guardian-vanilla / trainer_state.json
paulpacaud's picture
Initial release: guardian-vanilla checkpoint + model card
5246b65
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9996113486202876,
"eval_steps": 500,
"global_step": 1929,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0005182018396165306,
"grad_norm": 5.4082818031311035,
"learning_rate": 6.896551724137931e-07,
"loss": 0.2466,
"step": 1
},
{
"epoch": 0.0010364036792330613,
"grad_norm": 3.835052013397217,
"learning_rate": 1.3793103448275862e-06,
"loss": 0.2488,
"step": 2
},
{
"epoch": 0.001554605518849592,
"grad_norm": 5.501279354095459,
"learning_rate": 2.0689655172413796e-06,
"loss": 0.28,
"step": 3
},
{
"epoch": 0.0020728073584661225,
"grad_norm": 4.11079216003418,
"learning_rate": 2.7586206896551725e-06,
"loss": 0.245,
"step": 4
},
{
"epoch": 0.002591009198082653,
"grad_norm": 4.3664326667785645,
"learning_rate": 3.448275862068966e-06,
"loss": 0.2753,
"step": 5
},
{
"epoch": 0.003109211037699184,
"grad_norm": 5.692226409912109,
"learning_rate": 4.137931034482759e-06,
"loss": 0.234,
"step": 6
},
{
"epoch": 0.0036274128773157144,
"grad_norm": 4.383513450622559,
"learning_rate": 4.8275862068965525e-06,
"loss": 0.2831,
"step": 7
},
{
"epoch": 0.004145614716932245,
"grad_norm": 4.188362121582031,
"learning_rate": 5.517241379310345e-06,
"loss": 0.2912,
"step": 8
},
{
"epoch": 0.004663816556548776,
"grad_norm": 6.689868450164795,
"learning_rate": 6.206896551724138e-06,
"loss": 0.2132,
"step": 9
},
{
"epoch": 0.005182018396165306,
"grad_norm": 5.252689361572266,
"learning_rate": 6.896551724137932e-06,
"loss": 0.1583,
"step": 10
},
{
"epoch": 0.005700220235781837,
"grad_norm": 4.101102828979492,
"learning_rate": 7.586206896551724e-06,
"loss": 0.2285,
"step": 11
},
{
"epoch": 0.006218422075398368,
"grad_norm": 3.757774829864502,
"learning_rate": 8.275862068965518e-06,
"loss": 0.2126,
"step": 12
},
{
"epoch": 0.006736623915014899,
"grad_norm": 3.110372543334961,
"learning_rate": 8.965517241379312e-06,
"loss": 0.1666,
"step": 13
},
{
"epoch": 0.007254825754631429,
"grad_norm": 1.5280357599258423,
"learning_rate": 9.655172413793105e-06,
"loss": 0.1976,
"step": 14
},
{
"epoch": 0.00777302759424796,
"grad_norm": 2.1373836994171143,
"learning_rate": 1.0344827586206898e-05,
"loss": 0.1906,
"step": 15
},
{
"epoch": 0.00829122943386449,
"grad_norm": 1.0903468132019043,
"learning_rate": 1.103448275862069e-05,
"loss": 0.1151,
"step": 16
},
{
"epoch": 0.00880943127348102,
"grad_norm": 1.881485104560852,
"learning_rate": 1.1724137931034483e-05,
"loss": 0.1236,
"step": 17
},
{
"epoch": 0.009327633113097552,
"grad_norm": 1.7081018686294556,
"learning_rate": 1.2413793103448277e-05,
"loss": 0.1583,
"step": 18
},
{
"epoch": 0.009845834952714082,
"grad_norm": 1.3075495958328247,
"learning_rate": 1.310344827586207e-05,
"loss": 0.0965,
"step": 19
},
{
"epoch": 0.010364036792330613,
"grad_norm": 1.3644860982894897,
"learning_rate": 1.3793103448275863e-05,
"loss": 0.0981,
"step": 20
},
{
"epoch": 0.010882238631947143,
"grad_norm": 0.7825227379798889,
"learning_rate": 1.4482758620689657e-05,
"loss": 0.0718,
"step": 21
},
{
"epoch": 0.011400440471563675,
"grad_norm": 0.6305981874465942,
"learning_rate": 1.5172413793103448e-05,
"loss": 0.0577,
"step": 22
},
{
"epoch": 0.011918642311180205,
"grad_norm": 0.3038203716278076,
"learning_rate": 1.586206896551724e-05,
"loss": 0.0714,
"step": 23
},
{
"epoch": 0.012436844150796735,
"grad_norm": 0.2078988254070282,
"learning_rate": 1.6551724137931037e-05,
"loss": 0.0471,
"step": 24
},
{
"epoch": 0.012955045990413265,
"grad_norm": 0.20070360600948334,
"learning_rate": 1.7241379310344828e-05,
"loss": 0.0506,
"step": 25
},
{
"epoch": 0.013473247830029797,
"grad_norm": 0.2870819568634033,
"learning_rate": 1.7931034482758623e-05,
"loss": 0.0459,
"step": 26
},
{
"epoch": 0.013991449669646328,
"grad_norm": 0.5004845857620239,
"learning_rate": 1.8620689655172415e-05,
"loss": 0.0744,
"step": 27
},
{
"epoch": 0.014509651509262858,
"grad_norm": 0.32954147458076477,
"learning_rate": 1.931034482758621e-05,
"loss": 0.0553,
"step": 28
},
{
"epoch": 0.015027853348879388,
"grad_norm": 0.32314950227737427,
"learning_rate": 2e-05,
"loss": 0.0569,
"step": 29
},
{
"epoch": 0.01554605518849592,
"grad_norm": 0.41272085905075073,
"learning_rate": 2.0689655172413797e-05,
"loss": 0.0765,
"step": 30
},
{
"epoch": 0.01606425702811245,
"grad_norm": 0.27000927925109863,
"learning_rate": 2.1379310344827585e-05,
"loss": 0.0629,
"step": 31
},
{
"epoch": 0.01658245886772898,
"grad_norm": 0.28079530596733093,
"learning_rate": 2.206896551724138e-05,
"loss": 0.0569,
"step": 32
},
{
"epoch": 0.017100660707345512,
"grad_norm": 0.2628077268600464,
"learning_rate": 2.2758620689655175e-05,
"loss": 0.0516,
"step": 33
},
{
"epoch": 0.01761886254696204,
"grad_norm": 0.21723352372646332,
"learning_rate": 2.3448275862068967e-05,
"loss": 0.0464,
"step": 34
},
{
"epoch": 0.018137064386578573,
"grad_norm": 0.18918418884277344,
"learning_rate": 2.413793103448276e-05,
"loss": 0.0517,
"step": 35
},
{
"epoch": 0.018655266226195105,
"grad_norm": 0.188095822930336,
"learning_rate": 2.4827586206896553e-05,
"loss": 0.0429,
"step": 36
},
{
"epoch": 0.019173468065811633,
"grad_norm": 0.23603039979934692,
"learning_rate": 2.551724137931035e-05,
"loss": 0.0458,
"step": 37
},
{
"epoch": 0.019691669905428165,
"grad_norm": 0.21862974762916565,
"learning_rate": 2.620689655172414e-05,
"loss": 0.0586,
"step": 38
},
{
"epoch": 0.020209871745044693,
"grad_norm": 0.31196755170822144,
"learning_rate": 2.6896551724137935e-05,
"loss": 0.066,
"step": 39
},
{
"epoch": 0.020728073584661225,
"grad_norm": 0.24622637033462524,
"learning_rate": 2.7586206896551727e-05,
"loss": 0.0592,
"step": 40
},
{
"epoch": 0.021246275424277757,
"grad_norm": 0.3382416367530823,
"learning_rate": 2.8275862068965518e-05,
"loss": 0.0632,
"step": 41
},
{
"epoch": 0.021764477263894286,
"grad_norm": 0.19901607930660248,
"learning_rate": 2.8965517241379313e-05,
"loss": 0.0423,
"step": 42
},
{
"epoch": 0.022282679103510818,
"grad_norm": 0.21523790061473846,
"learning_rate": 2.965517241379311e-05,
"loss": 0.0476,
"step": 43
},
{
"epoch": 0.02280088094312735,
"grad_norm": 0.18428052961826324,
"learning_rate": 3.0344827586206897e-05,
"loss": 0.0515,
"step": 44
},
{
"epoch": 0.023319082782743878,
"grad_norm": 0.16002054512500763,
"learning_rate": 3.103448275862069e-05,
"loss": 0.043,
"step": 45
},
{
"epoch": 0.02383728462236041,
"grad_norm": 0.26238277554512024,
"learning_rate": 3.172413793103448e-05,
"loss": 0.0568,
"step": 46
},
{
"epoch": 0.02435548646197694,
"grad_norm": 0.18134547770023346,
"learning_rate": 3.2413793103448275e-05,
"loss": 0.0456,
"step": 47
},
{
"epoch": 0.02487368830159347,
"grad_norm": 0.1988374888896942,
"learning_rate": 3.310344827586207e-05,
"loss": 0.0491,
"step": 48
},
{
"epoch": 0.025391890141210002,
"grad_norm": 0.13628098368644714,
"learning_rate": 3.3793103448275865e-05,
"loss": 0.0408,
"step": 49
},
{
"epoch": 0.02591009198082653,
"grad_norm": 0.19092515110969543,
"learning_rate": 3.4482758620689657e-05,
"loss": 0.0569,
"step": 50
},
{
"epoch": 0.026428293820443063,
"grad_norm": 0.23275373876094818,
"learning_rate": 3.517241379310345e-05,
"loss": 0.0468,
"step": 51
},
{
"epoch": 0.026946495660059595,
"grad_norm": 0.28532490134239197,
"learning_rate": 3.586206896551725e-05,
"loss": 0.0536,
"step": 52
},
{
"epoch": 0.027464697499676123,
"grad_norm": 0.1914573460817337,
"learning_rate": 3.655172413793104e-05,
"loss": 0.0392,
"step": 53
},
{
"epoch": 0.027982899339292655,
"grad_norm": 0.17403478920459747,
"learning_rate": 3.724137931034483e-05,
"loss": 0.0509,
"step": 54
},
{
"epoch": 0.028501101178909184,
"grad_norm": 0.16115383803844452,
"learning_rate": 3.793103448275862e-05,
"loss": 0.0502,
"step": 55
},
{
"epoch": 0.029019303018525715,
"grad_norm": 0.17531007528305054,
"learning_rate": 3.862068965517242e-05,
"loss": 0.0466,
"step": 56
},
{
"epoch": 0.029537504858142247,
"grad_norm": 0.19580255448818207,
"learning_rate": 3.931034482758621e-05,
"loss": 0.0503,
"step": 57
},
{
"epoch": 0.030055706697758776,
"grad_norm": 0.15387947857379913,
