404953,
"learning_rate": 9.271667880553533e-05,
"loss": 2.2624,
"step": 174
},
{
"epoch": 0.1210444406017638,
"grad_norm": 0.5927959084510803,
"learning_rate": 9.264384559359069e-05,
"loss": 1.5911,
"step": 175
},
{
"epoch": 0.12173612311948816,
"grad_norm": 0.6512097716331482,
"learning_rate": 9.257101238164603e-05,
"loss": 1.3409,
"step": 176
},
{
"epoch": 0.12242780563721252,
"grad_norm": 0.4837232530117035,
"learning_rate": 9.24981791697014e-05,
"loss": 1.2417,
"step": 177
},
{
"epoch": 0.12311948815493688,
"grad_norm": 0.6251150369644165,
"learning_rate": 9.242534595775674e-05,
"loss": 2.1377,
"step": 178
},
{
"epoch": 0.12381117067266124,
"grad_norm": 0.5964468717575073,
"learning_rate": 9.235251274581209e-05,
"loss": 1.616,
"step": 179
},
{
"epoch": 0.1245028531903856,
"grad_norm": 0.8185293674468994,
"learning_rate": 9.227967953386745e-05,
"loss": 2.372,
"step": 180
},
{
"epoch": 0.12519453570810998,
"grad_norm": 0.7391765117645264,
"learning_rate": 9.22068463219228e-05,
"loss": 2.0473,
"step": 181
},
{
"epoch": 0.12588621822583435,
"grad_norm": 0.6867517828941345,
"learning_rate": 9.213401310997816e-05,
"loss": 2.2592,
"step": 182
},
{
"epoch": 0.1265779007435587,
"grad_norm": 0.6889393329620361,
"learning_rate": 9.206117989803351e-05,
"loss": 1.8863,
"step": 183
},
{
"epoch": 0.12726958326128307,
"grad_norm": 0.6404229402542114,
"learning_rate": 9.198834668608885e-05,
"loss": 1.5859,
"step": 184
},
{
"epoch": 0.12796126577900743,
"grad_norm": 2.46905255317688,
"learning_rate": 9.191551347414422e-05,
"loss": 2.4798,
"step": 185
},
{
"epoch": 0.1286529482967318,
"grad_norm": 0.7490037083625793,
"learning_rate": 9.184268026219956e-05,
"loss": 2.1623,
"step": 186
},
{
"epoch": 0.12934463081445616,
"grad_norm": 0.5947558283805847,
"learning_rate": 9.176984705025492e-05,
"loss": 2.2375,
"step": 187
},
{
"epoch": 0.13003631333218052,
"grad_norm": 0.4829116463661194,
"learning_rate": 9.169701383831027e-05,
"loss": 1.66,
"step": 188
},
{
"epoch": 0.13072799584990488,
"grad_norm": 0.5807337164878845,
"learning_rate": 9.162418062636563e-05,
"loss": 1.3098,
"step": 189
},
{
"epoch": 0.13141967836762924,
"grad_norm": 0.8253926038742065,
"learning_rate": 9.155134741442098e-05,
"loss": 1.135,
"step": 190
},
{
"epoch": 0.13211136088535363,
"grad_norm": 0.7517785429954529,
"learning_rate": 9.147851420247633e-05,
"loss": 1.6342,
"step": 191
},
{
"epoch": 0.132803043403078,
"grad_norm": 0.791852593421936,
"learning_rate": 9.140568099053169e-05,
"loss": 1.3013,
"step": 192
},
{
"epoch": 0.13349472592080236,
"grad_norm": 0.6530910730361938,
"learning_rate": 9.133284777858704e-05,
"loss": 2.4537,
"step": 193
},
{
"epoch": 0.13418640843852672,
"grad_norm": 0.8071674704551697,
"learning_rate": 9.12600145666424e-05,
"loss": 1.1286,
"step": 194
},
{
"epoch": 0.13487809095625108,
"grad_norm": 0.5800924301147461,
"learning_rate": 9.118718135469774e-05,
"loss": 1.913,
"step": 195
},
{
"epoch": 0.13556977347397545,
"grad_norm": 0.6227284073829651,
"learning_rate": 9.11143481427531e-05,
"loss": 2.1569,
"step": 196
},
{
"epoch": 0.1362614559916998,
"grad_norm": 0.7267847657203674,
"learning_rate": 9.104151493080845e-05,
"loss": 2.4129,
"step": 197
},
{
"epoch": 0.13695313850942417,
"grad_norm": 0.6935005784034729,
"learning_rate": 9.09686817188638e-05,
"loss": 1.2993,
"step": 198
},
{
"epoch": 0.13764482102714853,
"grad_norm": 0.5625126957893372,
"learning_rate": 9.089584850691916e-05,
"loss": 1.6137,
"step": 199
},
{
"epoch": 0.1383365035448729,
"grad_norm": 0.6450534462928772,
"learning_rate": 9.082301529497451e-05,
"loss": 2.0705,
"step": 200
},
{
"epoch": 0.13902818606259726,
"grad_norm": 0.7700170874595642,
"learning_rate": 9.075018208302987e-05,
"loss": 1.2858,
"step": 201
},
{
"epoch": 0.13971986858032162,
"grad_norm": 0.4889370799064636,
"learning_rate": 9.067734887108522e-05,
"loss": 1.1423,
"step": 202
},
{
"epoch": 0.14041155109804598,
"grad_norm": 0.5464507341384888,
"learning_rate": 9.060451565914058e-05,
"loss": 1.0928,
"step": 203
},
{
"epoch": 0.14110323361577037,
"grad_norm": 0.7002052664756775,
"learning_rate": 9.053168244719593e-05,
"loss": 2.154,
"step": 204
},
{
"epoch": 0.14179491613349474,
"grad_norm": 0.73245769739151,
"learning_rate": 9.045884923525127e-05,
"loss": 2.3437,
"step": 205
},
{
"epoch": 0.1424865986512191,
"grad_norm": 0.6388504505157471,
"learning_rate": 9.038601602330664e-05,
"loss": 1.6209,
"step": 206
},
{
"epoch": 0.14317828116894346,
"grad_norm": 0.5278307199478149,
"learning_rate": 9.031318281136198e-05,
"loss": 1.1904,
"step": 207
},
{
"epoch": 0.14386996368666782,
"grad_norm": 0.6711840033531189,
"learning_rate": 9.024034959941734e-05,
"loss": 1.7129,
"step": 208
},
{
"epoch": 0.14456164620439219,
"grad_norm": 0.6358078122138977,
"learning_rate": 9.016751638747269e-05,
"loss": 1.6655,
"step": 209
},
{
"epoch": 0.14525332872211655,
"grad_norm": 0.6745946407318115,
"learning_rate": 9.009468317552805e-05,
"loss": 1.8217,
"step": 210
},
{
"epoch": 0.1459450112398409,
"grad_norm": 2.227377414703369,
"learning_rate": 9.00218499635834e-05,
"loss": 3.1095,
"step": 211
},
{
"epoch": 0.14663669375756527,
"grad_norm": 0.5836493968963623,
"learning_rate": 8.994901675163875e-05,
"loss": 1.9259,
"step": 212
},
{
"epoch": 0.14732837627528964,
"grad_norm": 0.6573939323425293,
"learning_rate": 8.987618353969411e-05,
"loss": 1.9532,
"step": 213
},
{
"epoch": 0.148020058793014,
"grad_norm": 0.8865386843681335,
"learning_rate": 8.980335032774946e-05,
"loss": 1.7339,
"step": 214
},
{
"epoch": 0.14871174131073836,
"grad_norm": 0.7364823818206787,
"learning_rate": 8.973051711580482e-05,
"loss": 1.2418,
"step": 215
},
{
"epoch": 0.14940342382846272,
"grad_norm": 0.7573681473731995,
"learning_rate": 8.965768390386016e-05,
"loss": 1.5159,
"step": 216
},
{
"epoch": 0.1500951063461871,
"grad_norm": 0.6876881718635559,
"learning_rate": 8.958485069191553e-05,
"loss": 2.2091,
"step": 217
},
{
"epoch": 0.15078678886391148,
"grad_norm": 0.6091864705085754,
"learning_rate": 8.951201747997087e-05,
"loss": 1.2208,
"step": 218
},
{
"epoch": 0.15147847138163584,
"grad_norm": 0.7913519740104675,
"learning_rate": 8.943918426802622e-05,
"loss": 2.3047,
"step": 219
},
{
"epoch": 0.1521701538993602,
"grad_norm": 1.3490211963653564,
"learning_rate": 8.936635105608158e-05,
"loss": 1.7839,
"step": 220
},
{
"epoch": 0.15286183641708456,
"grad_norm": 0.8339107036590576,
"learning_rate": 8.929351784413693e-05,
"loss": 2.1117,
"step": 221
},
{
"epoch": 0.15355351893480892,
"grad_norm": 1.5490026473999023,
"learning_rate": 8.922068463219229e-05,
"loss": 1.8909,
"step": 222
},
{
"epoch": 0.1542452014525333,
"grad_norm": 0.7403117418289185,
"learning_rate": 8.914785142024764e-05,
"loss": 1.838,
"step": 223
},
{
"epoch": 0.15493688397025765,
"grad_norm": 0.9296958446502686,
"learning_rate": 8.907501820830299e-05,
"loss": 1.8148,
"step": 224
},
{
"epoch": 0.155628566487982,
"grad_norm": 0.6639947891235352,
"learning_rate": 8.900218499635835e-05,
"loss": 1.8576,
"step": 225
},
{
"epoch": 0.15632024900570637,
"grad_norm": 0.7341317534446716,
"learning_rate": 8.89293517844137e-05,
"loss": 1.9323,
"step": 226
},
{
"epoch": 0.15701193152343074,
"grad_norm": 0.5162379145622253,
"learning_rate": 8.885651857246906e-05,
"loss": 2.1562,
"step": 227
},
{
"epoch": 0.1577036140411551,
"grad_norm": 0.6806768774986267,
"learning_rate": 8.87836853605244e-05,
"loss": 1.8393,
"step": 228
},
{
"epoch": 0.15839529655887946,
"grad_norm": 0.5701708197593689,
"learning_rate": 8.871085214857976e-05,
"loss": 1.4365,
"step": 229
},
{
"epoch": 0.15908697907660385,
"grad_norm": 0.7863550186157227,
"learning_rate": 8.863801893663511e-05,
"loss": 2.071,
"step": 230
},
{
"epoch": 0.15977866159432821,
"grad_norm": 0.6164423823356628,
"learning_rate": 8.856518572469046e-05,
"loss": 2.0573,
"step": 231
},
{
"epoch": 0.16047034411205258,
"grad_norm": 0.8759453892707825,
"learning_rate": 8.849235251274582e-05,
"loss": 1.8708,
"step": 232
},
{
"epoch": 0.16116202662977694,
"grad_norm": 1.3226007223129272,
"learning_rate": 8.841951930080117e-05,
"loss": 1.4042,
"step": 233
},
{
"epoch": 0.1618537091475013,
"grad_norm": 0.5054848790168762,
"learning_rate": 8.834668608885653e-05,
"loss": 1.2018,
"step": 234
},
{
"epoch": 0.16254539166522566,
"grad_norm": 0.6680052876472473,
"learning_rate": 8.827385287691188e-05,
"loss": 2.1112,
"step": 235
},
{
"epoch": 0.16323707418295003,
"grad_norm": 0.838471531867981,
"learning_rate": 8.820101966496724e-05,
"loss": 2.3499,
"step": 236
},
{
"epoch": 0.1639287567006744,
"grad_norm": 0.6450280547142029,
"learning_rate": 8.812818645302258e-05,
"loss": 1.5009,
"step": 237
},
{
"epoch": 0.16462043921839875,
"grad_norm": 0.5540733933448792,
"learning_rate": 8.805535324107793e-05,
"loss": 1.7475,
"step": 238
},
{
"epoch": 0.1653121217361231,
"grad_norm": 0.6765146255493164,
"learning_rate": 8.798252002913329e-05,
"loss": 2.0798,
"step": 239
},
{
"epoch": 0.16600380425384748,
"grad_norm": 0.6925728917121887,
"learning_rate": 8.790968681718864e-05,
"loss": 1.9599,
"step": 240
},
{
"epoch": 0.16669548677157184,
"grad_norm": 0.787634015083313,
"learning_rate": 8.7836853605244e-05,
"loss": 1.5637,
"step": 241
},
{
"epoch": 0.1673871692892962,
"grad_norm": 0.7191415429115295,
"learning_rate": 8.776402039329935e-05,
"loss": 2.1707,
"step": 242
},
{
"epoch": 0.1680788518070206,
"grad_norm": 0.7958317995071411,
"learning_rate": 8.769118718135471e-05,
"loss": 2.147,
"step": 243
},
{
"epoch": 0.16877053432474495,
"grad_norm": 0.6804454922676086,
"learning_rate": 8.761835396941004e-05,
"loss": 1.9832,
"step": 244
},
{
"epoch": 0.16946221684246932,
"grad_norm": 0.8284922242164612,
"learning_rate": 8.75455207574654e-05,
"loss": 1.2555,
"step": 245
},
{
"epoch": 0.17015389936019368,
"grad_norm": 0.7180774211883545,
"learning_rate": 8.747268754552075e-05,
"loss": 1.8228,
"step": 246
},
{
"epoch": 0.17084558187791804,
"grad_norm": 0.8632107377052307,
"learning_rate": 8.739985433357611e-05,
"loss": 1.1625,
"step": 247
},
{
"epoch": 0.1715372643956424,
"grad_norm": 0.5004351139068604,
"learning_rate": 8.732702112163147e-05,
"loss": 1.6948,
"step": 248
},
{
"epoch": 0.17222894691336676,
"grad_norm": 0.6219859719276428,
"learning_rate": 8.725418790968682e-05,
"loss": 1.6748,
"step": 249
},
{
"epoch": 0.17292062943109113,
"grad_norm": 0.8690136075019836,
"learning_rate": 8.718135469774218e-05,
"loss": 1.7746,
"step": 250
},
{
"epoch": 0.1736123119488155,
"grad_norm": 0.6689561605453491,
"learning_rate": 8.710852148579752e-05,
"loss": 1.1883,
"step": 251
},
{
"epoch": 0.17430399446653985,
"grad_norm": 1.0592402219772339,
"learning_rate": 8.703568827385288e-05,
"loss": 1.6322,
"step": 252
},
{
"epoch": 0.17499567698426421,
"grad_norm": 0.721960723400116,
"learning_rate": 8.696285506190823e-05,
"loss": 1.8509,
"step": 253
},
{
"epoch": 0.17568735950198858,
"grad_norm": 0.7862136960029602,
"learning_rate": 8.689002184996359e-05,
"loss": 1.5465,
"step": 254
},
{
"epoch": 0.17637904201971294,
"grad_norm": 0.62255859375,
"learning_rate": 8.681718863801895e-05,
"loss": 1.7961,
"step": 255
},
{
"epoch": 0.17707072453743733,
"grad_norm": 0.8009974956512451,
"learning_rate": 8.67443554260743e-05,
"loss": 1.7475,
"step": 256
},
{
"epoch": 0.1777624070551617,
"grad_norm": 0.6300957202911377,
"learning_rate": 8.667152221412966e-05,
"loss": 1.8628,
"step": 257
},
{
"epoch": 0.17845408957288605,
"grad_norm": 0.5678866505622864,
"learning_rate": 8.659868900218499e-05,
"loss": 1.2443,
"step": 258
},
{
"epoch": 0.17914577209061042,
"grad_norm": 0.6319582462310791,
"learning_rate": 8.652585579024035e-05,
"loss": 1.3356,
"step": 259
},
{
"epoch": 0.17983745460833478,
"grad_norm": 0.5843179821968079,
"learning_rate": 8.64530225782957e-05,
"loss": 1.9961,
"step": 260
},
{
"epoch": 0.18052913712605914,
"grad_norm": 0.6211555600166321,
"learning_rate": 8.638018936635106e-05,
"loss": 1.1243,
"step": 261
},
{
"epoch": 0.1812208196437835,
"grad_norm": 0.7266420722007751,
"learning_rate": 8.630735615440642e-05,
"loss": 2.2338,
"step": 262
},
{
"epoch": 0.18191250216150787,
"grad_norm": 0.8187587857246399,
"learning_rate": 8.623452294246177e-05,
"loss": 2.0237,
"step": 263
},
{
"epoch": 0.18260418467923223,
"grad_norm": 0.9444418549537659,
"learning_rate": 8.616168973051712e-05,
"loss": 1.6278,
"step": 264
},
{
"epoch": 0.1832958671969566,
"grad_norm": 0.5889193415641785,
"learning_rate": 8.608885651857246e-05,
"loss": 2.1453,
"step": 265
},
{
"epoch": 0.18398754971468095,
"grad_norm": 0.6819151639938354,
"learning_rate": 8.601602330662782e-05,
"loss": 1.2219,
"step": 266
},
{
"epoch": 0.18467923223240532,
"grad_norm": 0.8448613286018372,
"learning_rate": 8.594319009468317e-05,
"loss": 2.1193,
"step": 267
},
{
"epoch": 0.18537091475012968,
"grad_norm": 0.7247136235237122,
"learning_rate": 8.587035688273853e-05,
"loss": 1.6155,
"step": 268
},
{
"epoch": 0.18606259726785407,
"grad_norm": 0.7371733784675598,
"learning_rate": 8.57975236707939e-05,
"loss": 1.5466,
"step": 269
},
{
"epoch": 0.18675427978557843,
"grad_norm": 0.6676841974258423,
"learning_rate": 8.572469045884924e-05,
"loss": 2.1909,
"step": 270
},
{
"epoch": 0.1874459623033028,
"grad_norm": 0.7091754078865051,
"learning_rate": 8.565185724690459e-05,
"loss": 2.3795,
"step": 271
},
{
"epoch": 0.18813764482102716,
"grad_norm": 0.7999275922775269,
"learning_rate": 8.557902403495994e-05,
"loss": 2.0659,
"step": 272
},
{
"epoch": 0.18882932733875152,
"grad_norm": 0.5176097750663757,
"learning_rate": 8.55061908230153e-05,
"loss": 1.4036,
"step": 273
},
{
"epoch": 0.18952100985647588,
"grad_norm": 0.6579016447067261,
"learning_rate": 8.543335761107065e-05,
"loss": 1.9072,
"step": 274
},
{
"epoch": 0.19021269237420024,
"grad_norm": 1.0604159832000732,
"learning_rate": 8.536052439912601e-05,
"loss": 2.1682,
"step": 275
},
{
"epoch": 0.1909043748919246,
"grad_norm": 0.7157823443412781,
"learning_rate": 8.528769118718137e-05,
"loss": 1.9188,
"step": 276
},
{
"epoch": 0.19159605740964897,
"grad_norm": 0.691941499710083,
"learning_rate": 8.521485797523672e-05,
"loss": 1.8626,
"step": 277
},
{
