Bespoke-Stratos-17k-v4 / trainer_state.json
sedrickkeh's picture
End of training
8bbabbc verified
raw
history blame
91.2 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.998563906175203,
"eval_steps": 500,
"global_step": 522,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0057443752991862135,
"grad_norm": 5.864684581756592,
"learning_rate": 1.886792452830189e-07,
"loss": 0.8955,
"step": 1
},
{
"epoch": 0.011488750598372427,
"grad_norm": 6.245960235595703,
"learning_rate": 3.773584905660378e-07,
"loss": 0.9318,
"step": 2
},
{
"epoch": 0.01723312589755864,
"grad_norm": 6.222021579742432,
"learning_rate": 5.660377358490567e-07,
"loss": 0.9555,
"step": 3
},
{
"epoch": 0.022977501196744854,
"grad_norm": 6.026363372802734,
"learning_rate": 7.547169811320755e-07,
"loss": 0.9198,
"step": 4
},
{
"epoch": 0.028721876495931067,
"grad_norm": 5.970179080963135,
"learning_rate": 9.433962264150944e-07,
"loss": 0.9159,
"step": 5
},
{
"epoch": 0.03446625179511728,
"grad_norm": 6.270013809204102,
"learning_rate": 1.1320754716981133e-06,
"loss": 0.9075,
"step": 6
},
{
"epoch": 0.040210627094303494,
"grad_norm": 5.812321186065674,
"learning_rate": 1.3207547169811322e-06,
"loss": 0.914,
"step": 7
},
{
"epoch": 0.04595500239348971,
"grad_norm": 4.669607639312744,
"learning_rate": 1.509433962264151e-06,
"loss": 0.8608,
"step": 8
},
{
"epoch": 0.05169937769267592,
"grad_norm": 4.368519306182861,
"learning_rate": 1.6981132075471698e-06,
"loss": 0.8335,
"step": 9
},
{
"epoch": 0.057443752991862135,
"grad_norm": 4.279673099517822,
"learning_rate": 1.8867924528301889e-06,
"loss": 0.8821,
"step": 10
},
{
"epoch": 0.06318812829104835,
"grad_norm": 2.5520718097686768,
"learning_rate": 2.075471698113208e-06,
"loss": 0.8347,
"step": 11
},
{
"epoch": 0.06893250359023456,
"grad_norm": 2.239227771759033,
"learning_rate": 2.2641509433962266e-06,
"loss": 0.76,
"step": 12
},
{
"epoch": 0.07467687888942078,
"grad_norm": 2.1608989238739014,
"learning_rate": 2.4528301886792453e-06,
"loss": 0.8344,
"step": 13
},
{
"epoch": 0.08042125418860699,
"grad_norm": 2.218498468399048,
"learning_rate": 2.6415094339622644e-06,
"loss": 0.8196,
"step": 14
},
{
"epoch": 0.0861656294877932,
"grad_norm": 3.5752766132354736,
"learning_rate": 2.830188679245283e-06,
"loss": 0.813,
"step": 15
},
{
"epoch": 0.09191000478697942,
"grad_norm": 3.9789814949035645,
"learning_rate": 3.018867924528302e-06,
"loss": 0.8194,
"step": 16
},
{
"epoch": 0.09765438008616563,
"grad_norm": 3.9097297191619873,
"learning_rate": 3.207547169811321e-06,
"loss": 0.7943,
"step": 17
},
{
"epoch": 0.10339875538535184,
"grad_norm": 3.6479055881500244,
"learning_rate": 3.3962264150943395e-06,
"loss": 0.7953,
"step": 18
},
{
"epoch": 0.10914313068453806,
"grad_norm": 3.101821184158325,
"learning_rate": 3.5849056603773586e-06,
"loss": 0.7265,
"step": 19
},
{
"epoch": 0.11488750598372427,
"grad_norm": 2.2246768474578857,
"learning_rate": 3.7735849056603777e-06,
"loss": 0.7448,
"step": 20
},
{
"epoch": 0.12063188128291048,
"grad_norm": 1.6926085948944092,
"learning_rate": 3.962264150943396e-06,
"loss": 0.716,
"step": 21
},
{
"epoch": 0.1263762565820967,
"grad_norm": 1.5090553760528564,
"learning_rate": 4.150943396226416e-06,
"loss": 0.7089,
"step": 22
},
{
"epoch": 0.13212063188128292,
"grad_norm": 1.2893575429916382,
"learning_rate": 4.339622641509435e-06,
"loss": 0.666,
"step": 23
},
{
"epoch": 0.13786500718046912,
"grad_norm": 1.1636651754379272,
"learning_rate": 4.528301886792453e-06,
"loss": 0.7036,
"step": 24
},
{
"epoch": 0.14360938247965535,
"grad_norm": 1.3487656116485596,
"learning_rate": 4.716981132075472e-06,
"loss": 0.6877,
"step": 25
},
{
"epoch": 0.14935375777884155,
"grad_norm": 1.198224425315857,
"learning_rate": 4.905660377358491e-06,
"loss": 0.677,
"step": 26
},
{
"epoch": 0.15509813307802778,
"grad_norm": 1.1632698774337769,
"learning_rate": 5.09433962264151e-06,
"loss": 0.7112,
"step": 27
},
{
"epoch": 0.16084250837721398,
"grad_norm": 0.9113264083862305,
"learning_rate": 5.283018867924529e-06,
"loss": 0.6699,
"step": 28
},
{
"epoch": 0.1665868836764002,
"grad_norm": 0.9448997378349304,
"learning_rate": 5.4716981132075475e-06,
"loss": 0.654,
"step": 29
},
{
"epoch": 0.1723312589755864,
"grad_norm": 1.0080927610397339,
"learning_rate": 5.660377358490566e-06,
"loss": 0.6587,
"step": 30
},
{
"epoch": 0.17807563427477263,
"grad_norm": 0.9000163674354553,
"learning_rate": 5.849056603773585e-06,
"loss": 0.6159,
"step": 31
},
{
"epoch": 0.18382000957395883,
"grad_norm": 0.8230825066566467,
"learning_rate": 6.037735849056604e-06,
"loss": 0.5924,
"step": 32
},
{
"epoch": 0.18956438487314506,
"grad_norm": 0.8905507326126099,
"learning_rate": 6.226415094339623e-06,
"loss": 0.6357,
"step": 33
},
{
"epoch": 0.19530876017233126,
"grad_norm": 0.9403512477874756,
"learning_rate": 6.415094339622642e-06,
"loss": 0.6548,
"step": 34
},
{
"epoch": 0.20105313547151749,
"grad_norm": 0.8337975144386292,
"learning_rate": 6.60377358490566e-06,
"loss": 0.639,
"step": 35
},
{
"epoch": 0.20679751077070369,
"grad_norm": 0.7238542437553406,
"learning_rate": 6.792452830188679e-06,
"loss": 0.6334,
"step": 36
},
{
"epoch": 0.2125418860698899,
"grad_norm": 0.7238707542419434,
"learning_rate": 6.981132075471699e-06,
"loss": 0.6154,
"step": 37
},
{
"epoch": 0.2182862613690761,
"grad_norm": 0.867655336856842,
"learning_rate": 7.169811320754717e-06,
"loss": 0.629,
"step": 38
},
{
"epoch": 0.22403063666826234,
"grad_norm": 0.7441483736038208,
"learning_rate": 7.358490566037736e-06,
"loss": 0.6283,
"step": 39
},
{
"epoch": 0.22977501196744854,
"grad_norm": 0.694157600402832,
"learning_rate": 7.5471698113207555e-06,
"loss": 0.6054,
"step": 40
},
{
"epoch": 0.23551938726663477,
"grad_norm": 0.7625733017921448,
"learning_rate": 7.735849056603775e-06,
"loss": 0.6478,
"step": 41
},
{
"epoch": 0.24126376256582097,
"grad_norm": 0.6908527612686157,
"learning_rate": 7.924528301886793e-06,
"loss": 0.5829,
"step": 42
},
{
"epoch": 0.2470081378650072,
"grad_norm": 0.6998976469039917,
"learning_rate": 8.113207547169812e-06,
"loss": 0.6201,
"step": 43
},
{
"epoch": 0.2527525131641934,
"grad_norm": 0.713438093662262,
"learning_rate": 8.301886792452832e-06,
"loss": 0.6351,
"step": 44
},
{
"epoch": 0.2584968884633796,
"grad_norm": 0.647293210029602,
"learning_rate": 8.49056603773585e-06,
"loss": 0.6185,
"step": 45
},
{
"epoch": 0.26424126376256585,
"grad_norm": 0.6884685158729553,
"learning_rate": 8.67924528301887e-06,
"loss": 0.5691,
"step": 46
},
{
"epoch": 0.26998563906175205,
"grad_norm": 0.7607399225234985,
"learning_rate": 8.867924528301887e-06,
"loss": 0.5987,
"step": 47
},
{
"epoch": 0.27573001436093825,
"grad_norm": 0.6738750338554382,
"learning_rate": 9.056603773584907e-06,
"loss": 0.5851,
"step": 48
},
{
"epoch": 0.28147438966012445,
"grad_norm": 0.7652818560600281,
"learning_rate": 9.245283018867926e-06,
"loss": 0.6191,
"step": 49
},
{
"epoch": 0.2872187649593107,
"grad_norm": 0.6622350811958313,
"learning_rate": 9.433962264150944e-06,
"loss": 0.5746,
"step": 50
},
{
"epoch": 0.2929631402584969,
"grad_norm": 0.6434982419013977,
"learning_rate": 9.622641509433963e-06,
"loss": 0.5605,
"step": 51
},
{
"epoch": 0.2987075155576831,
"grad_norm": 0.6086822748184204,
"learning_rate": 9.811320754716981e-06,
"loss": 0.6057,
"step": 52
},
{
"epoch": 0.3044518908568693,
"grad_norm": 0.6435369849205017,
"learning_rate": 1e-05,
"loss": 0.5778,
"step": 53
},
{
"epoch": 0.31019626615605556,
"grad_norm": 0.608729362487793,
"learning_rate": 9.999887825938495e-06,
"loss": 0.5676,
"step": 54
},
{
"epoch": 0.31594064145524176,
"grad_norm": 0.6421694755554199,
"learning_rate": 9.999551308787183e-06,
"loss": 0.5721,
"step": 55
},
{
"epoch": 0.32168501675442795,
"grad_norm": 0.6002755165100098,
"learning_rate": 9.998990463645464e-06,
"loss": 0.5793,
"step": 56
},
{
"epoch": 0.32742939205361415,
"grad_norm": 0.6116312742233276,
"learning_rate": 9.998205315678248e-06,
"loss": 0.5877,
"step": 57
},
{
"epoch": 0.3331737673528004,
"grad_norm": 0.6984326243400574,
"learning_rate": 9.997195900114833e-06,
"loss": 0.5807,
"step": 58
},
{
"epoch": 0.3389181426519866,
"grad_norm": 0.6070480942726135,
"learning_rate": 9.995962262247314e-06,
"loss": 0.5549,
"step": 59
},
{
"epoch": 0.3446625179511728,
"grad_norm": 0.6007083654403687,
"learning_rate": 9.994504457428557e-06,
"loss": 0.5635,
"step": 60
},
{
"epoch": 0.350406893250359,
"grad_norm": 0.5620129108428955,
"learning_rate": 9.99282255106972e-06,
"loss": 0.5793,
"step": 61
},
{
"epoch": 0.35615126854954526,
"grad_norm": 0.5313761830329895,
"learning_rate": 9.99091661863731e-06,
"loss": 0.5887,
"step": 62
},
{
"epoch": 0.36189564384873146,
"grad_norm": 0.5500160455703735,
"learning_rate": 9.988786745649798e-06,
"loss": 0.5727,
"step": 63
},
{
"epoch": 0.36764001914791766,
"grad_norm": 0.576194703578949,
"learning_rate": 9.986433027673786e-06,
"loss": 0.5562,
"step": 64
},
{
"epoch": 0.37338439444710386,
