Cartinoe5930's picture
Upload folder using huggingface_hub
2632a02 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.8042125418860699,
"eval_steps": 500,
"global_step": 420,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0019147917663954045,
"grad_norm": 1.17178213596344,
"learning_rate": 0.0,
"loss": 0.5878,
"step": 1
},
{
"epoch": 0.003829583532790809,
"grad_norm": 1.167578101158142,
"learning_rate": 3.846153846153847e-07,
"loss": 0.5815,
"step": 2
},
{
"epoch": 0.0057443752991862135,
"grad_norm": 1.2005791664123535,
"learning_rate": 7.692307692307694e-07,
"loss": 0.6327,
"step": 3
},
{
"epoch": 0.007659167065581618,
"grad_norm": 1.1366013288497925,
"learning_rate": 1.153846153846154e-06,
"loss": 0.5869,
"step": 4
},
{
"epoch": 0.009573958831977022,
"grad_norm": 1.0155231952667236,
"learning_rate": 1.5384615384615387e-06,
"loss": 0.5233,
"step": 5
},
{
"epoch": 0.011488750598372427,
"grad_norm": 1.2220700979232788,
"learning_rate": 1.9230769230769234e-06,
"loss": 0.5699,
"step": 6
},
{
"epoch": 0.013403542364767831,
"grad_norm": 1.1095085144042969,
"learning_rate": 2.307692307692308e-06,
"loss": 0.5378,
"step": 7
},
{
"epoch": 0.015318334131163236,
"grad_norm": 1.0518015623092651,
"learning_rate": 2.6923076923076923e-06,
"loss": 0.5432,
"step": 8
},
{
"epoch": 0.01723312589755864,
"grad_norm": 1.0267938375473022,
"learning_rate": 3.0769230769230774e-06,
"loss": 0.5702,
"step": 9
},
{
"epoch": 0.019147917663954045,
"grad_norm": 1.0278018712997437,
"learning_rate": 3.4615384615384617e-06,
"loss": 0.5411,
"step": 10
},
{
"epoch": 0.02106270943034945,
"grad_norm": 0.9818478226661682,
"learning_rate": 3.846153846153847e-06,
"loss": 0.5214,
"step": 11
},
{
"epoch": 0.022977501196744854,
"grad_norm": 0.9112760424613953,
"learning_rate": 4.230769230769231e-06,
"loss": 0.5235,
"step": 12
},
{
"epoch": 0.02489229296314026,
"grad_norm": 1.1367770433425903,
"learning_rate": 4.615384615384616e-06,
"loss": 0.5689,
"step": 13
},
{
"epoch": 0.026807084729535663,
"grad_norm": 0.887293815612793,
"learning_rate": 5e-06,
"loss": 0.5062,
"step": 14
},
{
"epoch": 0.028721876495931067,
"grad_norm": 0.8234561085700989,
"learning_rate": 5.384615384615385e-06,
"loss": 0.4938,
"step": 15
},
{
"epoch": 0.030636668262326472,
"grad_norm": 0.8301621079444885,
"learning_rate": 5.769230769230769e-06,
"loss": 0.4955,
"step": 16
},
{
"epoch": 0.032551460028721876,
"grad_norm": 0.7201998829841614,
"learning_rate": 6.153846153846155e-06,
"loss": 0.4481,
"step": 17
},
{
"epoch": 0.03446625179511728,
"grad_norm": 0.7492491602897644,
"learning_rate": 6.538461538461539e-06,
"loss": 0.4924,
"step": 18
},
{
"epoch": 0.036381043561512685,
"grad_norm": 0.8051350712776184,
"learning_rate": 6.923076923076923e-06,
"loss": 0.5109,
"step": 19
},
{
"epoch": 0.03829583532790809,
"grad_norm": 0.6143934726715088,
"learning_rate": 7.307692307692308e-06,
"loss": 0.4846,
"step": 20
},
{
"epoch": 0.040210627094303494,
"grad_norm": 0.7769045233726501,
"learning_rate": 7.692307692307694e-06,
"loss": 0.4761,
"step": 21
},
{
"epoch": 0.0421254188606989,
"grad_norm": 0.5961856842041016,
"learning_rate": 8.076923076923077e-06,
"loss": 0.467,
"step": 22
},
{
"epoch": 0.0440402106270943,
"grad_norm": 0.683874785900116,
"learning_rate": 8.461538461538462e-06,
"loss": 0.4847,
"step": 23
},
{
"epoch": 0.04595500239348971,
"grad_norm": 0.5849653482437134,
"learning_rate": 8.846153846153847e-06,
"loss": 0.4514,
"step": 24
},
{
"epoch": 0.04786979415988511,
"grad_norm": 0.6217619180679321,
"learning_rate": 9.230769230769232e-06,
"loss": 0.4733,
"step": 25
},
{
"epoch": 0.04978458592628052,
"grad_norm": 0.5796915888786316,
"learning_rate": 9.615384615384616e-06,
"loss": 0.4735,
"step": 26
},
{
"epoch": 0.05169937769267592,
"grad_norm": 0.5245983600616455,
"learning_rate": 1e-05,
"loss": 0.4398,
"step": 27
},
{
"epoch": 0.053614169459071326,
"grad_norm": 0.6732852458953857,
"learning_rate": 9.999899706000774e-06,
"loss": 0.4907,
"step": 28
},
{
"epoch": 0.05552896122546673,
"grad_norm": 0.5512049794197083,
"learning_rate": 9.999598828026644e-06,
"loss": 0.4723,
"step": 29
},
{
"epoch": 0.057443752991862135,
"grad_norm": 0.6110374331474304,
"learning_rate": 9.999097378148116e-06,
"loss": 0.4414,
"step": 30
},
{
"epoch": 0.05935854475825754,
"grad_norm": 0.5515976548194885,
"learning_rate": 9.998395376482152e-06,
"loss": 0.4314,
"step": 31
},
{
"epoch": 0.061273336524652944,
"grad_norm": 0.5074656009674072,
"learning_rate": 9.99749285119138e-06,
"loss": 0.4149,
"step": 32
},
{
"epoch": 0.06318812829104835,
"grad_norm": 0.6015526056289673,
"learning_rate": 9.996389838482942e-06,
"loss": 0.517,
"step": 33
},
{
"epoch": 0.06510292005744375,
"grad_norm": 0.5998480916023254,
"learning_rate": 9.995086382607064e-06,
"loss": 0.4695,
"step": 34
},
{
"epoch": 0.06701771182383916,
"grad_norm": 0.5995810627937317,
"learning_rate": 9.993582535855265e-06,
"loss": 0.4453,
"step": 35
},
{
"epoch": 0.06893250359023456,
"grad_norm": 0.5346629023551941,
"learning_rate": 9.991878358558267e-06,
"loss": 0.4667,
"step": 36
},
{
"epoch": 0.07084729535662997,
"grad_norm": 0.5550440549850464,
"learning_rate": 9.989973919083576e-06,
"loss": 0.4488,
"step": 37
},
{
"epoch": 0.07276208712302537,
"grad_norm": 0.5914222598075867,
"learning_rate": 9.987869293832727e-06,
"loss": 0.4473,
"step": 38
},
{
"epoch": 0.07467687888942078,
"grad_norm": 0.46610352396965027,
"learning_rate": 9.985564567238237e-06,
"loss": 0.4265,
"step": 39
},
{
"epoch": 0.07659167065581618,
"grad_norm": 0.6092692017555237,
"learning_rate": 9.983059831760205e-06,
"loss": 0.469,
"step": 40
},
{
"epoch": 0.07850646242221158,
"grad_norm": 0.48001110553741455,
"learning_rate": 9.980355187882606e-06,
"loss": 0.4219,
"step": 41
},
{
"epoch": 0.08042125418860699,
"grad_norm": 0.4854871332645416,
"learning_rate": 9.977450744109258e-06,
"loss": 0.3991,
"step": 42
},
{
"epoch": 0.0823360459550024,
"grad_norm": 0.47407639026641846,
"learning_rate": 9.974346616959476e-06,
"loss": 0.4177,
"step": 43
},
{
"epoch": 0.0842508377213978,
"grad_norm": 0.599115788936615,
"learning_rate": 9.97104293096339e-06,
"loss": 0.4401,
"step": 44
},
{
"epoch": 0.0861656294877932,
"grad_norm": 0.5324896574020386,
"learning_rate": 9.967539818656953e-06,
"loss": 0.4265,
"step": 45
},
{
"epoch": 0.0880804212541886,
"grad_norm": 0.7636860609054565,
"learning_rate": 9.96383742057662e-06,
"loss": 0.4957,
"step": 46
},
{
"epoch": 0.08999521302058401,
"grad_norm": 0.4660010039806366,
"learning_rate": 9.959935885253715e-06,
"loss": 0.4236,
"step": 47
},
{
"epoch": 0.09191000478697942,
"grad_norm": 0.4723781645298004,
"learning_rate": 9.955835369208475e-06,
"loss": 0.405,
"step": 48
},
{
"epoch": 0.09382479655337482,
"grad_norm": 0.473716139793396,
"learning_rate": 9.951536036943753e-06,
"loss": 0.4054,
"step": 49
},
{
"epoch": 0.09573958831977022,
"grad_norm": 0.8201298713684082,
"learning_rate": 9.94703806093845e-06,
"loss": 0.4437,
"step": 50
},
{
"epoch": 0.09765438008616563,
"grad_norm": 0.5334436297416687,
"learning_rate": 9.942341621640558e-06,
"loss": 0.4112,
"step": 51
},
{
"epoch": 0.09956917185256103,
"grad_norm": 0.4933352768421173,
"learning_rate": 9.937446907459954e-06,
