Spark2Scale / checkpoint-500 /trainer_state.json
Dohahemdann's picture
Checkpoint at step 500
801c16f verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.34584125886218225,
"eval_steps": 500,
"global_step": 500,
"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
}
],
"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": 1.596867200925696e+16,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}