"learning_rate": 4e-05,
"loss": 0.0365,
"step": 58
},
{
"epoch": 0.030573908537375308,
"grad_norm": 0.1713232398033142,
"learning_rate": 3.999997180630039e-05,
"loss": 0.035,
"step": 59
},
{
"epoch": 0.03109211037699184,
"grad_norm": 0.2288016676902771,
"learning_rate": 3.9999887225281024e-05,
"loss": 0.0388,
"step": 60
},
{
"epoch": 0.03161031221660837,
"grad_norm": 0.23962470889091492,
"learning_rate": 3.9999746257180374e-05,
"loss": 0.0483,
"step": 61
},
{
"epoch": 0.0321285140562249,
"grad_norm": 0.1429242640733719,
"learning_rate": 3.9999548902395895e-05,
"loss": 0.038,
"step": 62
},
{
"epoch": 0.03264671589584143,
"grad_norm": 0.2013135850429535,
"learning_rate": 3.999929516148398e-05,
"loss": 0.0515,
"step": 63
},
{
"epoch": 0.03316491773545796,
"grad_norm": 0.14072024822235107,
"learning_rate": 3.999898503516004e-05,
"loss": 0.0413,
"step": 64
},
{
"epoch": 0.03368311957507449,
"grad_norm": 0.201124906539917,
"learning_rate": 3.999861852429842e-05,
"loss": 0.0394,
"step": 65
},
{
"epoch": 0.034201321414691024,
"grad_norm": 0.1803581863641739,
"learning_rate": 3.999819562993246e-05,
"loss": 0.0443,
"step": 66
},
{
"epoch": 0.03471952325430755,
"grad_norm": 0.26812461018562317,
"learning_rate": 3.9997716353254456e-05,
"loss": 0.0505,
"step": 67
},
{
"epoch": 0.03523772509392408,
"grad_norm": 0.2087101936340332,
"learning_rate": 3.999718069561565e-05,
"loss": 0.0491,
"step": 68
},
{
"epoch": 0.03575592693354061,
"grad_norm": 0.145724818110466,
"learning_rate": 3.999658865852628e-05,
"loss": 0.0256,
"step": 69
},
{
"epoch": 0.036274128773157145,
"grad_norm": 0.5044896006584167,
"learning_rate": 3.999594024365551e-05,
"loss": 0.0593,
"step": 70
},
{
"epoch": 0.03679233061277368,
"grad_norm": 0.3319794535636902,
"learning_rate": 3.999523545283146e-05,
"loss": 0.0527,
"step": 71
},
{
"epoch": 0.03731053245239021,
"grad_norm": 0.16803030669689178,
"learning_rate": 3.999447428804119e-05,
"loss": 0.0406,
"step": 72
},
{
"epoch": 0.037828734292006734,
"grad_norm": 0.18300779163837433,
"learning_rate": 3.999365675143071e-05,
"loss": 0.0415,
"step": 73
},
{
"epoch": 0.038346936131623266,
"grad_norm": 0.26276132464408875,
"learning_rate": 3.999278284530498e-05,
"loss": 0.0488,
"step": 74
},
{
"epoch": 0.0388651379712398,
"grad_norm": 0.18061015009880066,
"learning_rate": 3.999185257212782e-05,
"loss": 0.0457,
"step": 75
},
{
"epoch": 0.03938333981085633,
"grad_norm": 0.15918686985969543,
"learning_rate": 3.999086593452205e-05,
"loss": 0.0315,
"step": 76
},
{
"epoch": 0.03990154165047286,
"grad_norm": 0.2396080493927002,
"learning_rate": 3.998982293526935e-05,
"loss": 0.054,
"step": 77
},
{
"epoch": 0.04041974349008939,
"grad_norm": 0.18997341394424438,
"learning_rate": 3.998872357731033e-05,
"loss": 0.0453,
"step": 78
},
{
"epoch": 0.04093794532970592,
"grad_norm": 0.16604867577552795,
"learning_rate": 3.998756786374448e-05,
"loss": 0.0412,
"step": 79
},
{
"epoch": 0.04145614716932245,
"grad_norm": 0.17019790410995483,
"learning_rate": 3.998635579783019e-05,
"loss": 0.0322,
"step": 80
},
{
"epoch": 0.04197434900893898,
"grad_norm": 0.17750458419322968,
"learning_rate": 3.9985087382984716e-05,
"loss": 0.0336,
"step": 81
},
{
"epoch": 0.042492550848555515,
"grad_norm": 0.33200982213020325,
"learning_rate": 3.998376262278419e-05,
"loss": 0.0478,
"step": 82
},
{
"epoch": 0.043010752688172046,
"grad_norm": 0.26823684573173523,
"learning_rate": 3.99823815209636e-05,
"loss": 0.0374,
"step": 83
},
{
"epoch": 0.04352895452778857,
"grad_norm": 0.1756100207567215,
"learning_rate": 3.99809440814168e-05,
"loss": 0.0259,
"step": 84
},
{
"epoch": 0.0440471563674051,
"grad_norm": 0.26577091217041016,
"learning_rate": 3.997945030819644e-05,
"loss": 0.0328,
"step": 85
},
{
"epoch": 0.044565358207021635,
"grad_norm": 0.2032586932182312,
"learning_rate": 3.997790020551403e-05,
"loss": 0.0323,
"step": 86
},
{
"epoch": 0.04508356004663817,
"grad_norm": 0.16400256752967834,
"learning_rate": 3.997629377773988e-05,
"loss": 0.0316,
"step": 87
},
{
"epoch": 0.0456017618862547,
"grad_norm": 0.13048818707466125,
"learning_rate": 3.997463102940311e-05,
"loss": 0.0227,
"step": 88
},
{
"epoch": 0.046119963725871224,
"grad_norm": 0.200350821018219,
"learning_rate": 3.997291196519161e-05,
"loss": 0.0152,
"step": 89
},
{
"epoch": 0.046638165565487756,
"grad_norm": 0.14069367945194244,
"learning_rate": 3.997113658995207e-05,
"loss": 0.0232,
"step": 90
},
{
"epoch": 0.04715636740510429,
"grad_norm": 0.2060389518737793,
"learning_rate": 3.9969304908689934e-05,
"loss": 0.0313,
"step": 91
},
{
"epoch": 0.04767456924472082,
"grad_norm": 0.5540250539779663,
"learning_rate": 3.996741692656938e-05,
"loss": 0.0637,
"step": 92
},
{
"epoch": 0.04819277108433735,
"grad_norm": 0.38743162155151367,
"learning_rate": 3.996547264891332e-05,
"loss": 0.0501,
"step": 93
},
{
"epoch": 0.04871097292395388,
"grad_norm": 0.33079805970191956,
"learning_rate": 3.996347208120341e-05,
"loss": 0.0491,
"step": 94
},
{
"epoch": 0.04922917476357041,
"grad_norm": 0.2402544617652893,
"learning_rate": 3.996141522907998e-05,
"loss": 0.0308,
"step": 95
},
{
"epoch": 0.04974737660318694,
"grad_norm": 0.151661679148674,
"learning_rate": 3.995930209834206e-05,
"loss": 0.0288,
"step": 96
},
{
"epoch": 0.05026557844280347,
"grad_norm": 0.28846490383148193,
"learning_rate": 3.995713269494734e-05,
"loss": 0.0417,
"step": 97
},
{
"epoch": 0.050783780282420005,
"grad_norm": 0.20232760906219482,
"learning_rate": 3.995490702501218e-05,
"loss": 0.0338,
"step": 98
},
{
"epoch": 0.05130198212203654,
"grad_norm": 0.15520133078098297,
"learning_rate": 3.995262509481157e-05,
"loss": 0.0269,
"step": 99
},
{
"epoch": 0.05182018396165306,
"grad_norm": 0.19639192521572113,
"learning_rate": 3.9950286910779106e-05,
"loss": 0.0391,
"step": 100
},
{
"epoch": 0.052338385801269594,
"grad_norm": 0.21983079612255096,
"learning_rate": 3.9947892479507e-05,
"loss": 0.0378,
"step": 101
},
{
"epoch": 0.052856587640886125,
"grad_norm": 0.21260757744312286,
"learning_rate": 3.994544180774603e-05,
"loss": 0.0379,
"step": 102
},
{
"epoch": 0.05337478948050266,
"grad_norm": 0.19602173566818237,
"learning_rate": 3.9942934902405564e-05,
"loss": 0.039,
"step": 103
},
{
"epoch": 0.05389299132011919,
"grad_norm": 0.1708235889673233,
"learning_rate": 3.9940371770553484e-05,
"loss": 0.0379,
"step": 104
},
{
"epoch": 0.054411193159735714,
"grad_norm": 0.18058320879936218,
"learning_rate": 3.99377524194162e-05,
"loss": 0.0337,
"step": 105
},
{
"epoch": 0.054929394999352246,
"grad_norm": 0.20318712294101715,
"learning_rate": 3.9935076856378646e-05,
"loss": 0.0309,
"step": 106
},
{
"epoch": 0.05544759683896878,
"grad_norm": 0.1709553748369217,
"learning_rate": 3.993234508898422e-05,
"loss": 0.0365,
"step": 107
},
{
"epoch": 0.05596579867858531,
"grad_norm": 0.3021198809146881,
"learning_rate": 3.992955712493477e-05,
"loss": 0.043,
"step": 108
},
{
"epoch": 0.05648400051820184,
"grad_norm": 0.20962610840797424,
"learning_rate": 3.9926712972090624e-05,
"loss": 0.0318,
"step": 109
},
{
"epoch": 0.05700220235781837,
"grad_norm": 0.12813960015773773,
"learning_rate": 3.992381263847048e-05,
"loss": 0.035,
"step": 110
},
{
"epoch": 0.0575204041974349,
"grad_norm": 0.16706594824790955,
"learning_rate": 3.992085613225147e-05,
"loss": 0.0339,
"step": 111
},
{
"epoch": 0.05803860603705143,
"grad_norm": 0.21272806823253632,
"learning_rate": 3.991784346176906e-05,
"loss": 0.0394,
"step": 112
},
{
"epoch": 0.05855680787666796,
"grad_norm": 0.11596448719501495,
"learning_rate": 3.99147746355171e-05,
"loss": 0.0333,
"step": 113
},
{
"epoch": 0.059075009716284495,
"grad_norm": 0.19818900525569916,
"learning_rate": 3.991164966214773e-05,
"loss": 0.0294,
"step": 114
},
{
"epoch": 0.05959321155590103,
"grad_norm": 0.4205951690673828,
"learning_rate": 3.990846855047141e-05,
"loss": 0.0628,
"step": 115
},
{
"epoch": 0.06011141339551755,
"grad_norm": 0.23930928111076355,
"learning_rate": 3.9905231309456884e-05,
"loss": 0.045,
"step": 116
},
{
"epoch": 0.060629615235134084,
"grad_norm": 0.3190346658229828,
"learning_rate": 3.9901937948231124e-05,
"loss": 0.044,
"step": 117
},