"grad_norm": 0.5685051679611206,
"learning_rate": 9.983855570319716e-06,
"loss": 0.5777,
"step": 65
},
{
"epoch": 0.3791287697462901,
"grad_norm": 0.5267539620399475,
"learning_rate": 9.981054489237132e-06,
"loss": 0.5713,
"step": 66
},
{
"epoch": 0.3848731450454763,
"grad_norm": 0.5135684013366699,
"learning_rate": 9.978029910109491e-06,
"loss": 0.532,
"step": 67
},
{
"epoch": 0.3906175203446625,
"grad_norm": 0.5370251536369324,
"learning_rate": 9.974781968648523e-06,
"loss": 0.5549,
"step": 68
},
{
"epoch": 0.3963618956438487,
"grad_norm": 0.5256316065788269,
"learning_rate": 9.971310810588141e-06,
"loss": 0.5817,
"step": 69
},
{
"epoch": 0.40210627094303497,
"grad_norm": 0.47199568152427673,
"learning_rate": 9.967616591677906e-06,
"loss": 0.5054,
"step": 70
},
{
"epoch": 0.40785064624222117,
"grad_norm": 0.64939284324646,
"learning_rate": 9.963699477676031e-06,
"loss": 0.5592,
"step": 71
},
{
"epoch": 0.41359502154140737,
"grad_norm": 0.5359857678413391,
"learning_rate": 9.959559644341954e-06,
"loss": 0.5388,
"step": 72
},
{
"epoch": 0.41933939684059357,
"grad_norm": 0.5025998950004578,
"learning_rate": 9.95519727742844e-06,
"loss": 0.5592,
"step": 73
},
{
"epoch": 0.4250837721397798,
"grad_norm": 0.601884126663208,
"learning_rate": 9.950612572673255e-06,
"loss": 0.5617,
"step": 74
},
{
"epoch": 0.430828147438966,
"grad_norm": 0.5677142143249512,
"learning_rate": 9.945805735790383e-06,
"loss": 0.5727,
"step": 75
},
{
"epoch": 0.4365725227381522,
"grad_norm": 0.4831050634384155,
"learning_rate": 9.940776982460787e-06,
"loss": 0.5487,
"step": 76
},
{
"epoch": 0.4423168980373384,
"grad_norm": 0.5099920630455017,
"learning_rate": 9.935526538322744e-06,
"loss": 0.5755,
"step": 77
},
{
"epoch": 0.4480612733365247,
"grad_norm": 0.5375122427940369,
"learning_rate": 9.930054638961709e-06,
"loss": 0.5522,
"step": 78
},
{
"epoch": 0.4538056486357109,
"grad_norm": 0.5247811675071716,
"learning_rate": 9.924361529899754e-06,
"loss": 0.5616,
"step": 79
},
{
"epoch": 0.4595500239348971,
"grad_norm": 0.5249711871147156,
"learning_rate": 9.918447466584545e-06,
"loss": 0.5434,
"step": 80
},
{
"epoch": 0.4652943992340833,
"grad_norm": 0.523030698299408,
"learning_rate": 9.91231271437788e-06,
"loss": 0.5475,
"step": 81
},
{
"epoch": 0.47103877453326953,
"grad_norm": 0.6253980994224548,
"learning_rate": 9.905957548543794e-06,
"loss": 0.5659,
"step": 82
},
{
"epoch": 0.47678314983245573,
"grad_norm": 0.48295021057128906,
"learning_rate": 9.899382254236186e-06,
"loss": 0.551,
"step": 83
},
{
"epoch": 0.48252752513164193,
"grad_norm": 0.6675193309783936,
"learning_rate": 9.892587126486046e-06,
"loss": 0.5792,
"step": 84
},
{
"epoch": 0.48827190043082813,
"grad_norm": 0.561299204826355,
"learning_rate": 9.885572470188207e-06,
"loss": 0.5362,
"step": 85
},
{
"epoch": 0.4940162757300144,
"grad_norm": 0.5058274269104004,
"learning_rate": 9.878338600087658e-06,
"loss": 0.5541,
"step": 86
},
{
"epoch": 0.4997606510292006,
"grad_norm": 0.6501106023788452,
"learning_rate": 9.87088584076544e-06,
"loss": 0.5708,
"step": 87
},
{
"epoch": 0.5055050263283868,
"grad_norm": 0.48683059215545654,
"learning_rate": 9.863214526624065e-06,
"loss": 0.5545,
"step": 88
},
{
"epoch": 0.511249401627573,
"grad_norm": 0.5247763395309448,
"learning_rate": 9.85532500187252e-06,
"loss": 0.5506,
"step": 89
},
{
"epoch": 0.5169937769267592,
"grad_norm": 0.5401856899261475,
"learning_rate": 9.847217620510815e-06,
"loss": 0.5595,
"step": 90
},
{
"epoch": 0.5227381522259454,
"grad_norm": 0.46576857566833496,
"learning_rate": 9.83889274631411e-06,
"loss": 0.5377,
"step": 91
},
{
"epoch": 0.5284825275251317,
"grad_norm": 0.5679385662078857,
"learning_rate": 9.830350752816386e-06,
"loss": 0.5438,
"step": 92
},
{
"epoch": 0.5342269028243178,
"grad_norm": 0.4763081967830658,
"learning_rate": 9.821592023293686e-06,
"loss": 0.5318,
"step": 93
},
{
"epoch": 0.5399712781235041,
"grad_norm": 0.5016634464263916,
"learning_rate": 9.81261695074691e-06,
"loss": 0.5611,
"step": 94
},
{
"epoch": 0.5457156534226902,
"grad_norm": 0.5415533781051636,
"learning_rate": 9.803425937884202e-06,
"loss": 0.5497,
"step": 95
},
{
"epoch": 0.5514600287218765,
"grad_norm": 0.4844048023223877,
"learning_rate": 9.794019397102852e-06,
"loss": 0.5628,
"step": 96
},
{
"epoch": 0.5572044040210627,
"grad_norm": 0.5687518119812012,
"learning_rate": 9.784397750470818e-06,
"loss": 0.5333,
"step": 97
},
{
"epoch": 0.5629487793202489,
"grad_norm": 0.5257644653320312,
"learning_rate": 9.774561429707769e-06,
"loss": 0.5756,
"step": 98
},
{
"epoch": 0.5686931546194351,
"grad_norm": 0.5053434371948242,
"learning_rate": 9.764510876165727e-06,
"loss": 0.5314,
"step": 99
},
{
"epoch": 0.5744375299186214,
"grad_norm": 0.5556429624557495,
"learning_rate": 9.754246540809257e-06,
"loss": 0.5345,
"step": 100
},
{
"epoch": 0.5801819052178075,
"grad_norm": 0.6641430854797363,
"learning_rate": 9.743768884195233e-06,
"loss": 0.5657,
"step": 101
},
{
"epoch": 0.5859262805169938,
"grad_norm": 0.6076303720474243,
"learning_rate": 9.733078376452172e-06,
"loss": 0.5623,
"step": 102
},
{
"epoch": 0.59167065581618,
"grad_norm": 0.5604638457298279,
"learning_rate": 9.722175497259145e-06,
"loss": 0.5472,
"step": 103
},
{
"epoch": 0.5974150311153662,
"grad_norm": 0.49825626611709595,
"learning_rate": 9.71106073582425e-06,
"loss": 0.5309,
"step": 104
},
{
"epoch": 0.6031594064145525,
"grad_norm": 0.48945000767707825,
"learning_rate": 9.699734590862655e-06,
"loss": 0.5317,
"step": 105
},
{
"epoch": 0.6089037817137386,
"grad_norm": 0.6487892270088196,
"learning_rate": 9.688197570574238e-06,
"loss": 0.5548,
"step": 106
},
{
"epoch": 0.6146481570129249,
"grad_norm": 0.5075920820236206,
"learning_rate": 9.676450192620767e-06,
"loss": 0.5486,
"step": 107
},
{
"epoch": 0.6203925323121111,
"grad_norm": 0.6226174235343933,
"learning_rate": 9.66449298410268e-06,
"loss": 0.5655,
"step": 108
},
{
"epoch": 0.6261369076112973,
"grad_norm": 0.5333593487739563,
"learning_rate": 9.652326481535434e-06,
"loss": 0.5118,
"step": 109
},
{
"epoch": 0.6318812829104835,
"grad_norm": 0.6098229885101318,
"learning_rate": 9.639951230825433e-06,
"loss": 0.5795,
"step": 110
},
{
"epoch": 0.6376256582096697,
"grad_norm": 0.6318515539169312,
"learning_rate": 9.62736778724553e-06,
"loss": 0.5582,
"step": 111
},
{
"epoch": 0.6433700335088559,
"grad_norm": 0.6139549016952515,
"learning_rate": 9.614576715410116e-06,
"loss": 0.5678,
"step": 112
},
{
"epoch": 0.6491144088080422,
"grad_norm": 0.682357907295227,
"learning_rate": 9.60157858924978e-06,
"loss": 0.5327,
"step": 113
},
{
"epoch": 0.6548587841072283,
"grad_norm": 0.6548144221305847,
"learning_rate": 9.588373991985566e-06,
"loss": 0.5571,
"step": 114
},
{
"epoch": 0.6606031594064146,
"grad_norm": 0.6865684986114502,
"learning_rate": 9.574963516102795e-06,
"loss": 0.564,
"step": 115
},
{
"epoch": 0.6663475347056008,
"grad_norm": 0.5575840473175049,
"learning_rate": 9.561347763324484e-06,
"loss": 0.528,
"step": 116
},
{
"epoch": 0.672091910004787,
"grad_norm": 0.5787408947944641,
"learning_rate": 9.547527344584353e-06,
"loss": 0.5466,
"step": 117
},
{
"epoch": 0.6778362853039732,
"grad_norm": 0.6304364204406738,
"learning_rate": 9.533502879999398e-06,
"loss": 0.5687,
"step": 118
},
{
"epoch": 0.6835806606031594,
"grad_norm": 0.5542575716972351,
"learning_rate": 9.519274998842084e-06,
"loss": 0.5721,
"step": 119
},
{
"epoch": 0.6893250359023456,
"grad_norm": 0.6041135191917419,
"learning_rate": 9.504844339512096e-06,
"loss": 0.5336,
"step": 120
},
{
"epoch": 0.6950694112015319,
"grad_norm": 0.48197638988494873,
"learning_rate": 9.490211549507701e-06,
"loss": 0.5557,
"step": 121
},
{
"epoch": 0.700813786500718,
"grad_norm": 0.53097003698349,
"learning_rate": 9.475377285396692e-06,
"loss": 0.5325,
"step": 122
},
{
"epoch": 0.7065581617999043,
"grad_norm": 0.4822128713130951,
"learning_rate": 9.460342212786933e-06,
"loss": 0.5243,
"step": 123
},
{
"epoch": 0.7123025370990905,
"grad_norm": 0.5756934881210327,
"learning_rate": 9.445107006296488e-06,
"loss": 0.5695,
"step": 124
},
{
"epoch": 0.7180469123982767,
"grad_norm": 0.5513118505477905,
"learning_rate": 9.42967234952335e-06,
"loss": 0.5685,
"step": 125
},
{
"epoch": 0.7237912876974629,
"grad_norm": 0.5595805644989014,
"learning_rate": 9.414038935014777e-06,
"loss": 0.5291,
"step": 126
},
{
"epoch": 0.7295356629966491,
"grad_norm": 0.5846650004386902,
"learning_rate": 9.398207464236209e-06,
"loss": 0.5287,
"step": 127
},
{
"epoch": 0.7352800382958353,
"grad_norm": 0.5306510925292969,
"learning_rate": 9.382178647539794e-06,
"loss": 0.541,
"step": 128
},
{
"epoch": 0.7410244135950216,
"grad_norm": 0.5361509323120117,
"learning_rate": 9.365953204132526e-06,
"loss": 0.5498,
"step": 129
},
{
"epoch": 0.7467687888942077,
"grad_norm": 0.46198606491088867,
"learning_rate": 9.349531862043952e-06,
"loss": 0.5631,
"step": 130
},
{
"epoch": 0.752513164193394,