"loss": 0.4303,
"step": 52
},
{
"epoch": 0.10148396361895644,
"grad_norm": 0.44361305236816406,
"learning_rate": 9.932354114760819e-06,
"loss": 0.4039,
"step": 53
},
{
"epoch": 0.10339875538535184,
"grad_norm": 0.5590924024581909,
"learning_rate": 9.92706344785377e-06,
"loss": 0.51,
"step": 54
},
{
"epoch": 0.10531354715174725,
"grad_norm": 0.3887651264667511,
"learning_rate": 9.921575118987672e-06,
"loss": 0.3709,
"step": 55
},
{
"epoch": 0.10722833891814265,
"grad_norm": 0.43217676877975464,
"learning_rate": 9.915889348341098e-06,
"loss": 0.4107,
"step": 56
},
{
"epoch": 0.10914313068453806,
"grad_norm": 0.46679773926734924,
"learning_rate": 9.910006364013522e-06,
"loss": 0.3832,
"step": 57
},
{
"epoch": 0.11105792245093346,
"grad_norm": 0.5174672603607178,
"learning_rate": 9.903926402016153e-06,
"loss": 0.397,
"step": 58
},
{
"epoch": 0.11297271421732887,
"grad_norm": 0.5932629704475403,
"learning_rate": 9.897649706262474e-06,
"loss": 0.4584,
"step": 59
},
{
"epoch": 0.11488750598372427,
"grad_norm": 0.5002132058143616,
"learning_rate": 9.891176528558451e-06,
"loss": 0.4115,
"step": 60
},
{
"epoch": 0.11680229775011967,
"grad_norm": 0.44753795862197876,
"learning_rate": 9.884507128592435e-06,
"loss": 0.417,
"step": 61
},
{
"epoch": 0.11871708951651508,
"grad_norm": 0.43308743834495544,
"learning_rate": 9.877641773924748e-06,
"loss": 0.3872,
"step": 62
},
{
"epoch": 0.12063188128291048,
"grad_norm": 0.44310009479522705,
"learning_rate": 9.870580739976936e-06,
"loss": 0.3877,
"step": 63
},
{
"epoch": 0.12254667304930589,
"grad_norm": 0.4362034797668457,
"learning_rate": 9.863324310020735e-06,
"loss": 0.3957,
"step": 64
},
{
"epoch": 0.12446146481570129,
"grad_norm": 0.46891993284225464,
"learning_rate": 9.855872775166696e-06,
"loss": 0.427,
"step": 65
},
{
"epoch": 0.1263762565820967,
"grad_norm": 0.47662439942359924,
"learning_rate": 9.848226434352513e-06,
"loss": 0.4041,
"step": 66
},
{
"epoch": 0.12829104834849211,
"grad_norm": 0.5884003043174744,
"learning_rate": 9.840385594331022e-06,
"loss": 0.4487,
"step": 67
},
{
"epoch": 0.1302058401148875,
"grad_norm": 0.5149067640304565,
"learning_rate": 9.83235056965791e-06,
"loss": 0.4567,
"step": 68
},
{
"epoch": 0.13212063188128292,
"grad_norm": 0.4653552174568176,
"learning_rate": 9.824121682679072e-06,
"loss": 0.4024,
"step": 69
},
{
"epoch": 0.13403542364767831,
"grad_norm": 0.47937148809432983,
"learning_rate": 9.815699263517712e-06,
"loss": 0.414,
"step": 70
},
{
"epoch": 0.13595021541407373,
"grad_norm": 0.4588964581489563,
"learning_rate": 9.807083650061063e-06,
"loss": 0.4199,
"step": 71
},
{
"epoch": 0.13786500718046912,
"grad_norm": 0.5369502902030945,
"learning_rate": 9.798275187946859e-06,
"loss": 0.4213,
"step": 72
},
{
"epoch": 0.13977979894686454,
"grad_norm": 0.48553192615509033,
"learning_rate": 9.789274230549456e-06,
"loss": 0.3694,
"step": 73
},
{
"epoch": 0.14169459071325993,
"grad_norm": 0.3765530586242676,
"learning_rate": 9.780081138965663e-06,
"loss": 0.3445,
"step": 74
},
{
"epoch": 0.14360938247965535,
"grad_norm": 0.48994067311286926,
"learning_rate": 9.770696282000245e-06,
"loss": 0.4202,
"step": 75
},
{
"epoch": 0.14552417424605074,
"grad_norm": 0.45693519711494446,
"learning_rate": 9.761120036151138e-06,
"loss": 0.384,
"step": 76
},
{
"epoch": 0.14743896601244616,
"grad_norm": 0.4083925485610962,
"learning_rate": 9.751352785594337e-06,
"loss": 0.3618,
"step": 77
},
{
"epoch": 0.14935375777884155,
"grad_norm": 0.5029312372207642,
"learning_rate": 9.741394922168495e-06,
"loss": 0.4446,
"step": 78
},
{
"epoch": 0.15126854954523697,
"grad_norm": 0.4641478359699249,
"learning_rate": 9.731246845359187e-06,
"loss": 0.3837,
"step": 79
},
{
"epoch": 0.15318334131163236,
"grad_norm": 0.42773452401161194,
"learning_rate": 9.720908962282893e-06,
"loss": 0.3747,
"step": 80
},
{
"epoch": 0.15509813307802778,
"grad_norm": 0.45671483874320984,
"learning_rate": 9.710381687670675e-06,
"loss": 0.396,
"step": 81
},
{
"epoch": 0.15701292484442317,
"grad_norm": 0.518532931804657,
"learning_rate": 9.699665443851518e-06,
"loss": 0.4207,
"step": 82
},
{
"epoch": 0.1589277166108186,
"grad_norm": 0.42763078212738037,
"learning_rate": 9.688760660735403e-06,
"loss": 0.3617,
"step": 83
},
{
"epoch": 0.16084250837721398,
"grad_norm": 0.4328898787498474,
"learning_rate": 9.677667775796052e-06,
"loss": 0.3612,
"step": 84
},
{
"epoch": 0.1627573001436094,
"grad_norm": 0.5515812635421753,
"learning_rate": 9.666387234053385e-06,
"loss": 0.4151,
"step": 85
},
{
"epoch": 0.1646720919100048,
"grad_norm": 0.4455530047416687,
"learning_rate": 9.654919488055656e-06,
"loss": 0.4048,
"step": 86
},
{
"epoch": 0.1665868836764002,
"grad_norm": 0.43045058846473694,
"learning_rate": 9.643264997861312e-06,
"loss": 0.3769,
"step": 87
},
{
"epoch": 0.1685016754427956,
"grad_norm": 0.44380709528923035,
"learning_rate": 9.631424231020523e-06,
"loss": 0.3963,
"step": 88
},
{
"epoch": 0.170416467209191,
"grad_norm": 0.5282372832298279,
"learning_rate": 9.619397662556434e-06,
"loss": 0.418,
"step": 89
},
{
"epoch": 0.1723312589755864,
"grad_norm": 0.5417309403419495,
"learning_rate": 9.607185774946106e-06,
"loss": 0.483,
"step": 90
},
{
"epoch": 0.17424605074198182,
"grad_norm": 0.48023760318756104,
"learning_rate": 9.594789058101154e-06,
"loss": 0.4041,
"step": 91
},
{
"epoch": 0.1761608425083772,
"grad_norm": 0.43183737993240356,
"learning_rate": 9.582208009348104e-06,
"loss": 0.3724,
"step": 92
},
{
"epoch": 0.17807563427477263,
"grad_norm": 0.4684963822364807,
"learning_rate": 9.569443133408434e-06,
"loss": 0.4199,
"step": 93
},
{
"epoch": 0.17999042604116802,
"grad_norm": 0.5076436400413513,
"learning_rate": 9.556494942378328e-06,
"loss": 0.4138,
"step": 94
},
{
"epoch": 0.18190521780756344,
"grad_norm": 0.527299165725708,
"learning_rate": 9.543363955708124e-06,
"loss": 0.4089,
"step": 95
},
{
"epoch": 0.18382000957395883,
"grad_norm": 0.4080921411514282,
"learning_rate": 9.530050700181499e-06,
"loss": 0.3337,
"step": 96
},
{
"epoch": 0.18573480134035425,
"grad_norm": 0.42223554849624634,
"learning_rate": 9.5165557098943e-06,
"loss": 0.3758,
"step": 97
},
{
"epoch": 0.18764959310674964,
"grad_norm": 0.4336399435997009,
"learning_rate": 9.502879526233151e-06,
"loss": 0.3664,
"step": 98
},
{
"epoch": 0.18956438487314506,
"grad_norm": 0.43095213174819946,
"learning_rate": 9.48902269785371e-06,
"loss": 0.379,
"step": 99
},
{
"epoch": 0.19147917663954045,
"grad_norm": 0.4262320399284363,
"learning_rate": 9.47498578065867e-06,
"loss": 0.3666,
"step": 100
},
{
"epoch": 0.19339396840593587,
"grad_norm": 0.4407063126564026,
"learning_rate": 9.460769337775461e-06,
"loss": 0.3857,
"step": 101
},
{
"epoch": 0.19530876017233126,
"grad_norm": 0.39989781379699707,
"learning_rate": 9.446373939533642e-06,
"loss": 0.3568,
"step": 102
},
{
"epoch": 0.19722355193872668,
"grad_norm": 0.43383434414863586,
"learning_rate": 9.431800163442043e-06,
"loss": 0.374,
"step": 103
},
{
"epoch": 0.19913834370512207,
"grad_norm": 0.5074713230133057,
"learning_rate": 9.417048594165572e-06,
"loss": 0.4137,
"step": 104
},
{
"epoch": 0.20105313547151749,
"grad_norm": 0.45141083002090454,
"learning_rate": 9.402119823501787e-06,