{
"epoch": 0.061147817074750616,
"grad_norm": 0.21698309481143951,
"learning_rate": 3.989858847607932e-05,
"loss": 0.0417,
"step": 118
},
{
"epoch": 0.06166601891436715,
"grad_norm": 0.14343884587287903,
"learning_rate": 3.9895182902444895e-05,
"loss": 0.0345,
"step": 119
},
{
"epoch": 0.06218422075398368,
"grad_norm": 0.2800877094268799,
"learning_rate": 3.989172123692941e-05,
"loss": 0.0431,
"step": 120
},
{
"epoch": 0.0627024225936002,
"grad_norm": 0.21430440247058868,
"learning_rate": 3.988820348929258e-05,
"loss": 0.0365,
"step": 121
},
{
"epoch": 0.06322062443321674,
"grad_norm": 0.18680894374847412,
"learning_rate": 3.988462966945224e-05,
"loss": 0.0277,
"step": 122
},
{
"epoch": 0.06373882627283327,
"grad_norm": 0.18713238835334778,
"learning_rate": 3.988099978748431e-05,
"loss": 0.0359,
"step": 123
},
{
"epoch": 0.0642570281124498,
"grad_norm": 0.16081608831882477,
"learning_rate": 3.987731385362277e-05,
"loss": 0.0256,
"step": 124
},
{
"epoch": 0.06477522995206633,
"grad_norm": 0.13975271582603455,
"learning_rate": 3.987357187825963e-05,
"loss": 0.0272,
"step": 125
},
{
"epoch": 0.06529343179168286,
"grad_norm": 0.18400432169437408,
"learning_rate": 3.98697738719449e-05,
"loss": 0.0295,
"step": 126
},
{
"epoch": 0.0658116336312994,
"grad_norm": 0.19633497297763824,
"learning_rate": 3.986591984538658e-05,
"loss": 0.0304,
"step": 127
},
{
"epoch": 0.06632983547091592,
"grad_norm": 0.12133795768022537,
"learning_rate": 3.986200980945057e-05,
"loss": 0.0253,
"step": 128
},
{
"epoch": 0.06684803731053245,
"grad_norm": 0.22974886000156403,
"learning_rate": 3.985804377516074e-05,
"loss": 0.0377,
"step": 129
},
{
"epoch": 0.06736623915014898,
"grad_norm": 0.1624593436717987,
"learning_rate": 3.985402175369878e-05,
"loss": 0.0252,
"step": 130
},
{
"epoch": 0.06788444098976551,
"grad_norm": 0.1626574844121933,
"learning_rate": 3.984994375640427e-05,
"loss": 0.03,
"step": 131
},
{
"epoch": 0.06840264282938205,
"grad_norm": 0.17271150648593903,
"learning_rate": 3.984580979477459e-05,
"loss": 0.0288,
"step": 132
},
{
"epoch": 0.06892084466899857,
"grad_norm": 0.38058826327323914,
"learning_rate": 3.984161988046491e-05,
"loss": 0.0392,
"step": 133
},
{
"epoch": 0.0694390465086151,
"grad_norm": 0.14003795385360718,
"learning_rate": 3.983737402528815e-05,
"loss": 0.0278,
"step": 134
},
{
"epoch": 0.06995724834823164,
"grad_norm": 0.22751788794994354,
"learning_rate": 3.983307224121494e-05,
"loss": 0.0273,
"step": 135
},
{
"epoch": 0.07047545018784816,
"grad_norm": 0.22172273695468903,
"learning_rate": 3.9828714540373605e-05,
"loss": 0.0306,
"step": 136
},
{
"epoch": 0.0709936520274647,
"grad_norm": 0.24642212688922882,
"learning_rate": 3.982430093505011e-05,
"loss": 0.0421,
"step": 137
},
{
"epoch": 0.07151185386708123,
"grad_norm": 0.15056243538856506,
"learning_rate": 3.9819831437688046e-05,
"loss": 0.029,
"step": 138
},
{
"epoch": 0.07203005570669777,
"grad_norm": 0.18077677488327026,
"learning_rate": 3.9815306060888585e-05,
"loss": 0.0306,
"step": 139
},
{
"epoch": 0.07254825754631429,
"grad_norm": 0.33231616020202637,
"learning_rate": 3.981072481741043e-05,
"loss": 0.0463,
"step": 140
},
{
"epoch": 0.07306645938593082,
"grad_norm": 0.4319797158241272,
"learning_rate": 3.980608772016981e-05,
"loss": 0.0507,
"step": 141
},
{
"epoch": 0.07358466122554735,
"grad_norm": 0.20595696568489075,
"learning_rate": 3.980139478224041e-05,
"loss": 0.0353,
"step": 142
},
{
"epoch": 0.07410286306516388,
"grad_norm": 0.21615833044052124,
"learning_rate": 3.979664601685336e-05,
"loss": 0.0386,
"step": 143
},
{
"epoch": 0.07462106490478042,
"grad_norm": 0.11346781253814697,
"learning_rate": 3.979184143739718e-05,
"loss": 0.0219,
"step": 144
},
{
"epoch": 0.07513926674439694,
"grad_norm": 0.4296473264694214,
"learning_rate": 3.978698105741777e-05,
"loss": 0.0483,
"step": 145
},
{
"epoch": 0.07565746858401347,
"grad_norm": 0.18602745234966278,
"learning_rate": 3.978206489061833e-05,
"loss": 0.0318,
"step": 146
},
{
"epoch": 0.07617567042363001,
"grad_norm": 0.1680869311094284,
"learning_rate": 3.9777092950859367e-05,
"loss": 0.0312,
"step": 147
},
{
"epoch": 0.07669387226324653,
"grad_norm": 0.29125967621803284,
"learning_rate": 3.9772065252158594e-05,
"loss": 0.0343,
"step": 148
},
{
"epoch": 0.07721207410286307,
"grad_norm": 0.19228827953338623,
"learning_rate": 3.976698180869097e-05,
"loss": 0.0336,
"step": 149
},
{
"epoch": 0.0777302759424796,
"grad_norm": 0.13979323208332062,
"learning_rate": 3.9761842634788606e-05,
"loss": 0.0217,
"step": 150
},
{
"epoch": 0.07824847778209612,
"grad_norm": 0.23602689802646637,
"learning_rate": 3.975664774494073e-05,
"loss": 0.0435,
"step": 151
},
{
"epoch": 0.07876667962171266,
"grad_norm": 0.19915208220481873,
"learning_rate": 3.975139715379364e-05,
"loss": 0.0248,
"step": 152
},
{
"epoch": 0.07928488146132918,
"grad_norm": 0.15488211810588837,
"learning_rate": 3.9746090876150736e-05,
"loss": 0.0226,
"step": 153
},
{
"epoch": 0.07980308330094572,
"grad_norm": 0.27361470460891724,
"learning_rate": 3.974072892697234e-05,
"loss": 0.0412,
"step": 154
},
{
"epoch": 0.08032128514056225,
"grad_norm": 0.12932442128658295,
"learning_rate": 3.973531132137579e-05,
"loss": 0.0214,
"step": 155
},
{
"epoch": 0.08083948698017877,
"grad_norm": 0.13963568210601807,
"learning_rate": 3.972983807463531e-05,
"loss": 0.0242,
"step": 156
},
{
"epoch": 0.08135768881979531,
"grad_norm": 0.22062024474143982,
"learning_rate": 3.9724309202182014e-05,
"loss": 0.0303,
"step": 157
},
{
"epoch": 0.08187589065941184,
"grad_norm": 0.3001038730144501,
"learning_rate": 3.9718724719603836e-05,
"loss": 0.0353,
"step": 158
},
{
"epoch": 0.08239409249902838,
"grad_norm": 0.2961874008178711,
"learning_rate": 3.9713084642645504e-05,
"loss": 0.0298,
"step": 159
},
{
"epoch": 0.0829122943386449,
"grad_norm": 0.30271297693252563,
"learning_rate": 3.970738898720847e-05,
"loss": 0.0353,
"step": 160
},
{
"epoch": 0.08343049617826143,
"grad_norm": 0.18537764251232147,
"learning_rate": 3.9701637769350906e-05,
"loss": 0.0276,
"step": 161
},
{
"epoch": 0.08394869801787797,
"grad_norm": 0.45981863141059875,
"learning_rate": 3.969583100528762e-05,
"loss": 0.0332,
"step": 162
},
{
"epoch": 0.08446689985749449,
"grad_norm": 0.3007035255432129,
"learning_rate": 3.968996871139002e-05,
"loss": 0.0417,
"step": 163
},
{
"epoch": 0.08498510169711103,
"grad_norm": 0.2484988421201706,
"learning_rate": 3.9684050904186094e-05,
"loss": 0.0332,
"step": 164
},
{
"epoch": 0.08550330353672755,
"grad_norm": 0.1978878676891327,
"learning_rate": 3.9678077600360316e-05,
"loss": 0.0267,
"step": 165
},
{
"epoch": 0.08602150537634409,
"grad_norm": 0.14420898258686066,
"learning_rate": 3.9672048816753654e-05,
"loss": 0.0224,
"step": 166
},
{
"epoch": 0.08653970721596062,
"grad_norm": 0.2479882538318634,
"learning_rate": 3.966596457036347e-05,
"loss": 0.0317,
"step": 167
},
{
"epoch": 0.08705790905557714,
"grad_norm": 0.2385600358247757,
"learning_rate": 3.96598248783435e-05,
"loss": 0.0362,
"step": 168
},
{
"epoch": 0.08757611089519368,
"grad_norm": 0.30355146527290344,
"learning_rate": 3.965362975800382e-05,
"loss": 0.0372,
"step": 169
},
{
"epoch": 0.0880943127348102,
"grad_norm": 0.440682590007782,
"learning_rate": 3.964737922681077e-05,
"loss": 0.0434,
"step": 170
},
{
"epoch": 0.08861251457442675,
"grad_norm": 0.1279064565896988,
"learning_rate": 3.964107330238689e-05,
"loss": 0.0164,
"step": 171
},
{
"epoch": 0.08913071641404327,
"grad_norm": 0.2853245735168457,
"learning_rate": 3.9634712002510925e-05,
"loss": 0.0453,
"step": 172
},
{
"epoch": 0.0896489182536598,
"grad_norm": 0.20429256558418274,
"learning_rate": 3.962829534511774e-05,
"loss": 0.024,
"step": 173
},
{
"epoch": 0.09016712009327633,
"grad_norm": 0.4442828595638275,
"learning_rate": 3.962182334829825e-05,
"loss": 0.0525,
"step": 174
},
{
"epoch": 0.09068532193289286,
"grad_norm": 0.2360175997018814,
"learning_rate": 3.961529603029942e-05,
"loss": 0.0282,
"step": 175
},
{
"epoch": 0.0912035237725094,
"grad_norm": 0.09913711994886398,
"learning_rate": 3.9608713409524175e-05,
"loss": 0.0217,
"step": 176
},
{
"epoch": 0.09172172561212592,