"grad_norm": 0.4705687463283539,
"learning_rate": 9.332915358093532e-06,
"loss": 0.5268,
"step": 131
},
{
"epoch": 0.7582575394925802,
"grad_norm": 0.5280994772911072,
"learning_rate": 9.316104437857561e-06,
"loss": 0.5322,
"step": 132
},
{
"epoch": 0.7640019147917664,
"grad_norm": 0.56460040807724,
"learning_rate": 9.299099855635716e-06,
"loss": 0.5244,
"step": 133
},
{
"epoch": 0.7697462900909526,
"grad_norm": 0.5753394365310669,
"learning_rate": 9.28190237441722e-06,
"loss": 0.5161,
"step": 134
},
{
"epoch": 0.7754906653901388,
"grad_norm": 0.6279768347740173,
"learning_rate": 9.2645127658466e-06,
"loss": 0.536,
"step": 135
},
{
"epoch": 0.781235040689325,
"grad_norm": 0.5690814256668091,
"learning_rate": 9.246931810189061e-06,
"loss": 0.5463,
"step": 136
},
{
"epoch": 0.7869794159885113,
"grad_norm": 0.5215439200401306,
"learning_rate": 9.229160296295488e-06,
"loss": 0.5093,
"step": 137
},
{
"epoch": 0.7927237912876974,
"grad_norm": 0.5325930714607239,
"learning_rate": 9.211199021567034e-06,
"loss": 0.5521,
"step": 138
},
{
"epoch": 0.7984681665868837,
"grad_norm": 0.6381127238273621,
"learning_rate": 9.193048791919357e-06,
"loss": 0.5596,
"step": 139
},
{
"epoch": 0.8042125418860699,
"grad_norm": 0.5356181263923645,
"learning_rate": 9.174710421746445e-06,
"loss": 0.5686,
"step": 140
},
{
"epoch": 0.8099569171852561,
"grad_norm": 0.5347083210945129,
"learning_rate": 9.156184733884084e-06,
"loss": 0.5251,
"step": 141
},
{
"epoch": 0.8157012924844423,
"grad_norm": 0.5627686977386475,
"learning_rate": 9.137472559572935e-06,
"loss": 0.5613,
"step": 142
},
{
"epoch": 0.8214456677836285,
"grad_norm": 0.5725522637367249,
"learning_rate": 9.118574738421236e-06,
"loss": 0.5179,
"step": 143
},
{
"epoch": 0.8271900430828147,
"grad_norm": 0.6335479617118835,
"learning_rate": 9.099492118367123e-06,
"loss": 0.558,
"step": 144
},
{
"epoch": 0.832934418382001,
"grad_norm": 0.5163323283195496,
"learning_rate": 9.080225555640601e-06,
"loss": 0.5401,
"step": 145
},
{
"epoch": 0.8386787936811871,
"grad_norm": 0.5082926750183105,
"learning_rate": 9.0607759147251e-06,
"loss": 0.5406,
"step": 146
},
{
"epoch": 0.8444231689803734,
"grad_norm": 0.612827718257904,
"learning_rate": 9.04114406831871e-06,
"loss": 0.5115,
"step": 147
},
{
"epoch": 0.8501675442795597,
"grad_norm": 0.4936908483505249,
"learning_rate": 9.021330897295011e-06,
"loss": 0.5385,
"step": 148
},
{
"epoch": 0.8559119195787458,
"grad_norm": 0.6068204641342163,
"learning_rate": 9.001337290663548e-06,
"loss": 0.5173,
"step": 149
},
{
"epoch": 0.861656294877932,
"grad_norm": 0.5877631902694702,
"learning_rate": 8.981164145529943e-06,
"loss": 0.5433,
"step": 150
},
{
"epoch": 0.8674006701771182,
"grad_norm": 0.49553361535072327,
"learning_rate": 8.960812367055646e-06,
"loss": 0.558,
"step": 151
},
{
"epoch": 0.8731450454763044,
"grad_norm": 0.5737854242324829,
"learning_rate": 8.940282868417321e-06,
"loss": 0.4894,
"step": 152
},
{
"epoch": 0.8788894207754907,
"grad_norm": 0.5427818894386292,
"learning_rate": 8.91957657076586e-06,
"loss": 0.5531,
"step": 153
},
{
"epoch": 0.8846337960746768,
"grad_norm": 0.5494779944419861,
"learning_rate": 8.898694403185066e-06,
"loss": 0.5155,
"step": 154
},
{
"epoch": 0.8903781713738631,
"grad_norm": 0.5184085369110107,
"learning_rate": 8.877637302649962e-06,
"loss": 0.5261,
"step": 155
},
{
"epoch": 0.8961225466730494,
"grad_norm": 0.4889534115791321,
"learning_rate": 8.856406213984743e-06,
"loss": 0.5443,
"step": 156
},
{
"epoch": 0.9018669219722355,
"grad_norm": 0.48349231481552124,
"learning_rate": 8.835002089820387e-06,
"loss": 0.5279,
"step": 157
},
{
"epoch": 0.9076112972714218,
"grad_norm": 0.546751856803894,
"learning_rate": 8.81342589055191e-06,
"loss": 0.5578,
"step": 158
},
{
"epoch": 0.9133556725706079,
"grad_norm": 0.454558789730072,
"learning_rate": 8.791678584295276e-06,
"loss": 0.5219,
"step": 159
},
{
"epoch": 0.9191000478697942,
"grad_norm": 0.6061405539512634,
"learning_rate": 8.76976114684395e-06,
"loss": 0.5285,
"step": 160
},
{
"epoch": 0.9248444231689804,
"grad_norm": 0.5467921495437622,
"learning_rate": 8.747674561625121e-06,
"loss": 0.5486,
"step": 161
},
{
"epoch": 0.9305887984681666,
"grad_norm": 0.48409557342529297,
"learning_rate": 8.725419819655582e-06,
"loss": 0.4822,
"step": 162
},
{
"epoch": 0.9363331737673528,
"grad_norm": 0.5776729583740234,
"learning_rate": 8.702997919497247e-06,
"loss": 0.5402,
"step": 163
},
{
"epoch": 0.9420775490665391,
"grad_norm": 0.5763126611709595,
"learning_rate": 8.680409867212359e-06,
"loss": 0.5563,
"step": 164
},
{
"epoch": 0.9478219243657252,
"grad_norm": 0.6214870810508728,
"learning_rate": 8.657656676318346e-06,
"loss": 0.5526,
"step": 165
},
{
"epoch": 0.9535662996649115,
"grad_norm": 0.47466009855270386,
"learning_rate": 8.634739367742341e-06,
"loss": 0.503,
"step": 166
},
{
"epoch": 0.9593106749640976,
"grad_norm": 0.7167166471481323,
"learning_rate": 8.611658969775378e-06,
"loss": 0.5246,
"step": 167
},
{
"epoch": 0.9650550502632839,
"grad_norm": 0.5475093126296997,
"learning_rate": 8.588416518026248e-06,
"loss": 0.5171,
"step": 168
},
{
"epoch": 0.9707994255624701,
"grad_norm": 0.6024914383888245,
"learning_rate": 8.565013055375035e-06,
"loss": 0.5351,
"step": 169
},
{
"epoch": 0.9765438008616563,
"grad_norm": 0.6403016448020935,
"learning_rate": 8.541449631926325e-06,
"loss": 0.5271,
"step": 170
},
{
"epoch": 0.9822881761608425,
"grad_norm": 0.47844192385673523,
"learning_rate": 8.51772730496208e-06,
"loss": 0.5609,
"step": 171
},
{
"epoch": 0.9880325514600288,
"grad_norm": 0.6408485174179077,
"learning_rate": 8.49384713889421e-06,
"loss": 0.5275,
"step": 172
},
{
"epoch": 0.9937769267592149,
"grad_norm": 0.6476163864135742,
"learning_rate": 8.469810205216795e-06,
"loss": 0.566,
"step": 173
},
{
"epoch": 0.9995213020584012,
"grad_norm": 0.5650629997253418,
"learning_rate": 8.445617582458033e-06,
"loss": 0.5474,
"step": 174
},
{
"epoch": 1.0052656773575874,
"grad_norm": 1.5358240604400635,
"learning_rate": 8.42127035613182e-06,
"loss": 0.9106,
"step": 175
},
{
"epoch": 1.0110100526567736,
"grad_norm": 0.532135009765625,
"learning_rate": 8.396769618689064e-06,
"loss": 0.4946,
"step": 176
},
{
"epoch": 1.0167544279559597,
"grad_norm": 0.5433627367019653,
"learning_rate": 8.372116469468654e-06,
"loss": 0.5218,
"step": 177
},
{
"epoch": 1.022498803255146,
"grad_norm": 0.597786009311676,
"learning_rate": 8.347312014648144e-06,
"loss": 0.4627,
"step": 178
},
{
"epoch": 1.0282431785543322,
"grad_norm": 0.5359785556793213,
"learning_rate": 8.32235736719411e-06,
"loss": 0.4429,
"step": 179
},
{
"epoch": 1.0339875538535184,
"grad_norm": 0.7234981656074524,
"learning_rate": 8.297253646812213e-06,
"loss": 0.5291,
"step": 180
},
{
"epoch": 1.0397319291527047,
"grad_norm": 0.5277758836746216,
"learning_rate": 8.272001979896962e-06,
"loss": 0.4785,
"step": 181
},
{
"epoch": 1.0454763044518909,
"grad_norm": 0.5918593406677246,
"learning_rate": 8.246603499481177e-06,
"loss": 0.4763,
"step": 182
},
{
"epoch": 1.051220679751077,
"grad_norm": 0.6785145401954651,
"learning_rate": 8.221059345185136e-06,
"loss": 0.4874,
"step": 183
},
{
"epoch": 1.0569650550502634,
"grad_norm": 0.5301976203918457,
"learning_rate": 8.195370663165455e-06,
"loss": 0.4795,
"step": 184
},
{
"epoch": 1.0627094303494495,
"grad_norm": 0.6745317578315735,
"learning_rate": 8.169538606063647e-06,
"loss": 0.4973,
"step": 185
},
{
"epoch": 1.0684538056486357,
"grad_norm": 0.6252505779266357,
"learning_rate": 8.143564332954426e-06,
"loss": 0.5057,
"step": 186
},
{
"epoch": 1.0741981809478218,
"grad_norm": 0.5269699692726135,
"learning_rate": 8.117449009293668e-06,
"loss": 0.5062,
"step": 187
},
{
"epoch": 1.0799425562470082,
"grad_norm": 0.5290875434875488,
"learning_rate": 8.091193806866147e-06,
"loss": 0.5019,
"step": 188
},
{
"epoch": 1.0856869315461943,
"grad_norm": 0.5496700406074524,
"learning_rate": 8.064799903732936e-06,
"loss": 0.4652,
"step": 189
},
{
"epoch": 1.0914313068453805,
"grad_norm": 0.48876017332077026,
"learning_rate": 8.038268484178566e-06,
"loss": 0.4918,
"step": 190
},
{
"epoch": 1.0971756821445668,
"grad_norm": 0.49750837683677673,
"learning_rate": 8.011600738657865e-06,
"loss": 0.4935,
"step": 191
},
{
"epoch": 1.102920057443753,
"grad_norm": 0.5250374674797058,
"learning_rate": 7.98479786374257e-06,
"loss": 0.4556,
"step": 192
},
{
"epoch": 1.1086644327429391,
"grad_norm": 0.659435510635376,
"learning_rate": 7.957861062067614e-06,
"loss": 0.4911,
"step": 193
},
{
"epoch": 1.1144088080421255,
"grad_norm": 0.5870724320411682,
"learning_rate": 7.930791542277175e-06,
"loss": 0.4896,
"step": 194
},
{
"epoch": 1.1201531833413116,
"grad_norm": 0.5560335516929626,
"learning_rate": 7.903590518970445e-06,
"loss": 0.5133,
"step": 195
},
{
"epoch": 1.1258975586404978,
"grad_norm": 0.4820505380630493,
"learning_rate": 7.876259212647129e-06,
"loss": 0.4414,
"step": 196
},
{
"epoch": 1.1316419339396842,