"epoch": 0.19228773992737333,
"grad_norm": 0.5831330418586731,
"learning_rate": 8.514202476329206e-05,
"loss": 1.9975,
"step": 278
},
{
"epoch": 0.1929794224450977,
"grad_norm": 0.6630810499191284,
"learning_rate": 8.506919155134741e-05,
"loss": 2.0809,
"step": 279
},
{
"epoch": 0.19367110496282205,
"grad_norm": 0.601836621761322,
"learning_rate": 8.499635833940277e-05,
"loss": 1.2491,
"step": 280
},
{
"epoch": 0.19436278748054642,
"grad_norm": 0.7532071471214294,
"learning_rate": 8.492352512745812e-05,
"loss": 2.0672,
"step": 281
},
{
"epoch": 0.1950544699982708,
"grad_norm": 0.70450758934021,
"learning_rate": 8.485069191551348e-05,
"loss": 2.1466,
"step": 282
},
{
"epoch": 0.19574615251599517,
"grad_norm": 0.6915740966796875,
"learning_rate": 8.477785870356884e-05,
"loss": 1.4086,
"step": 283
},
{
"epoch": 0.19643783503371953,
"grad_norm": 0.949400007724762,
"learning_rate": 8.470502549162418e-05,
"loss": 1.6344,
"step": 284
},
{
"epoch": 0.1971295175514439,
"grad_norm": 0.7051184177398682,
"learning_rate": 8.463219227967954e-05,
"loss": 1.4375,
"step": 285
},
{
"epoch": 0.19782120006916826,
"grad_norm": 0.5827087163925171,
"learning_rate": 8.455935906773488e-05,
"loss": 1.4801,
"step": 286
},
{
"epoch": 0.19851288258689262,
"grad_norm": 0.6058622002601624,
"learning_rate": 8.448652585579024e-05,
"loss": 1.4378,
"step": 287
},
{
"epoch": 0.19920456510461698,
"grad_norm": 0.7779908180236816,
"learning_rate": 8.441369264384559e-05,
"loss": 1.6377,
"step": 288
},
{
"epoch": 0.19989624762234134,
"grad_norm": 0.6972940564155579,
"learning_rate": 8.434085943190095e-05,
"loss": 1.9437,
"step": 289
},
{
"epoch": 0.2005879301400657,
"grad_norm": 0.7289730906486511,
"learning_rate": 8.426802621995631e-05,
"loss": 1.5402,
"step": 290
},
{
"epoch": 0.20127961265779007,
"grad_norm": 0.6566206216812134,
"learning_rate": 8.419519300801165e-05,
"loss": 1.4304,
"step": 291
},
{
"epoch": 0.20197129517551443,
"grad_norm": 0.97121262550354,
"learning_rate": 8.412235979606701e-05,
"loss": 1.7801,
"step": 292
},
{
"epoch": 0.2026629776932388,
"grad_norm": 0.7539506554603577,
"learning_rate": 8.404952658412236e-05,
"loss": 1.5326,
"step": 293
},
{
"epoch": 0.20335466021096316,
"grad_norm": 0.8037034869194031,
"learning_rate": 8.397669337217772e-05,
"loss": 1.5559,
"step": 294
},
{
"epoch": 0.20404634272868752,
"grad_norm": 1.2110254764556885,
"learning_rate": 8.390386016023307e-05,
"loss": 1.7529,
"step": 295
},
{
"epoch": 0.2047380252464119,
"grad_norm": 0.9396728873252869,
"learning_rate": 8.383102694828843e-05,
"loss": 2.0101,
"step": 296
},
{
"epoch": 0.20542970776413627,
"grad_norm": 0.7678546905517578,
"learning_rate": 8.375819373634379e-05,
"loss": 2.0231,
"step": 297
},
{
"epoch": 0.20612139028186063,
"grad_norm": 1.1070085763931274,
"learning_rate": 8.368536052439912e-05,
"loss": 1.7999,
"step": 298
},
{
"epoch": 0.206813072799585,
"grad_norm": 0.7589883804321289,
"learning_rate": 8.361252731245448e-05,
"loss": 1.6723,
"step": 299
},
{
"epoch": 0.20750475531730936,
"grad_norm": 0.6468162536621094,
"learning_rate": 8.353969410050983e-05,
"loss": 1.3266,
"step": 300
},
{
"epoch": 0.20819643783503372,
"grad_norm": 0.7720943689346313,
"learning_rate": 8.346686088856519e-05,
"loss": 1.5897,
"step": 301
},
{
"epoch": 0.20888812035275808,
"grad_norm": 0.7349941730499268,
"learning_rate": 8.339402767662054e-05,
"loss": 1.9242,
"step": 302
},
{
"epoch": 0.20957980287048245,
"grad_norm": 0.7902266979217529,
"learning_rate": 8.33211944646759e-05,
"loss": 2.2367,
"step": 303
},
{
"epoch": 0.2102714853882068,
"grad_norm": 0.822661280632019,
"learning_rate": 8.324836125273125e-05,
"loss": 2.02,
"step": 304
},
{
"epoch": 0.21096316790593117,
"grad_norm": 0.6001129746437073,
"learning_rate": 8.31755280407866e-05,
"loss": 1.4506,
"step": 305
},
{
"epoch": 0.21165485042365553,
"grad_norm": 0.7674906253814697,
"learning_rate": 8.310269482884196e-05,
"loss": 1.9657,
"step": 306
},
{
"epoch": 0.2123465329413799,
"grad_norm": 0.5808192491531372,
"learning_rate": 8.30298616168973e-05,
"loss": 1.8088,
"step": 307
},
{
"epoch": 0.21303821545910426,
"grad_norm": 0.7100119590759277,
"learning_rate": 8.295702840495266e-05,
"loss": 1.9353,
"step": 308
},
{
"epoch": 0.21372989797682865,
"grad_norm": 0.7359102964401245,
"learning_rate": 8.288419519300801e-05,
"loss": 1.7844,
"step": 309
},
{
"epoch": 0.214421580494553,
"grad_norm": 0.7489526867866516,
"learning_rate": 8.281136198106337e-05,
"loss": 1.6238,
"step": 310
},
{
"epoch": 0.21511326301227737,
"grad_norm": 0.8063543438911438,
"learning_rate": 8.273852876911872e-05,
"loss": 1.6634,
"step": 311
},
{
"epoch": 0.21580494553000173,
"grad_norm": 0.7559211850166321,
"learning_rate": 8.266569555717407e-05,
"loss": 1.9663,
"step": 312
},
{
"epoch": 0.2164966280477261,
"grad_norm": 0.7710449695587158,
"learning_rate": 8.259286234522943e-05,
"loss": 2.0552,
"step": 313
},
{
"epoch": 0.21718831056545046,
"grad_norm": 1.0111525058746338,
"learning_rate": 8.252002913328478e-05,
"loss": 1.668,
"step": 314
},
{
"epoch": 0.21787999308317482,
"grad_norm": 0.7205057144165039,
"learning_rate": 8.244719592134014e-05,
"loss": 2.1546,
"step": 315
},
{
"epoch": 0.21857167560089918,
"grad_norm": 0.9559133052825928,
"learning_rate": 8.237436270939549e-05,
"loss": 1.6799,
"step": 316
},
{
"epoch": 0.21926335811862355,
"grad_norm": 0.7418698668479919,
"learning_rate": 8.230152949745085e-05,
"loss": 1.2123,
"step": 317
},
{
"epoch": 0.2199550406363479,
"grad_norm": 0.7784197330474854,
"learning_rate": 8.22286962855062e-05,
"loss": 1.1066,
"step": 318
},
{
"epoch": 0.22064672315407227,
"grad_norm": 0.754654586315155,
"learning_rate": 8.215586307356154e-05,
"loss": 2.0185,
"step": 319
},
{
"epoch": 0.22133840567179663,
"grad_norm": 0.6510013341903687,
"learning_rate": 8.20830298616169e-05,
"loss": 1.8316,
"step": 320
},
{
"epoch": 0.222030088189521,
"grad_norm": 1.541455626487732,
"learning_rate": 8.201019664967225e-05,
"loss": 1.3254,
"step": 321
},
{
"epoch": 0.2227217707072454,
"grad_norm": 0.8958970308303833,
"learning_rate": 8.193736343772761e-05,
"loss": 1.9616,
"step": 322
},
{
"epoch": 0.22341345322496975,
"grad_norm": 0.6273507475852966,
"learning_rate": 8.186453022578296e-05,
"loss": 1.13,
"step": 323
},
{
"epoch": 0.2241051357426941,
"grad_norm": 0.7866650223731995,
"learning_rate": 8.17916970138383e-05,
"loss": 1.784,
"step": 324
},
{
"epoch": 0.22479681826041847,
"grad_norm": 0.9650758504867554,
"learning_rate": 8.171886380189367e-05,
"loss": 0.9438,
"step": 325
},
{
"epoch": 0.22548850077814284,
"grad_norm": 0.7491368055343628,
"learning_rate": 8.164603058994901e-05,
"loss": 1.1472,
"step": 326
},
{
"epoch": 0.2261801832958672,
"grad_norm": 0.7516188025474548,
"learning_rate": 8.157319737800438e-05,
"loss": 1.841,
"step": 327
},
{
"epoch": 0.22687186581359156,
"grad_norm": 0.8555276989936829,
"learning_rate": 8.150036416605972e-05,
"loss": 2.0819,
"step": 328
},
{
"epoch": 0.22756354833131592,
"grad_norm": 0.8025880455970764,
"learning_rate": 8.142753095411508e-05,
"loss": 1.871,
"step": 329
},
{
"epoch": 0.22825523084904029,
"grad_norm": 0.8006131052970886,
"learning_rate": 8.135469774217043e-05,
"loss": 1.8293,
"step": 330
},
{
"epoch": 0.22894691336676465,
"grad_norm": 0.6898099184036255,
"learning_rate": 8.128186453022578e-05,
"loss": 1.9701,
"step": 331
},
{
"epoch": 0.229638595884489,
"grad_norm": 0.7827674150466919,
"learning_rate": 8.120903131828114e-05,
"loss": 1.9634,
"step": 332
},
{
"epoch": 0.23033027840221337,
"grad_norm": 0.7763857245445251,
"learning_rate": 8.113619810633649e-05,
"loss": 1.4908,
"step": 333
},
{
"epoch": 0.23102196091993774,
"grad_norm": 0.7040197253227234,
"learning_rate": 8.106336489439185e-05,
"loss": 1.9854,
"step": 334
},
{
"epoch": 0.23171364343766213,
"grad_norm": 1.0337374210357666,
"learning_rate": 8.09905316824472e-05,
"loss": 2.1542,
"step": 335
},
{
"epoch": 0.2324053259553865,
"grad_norm": 0.7228108048439026,
"learning_rate": 8.091769847050256e-05,
"loss": 1.9267,
"step": 336
},
{
"epoch": 0.23309700847311085,
"grad_norm": 0.6841875910758972,
"learning_rate": 8.08448652585579e-05,
"loss": 1.769,
"step": 337
},
{
"epoch": 0.2337886909908352,
"grad_norm": 0.6749361753463745,
"learning_rate": 8.077203204661325e-05,
"loss": 1.7758,
"step": 338
},
{
"epoch": 0.23448037350855958,
"grad_norm": 0.6411470174789429,
"learning_rate": 8.069919883466861e-05,
"loss": 1.2938,
"step": 339
},
{
"epoch": 0.23517205602628394,
"grad_norm": 0.6457031965255737,
"learning_rate": 8.062636562272396e-05,
"loss": 1.3416,
"step": 340
},
{
"epoch": 0.2358637385440083,
"grad_norm": 0.7273378372192383,
"learning_rate": 8.055353241077932e-05,
"loss": 1.5468,
"step": 341
},
{
"epoch": 0.23655542106173266,
"grad_norm": 0.7910048365592957,
"learning_rate": 8.048069919883467e-05,
"loss": 1.8577,
"step": 342
},
{
"epoch": 0.23724710357945702,
"grad_norm": 0.7172124981880188,
"learning_rate": 8.040786598689003e-05,
"loss": 2.2877,
"step": 343
},
{
"epoch": 0.2379387860971814,
"grad_norm": 0.6511913537979126,
"learning_rate": 8.033503277494538e-05,
"loss": 1.0914,
"step": 344
},
{
"epoch": 0.23863046861490575,
"grad_norm": 0.6148791909217834,
"learning_rate": 8.026219956300073e-05,
"loss": 1.3388,
"step": 345
},
{
"epoch": 0.2393221511326301,
"grad_norm": 0.7384858131408691,
"learning_rate": 8.018936635105609e-05,
"loss": 1.8186,
"step": 346
},
{
"epoch": 0.24001383365035447,
"grad_norm": 0.793538510799408,
"learning_rate": 8.011653313911143e-05,
"loss": 1.5962,
"step": 347
},
{
"epoch": 0.24070551616807886,
"grad_norm": 0.6316525936126709,
"learning_rate": 8.00436999271668e-05,
"loss": 1.4712,
"step": 348
},
{
"epoch": 0.24139719868580323,
"grad_norm": 0.7671139240264893,
"learning_rate": 7.997086671522214e-05,
"loss": 2.2086,
"step": 349
},
{
"epoch": 0.2420888812035276,
"grad_norm": 0.7423043251037598,
"learning_rate": 7.98980335032775e-05,
"loss": 2.1409,
"step": 350
},
{
"epoch": 0.24278056372125195,
"grad_norm": 0.834575891494751,
"learning_rate": 7.982520029133285e-05,
"loss": 1.5166,
"step": 351
},
{
"epoch": 0.24347224623897631,
"grad_norm": 1.0157177448272705,
"learning_rate": 7.97523670793882e-05,
"loss": 1.7054,
"step": 352
},
{
"epoch": 0.24416392875670068,
"grad_norm": 0.8765892386436462,
"learning_rate": 7.967953386744356e-05,
"loss": 2.4397,
"step": 353
},
{
"epoch": 0.24485561127442504,
"grad_norm": 0.6456695795059204,
"learning_rate": 7.960670065549891e-05,
"loss": 1.2428,
"step": 354
},
{
"epoch": 0.2455472937921494,
"grad_norm": 0.813552737236023,
"learning_rate": 7.953386744355427e-05,
"loss": 1.7926,
"step": 355
},
{
"epoch": 0.24623897630987376,
"grad_norm": 0.6702203750610352,
"learning_rate": 7.946103423160962e-05,
"loss": 1.6654,
"step": 356
},
{
"epoch": 0.24693065882759813,
"grad_norm": 0.9113941788673401,
"learning_rate": 7.938820101966498e-05,
"loss": 2.1639,
"step": 357
},
{
"epoch": 0.2476223413453225,
"grad_norm": 0.6006965041160583,
"learning_rate": 7.931536780772032e-05,
"loss": 1.815,
"step": 358
},
{
"epoch": 0.24831402386304685,
"grad_norm": 0.9116283059120178,
"learning_rate": 7.924253459577567e-05,
"loss": 1.5244,
"step": 359
},
{
"epoch": 0.2490057063807712,
"grad_norm": 0.8563610315322876,
"learning_rate": 7.916970138383103e-05,
"loss": 2.1287,
"step": 360
},
{
"epoch": 0.2496973888984956,
"grad_norm": 0.7254488468170166,
"learning_rate": 7.909686817188638e-05,
"loss": 1.1997,
"step": 361
},
{
"epoch": 0.25038907141621997,
"grad_norm": 0.8156582117080688,
"learning_rate": 7.902403495994174e-05,
"loss": 2.1275,
"step": 362
},
{
"epoch": 0.2510807539339443,
"grad_norm": 0.7504958510398865,
"learning_rate": 7.895120174799709e-05,
"loss": 2.1381,
"step": 363
},
{
"epoch": 0.2517724364516687,
"grad_norm": 0.6650155782699585,
"learning_rate": 7.887836853605244e-05,
"loss": 0.9823,
"step": 364
},
{
"epoch": 0.252464118969393,
"grad_norm": 0.6706410646438599,
"learning_rate": 7.88055353241078e-05,
"loss": 1.1854,
"step": 365
},
{
"epoch": 0.2531558014871174,
"grad_norm": 1.0841103792190552,
"learning_rate": 7.873270211216315e-05,
"loss": 1.4933,
"step": 366
},
{
"epoch": 0.25384748400484175,
"grad_norm": 0.7275784611701965,
"learning_rate": 7.86598689002185e-05,
"loss": 2.162,
"step": 367
},
{
"epoch": 0.25453916652256614,
"grad_norm": 0.994987964630127,
"learning_rate": 7.858703568827385e-05,
"loss": 1.6992,
"step": 368
},
{
"epoch": 0.25523084904029053,
"grad_norm": 1.160749912261963,
"learning_rate": 7.851420247632922e-05,
"loss": 1.2521,
"step": 369
},
{
"epoch": 0.25592253155801487,
"grad_norm": 0.8106790781021118,
"learning_rate": 7.844136926438456e-05,
"loss": 1.5949,
"step": 370
},
{
"epoch": 0.25661421407573926,
"grad_norm": 0.5805374383926392,
"learning_rate": 7.836853605243991e-05,
"loss": 1.713,
"step": 371
},
{
"epoch": 0.2573058965934636,
"grad_norm": 0.6805906295776367,
"learning_rate": 7.829570284049527e-05,
"loss": 1.7713,
"step": 372
},
{
"epoch": 0.257997579111188,
"grad_norm": 0.6964860558509827,
"learning_rate": 7.822286962855062e-05,
"loss": 1.8013,
"step": 373
},
{
"epoch": 0.2586892616289123,
"grad_norm": 1.2079120874404907,
"learning_rate": 7.815003641660598e-05,
"loss": 2.0171,
"step": 374
},
{
"epoch": 0.2593809441466367,
"grad_norm": 0.689630925655365,
"learning_rate": 7.807720320466133e-05,
"loss": 1.8816,
"step": 375
},
{
"epoch": 0.26007262666436104,
"grad_norm": 0.9079475402832031,
"learning_rate": 7.800436999271669e-05,
"loss": 1.5992,
"step": 376
},
{
"epoch": 0.26076430918208543,
"grad_norm": 0.4846475422382355,
"learning_rate": 7.793153678077204e-05,
"loss": 0.9038,
"step": 377
},
{
"epoch": 0.26145599169980976,
"grad_norm": 0.6311324238777161,
"learning_rate": 7.785870356882738e-05,
"loss": 1.757,
"step": 378
},
{
"epoch": 0.26214767421753415,
"grad_norm": 0.6497640609741211,
"learning_rate": 7.778587035688274e-05,
"loss": 1.799,
"step": 379
},
{
"epoch": 0.2628393567352585,
"grad_norm": 0.6967574954032898,
"learning_rate": 7.771303714493809e-05,
"loss": 2.0541,
"step": 380
},
{
"epoch": 0.2635310392529829,
"grad_norm": 0.8538269400596619,
"learning_rate": 7.764020393299345e-05,
"loss": 1.4781,