"grad_norm": 0.18438151478767395,
"learning_rate": 3.9602075504531356e-05,
"loss": 0.0312,
"step": 177
},
{
"epoch": 0.09223992745174245,
"grad_norm": 0.23954534530639648,
"learning_rate": 3.959538233403567e-05,
"loss": 0.0206,
"step": 178
},
{
"epoch": 0.09275812929135899,
"grad_norm": 0.28380924463272095,
"learning_rate": 3.9588633916907635e-05,
"loss": 0.0388,
"step": 179
},
{
"epoch": 0.09327633113097551,
"grad_norm": 0.16261892020702362,
"learning_rate": 3.9581830272173556e-05,
"loss": 0.0191,
"step": 180
},
{
"epoch": 0.09379453297059205,
"grad_norm": 0.1379668265581131,
"learning_rate": 3.95749714190154e-05,
"loss": 0.0289,
"step": 181
},
{
"epoch": 0.09431273481020858,
"grad_norm": 0.29320958256721497,
"learning_rate": 3.9568057376770826e-05,
"loss": 0.0227,
"step": 182
},
{
"epoch": 0.0948309366498251,
"grad_norm": 0.3081492781639099,
"learning_rate": 3.956108816493309e-05,
"loss": 0.0388,
"step": 183
},
{
"epoch": 0.09534913848944164,
"grad_norm": 0.14883936941623688,
"learning_rate": 3.955406380315094e-05,
"loss": 0.0262,
"step": 184
},
{
"epoch": 0.09586734032905816,
"grad_norm": 0.18635591864585876,
"learning_rate": 3.954698431122868e-05,
"loss": 0.0204,
"step": 185
},
{
"epoch": 0.0963855421686747,
"grad_norm": 0.2801106572151184,
"learning_rate": 3.953984970912601e-05,
"loss": 0.0223,
"step": 186
},
{
"epoch": 0.09690374400829123,
"grad_norm": 0.4112246334552765,
"learning_rate": 3.953266001695802e-05,
"loss": 0.0434,
"step": 187
},
{
"epoch": 0.09742194584790775,
"grad_norm": 0.19749994575977325,
"learning_rate": 3.95254152549951e-05,
"loss": 0.0266,
"step": 188
},
{
"epoch": 0.09794014768752429,
"grad_norm": 0.1635226160287857,
"learning_rate": 3.951811544366292e-05,
"loss": 0.0216,
"step": 189
},
{
"epoch": 0.09845834952714082,
"grad_norm": 0.15637846291065216,
"learning_rate": 3.9510760603542354e-05,
"loss": 0.0294,
"step": 190
},
{
"epoch": 0.09897655136675736,
"grad_norm": 0.376310259103775,
"learning_rate": 3.950335075536941e-05,
"loss": 0.0426,
"step": 191
},
{
"epoch": 0.09949475320637388,
"grad_norm": 0.32836809754371643,
"learning_rate": 3.949588592003518e-05,
"loss": 0.023,
"step": 192
},
{
"epoch": 0.1000129550459904,
"grad_norm": 0.14829964935779572,
"learning_rate": 3.948836611858582e-05,
"loss": 0.0296,
"step": 193
},
{
"epoch": 0.10053115688560695,
"grad_norm": 0.13914668560028076,
"learning_rate": 3.948079137222242e-05,
"loss": 0.0209,
"step": 194
},
{
"epoch": 0.10104935872522347,
"grad_norm": 0.28442588448524475,
"learning_rate": 3.9473161702301e-05,
"loss": 0.0308,
"step": 195
},
{
"epoch": 0.10156756056484001,
"grad_norm": 0.1649635136127472,
"learning_rate": 3.9465477130332405e-05,
"loss": 0.0183,
"step": 196
},
{
"epoch": 0.10208576240445653,
"grad_norm": 0.22778348624706268,
"learning_rate": 3.945773767798231e-05,
"loss": 0.0298,
"step": 197
},
{
"epoch": 0.10260396424407307,
"grad_norm": 0.5452020764350891,
"learning_rate": 3.944994336707107e-05,
"loss": 0.0299,
"step": 198
},
{
"epoch": 0.1031221660836896,
"grad_norm": 0.20926618576049805,
"learning_rate": 3.944209421957375e-05,
"loss": 0.0268,
"step": 199
},
{
"epoch": 0.10364036792330612,
"grad_norm": 0.15481847524642944,
"learning_rate": 3.943419025762e-05,
"loss": 0.021,
"step": 200
},
{
"epoch": 0.10415856976292266,
"grad_norm": 0.8367764353752136,
"learning_rate": 3.942623150349399e-05,
"loss": 0.0549,
"step": 201
},
{
"epoch": 0.10467677160253919,
"grad_norm": 0.37851259112358093,
"learning_rate": 3.9418217979634425e-05,
"loss": 0.0382,
"step": 202
},
{
"epoch": 0.10519497344215573,
"grad_norm": 0.506738543510437,
"learning_rate": 3.941014970863437e-05,
"loss": 0.0491,
"step": 203
},
{
"epoch": 0.10571317528177225,
"grad_norm": 0.37913596630096436,
"learning_rate": 3.940202671324127e-05,
"loss": 0.0275,
"step": 204
},
{
"epoch": 0.10623137712138878,
"grad_norm": 0.20694205164909363,
"learning_rate": 3.9393849016356866e-05,
"loss": 0.0182,
"step": 205
},
{
"epoch": 0.10674957896100531,
"grad_norm": 0.2608841061592102,
"learning_rate": 3.9385616641037104e-05,
"loss": 0.0296,
"step": 206
},
{
"epoch": 0.10726778080062184,
"grad_norm": 0.18457987904548645,
"learning_rate": 3.9377329610492084e-05,
"loss": 0.0202,
"step": 207
},
{
"epoch": 0.10778598264023838,
"grad_norm": 0.34302592277526855,
"learning_rate": 3.9368987948086024e-05,
"loss": 0.0348,
"step": 208
},
{
"epoch": 0.1083041844798549,
"grad_norm": 0.25883880257606506,
"learning_rate": 3.9360591677337166e-05,
"loss": 0.0264,
"step": 209
},
{
"epoch": 0.10882238631947143,
"grad_norm": 0.1379159390926361,
"learning_rate": 3.935214082191768e-05,
"loss": 0.0198,
"step": 210
},
{
"epoch": 0.10934058815908797,
"grad_norm": 0.1821633279323578,
"learning_rate": 3.934363540565366e-05,
"loss": 0.0215,
"step": 211
},
{
"epoch": 0.10985878999870449,
"grad_norm": 0.18341860175132751,
"learning_rate": 3.933507545252504e-05,
"loss": 0.0209,
"step": 212
},
{
"epoch": 0.11037699183832103,
"grad_norm": 0.22454297542572021,
"learning_rate": 3.9326460986665476e-05,
"loss": 0.0294,
"step": 213
},
{
"epoch": 0.11089519367793756,
"grad_norm": 0.16194742918014526,
"learning_rate": 3.931779203236233e-05,
"loss": 0.0254,
"step": 214
},
{
"epoch": 0.11141339551755408,
"grad_norm": 0.47000735998153687,
"learning_rate": 3.9309068614056615e-05,
"loss": 0.048,
"step": 215
},
{
"epoch": 0.11193159735717062,
"grad_norm": 0.28025323152542114,
"learning_rate": 3.930029075634286e-05,
"loss": 0.0253,
"step": 216
},
{
"epoch": 0.11244979919678715,
"grad_norm": 0.5879072546958923,
"learning_rate": 3.9291458483969086e-05,
"loss": 0.062,
"step": 217
},
{
"epoch": 0.11296800103640368,
"grad_norm": 0.18934841454029083,
"learning_rate": 3.928257182183674e-05,
"loss": 0.0318,
"step": 218
},
{
"epoch": 0.11348620287602021,
"grad_norm": 0.21203094720840454,
"learning_rate": 3.9273630795000626e-05,
"loss": 0.026,
"step": 219
},
{
"epoch": 0.11400440471563673,
"grad_norm": 0.25133925676345825,
"learning_rate": 3.926463542866879e-05,
"loss": 0.0183,
"step": 220
},
{
"epoch": 0.11452260655525327,
"grad_norm": 0.25150206685066223,
"learning_rate": 3.92555857482025e-05,
"loss": 0.0223,
"step": 221
},
{
"epoch": 0.1150408083948698,
"grad_norm": 0.20962639153003693,
"learning_rate": 3.924648177911616e-05,
"loss": 0.0309,
"step": 222
},
{
"epoch": 0.11555901023448634,
"grad_norm": 0.21745115518569946,
"learning_rate": 3.9237323547077216e-05,
"loss": 0.0257,
"step": 223
},
{
"epoch": 0.11607721207410286,
"grad_norm": 0.2105790674686432,
"learning_rate": 3.922811107790613e-05,
"loss": 0.0251,
"step": 224
},
{
"epoch": 0.11659541391371939,
"grad_norm": 0.1776510328054428,
"learning_rate": 3.921884439757624e-05,
"loss": 0.0211,
"step": 225
},
{
"epoch": 0.11711361575333593,
"grad_norm": 0.4401101768016815,
"learning_rate": 3.920952353221376e-05,
"loss": 0.0413,
"step": 226
},
{
"epoch": 0.11763181759295245,
"grad_norm": 0.33000439405441284,
"learning_rate": 3.9200148508097656e-05,
"loss": 0.0378,
"step": 227
},
{
"epoch": 0.11815001943256899,
"grad_norm": 0.22733037173748016,
"learning_rate": 3.919071935165958e-05,
"loss": 0.0309,
"step": 228
},
{
"epoch": 0.11866822127218551,
"grad_norm": 0.27271509170532227,
"learning_rate": 3.918123608948382e-05,
"loss": 0.0332,
"step": 229
},
{
"epoch": 0.11918642311180205,
"grad_norm": 0.28250646591186523,
"learning_rate": 3.91716987483072e-05,
"loss": 0.0274,
"step": 230
},
{
"epoch": 0.11970462495141858,
"grad_norm": 0.18986068665981293,
"learning_rate": 3.916210735501902e-05,
"loss": 0.0188,
"step": 231
},
{
"epoch": 0.1202228267910351,
"grad_norm": 0.13927146792411804,
"learning_rate": 3.915246193666096e-05,
"loss": 0.0202,
"step": 232
},
{
"epoch": 0.12074102863065164,
"grad_norm": 0.31036004424095154,
"learning_rate": 3.914276252042702e-05,
"loss": 0.024,
"step": 233
},
{
"epoch": 0.12125923047026817,
"grad_norm": 0.21623092889785767,
"learning_rate": 3.913300913366344e-05,
"loss": 0.024,
"step": 234
},
{
"epoch": 0.1217774323098847,
"grad_norm": 0.17337322235107422,
"learning_rate": 3.9123201803868644e-05,
"loss": 0.0162,
"step": 235
},
{
"epoch": 0.12229563414950123,