"grad_norm": 0.5640401840209961,
"learning_rate": 7.848798849652684e-06,
"loss": 0.5015,
"step": 197
},
{
"epoch": 1.1373863092388703,
"grad_norm": 0.47365468740463257,
"learning_rate": 7.821210662123284e-06,
"loss": 0.474,
"step": 198
},
{
"epoch": 1.1431306845380564,
"grad_norm": 0.4788399338722229,
"learning_rate": 7.793495887930551e-06,
"loss": 0.4463,
"step": 199
},
{
"epoch": 1.1488750598372426,
"grad_norm": 0.5238497257232666,
"learning_rate": 7.765655770625997e-06,
"loss": 0.4672,
"step": 200
},
{
"epoch": 1.154619435136429,
"grad_norm": 0.6221843957901001,
"learning_rate": 7.737691559385237e-06,
"loss": 0.4986,
"step": 201
},
{
"epoch": 1.160363810435615,
"grad_norm": 0.47932368516921997,
"learning_rate": 7.709604508951927e-06,
"loss": 0.51,
"step": 202
},
{
"epoch": 1.1661081857348012,
"grad_norm": 0.5288423299789429,
"learning_rate": 7.68139587958148e-06,
"loss": 0.4683,
"step": 203
},
{
"epoch": 1.1718525610339876,
"grad_norm": 0.5743793845176697,
"learning_rate": 7.653066936984504e-06,
"loss": 0.5043,
"step": 204
},
{
"epoch": 1.1775969363331737,
"grad_norm": 0.5184981822967529,
"learning_rate": 7.6246189522700205e-06,
"loss": 0.5199,
"step": 205
},
{
"epoch": 1.18334131163236,
"grad_norm": 0.4760245382785797,
"learning_rate": 7.596053201888425e-06,
"loss": 0.4666,
"step": 206
},
{
"epoch": 1.1890856869315463,
"grad_norm": 0.5408570170402527,
"learning_rate": 7.56737096757421e-06,
"loss": 0.467,
"step": 207
},
{
"epoch": 1.1948300622307324,
"grad_norm": 0.47764548659324646,
"learning_rate": 7.538573536288466e-06,
"loss": 0.4847,
"step": 208
},
{
"epoch": 1.2005744375299185,
"grad_norm": 0.4249972403049469,
"learning_rate": 7.509662200161122e-06,
"loss": 0.4524,
"step": 209
},
{
"epoch": 1.206318812829105,
"grad_norm": 0.539272665977478,
"learning_rate": 7.480638256432977e-06,
"loss": 0.5122,
"step": 210
},
{
"epoch": 1.212063188128291,
"grad_norm": 0.4662354290485382,
"learning_rate": 7.4515030073974915e-06,
"loss": 0.4921,
"step": 211
},
{
"epoch": 1.2178075634274772,
"grad_norm": 0.4967215657234192,
"learning_rate": 7.422257760342351e-06,
"loss": 0.4924,
"step": 212
},
{
"epoch": 1.2235519387266636,
"grad_norm": 0.5155688524246216,
"learning_rate": 7.392903827490814e-06,
"loss": 0.4601,
"step": 213
},
{
"epoch": 1.2292963140258497,
"grad_norm": 0.5294659733772278,
"learning_rate": 7.363442525942827e-06,
"loss": 0.4475,
"step": 214
},
{
"epoch": 1.2350406893250359,
"grad_norm": 0.46415042877197266,
"learning_rate": 7.333875177615931e-06,
"loss": 0.4933,
"step": 215
},
{
"epoch": 1.2407850646242222,
"grad_norm": 0.6214727163314819,
"learning_rate": 7.304203109185947e-06,
"loss": 0.501,
"step": 216
},
{
"epoch": 1.2465294399234084,
"grad_norm": 0.452462762594223,
"learning_rate": 7.274427652027444e-06,
"loss": 0.4457,
"step": 217
},
{
"epoch": 1.2522738152225945,
"grad_norm": 0.5127636790275574,
"learning_rate": 7.244550142154009e-06,
"loss": 0.4801,
"step": 218
},
{
"epoch": 1.2580181905217809,
"grad_norm": 0.5973929762840271,
"learning_rate": 7.214571920158293e-06,
"loss": 0.4854,
"step": 219
},
{
"epoch": 1.263762565820967,
"grad_norm": 0.5596358776092529,
"learning_rate": 7.1844943311518665e-06,
"loss": 0.5256,
"step": 220
},
{
"epoch": 1.2695069411201532,
"grad_norm": 0.46867719292640686,
"learning_rate": 7.1543187247048525e-06,
"loss": 0.5043,
"step": 221
},
{
"epoch": 1.2752513164193395,
"grad_norm": 0.5118942260742188,
"learning_rate": 7.124046454785387e-06,
"loss": 0.4747,
"step": 222
},
{
"epoch": 1.2809956917185257,
"grad_norm": 0.5921294093132019,
"learning_rate": 7.093678879698858e-06,
"loss": 0.5213,
"step": 223
},
{
"epoch": 1.2867400670177118,
"grad_norm": 0.4742209017276764,
"learning_rate": 7.063217362026957e-06,
"loss": 0.4703,
"step": 224
},
{
"epoch": 1.292484442316898,
"grad_norm": 0.4378364086151123,
"learning_rate": 7.032663268566547e-06,
"loss": 0.4582,
"step": 225
},
{
"epoch": 1.2982288176160843,
"grad_norm": 0.5260799527168274,
"learning_rate": 7.002017970268336e-06,
"loss": 0.4656,
"step": 226
},
{
"epoch": 1.3039731929152705,
"grad_norm": 0.5427009463310242,
"learning_rate": 6.97128284217535e-06,
"loss": 0.5092,
"step": 227
},
{
"epoch": 1.3097175682144566,
"grad_norm": 0.5533111095428467,
"learning_rate": 6.9404592633612486e-06,
"loss": 0.5472,
"step": 228
},
{
"epoch": 1.3154619435136428,
"grad_norm": 0.5076018571853638,
"learning_rate": 6.909548616868444e-06,
"loss": 0.4966,
"step": 229
},
{
"epoch": 1.3212063188128291,
"grad_norm": 0.47278183698654175,
"learning_rate": 6.878552289646041e-06,
"loss": 0.4822,
"step": 230
},
{
"epoch": 1.3269506941120153,
"grad_norm": 0.45742684602737427,
"learning_rate": 6.847471672487607e-06,
"loss": 0.4365,
"step": 231
},
{
"epoch": 1.3326950694112014,
"grad_norm": 0.4847550392150879,
"learning_rate": 6.816308159968761e-06,
"loss": 0.5396,
"step": 232
},
{
"epoch": 1.3384394447103878,
"grad_norm": 0.5120888352394104,
"learning_rate": 6.7850631503846165e-06,
"loss": 0.4811,
"step": 233
},
{
"epoch": 1.344183820009574,
"grad_norm": 0.4778832197189331,
"learning_rate": 6.753738045687021e-06,
"loss": 0.4946,
"step": 234
},
{
"epoch": 1.34992819530876,
"grad_norm": 0.41883185505867004,
"learning_rate": 6.722334251421665e-06,
"loss": 0.4706,
"step": 235
},
{
"epoch": 1.3556725706079464,
"grad_norm": 0.5374260544776917,
"learning_rate": 6.690853176665007e-06,
"loss": 0.5349,
"step": 236
},
{
"epoch": 1.3614169459071326,
"grad_norm": 0.4322339594364166,
"learning_rate": 6.659296233961055e-06,
"loss": 0.4529,
"step": 237
},
{
"epoch": 1.3671613212063187,
"grad_norm": 0.46204888820648193,
"learning_rate": 6.627664839257979e-06,
"loss": 0.4836,
"step": 238
},
{
"epoch": 1.372905696505505,
"grad_norm": 0.4843834340572357,
"learning_rate": 6.595960411844589e-06,
"loss": 0.5056,
"step": 239
},
{
"epoch": 1.3786500718046912,
"grad_norm": 0.4898874759674072,
"learning_rate": 6.564184374286636e-06,
"loss": 0.5039,
"step": 240
},
{
"epoch": 1.3843944471038774,
"grad_norm": 0.40900927782058716,
"learning_rate": 6.532338152363001e-06,
"loss": 0.4653,
"step": 241
},
{
"epoch": 1.3901388224030637,
"grad_norm": 0.46595582365989685,
"learning_rate": 6.500423175001705e-06,
"loss": 0.4342,
"step": 242
},
{
"epoch": 1.39588319770225,
"grad_norm": 0.5343397259712219,
"learning_rate": 6.468440874215801e-06,
"loss": 0.5061,
"step": 243
},
{
"epoch": 1.401627573001436,
"grad_norm": 0.47447115182876587,
"learning_rate": 6.43639268503912e-06,
"loss": 0.4998,
"step": 244
},
{
"epoch": 1.4073719483006224,
"grad_norm": 0.5101116299629211,
"learning_rate": 6.40428004546188e-06,
"loss": 0.5072,
"step": 245
},
{
"epoch": 1.4131163235998085,
"grad_norm": 0.47331035137176514,
"learning_rate": 6.372104396366162e-06,
"loss": 0.477,
"step": 246
},
{
"epoch": 1.4188606988989947,
"grad_norm": 0.45054635405540466,
"learning_rate": 6.339867181461265e-06,
"loss": 0.5065,
"step": 247
},
{
"epoch": 1.424605074198181,
"grad_norm": 0.42182686924934387,
"learning_rate": 6.307569847218917e-06,
"loss": 0.4668,
"step": 248
},
{
"epoch": 1.4303494494973672,
"grad_norm": 0.44658786058425903,
"learning_rate": 6.275213842808383e-06,
"loss": 0.4838,
"step": 249
},
{
"epoch": 1.4360938247965533,
"grad_norm": 0.47851428389549255,
"learning_rate": 6.242800620031434e-06,
"loss": 0.4706,
"step": 250
},
{
"epoch": 1.4418382000957397,
"grad_norm": 0.43599647283554077,
"learning_rate": 6.2103316332572095e-06,
"loss": 0.5027,
"step": 251
},
{
"epoch": 1.4475825753949259,
"grad_norm": 0.47949090600013733,
"learning_rate": 6.177808339356954e-06,
"loss": 0.5254,
"step": 252
},
{
"epoch": 1.453326950694112,
"grad_norm": 0.42577695846557617,
"learning_rate": 6.14523219763866e-06,
"loss": 0.4559,
"step": 253
},
{
"epoch": 1.4590713259932984,
"grad_norm": 0.4304141104221344,
"learning_rate": 6.112604669781572e-06,
"loss": 0.4851,
"step": 254
},
{
"epoch": 1.4648157012924845,
"grad_norm": 0.46502649784088135,
"learning_rate": 6.079927219770623e-06,
"loss": 0.4963,
"step": 255
},
{
"epoch": 1.4705600765916707,
"grad_norm": 0.4513600170612335,
"learning_rate": 6.047201313830724e-06,
"loss": 0.4864,
"step": 256
},
{
"epoch": 1.4763044518908568,
"grad_norm": 0.44206321239471436,
"learning_rate": 6.014428420360987e-06,
"loss": 0.5142,
"step": 257
},
{
"epoch": 1.4820488271900432,
"grad_norm": 0.42616114020347595,
"learning_rate": 5.9816100098688456e-06,
"loss": 0.4622,
"step": 258
},
{
"epoch": 1.4877932024892293,
"grad_norm": 0.46303677558898926,
"learning_rate": 5.948747554904054e-06,
"loss": 0.5215,
"step": 259
},
{
"epoch": 1.4935375777884154,
"grad_norm": 0.48208341002464294,
"learning_rate": 5.915842529992632e-06,
"loss": 0.4846,
"step": 260
},
{
"epoch": 1.4992819530876016,
"grad_norm": 0.4425645172595978,
"learning_rate": 5.8828964115706925e-06,
"loss": 0.4288,
"step": 261
},
{
"epoch": 1.505026328386788,
"grad_norm": 0.4504961371421814,
"learning_rate": 5.849910677918205e-06,
"loss": 0.4554,
"step": 262
},
{