"loss": 0.3977,
"step": 105
},
{
"epoch": 0.20296792723791288,
"grad_norm": 0.44136589765548706,
"learning_rate": 9.387014450357128e-06,
"loss": 0.3808,
"step": 106
},
{
"epoch": 0.2048827190043083,
"grad_norm": 0.6126969456672668,
"learning_rate": 9.371733080722911e-06,
"loss": 0.4527,
"step": 107
},
{
"epoch": 0.20679751077070369,
"grad_norm": 0.48484617471694946,
"learning_rate": 9.356276327651006e-06,
"loss": 0.4032,
"step": 108
},
{
"epoch": 0.2087123025370991,
"grad_norm": 0.4131053388118744,
"learning_rate": 9.340644811229243e-06,
"loss": 0.3594,
"step": 109
},
{
"epoch": 0.2106270943034945,
"grad_norm": 0.3629656434059143,
"learning_rate": 9.324839158556542e-06,
"loss": 0.3337,
"step": 110
},
{
"epoch": 0.2125418860698899,
"grad_norm": 0.4247240424156189,
"learning_rate": 9.308860003717748e-06,
"loss": 0.3809,
"step": 111
},
{
"epoch": 0.2144566778362853,
"grad_norm": 0.4432498812675476,
"learning_rate": 9.292707987758202e-06,
"loss": 0.398,
"step": 112
},
{
"epoch": 0.21637146960268072,
"grad_norm": 0.45140188932418823,
"learning_rate": 9.27638375865801e-06,
"loss": 0.3938,
"step": 113
},
{
"epoch": 0.2182862613690761,
"grad_norm": 0.5100187063217163,
"learning_rate": 9.259887971306064e-06,
"loss": 0.4427,
"step": 114
},
{
"epoch": 0.22020105313547153,
"grad_norm": 0.5069835782051086,
"learning_rate": 9.243221287473755e-06,
"loss": 0.4131,
"step": 115
},
{
"epoch": 0.22211584490186692,
"grad_norm": 0.4621468782424927,
"learning_rate": 9.226384375788435e-06,
"loss": 0.3788,
"step": 116
},
{
"epoch": 0.22403063666826234,
"grad_norm": 0.4977872371673584,
"learning_rate": 9.209377911706585e-06,
"loss": 0.3719,
"step": 117
},
{
"epoch": 0.22594542843465773,
"grad_norm": 0.4791455566883087,
"learning_rate": 9.192202577486725e-06,
"loss": 0.3562,
"step": 118
},
{
"epoch": 0.22786022020105315,
"grad_norm": 0.4171106219291687,
"learning_rate": 9.174859062162037e-06,
"loss": 0.3568,
"step": 119
},
{
"epoch": 0.22977501196744854,
"grad_norm": 0.4501950740814209,
"learning_rate": 9.157348061512728e-06,
"loss": 0.4015,
"step": 120
},
{
"epoch": 0.23168980373384396,
"grad_norm": 0.4359966814517975,
"learning_rate": 9.139670278038109e-06,
"loss": 0.4014,
"step": 121
},
{
"epoch": 0.23360459550023935,
"grad_norm": 0.3942509889602661,
"learning_rate": 9.121826420928421e-06,
"loss": 0.3328,
"step": 122
},
{
"epoch": 0.23551938726663477,
"grad_norm": 0.4340513050556183,
"learning_rate": 9.103817206036383e-06,
"loss": 0.3628,
"step": 123
},
{
"epoch": 0.23743417903303016,
"grad_norm": 0.44619980454444885,
"learning_rate": 9.085643355848468e-06,
"loss": 0.4003,
"step": 124
},
{
"epoch": 0.23934897079942558,
"grad_norm": 0.43487605452537537,
"learning_rate": 9.06730559945592e-06,
"loss": 0.3558,
"step": 125
},
{
"epoch": 0.24126376256582097,
"grad_norm": 0.4643803834915161,
"learning_rate": 9.048804672525513e-06,
"loss": 0.375,
"step": 126
},
{
"epoch": 0.24317855433221638,
"grad_norm": 0.44975554943084717,
"learning_rate": 9.030141317270026e-06,
"loss": 0.3602,
"step": 127
},
{
"epoch": 0.24509334609861178,
"grad_norm": 0.45412832498550415,
"learning_rate": 9.011316282418474e-06,
"loss": 0.3724,
"step": 128
},
{
"epoch": 0.2470081378650072,
"grad_norm": 0.4402771592140198,
"learning_rate": 8.992330323186069e-06,
"loss": 0.3945,
"step": 129
},
{
"epoch": 0.24892292963140258,
"grad_norm": 0.46597278118133545,
"learning_rate": 8.973184201243922e-06,
"loss": 0.4407,
"step": 130
},
{
"epoch": 0.250837721397798,
"grad_norm": 0.44561654329299927,
"learning_rate": 8.953878684688492e-06,
"loss": 0.3814,
"step": 131
},
{
"epoch": 0.2527525131641934,
"grad_norm": 0.4216248095035553,
"learning_rate": 8.934414548010764e-06,
"loss": 0.3647,
"step": 132
},
{
"epoch": 0.2546673049305888,
"grad_norm": 0.4280121326446533,
"learning_rate": 8.914792572065178e-06,
"loss": 0.3733,
"step": 133
},
{
"epoch": 0.25658209669698423,
"grad_norm": 0.5127595067024231,
"learning_rate": 8.89501354403831e-06,
"loss": 0.4195,
"step": 134
},
{
"epoch": 0.2584968884633796,
"grad_norm": 0.5074846744537354,
"learning_rate": 8.875078257417294e-06,
"loss": 0.4204,
"step": 135
},
{
"epoch": 0.260411680229775,
"grad_norm": 0.4372941553592682,
"learning_rate": 8.854987511957974e-06,
"loss": 0.3924,
"step": 136
},
{
"epoch": 0.26232647199617043,
"grad_norm": 0.482693076133728,
"learning_rate": 8.834742113652835e-06,
"loss": 0.3989,
"step": 137
},
{
"epoch": 0.26424126376256585,
"grad_norm": 0.40820303559303284,
"learning_rate": 8.81434287469866e-06,
"loss": 0.3339,
"step": 138
},
{
"epoch": 0.2661560555289612,
"grad_norm": 0.5613487362861633,
"learning_rate": 8.793790613463956e-06,
"loss": 0.3689,
"step": 139
},
{
"epoch": 0.26807084729535663,
"grad_norm": 0.4296424686908722,
"learning_rate": 8.773086154456106e-06,
"loss": 0.3781,
"step": 140
},
{
"epoch": 0.26998563906175205,
"grad_norm": 0.4859068989753723,
"learning_rate": 8.752230328288314e-06,
"loss": 0.4299,
"step": 141
},
{
"epoch": 0.27190043082814747,
"grad_norm": 0.4237840175628662,
"learning_rate": 8.731223971646261e-06,
"loss": 0.3507,
"step": 142
},
{
"epoch": 0.27381522259454283,
"grad_norm": 0.40383824706077576,
"learning_rate": 8.710067927254555e-06,
"loss": 0.344,
"step": 143
},
{
"epoch": 0.27573001436093825,
"grad_norm": 0.40671247243881226,
"learning_rate": 8.688763043842916e-06,
"loss": 0.361,
"step": 144
},
{
"epoch": 0.27764480612733367,
"grad_norm": 0.4422585666179657,
"learning_rate": 8.66731017611213e-06,
"loss": 0.3884,
"step": 145
},
{
"epoch": 0.2795595978937291,
"grad_norm": 0.4300481081008911,
"learning_rate": 8.645710184699756e-06,
"loss": 0.3804,
"step": 146
},
{
"epoch": 0.28147438966012445,
"grad_norm": 0.43995732069015503,
"learning_rate": 8.6239639361456e-06,
"loss": 0.3715,
"step": 147
},
{
"epoch": 0.28338918142651986,
"grad_norm": 0.39543983340263367,
"learning_rate": 8.602072302856961e-06,
"loss": 0.3664,
"step": 148
},
{
"epoch": 0.2853039731929153,
"grad_norm": 0.4740181863307953,
"learning_rate": 8.580036163073615e-06,
"loss": 0.3998,
"step": 149
},
{
"epoch": 0.2872187649593107,
"grad_norm": 0.4195634126663208,
"learning_rate": 8.5578564008326e-06,
"loss": 0.3514,
"step": 150
},
{
"epoch": 0.28913355672570606,
"grad_norm": 0.4382091164588928,
"learning_rate": 8.535533905932739e-06,
"loss": 0.3645,
"step": 151
},
{
"epoch": 0.2910483484921015,
"grad_norm": 0.47153332829475403,
"learning_rate": 8.513069573898944e-06,
"loss": 0.3709,
"step": 152
},
{
"epoch": 0.2929631402584969,
"grad_norm": 0.428423672914505,
"learning_rate": 8.490464305946296e-06,
"loss": 0.377,
"step": 153
},
{
"epoch": 0.2948779320248923,
"grad_norm": 0.4477589428424835,
"learning_rate": 8.467719008943886e-06,
"loss": 0.3696,
"step": 154
},
{
"epoch": 0.2967927237912877,
"grad_norm": 0.44177618622779846,
"learning_rate": 8.444834595378434e-06,
"loss": 0.3793,
"step": 155
},
{
"epoch": 0.2987075155576831,
"grad_norm": 0.40658193826675415,
"learning_rate": 8.421811983317682e-06,
"loss": 0.3564,
"step": 156
},
{
"epoch": 0.3006223073240785,
"grad_norm": 0.40630021691322327,
"learning_rate": 8.398652096373566e-06,
"loss": 0.3743,
"step": 157
},
{
"epoch": 0.30253709909047394,
"grad_norm": 0.419085294008255,