"grad_norm": 0.2597239315509796,
"learning_rate": 3.911334055869309e-05,
"loss": 0.0277,
"step": 236
},
{
"epoch": 0.12281383598911776,
"grad_norm": 0.23021075129508972,
"learning_rate": 3.910342542593932e-05,
"loss": 0.0263,
"step": 237
},
{
"epoch": 0.1233320378287343,
"grad_norm": 0.38447242975234985,
"learning_rate": 3.909345643356172e-05,
"loss": 0.0293,
"step": 238
},
{
"epoch": 0.12385023966835082,
"grad_norm": 0.20466025173664093,
"learning_rate": 3.908343360966659e-05,
"loss": 0.021,
"step": 239
},
{
"epoch": 0.12436844150796736,
"grad_norm": 0.35317835211753845,
"learning_rate": 3.9073356982511975e-05,
"loss": 0.0427,
"step": 240
},
{
"epoch": 0.12488664334758388,
"grad_norm": 0.5810201168060303,
"learning_rate": 3.906322658050761e-05,
"loss": 0.05,
"step": 241
},
{
"epoch": 0.1254048451872004,
"grad_norm": 0.3113636374473572,
"learning_rate": 3.905304243221485e-05,
"loss": 0.0313,
"step": 242
},
{
"epoch": 0.12592304702681695,
"grad_norm": 0.1605708748102188,
"learning_rate": 3.904280456634659e-05,
"loss": 0.0222,
"step": 243
},
{
"epoch": 0.1264412488664335,
"grad_norm": 0.35602137446403503,
"learning_rate": 3.9032513011767136e-05,
"loss": 0.0224,
"step": 244
},
{
"epoch": 0.12695945070605,
"grad_norm": 0.42479732632637024,
"learning_rate": 3.90221677974922e-05,
"loss": 0.0314,
"step": 245
},
{
"epoch": 0.12747765254566654,
"grad_norm": 0.24160167574882507,
"learning_rate": 3.901176895268877e-05,
"loss": 0.0236,
"step": 246
},
{
"epoch": 0.12799585438528308,
"grad_norm": 0.5197824835777283,
"learning_rate": 3.900131650667504e-05,
"loss": 0.0314,
"step": 247
},
{
"epoch": 0.1285140562248996,
"grad_norm": 0.1657523363828659,
"learning_rate": 3.899081048892031e-05,
"loss": 0.0207,
"step": 248
},
{
"epoch": 0.12903225806451613,
"grad_norm": 0.21127645671367645,
"learning_rate": 3.898025092904494e-05,
"loss": 0.0259,
"step": 249
},
{
"epoch": 0.12955045990413266,
"grad_norm": 0.19938185811042786,
"learning_rate": 3.896963785682024e-05,
"loss": 0.024,
"step": 250
},
{
"epoch": 0.1300686617437492,
"grad_norm": 0.6112692952156067,
"learning_rate": 3.895897130216838e-05,
"loss": 0.0594,
"step": 251
},
{
"epoch": 0.13058686358336571,
"grad_norm": 0.29001185297966003,
"learning_rate": 3.894825129516232e-05,
"loss": 0.0308,
"step": 252
},
{
"epoch": 0.13110506542298225,
"grad_norm": 0.12033861130475998,
"learning_rate": 3.893747786602573e-05,
"loss": 0.0173,
"step": 253
},
{
"epoch": 0.1316232672625988,
"grad_norm": 0.21433091163635254,
"learning_rate": 3.89266510451329e-05,
"loss": 0.0299,
"step": 254
},
{
"epoch": 0.1321414691022153,
"grad_norm": 0.2604752779006958,
"learning_rate": 3.8915770863008635e-05,
"loss": 0.0209,
"step": 255
},
{
"epoch": 0.13265967094183184,
"grad_norm": 0.2170797735452652,
"learning_rate": 3.8904837350328194e-05,
"loss": 0.0221,
"step": 256
},
{
"epoch": 0.13317787278144838,
"grad_norm": 0.23096159100532532,
"learning_rate": 3.88938505379172e-05,
"loss": 0.0337,
"step": 257
},
{
"epoch": 0.1336960746210649,
"grad_norm": 0.1898171603679657,
"learning_rate": 3.888281045675153e-05,
"loss": 0.0211,
"step": 258
},
{
"epoch": 0.13421427646068143,
"grad_norm": 0.18458768725395203,
"learning_rate": 3.887171713795727e-05,
"loss": 0.0211,
"step": 259
},
{
"epoch": 0.13473247830029797,
"grad_norm": 0.17833667993545532,
"learning_rate": 3.886057061281058e-05,
"loss": 0.0318,
"step": 260
},
{
"epoch": 0.1352506801399145,
"grad_norm": 0.1271020621061325,
"learning_rate": 3.8849370912737644e-05,
"loss": 0.0161,
"step": 261
},
{
"epoch": 0.13576888197953102,
"grad_norm": 0.23673047125339508,
"learning_rate": 3.883811806931456e-05,
"loss": 0.038,
"step": 262
},
{
"epoch": 0.13628708381914756,
"grad_norm": 0.16474123299121857,
"learning_rate": 3.8826812114267264e-05,
"loss": 0.0258,
"step": 263
},
{
"epoch": 0.1368052856587641,
"grad_norm": 0.2002090960741043,
"learning_rate": 3.881545307947141e-05,
"loss": 0.0238,
"step": 264
},
{
"epoch": 0.1373234874983806,
"grad_norm": 0.19066838920116425,
"learning_rate": 3.880404099695232e-05,
"loss": 0.0214,
"step": 265
},
{
"epoch": 0.13784168933799715,
"grad_norm": 0.245769664645195,
"learning_rate": 3.879257589888489e-05,
"loss": 0.0345,
"step": 266
},
{
"epoch": 0.1383598911776137,
"grad_norm": 0.27019011974334717,
"learning_rate": 3.878105781759347e-05,
"loss": 0.0387,
"step": 267
},
{
"epoch": 0.1388780930172302,
"grad_norm": 0.20554135739803314,
"learning_rate": 3.876948678555179e-05,
"loss": 0.0209,
"step": 268
},
{
"epoch": 0.13939629485684674,
"grad_norm": 0.273275762796402,
"learning_rate": 3.875786283538287e-05,
"loss": 0.0209,
"step": 269
},
{
"epoch": 0.13991449669646328,
"grad_norm": 0.1820681393146515,
"learning_rate": 3.8746185999858925e-05,
"loss": 0.0235,
"step": 270
},
{
"epoch": 0.14043269853607981,
"grad_norm": 0.1960431933403015,
"learning_rate": 3.873445631190127e-05,
"loss": 0.0217,
"step": 271
},
{
"epoch": 0.14095090037569633,
"grad_norm": 0.3486228883266449,
"learning_rate": 3.872267380458024e-05,
"loss": 0.0279,
"step": 272
},
{
"epoch": 0.14146910221531286,
"grad_norm": 0.25267505645751953,
"learning_rate": 3.871083851111508e-05,
"loss": 0.0269,
"step": 273
},
{
"epoch": 0.1419873040549294,
"grad_norm": 0.26706644892692566,
"learning_rate": 3.8698950464873874e-05,
"loss": 0.0253,
"step": 274
},
{
"epoch": 0.14250550589454591,
"grad_norm": 0.271297812461853,
"learning_rate": 3.86870096993734e-05,
"loss": 0.0362,
"step": 275
},
{
"epoch": 0.14302370773416245,
"grad_norm": 0.3143746554851532,
"learning_rate": 3.867501624827911e-05,
"loss": 0.0286,
"step": 276
},
{
"epoch": 0.143541909573779,
"grad_norm": 0.28915753960609436,
"learning_rate": 3.866297014540497e-05,
"loss": 0.0223,
"step": 277
},
{
"epoch": 0.14406011141339553,
"grad_norm": 0.24351872503757477,
"learning_rate": 3.8650871424713406e-05,
"loss": 0.0204,
"step": 278
},
{
"epoch": 0.14457831325301204,
"grad_norm": 0.22745080292224884,
"learning_rate": 3.863872012031519e-05,
"loss": 0.0201,
"step": 279
},
{
"epoch": 0.14509651509262858,
"grad_norm": 0.17342780530452728,
"learning_rate": 3.8626516266469346e-05,
"loss": 0.0154,
"step": 280
},
{
"epoch": 0.14561471693224512,
"grad_norm": 0.32193371653556824,
"learning_rate": 3.8614259897583035e-05,
"loss": 0.0296,
"step": 281
},
{
"epoch": 0.14613291877186163,
"grad_norm": 0.23850269615650177,
"learning_rate": 3.8601951048211516e-05,
"loss": 0.0382,
"step": 282
},
{
"epoch": 0.14665112061147817,
"grad_norm": 0.13823804259300232,
"learning_rate": 3.8589589753057986e-05,
"loss": 0.0142,
"step": 283
},
{
"epoch": 0.1471693224510947,
"grad_norm": 0.5433732271194458,
"learning_rate": 3.85771760469735e-05,
"loss": 0.0404,
"step": 284
},
{
"epoch": 0.14768752429071122,
"grad_norm": 0.43869102001190186,
"learning_rate": 3.856470996495689e-05,
"loss": 0.0531,
"step": 285
},
{
"epoch": 0.14820572613032776,
"grad_norm": 0.19134990870952606,
"learning_rate": 3.855219154215466e-05,
"loss": 0.0205,
"step": 286
},
{
"epoch": 0.1487239279699443,
"grad_norm": 0.35158365964889526,
"learning_rate": 3.8539620813860875e-05,
"loss": 0.0283,
"step": 287
},
{
"epoch": 0.14924212980956084,
"grad_norm": 0.18466360867023468,
"learning_rate": 3.8526997815517064e-05,
"loss": 0.02,
"step": 288
},
{
"epoch": 0.14976033164917735,
"grad_norm": 0.1774568259716034,
"learning_rate": 3.851432258271213e-05,
"loss": 0.023,
"step": 289
},
{
"epoch": 0.1502785334887939,
"grad_norm": 0.10834086686372757,
"learning_rate": 3.850159515118224e-05,
"loss": 0.0119,
"step": 290
},
{
"epoch": 0.15079673532841043,
"grad_norm": 0.3568941056728363,
"learning_rate": 3.8488815556810746e-05,
"loss": 0.035,
"step": 291
},
{
"epoch": 0.15131493716802694,
"grad_norm": 0.09321002662181854,
"learning_rate": 3.847598383562803e-05,
"loss": 0.0167,
"step": 292
},
{
"epoch": 0.15183313900764348,
"grad_norm": 0.24882569909095764,
"learning_rate": 3.846310002381148e-05,
"loss": 0.032,
"step": 293
},
{
"epoch": 0.15235134084726001,
"grad_norm": 0.4001385271549225,
"learning_rate": 3.845016415768532e-05,
"loss": 0.0365,
"step": 294
},
{
"epoch": 0.15286954268687653,
"grad_norm": 0.08750130981206894,
"learning_rate": 3.8437176273720546e-05,