"epoch": 1.510770703685974,
"grad_norm": 0.46808743476867676,
"learning_rate": 5.816886809092651e-06,
"loss": 0.4694,
"step": 263
},
{
"epoch": 1.5165150789851602,
"grad_norm": 0.5024488568305969,
"learning_rate": 5.783826286862631e-06,
"loss": 0.5115,
"step": 264
},
{
"epoch": 1.5222594542843466,
"grad_norm": 0.40128159523010254,
"learning_rate": 5.750730594641367e-06,
"loss": 0.4581,
"step": 265
},
{
"epoch": 1.5280038295835328,
"grad_norm": 0.45904093980789185,
"learning_rate": 5.717601217420143e-06,
"loss": 0.4697,
"step": 266
},
{
"epoch": 1.533748204882719,
"grad_norm": 0.4493611752986908,
"learning_rate": 5.68443964170168e-06,
"loss": 0.4777,
"step": 267
},
{
"epoch": 1.5394925801819053,
"grad_norm": 0.49274754524230957,
"learning_rate": 5.6512473554334294e-06,
"loss": 0.4905,
"step": 268
},
{
"epoch": 1.5452369554810914,
"grad_norm": 0.4549119770526886,
"learning_rate": 5.618025847940817e-06,
"loss": 0.4858,
"step": 269
},
{
"epoch": 1.5509813307802776,
"grad_norm": 0.4543653130531311,
"learning_rate": 5.584776609860414e-06,
"loss": 0.4701,
"step": 270
},
{
"epoch": 1.556725706079464,
"grad_norm": 0.49982017278671265,
"learning_rate": 5.551501133073048e-06,
"loss": 0.545,
"step": 271
},
{
"epoch": 1.56247008137865,
"grad_norm": 0.40904542803764343,
"learning_rate": 5.518200910636875e-06,
"loss": 0.468,
"step": 272
},
{
"epoch": 1.5682144566778362,
"grad_norm": 0.4438726603984833,
"learning_rate": 5.4848774367203715e-06,
"loss": 0.4999,
"step": 273
},
{
"epoch": 1.5739588319770226,
"grad_norm": 0.4355951249599457,
"learning_rate": 5.451532206535306e-06,
"loss": 0.5067,
"step": 274
},
{
"epoch": 1.5797032072762087,
"grad_norm": 0.5013201236724854,
"learning_rate": 5.418166716269636e-06,
"loss": 0.4963,
"step": 275
},
{
"epoch": 1.5854475825753949,
"grad_norm": 0.42678409814834595,
"learning_rate": 5.384782463020385e-06,
"loss": 0.4908,
"step": 276
},
{
"epoch": 1.5911919578745812,
"grad_norm": 0.396161824464798,
"learning_rate": 5.351380944726465e-06,
"loss": 0.4867,
"step": 277
},
{
"epoch": 1.5969363331737674,
"grad_norm": 0.36465850472450256,
"learning_rate": 5.317963660101464e-06,
"loss": 0.4222,
"step": 278
},
{
"epoch": 1.6026807084729535,
"grad_norm": 0.47781455516815186,
"learning_rate": 5.284532108566396e-06,
"loss": 0.4913,
"step": 279
},
{
"epoch": 1.6084250837721399,
"grad_norm": 0.4402898848056793,
"learning_rate": 5.251087790182428e-06,
"loss": 0.4912,
"step": 280
},
{
"epoch": 1.614169459071326,
"grad_norm": 0.4674187898635864,
"learning_rate": 5.217632205583574e-06,
"loss": 0.5057,
"step": 281
},
{
"epoch": 1.6199138343705122,
"grad_norm": 0.41630300879478455,
"learning_rate": 5.184166855909355e-06,
"loss": 0.4883,
"step": 282
},
{
"epoch": 1.6256582096696985,
"grad_norm": 0.42907628417015076,
"learning_rate": 5.150693242737444e-06,
"loss": 0.5034,
"step": 283
},
{
"epoch": 1.6314025849688847,
"grad_norm": 0.43026891350746155,
"learning_rate": 5.117212868016303e-06,
"loss": 0.5234,
"step": 284
},
{
"epoch": 1.6371469602680708,
"grad_norm": 0.5048802495002747,
"learning_rate": 5.083727233997775e-06,
"loss": 0.4608,
"step": 285
},
{
"epoch": 1.6428913355672572,
"grad_norm": 0.4373928904533386,
"learning_rate": 5.05023784316969e-06,
"loss": 0.4796,
"step": 286
},
{
"epoch": 1.6486357108664431,
"grad_norm": 0.4595464766025543,
"learning_rate": 5.016746198188439e-06,
"loss": 0.5178,
"step": 287
},
{
"epoch": 1.6543800861656295,
"grad_norm": 0.36910781264305115,
"learning_rate": 4.983253801811562e-06,
"loss": 0.4475,
"step": 288
},
{
"epoch": 1.6601244614648158,
"grad_norm": 0.4128073751926422,
"learning_rate": 4.949762156830312e-06,
"loss": 0.5042,
"step": 289
},
{
"epoch": 1.6658688367640018,
"grad_norm": 0.45801302790641785,
"learning_rate": 4.916272766002227e-06,
"loss": 0.4878,
"step": 290
},
{
"epoch": 1.6716132120631881,
"grad_norm": 0.45329952239990234,
"learning_rate": 4.882787131983698e-06,
"loss": 0.4534,
"step": 291
},
{
"epoch": 1.6773575873623745,
"grad_norm": 0.4560602605342865,
"learning_rate": 4.849306757262558e-06,
"loss": 0.4711,
"step": 292
},
{
"epoch": 1.6831019626615604,
"grad_norm": 0.5043655037879944,
"learning_rate": 4.8158331440906466e-06,
"loss": 0.5153,
"step": 293
},
{
"epoch": 1.6888463379607468,
"grad_norm": 0.47163787484169006,
"learning_rate": 4.7823677944164285e-06,
"loss": 0.5405,
"step": 294
},
{
"epoch": 1.694590713259933,
"grad_norm": 0.4477205276489258,
"learning_rate": 4.748912209817572e-06,
"loss": 0.4782,
"step": 295
},
{
"epoch": 1.700335088559119,
"grad_norm": 0.39727628231048584,
"learning_rate": 4.715467891433607e-06,
"loss": 0.4481,
"step": 296
},
{
"epoch": 1.7060794638583054,
"grad_norm": 0.40849247574806213,
"learning_rate": 4.682036339898537e-06,
"loss": 0.4783,
"step": 297
},
{
"epoch": 1.7118238391574916,
"grad_norm": 0.4352889358997345,
"learning_rate": 4.6486190552735375e-06,
"loss": 0.5101,
"step": 298
},
{
"epoch": 1.7175682144566777,
"grad_norm": 0.4444569945335388,
"learning_rate": 4.615217536979616e-06,
"loss": 0.481,
"step": 299
},
{
"epoch": 1.723312589755864,
"grad_norm": 0.39897894859313965,
"learning_rate": 4.581833283730367e-06,
"loss": 0.4479,
"step": 300
},
{
"epoch": 1.7290569650550502,
"grad_norm": 0.48069655895233154,
"learning_rate": 4.548467793464696e-06,
"loss": 0.5243,
"step": 301
},
{
"epoch": 1.7348013403542364,
"grad_norm": 0.39833492040634155,
"learning_rate": 4.515122563279631e-06,
"loss": 0.4563,
"step": 302
},
{
"epoch": 1.7405457156534228,
"grad_norm": 0.40915027260780334,
"learning_rate": 4.481799089363127e-06,
"loss": 0.459,
"step": 303
},
{
"epoch": 1.746290090952609,
"grad_norm": 0.45061710476875305,
"learning_rate": 4.448498866926952e-06,
"loss": 0.4973,
"step": 304
},
{
"epoch": 1.752034466251795,
"grad_norm": 0.40321865677833557,
"learning_rate": 4.415223390139588e-06,
"loss": 0.5133,
"step": 305
},
{
"epoch": 1.7577788415509814,
"grad_norm": 0.4163152575492859,
"learning_rate": 4.381974152059184e-06,
"loss": 0.4906,
"step": 306
},
{
"epoch": 1.7635232168501676,
"grad_norm": 0.3934337794780731,
"learning_rate": 4.348752644566573e-06,
"loss": 0.4561,
"step": 307
},
{
"epoch": 1.7692675921493537,
"grad_norm": 0.4110010862350464,
"learning_rate": 4.315560358298321e-06,
"loss": 0.499,
"step": 308
},
{
"epoch": 1.77501196744854,
"grad_norm": 0.4093308448791504,
"learning_rate": 4.2823987825798575e-06,
"loss": 0.4681,
"step": 309
},
{
"epoch": 1.7807563427477262,
"grad_norm": 0.3898443281650543,
"learning_rate": 4.249269405358634e-06,
"loss": 0.4545,
"step": 310
},
{
"epoch": 1.7865007180469124,
"grad_norm": 0.4322449564933777,
"learning_rate": 4.2161737131373695e-06,
"loss": 0.5156,
"step": 311
},
{
"epoch": 1.7922450933460987,
"grad_norm": 0.41032981872558594,
"learning_rate": 4.183113190907349e-06,
"loss": 0.4756,
"step": 312
},
{
"epoch": 1.7979894686452849,
"grad_norm": 0.43316203355789185,
"learning_rate": 4.150089322081797e-06,
"loss": 0.4614,
"step": 313
},
{
"epoch": 1.803733843944471,
"grad_norm": 0.3737150728702545,
"learning_rate": 4.1171035884293075e-06,
"loss": 0.4491,
"step": 314
},
{
"epoch": 1.8094782192436574,
"grad_norm": 0.3852100074291229,
"learning_rate": 4.084157470007371e-06,
"loss": 0.4828,
"step": 315
},
{
"epoch": 1.8152225945428435,
"grad_norm": 0.44559043645858765,
"learning_rate": 4.051252445095946e-06,
"loss": 0.5271,
"step": 316
},
{
"epoch": 1.8209669698420297,
"grad_norm": 0.3891163766384125,
"learning_rate": 4.018389990131156e-06,
"loss": 0.4841,
"step": 317
},
{
"epoch": 1.826711345141216,
"grad_norm": 0.3703552782535553,
"learning_rate": 3.985571579639013e-06,
"loss": 0.4684,
"step": 318
},
{
"epoch": 1.832455720440402,
"grad_norm": 0.38720962405204773,
"learning_rate": 3.952798686169279e-06,
"loss": 0.4684,
"step": 319
},
{
"epoch": 1.8382000957395883,
"grad_norm": 0.4561994671821594,
"learning_rate": 3.920072780229378e-06,
"loss": 0.4763,
"step": 320
},
{
"epoch": 1.8439444710387747,
"grad_norm": 0.4258320927619934,
"learning_rate": 3.887395330218429e-06,
"loss": 0.504,
"step": 321
},
{
"epoch": 1.8496888463379606,
"grad_norm": 0.3646913766860962,
"learning_rate": 3.854767802361342e-06,
"loss": 0.4414,
"step": 322
},
{
"epoch": 1.855433221637147,
"grad_norm": 0.44896551966667175,
"learning_rate": 3.822191660643047e-06,
"loss": 0.5013,
"step": 323
},
{
"epoch": 1.8611775969363333,
"grad_norm": 0.38893213868141174,
"learning_rate": 3.789668366742792e-06,
"loss": 0.5048,
"step": 324
},
{
"epoch": 1.8669219722355193,
"grad_norm": 0.40248069167137146,
"learning_rate": 3.7571993799685675e-06,
"loss": 0.4853,
"step": 325
},
{
"epoch": 1.8726663475347056,
"grad_norm": 0.4159008860588074,
"learning_rate": 3.7247861571916183e-06,
"loss": 0.465,
"step": 326
},
{
"epoch": 1.8784107228338918,
"grad_norm": 0.4135017395019531,
"learning_rate": 3.6924301527810856e-06,
"loss": 0.4831,
"step": 327
},
{
"epoch": 1.884155098133078,
"grad_norm": 0.42082080245018005,
"learning_rate": 3.6601328185387364e-06,