"step": 381
},
{
"epoch": 0.26422272177070727,
"grad_norm": 0.747353196144104,
"learning_rate": 7.75673707210488e-05,
"loss": 1.063,
"step": 382
},
{
"epoch": 0.2649144042884316,
"grad_norm": 0.7110087871551514,
"learning_rate": 7.749453750910416e-05,
"loss": 1.6301,
"step": 383
},
{
"epoch": 0.265606086806156,
"grad_norm": 0.8608129620552063,
"learning_rate": 7.742170429715951e-05,
"loss": 1.7459,
"step": 384
},
{
"epoch": 0.26629776932388033,
"grad_norm": 0.903325617313385,
"learning_rate": 7.734887108521486e-05,
"loss": 1.0401,
"step": 385
},
{
"epoch": 0.2669894518416047,
"grad_norm": 0.6887355446815491,
"learning_rate": 7.727603787327022e-05,
"loss": 1.6558,
"step": 386
},
{
"epoch": 0.26768113435932905,
"grad_norm": 0.8748055100440979,
"learning_rate": 7.720320466132557e-05,
"loss": 2.2414,
"step": 387
},
{
"epoch": 0.26837281687705344,
"grad_norm": 0.7315294146537781,
"learning_rate": 7.713037144938093e-05,
"loss": 1.7651,
"step": 388
},
{
"epoch": 0.2690644993947778,
"grad_norm": 0.7685129642486572,
"learning_rate": 7.705753823743627e-05,
"loss": 1.64,
"step": 389
},
{
"epoch": 0.26975618191250217,
"grad_norm": 0.7387799620628357,
"learning_rate": 7.698470502549163e-05,
"loss": 1.655,
"step": 390
},
{
"epoch": 0.2704478644302265,
"grad_norm": 0.6962683200836182,
"learning_rate": 7.691187181354698e-05,
"loss": 2.0759,
"step": 391
},
{
"epoch": 0.2711395469479509,
"grad_norm": 0.6928436160087585,
"learning_rate": 7.683903860160233e-05,
"loss": 1.902,
"step": 392
},
{
"epoch": 0.27183122946567523,
"grad_norm": 0.6286187767982483,
"learning_rate": 7.676620538965769e-05,
"loss": 1.1205,
"step": 393
},
{
"epoch": 0.2725229119833996,
"grad_norm": 0.8087384104728699,
"learning_rate": 7.669337217771304e-05,
"loss": 1.5928,
"step": 394
},
{
"epoch": 0.273214594501124,
"grad_norm": 0.6950211524963379,
"learning_rate": 7.66205389657684e-05,
"loss": 1.9022,
"step": 395
},
{
"epoch": 0.27390627701884834,
"grad_norm": 0.716234028339386,
"learning_rate": 7.654770575382375e-05,
"loss": 1.9103,
"step": 396
},
{
"epoch": 0.27459795953657273,
"grad_norm": 1.0911515951156616,
"learning_rate": 7.647487254187911e-05,
"loss": 2.1727,
"step": 397
},
{
"epoch": 0.27528964205429707,
"grad_norm": 1.0012110471725464,
"learning_rate": 7.640203932993446e-05,
"loss": 2.0671,
"step": 398
},
{
"epoch": 0.27598132457202146,
"grad_norm": 0.7271398305892944,
"learning_rate": 7.63292061179898e-05,
"loss": 1.8487,
"step": 399
},
{
"epoch": 0.2766730070897458,
"grad_norm": 0.9112443923950195,
"learning_rate": 7.625637290604516e-05,
"loss": 1.6583,
"step": 400
},
{
"epoch": 0.2773646896074702,
"grad_norm": 0.8109308481216431,
"learning_rate": 7.618353969410051e-05,
"loss": 1.6202,
"step": 401
},
{
"epoch": 0.2780563721251945,
"grad_norm": 0.7934839129447937,
"learning_rate": 7.611070648215587e-05,
"loss": 2.2605,
"step": 402
},
{
"epoch": 0.2787480546429189,
"grad_norm": 0.7166281342506409,
"learning_rate": 7.603787327021122e-05,
"loss": 1.5321,
"step": 403
},
{
"epoch": 0.27943973716064324,
"grad_norm": 0.6386390924453735,
"learning_rate": 7.596504005826657e-05,
"loss": 0.9431,
"step": 404
},
{
"epoch": 0.28013141967836763,
"grad_norm": 1.2669974565505981,
"learning_rate": 7.589220684632193e-05,
"loss": 0.7634,
"step": 405
},
{
"epoch": 0.28082310219609197,
"grad_norm": 0.7032380700111389,
"learning_rate": 7.581937363437728e-05,
"loss": 1.9926,
"step": 406
},
{
"epoch": 0.28151478471381636,
"grad_norm": 0.648324191570282,
"learning_rate": 7.574654042243264e-05,
"loss": 1.2027,
"step": 407
},
{
"epoch": 0.28220646723154075,
"grad_norm": 0.7746976613998413,
"learning_rate": 7.567370721048798e-05,
"loss": 1.3615,
"step": 408
},
{
"epoch": 0.2828981497492651,
"grad_norm": 0.8174494504928589,
"learning_rate": 7.560087399854335e-05,
"loss": 1.9945,
"step": 409
},
{
"epoch": 0.28358983226698947,
"grad_norm": 0.9701315760612488,
"learning_rate": 7.55280407865987e-05,
"loss": 1.4894,
"step": 410
},
{
"epoch": 0.2842815147847138,
"grad_norm": 0.9906868934631348,
"learning_rate": 7.545520757465404e-05,
"loss": 1.7937,
"step": 411
},
{
"epoch": 0.2849731973024382,
"grad_norm": 0.6380108594894409,
"learning_rate": 7.53823743627094e-05,
"loss": 1.6794,
"step": 412
},
{
"epoch": 0.28566487982016253,
"grad_norm": 0.6367055177688599,
"learning_rate": 7.530954115076475e-05,
"loss": 1.4019,
"step": 413
},
{
"epoch": 0.2863565623378869,
"grad_norm": 0.9779865741729736,
"learning_rate": 7.523670793882011e-05,
"loss": 1.8048,
"step": 414
},
{
"epoch": 0.28704824485561126,
"grad_norm": 0.6246545910835266,
"learning_rate": 7.516387472687546e-05,
"loss": 1.7138,
"step": 415
},
{
"epoch": 0.28773992737333565,
"grad_norm": 0.7418220639228821,
"learning_rate": 7.509104151493082e-05,
"loss": 1.9917,
"step": 416
},
{
"epoch": 0.28843160989106,
"grad_norm": 0.7883859276771545,
"learning_rate": 7.501820830298617e-05,
"loss": 1.2134,
"step": 417
},
{
"epoch": 0.28912329240878437,
"grad_norm": 0.8116041421890259,
"learning_rate": 7.494537509104151e-05,
"loss": 1.1634,
"step": 418
},
{
"epoch": 0.2898149749265087,
"grad_norm": 0.7708969712257385,
"learning_rate": 7.487254187909688e-05,
"loss": 1.8895,
"step": 419
},
{
"epoch": 0.2905066574442331,
"grad_norm": 1.3700909614562988,
"learning_rate": 7.479970866715222e-05,
"loss": 1.1294,
"step": 420
},
{
"epoch": 0.2911983399619575,
"grad_norm": 0.8107597827911377,
"learning_rate": 7.472687545520758e-05,
"loss": 1.0097,
"step": 421
},
{
"epoch": 0.2918900224796818,
"grad_norm": 0.8277159929275513,
"learning_rate": 7.465404224326293e-05,
"loss": 1.0633,
"step": 422
},
{
"epoch": 0.2925817049974062,
"grad_norm": 0.75505131483078,
"learning_rate": 7.458120903131829e-05,
"loss": 1.4971,
"step": 423
},
{
"epoch": 0.29327338751513055,
"grad_norm": 0.7761143445968628,
"learning_rate": 7.450837581937363e-05,
"loss": 1.569,
"step": 424
},
{
"epoch": 0.29396507003285494,
"grad_norm": 0.8429370522499084,
"learning_rate": 7.443554260742899e-05,
"loss": 2.147,
"step": 425
},
{
"epoch": 0.29465675255057927,
"grad_norm": 0.7506592273712158,
"learning_rate": 7.436270939548435e-05,
"loss": 2.1514,
"step": 426
},
{
"epoch": 0.29534843506830366,
"grad_norm": 0.6654348969459534,
"learning_rate": 7.42898761835397e-05,
"loss": 1.7697,
"step": 427
},
{
"epoch": 0.296040117586028,
"grad_norm": 1.179040551185608,
"learning_rate": 7.421704297159506e-05,
"loss": 1.6906,
"step": 428
},
{
"epoch": 0.2967318001037524,
"grad_norm": 0.6145846247673035,
"learning_rate": 7.41442097596504e-05,
"loss": 1.8308,
"step": 429
},
{
"epoch": 0.2974234826214767,
"grad_norm": 0.7995002269744873,
"learning_rate": 7.407137654770577e-05,
"loss": 1.7942,
"step": 430
},
{
"epoch": 0.2981151651392011,
"grad_norm": 0.707095742225647,
"learning_rate": 7.39985433357611e-05,
"loss": 1.6325,
"step": 431
},
{
"epoch": 0.29880684765692545,
"grad_norm": 0.5955487489700317,
"learning_rate": 7.392571012381646e-05,
"loss": 1.3466,
"step": 432
},
{
"epoch": 0.29949853017464984,
"grad_norm": 0.9258183240890503,
"learning_rate": 7.385287691187182e-05,
"loss": 1.8544,
"step": 433
},
{
"epoch": 0.3001902126923742,
"grad_norm": 0.8435434103012085,
"learning_rate": 7.378004369992717e-05,
"loss": 1.8884,
"step": 434
},
{
"epoch": 0.30088189521009856,
"grad_norm": 0.7008845210075378,
"learning_rate": 7.370721048798253e-05,
"loss": 1.8409,
"step": 435
},
{
"epoch": 0.30157357772782295,
"grad_norm": 0.7846685647964478,
"learning_rate": 7.363437727603788e-05,
"loss": 1.8821,
"step": 436
},
{
"epoch": 0.3022652602455473,
"grad_norm": 0.7767228484153748,
"learning_rate": 7.356154406409324e-05,
"loss": 2.0034,
"step": 437
},
{
"epoch": 0.3029569427632717,
"grad_norm": 0.787028431892395,
"learning_rate": 7.348871085214857e-05,
"loss": 1.8664,
"step": 438
},
{
"epoch": 0.303648625280996,
"grad_norm": 0.6583303809165955,
"learning_rate": 7.341587764020393e-05,
"loss": 1.1991,
"step": 439
},
{
"epoch": 0.3043403077987204,
"grad_norm": 0.8681322932243347,
"learning_rate": 7.33430444282593e-05,
"loss": 1.7124,
"step": 440
},
{
"epoch": 0.30503199031644473,
"grad_norm": 0.9781901240348816,
"learning_rate": 7.327021121631464e-05,
"loss": 2.0128,
"step": 441
},
{
"epoch": 0.3057236728341691,
"grad_norm": 0.6712149381637573,
"learning_rate": 7.319737800437e-05,
"loss": 1.4835,
"step": 442
},
{
"epoch": 0.30641535535189346,
"grad_norm": 0.7322747111320496,
"learning_rate": 7.312454479242535e-05,
"loss": 1.1381,
"step": 443
},
{
"epoch": 0.30710703786961785,
"grad_norm": 0.8576450347900391,
"learning_rate": 7.30517115804807e-05,
"loss": 1.8142,
"step": 444
},
{
"epoch": 0.3077987203873422,
"grad_norm": 0.6087522506713867,
"learning_rate": 7.297887836853605e-05,
"loss": 1.2711,
"step": 445
},
{
"epoch": 0.3084904029050666,
"grad_norm": 0.8685612082481384,
"learning_rate": 7.290604515659141e-05,
"loss": 1.2365,
"step": 446
},
{
"epoch": 0.30918208542279096,
"grad_norm": 0.7040896415710449,
"learning_rate": 7.283321194464677e-05,
"loss": 1.6385,
"step": 447
},
{
"epoch": 0.3098737679405153,
"grad_norm": 0.943147599697113,
"learning_rate": 7.276037873270212e-05,
"loss": 1.4699,
"step": 448
},
{
"epoch": 0.3105654504582397,
"grad_norm": 0.6641246676445007,
"learning_rate": 7.268754552075748e-05,
"loss": 1.6489,
"step": 449
},
{
"epoch": 0.311257132975964,
"grad_norm": 0.7420269250869751,
"learning_rate": 7.261471230881282e-05,
"loss": 2.1525,
"step": 450
},
{
"epoch": 0.3119488154936884,
"grad_norm": 0.7198522090911865,
"learning_rate": 7.254187909686817e-05,
"loss": 2.01,
"step": 451
},
{
"epoch": 0.31264049801141275,
"grad_norm": 0.8209108114242554,
"learning_rate": 7.246904588492352e-05,
"loss": 1.5205,
"step": 452
},
{
"epoch": 0.31333218052913714,
"grad_norm": 1.0207217931747437,
"learning_rate": 7.239621267297888e-05,
"loss": 1.589,
"step": 453
},
{
"epoch": 0.3140238630468615,
"grad_norm": 0.7939510345458984,
"learning_rate": 7.232337946103424e-05,
"loss": 1.9908,
"step": 454
},
{
"epoch": 0.31471554556458586,
"grad_norm": 0.6668927073478699,
"learning_rate": 7.225054624908959e-05,
"loss": 1.7476,
"step": 455
},
{
"epoch": 0.3154072280823102,
"grad_norm": 0.8068589568138123,
"learning_rate": 7.217771303714495e-05,
"loss": 1.7593,
"step": 456
},
{
"epoch": 0.3160989106000346,
"grad_norm": 0.6828214526176453,
"learning_rate": 7.21048798252003e-05,
"loss": 1.5636,
"step": 457
},
{
"epoch": 0.3167905931177589,
"grad_norm": 0.7853332757949829,
"learning_rate": 7.203204661325565e-05,
"loss": 1.1448,
"step": 458
},
{
"epoch": 0.3174822756354833,
"grad_norm": 1.0938345193862915,
"learning_rate": 7.195921340131099e-05,
"loss": 1.7145,
"step": 459
},
{
"epoch": 0.3181739581532077,
"grad_norm": 0.7735335826873779,
"learning_rate": 7.188638018936635e-05,
"loss": 1.2118,
"step": 460
},
{
"epoch": 0.31886564067093204,
"grad_norm": 0.6759296655654907,
"learning_rate": 7.181354697742171e-05,
"loss": 1.1214,
"step": 461
},
{
"epoch": 0.31955732318865643,
"grad_norm": 0.8163008689880371,
"learning_rate": 7.174071376547706e-05,
"loss": 1.9071,
"step": 462
},
{
"epoch": 0.32024900570638076,
"grad_norm": 0.8205277919769287,
"learning_rate": 7.166788055353242e-05,
"loss": 1.0775,
"step": 463
},
{
"epoch": 0.32094068822410515,
"grad_norm": 0.712774395942688,
"learning_rate": 7.159504734158776e-05,
"loss": 1.8912,
"step": 464
},
{
"epoch": 0.3216323707418295,
"grad_norm": 0.7423759698867798,
"learning_rate": 7.152221412964312e-05,
"loss": 2.0655,
"step": 465
},
{
"epoch": 0.3223240532595539,
"grad_norm": 0.6948613524436951,
"learning_rate": 7.144938091769847e-05,
"loss": 2.1957,
"step": 466
},
{
"epoch": 0.3230157357772782,
"grad_norm": 0.6164250373840332,
"learning_rate": 7.137654770575383e-05,
"loss": 0.9997,
"step": 467
},
{
"epoch": 0.3237074182950026,
"grad_norm": 0.7659111022949219,
"learning_rate": 7.130371449380919e-05,
"loss": 1.4006,
"step": 468
},
{
"epoch": 0.32439910081272694,
"grad_norm": 0.830442488193512,
"learning_rate": 7.123088128186454e-05,
"loss": 1.6562,
"step": 469
},
{
"epoch": 0.3250907833304513,
"grad_norm": 0.649169921875,
"learning_rate": 7.11580480699199e-05,
"loss": 1.8275,
"step": 470
},
{
"epoch": 0.32578246584817566,
"grad_norm": 0.7854671478271484,
"learning_rate": 7.108521485797523e-05,
"loss": 0.3147,
"step": 471
},
{
"epoch": 0.32647414836590005,
"grad_norm": 1.0381858348846436,
"learning_rate": 7.101238164603059e-05,
"loss": 2.0842,
"step": 472
},
{
"epoch": 0.32716583088362444,
"grad_norm": 0.7317079305648804,
"learning_rate": 7.093954843408594e-05,
"loss": 2.1326,
"step": 473
},
{
"epoch": 0.3278575134013488,
"grad_norm": 0.73940110206604,
"learning_rate": 7.08667152221413e-05,
"loss": 1.9647,
"step": 474
},
{
"epoch": 0.32854919591907317,
"grad_norm": 0.8855507969856262,
"learning_rate": 7.079388201019666e-05,
"loss": 1.9308,
"step": 475
},
{
"epoch": 0.3292408784367975,
"grad_norm": 0.7795459628105164,
"learning_rate": 7.072104879825201e-05,
"loss": 1.9297,
"step": 476
},
{
"epoch": 0.3299325609545219,
"grad_norm": 0.765084981918335,
"learning_rate": 7.064821558630737e-05,
"loss": 2.3063,
"step": 477
},
{
"epoch": 0.3306242434722462,
"grad_norm": 0.7164610624313354,
"learning_rate": 7.05753823743627e-05,
"loss": 1.8509,
"step": 478
},
{
"epoch": 0.3313159259899706,
"grad_norm": 0.6976935863494873,
"learning_rate": 7.050254916241806e-05,
"loss": 1.8351,
"step": 479
},
{
"epoch": 0.33200760850769495,
"grad_norm": 1.1228684186935425,
"learning_rate": 7.042971595047341e-05,
"loss": 1.4274,
"step": 480
},
{
"epoch": 0.33269929102541934,
"grad_norm": 0.7229692935943604,
"learning_rate": 7.035688273852877e-05,
"loss": 1.5104,
"step": 481
},
{
"epoch": 0.3333909735431437,
"grad_norm": 0.7485585808753967,
"learning_rate": 7.028404952658413e-05,
"loss": 1.1417,
"step": 482
},
{
"epoch": 0.33408265606086807,
"grad_norm": 1.7345014810562134,
"learning_rate": 7.021121631463948e-05,
"loss": 1.6614,
"step": 483
},
{
"epoch": 0.3347743385785924,
"grad_norm": 0.7559076547622681,
"learning_rate": 7.013838310269483e-05,
"loss": 1.8508,
"step": 484
},
{
"epoch": 0.3354660210963168,
"grad_norm": 1.2972337007522583,
"learning_rate": 7.006554989075018e-05,