"learning_rate": 8.375355863665155e-06,
"loss": 0.3609,
"step": 158
},
{
"epoch": 0.3044518908568693,
"grad_norm": 0.4741552174091339,
"learning_rate": 8.351924219781393e-06,
"loss": 0.4016,
"step": 159
},
{
"epoch": 0.3063666826232647,
"grad_norm": 0.445846825838089,
"learning_rate": 8.328358104743588e-06,
"loss": 0.3714,
"step": 160
},
{
"epoch": 0.30828147438966014,
"grad_norm": 0.4190356135368347,
"learning_rate": 8.304658463967705e-06,
"loss": 0.3718,
"step": 161
},
{
"epoch": 0.31019626615605556,
"grad_norm": 0.4018896222114563,
"learning_rate": 8.28082624822645e-06,
"loss": 0.3352,
"step": 162
},
{
"epoch": 0.3121110579224509,
"grad_norm": 0.41169753670692444,
"learning_rate": 8.256862413611113e-06,
"loss": 0.3439,
"step": 163
},
{
"epoch": 0.31402584968884634,
"grad_norm": 0.4585376977920532,
"learning_rate": 8.232767921493216e-06,
"loss": 0.3901,
"step": 164
},
{
"epoch": 0.31594064145524176,
"grad_norm": 0.4590586721897125,
"learning_rate": 8.20854373848595e-06,
"loss": 0.4042,
"step": 165
},
{
"epoch": 0.3178554332216372,
"grad_norm": 0.42637982964515686,
"learning_rate": 8.184190836405394e-06,
"loss": 0.3295,
"step": 166
},
{
"epoch": 0.31977022498803254,
"grad_norm": 0.45910605788230896,
"learning_rate": 8.15971019223152e-06,
"loss": 0.3759,
"step": 167
},
{
"epoch": 0.32168501675442795,
"grad_norm": 0.3774631917476654,
"learning_rate": 8.135102788069015e-06,
"loss": 0.3142,
"step": 168
},
{
"epoch": 0.3235998085208234,
"grad_norm": 0.3829216957092285,
"learning_rate": 8.110369611107869e-06,
"loss": 0.3288,
"step": 169
},
{
"epoch": 0.3255146002872188,
"grad_norm": 0.42439737915992737,
"learning_rate": 8.085511653583772e-06,
"loss": 0.3452,
"step": 170
},
{
"epoch": 0.32742939205361415,
"grad_norm": 0.4537997543811798,
"learning_rate": 8.060529912738316e-06,
"loss": 0.4024,
"step": 171
},
{
"epoch": 0.3293441838200096,
"grad_norm": 0.5059738755226135,
"learning_rate": 8.035425390778975e-06,
"loss": 0.4093,
"step": 172
},
{
"epoch": 0.331258975586405,
"grad_norm": 0.41058847308158875,
"learning_rate": 8.010199094838915e-06,
"loss": 0.3704,
"step": 173
},
{
"epoch": 0.3331737673528004,
"grad_norm": 0.43019795417785645,
"learning_rate": 7.984852036936578e-06,
"loss": 0.3463,
"step": 174
},
{
"epoch": 0.3350885591191958,
"grad_norm": 0.4632018506526947,
"learning_rate": 7.959385233935087e-06,
"loss": 0.4076,
"step": 175
},
{
"epoch": 0.3370033508855912,
"grad_norm": 0.363118976354599,
"learning_rate": 7.933799707501448e-06,
"loss": 0.3089,
"step": 176
},
{
"epoch": 0.3389181426519866,
"grad_norm": 0.414247065782547,
"learning_rate": 7.908096484065569e-06,
"loss": 0.3691,
"step": 177
},
{
"epoch": 0.340832934418382,
"grad_norm": 0.4680274724960327,
"learning_rate": 7.88227659477908e-06,
"loss": 0.3742,
"step": 178
},
{
"epoch": 0.3427477261847774,
"grad_norm": 0.4838680624961853,
"learning_rate": 7.856341075473963e-06,
"loss": 0.3509,
"step": 179
},
{
"epoch": 0.3446625179511728,
"grad_norm": 0.39406099915504456,
"learning_rate": 7.830290966620997e-06,
"loss": 0.327,
"step": 180
},
{
"epoch": 0.3465773097175682,
"grad_norm": 0.4498644471168518,
"learning_rate": 7.804127313288023e-06,
"loss": 0.3665,
"step": 181
},
{
"epoch": 0.34849210148396365,
"grad_norm": 0.42900529503822327,
"learning_rate": 7.777851165098012e-06,
"loss": 0.3779,
"step": 182
},
{
"epoch": 0.350406893250359,
"grad_norm": 0.3877162039279938,
"learning_rate": 7.751463576186957e-06,
"loss": 0.3394,
"step": 183
},
{
"epoch": 0.3523216850167544,
"grad_norm": 0.6071410179138184,
"learning_rate": 7.72496560516159e-06,
"loss": 0.3813,
"step": 184
},
{
"epoch": 0.35423647678314985,
"grad_norm": 0.5272730588912964,
"learning_rate": 7.6983583150569e-06,
"loss": 0.3817,
"step": 185
},
{
"epoch": 0.35615126854954526,
"grad_norm": 0.4268261194229126,
"learning_rate": 7.671642773293506e-06,
"loss": 0.3415,
"step": 186
},
{
"epoch": 0.3580660603159406,
"grad_norm": 0.4886492192745209,
"learning_rate": 7.644820051634813e-06,
"loss": 0.3846,
"step": 187
},
{
"epoch": 0.35998085208233604,
"grad_norm": 0.4869603216648102,
"learning_rate": 7.617891226144034e-06,
"loss": 0.4274,
"step": 188
},
{
"epoch": 0.36189564384873146,
"grad_norm": 0.42415544390678406,
"learning_rate": 7.59085737714101e-06,
"loss": 0.3438,
"step": 189
},
{
"epoch": 0.3638104356151269,
"grad_norm": 0.5172827243804932,
"learning_rate": 7.563719589158874e-06,
"loss": 0.3836,
"step": 190
},
{
"epoch": 0.36572522738152224,
"grad_norm": 0.4153439998626709,
"learning_rate": 7.536478950900537e-06,
"loss": 0.3636,
"step": 191
},
{
"epoch": 0.36764001914791766,
"grad_norm": 0.42014452815055847,
"learning_rate": 7.509136555195025e-06,
"loss": 0.3501,
"step": 192
},
{
"epoch": 0.3695548109143131,
"grad_norm": 0.4194999933242798,
"learning_rate": 7.481693498953621e-06,
"loss": 0.3745,
"step": 193
},
{
"epoch": 0.3714696026807085,
"grad_norm": 0.4361554980278015,
"learning_rate": 7.4541508831258695e-06,
"loss": 0.3469,
"step": 194
},
{
"epoch": 0.37338439444710386,
"grad_norm": 0.3900616765022278,
"learning_rate": 7.4265098126554065e-06,
"loss": 0.3354,
"step": 195
},
{
"epoch": 0.3752991862134993,
"grad_norm": 0.47491058707237244,
"learning_rate": 7.3987713964356335e-06,
"loss": 0.4077,
"step": 196
},
{
"epoch": 0.3772139779798947,
"grad_norm": 0.5020147562026978,
"learning_rate": 7.370936747265226e-06,
"loss": 0.4198,
"step": 197
},
{
"epoch": 0.3791287697462901,
"grad_norm": 0.4094443619251251,
"learning_rate": 7.3430069818035e-06,
"loss": 0.3482,
"step": 198
},
{
"epoch": 0.3810435615126855,
"grad_norm": 0.4021066427230835,
"learning_rate": 7.314983220525604e-06,
"loss": 0.3561,
"step": 199
},
{
"epoch": 0.3829583532790809,
"grad_norm": 0.3932279348373413,
"learning_rate": 7.286866587677576e-06,
"loss": 0.3414,
"step": 200
},
{
"epoch": 0.3848731450454763,
"grad_norm": 0.466545045375824,
"learning_rate": 7.2586582112312355e-06,
"loss": 0.3868,
"step": 201
},
{
"epoch": 0.38678793681187174,
"grad_norm": 0.4097374677658081,
"learning_rate": 7.230359222838939e-06,
"loss": 0.3468,
"step": 202
},
{
"epoch": 0.3887027285782671,
"grad_norm": 0.4792751967906952,
"learning_rate": 7.201970757788172e-06,
"loss": 0.4131,
"step": 203
},
{
"epoch": 0.3906175203446625,
"grad_norm": 0.40435218811035156,
"learning_rate": 7.173493954956012e-06,
"loss": 0.3456,
"step": 204
},
{
"epoch": 0.39253231211105793,
"grad_norm": 0.45885634422302246,
"learning_rate": 7.144929956763438e-06,
"loss": 0.3672,
"step": 205
},
{
"epoch": 0.39444710387745335,
"grad_norm": 0.5055181384086609,
"learning_rate": 7.116279909129492e-06,
"loss": 0.4122,
"step": 206
},
{
"epoch": 0.3963618956438487,
"grad_norm": 0.42308080196380615,
"learning_rate": 7.087544961425317e-06,
"loss": 0.3621,
"step": 207
},
{
"epoch": 0.39827668741024413,
"grad_norm": 0.41654130816459656,
"learning_rate": 7.058726266428042e-06,
"loss": 0.3623,
"step": 208
},
{
"epoch": 0.40019147917663955,
"grad_norm": 0.3884025812149048,
"learning_rate": 7.029824980274536e-06,
"loss": 0.3302,
"step": 209
},
{
"epoch": 0.40210627094303497,
"grad_norm": 0.43903034925460815,
"learning_rate": 7.0008422624150285e-06,
"loss": 0.3728,
"step": 210
},
{
"epoch": 0.40402106270943033,
"grad_norm": 0.4426249563694,