"loss": 0.4992,
"step": 328
},
{
"epoch": 1.8898994734322643,
"grad_norm": 0.3914032578468323,
"learning_rate": 3.6278956036338397e-06,
"loss": 0.4611,
"step": 329
},
{
"epoch": 1.8956438487314504,
"grad_norm": 0.40370723605155945,
"learning_rate": 3.5957199545381216e-06,
"loss": 0.4542,
"step": 330
},
{
"epoch": 1.9013882240306366,
"grad_norm": 0.423311710357666,
"learning_rate": 3.5636073149608824e-06,
"loss": 0.4777,
"step": 331
},
{
"epoch": 1.907132599329823,
"grad_norm": 0.39622029662132263,
"learning_rate": 3.5315591257842e-06,
"loss": 0.4278,
"step": 332
},
{
"epoch": 1.912876974629009,
"grad_norm": 0.4199773669242859,
"learning_rate": 3.4995768249982975e-06,
"loss": 0.459,
"step": 333
},
{
"epoch": 1.9186213499281952,
"grad_norm": 0.40625452995300293,
"learning_rate": 3.467661847637001e-06,
"loss": 0.502,
"step": 334
},
{
"epoch": 1.9243657252273816,
"grad_norm": 0.3874070644378662,
"learning_rate": 3.4358156257133644e-06,
"loss": 0.5028,
"step": 335
},
{
"epoch": 1.9301101005265677,
"grad_norm": 0.4368024170398712,
"learning_rate": 3.404039588155413e-06,
"loss": 0.4439,
"step": 336
},
{
"epoch": 1.9358544758257539,
"grad_norm": 0.37655797600746155,
"learning_rate": 3.372335160742022e-06,
"loss": 0.4325,
"step": 337
},
{
"epoch": 1.9415988511249402,
"grad_norm": 0.37845367193222046,
"learning_rate": 3.3407037660389474e-06,
"loss": 0.4567,
"step": 338
},
{
"epoch": 1.9473432264241264,
"grad_norm": 0.38552960753440857,
"learning_rate": 3.3091468233349934e-06,
"loss": 0.5217,
"step": 339
},
{
"epoch": 1.9530876017233125,
"grad_norm": 0.4249011278152466,
"learning_rate": 3.2776657485783357e-06,
"loss": 0.484,
"step": 340
},
{
"epoch": 1.958831977022499,
"grad_norm": 0.42907825112342834,
"learning_rate": 3.246261954312979e-06,
"loss": 0.496,
"step": 341
},
{
"epoch": 1.964576352321685,
"grad_norm": 0.39993909001350403,
"learning_rate": 3.2149368496153856e-06,
"loss": 0.5015,
"step": 342
},
{
"epoch": 1.9703207276208712,
"grad_norm": 0.41646891832351685,
"learning_rate": 3.1836918400312387e-06,
"loss": 0.4545,
"step": 343
},
{
"epoch": 1.9760651029200575,
"grad_norm": 0.41628479957580566,
"learning_rate": 3.152528327512395e-06,
"loss": 0.4743,
"step": 344
},
{
"epoch": 1.9818094782192437,
"grad_norm": 0.4280012249946594,
"learning_rate": 3.1214477103539585e-06,
"loss": 0.4885,
"step": 345
},
{
"epoch": 1.9875538535184298,
"grad_norm": 0.41522058844566345,
"learning_rate": 3.0904513831315563e-06,
"loss": 0.5036,
"step": 346
},
{
"epoch": 1.9932982288176162,
"grad_norm": 0.4009215831756592,
"learning_rate": 3.059540736638751e-06,
"loss": 0.5056,
"step": 347
},
{
"epoch": 1.9990426041168023,
"grad_norm": 0.3924543857574463,
"learning_rate": 3.028717157824652e-06,
"loss": 0.4316,
"step": 348
},
{
"epoch": 2.0047869794159885,
"grad_norm": 1.5738595724105835,
"learning_rate": 2.9979820297316652e-06,
"loss": 0.7748,
"step": 349
},
{
"epoch": 2.010531354715175,
"grad_norm": 0.5284481048583984,
"learning_rate": 2.9673367314334533e-06,
"loss": 0.5035,
"step": 350
},
{
"epoch": 2.016275730014361,
"grad_norm": 0.4917297959327698,
"learning_rate": 2.936782637973044e-06,
"loss": 0.4493,
"step": 351
},
{
"epoch": 2.022020105313547,
"grad_norm": 0.4824795424938202,
"learning_rate": 2.9063211203011443e-06,
"loss": 0.4554,
"step": 352
},
{
"epoch": 2.0277644806127335,
"grad_norm": 0.407044380903244,
"learning_rate": 2.8759535452146128e-06,
"loss": 0.4531,
"step": 353
},
{
"epoch": 2.0335088559119194,
"grad_norm": 0.4383600354194641,
"learning_rate": 2.8456812752951483e-06,
"loss": 0.4553,
"step": 354
},
{
"epoch": 2.039253231211106,
"grad_norm": 0.4619106948375702,
"learning_rate": 2.815505668848136e-06,
"loss": 0.4195,
"step": 355
},
{
"epoch": 2.044997606510292,
"grad_norm": 0.4435921907424927,
"learning_rate": 2.785428079841709e-06,
"loss": 0.4514,
"step": 356
},
{
"epoch": 2.050741981809478,
"grad_norm": 0.41282203793525696,
"learning_rate": 2.755449857845992e-06,
"loss": 0.4489,
"step": 357
},
{
"epoch": 2.0564863571086645,
"grad_norm": 0.4201320707798004,
"learning_rate": 2.725572347972558e-06,
"loss": 0.42,
"step": 358
},
{
"epoch": 2.062230732407851,
"grad_norm": 0.40200960636138916,
"learning_rate": 2.6957968908140546e-06,
"loss": 0.4273,
"step": 359
},
{
"epoch": 2.0679751077070367,
"grad_norm": 0.45121583342552185,
"learning_rate": 2.666124822384071e-06,
"loss": 0.4482,
"step": 360
},
{
"epoch": 2.073719483006223,
"grad_norm": 0.38477957248687744,
"learning_rate": 2.636557474057173e-06,
"loss": 0.4017,
"step": 361
},
{
"epoch": 2.0794638583054095,
"grad_norm": 0.3674720823764801,
"learning_rate": 2.607096172509187e-06,
"loss": 0.4402,
"step": 362
},
{
"epoch": 2.0852082336045954,
"grad_norm": 0.43862369656562805,
"learning_rate": 2.5777422396576503e-06,
"loss": 0.492,
"step": 363
},
{
"epoch": 2.0909526089037818,
"grad_norm": 0.36172252893447876,
"learning_rate": 2.5484969926025114e-06,
"loss": 0.4384,
"step": 364
},
{
"epoch": 2.096696984202968,
"grad_norm": 0.42003771662712097,
"learning_rate": 2.5193617435670244e-06,
"loss": 0.4897,
"step": 365
},
{
"epoch": 2.102441359502154,
"grad_norm": 0.38633331656455994,
"learning_rate": 2.4903377998388783e-06,
"loss": 0.4403,
"step": 366
},
{
"epoch": 2.1081857348013404,
"grad_norm": 0.3912460207939148,
"learning_rate": 2.461426463711535e-06,
"loss": 0.4439,
"step": 367
},
{
"epoch": 2.113930110100527,
"grad_norm": 0.35407376289367676,
"learning_rate": 2.4326290324257896e-06,
"loss": 0.4234,
"step": 368
},
{
"epoch": 2.1196744853997127,
"grad_norm": 0.36607515811920166,
"learning_rate": 2.403946798111576e-06,
"loss": 0.4148,
"step": 369
},
{
"epoch": 2.125418860698899,
"grad_norm": 0.4565516710281372,
"learning_rate": 2.37538104772998e-06,
"loss": 0.486,
"step": 370
},
{
"epoch": 2.131163235998085,
"grad_norm": 0.3397116959095001,
"learning_rate": 2.3469330630154974e-06,
"loss": 0.3835,
"step": 371
},
{
"epoch": 2.1369076112972714,
"grad_norm": 0.39572739601135254,
"learning_rate": 2.318604120418521e-06,
"loss": 0.4679,
"step": 372
},
{
"epoch": 2.1426519865964577,
"grad_norm": 0.3581886887550354,
"learning_rate": 2.2903954910480746e-06,
"loss": 0.4637,
"step": 373
},
{
"epoch": 2.1483963618956436,
"grad_norm": 0.3759213387966156,
"learning_rate": 2.2623084406147643e-06,
"loss": 0.4316,
"step": 374
},
{
"epoch": 2.15414073719483,
"grad_norm": 0.38246414065361023,
"learning_rate": 2.234344229374003e-06,
"loss": 0.4416,
"step": 375
},
{
"epoch": 2.1598851124940164,
"grad_norm": 0.376297265291214,
"learning_rate": 2.2065041120694487e-06,
"loss": 0.4426,
"step": 376
},
{
"epoch": 2.1656294877932023,
"grad_norm": 0.3699423670768738,
"learning_rate": 2.178789337876716e-06,
"loss": 0.4452,
"step": 377
},
{
"epoch": 2.1713738630923887,
"grad_norm": 0.39869898557662964,
"learning_rate": 2.151201150347318e-06,
"loss": 0.437,
"step": 378
},
{
"epoch": 2.177118238391575,
"grad_norm": 0.3581327795982361,
"learning_rate": 2.123740787352872e-06,
"loss": 0.4054,
"step": 379
},
{
"epoch": 2.182862613690761,
"grad_norm": 0.3862365782260895,
"learning_rate": 2.096409481029556e-06,
"loss": 0.443,
"step": 380
},
{
"epoch": 2.1886069889899473,
"grad_norm": 0.38961994647979736,
"learning_rate": 2.069208457722828e-06,
"loss": 0.4219,
"step": 381
},
{
"epoch": 2.1943513642891337,
"grad_norm": 0.41970494389533997,
"learning_rate": 2.042138937932388e-06,
"loss": 0.4259,
"step": 382
},
{
"epoch": 2.2000957395883196,
"grad_norm": 0.39905792474746704,
"learning_rate": 2.015202136257432e-06,
"loss": 0.4904,
"step": 383
},
{
"epoch": 2.205840114887506,
"grad_norm": 0.36205026507377625,
"learning_rate": 1.988399261342135e-06,
"loss": 0.4056,
"step": 384
},
{
"epoch": 2.2115844901866923,
"grad_norm": 0.38249823451042175,
"learning_rate": 1.9617315158214363e-06,
"loss": 0.4665,
"step": 385
},
{
"epoch": 2.2173288654858783,
"grad_norm": 0.36974844336509705,
"learning_rate": 1.935200096267064e-06,
"loss": 0.4481,
"step": 386
},
{
"epoch": 2.2230732407850646,
"grad_norm": 0.38684070110321045,
"learning_rate": 1.908806193133855e-06,
"loss": 0.4327,
"step": 387
},
{
"epoch": 2.228817616084251,
"grad_norm": 0.3696470856666565,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.4609,
"step": 388
},
{
"epoch": 2.234561991383437,
"grad_norm": 0.4029988646507263,
"learning_rate": 1.856435667045577e-06,
"loss": 0.4576,
"step": 389
},
{
"epoch": 2.2403063666826233,
"grad_norm": 0.4148859679698944,
"learning_rate": 1.8304613939363531e-06,
"loss": 0.4519,
"step": 390
},
{
"epoch": 2.2460507419818097,
"grad_norm": 0.3689860701560974,
"learning_rate": 1.8046293368345485e-06,
"loss": 0.434,
"step": 391
},
{
"epoch": 2.2517951172809956,
"grad_norm": 0.331613689661026,
"learning_rate": 1.7789406548148647e-06,
"loss": 0.4185,
"step": 392
},
{
"epoch": 2.257539492580182,
"grad_norm": 0.41019827127456665,
"learning_rate": 1.7533965005188242e-06,
"loss": 0.522,
"step": 393
},
{
"epoch": 2.2632838678793683,