"loss": 1.9322,
"step": 485
},
{
"epoch": 0.3361577036140412,
"grad_norm": 0.9719124436378479,
"learning_rate": 6.999271667880554e-05,
"loss": 1.6138,
"step": 486
},
{
"epoch": 0.3368493861317655,
"grad_norm": 0.7772285342216492,
"learning_rate": 6.991988346686089e-05,
"loss": 2.1904,
"step": 487
},
{
"epoch": 0.3375410686494899,
"grad_norm": 0.8187295198440552,
"learning_rate": 6.984705025491625e-05,
"loss": 1.4893,
"step": 488
},
{
"epoch": 0.33823275116721424,
"grad_norm": 0.8635819554328918,
"learning_rate": 6.977421704297161e-05,
"loss": 2.0885,
"step": 489
},
{
"epoch": 0.33892443368493863,
"grad_norm": 1.2180774211883545,
"learning_rate": 6.970138383102696e-05,
"loss": 1.1084,
"step": 490
},
{
"epoch": 0.33961611620266297,
"grad_norm": 0.8329484462738037,
"learning_rate": 6.96285506190823e-05,
"loss": 1.3746,
"step": 491
},
{
"epoch": 0.34030779872038736,
"grad_norm": 0.7301535606384277,
"learning_rate": 6.955571740713765e-05,
"loss": 1.2109,
"step": 492
},
{
"epoch": 0.3409994812381117,
"grad_norm": 0.740479052066803,
"learning_rate": 6.948288419519301e-05,
"loss": 1.9807,
"step": 493
},
{
"epoch": 0.3416911637558361,
"grad_norm": 0.9610648155212402,
"learning_rate": 6.941005098324836e-05,
"loss": 1.5924,
"step": 494
},
{
"epoch": 0.3423828462735604,
"grad_norm": 0.7890329360961914,
"learning_rate": 6.933721777130372e-05,
"loss": 1.7208,
"step": 495
},
{
"epoch": 0.3430745287912848,
"grad_norm": 0.7466354370117188,
"learning_rate": 6.926438455935908e-05,
"loss": 1.8483,
"step": 496
},
{
"epoch": 0.34376621130900914,
"grad_norm": 0.8114244341850281,
"learning_rate": 6.919155134741443e-05,
"loss": 1.9642,
"step": 497
},
{
"epoch": 0.34445789382673353,
"grad_norm": 0.8441615104675293,
"learning_rate": 6.911871813546978e-05,
"loss": 1.6589,
"step": 498
},
{
"epoch": 0.3451495763444579,
"grad_norm": 0.771629273891449,
"learning_rate": 6.904588492352512e-05,
"loss": 2.1515,
"step": 499
},
{
"epoch": 0.34584125886218225,
"grad_norm": 0.7773415446281433,
"learning_rate": 6.897305171158048e-05,
"loss": 1.1782,
"step": 500
},
{
"epoch": 0.34584125886218225,
"eval_loss": 1.5188031196594238,
"eval_runtime": 636.0478,
"eval_samples_per_second": 2.02,
"eval_steps_per_second": 1.011,
"step": 500
},
{
"epoch": 0.34653294137990664,
"grad_norm": 0.8539692759513855,
"learning_rate": 6.890021849963583e-05,
"loss": 2.2602,
"step": 501
},
{
"epoch": 0.347224623897631,
"grad_norm": 0.7637029886245728,
"learning_rate": 6.882738528769119e-05,
"loss": 1.8441,
"step": 502
},
{
"epoch": 0.34791630641535537,
"grad_norm": 0.7741591334342957,
"learning_rate": 6.875455207574655e-05,
"loss": 1.876,
"step": 503
},
{
"epoch": 0.3486079889330797,
"grad_norm": 0.7942149639129639,
"learning_rate": 6.868171886380189e-05,
"loss": 1.2218,
"step": 504
},
{
"epoch": 0.3492996714508041,
"grad_norm": 0.7626279592514038,
"learning_rate": 6.860888565185725e-05,
"loss": 1.563,
"step": 505
},
{
"epoch": 0.34999135396852843,
"grad_norm": 0.6218715310096741,
"learning_rate": 6.85360524399126e-05,
"loss": 1.7303,
"step": 506
},
{
"epoch": 0.3506830364862528,
"grad_norm": 0.8599730134010315,
"learning_rate": 6.846321922796796e-05,
"loss": 1.3347,
"step": 507
},
{
"epoch": 0.35137471900397715,
"grad_norm": 0.5834329128265381,
"learning_rate": 6.83903860160233e-05,
"loss": 1.1654,
"step": 508
},
{
"epoch": 0.35206640152170154,
"grad_norm": 0.8399999737739563,
"learning_rate": 6.831755280407867e-05,
"loss": 1.4099,
"step": 509
},
{
"epoch": 0.3527580840394259,
"grad_norm": 0.7133538126945496,
"learning_rate": 6.824471959213403e-05,
"loss": 0.9115,
"step": 510
},
{
"epoch": 0.35344976655715027,
"grad_norm": 0.7058927416801453,
"learning_rate": 6.817188638018936e-05,
"loss": 1.4206,
"step": 511
},
{
"epoch": 0.35414144907487466,
"grad_norm": 1.1560832262039185,
"learning_rate": 6.809905316824472e-05,
"loss": 1.5635,
"step": 512
},
{
"epoch": 0.354833131592599,
"grad_norm": 0.7324162721633911,
"learning_rate": 6.802621995630007e-05,
"loss": 2.045,
"step": 513
},
{
"epoch": 0.3555248141103234,
"grad_norm": 0.6711490750312805,
"learning_rate": 6.795338674435543e-05,
"loss": 2.0891,
"step": 514
},
{
"epoch": 0.3562164966280477,
"grad_norm": 0.795870304107666,
"learning_rate": 6.788055353241078e-05,
"loss": 2.1285,
"step": 515
},
{
"epoch": 0.3569081791457721,
"grad_norm": 1.913321852684021,
"learning_rate": 6.780772032046614e-05,
"loss": 2.1861,
"step": 516
},
{
"epoch": 0.35759986166349644,
"grad_norm": 0.7828819751739502,
"learning_rate": 6.77348871085215e-05,
"loss": 1.4283,
"step": 517
},
{
"epoch": 0.35829154418122083,
"grad_norm": 0.7590184211730957,
"learning_rate": 6.766205389657683e-05,
"loss": 1.6581,
"step": 518
},
{
"epoch": 0.35898322669894517,
"grad_norm": 0.8273248672485352,
"learning_rate": 6.75892206846322e-05,
"loss": 2.0783,
"step": 519
},
{
"epoch": 0.35967490921666956,
"grad_norm": 0.791447639465332,
"learning_rate": 6.751638747268754e-05,
"loss": 1.2507,
"step": 520
},
{
"epoch": 0.3603665917343939,
"grad_norm": 0.7864679098129272,
"learning_rate": 6.74435542607429e-05,
"loss": 1.4686,
"step": 521
},
{
"epoch": 0.3610582742521183,
"grad_norm": 0.7522057294845581,
"learning_rate": 6.737072104879825e-05,
"loss": 1.4506,
"step": 522
},
{
"epoch": 0.3617499567698426,
"grad_norm": 0.6190770864486694,
"learning_rate": 6.729788783685361e-05,
"loss": 2.0413,
"step": 523
},
{
"epoch": 0.362441639287567,
"grad_norm": 0.9738603830337524,
"learning_rate": 6.722505462490896e-05,
"loss": 1.315,
"step": 524
},
{
"epoch": 0.3631333218052914,
"grad_norm": 0.7005712389945984,
"learning_rate": 6.715222141296431e-05,
"loss": 1.528,
"step": 525
},
{
"epoch": 0.36382500432301573,
"grad_norm": 1.2126790285110474,
"learning_rate": 6.707938820101967e-05,
"loss": 2.0808,
"step": 526
},
{
"epoch": 0.3645166868407401,
"grad_norm": 0.9661760330200195,
"learning_rate": 6.700655498907502e-05,
"loss": 2.0187,
"step": 527
},
{
"epoch": 0.36520836935846446,
"grad_norm": 0.7546457648277283,
"learning_rate": 6.693372177713038e-05,
"loss": 1.4964,
"step": 528
},
{
"epoch": 0.36590005187618885,
"grad_norm": 0.8270094394683838,
"learning_rate": 6.686088856518573e-05,
"loss": 1.7801,
"step": 529
},
{
"epoch": 0.3665917343939132,
"grad_norm": 0.7802672386169434,
"learning_rate": 6.678805535324109e-05,
"loss": 1.4457,
"step": 530
},
{
"epoch": 0.3672834169116376,
"grad_norm": 0.7970958948135376,
"learning_rate": 6.671522214129643e-05,
"loss": 1.5796,
"step": 531
},
{
"epoch": 0.3679750994293619,
"grad_norm": 0.5904183983802795,
"learning_rate": 6.664238892935178e-05,
"loss": 1.013,
"step": 532
},
{
"epoch": 0.3686667819470863,
"grad_norm": 0.8535469174385071,
"learning_rate": 6.656955571740714e-05,
"loss": 1.3146,
"step": 533
},
{
"epoch": 0.36935846446481063,
"grad_norm": 0.722558319568634,
"learning_rate": 6.649672250546249e-05,
"loss": 1.2695,
"step": 534
},
{
"epoch": 0.370050146982535,
"grad_norm": 0.7470422983169556,
"learning_rate": 6.642388929351785e-05,
"loss": 2.1384,
"step": 535
},
{
"epoch": 0.37074182950025936,
"grad_norm": 1.035486102104187,
"learning_rate": 6.63510560815732e-05,
"loss": 1.6666,
"step": 536
},
{
"epoch": 0.37143351201798375,
"grad_norm": 0.8261759281158447,
"learning_rate": 6.627822286962856e-05,
"loss": 1.9752,
"step": 537
},
{
"epoch": 0.37212519453570814,
"grad_norm": 0.8254296183586121,
"learning_rate": 6.620538965768391e-05,
"loss": 1.9392,
"step": 538
},
{
"epoch": 0.37281687705343247,
"grad_norm": 0.9176802635192871,
"learning_rate": 6.613255644573925e-05,
"loss": 2.1566,
"step": 539
},
{
"epoch": 0.37350855957115686,
"grad_norm": 0.6909226179122925,
"learning_rate": 6.605972323379462e-05,
"loss": 1.4706,
"step": 540
},
{
"epoch": 0.3742002420888812,
"grad_norm": 1.0335818529129028,
"learning_rate": 6.598689002184996e-05,
"loss": 2.4501,
"step": 541
},
{
"epoch": 0.3748919246066056,
"grad_norm": 0.9764876365661621,
"learning_rate": 6.591405680990532e-05,
"loss": 2.2261,
"step": 542
},
{
"epoch": 0.3755836071243299,
"grad_norm": 0.8037734627723694,
"learning_rate": 6.584122359796067e-05,
"loss": 1.6487,
"step": 543
},
{
"epoch": 0.3762752896420543,
"grad_norm": 0.9658376574516296,
"learning_rate": 6.576839038601602e-05,
"loss": 2.0596,
"step": 544
},
{
"epoch": 0.37696697215977865,
"grad_norm": 0.7336639165878296,
"learning_rate": 6.569555717407138e-05,
"loss": 1.9689,
"step": 545
},
{
"epoch": 0.37765865467750304,
"grad_norm": 0.6894165873527527,
"learning_rate": 6.562272396212673e-05,
"loss": 1.8499,
"step": 546
},
{
"epoch": 0.37835033719522737,
"grad_norm": 0.7265276908874512,
"learning_rate": 6.554989075018209e-05,
"loss": 2.111,
"step": 547
},
{
"epoch": 0.37904201971295176,
"grad_norm": 0.8299263715744019,
"learning_rate": 6.547705753823744e-05,
"loss": 2.0565,
"step": 548
},
{
"epoch": 0.3797337022306761,
"grad_norm": 0.8597300052642822,
"learning_rate": 6.54042243262928e-05,
"loss": 1.7233,
"step": 549
},
{
"epoch": 0.3804253847484005,
"grad_norm": 0.7230535745620728,
"learning_rate": 6.533139111434814e-05,
"loss": 1.2764,
"step": 550
},
{
"epoch": 0.3811170672661249,
"grad_norm": 0.8922619819641113,
"learning_rate": 6.525855790240349e-05,
"loss": 2.1088,
"step": 551
},
{
"epoch": 0.3818087497838492,
"grad_norm": 0.7452985048294067,
"learning_rate": 6.518572469045885e-05,
"loss": 1.5725,
"step": 552
},
{
"epoch": 0.3825004323015736,
"grad_norm": 0.8344093561172485,
"learning_rate": 6.51128914785142e-05,
"loss": 1.8517,
"step": 553
},
{
"epoch": 0.38319211481929794,
"grad_norm": 0.6585731506347656,
"learning_rate": 6.504005826656956e-05,
"loss": 1.4584,
"step": 554
},
{
"epoch": 0.3838837973370223,
"grad_norm": 0.8440024256706238,
"learning_rate": 6.496722505462491e-05,
"loss": 1.547,
"step": 555
},
{
"epoch": 0.38457547985474666,
"grad_norm": 0.8570727705955505,
"learning_rate": 6.489439184268027e-05,
"loss": 1.7777,
"step": 556
},
{
"epoch": 0.38526716237247105,
"grad_norm": 0.8458102345466614,
"learning_rate": 6.482155863073562e-05,
"loss": 2.1761,
"step": 557
},
{
"epoch": 0.3859588448901954,
"grad_norm": 0.9602516293525696,
"learning_rate": 6.474872541879097e-05,
"loss": 1.224,
"step": 558
},
{
"epoch": 0.3866505274079198,
"grad_norm": 0.7018793821334839,
"learning_rate": 6.467589220684633e-05,
"loss": 1.8217,
"step": 559
},
{
"epoch": 0.3873422099256441,
"grad_norm": 0.796327531337738,
"learning_rate": 6.460305899490167e-05,
"loss": 1.9399,
"step": 560
},
{
"epoch": 0.3880338924433685,
"grad_norm": 0.6899397373199463,
"learning_rate": 6.453022578295704e-05,
"loss": 0.8488,
"step": 561
},
{
"epoch": 0.38872557496109283,
"grad_norm": 0.6222304701805115,
"learning_rate": 6.445739257101238e-05,
"loss": 0.7811,
"step": 562
},
{
"epoch": 0.3894172574788172,
"grad_norm": 0.7914639115333557,
"learning_rate": 6.438455935906774e-05,
"loss": 2.0951,
"step": 563
},
{
"epoch": 0.3901089399965416,
"grad_norm": 0.8506953120231628,
"learning_rate": 6.431172614712309e-05,
"loss": 1.3678,
"step": 564
},
{
"epoch": 0.39080062251426595,
"grad_norm": 0.6404629349708557,
"learning_rate": 6.423889293517844e-05,
"loss": 1.0092,
"step": 565
},
{
"epoch": 0.39149230503199034,
"grad_norm": 0.7604504227638245,
"learning_rate": 6.41660597232338e-05,
"loss": 1.931,
"step": 566
},
{
"epoch": 0.3921839875497147,
"grad_norm": 0.7102979421615601,
"learning_rate": 6.409322651128915e-05,
"loss": 1.3211,
"step": 567
},
{
"epoch": 0.39287567006743906,
"grad_norm": 0.7402350902557373,
"learning_rate": 6.402039329934451e-05,
"loss": 1.5225,
"step": 568
},
{
"epoch": 0.3935673525851634,
"grad_norm": 0.9871742129325867,
"learning_rate": 6.394756008739986e-05,
"loss": 0.7057,
"step": 569
},
{
"epoch": 0.3942590351028878,
"grad_norm": 0.7585322856903076,
"learning_rate": 6.387472687545522e-05,
"loss": 1.5484,
"step": 570
},
{
"epoch": 0.3949507176206121,
"grad_norm": 0.746465265750885,
"learning_rate": 6.380189366351056e-05,
"loss": 1.4404,
"step": 571
},
{
"epoch": 0.3956424001383365,
"grad_norm": 0.7660443186759949,
"learning_rate": 6.372906045156591e-05,
"loss": 1.8797,
"step": 572
},
{
"epoch": 0.39633408265606085,
"grad_norm": 0.7403152585029602,
"learning_rate": 6.365622723962127e-05,
"loss": 1.8291,
"step": 573
},
{
"epoch": 0.39702576517378524,
"grad_norm": 1.007710576057434,
"learning_rate": 6.358339402767662e-05,
"loss": 1.3267,
"step": 574
},
{
"epoch": 0.3977174476915096,
"grad_norm": 0.9685802459716797,
"learning_rate": 6.351056081573198e-05,
"loss": 2.0822,
"step": 575
},
{
"epoch": 0.39840913020923396,
"grad_norm": 0.8278066515922546,
"learning_rate": 6.343772760378733e-05,
"loss": 1.3005,
"step": 576
},
{
"epoch": 0.39910081272695835,
"grad_norm": 0.7012304067611694,
"learning_rate": 6.336489439184269e-05,
"loss": 0.9913,
"step": 577
},
{
"epoch": 0.3997924952446827,
"grad_norm": 1.449615716934204,
"learning_rate": 6.329206117989804e-05,
"loss": 1.3947,
"step": 578
},
{
"epoch": 0.4004841777624071,
"grad_norm": 1.1024595499038696,
"learning_rate": 6.321922796795339e-05,
"loss": 1.683,
"step": 579
},
{
"epoch": 0.4011758602801314,
"grad_norm": 1.3790334463119507,
"learning_rate": 6.314639475600875e-05,
"loss": 1.3824,
"step": 580
},
{
"epoch": 0.4018675427978558,
"grad_norm": 0.754559338092804,
"learning_rate": 6.30735615440641e-05,
"loss": 1.9189,
"step": 581
},
{
"epoch": 0.40255922531558014,
"grad_norm": 0.9193564653396606,
"learning_rate": 6.300072833211945e-05,
"loss": 1.7524,
"step": 582
},
{
"epoch": 0.40325090783330453,
"grad_norm": 0.6305272579193115,
"learning_rate": 6.29278951201748e-05,
"loss": 1.0415,
"step": 583
},
{
"epoch": 0.40394259035102886,
"grad_norm": 0.7553091645240784,
"learning_rate": 6.285506190823015e-05,
"loss": 1.4771,
"step": 584
},
{
"epoch": 0.40463427286875325,
"grad_norm": 0.9118616580963135,
"learning_rate": 6.278222869628551e-05,
"loss": 0.9804,
"step": 585
},
{
"epoch": 0.4053259553864776,
"grad_norm": 0.9057390093803406,
"learning_rate": 6.270939548434086e-05,
"loss": 1.8537,
"step": 586
},
{
"epoch": 0.406017637904202,
"grad_norm": 0.808927595615387,
"learning_rate": 6.263656227239622e-05,
"loss": 1.3161,
"step": 587
},
{
"epoch": 0.4067093204219263,
"grad_norm": 0.7657842040061951,