"loss": 0.0161,
"step": 295
},
{
"epoch": 0.15338774452649306,
"grad_norm": 0.158023864030838,
"learning_rate": 3.8424136408534814e-05,
"loss": 0.0181,
"step": 296
},
{
"epoch": 0.1539059463661096,
"grad_norm": 0.12722338736057281,
"learning_rate": 3.841104459889232e-05,
"loss": 0.0193,
"step": 297
},
{
"epoch": 0.15442414820572614,
"grad_norm": 0.16069327294826508,
"learning_rate": 3.839790088170371e-05,
"loss": 0.0202,
"step": 298
},
{
"epoch": 0.15494235004534265,
"grad_norm": 0.16501739621162415,
"learning_rate": 3.838470529402599e-05,
"loss": 0.0223,
"step": 299
},
{
"epoch": 0.1554605518849592,
"grad_norm": 0.1982022225856781,
"learning_rate": 3.8371457873062425e-05,
"loss": 0.0207,
"step": 300
},
{
"epoch": 0.15597875372457573,
"grad_norm": 0.22107556462287903,
"learning_rate": 3.835815865616237e-05,
"loss": 0.0371,
"step": 301
},
{
"epoch": 0.15649695556419224,
"grad_norm": 0.2059648483991623,
"learning_rate": 3.834480768082125e-05,
"loss": 0.0198,
"step": 302
},
{
"epoch": 0.15701515740380878,
"grad_norm": 0.24241840839385986,
"learning_rate": 3.83314049846804e-05,
"loss": 0.0216,
"step": 303
},
{
"epoch": 0.15753335924342532,
"grad_norm": 0.1592680811882019,
"learning_rate": 3.831795060552698e-05,
"loss": 0.0274,
"step": 304
},
{
"epoch": 0.15805156108304186,
"grad_norm": 0.16157729923725128,
"learning_rate": 3.830444458129386e-05,
"loss": 0.0197,
"step": 305
},
{
"epoch": 0.15856976292265837,
"grad_norm": 0.1381078064441681,
"learning_rate": 3.829088695005952e-05,
"loss": 0.028,
"step": 306
},
{
"epoch": 0.1590879647622749,
"grad_norm": 0.1145080178976059,
"learning_rate": 3.827727775004794e-05,
"loss": 0.0151,
"step": 307
},
{
"epoch": 0.15960616660189145,
"grad_norm": 0.12582014501094818,
"learning_rate": 3.8263617019628495e-05,
"loss": 0.017,
"step": 308
},
{
"epoch": 0.16012436844150796,
"grad_norm": 0.11775611340999603,
"learning_rate": 3.8249904797315825e-05,
"loss": 0.021,
"step": 309
},
{
"epoch": 0.1606425702811245,
"grad_norm": 0.18558956682682037,
"learning_rate": 3.823614112176977e-05,
"loss": 0.0229,
"step": 310
},
{
"epoch": 0.16116077212074104,
"grad_norm": 0.2732170820236206,
"learning_rate": 3.822232603179521e-05,
"loss": 0.0351,
"step": 311
},
{
"epoch": 0.16167897396035755,
"grad_norm": 0.36403414607048035,
"learning_rate": 3.820845956634201e-05,
"loss": 0.0217,
"step": 312
},
{
"epoch": 0.16219717579997409,
"grad_norm": 0.16710692644119263,
"learning_rate": 3.819454176450486e-05,
"loss": 0.0211,
"step": 313
},
{
"epoch": 0.16271537763959062,
"grad_norm": 0.11457782238721848,
"learning_rate": 3.818057266552319e-05,
"loss": 0.0265,
"step": 314
},
{
"epoch": 0.16323357947920716,
"grad_norm": 0.2501565217971802,
"learning_rate": 3.816655230878106e-05,
"loss": 0.0323,
"step": 315
},
{
"epoch": 0.16375178131882367,
"grad_norm": 0.13976308703422546,
"learning_rate": 3.815248073380704e-05,
"loss": 0.0215,
"step": 316
},
{
"epoch": 0.1642699831584402,
"grad_norm": 0.14425045251846313,
"learning_rate": 3.8138357980274106e-05,
"loss": 0.0209,
"step": 317
},
{
"epoch": 0.16478818499805675,
"grad_norm": 0.17892557382583618,
"learning_rate": 3.812418408799953e-05,
"loss": 0.0249,
"step": 318
},
{
"epoch": 0.16530638683767326,
"grad_norm": 0.14994150400161743,
"learning_rate": 3.810995909694476e-05,
"loss": 0.0212,
"step": 319
},
{
"epoch": 0.1658245886772898,
"grad_norm": 0.2975420355796814,
"learning_rate": 3.80956830472153e-05,
"loss": 0.0269,
"step": 320
},
{
"epoch": 0.16634279051690634,
"grad_norm": 0.16391150653362274,
"learning_rate": 3.808135597906061e-05,
"loss": 0.0173,
"step": 321
},
{
"epoch": 0.16686099235652285,
"grad_norm": 0.4943293631076813,
"learning_rate": 3.8066977932874014e-05,
"loss": 0.0466,
"step": 322
},
{
"epoch": 0.1673791941961394,
"grad_norm": 0.2048560231924057,
"learning_rate": 3.8052548949192534e-05,
"loss": 0.0213,
"step": 323
},
{
"epoch": 0.16789739603575593,
"grad_norm": 0.3236941397190094,
"learning_rate": 3.803806906869682e-05,
"loss": 0.0385,
"step": 324
},
{
"epoch": 0.16841559787537247,
"grad_norm": 0.10196578502655029,
"learning_rate": 3.8023538332210994e-05,
"loss": 0.0137,
"step": 325
},
{
"epoch": 0.16893379971498898,
"grad_norm": 0.1597544252872467,
"learning_rate": 3.80089567807026e-05,
"loss": 0.0226,
"step": 326
},
{
"epoch": 0.16945200155460552,
"grad_norm": 0.12884768843650818,
"learning_rate": 3.799432445528241e-05,
"loss": 0.0193,
"step": 327
},
{
"epoch": 0.16997020339422206,
"grad_norm": 0.12846127152442932,
"learning_rate": 3.797964139720437e-05,
"loss": 0.0194,
"step": 328
},
{
"epoch": 0.17048840523383857,
"grad_norm": 0.2257414311170578,
"learning_rate": 3.796490764786545e-05,
"loss": 0.0258,
"step": 329
},
{
"epoch": 0.1710066070734551,
"grad_norm": 0.1655774712562561,
"learning_rate": 3.795012324880554e-05,
"loss": 0.0261,
"step": 330
},
{
"epoch": 0.17152480891307165,
"grad_norm": 0.8630039095878601,
"learning_rate": 3.793528824170733e-05,
"loss": 0.0519,
"step": 331
},
{
"epoch": 0.17204301075268819,
"grad_norm": 0.3237569332122803,
"learning_rate": 3.792040266839621e-05,
"loss": 0.018,
"step": 332
},
{
"epoch": 0.1725612125923047,
"grad_norm": 0.1244291141629219,
"learning_rate": 3.79054665708401e-05,
"loss": 0.0224,
"step": 333
},
{
"epoch": 0.17307941443192124,
"grad_norm": 0.12003582715988159,
"learning_rate": 3.7890479991149384e-05,
"loss": 0.0191,
"step": 334
},
{
"epoch": 0.17359761627153777,
"grad_norm": 0.1280771940946579,
"learning_rate": 3.787544297157678e-05,
"loss": 0.0161,
"step": 335
},
{
"epoch": 0.17411581811115429,
"grad_norm": 0.5037534236907959,
"learning_rate": 3.786035555451721e-05,
"loss": 0.0544,
"step": 336
},
{
"epoch": 0.17463401995077082,
"grad_norm": 0.13476350903511047,
"learning_rate": 3.7845217782507686e-05,
"loss": 0.0218,
"step": 337
},
{
"epoch": 0.17515222179038736,
"grad_norm": 0.11884070932865143,
"learning_rate": 3.783002969822718e-05,
"loss": 0.0228,
"step": 338
},
{
"epoch": 0.17567042363000387,
"grad_norm": 0.1743396818637848,
"learning_rate": 3.781479134449654e-05,
"loss": 0.0251,
"step": 339
},
{
"epoch": 0.1761886254696204,
"grad_norm": 0.26071441173553467,
"learning_rate": 3.779950276427829e-05,
"loss": 0.0266,
"step": 340
},
{
"epoch": 0.17670682730923695,
"grad_norm": 0.16687913239002228,
"learning_rate": 3.778416400067662e-05,
"loss": 0.0176,
"step": 341
},
{
"epoch": 0.1772250291488535,
"grad_norm": 0.11515403538942337,
"learning_rate": 3.7768775096937176e-05,
"loss": 0.0186,
"step": 342
},
{
"epoch": 0.17774323098847,
"grad_norm": 0.2612086534500122,
"learning_rate": 3.7753336096446954e-05,
"loss": 0.0262,
"step": 343
},
{
"epoch": 0.17826143282808654,
"grad_norm": 0.2220304310321808,
"learning_rate": 3.773784704273424e-05,
"loss": 0.0376,
"step": 344
},
{
"epoch": 0.17877963466770308,
"grad_norm": 0.2760023772716522,
"learning_rate": 3.7722307979468374e-05,
"loss": 0.0227,
"step": 345
},
{
"epoch": 0.1792978365073196,
"grad_norm": 0.1418105512857437,
"learning_rate": 3.770671895045974e-05,
"loss": 0.0256,
"step": 346
},
{
"epoch": 0.17981603834693613,
"grad_norm": 0.17588967084884644,
"learning_rate": 3.769107999965958e-05,
"loss": 0.0184,
"step": 347
},
{
"epoch": 0.18033424018655267,
"grad_norm": 0.17219921946525574,
"learning_rate": 3.767539117115988e-05,
"loss": 0.0195,
"step": 348
},
{
"epoch": 0.18085244202616918,
"grad_norm": 0.22888055443763733,
"learning_rate": 3.765965250919324e-05,
"loss": 0.0276,
"step": 349
},
{
"epoch": 0.18137064386578572,
"grad_norm": 0.15111926198005676,
"learning_rate": 3.764386405813278e-05,
"loss": 0.0222,
"step": 350
},
{
"epoch": 0.18188884570540226,
"grad_norm": 0.10396119952201843,
"learning_rate": 3.7628025862492e-05,
"loss": 0.0169,
"step": 351
},
{
"epoch": 0.1824070475450188,
"grad_norm": 0.1553802788257599,
"learning_rate": 3.7612137966924606e-05,
"loss": 0.0205,
"step": 352
},
{
"epoch": 0.1829252493846353,
"grad_norm": 0.21649691462516785,
"learning_rate": 3.759620041622447e-05,
"loss": 0.0311,
"step": 353
},
{
"epoch": 0.18344345122425185,
"grad_norm": 0.1523357480764389,
"learning_rate": 3.758021325532544e-05,
"loss": 0.0263,
"step": 354
},
{