"grad_norm": 0.346804678440094,
"learning_rate": 1.7279980201030382e-06,
"loss": 0.4143,
"step": 394
},
{
"epoch": 2.2690282431785542,
"grad_norm": 0.3892984688282013,
"learning_rate": 1.7027463531877897e-06,
"loss": 0.4446,
"step": 395
},
{
"epoch": 2.2747726184777406,
"grad_norm": 0.3754962980747223,
"learning_rate": 1.677642632805892e-06,
"loss": 0.4522,
"step": 396
},
{
"epoch": 2.280516993776927,
"grad_norm": 0.39690718054771423,
"learning_rate": 1.6526879853518558e-06,
"loss": 0.447,
"step": 397
},
{
"epoch": 2.286261369076113,
"grad_norm": 0.37459442019462585,
"learning_rate": 1.6278835305313462e-06,
"loss": 0.4545,
"step": 398
},
{
"epoch": 2.2920057443752992,
"grad_norm": 0.347006231546402,
"learning_rate": 1.6032303813109368e-06,
"loss": 0.4486,
"step": 399
},
{
"epoch": 2.297750119674485,
"grad_norm": 0.3818560540676117,
"learning_rate": 1.578729643868181e-06,
"loss": 0.4554,
"step": 400
},
{
"epoch": 2.3034944949736715,
"grad_norm": 0.3954823613166809,
"learning_rate": 1.5543824175419691e-06,
"loss": 0.4624,
"step": 401
},
{
"epoch": 2.309238870272858,
"grad_norm": 0.33383116126060486,
"learning_rate": 1.5301897947832063e-06,
"loss": 0.4478,
"step": 402
},
{
"epoch": 2.3149832455720443,
"grad_norm": 0.36856454610824585,
"learning_rate": 1.5061528611057917e-06,
"loss": 0.4083,
"step": 403
},
{
"epoch": 2.32072762087123,
"grad_norm": 0.3847210705280304,
"learning_rate": 1.4822726950379207e-06,
"loss": 0.4766,
"step": 404
},
{
"epoch": 2.3264719961704166,
"grad_norm": 0.36447674036026,
"learning_rate": 1.4585503680736756e-06,
"loss": 0.436,
"step": 405
},
{
"epoch": 2.3322163714696025,
"grad_norm": 0.3663654327392578,
"learning_rate": 1.4349869446249664e-06,
"loss": 0.431,
"step": 406
},
{
"epoch": 2.337960746768789,
"grad_norm": 0.36485815048217773,
"learning_rate": 1.4115834819737534e-06,
"loss": 0.3955,
"step": 407
},
{
"epoch": 2.343705122067975,
"grad_norm": 0.39638105034828186,
"learning_rate": 1.3883410302246237e-06,
"loss": 0.4395,
"step": 408
},
{
"epoch": 2.349449497367161,
"grad_norm": 0.3715421259403229,
"learning_rate": 1.3652606322576606e-06,
"loss": 0.402,
"step": 409
},
{
"epoch": 2.3551938726663475,
"grad_norm": 0.3838759660720825,
"learning_rate": 1.3423433236816563e-06,
"loss": 0.4536,
"step": 410
},
{
"epoch": 2.360938247965534,
"grad_norm": 0.38060757517814636,
"learning_rate": 1.3195901327876426e-06,
"loss": 0.4411,
"step": 411
},
{
"epoch": 2.36668262326472,
"grad_norm": 0.35738280415534973,
"learning_rate": 1.2970020805027555e-06,
"loss": 0.4181,
"step": 412
},
{
"epoch": 2.372426998563906,
"grad_norm": 0.38840875029563904,
"learning_rate": 1.2745801803444192e-06,
"loss": 0.4701,
"step": 413
},
{
"epoch": 2.3781713738630925,
"grad_norm": 0.4019489884376526,
"learning_rate": 1.25232543837488e-06,
"loss": 0.4689,
"step": 414
},
{
"epoch": 2.3839157491622784,
"grad_norm": 0.35275644063949585,
"learning_rate": 1.2302388531560515e-06,
"loss": 0.434,
"step": 415
},
{
"epoch": 2.389660124461465,
"grad_norm": 0.34680259227752686,
"learning_rate": 1.2083214157047257e-06,
"loss": 0.4643,
"step": 416
},
{
"epoch": 2.395404499760651,
"grad_norm": 0.3637407422065735,
"learning_rate": 1.186574109448091e-06,
"loss": 0.4502,
"step": 417
},
{
"epoch": 2.401148875059837,
"grad_norm": 0.35043832659721375,
"learning_rate": 1.164997910179615e-06,
"loss": 0.4668,
"step": 418
},
{
"epoch": 2.4068932503590235,
"grad_norm": 0.3933018445968628,
"learning_rate": 1.1435937860152579e-06,
"loss": 0.4543,
"step": 419
},
{
"epoch": 2.41263762565821,
"grad_norm": 0.34680336713790894,
"learning_rate": 1.1223626973500395e-06,
"loss": 0.4085,
"step": 420
},
{
"epoch": 2.4183820009573957,
"grad_norm": 0.36302450299263,
"learning_rate": 1.1013055968149343e-06,
"loss": 0.4495,
"step": 421
},
{
"epoch": 2.424126376256582,
"grad_norm": 0.35072633624076843,
"learning_rate": 1.0804234292341426e-06,
"loss": 0.4707,
"step": 422
},
{
"epoch": 2.4298707515557685,
"grad_norm": 0.3434313237667084,
"learning_rate": 1.0597171315826805e-06,
"loss": 0.4034,
"step": 423
},
{
"epoch": 2.4356151268549544,
"grad_norm": 0.36498257517814636,
"learning_rate": 1.0391876329443534e-06,
"loss": 0.4223,
"step": 424
},
{
"epoch": 2.4413595021541408,
"grad_norm": 0.4297119975090027,
"learning_rate": 1.0188358544700583e-06,
"loss": 0.4474,
"step": 425
},
{
"epoch": 2.447103877453327,
"grad_norm": 0.38329073786735535,
"learning_rate": 9.986627093364542e-07,
"loss": 0.4795,
"step": 426
},
{
"epoch": 2.452848252752513,
"grad_norm": 0.3241833746433258,
"learning_rate": 9.786691027049893e-07,
"loss": 0.395,
"step": 427
},
{
"epoch": 2.4585926280516994,
"grad_norm": 0.41434356570243835,
"learning_rate": 9.588559316812906e-07,
"loss": 0.4971,
"step": 428
},
{
"epoch": 2.4643370033508853,
"grad_norm": 0.341575562953949,
"learning_rate": 9.392240852749007e-07,
"loss": 0.453,
"step": 429
},
{
"epoch": 2.4700813786500717,
"grad_norm": 0.3556068539619446,
"learning_rate": 9.197744443594003e-07,
"loss": 0.4153,
"step": 430
},
{
"epoch": 2.475825753949258,
"grad_norm": 0.37580549716949463,
"learning_rate": 9.005078816328772e-07,
"loss": 0.4842,
"step": 431
},
{
"epoch": 2.4815701292484444,
"grad_norm": 0.3497409224510193,
"learning_rate": 8.814252615787661e-07,
"loss": 0.4325,
"step": 432
},
{
"epoch": 2.4873145045476304,
"grad_norm": 0.3454315960407257,
"learning_rate": 8.625274404270662e-07,
"loss": 0.4739,
"step": 433
},
{
"epoch": 2.4930588798468167,
"grad_norm": 0.36795657873153687,
"learning_rate": 8.438152661159165e-07,
"loss": 0.4202,
"step": 434
},
{
"epoch": 2.4988032551460027,
"grad_norm": 0.39623910188674927,
"learning_rate": 8.252895782535569e-07,
"loss": 0.4704,
"step": 435
},
{
"epoch": 2.504547630445189,
"grad_norm": 0.31534871459007263,
"learning_rate": 8.069512080806441e-07,
"loss": 0.393,
"step": 436
},
{
"epoch": 2.5102920057443754,
"grad_norm": 0.3707200288772583,
"learning_rate": 7.88800978432967e-07,
"loss": 0.465,
"step": 437
},
{
"epoch": 2.5160363810435618,
"grad_norm": 0.43032369017601013,
"learning_rate": 7.708397037045129e-07,
"loss": 0.4779,
"step": 438
},
{
"epoch": 2.5217807563427477,
"grad_norm": 0.3233213722705841,
"learning_rate": 7.530681898109393e-07,
"loss": 0.4145,
"step": 439
},
{
"epoch": 2.527525131641934,
"grad_norm": 0.3708878755569458,
"learning_rate": 7.35487234153402e-07,
"loss": 0.4811,
"step": 440
},
{
"epoch": 2.53326950694112,
"grad_norm": 0.36677300930023193,
"learning_rate": 7.180976255827809e-07,
"loss": 0.4763,
"step": 441
},
{
"epoch": 2.5390138822403063,
"grad_norm": 0.32994070649147034,
"learning_rate": 7.009001443642843e-07,
"loss": 0.3945,
"step": 442
},
{
"epoch": 2.5447582575394927,
"grad_norm": 0.3282773494720459,
"learning_rate": 6.838955621424404e-07,
"loss": 0.3903,
"step": 443
},
{
"epoch": 2.550502632838679,
"grad_norm": 0.3630821108818054,
"learning_rate": 6.67084641906468e-07,
"loss": 0.4785,
"step": 444
},
{
"epoch": 2.556247008137865,
"grad_norm": 0.36368730664253235,
"learning_rate": 6.50468137956049e-07,
"loss": 0.4441,
"step": 445
},
{
"epoch": 2.5619913834370514,
"grad_norm": 0.340044766664505,
"learning_rate": 6.340467958674762e-07,
"loss": 0.427,
"step": 446
},
{
"epoch": 2.5677357587362373,
"grad_norm": 0.37686091661453247,
"learning_rate": 6.178213524602061e-07,
"loss": 0.4678,
"step": 447
},
{
"epoch": 2.5734801340354236,
"grad_norm": 0.36389777064323425,
"learning_rate": 6.017925357637932e-07,
"loss": 0.4194,
"step": 448
},
{
"epoch": 2.57922450933461,
"grad_norm": 0.3633215129375458,
"learning_rate": 5.859610649852249e-07,
"loss": 0.427,
"step": 449
},
{
"epoch": 2.584968884633796,
"grad_norm": 0.3677927553653717,
"learning_rate": 5.703276504766514e-07,
"loss": 0.4147,
"step": 450
},
{
"epoch": 2.5907132599329823,
"grad_norm": 0.32779526710510254,
"learning_rate": 5.548929937035147e-07,
"loss": 0.3927,
"step": 451
},
{
"epoch": 2.5964576352321687,
"grad_norm": 0.343524694442749,
"learning_rate": 5.396577872130676e-07,
"loss": 0.4481,
"step": 452
},
{
"epoch": 2.6022020105313546,
"grad_norm": 0.3486615717411041,
"learning_rate": 5.246227146033089e-07,
"loss": 0.4717,
"step": 453
},
{
"epoch": 2.607946385830541,
"grad_norm": 0.32182297110557556,
"learning_rate": 5.097884504922996e-07,
"loss": 0.4107,
"step": 454
},
{
"epoch": 2.6136907611297273,
"grad_norm": 0.369724303483963,
"learning_rate": 4.951556604879049e-07,
"loss": 0.514,
"step": 455
},
{
"epoch": 2.6194351364289132,
"grad_norm": 0.3562856614589691,
"learning_rate": 4.807250011579168e-07,
"loss": 0.4429,
"step": 456
},
{
"epoch": 2.6251795117280996,
"grad_norm": 0.3348632752895355,
"learning_rate": 4.6649712000060297e-07,
"loss": 0.3847,
"step": 457
},
{
"epoch": 2.6309238870272855,
"grad_norm": 0.3391302525997162,
"learning_rate": 4.5247265541564836e-07,
"loss": 0.4419,
"step": 458
},
{
"epoch": 2.636668262326472,
"grad_norm": 0.35214006900787354,
"learning_rate": 4.386522366755169e-07,
"loss": 0.4397,
"step": 459
},
{