"learning_rate": 6.971779275566593e-06,
"loss": 0.3756,
"step": 211
},
{
"epoch": 0.40593585447582575,
"grad_norm": 0.42531484365463257,
"learning_rate": 6.9426371856665005e-06,
"loss": 0.3782,
"step": 212
},
{
"epoch": 0.40785064624222117,
"grad_norm": 0.4498990476131439,
"learning_rate": 6.913417161825449e-06,
"loss": 0.3805,
"step": 213
},
{
"epoch": 0.4097654380086166,
"grad_norm": 0.43041449785232544,
"learning_rate": 6.884120376280658e-06,
"loss": 0.3514,
"step": 214
},
{
"epoch": 0.41168022977501195,
"grad_norm": 0.44849497079849243,
"learning_rate": 6.85474800434884e-06,
"loss": 0.3838,
"step": 215
},
{
"epoch": 0.41359502154140737,
"grad_norm": 0.3875875473022461,
"learning_rate": 6.8253012243790565e-06,
"loss": 0.3536,
"step": 216
},
{
"epoch": 0.4155098133078028,
"grad_norm": 0.4481753706932068,
"learning_rate": 6.795781217705436e-06,
"loss": 0.4041,
"step": 217
},
{
"epoch": 0.4174246050741982,
"grad_norm": 0.44694238901138306,
"learning_rate": 6.76618916859979e-06,
"loss": 0.3371,
"step": 218
},
{
"epoch": 0.41933939684059357,
"grad_norm": 0.3857312798500061,
"learning_rate": 6.736526264224101e-06,
"loss": 0.3386,
"step": 219
},
{
"epoch": 0.421254188606989,
"grad_norm": 0.4125332832336426,
"learning_rate": 6.706793694582892e-06,
"loss": 0.333,
"step": 220
},
{
"epoch": 0.4231689803733844,
"grad_norm": 0.4144175350666046,
"learning_rate": 6.676992652475487e-06,
"loss": 0.3735,
"step": 221
},
{
"epoch": 0.4250837721397798,
"grad_norm": 0.4129754304885864,
"learning_rate": 6.647124333448165e-06,
"loss": 0.324,
"step": 222
},
{
"epoch": 0.4269985639061752,
"grad_norm": 0.43009379506111145,
"learning_rate": 6.617189935746191e-06,
"loss": 0.3662,
"step": 223
},
{
"epoch": 0.4289133556725706,
"grad_norm": 0.43365833163261414,
"learning_rate": 6.587190660265752e-06,
"loss": 0.3737,
"step": 224
},
{
"epoch": 0.430828147438966,
"grad_norm": 0.4308427572250366,
"learning_rate": 6.55712771050577e-06,
"loss": 0.3681,
"step": 225
},
{
"epoch": 0.43274293920536144,
"grad_norm": 0.41205909848213196,
"learning_rate": 6.52700229251963e-06,
"loss": 0.3607,
"step": 226
},
{
"epoch": 0.4346577309717568,
"grad_norm": 0.44770729541778564,
"learning_rate": 6.496815614866792e-06,
"loss": 0.3564,
"step": 227
},
{
"epoch": 0.4365725227381522,
"grad_norm": 0.4543178677558899,
"learning_rate": 6.466568888564303e-06,
"loss": 0.4141,
"step": 228
},
{
"epoch": 0.43848731450454764,
"grad_norm": 0.4294315576553345,
"learning_rate": 6.436263327038225e-06,
"loss": 0.3584,
"step": 229
},
{
"epoch": 0.44040210627094306,
"grad_norm": 0.413022518157959,
"learning_rate": 6.405900146074941e-06,
"loss": 0.343,
"step": 230
},
{
"epoch": 0.4423168980373384,
"grad_norm": 0.4493683874607086,
"learning_rate": 6.375480563772391e-06,
"loss": 0.3954,
"step": 231
},
{
"epoch": 0.44423168980373384,
"grad_norm": 0.40023738145828247,
"learning_rate": 6.3450058004912004e-06,
"loss": 0.3218,
"step": 232
},
{
"epoch": 0.44614648157012926,
"grad_norm": 0.44051283597946167,
"learning_rate": 6.314477078805724e-06,
"loss": 0.3625,
"step": 233
},
{
"epoch": 0.4480612733365247,
"grad_norm": 0.39243170619010925,
"learning_rate": 6.283895623454997e-06,
"loss": 0.345,
"step": 234
},
{
"epoch": 0.44997606510292004,
"grad_norm": 0.4121267795562744,
"learning_rate": 6.2532626612936035e-06,
"loss": 0.3666,
"step": 235
},
{
"epoch": 0.45189085686931546,
"grad_norm": 0.4519788324832916,
"learning_rate": 6.2225794212424565e-06,
"loss": 0.3913,
"step": 236
},
{
"epoch": 0.4538056486357109,
"grad_norm": 0.4214268624782562,
"learning_rate": 6.191847134239496e-06,
"loss": 0.3558,
"step": 237
},
{
"epoch": 0.4557204404021063,
"grad_norm": 0.8309491872787476,
"learning_rate": 6.161067033190311e-06,
"loss": 0.3823,
"step": 238
},
{
"epoch": 0.45763523216850166,
"grad_norm": 0.4205809235572815,
"learning_rate": 6.130240352918675e-06,
"loss": 0.3568,
"step": 239
},
{
"epoch": 0.4595500239348971,
"grad_norm": 0.47291141748428345,
"learning_rate": 6.0993683301170046e-06,
"loss": 0.3873,
"step": 240
},
{
"epoch": 0.4614648157012925,
"grad_norm": 0.43636125326156616,
"learning_rate": 6.068452203296754e-06,
"loss": 0.3465,
"step": 241
},
{
"epoch": 0.4633796074676879,
"grad_norm": 0.4022378921508789,
"learning_rate": 6.0374932127387234e-06,
"loss": 0.3512,
"step": 242
},
{
"epoch": 0.4652943992340833,
"grad_norm": 0.4230445623397827,
"learning_rate": 6.006492600443301e-06,
"loss": 0.357,
"step": 243
},
{
"epoch": 0.4672091910004787,
"grad_norm": 0.40407922863960266,
"learning_rate": 5.975451610080643e-06,
"loss": 0.3417,
"step": 244
},
{
"epoch": 0.4691239827668741,
"grad_norm": 0.46995967626571655,
"learning_rate": 5.944371486940772e-06,
"loss": 0.3619,
"step": 245
},
{
"epoch": 0.47103877453326953,
"grad_norm": 0.4951625168323517,
"learning_rate": 5.913253477883629e-06,
"loss": 0.3982,
"step": 246
},
{
"epoch": 0.4729535662996649,
"grad_norm": 0.39868447184562683,
"learning_rate": 5.882098831289044e-06,
"loss": 0.3482,
"step": 247
},
{
"epoch": 0.4748683580660603,
"grad_norm": 0.42509621381759644,
"learning_rate": 5.850908797006656e-06,
"loss": 0.3585,
"step": 248
},
{
"epoch": 0.47678314983245573,
"grad_norm": 0.4544394314289093,
"learning_rate": 5.819684626305776e-06,
"loss": 0.3967,
"step": 249
},
{
"epoch": 0.47869794159885115,
"grad_norm": 0.42766261100769043,
"learning_rate": 5.788427571825186e-06,
"loss": 0.356,
"step": 250
},
{
"epoch": 0.4806127333652465,
"grad_norm": 0.4494774639606476,
"learning_rate": 5.757138887522884e-06,
"loss": 0.365,
"step": 251
},
{
"epoch": 0.48252752513164193,
"grad_norm": 0.4610314965248108,
"learning_rate": 5.725819828625782e-06,
"loss": 0.3633,
"step": 252
},
{
"epoch": 0.48444231689803735,
"grad_norm": 0.41364309191703796,
"learning_rate": 5.694471651579346e-06,
"loss": 0.3646,
"step": 253
},
{
"epoch": 0.48635710866443277,
"grad_norm": 0.39207273721694946,
"learning_rate": 5.663095613997196e-06,
"loss": 0.3417,
"step": 254
},
{
"epoch": 0.48827190043082813,
"grad_norm": 0.39583784341812134,
"learning_rate": 5.631692974610647e-06,
"loss": 0.3362,
"step": 255
},
{
"epoch": 0.49018669219722355,
"grad_norm": 0.44171953201293945,
"learning_rate": 5.600264993218215e-06,
"loss": 0.3738,
"step": 256
},
{
"epoch": 0.49210148396361897,
"grad_norm": 0.43829062581062317,
"learning_rate": 5.568812930635076e-06,
"loss": 0.3539,
"step": 257
},
{
"epoch": 0.4940162757300144,
"grad_norm": 0.469128280878067,
"learning_rate": 5.537338048642487e-06,
"loss": 0.3923,
"step": 258
},
{
"epoch": 0.49593106749640975,
"grad_norm": 0.4256131649017334,
"learning_rate": 5.505841609937162e-06,
"loss": 0.3386,
"step": 259
},
{
"epoch": 0.49784585926280517,
"grad_norm": 0.3736073672771454,
"learning_rate": 5.474324878080623e-06,
"loss": 0.2973,
"step": 260
},
{
"epoch": 0.4997606510292006,
"grad_norm": 0.41736721992492676,
"learning_rate": 5.4427891174485014e-06,
"loss": 0.3374,
"step": 261
},
{
"epoch": 0.501675442795596,
"grad_norm": 0.42438390851020813,
"learning_rate": 5.41123559317982e-06,
"loss": 0.3703,
"step": 262
},
{
"epoch": 0.5035902345619914,
"grad_norm": 0.4202900826931,
"learning_rate": 5.379665571126232e-06,
"loss": 0.3393,
"step": 263
},
{