"learning_rate": 6.256372906045157e-05,
"loss": 1.3956,
"step": 588
},
{
"epoch": 0.4074010029396507,
"grad_norm": 0.7779245376586914,
"learning_rate": 6.249089584850693e-05,
"loss": 1.5218,
"step": 589
},
{
"epoch": 0.40809268545737504,
"grad_norm": 0.813758373260498,
"learning_rate": 6.241806263656228e-05,
"loss": 1.6086,
"step": 590
},
{
"epoch": 0.4087843679750994,
"grad_norm": 0.9651972055435181,
"learning_rate": 6.234522942461762e-05,
"loss": 2.0356,
"step": 591
},
{
"epoch": 0.4094760504928238,
"grad_norm": 0.9631683826446533,
"learning_rate": 6.227239621267298e-05,
"loss": 2.1457,
"step": 592
},
{
"epoch": 0.41016773301054815,
"grad_norm": 0.6648775339126587,
"learning_rate": 6.219956300072833e-05,
"loss": 1.3306,
"step": 593
},
{
"epoch": 0.41085941552827254,
"grad_norm": 1.3052250146865845,
"learning_rate": 6.212672978878369e-05,
"loss": 1.8813,
"step": 594
},
{
"epoch": 0.4115510980459969,
"grad_norm": 0.8607529401779175,
"learning_rate": 6.205389657683904e-05,
"loss": 1.9435,
"step": 595
},
{
"epoch": 0.41224278056372127,
"grad_norm": 0.7908481359481812,
"learning_rate": 6.19810633648944e-05,
"loss": 1.7544,
"step": 596
},
{
"epoch": 0.4129344630814456,
"grad_norm": 0.8440048694610596,
"learning_rate": 6.190823015294975e-05,
"loss": 1.8535,
"step": 597
},
{
"epoch": 0.41362614559917,
"grad_norm": 1.0176901817321777,
"learning_rate": 6.18353969410051e-05,
"loss": 1.5932,
"step": 598
},
{
"epoch": 0.4143178281168943,
"grad_norm": 0.8359641432762146,
"learning_rate": 6.176256372906046e-05,
"loss": 1.6704,
"step": 599
},
{
"epoch": 0.4150095106346187,
"grad_norm": 0.8534523248672485,
"learning_rate": 6.16897305171158e-05,
"loss": 1.7888,
"step": 600
},
{
"epoch": 0.41570119315234305,
"grad_norm": 0.8510312438011169,
"learning_rate": 6.161689730517117e-05,
"loss": 1.492,
"step": 601
},
{
"epoch": 0.41639287567006744,
"grad_norm": 0.9183124303817749,
"learning_rate": 6.154406409322651e-05,
"loss": 1.1547,
"step": 602
},
{
"epoch": 0.4170845581877918,
"grad_norm": 0.7638151049613953,
"learning_rate": 6.147123088128187e-05,
"loss": 1.198,
"step": 603
},
{
"epoch": 0.41777624070551617,
"grad_norm": 0.835200846195221,
"learning_rate": 6.139839766933721e-05,
"loss": 1.8262,
"step": 604
},
{
"epoch": 0.41846792322324056,
"grad_norm": 0.961685836315155,
"learning_rate": 6.132556445739257e-05,
"loss": 2.0815,
"step": 605
},
{
"epoch": 0.4191596057409649,
"grad_norm": 1.1949551105499268,
"learning_rate": 6.125273124544793e-05,
"loss": 2.0426,
"step": 606
},
{
"epoch": 0.4198512882586893,
"grad_norm": 0.6756623983383179,
"learning_rate": 6.117989803350328e-05,
"loss": 1.1775,
"step": 607
},
{
"epoch": 0.4205429707764136,
"grad_norm": 0.8606211543083191,
"learning_rate": 6.110706482155864e-05,
"loss": 1.1293,
"step": 608
},
{
"epoch": 0.421234653294138,
"grad_norm": 0.823352575302124,
"learning_rate": 6.103423160961399e-05,
"loss": 1.7001,
"step": 609
},
{
"epoch": 0.42192633581186234,
"grad_norm": 1.0490659475326538,
"learning_rate": 6.096139839766935e-05,
"loss": 1.6179,
"step": 610
},
{
"epoch": 0.42261801832958673,
"grad_norm": 0.972795307636261,
"learning_rate": 6.088856518572469e-05,
"loss": 2.1116,
"step": 611
},
{
"epoch": 0.42330970084731107,
"grad_norm": 0.795642077922821,
"learning_rate": 6.081573197378004e-05,
"loss": 1.7389,
"step": 612
},
{
"epoch": 0.42400138336503546,
"grad_norm": 0.796417772769928,
"learning_rate": 6.07428987618354e-05,
"loss": 1.2999,
"step": 613
},
{
"epoch": 0.4246930658827598,
"grad_norm": 0.8704690933227539,
"learning_rate": 6.067006554989075e-05,
"loss": 1.9508,
"step": 614
},
{
"epoch": 0.4253847484004842,
"grad_norm": 0.899750292301178,
"learning_rate": 6.0597232337946106e-05,
"loss": 1.9528,
"step": 615
},
{
"epoch": 0.4260764309182085,
"grad_norm": 0.7315880656242371,
"learning_rate": 6.052439912600146e-05,
"loss": 1.6963,
"step": 616
},
{
"epoch": 0.4267681134359329,
"grad_norm": 0.5647531151771545,
"learning_rate": 6.045156591405682e-05,
"loss": 1.1703,
"step": 617
},
{
"epoch": 0.4274597959536573,
"grad_norm": 0.9686012268066406,
"learning_rate": 6.037873270211216e-05,
"loss": 1.9199,
"step": 618
},
{
"epoch": 0.42815147847138163,
"grad_norm": 0.692129909992218,
"learning_rate": 6.0305899490167516e-05,
"loss": 0.9099,
"step": 619
},
{
"epoch": 0.428843160989106,
"grad_norm": 0.823519229888916,
"learning_rate": 6.023306627822287e-05,
"loss": 1.8828,
"step": 620
},
{
"epoch": 0.42953484350683035,
"grad_norm": 0.8990175127983093,
"learning_rate": 6.0160233066278225e-05,
"loss": 1.4318,
"step": 621
},
{
"epoch": 0.43022652602455475,
"grad_norm": 0.8949781656265259,
"learning_rate": 6.008739985433358e-05,
"loss": 2.0491,
"step": 622
},
{
"epoch": 0.4309182085422791,
"grad_norm": 1.0537045001983643,
"learning_rate": 6.001456664238893e-05,
"loss": 1.8443,
"step": 623
},
{
"epoch": 0.43160989106000347,
"grad_norm": 0.647614598274231,
"learning_rate": 5.994173343044428e-05,
"loss": 1.7638,
"step": 624
},
{
"epoch": 0.4323015735777278,
"grad_norm": 0.9621549844741821,
"learning_rate": 5.9868900218499635e-05,
"loss": 1.6114,
"step": 625
},
{
"epoch": 0.4329932560954522,
"grad_norm": 1.032599925994873,
"learning_rate": 5.979606700655499e-05,
"loss": 1.9123,
"step": 626
},
{
"epoch": 0.43368493861317653,
"grad_norm": 0.7238906621932983,
"learning_rate": 5.9723233794610344e-05,
"loss": 1.4896,
"step": 627
},
{
"epoch": 0.4343766211309009,
"grad_norm": 0.7447595596313477,
"learning_rate": 5.96504005826657e-05,
"loss": 1.5476,
"step": 628
},
{
"epoch": 0.43506830364862525,
"grad_norm": 0.6987977027893066,
"learning_rate": 5.957756737072105e-05,
"loss": 1.4778,
"step": 629
},
{
"epoch": 0.43575998616634964,
"grad_norm": 0.8449330925941467,
"learning_rate": 5.950473415877641e-05,
"loss": 1.8849,
"step": 630
},
{
"epoch": 0.43645166868407403,
"grad_norm": 1.0556013584136963,
"learning_rate": 5.9431900946831754e-05,
"loss": 2.2383,
"step": 631
},
{
"epoch": 0.43714335120179837,
"grad_norm": 0.6370812654495239,
"learning_rate": 5.935906773488711e-05,
"loss": 1.5145,
"step": 632
},
{
"epoch": 0.43783503371952276,
"grad_norm": 0.7922269701957703,
"learning_rate": 5.928623452294246e-05,
"loss": 1.5945,
"step": 633
},
{
"epoch": 0.4385267162372471,
"grad_norm": 0.9030670523643494,
"learning_rate": 5.921340131099782e-05,
"loss": 1.9727,
"step": 634
},
{
"epoch": 0.4392183987549715,
"grad_norm": 0.5330315828323364,
"learning_rate": 5.914056809905317e-05,
"loss": 1.1115,
"step": 635
},
{
"epoch": 0.4399100812726958,
"grad_norm": 0.7586291432380676,
"learning_rate": 5.9067734887108526e-05,
"loss": 1.3721,
"step": 636
},
{
"epoch": 0.4406017637904202,
"grad_norm": 0.8726906776428223,
"learning_rate": 5.899490167516388e-05,
"loss": 1.6429,
"step": 637
},
{
"epoch": 0.44129344630814454,
"grad_norm": 0.8924012184143066,
"learning_rate": 5.892206846321923e-05,
"loss": 1.7596,
"step": 638
},
{
"epoch": 0.44198512882586893,
"grad_norm": 0.9185534715652466,
"learning_rate": 5.884923525127458e-05,
"loss": 1.8224,
"step": 639
},
{
"epoch": 0.44267681134359327,
"grad_norm": 0.8137707114219666,
"learning_rate": 5.8776402039329936e-05,
"loss": 1.7631,
"step": 640
},
{
"epoch": 0.44336849386131766,
"grad_norm": 0.7134371399879456,
"learning_rate": 5.870356882738529e-05,
"loss": 1.8246,
"step": 641
},
{
"epoch": 0.444060176379042,
"grad_norm": 0.7280315160751343,
"learning_rate": 5.8630735615440645e-05,
"loss": 1.5007,
"step": 642
},
{
"epoch": 0.4447518588967664,
"grad_norm": 0.6889392733573914,
"learning_rate": 5.8557902403496e-05,
"loss": 1.2957,
"step": 643
},
{
"epoch": 0.4454435414144908,
"grad_norm": 0.6772429943084717,
"learning_rate": 5.8485069191551346e-05,
"loss": 1.2119,
"step": 644
},
{
"epoch": 0.4461352239322151,
"grad_norm": 0.7719807028770447,
"learning_rate": 5.84122359796067e-05,
"loss": 1.0095,
"step": 645
},
{
"epoch": 0.4468269064499395,
"grad_norm": 0.9226011037826538,
"learning_rate": 5.8339402767662055e-05,
"loss": 1.7821,
"step": 646
},
{
"epoch": 0.44751858896766383,
"grad_norm": 0.6220186948776245,
"learning_rate": 5.826656955571741e-05,
"loss": 1.1275,
"step": 647
},
{
"epoch": 0.4482102714853882,
"grad_norm": 0.6376112699508667,
"learning_rate": 5.8193736343772763e-05,
"loss": 1.4188,
"step": 648
},
{
"epoch": 0.44890195400311256,
"grad_norm": 0.7491716146469116,
"learning_rate": 5.812090313182812e-05,
"loss": 1.5822,
"step": 649
},
{
"epoch": 0.44959363652083695,
"grad_norm": 0.6484271287918091,
"learning_rate": 5.804806991988347e-05,
"loss": 1.6037,
"step": 650
},
{
"epoch": 0.4502853190385613,
"grad_norm": 1.5858505964279175,
"learning_rate": 5.797523670793882e-05,
"loss": 1.4725,
"step": 651
},
{
"epoch": 0.4509770015562857,
"grad_norm": 1.0042520761489868,
"learning_rate": 5.7902403495994174e-05,
"loss": 1.6445,
"step": 652
},
{
"epoch": 0.45166868407401,
"grad_norm": 1.0305542945861816,
"learning_rate": 5.782957028404953e-05,
"loss": 1.4438,
"step": 653
},
{
"epoch": 0.4523603665917344,
"grad_norm": 0.7213695049285889,
"learning_rate": 5.775673707210488e-05,
"loss": 1.6495,
"step": 654
},
{
"epoch": 0.45305204910945873,
"grad_norm": 0.8074948191642761,
"learning_rate": 5.768390386016024e-05,
"loss": 1.1349,
"step": 655
},
{
"epoch": 0.4537437316271831,
"grad_norm": 0.9579366445541382,
"learning_rate": 5.761107064821559e-05,
"loss": 1.421,
"step": 656
},
{
"epoch": 0.4544354141449075,
"grad_norm": 0.7831091284751892,
"learning_rate": 5.7538237436270945e-05,
"loss": 1.8771,
"step": 657
},
{
"epoch": 0.45512709666263185,
"grad_norm": 0.7758937478065491,
"learning_rate": 5.746540422432629e-05,
"loss": 1.7355,
"step": 658
},
{
"epoch": 0.45581877918035624,
"grad_norm": 0.8883340358734131,
"learning_rate": 5.739257101238165e-05,
"loss": 1.7985,
"step": 659
},
{
"epoch": 0.45651046169808057,
"grad_norm": 0.7148769497871399,
"learning_rate": 5.7319737800437e-05,
"loss": 2.0838,
"step": 660
},
{
"epoch": 0.45720214421580496,
"grad_norm": 0.7475634217262268,
"learning_rate": 5.7246904588492356e-05,
"loss": 1.1599,
"step": 661
},
{
"epoch": 0.4578938267335293,
"grad_norm": 0.9443761706352234,
"learning_rate": 5.717407137654771e-05,
"loss": 1.9477,
"step": 662
},
{
"epoch": 0.4585855092512537,
"grad_norm": 0.8133770823478699,
"learning_rate": 5.7101238164603064e-05,
"loss": 1.8349,
"step": 663
},
{
"epoch": 0.459277191768978,
"grad_norm": 0.7030238509178162,
"learning_rate": 5.702840495265841e-05,
"loss": 0.7075,
"step": 664
},
{
"epoch": 0.4599688742867024,
"grad_norm": 0.6615720987319946,
"learning_rate": 5.6955571740713766e-05,
"loss": 1.4996,
"step": 665
},
{
"epoch": 0.46066055680442675,
"grad_norm": 0.9358487725257874,
"learning_rate": 5.688273852876912e-05,
"loss": 1.4268,
"step": 666
},
{
"epoch": 0.46135223932215114,
"grad_norm": 0.6841732859611511,
"learning_rate": 5.6809905316824475e-05,
"loss": 1.7628,
"step": 667
},
{
"epoch": 0.46204392183987547,
"grad_norm": 0.7229800820350647,
"learning_rate": 5.673707210487983e-05,
"loss": 1.3031,
"step": 668
},
{
"epoch": 0.46273560435759986,
"grad_norm": 0.655465304851532,
"learning_rate": 5.666423889293518e-05,
"loss": 1.9593,
"step": 669
},
{
"epoch": 0.46342728687532425,
"grad_norm": 0.6893927454948425,
"learning_rate": 5.659140568099054e-05,
"loss": 0.9872,
"step": 670
},
{
"epoch": 0.4641189693930486,
"grad_norm": 1.0941834449768066,
"learning_rate": 5.6518572469045885e-05,
"loss": 1.4924,
"step": 671
},
{
"epoch": 0.464810651910773,
"grad_norm": 0.893545925617218,
"learning_rate": 5.644573925710124e-05,
"loss": 1.531,
"step": 672
},
{
"epoch": 0.4655023344284973,
"grad_norm": 0.7984673976898193,
"learning_rate": 5.6372906045156594e-05,
"loss": 1.79,
"step": 673
},
{
"epoch": 0.4661940169462217,
"grad_norm": 0.5964581370353699,
"learning_rate": 5.630007283321195e-05,
"loss": 1.1119,
"step": 674
},
{
"epoch": 0.46688569946394604,
"grad_norm": 0.7665301561355591,
"learning_rate": 5.62272396212673e-05,
"loss": 1.3566,
"step": 675
},
{
"epoch": 0.4675773819816704,
"grad_norm": 0.8093408346176147,
"learning_rate": 5.6154406409322656e-05,
"loss": 1.5565,
"step": 676
},
{
"epoch": 0.46826906449939476,
"grad_norm": 0.8169633150100708,
"learning_rate": 5.608157319737801e-05,
"loss": 2.2779,
"step": 677
},
{
"epoch": 0.46896074701711915,
"grad_norm": 0.8401588797569275,
"learning_rate": 5.600873998543336e-05,
"loss": 1.9076,
"step": 678
},
{
"epoch": 0.4696524295348435,
"grad_norm": 1.4105268716812134,
"learning_rate": 5.593590677348871e-05,
"loss": 1.4305,
"step": 679
},
{
"epoch": 0.4703441120525679,
"grad_norm": 1.0228116512298584,
"learning_rate": 5.586307356154407e-05,
"loss": 2.0295,
"step": 680
},
{
"epoch": 0.4710357945702922,
"grad_norm": 0.8564938306808472,
"learning_rate": 5.579024034959942e-05,
"loss": 1.6077,
"step": 681
},
{
"epoch": 0.4717274770880166,
"grad_norm": 0.8669772148132324,
"learning_rate": 5.5717407137654775e-05,
"loss": 1.584,
"step": 682
},
{
"epoch": 0.472419159605741,
"grad_norm": 0.5886315703392029,
"learning_rate": 5.564457392571013e-05,
"loss": 1.4101,
"step": 683
},
{
"epoch": 0.4731108421234653,
"grad_norm": 0.8294202089309692,
"learning_rate": 5.557174071376548e-05,
"loss": 1.8993,
"step": 684
},
{
"epoch": 0.4738025246411897,
"grad_norm": 1.0754139423370361,
"learning_rate": 5.549890750182083e-05,
"loss": 1.654,
"step": 685
},
{
"epoch": 0.47449420715891405,
"grad_norm": 0.7597669363021851,
"learning_rate": 5.5426074289876186e-05,
"loss": 2.0938,
"step": 686
},
{
"epoch": 0.47518588967663844,
"grad_norm": 0.7472283244132996,
"learning_rate": 5.535324107793154e-05,
"loss": 1.6874,
"step": 687
},
{
"epoch": 0.4758775721943628,
"grad_norm": 1.1373130083084106,
"learning_rate": 5.5280407865986894e-05,
"loss": 1.5681,
"step": 688
},
{
"epoch": 0.47656925471208716,
"grad_norm": 0.6962531805038452,
"learning_rate": 5.520757465404225e-05,
"loss": 1.2423,
"step": 689
},
{
"epoch": 0.4772609372298115,
"grad_norm": 0.8981777429580688,
"learning_rate": 5.51347414420976e-05,
"loss": 1.4854,
"step": 690
},
{
"epoch": 0.4779526197475359,
"grad_norm": 0.7226567268371582,
"learning_rate": 5.506190823015295e-05,
"loss": 2.2022,
"step": 691
},
{