"epoch": 2.6424126376256583,
"grad_norm": 0.37531718611717224,
"learning_rate": 4.250364838972065e-07,
"loss": 0.4695,
"step": 460
},
{
"epoch": 2.6481570129248446,
"grad_norm": 0.3308485746383667,
"learning_rate": 4.116260080144352e-07,
"loss": 0.4502,
"step": 461
},
{
"epoch": 2.6539013882240305,
"grad_norm": 0.3776616156101227,
"learning_rate": 3.98421410750221e-07,
"loss": 0.4571,
"step": 462
},
{
"epoch": 2.659645763523217,
"grad_norm": 0.35296255350112915,
"learning_rate": 3.854232845898859e-07,
"loss": 0.4419,
"step": 463
},
{
"epoch": 2.665390138822403,
"grad_norm": 0.33054107427597046,
"learning_rate": 3.7263221275447125e-07,
"loss": 0.4384,
"step": 464
},
{
"epoch": 2.671134514121589,
"grad_norm": 0.3511356711387634,
"learning_rate": 3.60048769174568e-07,
"loss": 0.4389,
"step": 465
},
{
"epoch": 2.6768788894207756,
"grad_norm": 0.3672967255115509,
"learning_rate": 3.4767351846456744e-07,
"loss": 0.4722,
"step": 466
},
{
"epoch": 2.682623264719962,
"grad_norm": 0.35613754391670227,
"learning_rate": 3.355070158973212e-07,
"loss": 0.4448,
"step": 467
},
{
"epoch": 2.688367640019148,
"grad_norm": 0.33944112062454224,
"learning_rate": 3.235498073792342e-07,
"loss": 0.4444,
"step": 468
},
{
"epoch": 2.694112015318334,
"grad_norm": 0.36371228098869324,
"learning_rate": 3.118024294257621e-07,
"loss": 0.4623,
"step": 469
},
{
"epoch": 2.69985639061752,
"grad_norm": 0.34890905022621155,
"learning_rate": 3.002654091373453e-07,
"loss": 0.4186,
"step": 470
},
{
"epoch": 2.7056007659167065,
"grad_norm": 0.3459916114807129,
"learning_rate": 2.889392641757527e-07,
"loss": 0.4776,
"step": 471
},
{
"epoch": 2.711345141215893,
"grad_norm": 0.3279518485069275,
"learning_rate": 2.778245027408566e-07,
"loss": 0.427,
"step": 472
},
{
"epoch": 2.7170895165150792,
"grad_norm": 0.37685245275497437,
"learning_rate": 2.669216235478295e-07,
"loss": 0.4889,
"step": 473
},
{
"epoch": 2.722833891814265,
"grad_norm": 0.36954060196876526,
"learning_rate": 2.562311158047692e-07,
"loss": 0.4585,
"step": 474
},
{
"epoch": 2.7285782671134515,
"grad_norm": 0.35874509811401367,
"learning_rate": 2.45753459190744e-07,
"loss": 0.4486,
"step": 475
},
{
"epoch": 2.7343226424126374,
"grad_norm": 0.3622611463069916,
"learning_rate": 2.354891238342738e-07,
"loss": 0.4258,
"step": 476
},
{
"epoch": 2.740067017711824,
"grad_norm": 0.3557218015193939,
"learning_rate": 2.254385702922318e-07,
"loss": 0.4552,
"step": 477
},
{
"epoch": 2.74581139301101,
"grad_norm": 0.3304785192012787,
"learning_rate": 2.1560224952918373e-07,
"loss": 0.3928,
"step": 478
},
{
"epoch": 2.751555768310196,
"grad_norm": 0.33311760425567627,
"learning_rate": 2.0598060289714893e-07,
"loss": 0.3941,
"step": 479
},
{
"epoch": 2.7573001436093825,
"grad_norm": 0.3741648197174072,
"learning_rate": 1.9657406211579966e-07,
"loss": 0.4783,
"step": 480
},
{
"epoch": 2.763044518908569,
"grad_norm": 0.32854151725769043,
"learning_rate": 1.8738304925308926e-07,
"loss": 0.4619,
"step": 481
},
{
"epoch": 2.7687888942077548,
"grad_norm": 0.3276991844177246,
"learning_rate": 1.7840797670631572e-07,
"loss": 0.4592,
"step": 482
},
{
"epoch": 2.774533269506941,
"grad_norm": 0.3500697910785675,
"learning_rate": 1.6964924718361364e-07,
"loss": 0.4425,
"step": 483
},
{
"epoch": 2.7802776448061275,
"grad_norm": 0.33605843782424927,
"learning_rate": 1.6110725368589041e-07,
"loss": 0.4183,
"step": 484
},
{
"epoch": 2.7860220201053134,
"grad_norm": 0.36614635586738586,
"learning_rate": 1.5278237948918585e-07,
"loss": 0.4841,
"step": 485
},
{
"epoch": 2.7917663954045,
"grad_norm": 0.3272748589515686,
"learning_rate": 1.4467499812748143e-07,
"loss": 0.443,
"step": 486
},
{
"epoch": 2.7975107707036857,
"grad_norm": 0.3265168368816376,
"learning_rate": 1.3678547337593494e-07,
"loss": 0.4546,
"step": 487
},
{
"epoch": 2.803255146002872,
"grad_norm": 0.34403395652770996,
"learning_rate": 1.2911415923456017e-07,
"loss": 0.4488,
"step": 488
},
{
"epoch": 2.8089995213020584,
"grad_norm": 0.35020750761032104,
"learning_rate": 1.2166139991234227e-07,
"loss": 0.4165,
"step": 489
},
{
"epoch": 2.814743896601245,
"grad_norm": 0.35353976488113403,
"learning_rate": 1.1442752981179527e-07,
"loss": 0.4639,
"step": 490
},
{
"epoch": 2.8204882719004307,
"grad_norm": 0.34721872210502625,
"learning_rate": 1.0741287351395402e-07,
"loss": 0.4842,
"step": 491
},
{
"epoch": 2.826232647199617,
"grad_norm": 0.339238703250885,
"learning_rate": 1.0061774576381411e-07,
"loss": 0.4441,
"step": 492
},
{
"epoch": 2.831977022498803,
"grad_norm": 0.3654729127883911,
"learning_rate": 9.404245145620717e-08,
"loss": 0.4833,
"step": 493
},
{
"epoch": 2.8377213977979894,
"grad_norm": 0.3376854956150055,
"learning_rate": 8.768728562211948e-08,
"loss": 0.3834,
"step": 494
},
{
"epoch": 2.8434657730971757,
"grad_norm": 0.345510333776474,
"learning_rate": 8.155253341545655e-08,
"loss": 0.4836,
"step": 495
},
{
"epoch": 2.849210148396362,
"grad_norm": 0.36857500672340393,
"learning_rate": 7.563847010024716e-08,
"loss": 0.4415,
"step": 496
},
{
"epoch": 2.854954523695548,
"grad_norm": 0.3405042886734009,
"learning_rate": 6.994536103829164e-08,
"loss": 0.4478,
"step": 497
},
{
"epoch": 2.8606988989947344,
"grad_norm": 0.3312218189239502,
"learning_rate": 6.447346167725688e-08,
"loss": 0.4068,
"step": 498
},
{
"epoch": 2.8664432742939203,
"grad_norm": 0.34017452597618103,
"learning_rate": 5.9223017539213335e-08,
"loss": 0.4854,
"step": 499
},
{
"epoch": 2.8721876495931067,
"grad_norm": 0.30581504106521606,
"learning_rate": 5.4194264209617705e-08,
"loss": 0.412,
"step": 500
},
{
"epoch": 2.877932024892293,
"grad_norm": 0.35008949041366577,
"learning_rate": 4.9387427326745287e-08,
"loss": 0.4657,
"step": 501
},
{
"epoch": 2.8836764001914794,
"grad_norm": 0.35655391216278076,
"learning_rate": 4.4802722571561374e-08,
"loss": 0.4724,
"step": 502
},
{
"epoch": 2.8894207754906653,
"grad_norm": 0.36218518018722534,
"learning_rate": 4.044035565804793e-08,
"loss": 0.4222,
"step": 503
},
{
"epoch": 2.8951651507898517,
"grad_norm": 0.34190085530281067,
"learning_rate": 3.6300522323969855e-08,
"loss": 0.3833,
"step": 504
},
{
"epoch": 2.9009095260890376,
"grad_norm": 0.3299616575241089,
"learning_rate": 3.2383408322095856e-08,
"loss": 0.4841,
"step": 505
},
{
"epoch": 2.906653901388224,
"grad_norm": 0.32240092754364014,
"learning_rate": 2.8689189411859607e-08,
"loss": 0.4119,
"step": 506
},
{
"epoch": 2.9123982766874104,
"grad_norm": 0.36698001623153687,
"learning_rate": 2.5218031351478268e-08,
"loss": 0.5207,
"step": 507
},
{
"epoch": 2.9181426519865967,
"grad_norm": 0.36925238370895386,
"learning_rate": 2.1970089890509527e-08,
"loss": 0.4163,
"step": 508
},
{
"epoch": 2.9238870272857826,
"grad_norm": 0.368062287569046,
"learning_rate": 1.8945510762868325e-08,
"loss": 0.4461,
"step": 509
},
{
"epoch": 2.929631402584969,
"grad_norm": 0.33701467514038086,
"learning_rate": 1.614442968028429e-08,
"loss": 0.455,
"step": 510
},
{
"epoch": 2.935375777884155,
"grad_norm": 0.33833688497543335,
"learning_rate": 1.3566972326214956e-08,
"loss": 0.4412,
"step": 511
},
{
"epoch": 2.9411201531833413,
"grad_norm": 0.3553823232650757,
"learning_rate": 1.1213254350202486e-08,
"loss": 0.4247,
"step": 512
},
{
"epoch": 2.9468645284825277,
"grad_norm": 0.3272307813167572,
"learning_rate": 9.083381362690603e-09,
"loss": 0.4326,
"step": 513
},
{
"epoch": 2.9526089037817136,
"grad_norm": 0.34332549571990967,
"learning_rate": 7.177448930279496e-09,
"loss": 0.4514,
"step": 514
},
{
"epoch": 2.9583532790809,
"grad_norm": 0.35976776480674744,
"learning_rate": 5.495542571443135e-09,
"loss": 0.4163,
"step": 515
},
{
"epoch": 2.9640976543800863,
"grad_norm": 0.3729633390903473,
"learning_rate": 4.037737752686788e-09,
"loss": 0.441,
"step": 516
},
{
"epoch": 2.9698420296792722,
"grad_norm": 0.34086230397224426,
"learning_rate": 2.8040998851674996e-09,
"loss": 0.4451,
"step": 517
},
{
"epoch": 2.9755864049784586,
"grad_norm": 0.33816850185394287,
"learning_rate": 1.7946843217514498e-09,
"loss": 0.4208,
"step": 518
},
{
"epoch": 2.981330780277645,
"grad_norm": 0.3527694344520569,
"learning_rate": 1.009536354537044e-09,
"loss": 0.393,
"step": 519
},
{
"epoch": 2.987075155576831,
"grad_norm": 0.36905810236930847,
"learning_rate": 4.486912128182086e-10,
"loss": 0.5064,
"step": 520
},
{
"epoch": 2.9928195308760173,
"grad_norm": 0.3305312395095825,
"learning_rate": 1.1217406150676457e-10,
"loss": 0.4423,
"step": 521
},
{
"epoch": 2.998563906175203,
"grad_norm": 0.33031007647514343,
"learning_rate": 0.0,
"loss": 0.4217,
"step": 522
},
{
"epoch": 2.998563906175203,
"step": 522,
"total_flos": 516591112486912.0,
"train_loss": 0.5103290048481404,
"train_runtime": 26646.5751,
"train_samples_per_second": 1.881,
"train_steps_per_second": 0.02
}
],
"logging_steps": 1.0,
"max_steps": 522,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 516591112486912.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}