"epoch": 0.5055050263283868,
"grad_norm": 0.3746366798877716,
"learning_rate": 5.348080317801244e-06,
"loss": 0.3235,
"step": 264
},
{
"epoch": 0.5074198180947822,
"grad_norm": 0.4379538595676422,
"learning_rate": 5.316481100329408e-06,
"loss": 0.3858,
"step": 265
},
{
"epoch": 0.5093346098611776,
"grad_norm": 0.416400671005249,
"learning_rate": 5.284869186395478e-06,
"loss": 0.3696,
"step": 266
},
{
"epoch": 0.511249401627573,
"grad_norm": 0.4202958047389984,
"learning_rate": 5.253245844193564e-06,
"loss": 0.3625,
"step": 267
},
{
"epoch": 0.5131641933939685,
"grad_norm": 0.41637325286865234,
"learning_rate": 5.22161234237625e-06,
"loss": 0.3568,
"step": 268
},
{
"epoch": 0.5150789851603638,
"grad_norm": 0.40750014781951904,
"learning_rate": 5.189969950003697e-06,
"loss": 0.3546,
"step": 269
},
{
"epoch": 0.5169937769267592,
"grad_norm": 0.4708370864391327,
"learning_rate": 5.158319936492736e-06,
"loss": 0.3867,
"step": 270
},
{
"epoch": 0.5189085686931546,
"grad_norm": 0.40184107422828674,
"learning_rate": 5.12666357156594e-06,
"loss": 0.3478,
"step": 271
},
{
"epoch": 0.52082336045955,
"grad_norm": 0.46964147686958313,
"learning_rate": 5.0950021252006845e-06,
"loss": 0.3854,
"step": 272
},
{
"epoch": 0.5227381522259454,
"grad_norm": 0.4432660937309265,
"learning_rate": 5.063336867578201e-06,
"loss": 0.3963,
"step": 273
},
{
"epoch": 0.5246529439923409,
"grad_norm": 0.41775625944137573,
"learning_rate": 5.0316690690326175e-06,
"loss": 0.3388,
"step": 274
},
{
"epoch": 0.5265677357587363,
"grad_norm": 0.4029581844806671,
"learning_rate": 5e-06,
"loss": 0.331,
"step": 275
},
{
"epoch": 0.5284825275251317,
"grad_norm": 0.4377078115940094,
"learning_rate": 4.9683309309673825e-06,
"loss": 0.3436,
"step": 276
},
{
"epoch": 0.530397319291527,
"grad_norm": 0.38507312536239624,
"learning_rate": 4.936663132421801e-06,
"loss": 0.3239,
"step": 277
},
{
"epoch": 0.5323121110579224,
"grad_norm": 0.39098119735717773,
"learning_rate": 4.904997874799316e-06,
"loss": 0.3395,
"step": 278
},
{
"epoch": 0.5342269028243178,
"grad_norm": 0.41355735063552856,
"learning_rate": 4.873336428434062e-06,
"loss": 0.3459,
"step": 279
},
{
"epoch": 0.5361416945907133,
"grad_norm": 0.40377330780029297,
"learning_rate": 4.841680063507265e-06,
"loss": 0.3755,
"step": 280
},
{
"epoch": 0.5380564863571087,
"grad_norm": 0.4401605427265167,
"learning_rate": 4.8100300499963045e-06,
"loss": 0.3749,
"step": 281
},
{
"epoch": 0.5399712781235041,
"grad_norm": 0.3882392644882202,
"learning_rate": 4.778387657623751e-06,
"loss": 0.3314,
"step": 282
},
{
"epoch": 0.5418860698898995,
"grad_norm": 0.47119903564453125,
"learning_rate": 4.746754155806437e-06,
"loss": 0.4131,
"step": 283
},
{
"epoch": 0.5438008616562949,
"grad_norm": 0.41662779450416565,
"learning_rate": 4.715130813604522e-06,
"loss": 0.3572,
"step": 284
},
{
"epoch": 0.5457156534226902,
"grad_norm": 0.409316748380661,
"learning_rate": 4.683518899670594e-06,
"loss": 0.3488,
"step": 285
},
{
"epoch": 0.5476304451890857,
"grad_norm": 0.4336586892604828,
"learning_rate": 4.651919682198756e-06,
"loss": 0.3503,
"step": 286
},
{
"epoch": 0.5495452369554811,
"grad_norm": 0.4213099777698517,
"learning_rate": 4.62033442887377e-06,
"loss": 0.3584,
"step": 287
},
{
"epoch": 0.5514600287218765,
"grad_norm": 0.43022260069847107,
"learning_rate": 4.588764406820181e-06,
"loss": 0.3884,
"step": 288
},
{
"epoch": 0.5533748204882719,
"grad_norm": 0.38072916865348816,
"learning_rate": 4.5572108825515e-06,
"loss": 0.324,
"step": 289
},
{
"epoch": 0.5552896122546673,
"grad_norm": 0.444703608751297,
"learning_rate": 4.5256751219193784e-06,
"loss": 0.3479,
"step": 290
},
{
"epoch": 0.5572044040210627,
"grad_norm": 0.42911389470100403,
"learning_rate": 4.49415839006284e-06,
"loss": 0.3643,
"step": 291
},
{
"epoch": 0.5591191957874582,
"grad_norm": 0.4525517225265503,
"learning_rate": 4.462661951357515e-06,
"loss": 0.3678,
"step": 292
},
{
"epoch": 0.5610339875538535,
"grad_norm": 0.44527730345726013,
"learning_rate": 4.431187069364927e-06,
"loss": 0.3711,
"step": 293
},
{
"epoch": 0.5629487793202489,
"grad_norm": 0.42514294385910034,
"learning_rate": 4.3997350067817866e-06,
"loss": 0.358,
"step": 294
},
{
"epoch": 0.5648635710866443,
"grad_norm": 0.44586825370788574,
"learning_rate": 4.368307025389355e-06,
"loss": 0.3789,
"step": 295
},
{
"epoch": 0.5667783628530397,
"grad_norm": 0.4211961627006531,
"learning_rate": 4.336904386002805e-06,
"loss": 0.3443,
"step": 296
},
{
"epoch": 0.5686931546194351,
"grad_norm": 0.42788630723953247,
"learning_rate": 4.3055283484206565e-06,
"loss": 0.3791,
"step": 297
},
{
"epoch": 0.5706079463858306,
"grad_norm": 0.3960091173648834,
"learning_rate": 4.27418017137422e-06,
"loss": 0.33,
"step": 298
},
{
"epoch": 0.572522738152226,
"grad_norm": 0.3952198326587677,
"learning_rate": 4.2428611124771184e-06,
"loss": 0.3283,
"step": 299
},
{
"epoch": 0.5744375299186214,
"grad_norm": 0.40487802028656006,
"learning_rate": 4.211572428174816e-06,
"loss": 0.3251,
"step": 300
},
{
"epoch": 0.5763523216850167,
"grad_norm": 0.4621725380420685,
"learning_rate": 4.180315373694225e-06,
"loss": 0.3669,
"step": 301
},
{
"epoch": 0.5782671134514121,
"grad_norm": 0.41292381286621094,
"learning_rate": 4.149091202993345e-06,
"loss": 0.3132,
"step": 302
},
{
"epoch": 0.5801819052178075,
"grad_norm": 0.4315532147884369,
"learning_rate": 4.11790116871096e-06,
"loss": 0.3732,
"step": 303
},
{
"epoch": 0.582096696984203,
"grad_norm": 0.48196038603782654,
"learning_rate": 4.086746522116372e-06,
"loss": 0.4136,
"step": 304
},
{
"epoch": 0.5840114887505984,
"grad_norm": 0.38967981934547424,
"learning_rate": 4.055628513059231e-06,
"loss": 0.3445,
"step": 305
},
{
"epoch": 0.5859262805169938,
"grad_norm": 0.4445069134235382,
"learning_rate": 4.02454838991936e-06,
"loss": 0.3391,
"step": 306
},
{
"epoch": 0.5878410722833892,
"grad_norm": 0.4397939443588257,
"learning_rate": 3.993507399556699e-06,
"loss": 0.3787,
"step": 307
},
{
"epoch": 0.5897558640497846,
"grad_norm": 0.38508886098861694,
"learning_rate": 3.962506787261278e-06,
"loss": 0.3472,
"step": 308
},
{
"epoch": 0.59167065581618,
"grad_norm": 0.4952245056629181,
"learning_rate": 3.931547796703245e-06,
"loss": 0.3421,
"step": 309
},
{
"epoch": 0.5935854475825754,
"grad_norm": 0.4101730287075043,
"learning_rate": 3.900631669882996e-06,
"loss": 0.3509,
"step": 310
},
{
"epoch": 0.5955002393489708,
"grad_norm": 0.40140795707702637,
"learning_rate": 3.869759647081326e-06,
"loss": 0.3274,
"step": 311
},
{
"epoch": 0.5974150311153662,
"grad_norm": 0.3836864233016968,
"learning_rate": 3.83893296680969e-06,
"loss": 0.3425,
"step": 312
},
{
"epoch": 0.5993298228817616,
"grad_norm": 0.4107282757759094,
"learning_rate": 3.8081528657605045e-06,
"loss": 0.3352,
"step": 313
},
{
"epoch": 0.601244614648157,
"grad_norm": 0.39755916595458984,
"learning_rate": 3.7774205787575455e-06,
"loss": 0.345,
"step": 314
},
{
"epoch": 0.6031594064145525,
"grad_norm": 0.41993725299835205,
"learning_rate": 3.7467373387063973e-06,
"loss": 0.3852,
"step": 315
},
{
"epoch": 0.6050741981809479,
"grad_norm": 0.47688186168670654,
"learning_rate": 3.7161043765450044e-06,
"loss": 0.3707,
"step": 316
},
{