"epoch": 0.4786443022652602,
"grad_norm": 0.9245135188102722,
"learning_rate": 5.4989075018208305e-05,
"loss": 1.3642,
"step": 692
},
{
"epoch": 0.4793359847829846,
"grad_norm": 1.0049363374710083,
"learning_rate": 5.491624180626366e-05,
"loss": 1.9998,
"step": 693
},
{
"epoch": 0.48002766730070895,
"grad_norm": 0.6878874897956848,
"learning_rate": 5.484340859431901e-05,
"loss": 1.9002,
"step": 694
},
{
"epoch": 0.48071934981843334,
"grad_norm": 1.1163276433944702,
"learning_rate": 5.477057538237437e-05,
"loss": 1.1385,
"step": 695
},
{
"epoch": 0.48141103233615773,
"grad_norm": 0.7806943655014038,
"learning_rate": 5.469774217042972e-05,
"loss": 1.0122,
"step": 696
},
{
"epoch": 0.48210271485388206,
"grad_norm": 0.9799820184707642,
"learning_rate": 5.4624908958485076e-05,
"loss": 2.0761,
"step": 697
},
{
"epoch": 0.48279439737160645,
"grad_norm": 0.5801637768745422,
"learning_rate": 5.4552075746540424e-05,
"loss": 0.8659,
"step": 698
},
{
"epoch": 0.4834860798893308,
"grad_norm": 0.7793147563934326,
"learning_rate": 5.447924253459578e-05,
"loss": 1.4955,
"step": 699
},
{
"epoch": 0.4841777624070552,
"grad_norm": 0.7253298759460449,
"learning_rate": 5.440640932265113e-05,
"loss": 0.8938,
"step": 700
},
{
"epoch": 0.4848694449247795,
"grad_norm": 0.8305007219314575,
"learning_rate": 5.4333576110706487e-05,
"loss": 1.3896,
"step": 701
},
{
"epoch": 0.4855611274425039,
"grad_norm": 0.8099757432937622,
"learning_rate": 5.426074289876184e-05,
"loss": 1.7134,
"step": 702
},
{
"epoch": 0.48625280996022824,
"grad_norm": 0.7706416845321655,
"learning_rate": 5.4187909686817195e-05,
"loss": 0.8483,
"step": 703
},
{
"epoch": 0.48694449247795263,
"grad_norm": 2.7785534858703613,
"learning_rate": 5.4115076474872536e-05,
"loss": 1.5148,
"step": 704
},
{
"epoch": 0.48763617499567696,
"grad_norm": 0.7698436975479126,
"learning_rate": 5.40422432629279e-05,
"loss": 2.031,
"step": 705
},
{
"epoch": 0.48832785751340135,
"grad_norm": 0.8159623742103577,
"learning_rate": 5.396941005098325e-05,
"loss": 2.0041,
"step": 706
},
{
"epoch": 0.4890195400311257,
"grad_norm": 0.6865976452827454,
"learning_rate": 5.3896576839038606e-05,
"loss": 1.2877,
"step": 707
},
{
"epoch": 0.4897112225488501,
"grad_norm": 0.9447488784790039,
"learning_rate": 5.382374362709396e-05,
"loss": 2.1087,
"step": 708
},
{
"epoch": 0.49040290506657447,
"grad_norm": 0.9315760731697083,
"learning_rate": 5.3750910415149314e-05,
"loss": 1.5459,
"step": 709
},
{
"epoch": 0.4910945875842988,
"grad_norm": 0.7628729939460754,
"learning_rate": 5.367807720320467e-05,
"loss": 1.6846,
"step": 710
},
{
"epoch": 0.4917862701020232,
"grad_norm": 0.7759138345718384,
"learning_rate": 5.360524399126001e-05,
"loss": 1.6179,
"step": 711
},
{
"epoch": 0.4924779526197475,
"grad_norm": 0.7866867184638977,
"learning_rate": 5.353241077931537e-05,
"loss": 2.2264,
"step": 712
},
{
"epoch": 0.4931696351374719,
"grad_norm": 0.7560875415802002,
"learning_rate": 5.3459577567370724e-05,
"loss": 1.5638,
"step": 713
},
{
"epoch": 0.49386131765519625,
"grad_norm": 0.6578519940376282,
"learning_rate": 5.338674435542608e-05,
"loss": 1.0194,
"step": 714
},
{
"epoch": 0.49455300017292064,
"grad_norm": 0.8581376671791077,
"learning_rate": 5.331391114348143e-05,
"loss": 1.7412,
"step": 715
},
{
"epoch": 0.495244682690645,
"grad_norm": 0.8838049173355103,
"learning_rate": 5.324107793153679e-05,
"loss": 2.0492,
"step": 716
},
{
"epoch": 0.49593636520836937,
"grad_norm": 0.7301944494247437,
"learning_rate": 5.316824471959214e-05,
"loss": 1.8718,
"step": 717
},
{
"epoch": 0.4966280477260937,
"grad_norm": 0.8501876592636108,
"learning_rate": 5.309541150764748e-05,
"loss": 1.9824,
"step": 718
},
{
"epoch": 0.4973197302438181,
"grad_norm": 0.8642913699150085,
"learning_rate": 5.3022578295702843e-05,
"loss": 2.0727,
"step": 719
},
{
"epoch": 0.4980114127615424,
"grad_norm": 0.7409586310386658,
"learning_rate": 5.29497450837582e-05,
"loss": 2.0216,
"step": 720
},
{
"epoch": 0.4987030952792668,
"grad_norm": 1.2024013996124268,
"learning_rate": 5.287691187181355e-05,
"loss": 1.534,
"step": 721
},
{
"epoch": 0.4993947777969912,
"grad_norm": 0.7344438433647156,
"learning_rate": 5.2804078659868906e-05,
"loss": 1.7692,
"step": 722
},
{
"epoch": 0.5000864603147156,
"grad_norm": 0.8432995080947876,
"learning_rate": 5.273124544792426e-05,
"loss": 1.1382,
"step": 723
},
{
"epoch": 0.5007781428324399,
"grad_norm": 0.8637827038764954,
"learning_rate": 5.26584122359796e-05,
"loss": 2.0577,
"step": 724
},
{
"epoch": 0.5014698253501643,
"grad_norm": 0.7211031913757324,
"learning_rate": 5.2585579024034956e-05,
"loss": 1.0286,
"step": 725
},
{
"epoch": 0.5021615078678886,
"grad_norm": 0.8980405330657959,
"learning_rate": 5.251274581209032e-05,
"loss": 1.6481,
"step": 726
},
{
"epoch": 0.502853190385613,
"grad_norm": 1.0838844776153564,
"learning_rate": 5.243991260014567e-05,
"loss": 1.8133,
"step": 727
},
{
"epoch": 0.5035448729033374,
"grad_norm": 0.7398082613945007,
"learning_rate": 5.2367079388201025e-05,
"loss": 1.9674,
"step": 728
},
{
"epoch": 0.5042365554210617,
"grad_norm": 0.9297565817832947,
"learning_rate": 5.229424617625638e-05,
"loss": 1.5748,
"step": 729
},
{
"epoch": 0.504928237938786,
"grad_norm": 1.0690850019454956,
"learning_rate": 5.2221412964311734e-05,
"loss": 1.5343,
"step": 730
},
{
"epoch": 0.5056199204565105,
"grad_norm": 0.8005755543708801,
"learning_rate": 5.2148579752367075e-05,
"loss": 1.7446,
"step": 731
},
{
"epoch": 0.5063116029742348,
"grad_norm": 0.9279308915138245,
"learning_rate": 5.207574654042243e-05,
"loss": 2.0124,
"step": 732
},
{
"epoch": 0.5070032854919592,
"grad_norm": 0.8205484747886658,
"learning_rate": 5.200291332847779e-05,
"loss": 1.9278,
"step": 733
},
{
"epoch": 0.5076949680096835,
"grad_norm": 1.0283253192901611,
"learning_rate": 5.1930080116533144e-05,
"loss": 1.6584,
"step": 734
},
{
"epoch": 0.508386650527408,
"grad_norm": 0.9164149165153503,
"learning_rate": 5.18572469045885e-05,
"loss": 1.5658,
"step": 735
},
{
"epoch": 0.5090783330451323,
"grad_norm": 0.8904675841331482,
"learning_rate": 5.178441369264385e-05,
"loss": 2.054,
"step": 736
},
{
"epoch": 0.5097700155628566,
"grad_norm": 0.731334924697876,
"learning_rate": 5.171158048069921e-05,
"loss": 1.8062,
"step": 737
},
{
"epoch": 0.5104616980805811,
"grad_norm": 1.436216115951538,
"learning_rate": 5.163874726875455e-05,
"loss": 2.4866,
"step": 738
},
{
"epoch": 0.5111533805983054,
"grad_norm": 0.7042586803436279,
"learning_rate": 5.15659140568099e-05,
"loss": 1.6228,
"step": 739
},
{
"epoch": 0.5118450631160297,
"grad_norm": 0.63124018907547,
"learning_rate": 5.149308084486526e-05,
"loss": 0.9537,
"step": 740
},
{
"epoch": 0.5125367456337541,
"grad_norm": 0.8653015494346619,
"learning_rate": 5.142024763292062e-05,
"loss": 2.0306,
"step": 741
},
{
"epoch": 0.5132284281514785,
"grad_norm": 0.8902332782745361,
"learning_rate": 5.134741442097597e-05,
"loss": 2.0709,
"step": 742
},
{
"epoch": 0.5139201106692028,
"grad_norm": 0.8607166409492493,
"learning_rate": 5.1274581209031326e-05,
"loss": 1.635,
"step": 743
},
{
"epoch": 0.5146117931869272,
"grad_norm": 0.8155879974365234,
"learning_rate": 5.120174799708667e-05,
"loss": 1.0902,
"step": 744
},
{
"epoch": 0.5153034757046515,
"grad_norm": 0.8006990551948547,
"learning_rate": 5.112891478514202e-05,
"loss": 1.2486,
"step": 745
},
{
"epoch": 0.515995158222376,
"grad_norm": 0.7772015929222107,
"learning_rate": 5.1056081573197375e-05,
"loss": 2.1048,
"step": 746
},
{
"epoch": 0.5166868407401003,
"grad_norm": 0.7376934289932251,
"learning_rate": 5.0983248361252736e-05,
"loss": 1.9748,
"step": 747
},
{
"epoch": 0.5173785232578246,
"grad_norm": 0.9872801899909973,
"learning_rate": 5.091041514930809e-05,
"loss": 1.3626,
"step": 748
},
{
"epoch": 0.5180702057755491,
"grad_norm": 0.6437541246414185,
"learning_rate": 5.0837581937363445e-05,
"loss": 1.8523,
"step": 749
},
{
"epoch": 0.5187618882932734,
"grad_norm": 0.8643835186958313,
"learning_rate": 5.07647487254188e-05,
"loss": 1.5464,
"step": 750
},
{
"epoch": 0.5194535708109977,
"grad_norm": 0.9346618056297302,
"learning_rate": 5.069191551347414e-05,
"loss": 1.7027,
"step": 751
},
{
"epoch": 0.5201452533287221,
"grad_norm": 0.8096154928207397,
"learning_rate": 5.0619082301529494e-05,
"loss": 1.4638,
"step": 752
},
{
"epoch": 0.5208369358464465,
"grad_norm": 0.8895930051803589,
"learning_rate": 5.054624908958485e-05,
"loss": 1.3399,
"step": 753
},
{
"epoch": 0.5215286183641709,
"grad_norm": 0.8496235013008118,
"learning_rate": 5.047341587764021e-05,
"loss": 0.9693,
"step": 754
},
{
"epoch": 0.5222203008818952,
"grad_norm": 0.7915313243865967,
"learning_rate": 5.0400582665695564e-05,
"loss": 1.4108,
"step": 755
},
{
"epoch": 0.5229119833996195,
"grad_norm": 0.8515884876251221,
"learning_rate": 5.032774945375092e-05,
"loss": 1.7595,
"step": 756
},
{
"epoch": 0.523603665917344,
"grad_norm": 0.7750046849250793,
"learning_rate": 5.025491624180627e-05,
"loss": 1.73,
"step": 757
},
{
"epoch": 0.5242953484350683,
"grad_norm": 0.8660963177680969,
"learning_rate": 5.018208302986161e-05,
"loss": 1.8401,
"step": 758
},
{
"epoch": 0.5249870309527926,
"grad_norm": 1.214325189590454,
"learning_rate": 5.010924981791697e-05,
"loss": 1.5471,
"step": 759
},
{
"epoch": 0.525678713470517,
"grad_norm": 1.0521632432937622,
"learning_rate": 5.003641660597232e-05,
"loss": 2.0549,
"step": 760
},
{
"epoch": 0.5263703959882414,
"grad_norm": 0.8383963108062744,
"learning_rate": 4.996358339402768e-05,
"loss": 1.9966,
"step": 761
},
{
"epoch": 0.5270620785059658,
"grad_norm": 0.9236989617347717,
"learning_rate": 4.989075018208304e-05,
"loss": 1.3734,
"step": 762
},
{
"epoch": 0.5277537610236901,
"grad_norm": 0.7525227665901184,
"learning_rate": 4.9817916970138385e-05,
"loss": 1.2582,
"step": 763
},
{
"epoch": 0.5284454435414145,
"grad_norm": 0.8061596751213074,
"learning_rate": 4.974508375819374e-05,
"loss": 2.0085,
"step": 764
},
{
"epoch": 0.5291371260591389,
"grad_norm": 1.0033149719238281,
"learning_rate": 4.967225054624909e-05,
"loss": 1.5858,
"step": 765
},
{
"epoch": 0.5298288085768632,
"grad_norm": 1.3156208992004395,
"learning_rate": 4.959941733430444e-05,
"loss": 1.8546,
"step": 766
},
{
"epoch": 0.5305204910945875,
"grad_norm": 0.7604823112487793,
"learning_rate": 4.9526584122359795e-05,
"loss": 1.7595,
"step": 767
},
{
"epoch": 0.531212173612312,
"grad_norm": 0.9594501852989197,
"learning_rate": 4.9453750910415156e-05,
"loss": 1.507,
"step": 768
},
{
"epoch": 0.5319038561300363,
"grad_norm": 1.176438808441162,
"learning_rate": 4.9380917698470504e-05,
"loss": 1.6748,
"step": 769
},
{
"epoch": 0.5325955386477607,
"grad_norm": 0.6893585324287415,
"learning_rate": 4.930808448652586e-05,
"loss": 1.326,
"step": 770
},
{
"epoch": 0.533287221165485,
"grad_norm": 0.8642687797546387,
"learning_rate": 4.923525127458121e-05,
"loss": 1.1461,
"step": 771
},
{
"epoch": 0.5339789036832094,
"grad_norm": 0.6313591003417969,
"learning_rate": 4.9162418062636567e-05,
"loss": 1.5494,
"step": 772
},
{
"epoch": 0.5346705862009338,
"grad_norm": 0.7172280550003052,
"learning_rate": 4.9089584850691914e-05,
"loss": 1.4831,
"step": 773
},
{
"epoch": 0.5353622687186581,
"grad_norm": 0.8146588206291199,
"learning_rate": 4.901675163874727e-05,
"loss": 1.913,
"step": 774
},
{
"epoch": 0.5360539512363826,
"grad_norm": 0.9037981033325195,
"learning_rate": 4.894391842680263e-05,
"loss": 1.7836,
"step": 775
},
{
"epoch": 0.5367456337541069,
"grad_norm": 0.7782710194587708,
"learning_rate": 4.887108521485798e-05,
"loss": 1.4689,
"step": 776
},
{
"epoch": 0.5374373162718312,
"grad_norm": 1.111527681350708,
"learning_rate": 4.879825200291333e-05,
"loss": 1.6938,
"step": 777
},
{
"epoch": 0.5381289987895556,
"grad_norm": 1.2994364500045776,
"learning_rate": 4.8725418790968686e-05,
"loss": 1.529,
"step": 778
},
{
"epoch": 0.53882068130728,
"grad_norm": 0.7461956143379211,
"learning_rate": 4.865258557902403e-05,
"loss": 1.1475,
"step": 779
},
{
"epoch": 0.5395123638250043,
"grad_norm": 0.6900501847267151,
"learning_rate": 4.857975236707939e-05,
"loss": 1.8377,
"step": 780
},
{
"epoch": 0.5402040463427287,
"grad_norm": 0.7007532715797424,
"learning_rate": 4.850691915513474e-05,
"loss": 1.4117,
"step": 781
},
{
"epoch": 0.540895728860453,
"grad_norm": 0.8662421703338623,
"learning_rate": 4.84340859431901e-05,
"loss": 1.7831,
"step": 782
},
{
"epoch": 0.5415874113781775,
"grad_norm": 1.2154408693313599,
"learning_rate": 4.836125273124545e-05,
"loss": 1.9868,
"step": 783
},
{
"epoch": 0.5422790938959018,
"grad_norm": 0.822338342666626,
"learning_rate": 4.8288419519300804e-05,
"loss": 1.8789,
"step": 784
},
{
"epoch": 0.5429707764136261,
"grad_norm": 0.9804028868675232,
"learning_rate": 4.821558630735616e-05,
"loss": 2.2036,
"step": 785
},
{
"epoch": 0.5436624589313505,
"grad_norm": 0.9605854749679565,
"learning_rate": 4.8142753095411506e-05,
"loss": 2.0455,
"step": 786
},
{
"epoch": 0.5443541414490749,
"grad_norm": 0.7948291301727295,
"learning_rate": 4.806991988346686e-05,
"loss": 1.529,
"step": 787
},
{
"epoch": 0.5450458239667992,
"grad_norm": 0.7471349239349365,
"learning_rate": 4.7997086671522215e-05,
"loss": 1.5366,
"step": 788
},
{
"epoch": 0.5457375064845236,
"grad_norm": 1.0526931285858154,
"learning_rate": 4.792425345957757e-05,
"loss": 1.2396,
"step": 789
},
{
"epoch": 0.546429189002248,
"grad_norm": 1.0676295757293701,
"learning_rate": 4.7851420247632923e-05,
"loss": 1.6639,
"step": 790
},
{
"epoch": 0.5471208715199724,
"grad_norm": 0.9911693334579468,
"learning_rate": 4.777858703568828e-05,
"loss": 1.8713,
"step": 791
},
{
"epoch": 0.5478125540376967,
"grad_norm": 0.5897760391235352,
"learning_rate": 4.770575382374363e-05,
"loss": 1.4593,
"step": 792
},
{
"epoch": 0.548504236555421,
"grad_norm": 0.836234986782074,
"learning_rate": 4.763292061179898e-05,
"loss": 1.3373,
"step": 793
},
{
"epoch": 0.5491959190731455,
"grad_norm": 1.0234719514846802,
"learning_rate": 4.7560087399854334e-05,
"loss": 1.3758,
"step": 794
},
{
"epoch": 0.5498876015908698,
"grad_norm": 0.6555935144424438,
"learning_rate": 4.748725418790969e-05,
"loss": 1.0075,
"step": 795