"epoch": 0.18396165306386839,
"grad_norm": 0.20473290979862213,
"learning_rate": 3.756417652930124e-05,
"loss": 0.0274,
"step": 355
},
{
"epoch": 0.1844798549034849,
"grad_norm": 0.3731272220611572,
"learning_rate": 3.7548090283365325e-05,
"loss": 0.0481,
"step": 356
},
{
"epoch": 0.18499805674310144,
"grad_norm": 0.2392759472131729,
"learning_rate": 3.753195456287079e-05,
"loss": 0.0188,
"step": 357
},
{
"epoch": 0.18551625858271797,
"grad_norm": 0.1472381055355072,
"learning_rate": 3.751576941331017e-05,
"loss": 0.0166,
"step": 358
},
{
"epoch": 0.18603446042233449,
"grad_norm": 0.15394793450832367,
"learning_rate": 3.749953488031542e-05,
"loss": 0.0122,
"step": 359
},
{
"epoch": 0.18655266226195102,
"grad_norm": 0.13148321211338043,
"learning_rate": 3.748325100965769e-05,
"loss": 0.0237,
"step": 360
},
{
"epoch": 0.18707086410156756,
"grad_norm": 0.22537775337696075,
"learning_rate": 3.7466917847247225e-05,
"loss": 0.0192,
"step": 361
},
{
"epoch": 0.1875890659411841,
"grad_norm": 0.10962450504302979,
"learning_rate": 3.745053543913325e-05,
"loss": 0.0209,
"step": 362
},
{
"epoch": 0.1881072677808006,
"grad_norm": 0.19166776537895203,
"learning_rate": 3.7434103831503855e-05,
"loss": 0.0303,
"step": 363
},
{
"epoch": 0.18862546962041715,
"grad_norm": 0.20966096222400665,
"learning_rate": 3.7417623070685796e-05,
"loss": 0.0276,
"step": 364
},
{
"epoch": 0.1891436714600337,
"grad_norm": 0.38217735290527344,
"learning_rate": 3.740109320314445e-05,
"loss": 0.0349,
"step": 365
},
{
"epoch": 0.1896618732996502,
"grad_norm": 0.15686549246311188,
"learning_rate": 3.738451427548363e-05,
"loss": 0.0186,
"step": 366
},
{
"epoch": 0.19018007513926674,
"grad_norm": 0.34243685007095337,
"learning_rate": 3.736788633444546e-05,
"loss": 0.0335,
"step": 367
},
{
"epoch": 0.19069827697888328,
"grad_norm": 0.23040959239006042,
"learning_rate": 3.735120942691025e-05,
"loss": 0.0239,
"step": 368
},
{
"epoch": 0.19121647881849982,
"grad_norm": 0.1737140417098999,
"learning_rate": 3.7334483599896394e-05,
"loss": 0.0352,
"step": 369
},
{
"epoch": 0.19173468065811633,
"grad_norm": 0.09808428585529327,
"learning_rate": 3.731770890056017e-05,
"loss": 0.0134,
"step": 370
},
{
"epoch": 0.19225288249773287,
"grad_norm": 0.1002270057797432,
"learning_rate": 3.730088537619566e-05,
"loss": 0.0189,
"step": 371
},
{
"epoch": 0.1927710843373494,
"grad_norm": 0.1760372668504715,
"learning_rate": 3.728401307423462e-05,
"loss": 0.0191,
"step": 372
},
{
"epoch": 0.19328928617696592,
"grad_norm": 0.2071564793586731,
"learning_rate": 3.7267092042246296e-05,
"loss": 0.0352,
"step": 373
},
{
"epoch": 0.19380748801658246,
"grad_norm": 0.11756689846515656,
"learning_rate": 3.725012232793734e-05,
"loss": 0.0273,
"step": 374
},
{
"epoch": 0.194325689856199,
"grad_norm": 0.20388537645339966,
"learning_rate": 3.723310397915166e-05,
"loss": 0.0346,
"step": 375
},
{
"epoch": 0.1948438916958155,
"grad_norm": 0.17640188336372375,
"learning_rate": 3.721603704387026e-05,
"loss": 0.0243,
"step": 376
},
{
"epoch": 0.19536209353543205,
"grad_norm": 0.25531336665153503,
"learning_rate": 3.7198921570211174e-05,
"loss": 0.0293,
"step": 377
},
{
"epoch": 0.19588029537504859,
"grad_norm": 0.36133551597595215,
"learning_rate": 3.718175760642923e-05,
"loss": 0.0207,
"step": 378
},
{
"epoch": 0.19639849721466512,
"grad_norm": 0.1869734823703766,
"learning_rate": 3.716454520091601e-05,
"loss": 0.0257,
"step": 379
},
{
"epoch": 0.19691669905428164,
"grad_norm": 0.32143622636795044,
"learning_rate": 3.714728440219963e-05,
"loss": 0.0264,
"step": 380
},
{
"epoch": 0.19743490089389817,
"grad_norm": 0.24964651465415955,
"learning_rate": 3.712997525894469e-05,
"loss": 0.0188,
"step": 381
},
{
"epoch": 0.1979531027335147,
"grad_norm": 0.2554227113723755,
"learning_rate": 3.711261781995206e-05,
"loss": 0.0203,
"step": 382
},
{
"epoch": 0.19847130457313122,
"grad_norm": 0.09913673996925354,
"learning_rate": 3.709521213415878e-05,
"loss": 0.0148,
"step": 383
},
{
"epoch": 0.19898950641274776,
"grad_norm": 0.12040822952985764,
"learning_rate": 3.7077758250637914e-05,
"loss": 0.0155,
"step": 384
},
{
"epoch": 0.1995077082523643,
"grad_norm": 0.17180348932743073,
"learning_rate": 3.706025621859842e-05,
"loss": 0.0144,
"step": 385
},
{
"epoch": 0.2000259100919808,
"grad_norm": 0.11793264001607895,
"learning_rate": 3.704270608738502e-05,
"loss": 0.0184,
"step": 386
},
{
"epoch": 0.20054411193159735,
"grad_norm": 0.27998659014701843,
"learning_rate": 3.7025107906477995e-05,
"loss": 0.022,
"step": 387
},
{
"epoch": 0.2010623137712139,
"grad_norm": 0.16252531111240387,
"learning_rate": 3.700746172549315e-05,
"loss": 0.0151,
"step": 388
},
{
"epoch": 0.20158051561083043,
"grad_norm": 0.1505972445011139,
"learning_rate": 3.698976759418159e-05,
"loss": 0.0152,
"step": 389
},
{
"epoch": 0.20209871745044694,
"grad_norm": 0.34772515296936035,
"learning_rate": 3.697202556242961e-05,
"loss": 0.0403,
"step": 390
},
{
"epoch": 0.20261691929006348,
"grad_norm": 0.19800731539726257,
"learning_rate": 3.695423568025858e-05,
"loss": 0.027,
"step": 391
},
{
"epoch": 0.20313512112968002,
"grad_norm": 0.16485491394996643,
"learning_rate": 3.693639799782475e-05,
"loss": 0.0303,
"step": 392
},
{
"epoch": 0.20365332296929653,
"grad_norm": 0.13744878768920898,
"learning_rate": 3.691851256541913e-05,
"loss": 0.0125,
"step": 393
},
{
"epoch": 0.20417152480891307,
"grad_norm": 0.3705524802207947,
"learning_rate": 3.69005794334674e-05,
"loss": 0.03,
"step": 394
},
{
"epoch": 0.2046897266485296,
"grad_norm": 0.21377843618392944,
"learning_rate": 3.688259865252968e-05,
"loss": 0.0152,
"step": 395
},
{
"epoch": 0.20520792848814615,
"grad_norm": 0.20547166466712952,
"learning_rate": 3.6864570273300436e-05,
"loss": 0.0277,
"step": 396
},
{
"epoch": 0.20572613032776266,
"grad_norm": 0.4766762852668762,
"learning_rate": 3.6846494346608346e-05,
"loss": 0.0287,
"step": 397
},
{
"epoch": 0.2062443321673792,
"grad_norm": 0.3239550292491913,
"learning_rate": 3.6828370923416135e-05,
"loss": 0.0234,
"step": 398
},
{
"epoch": 0.20676253400699574,
"grad_norm": 0.13522012531757355,
"learning_rate": 3.6810200054820435e-05,
"loss": 0.0119,
"step": 399
},
{
"epoch": 0.20728073584661225,
"grad_norm": 0.366756409406662,
"learning_rate": 3.679198179205165e-05,
"loss": 0.0421,
"step": 400
},
{
"epoch": 0.20779893768622879,
"grad_norm": 0.44114238023757935,
"learning_rate": 3.677371618647381e-05,
"loss": 0.0191,
"step": 401
},
{
"epoch": 0.20831713952584532,
"grad_norm": 0.1942875236272812,
"learning_rate": 3.67554032895844e-05,
"loss": 0.0175,
"step": 402
},
{
"epoch": 0.20883534136546184,
"grad_norm": 0.4334729313850403,
"learning_rate": 3.673704315301426e-05,
"loss": 0.0341,
"step": 403
},
{
"epoch": 0.20935354320507837,
"grad_norm": 0.20343036949634552,
"learning_rate": 3.6718635828527405e-05,
"loss": 0.0169,
"step": 404
},
{
"epoch": 0.2098717450446949,
"grad_norm": 0.2867695391178131,
"learning_rate": 3.67001813680209e-05,
"loss": 0.0182,
"step": 405
},
{
"epoch": 0.21038994688431145,
"grad_norm": 0.375577449798584,
"learning_rate": 3.668167982352469e-05,
"loss": 0.0249,
"step": 406
},
{
"epoch": 0.21090814872392796,
"grad_norm": 0.33300960063934326,
"learning_rate": 3.666313124720146e-05,
"loss": 0.0283,
"step": 407
},
{
"epoch": 0.2114263505635445,
"grad_norm": 0.44241946935653687,
"learning_rate": 3.664453569134654e-05,
"loss": 0.0258,
"step": 408
},
{
"epoch": 0.21194455240316104,
"grad_norm": 0.21153447031974792,
"learning_rate": 3.6625893208387654e-05,
"loss": 0.0196,
"step": 409
},
{
"epoch": 0.21246275424277755,
"grad_norm": 0.2204204499721527,
"learning_rate": 3.660720385088487e-05,
"loss": 0.019,
"step": 410
},
{
"epoch": 0.2129809560823941,
"grad_norm": 0.329979807138443,
"learning_rate": 3.6588467671530404e-05,
"loss": 0.0255,
"step": 411
},
{
"epoch": 0.21349915792201063,
"grad_norm": 0.4161871373653412,
"learning_rate": 3.656968472314847e-05,
"loss": 0.013,
"step": 412
},
{
"epoch": 0.21401735976162714,
"grad_norm": 0.3123794496059418,
"learning_rate": 3.655085505869516e-05,
"loss": 0.0248,
"step": 413
},
{
"epoch": 0.21453556160124368,