"epoch": 0.6069889899473432,
"grad_norm": 0.40953752398490906,
"learning_rate": 3.685522921194276e-06,
"loss": 0.3341,
"step": 317
},
{
"epoch": 0.6089037817137386,
"grad_norm": 0.4336431324481964,
"learning_rate": 3.6549941995088012e-06,
"loss": 0.3549,
"step": 318
},
{
"epoch": 0.610818573480134,
"grad_norm": 0.4835233986377716,
"learning_rate": 3.62451943622761e-06,
"loss": 0.3857,
"step": 319
},
{
"epoch": 0.6127333652465294,
"grad_norm": 0.3688715994358063,
"learning_rate": 3.5940998539250614e-06,
"loss": 0.3087,
"step": 320
},
{
"epoch": 0.6146481570129249,
"grad_norm": 0.40031322836875916,
"learning_rate": 3.5637366729617766e-06,
"loss": 0.3299,
"step": 321
},
{
"epoch": 0.6165629487793203,
"grad_norm": 0.43611854314804077,
"learning_rate": 3.5334311114356983e-06,
"loss": 0.3744,
"step": 322
},
{
"epoch": 0.6184777405457157,
"grad_norm": 0.4520419239997864,
"learning_rate": 3.5031843851332105e-06,
"loss": 0.3774,
"step": 323
},
{
"epoch": 0.6203925323121111,
"grad_norm": 0.4471440315246582,
"learning_rate": 3.472997707480372e-06,
"loss": 0.3639,
"step": 324
},
{
"epoch": 0.6223073240785064,
"grad_norm": 0.47213125228881836,
"learning_rate": 3.4428722894942313e-06,
"loss": 0.3765,
"step": 325
},
{
"epoch": 0.6242221158449018,
"grad_norm": 0.4469670355319977,
"learning_rate": 3.4128093397342508e-06,
"loss": 0.3922,
"step": 326
},
{
"epoch": 0.6261369076112973,
"grad_norm": 0.40332043170928955,
"learning_rate": 3.3828100642538097e-06,
"loss": 0.3361,
"step": 327
},
{
"epoch": 0.6280516993776927,
"grad_norm": 0.3857273757457733,
"learning_rate": 3.352875666551838e-06,
"loss": 0.328,
"step": 328
},
{
"epoch": 0.6299664911440881,
"grad_norm": 0.4381800591945648,
"learning_rate": 3.323007347524515e-06,
"loss": 0.3628,
"step": 329
},
{
"epoch": 0.6318812829104835,
"grad_norm": 0.36643126606941223,
"learning_rate": 3.2932063054171108e-06,
"loss": 0.3401,
"step": 330
},
{
"epoch": 0.6337960746768789,
"grad_norm": 0.395336389541626,
"learning_rate": 3.2634737357758994e-06,
"loss": 0.3383,
"step": 331
},
{
"epoch": 0.6357108664432743,
"grad_norm": 0.4047113358974457,
"learning_rate": 3.2338108314002102e-06,
"loss": 0.3371,
"step": 332
},
{
"epoch": 0.6376256582096697,
"grad_norm": 0.4104958474636078,
"learning_rate": 3.204218782294565e-06,
"loss": 0.3278,
"step": 333
},
{
"epoch": 0.6395404499760651,
"grad_norm": 0.39561185240745544,
"learning_rate": 3.174698775620947e-06,
"loss": 0.3509,
"step": 334
},
{
"epoch": 0.6414552417424605,
"grad_norm": 0.4158308804035187,
"learning_rate": 3.145251995651162e-06,
"loss": 0.3574,
"step": 335
},
{
"epoch": 0.6433700335088559,
"grad_norm": 0.39432471990585327,
"learning_rate": 3.1158796237193444e-06,
"loss": 0.3403,
"step": 336
},
{
"epoch": 0.6452848252752513,
"grad_norm": 0.39117756485939026,
"learning_rate": 3.0865828381745515e-06,
"loss": 0.3434,
"step": 337
},
{
"epoch": 0.6471996170416467,
"grad_norm": 0.39029520750045776,
"learning_rate": 3.0573628143334986e-06,
"loss": 0.3233,
"step": 338
},
{
"epoch": 0.6491144088080422,
"grad_norm": 0.40311500430107117,
"learning_rate": 3.0282207244334084e-06,
"loss": 0.3379,
"step": 339
},
{
"epoch": 0.6510292005744376,
"grad_norm": 0.3994643986225128,
"learning_rate": 2.999157737584971e-06,
"loss": 0.3369,
"step": 340
},
{
"epoch": 0.6529439923408329,
"grad_norm": 0.42533010244369507,
"learning_rate": 2.970175019725465e-06,
"loss": 0.3764,
"step": 341
},
{
"epoch": 0.6548587841072283,
"grad_norm": 0.48993390798568726,
"learning_rate": 2.94127373357196e-06,
"loss": 0.4002,
"step": 342
},
{
"epoch": 0.6567735758736237,
"grad_norm": 0.4385669231414795,
"learning_rate": 2.912455038574686e-06,
"loss": 0.3495,
"step": 343
},
{
"epoch": 0.6586883676400191,
"grad_norm": 0.4084681570529938,
"learning_rate": 2.88372009087051e-06,
"loss": 0.3595,
"step": 344
},
{
"epoch": 0.6606031594064146,
"grad_norm": 0.413422167301178,
"learning_rate": 2.8550700432365647e-06,
"loss": 0.3222,
"step": 345
},
{
"epoch": 0.66251795117281,
"grad_norm": 0.465225487947464,
"learning_rate": 2.8265060450439887e-06,
"loss": 0.3949,
"step": 346
},
{
"epoch": 0.6644327429392054,
"grad_norm": 0.4203803539276123,
"learning_rate": 2.7980292422118282e-06,
"loss": 0.3414,
"step": 347
},
{
"epoch": 0.6663475347056008,
"grad_norm": 0.40917420387268066,
"learning_rate": 2.769640777161063e-06,
"loss": 0.3481,
"step": 348
},
{
"epoch": 0.6682623264719961,
"grad_norm": 0.4400990605354309,
"learning_rate": 2.7413417887687644e-06,
"loss": 0.3814,
"step": 349
},
{
"epoch": 0.6701771182383915,
"grad_norm": 0.48319095373153687,
"learning_rate": 2.713133412322424e-06,
"loss": 0.3562,
"step": 350
},
{
"epoch": 0.672091910004787,
"grad_norm": 0.45311659574508667,
"learning_rate": 2.6850167794743966e-06,
"loss": 0.3776,
"step": 351
},
{
"epoch": 0.6740067017711824,
"grad_norm": 0.41590750217437744,
"learning_rate": 2.6569930181965e-06,
"loss": 0.3451,
"step": 352
},
{
"epoch": 0.6759214935375778,
"grad_norm": 0.40761658549308777,
"learning_rate": 2.629063252734775e-06,
"loss": 0.3571,
"step": 353
},
{
"epoch": 0.6778362853039732,
"grad_norm": 0.4324440062046051,
"learning_rate": 2.601228603564368e-06,
"loss": 0.3639,
"step": 354
},
{
"epoch": 0.6797510770703686,
"grad_norm": 0.4608740508556366,
"learning_rate": 2.573490187344596e-06,
"loss": 0.3437,
"step": 355
},
{
"epoch": 0.681665868836764,
"grad_norm": 0.386198490858078,
"learning_rate": 2.545849116874132e-06,
"loss": 0.3035,
"step": 356
},
{
"epoch": 0.6835806606031594,
"grad_norm": 0.376306414604187,
"learning_rate": 2.5183065010463813e-06,
"loss": 0.3209,
"step": 357
},
{
"epoch": 0.6854954523695548,
"grad_norm": 0.36713552474975586,
"learning_rate": 2.490863444804976e-06,
"loss": 0.3192,
"step": 358
},
{
"epoch": 0.6874102441359502,
"grad_norm": 0.4572632908821106,
"learning_rate": 2.4635210490994648e-06,
"loss": 0.3667,
"step": 359
},
{
"epoch": 0.6893250359023456,
"grad_norm": 0.39816582202911377,
"learning_rate": 2.436280410841128e-06,
"loss": 0.3302,
"step": 360
},
{
"epoch": 0.691239827668741,
"grad_norm": 0.4200025200843811,
"learning_rate": 2.409142622858992e-06,
"loss": 0.3416,
"step": 361
},
{
"epoch": 0.6931546194351365,
"grad_norm": 0.423297256231308,
"learning_rate": 2.3821087738559674e-06,
"loss": 0.3366,
"step": 362
},
{
"epoch": 0.6950694112015319,
"grad_norm": 0.40153416991233826,
"learning_rate": 2.3551799483651894e-06,
"loss": 0.339,
"step": 363
},
{
"epoch": 0.6969842029679273,
"grad_norm": 0.4924335479736328,
"learning_rate": 2.3283572267064963e-06,
"loss": 0.4113,
"step": 364
},
{
"epoch": 0.6988989947343226,
"grad_norm": 0.421258807182312,
"learning_rate": 2.3016416849431023e-06,
"loss": 0.3621,
"step": 365
},
{
"epoch": 0.700813786500718,
"grad_norm": 0.45369982719421387,
"learning_rate": 2.275034394838413e-06,
"loss": 0.3649,
"step": 366
},
{
"epoch": 0.7027285782671134,
"grad_norm": 0.4841824471950531,
"learning_rate": 2.2485364238130435e-06,
"loss": 0.3634,
"step": 367
},
{
"epoch": 0.7046433700335089,
"grad_norm": 1.3670042753219604,
"learning_rate": 2.2221488349019903e-06,
"loss": 0.3491,
"step": 368
},
{
"epoch": 0.7065581617999043,
"grad_norm": 0.4118163585662842,
"learning_rate": 2.1958726867119785e-06,
"loss": 0.3769,