},
{
"epoch": 0.5505792841085941,
"grad_norm": 0.7756921052932739,
"learning_rate": 4.741442097596504e-05,
"loss": 1.5594,
"step": 796
},
{
"epoch": 0.5512709666263185,
"grad_norm": 0.8114323616027832,
"learning_rate": 4.73415877640204e-05,
"loss": 1.9788,
"step": 797
},
{
"epoch": 0.5519626491440429,
"grad_norm": 1.0409117937088013,
"learning_rate": 4.726875455207575e-05,
"loss": 1.4971,
"step": 798
},
{
"epoch": 0.5526543316617673,
"grad_norm": 0.963651716709137,
"learning_rate": 4.71959213401311e-05,
"loss": 1.8012,
"step": 799
},
{
"epoch": 0.5533460141794916,
"grad_norm": 0.7546253204345703,
"learning_rate": 4.712308812818645e-05,
"loss": 2.0704,
"step": 800
},
{
"epoch": 0.554037696697216,
"grad_norm": 0.7837393283843994,
"learning_rate": 4.705025491624181e-05,
"loss": 1.9567,
"step": 801
},
{
"epoch": 0.5547293792149404,
"grad_norm": 0.9750471711158752,
"learning_rate": 4.697742170429716e-05,
"loss": 1.8202,
"step": 802
},
{
"epoch": 0.5554210617326647,
"grad_norm": 0.656570315361023,
"learning_rate": 4.6904588492352516e-05,
"loss": 1.9521,
"step": 803
},
{
"epoch": 0.556112744250389,
"grad_norm": 0.7636544108390808,
"learning_rate": 4.683175528040787e-05,
"loss": 1.6885,
"step": 804
},
{
"epoch": 0.5568044267681135,
"grad_norm": 0.6734775900840759,
"learning_rate": 4.6758922068463224e-05,
"loss": 1.4624,
"step": 805
},
{
"epoch": 0.5574961092858378,
"grad_norm": 0.7191459536552429,
"learning_rate": 4.668608885651857e-05,
"loss": 1.9275,
"step": 806
},
{
"epoch": 0.5581877918035621,
"grad_norm": 0.7927097082138062,
"learning_rate": 4.6613255644573926e-05,
"loss": 1.3038,
"step": 807
},
{
"epoch": 0.5588794743212865,
"grad_norm": 0.7056376934051514,
"learning_rate": 4.654042243262928e-05,
"loss": 1.4776,
"step": 808
},
{
"epoch": 0.5595711568390109,
"grad_norm": 0.6944011449813843,
"learning_rate": 4.6467589220684635e-05,
"loss": 1.6238,
"step": 809
},
{
"epoch": 0.5602628393567353,
"grad_norm": 0.6770850419998169,
"learning_rate": 4.639475600873999e-05,
"loss": 2.0007,
"step": 810
},
{
"epoch": 0.5609545218744596,
"grad_norm": 0.6012453436851501,
"learning_rate": 4.632192279679534e-05,
"loss": 1.3431,
"step": 811
},
{
"epoch": 0.5616462043921839,
"grad_norm": 0.8556742072105408,
"learning_rate": 4.62490895848507e-05,
"loss": 1.9744,
"step": 812
},
{
"epoch": 0.5623378869099084,
"grad_norm": 0.9271799921989441,
"learning_rate": 4.6176256372906045e-05,
"loss": 2.0651,
"step": 813
},
{
"epoch": 0.5630295694276327,
"grad_norm": 0.7125703692436218,
"learning_rate": 4.61034231609614e-05,
"loss": 1.519,
"step": 814
},
{
"epoch": 0.563721251945357,
"grad_norm": 0.721099853515625,
"learning_rate": 4.6030589949016754e-05,
"loss": 1.4397,
"step": 815
},
{
"epoch": 0.5644129344630815,
"grad_norm": 0.8622443675994873,
"learning_rate": 4.595775673707211e-05,
"loss": 1.5684,
"step": 816
},
{
"epoch": 0.5651046169808058,
"grad_norm": 0.8319584727287292,
"learning_rate": 4.588492352512746e-05,
"loss": 1.8106,
"step": 817
},
{
"epoch": 0.5657962994985302,
"grad_norm": 0.7659180760383606,
"learning_rate": 4.5812090313182816e-05,
"loss": 1.976,
"step": 818
},
{
"epoch": 0.5664879820162545,
"grad_norm": 0.658638060092926,
"learning_rate": 4.5739257101238164e-05,
"loss": 1.6156,
"step": 819
},
{
"epoch": 0.5671796645339789,
"grad_norm": 0.7831644415855408,
"learning_rate": 4.566642388929352e-05,
"loss": 1.7169,
"step": 820
},
{
"epoch": 0.5678713470517033,
"grad_norm": 0.8587795495986938,
"learning_rate": 4.559359067734887e-05,
"loss": 1.8784,
"step": 821
},
{
"epoch": 0.5685630295694276,
"grad_norm": 0.6490662693977356,
"learning_rate": 4.552075746540423e-05,
"loss": 1.2044,
"step": 822
},
{
"epoch": 0.569254712087152,
"grad_norm": 0.8098453879356384,
"learning_rate": 4.544792425345958e-05,
"loss": 2.0041,
"step": 823
},
{
"epoch": 0.5699463946048764,
"grad_norm": 0.9287192225456238,
"learning_rate": 4.5375091041514935e-05,
"loss": 2.0794,
"step": 824
},
{
"epoch": 0.5706380771226007,
"grad_norm": 0.6558881998062134,
"learning_rate": 4.530225782957029e-05,
"loss": 1.1153,
"step": 825
},
{
"epoch": 0.5713297596403251,
"grad_norm": 0.9448340535163879,
"learning_rate": 4.522942461762564e-05,
"loss": 1.8924,
"step": 826
},
{
"epoch": 0.5720214421580495,
"grad_norm": 0.9030715823173523,
"learning_rate": 4.515659140568099e-05,
"loss": 1.487,
"step": 827
},
{
"epoch": 0.5727131246757738,
"grad_norm": 0.8691999316215515,
"learning_rate": 4.5083758193736346e-05,
"loss": 1.709,
"step": 828
},
{
"epoch": 0.5734048071934982,
"grad_norm": 0.802944540977478,
"learning_rate": 4.50109249817917e-05,
"loss": 1.5163,
"step": 829
},
{
"epoch": 0.5740964897112225,
"grad_norm": 0.7917433381080627,
"learning_rate": 4.4938091769847054e-05,
"loss": 1.5238,
"step": 830
},
{
"epoch": 0.574788172228947,
"grad_norm": 0.8852903842926025,
"learning_rate": 4.486525855790241e-05,
"loss": 1.4459,
"step": 831
},
{
"epoch": 0.5754798547466713,
"grad_norm": 0.8966250419616699,
"learning_rate": 4.479242534595776e-05,
"loss": 1.4654,
"step": 832
},
{
"epoch": 0.5761715372643956,
"grad_norm": 1.3262107372283936,
"learning_rate": 4.471959213401311e-05,
"loss": 1.5453,
"step": 833
},
{
"epoch": 0.57686321978212,
"grad_norm": 0.7559188604354858,
"learning_rate": 4.4646758922068465e-05,
"loss": 1.5227,
"step": 834
},
{
"epoch": 0.5775549022998444,
"grad_norm": 0.9680230617523193,
"learning_rate": 4.457392571012382e-05,
"loss": 1.8038,
"step": 835
},
{
"epoch": 0.5782465848175687,
"grad_norm": 0.8509007096290588,
"learning_rate": 4.450109249817917e-05,
"loss": 1.8162,
"step": 836
},
{
"epoch": 0.5789382673352931,
"grad_norm": 1.4922651052474976,
"learning_rate": 4.442825928623453e-05,
"loss": 1.5922,
"step": 837
},
{
"epoch": 0.5796299498530174,
"grad_norm": 1.0057311058044434,
"learning_rate": 4.435542607428988e-05,
"loss": 1.6555,
"step": 838
},
{
"epoch": 0.5803216323707419,
"grad_norm": 0.7542306184768677,
"learning_rate": 4.428259286234523e-05,
"loss": 1.7829,
"step": 839
},
{
"epoch": 0.5810133148884662,
"grad_norm": 0.7741349935531616,
"learning_rate": 4.4209759650400584e-05,
"loss": 1.6429,
"step": 840
},
{
"epoch": 0.5817049974061905,
"grad_norm": 0.6510467529296875,
"learning_rate": 4.413692643845594e-05,
"loss": 1.6901,
"step": 841
},
{
"epoch": 0.582396679923915,
"grad_norm": 0.5973620414733887,
"learning_rate": 4.406409322651129e-05,
"loss": 1.8884,
"step": 842
},
{
"epoch": 0.5830883624416393,
"grad_norm": 0.7993059158325195,
"learning_rate": 4.3991260014566647e-05,
"loss": 1.9353,
"step": 843
},
{
"epoch": 0.5837800449593636,
"grad_norm": 0.9664002656936646,
"learning_rate": 4.3918426802622e-05,
"loss": 1.5514,
"step": 844
},
{
"epoch": 0.584471727477088,
"grad_norm": 1.7549793720245361,
"learning_rate": 4.3845593590677355e-05,
"loss": 1.3181,
"step": 845
},
{
"epoch": 0.5851634099948124,
"grad_norm": 0.681312084197998,
"learning_rate": 4.37727603787327e-05,
"loss": 1.3649,
"step": 846
},
{
"epoch": 0.5858550925125368,
"grad_norm": 1.23701012134552,
"learning_rate": 4.369992716678806e-05,
"loss": 1.7591,
"step": 847
},
{
"epoch": 0.5865467750302611,
"grad_norm": 0.6490361094474792,
"learning_rate": 4.362709395484341e-05,
"loss": 0.9716,
"step": 848
},
{
"epoch": 0.5872384575479854,
"grad_norm": 1.2864362001419067,
"learning_rate": 4.355426074289876e-05,
"loss": 1.2925,
"step": 849
},
{
"epoch": 0.5879301400657099,
"grad_norm": 0.8159788250923157,
"learning_rate": 4.348142753095411e-05,
"loss": 1.6182,
"step": 850
},
{
"epoch": 0.5886218225834342,
"grad_norm": 0.9866076111793518,
"learning_rate": 4.3408594319009474e-05,
"loss": 1.0921,
"step": 851
},
{
"epoch": 0.5893135051011585,
"grad_norm": 0.7778481841087341,
"learning_rate": 4.333576110706483e-05,
"loss": 1.8508,
"step": 852
},
{
"epoch": 0.590005187618883,
"grad_norm": 0.7202186584472656,
"learning_rate": 4.3262927895120176e-05,
"loss": 2.0207,
"step": 853
},
{
"epoch": 0.5906968701366073,
"grad_norm": 0.764106035232544,
"learning_rate": 4.319009468317553e-05,
"loss": 1.6755,
"step": 854
},
{
"epoch": 0.5913885526543317,
"grad_norm": 0.993752121925354,
"learning_rate": 4.3117261471230884e-05,
"loss": 2.061,
"step": 855
},
{
"epoch": 0.592080235172056,
"grad_norm": 0.9889275431632996,
"learning_rate": 4.304442825928623e-05,
"loss": 2.0183,
"step": 856
},
{
"epoch": 0.5927719176897804,
"grad_norm": 0.7371869683265686,
"learning_rate": 4.2971595047341586e-05,
"loss": 1.8013,
"step": 857
},
{
"epoch": 0.5934636002075048,
"grad_norm": 0.8173306584358215,
"learning_rate": 4.289876183539695e-05,
"loss": 1.8409,
"step": 858
},
{
"epoch": 0.5941552827252291,
"grad_norm": 0.8699966669082642,
"learning_rate": 4.2825928623452295e-05,
"loss": 1.1185,
"step": 859
},
{
"epoch": 0.5948469652429534,
"grad_norm": 0.8679646253585815,
"learning_rate": 4.275309541150765e-05,
"loss": 1.7849,
"step": 860
},
{
"epoch": 0.5955386477606779,
"grad_norm": 0.7400544881820679,
"learning_rate": 4.2680262199563003e-05,
"loss": 1.4845,
"step": 861
},
{
"epoch": 0.5962303302784022,
"grad_norm": 0.9016357064247131,
"learning_rate": 4.260742898761836e-05,
"loss": 1.6556,
"step": 862
},
{
"epoch": 0.5969220127961266,
"grad_norm": 1.0750559568405151,
"learning_rate": 4.2534595775673705e-05,
"loss": 1.2046,
"step": 863
},
{
"epoch": 0.5976136953138509,
"grad_norm": 0.9797933101654053,
"learning_rate": 4.246176256372906e-05,
"loss": 1.9871,
"step": 864
},
{
"epoch": 0.5983053778315753,
"grad_norm": 0.9287171959877014,
"learning_rate": 4.238892935178442e-05,
"loss": 2.0108,
"step": 865
},
{
"epoch": 0.5989970603492997,
"grad_norm": 1.2748676538467407,
"learning_rate": 4.231609613983977e-05,
"loss": 1.8482,
"step": 866
},
{
"epoch": 0.599688742867024,
"grad_norm": 0.7953082919120789,
"learning_rate": 4.224326292789512e-05,
"loss": 1.9459,
"step": 867
},
{
"epoch": 0.6003804253847485,
"grad_norm": 0.6961773037910461,
"learning_rate": 4.217042971595048e-05,
"loss": 1.4124,
"step": 868
},
{
"epoch": 0.6010721079024728,
"grad_norm": 0.7349817156791687,
"learning_rate": 4.2097596504005824e-05,
"loss": 1.3384,
"step": 869
},
{
"epoch": 0.6017637904201971,
"grad_norm": 0.8988999724388123,
"learning_rate": 4.202476329206118e-05,
"loss": 2.1146,
"step": 870
},
{
"epoch": 0.6024554729379215,
"grad_norm": 0.7482579946517944,
"learning_rate": 4.195193008011653e-05,
"loss": 1.9195,
"step": 871
},
{
"epoch": 0.6031471554556459,
"grad_norm": 0.584476888179779,
"learning_rate": 4.1879096868171894e-05,
"loss": 1.1254,
"step": 872
},
{
"epoch": 0.6038388379733702,
"grad_norm": 0.9063954949378967,
"learning_rate": 4.180626365622724e-05,
"loss": 1.655,
"step": 873
},
{
"epoch": 0.6045305204910946,
"grad_norm": 1.784583568572998,
"learning_rate": 4.1733430444282596e-05,
"loss": 1.6756,
"step": 874
},
{
"epoch": 0.6052222030088189,
"grad_norm": 1.1327074766159058,
"learning_rate": 4.166059723233795e-05,
"loss": 1.6496,
"step": 875
},
{
"epoch": 0.6059138855265433,
"grad_norm": 0.6708528399467468,
"learning_rate": 4.15877640203933e-05,
"loss": 1.4513,
"step": 876
},
{
"epoch": 0.6066055680442677,
"grad_norm": 0.8013389110565186,
"learning_rate": 4.151493080844865e-05,
"loss": 1.2203,
"step": 877
},
{
"epoch": 0.607297250561992,
"grad_norm": 0.9567274451255798,
"learning_rate": 4.1442097596504006e-05,
"loss": 1.6625,
"step": 878
},
{
"epoch": 0.6079889330797165,
"grad_norm": 0.7005029916763306,
"learning_rate": 4.136926438455936e-05,
"loss": 1.7659,
"step": 879
},
{
"epoch": 0.6086806155974408,
"grad_norm": 0.7626811861991882,
"learning_rate": 4.1296431172614715e-05,
"loss": 1.9302,
"step": 880
},
{
"epoch": 0.6093722981151651,
"grad_norm": 0.8964532017707825,
"learning_rate": 4.122359796067007e-05,
"loss": 1.5873,
"step": 881
},
{
"epoch": 0.6100639806328895,
"grad_norm": 1.0764849185943604,
"learning_rate": 4.115076474872542e-05,
"loss": 1.7932,
"step": 882
},
{
"epoch": 0.6107556631506139,
"grad_norm": 0.6796227097511292,
"learning_rate": 4.107793153678077e-05,
"loss": 1.4911,
"step": 883
},
{
"epoch": 0.6114473456683382,
"grad_norm": 0.8213468790054321,
"learning_rate": 4.1005098324836125e-05,
"loss": 1.101,
"step": 884
},
{
"epoch": 0.6121390281860626,
"grad_norm": 0.7105699777603149,
"learning_rate": 4.093226511289148e-05,
"loss": 1.583,
"step": 885
},
{
"epoch": 0.6128307107037869,
"grad_norm": 0.8570541143417358,
"learning_rate": 4.0859431900946834e-05,
"loss": 1.4963,
"step": 886
},
{
"epoch": 0.6135223932215114,
"grad_norm": 0.9847023487091064,
"learning_rate": 4.078659868900219e-05,
"loss": 1.7706,
"step": 887
},
{
"epoch": 0.6142140757392357,
"grad_norm": 0.7415423393249512,
"learning_rate": 4.071376547705754e-05,
"loss": 1.7496,
"step": 888
},
{
"epoch": 0.61490575825696,
"grad_norm": 0.8785969614982605,
"learning_rate": 4.064093226511289e-05,
"loss": 1.3073,
"step": 889
},
{
"epoch": 0.6155974407746844,
"grad_norm": 0.8583815097808838,
"learning_rate": 4.0568099053168244e-05,
"loss": 1.5086,
"step": 890
},
{
"epoch": 0.6162891232924088,
"grad_norm": 4.068897724151611,
"learning_rate": 4.04952658412236e-05,
"loss": 1.9179,
"step": 891
},
{
"epoch": 0.6169808058101331,
"grad_norm": 0.781101644039154,
"learning_rate": 4.042243262927895e-05,
"loss": 1.7548,
"step": 892
},
{
"epoch": 0.6176724883278575,
"grad_norm": 0.7089797854423523,
"learning_rate": 4.034959941733431e-05,
"loss": 0.8754,
"step": 893
},
{
"epoch": 0.6183641708455819,
"grad_norm": 0.8193789124488831,
"learning_rate": 4.027676620538966e-05,
"loss": 1.8492,
"step": 894
},
{
"epoch": 0.6190558533633063,
"grad_norm": 1.1896588802337646,
"learning_rate": 4.0203932993445015e-05,
"loss": 1.3831,
"step": 895
},
{
"epoch": 0.6197475358810306,
"grad_norm": 0.7735778093338013,
"learning_rate": 4.013109978150036e-05,
"loss": 1.3693,
"step": 896
},
{
"epoch": 0.6204392183987549,
"grad_norm": 0.8395145535469055,
"learning_rate": 4.005826656955572e-05,
"loss": 2.0921,
"step": 897
},
{
"epoch": 0.6211309009164794,
"grad_norm": 0.8883162140846252,
"learning_rate": 3.998543335761107e-05,
"loss": 1.1215,
"step": 898
},
{
"epoch": 0.6218225834342037,
"grad_norm": 0.7758648991584778,
"learning_rate": 3.9912600145666426e-05,
"loss": 2.1886,
"step": 899
},
{
"epoch": 0.622514265951928,
"grad_norm": 1.0782335996627808,
"learning_rate": 3.983976693372178e-05,
"loss": 1.6127,
"step": 900
},
{
"epoch": 0.6232059484696524,
"grad_norm": 0.7295915484428406,
"learning_rate": 3.9766933721777134e-05,
"loss": 1.4735,
"step": 901
},
{
"epoch": 0.6238976309873768,
"grad_norm": 0.8186253309249878,
"learning_rate": 3.969410050983249e-05,
"loss": 1.9706,
"step": 902
},
{
"epoch": 0.6245893135051012,
"grad_norm": 0.9425630569458008,
"learning_rate": 3.9621267297887836e-05,
"loss": 1.7289,
"step": 903
},
{
"epoch": 0.6252809960228255,
"grad_norm": 1.7592432498931885,
"learning_rate": 3.954843408594319e-05,
"loss": 1.7906,
"step": 904
},
{
"epoch": 0.6259726785405499,
"grad_norm": 0.8689557909965515,
"learning_rate": 3.9475600873998545e-05,
"loss": 1.7125,
"step": 905
},
{
"epoch": 0.6266643610582743,
"grad_norm": 0.7336760759353638,
"learning_rate": 3.94027676620539e-05,
"loss": 1.2892,
"step": 906
},
{
"epoch": 0.6273560435759986,
"grad_norm": 1.0234981775283813,
"learning_rate": 3.932993445010925e-05,
"loss": 1.4871,
"step": 907
},
{
"epoch": 0.628047726093723,
"grad_norm": 0.62737637758255,
"learning_rate": 3.925710123816461e-05,
"loss": 1.4537,
"step": 908
},
{
"epoch": 0.6287394086114474,
"grad_norm": 0.8133891224861145,
"learning_rate": 3.9184268026219955e-05,
"loss": 1.9147,
"step": 909
},
{
"epoch": 0.6294310911291717,
"grad_norm": 0.8863133192062378,
"learning_rate": 3.911143481427531e-05,
"loss": 1.4371,
"step": 910
},
{
"epoch": 0.6301227736468961,
"grad_norm": 0.8457992076873779,
"learning_rate": 3.9038601602330664e-05,
"loss": 1.4767,
"step": 911
},
{
"epoch": 0.6308144561646204,
"grad_norm": 0.9177572727203369,
"learning_rate": 3.896576839038602e-05,
"loss": 0.9703,
"step": 912
},
{
"epoch": 0.6315061386823448,
"grad_norm": 0.798324465751648,
"learning_rate": 3.889293517844137e-05,
"loss": 1.21,
"step": 913
},
{
"epoch": 0.6321978212000692,
"grad_norm": 0.8874804973602295,
"learning_rate": 3.8820101966496726e-05,
"loss": 1.5766,
"step": 914
},
{
"epoch": 0.6328895037177935,
"grad_norm": 0.7579106092453003,
"learning_rate": 3.874726875455208e-05,
"loss": 2.0235,
"step": 915
},
{
"epoch": 0.6335811862355178,
"grad_norm": 0.8905123472213745,
"learning_rate": 3.867443554260743e-05,
"loss": 1.4235,
"step": 916
},
{
"epoch": 0.6342728687532423,
"grad_norm": 0.7457150816917419,
"learning_rate": 3.860160233066278e-05,
"loss": 1.005,
"step": 917
},
{
"epoch": 0.6349645512709666,
"grad_norm": 0.7743381261825562,
"learning_rate": 3.852876911871814e-05,
"loss": 1.5984,
"step": 918
},
{
"epoch": 0.635656233788691,
"grad_norm": 0.9183463454246521,
"learning_rate": 3.845593590677349e-05,
"loss": 1.9001,
"step": 919
},
{
"epoch": 0.6363479163064154,
"grad_norm": 0.9490715861320496,
"learning_rate": 3.8383102694828845e-05,
"loss": 1.6814,
"step": 920
},
{
"epoch": 0.6370395988241397,
"grad_norm": 0.7596708536148071,
"learning_rate": 3.83102694828842e-05,
"loss": 1.3015,
"step": 921
},
{
"epoch": 0.6377312813418641,
"grad_norm": 0.8084985017776489,
"learning_rate": 3.8237436270939554e-05,
"loss": 1.4776,
"step": 922
},
{
"epoch": 0.6384229638595884,
"grad_norm": 1.5200563669204712,
"learning_rate": 3.81646030589949e-05,
"loss": 1.5271,
"step": 923
},
{
"epoch": 0.6391146463773129,
"grad_norm": 0.9168257117271423,
"learning_rate": 3.8091769847050256e-05,
"loss": 0.865,
"step": 924
},
{
"epoch": 0.6398063288950372,
"grad_norm": 0.8897766470909119,
"learning_rate": 3.801893663510561e-05,
"loss": 1.7975,
"step": 925
},
{
"epoch": 0.6404980114127615,
"grad_norm": 0.8529428839683533,
"learning_rate": 3.7946103423160964e-05,
"loss": 1.5159,
"step": 926
},
{
"epoch": 0.6411896939304859,
"grad_norm": 0.9676948189735413,
"learning_rate": 3.787327021121632e-05,
"loss": 0.9847,
"step": 927
},
{
"epoch": 0.6418813764482103,
"grad_norm": 0.8767004013061523,
"learning_rate": 3.780043699927167e-05,
"loss": 2.0668,
"step": 928
},
{
"epoch": 0.6425730589659346,
"grad_norm": 0.7345457673072815,
"learning_rate": 3.772760378732702e-05,
"loss": 1.8291,
"step": 929
},
{
"epoch": 0.643264741483659,
"grad_norm": 1.1258862018585205,
"learning_rate": 3.7654770575382375e-05,
"loss": 1.9181,
"step": 930
},
{
"epoch": 0.6439564240013833,
"grad_norm": 1.059696078300476,
"learning_rate": 3.758193736343773e-05,
"loss": 1.1648,
"step": 931
},
{
"epoch": 0.6446481065191078,
"grad_norm": 0.875996470451355,
"learning_rate": 3.750910415149308e-05,
"loss": 1.8461,
"step": 932
},
{
"epoch": 0.6453397890368321,
"grad_norm": 0.7454437017440796,
"learning_rate": 3.743627093954844e-05,
"loss": 1.8346,
"step": 933
},
{
"epoch": 0.6460314715545564,
"grad_norm": 1.0619977712631226,
"learning_rate": 3.736343772760379e-05,
"loss": 1.1181,
"step": 934
},
{
"epoch": 0.6467231540722809,
"grad_norm": 0.7725276350975037,
"learning_rate": 3.7290604515659146e-05,
"loss": 1.3964,
"step": 935
},
{
"epoch": 0.6474148365900052,
"grad_norm": 0.9066617488861084,
"learning_rate": 3.7217771303714494e-05,
"loss": 1.8135,
"step": 936
},
{
"epoch": 0.6481065191077295,
"grad_norm": 0.7097616195678711,
"learning_rate": 3.714493809176985e-05,
"loss": 2.001,
"step": 937
},
{
"epoch": 0.6487982016254539,
"grad_norm": 0.9724728465080261,
"learning_rate": 3.70721048798252e-05,
"loss": 2.0217,
"step": 938
},
{
"epoch": 0.6494898841431783,
"grad_norm": 0.7304125428199768,
"learning_rate": 3.699927166788055e-05,
"loss": 1.5426,
"step": 939
},
{
"epoch": 0.6501815666609027,
"grad_norm": 0.886914074420929,
"learning_rate": 3.692643845593591e-05,
"loss": 1.6769,
"step": 940
},
{
"epoch": 0.650873249178627,
"grad_norm": 1.057144284248352,
"learning_rate": 3.6853605243991265e-05,
"loss": 1.4546,
"step": 941
},
{
"epoch": 0.6515649316963513,
"grad_norm": 0.9073127508163452,
"learning_rate": 3.678077203204662e-05,
"loss": 1.6194,
"step": 942
},
{
"epoch": 0.6522566142140758,
"grad_norm": 0.9237063527107239,
"learning_rate": 3.670793882010197e-05,
"loss": 1.4516,
"step": 943
},
{
"epoch": 0.6529482967318001,
"grad_norm": 0.7701244354248047,
"learning_rate": 3.663510560815732e-05,
"loss": 1.7599,
"step": 944
},
{
"epoch": 0.6536399792495244,
"grad_norm": 0.7972707152366638,
"learning_rate": 3.6562272396212676e-05,
"loss": 1.9227,
"step": 945
},
{
"epoch": 0.6543316617672489,
"grad_norm": 0.6878631711006165,
"learning_rate": 3.648943918426802e-05,
"loss": 1.2225,
"step": 946
},
{
"epoch": 0.6550233442849732,
"grad_norm": 0.6954039335250854,
"learning_rate": 3.6416605972323384e-05,
"loss": 1.3706,
"step": 947
},
{
"epoch": 0.6557150268026976,
"grad_norm": 0.8597190380096436,
"learning_rate": 3.634377276037874e-05,
"loss": 2.0226,
"step": 948
},
{
"epoch": 0.6564067093204219,
"grad_norm": 0.8058347702026367,
"learning_rate": 3.6270939548434086e-05,
"loss": 1.8904,
"step": 949
},
{
"epoch": 0.6570983918381463,
"grad_norm": 0.7838011384010315,
"learning_rate": 3.619810633648944e-05,
"loss": 1.1612,
"step": 950
},
{
"epoch": 0.6577900743558707,
"grad_norm": 0.9218417406082153,
"learning_rate": 3.6125273124544795e-05,
"loss": 1.9737,
"step": 951
},
{
"epoch": 0.658481756873595,
"grad_norm": 0.7816157937049866,
"learning_rate": 3.605243991260015e-05,
"loss": 0.86,
"step": 952
},
{
"epoch": 0.6591734393913193,
"grad_norm": 0.8489840626716614,
"learning_rate": 3.5979606700655496e-05,
"loss": 2.0043,
"step": 953
},
{
"epoch": 0.6598651219090438,
"grad_norm": 1.0207152366638184,
"learning_rate": 3.590677348871086e-05,
"loss": 1.4315,
"step": 954
},
{
"epoch": 0.6605568044267681,
"grad_norm": 0.8000298738479614,
"learning_rate": 3.583394027676621e-05,
"loss": 1.9495,
"step": 955
},
{
"epoch": 0.6612484869444925,
"grad_norm": 0.8676748871803284,
"learning_rate": 3.576110706482156e-05,
"loss": 1.7678,
"step": 956
},
{
"epoch": 0.6619401694622168,
"grad_norm": 0.9304187297821045,
"learning_rate": 3.5688273852876913e-05,
"loss": 2.0562,
"step": 957
},
{
"epoch": 0.6626318519799412,
"grad_norm": 1.0136486291885376,
"learning_rate": 3.561544064093227e-05,
"loss": 1.7028,
"step": 958
},
{
"epoch": 0.6633235344976656,
"grad_norm": 0.735945463180542,
"learning_rate": 3.5542607428987615e-05,
"loss": 1.3553,
"step": 959
},
{
"epoch": 0.6640152170153899,
"grad_norm": 0.7199145555496216,
"learning_rate": 3.546977421704297e-05,
"loss": 1.3213,
"step": 960
},
{
"epoch": 0.6647068995331143,
"grad_norm": 0.8235880732536316,
"learning_rate": 3.539694100509833e-05,
"loss": 2.0212,
"step": 961
},
{
"epoch": 0.6653985820508387,
"grad_norm": 1.330508828163147,
"learning_rate": 3.5324107793153685e-05,
"loss": 1.8431,
"step": 962
},
{
"epoch": 0.666090264568563,
"grad_norm": 0.7981826663017273,
"learning_rate": 3.525127458120903e-05,
"loss": 1.1926,
"step": 963
},
{
"epoch": 0.6667819470862874,
"grad_norm": 0.9556028842926025,
"learning_rate": 3.517844136926439e-05,
"loss": 1.1749,
"step": 964
},
{
"epoch": 0.6674736296040118,
"grad_norm": 0.9828068614006042,
"learning_rate": 3.510560815731974e-05,
"loss": 1.4142,
"step": 965
},
{
"epoch": 0.6681653121217361,
"grad_norm": 1.730177879333496,
"learning_rate": 3.503277494537509e-05,
"loss": 1.8473,
"step": 966
},
{
"epoch": 0.6688569946394605,
"grad_norm": 0.7303544878959656,
"learning_rate": 3.495994173343044e-05,
"loss": 1.3033,
"step": 967
},
{
"epoch": 0.6695486771571848,
"grad_norm": 0.7854034900665283,
"learning_rate": 3.4887108521485804e-05,
"loss": 1.2283,
"step": 968
},
{
"epoch": 0.6702403596749092,
"grad_norm": 0.8185420036315918,
"learning_rate": 3.481427530954115e-05,
"loss": 2.0494,
"step": 969
},
{
"epoch": 0.6709320421926336,
"grad_norm": 0.7962689995765686,
"learning_rate": 3.4741442097596506e-05,
"loss": 1.9523,
"step": 970
},
{
"epoch": 0.6716237247103579,
"grad_norm": 0.9332857131958008,
"learning_rate": 3.466860888565186e-05,
"loss": 1.6991,
"step": 971
},
{
"epoch": 0.6723154072280824,
"grad_norm": 0.7964072823524475,
"learning_rate": 3.4595775673707214e-05,
"loss": 1.6715,
"step": 972
},
{
"epoch": 0.6730070897458067,
"grad_norm": 0.8391403555870056,
"learning_rate": 3.452294246176256e-05,
"loss": 1.8878,
"step": 973
},
{
"epoch": 0.673698772263531,
"grad_norm": 0.8566287159919739,
"learning_rate": 3.4450109249817916e-05,
"loss": 1.7805,
"step": 974
},
{
"epoch": 0.6743904547812554,
"grad_norm": 0.9097205400466919,
"learning_rate": 3.437727603787328e-05,
"loss": 1.4648,
"step": 975
},
{
"epoch": 0.6750821372989798,
"grad_norm": 0.9566125273704529,
"learning_rate": 3.4304442825928625e-05,
"loss": 1.4794,
"step": 976
},
{
"epoch": 0.6757738198167041,
"grad_norm": 0.89398193359375,
"learning_rate": 3.423160961398398e-05,
"loss": 2.0932,
"step": 977
},
{
"epoch": 0.6764655023344285,
"grad_norm": 0.8663723468780518,
"learning_rate": 3.415877640203933e-05,
"loss": 1.8625,
"step": 978
},
{
"epoch": 0.6771571848521528,
"grad_norm": 0.751266598701477,
"learning_rate": 3.408594319009468e-05,
"loss": 1.0625,
"step": 979
},
{
"epoch": 0.6778488673698773,
"grad_norm": 0.8049336671829224,
"learning_rate": 3.4013109978150035e-05,
"loss": 1.3388,
"step": 980
},
{
"epoch": 0.6785405498876016,
"grad_norm": 0.8025321364402771,
"learning_rate": 3.394027676620539e-05,
"loss": 1.4493,
"step": 981
},
{
"epoch": 0.6792322324053259,
"grad_norm": 0.8283219933509827,
"learning_rate": 3.386744355426075e-05,
"loss": 1.1765,
"step": 982
},
{
"epoch": 0.6799239149230503,
"grad_norm": 1.0506592988967896,
"learning_rate": 3.37946103423161e-05,
"loss": 1.2619,
"step": 983
},
{
"epoch": 0.6806155974407747,
"grad_norm": 0.5937513709068298,
"learning_rate": 3.372177713037145e-05,
"loss": 0.9384,
"step": 984
},
{
"epoch": 0.681307279958499,
"grad_norm": 0.6834938526153564,
"learning_rate": 3.3648943918426806e-05,
"loss": 1.0192,
"step": 985
},
{
"epoch": 0.6819989624762234,
"grad_norm": 1.1329364776611328,
"learning_rate": 3.3576110706482154e-05,
"loss": 2.2567,
"step": 986
},
{
"epoch": 0.6826906449939478,
"grad_norm": 0.7666782736778259,
"learning_rate": 3.350327749453751e-05,
"loss": 1.8783,
"step": 987
},
{
"epoch": 0.6833823275116722,
"grad_norm": 1.2282352447509766,
"learning_rate": 3.343044428259286e-05,
"loss": 1.5103,
"step": 988
},
{
"epoch": 0.6840740100293965,
"grad_norm": 0.7626004219055176,
"learning_rate": 3.335761107064822e-05,
"loss": 1.3147,
"step": 989
},
{
"epoch": 0.6847656925471208,
"grad_norm": 1.0253794193267822,
"learning_rate": 3.328477785870357e-05,
"loss": 1.7115,
"step": 990
},
{
"epoch": 0.6854573750648453,
"grad_norm": 0.890225887298584,
"learning_rate": 3.3211944646758925e-05,
"loss": 1.6278,
"step": 991
},
{
"epoch": 0.6861490575825696,
"grad_norm": 1.0109024047851562,
"learning_rate": 3.313911143481428e-05,
"loss": 1.5538,
"step": 992
},
{
"epoch": 0.686840740100294,
"grad_norm": 0.8567763566970825,
"learning_rate": 3.306627822286963e-05,
"loss": 1.7241,
"step": 993
},
{
"epoch": 0.6875324226180183,
"grad_norm": 0.7532871961593628,
"learning_rate": 3.299344501092498e-05,
"loss": 1.3609,
"step": 994
},
{
"epoch": 0.6882241051357427,
"grad_norm": 0.9524408578872681,
"learning_rate": 3.2920611798980336e-05,
"loss": 1.9429,
"step": 995
},
{
"epoch": 0.6889157876534671,
"grad_norm": 0.6177567839622498,
"learning_rate": 3.284777858703569e-05,
"loss": 0.9766,
"step": 996
},
{
"epoch": 0.6896074701711914,
"grad_norm": 1.1738656759262085,
"learning_rate": 3.2774945375091044e-05,
"loss": 1.6276,
"step": 997
},
{
"epoch": 0.6902991526889158,
"grad_norm": 1.018707036972046,
"learning_rate": 3.27021121631464e-05,
"loss": 0.8201,
"step": 998
},
{
"epoch": 0.6909908352066402,
"grad_norm": 0.8498096466064453,
"learning_rate": 3.2629278951201746e-05,
"loss": 0.9932,
"step": 999
},
{
"epoch": 0.6916825177243645,
"grad_norm": 0.8145414590835571,
"learning_rate": 3.25564457392571e-05,
"loss": 1.9895,
"step": 1000
},
{
"epoch": 0.6916825177243645,
"eval_loss": 1.4734511375427246,
"eval_runtime": 636.6322,
"eval_samples_per_second": 2.018,
"eval_steps_per_second": 1.01,
"step": 1000
}
],
"logging_steps": 1,
"max_steps": 1446,
"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": false
},
"attributes": {}
}
},
"total_flos": 3.178081149272064e+16,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}