"step": 369
},
{
"epoch": 0.7084729535662997,
"grad_norm": 0.4216058850288391,
"learning_rate": 2.169709033379004e-06,
"loss": 0.3592,
"step": 370
},
{
"epoch": 0.7103877453326951,
"grad_norm": 0.42444273829460144,
"learning_rate": 2.1436589245260375e-06,
"loss": 0.3766,
"step": 371
},
{
"epoch": 0.7123025370990905,
"grad_norm": 0.4765802323818207,
"learning_rate": 2.1177234052209208e-06,
"loss": 0.399,
"step": 372
},
{
"epoch": 0.7142173288654858,
"grad_norm": 0.46905720233917236,
"learning_rate": 2.09190351593443e-06,
"loss": 0.3855,
"step": 373
},
{
"epoch": 0.7161321206318813,
"grad_norm": 0.42056676745414734,
"learning_rate": 2.066200292498553e-06,
"loss": 0.3512,
"step": 374
},
{
"epoch": 0.7180469123982767,
"grad_norm": 0.43623343110084534,
"learning_rate": 2.040614766064913e-06,
"loss": 0.3491,
"step": 375
},
{
"epoch": 0.7199617041646721,
"grad_norm": 0.4116828143596649,
"learning_rate": 2.0151479630634225e-06,
"loss": 0.3719,
"step": 376
},
{
"epoch": 0.7218764959310675,
"grad_norm": 0.4012473523616791,
"learning_rate": 1.9898009051610847e-06,
"loss": 0.3354,
"step": 377
},
{
"epoch": 0.7237912876974629,
"grad_norm": 0.416290819644928,
"learning_rate": 1.964574609221026e-06,
"loss": 0.353,
"step": 378
},
{
"epoch": 0.7257060794638583,
"grad_norm": 0.38975054025650024,
"learning_rate": 1.9394700872616856e-06,
"loss": 0.3247,
"step": 379
},
{
"epoch": 0.7276208712302538,
"grad_norm": 0.4278734624385834,
"learning_rate": 1.914488346416229e-06,
"loss": 0.3607,
"step": 380
},
{
"epoch": 0.7295356629966491,
"grad_norm": 0.4305007755756378,
"learning_rate": 1.8896303888921313e-06,
"loss": 0.3795,
"step": 381
},
{
"epoch": 0.7314504547630445,
"grad_norm": 0.37711989879608154,
"learning_rate": 1.8648972119309854e-06,
"loss": 0.3225,
"step": 382
},
{
"epoch": 0.7333652465294399,
"grad_norm": 0.4459383189678192,
"learning_rate": 1.8402898077684806e-06,
"loss": 0.3773,
"step": 383
},
{
"epoch": 0.7352800382958353,
"grad_norm": 0.4332491159439087,
"learning_rate": 1.815809163594609e-06,
"loss": 0.3714,
"step": 384
},
{
"epoch": 0.7371948300622307,
"grad_norm": 0.4447531998157501,
"learning_rate": 1.7914562615140507e-06,
"loss": 0.3925,
"step": 385
},
{
"epoch": 0.7391096218286262,
"grad_norm": 0.42695391178131104,
"learning_rate": 1.7672320785067871e-06,
"loss": 0.3565,
"step": 386
},
{
"epoch": 0.7410244135950216,
"grad_norm": 0.42980119585990906,
"learning_rate": 1.74313758638889e-06,
"loss": 0.325,
"step": 387
},
{
"epoch": 0.742939205361417,
"grad_norm": 0.4298250377178192,
"learning_rate": 1.7191737517735513e-06,
"loss": 0.3431,
"step": 388
},
{
"epoch": 0.7448539971278123,
"grad_norm": 0.3868040144443512,
"learning_rate": 1.6953415360322972e-06,
"loss": 0.3399,
"step": 389
},
{
"epoch": 0.7467687888942077,
"grad_norm": 0.4056515693664551,
"learning_rate": 1.6716418952564145e-06,
"loss": 0.3302,
"step": 390
},
{
"epoch": 0.7486835806606031,
"grad_norm": 0.3780469596385956,
"learning_rate": 1.648075780218607e-06,
"loss": 0.3135,
"step": 391
},
{
"epoch": 0.7505983724269986,
"grad_norm": 0.4133198857307434,
"learning_rate": 1.6246441363348453e-06,
"loss": 0.3466,
"step": 392
},
{
"epoch": 0.752513164193394,
"grad_norm": 0.4301663339138031,
"learning_rate": 1.6013479036264358e-06,
"loss": 0.3355,
"step": 393
},
{
"epoch": 0.7544279559597894,
"grad_norm": 0.39125901460647583,
"learning_rate": 1.57818801668232e-06,
"loss": 0.302,
"step": 394
},
{
"epoch": 0.7563427477261848,
"grad_norm": 0.4187367260456085,
"learning_rate": 1.555165404621567e-06,
"loss": 0.3554,
"step": 395
},
{
"epoch": 0.7582575394925802,
"grad_norm": 0.49991196393966675,
"learning_rate": 1.532280991056116e-06,
"loss": 0.4163,
"step": 396
},
{
"epoch": 0.7601723312589755,
"grad_norm": 0.4100117087364197,
"learning_rate": 1.5095356940537053e-06,
"loss": 0.3455,
"step": 397
},
{
"epoch": 0.762087123025371,
"grad_norm": 0.4645809829235077,
"learning_rate": 1.4869304261010586e-06,
"loss": 0.3605,
"step": 398
},
{
"epoch": 0.7640019147917664,
"grad_norm": 0.8521024584770203,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.4221,
"step": 399
},
{
"epoch": 0.7659167065581618,
"grad_norm": 0.42581960558891296,
"learning_rate": 1.4421435991674e-06,
"loss": 0.3414,
"step": 400
},
{
"epoch": 0.7678314983245572,
"grad_norm": 0.3728087246417999,
"learning_rate": 1.4199638369263858e-06,
"loss": 0.2917,
"step": 401
},
{
"epoch": 0.7697462900909526,
"grad_norm": 0.3792401850223541,
"learning_rate": 1.3979276971430406e-06,
"loss": 0.3017,
"step": 402
},
{
"epoch": 0.771661081857348,
"grad_norm": 0.412489116191864,
"learning_rate": 1.3760360638544012e-06,
"loss": 0.356,
"step": 403
},
{
"epoch": 0.7735758736237435,
"grad_norm": 0.3927983045578003,
"learning_rate": 1.3542898153002453e-06,
"loss": 0.3432,
"step": 404
},
{
"epoch": 0.7754906653901388,
"grad_norm": 0.42460405826568604,
"learning_rate": 1.3326898238878716e-06,
"loss": 0.3497,
"step": 405
},
{
"epoch": 0.7774054571565342,
"grad_norm": 0.4294438064098358,
"learning_rate": 1.3112369561570842e-06,
"loss": 0.3531,
"step": 406
},
{
"epoch": 0.7793202489229296,
"grad_norm": 0.40134039521217346,
"learning_rate": 1.2899320727454472e-06,
"loss": 0.3257,
"step": 407
},
{
"epoch": 0.781235040689325,
"grad_norm": 0.4099023938179016,
"learning_rate": 1.2687760283537414e-06,
"loss": 0.3448,
"step": 408
},
{
"epoch": 0.7831498324557205,
"grad_norm": 0.4487830400466919,
"learning_rate": 1.2477696717116878e-06,
"loss": 0.4018,
"step": 409
},
{
"epoch": 0.7850646242221159,
"grad_norm": 0.413581907749176,
"learning_rate": 1.226913845543895e-06,
"loss": 0.3398,
"step": 410
},
{
"epoch": 0.7869794159885113,
"grad_norm": 0.4662453234195709,
"learning_rate": 1.2062093865360458e-06,
"loss": 0.4084,
"step": 411
},
{
"epoch": 0.7888942077549067,
"grad_norm": 0.3948284387588501,
"learning_rate": 1.1856571253013393e-06,
"loss": 0.3275,
"step": 412
},
{
"epoch": 0.790808999521302,
"grad_norm": 0.427971214056015,
"learning_rate": 1.1652578863471664e-06,
"loss": 0.3693,
"step": 413
},
{
"epoch": 0.7927237912876974,
"grad_norm": 0.4065532386302948,
"learning_rate": 1.145012488042026e-06,
"loss": 0.3253,
"step": 414
},
{
"epoch": 0.7946385830540929,
"grad_norm": 0.5054202079772949,
"learning_rate": 1.1249217425827063e-06,
"loss": 0.4227,
"step": 415
},
{
"epoch": 0.7965533748204883,
"grad_norm": 0.4664987325668335,
"learning_rate": 1.1049864559616885e-06,
"loss": 0.3697,
"step": 416
},
{
"epoch": 0.7984681665868837,
"grad_norm": 0.41670548915863037,
"learning_rate": 1.0852074279348234e-06,
"loss": 0.3322,
"step": 417
},
{
"epoch": 0.8003829583532791,
"grad_norm": 0.4049803912639618,
"learning_rate": 1.0655854519892367e-06,
"loss": 0.3407,
"step": 418
},
{
"epoch": 0.8022977501196745,
"grad_norm": 0.40564632415771484,
"learning_rate": 1.046121315311508e-06,
"loss": 0.3513,
"step": 419
},
{
"epoch": 0.8042125418860699,
"grad_norm": 0.4141291379928589,
"learning_rate": 1.0268157987560773e-06,
"loss": 0.3565,
"step": 420
}
],
"logging_steps": 1,
"max_steps": 522,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 105,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2.2094028312097587e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}