"grad_norm": 0.13111092150211334,
"learning_rate": 3.653197873125825e-05,
"loss": 0.0072,
"step": 414
},
{
"epoch": 0.21505376344086022,
"grad_norm": 0.30517759919166565,
"learning_rate": 3.651305579405709e-05,
"loss": 0.0332,
"step": 415
},
{
"epoch": 0.21557196528047676,
"grad_norm": 0.2333979308605194,
"learning_rate": 3.649408630044246e-05,
"loss": 0.0117,
"step": 416
},
{
"epoch": 0.21609016712009327,
"grad_norm": 0.1910547912120819,
"learning_rate": 3.6475070303896364e-05,
"loss": 0.0235,
"step": 417
},
{
"epoch": 0.2166083689597098,
"grad_norm": 0.3474821448326111,
"learning_rate": 3.645600785803193e-05,
"loss": 0.0322,
"step": 418
},
{
"epoch": 0.21712657079932635,
"grad_norm": 0.33099234104156494,
"learning_rate": 3.643689901659326e-05,
"loss": 0.0222,
"step": 419
},
{
"epoch": 0.21764477263894286,
"grad_norm": 0.7437701225280762,
"learning_rate": 3.641774383345523e-05,
"loss": 0.0412,
"step": 420
},
{
"epoch": 0.2181629744785594,
"grad_norm": 0.21985647082328796,
"learning_rate": 3.6398542362623406e-05,
"loss": 0.0197,
"step": 421
},
{
"epoch": 0.21868117631817593,
"grad_norm": 0.3680473566055298,
"learning_rate": 3.637929465823382e-05,
"loss": 0.024,
"step": 422
},
{
"epoch": 0.21919937815779247,
"grad_norm": 0.20354367792606354,
"learning_rate": 3.6360000774552884e-05,
"loss": 0.0261,
"step": 423
},
{
"epoch": 0.21971757999740898,
"grad_norm": 0.3228227496147156,
"learning_rate": 3.6340660765977186e-05,
"loss": 0.0239,
"step": 424
},
{
"epoch": 0.22023578183702552,
"grad_norm": 0.17572440207004547,
"learning_rate": 3.632127468703337e-05,
"loss": 0.0125,
"step": 425
},
{
"epoch": 0.22075398367664206,
"grad_norm": 0.2067795842885971,
"learning_rate": 3.630184259237797e-05,
"loss": 0.0196,
"step": 426
},
{
"epoch": 0.22127218551625857,
"grad_norm": 0.17999672889709473,
"learning_rate": 3.628236453679724e-05,
"loss": 0.0222,
"step": 427
},
{
"epoch": 0.2217903873558751,
"grad_norm": 0.17415858805179596,
"learning_rate": 3.6262840575207034e-05,
"loss": 0.0218,
"step": 428
},
{
"epoch": 0.22230858919549165,
"grad_norm": 0.21333667635917664,
"learning_rate": 3.624327076265261e-05,
"loss": 0.0228,
"step": 429
},
{
"epoch": 0.22282679103510816,
"grad_norm": 0.29897600412368774,
"learning_rate": 3.622365515430851e-05,
"loss": 0.0298,
"step": 430
},
{
"epoch": 0.2233449928747247,
"grad_norm": 0.17251932621002197,
"learning_rate": 3.6203993805478414e-05,
"loss": 0.0211,
"step": 431
},
{
"epoch": 0.22386319471434124,
"grad_norm": 0.11624428629875183,
"learning_rate": 3.618428677159492e-05,
"loss": 0.0104,
"step": 432
},
{
"epoch": 0.22438139655395778,
"grad_norm": 0.20840483903884888,
"learning_rate": 3.6164534108219445e-05,
"loss": 0.028,
"step": 433
},
{
"epoch": 0.2248995983935743,
"grad_norm": 0.17955373227596283,
"learning_rate": 3.614473587104206e-05,
"loss": 0.0206,
"step": 434
},
{
"epoch": 0.22541780023319083,
"grad_norm": 0.32150277495384216,
"learning_rate": 3.6124892115881326e-05,
"loss": 0.0268,
"step": 435
},
{
"epoch": 0.22593600207280737,
"grad_norm": 0.38776031136512756,
"learning_rate": 3.6105002898684115e-05,
"loss": 0.0323,
"step": 436
},
{
"epoch": 0.22645420391242388,
"grad_norm": 0.28147077560424805,
"learning_rate": 3.60850682755255e-05,
"loss": 0.0329,
"step": 437
},
{
"epoch": 0.22697240575204042,
"grad_norm": 0.15004199743270874,
"learning_rate": 3.606508830260856e-05,
"loss": 0.0156,
"step": 438
},
{
"epoch": 0.22749060759165696,
"grad_norm": 0.13495272397994995,
"learning_rate": 3.604506303626423e-05,
"loss": 0.0142,
"step": 439
},
{
"epoch": 0.22800880943127347,
"grad_norm": 0.23692473769187927,
"learning_rate": 3.602499253295113e-05,
"loss": 0.0265,
"step": 440
},
{
"epoch": 0.22852701127089,
"grad_norm": 0.15593603253364563,
"learning_rate": 3.600487684925545e-05,
"loss": 0.017,
"step": 441
},
{
"epoch": 0.22904521311050655,
"grad_norm": 0.27002614736557007,
"learning_rate": 3.5984716041890745e-05,
"loss": 0.0345,
"step": 442
},
{
"epoch": 0.22956341495012308,
"grad_norm": 0.2901112139225006,
"learning_rate": 3.596451016769778e-05,
"loss": 0.0421,
"step": 443
},
{
"epoch": 0.2300816167897396,
"grad_norm": 0.1251426786184311,
"learning_rate": 3.5944259283644394e-05,
"loss": 0.0156,
"step": 444
},
{
"epoch": 0.23059981862935613,
"grad_norm": 0.3259553015232086,
"learning_rate": 3.5923963446825325e-05,
"loss": 0.0325,
"step": 445
},
{
"epoch": 0.23111802046897267,
"grad_norm": 0.1895984560251236,
"learning_rate": 3.5903622714462045e-05,
"loss": 0.0153,
"step": 446
},
{
"epoch": 0.23163622230858918,
"grad_norm": 0.4781612157821655,
"learning_rate": 3.5883237143902594e-05,
"loss": 0.0491,
"step": 447
},
{
"epoch": 0.23215442414820572,
"grad_norm": 0.3092973530292511,
"learning_rate": 3.586280679262144e-05,
"loss": 0.0321,
"step": 448
},
{
"epoch": 0.23267262598782226,
"grad_norm": 0.11183629184961319,
"learning_rate": 3.584233171821931e-05,
"loss": 0.0121,
"step": 449
},
{
"epoch": 0.23319082782743877,
"grad_norm": 0.26677051186561584,
"learning_rate": 3.582181197842301e-05,
"loss": 0.0261,
"step": 450
},
{
"epoch": 0.2337090296670553,
"grad_norm": 0.19713018834590912,
"learning_rate": 3.5801247631085286e-05,
"loss": 0.0273,
"step": 451
},
{
"epoch": 0.23422723150667185,
"grad_norm": 0.28130441904067993,
"learning_rate": 3.578063873418462e-05,
"loss": 0.0431,
"step": 452
},
{
"epoch": 0.2347454333462884,
"grad_norm": 0.24492007493972778,
"learning_rate": 3.5759985345825143e-05,
"loss": 0.023,
"step": 453
},
{
"epoch": 0.2352636351859049,
"grad_norm": 0.2576156258583069,
"learning_rate": 3.573928752423638e-05,
"loss": 0.0276,
"step": 454
},
{
"epoch": 0.23578183702552144,
"grad_norm": 0.2688496708869934,
"learning_rate": 3.5718545327773145e-05,
"loss": 0.0241,
"step": 455
},
{
"epoch": 0.23630003886513798,
"grad_norm": 0.2165881246328354,
"learning_rate": 3.5697758814915376e-05,
"loss": 0.0184,
"step": 456
},
{
"epoch": 0.2368182407047545,
"grad_norm": 0.10839606076478958,
"learning_rate": 3.5676928044267935e-05,
"loss": 0.0148,
"step": 457
},
{
"epoch": 0.23733644254437103,
"grad_norm": 0.13416044414043427,
"learning_rate": 3.565605307456047e-05,
"loss": 0.0174,
"step": 458
},
{
"epoch": 0.23785464438398757,
"grad_norm": 0.19393199682235718,
"learning_rate": 3.563513396464726e-05,
"loss": 0.0198,
"step": 459
},
{
"epoch": 0.2383728462236041,
"grad_norm": 0.1913292557001114,
"learning_rate": 3.561417077350699e-05,
"loss": 0.0247,
"step": 460
},
{
"epoch": 0.23889104806322062,
"grad_norm": 0.15598498284816742,
"learning_rate": 3.559316356024267e-05,
"loss": 0.0212,
"step": 461
},
{
"epoch": 0.23940924990283716,
"grad_norm": 0.2367229461669922,
"learning_rate": 3.5572112384081395e-05,
"loss": 0.03,
"step": 462
},
{
"epoch": 0.2399274517424537,
"grad_norm": 0.15201199054718018,
"learning_rate": 3.555101730437423e-05,
"loss": 0.0226,
"step": 463
},
{
"epoch": 0.2404456535820702,
"grad_norm": 0.20422053337097168,
"learning_rate": 3.5529878380596e-05,
"loss": 0.0195,
"step": 464
},
{
"epoch": 0.24096385542168675,
"grad_norm": 0.2784031629562378,
"learning_rate": 3.5508695672345156e-05,
"loss": 0.0218,
"step": 465
},
{
"epoch": 0.24148205726130328,
"grad_norm": 0.17994067072868347,
"learning_rate": 3.5487469239343576e-05,
"loss": 0.0207,
"step": 466
},
{
"epoch": 0.2420002591009198,
"grad_norm": 0.21507667005062103,
"learning_rate": 3.546619914143645e-05,
"loss": 0.0201,
"step": 467
},
{
"epoch": 0.24251846094053633,
"grad_norm": 0.10343655943870544,
"learning_rate": 3.5444885438592036e-05,
"loss": 0.0124,
"step": 468
},
{
"epoch": 0.24303666278015287,
"grad_norm": 0.14296704530715942,
"learning_rate": 3.542352819090156e-05,
"loss": 0.02,
"step": 469
},
{
"epoch": 0.2435548646197694,
"grad_norm": 0.26198479533195496,
"learning_rate": 3.5402127458578994e-05,
"loss": 0.0297,
"step": 470
},
{
"epoch": 0.24407306645938592,
"grad_norm": 0.21289674937725067,
"learning_rate": 3.538068330196093e-05,
"loss": 0.0275,
"step": 471
},
{
"epoch": 0.24459126829900246,
"grad_norm": 0.25298917293548584,
"learning_rate": 3.535919578150637e-05,
"loss": 0.0202,
"step": 472
},
{
"epoch": 0.245109470138619,
"grad_norm": 0.2316911369562149,
"learning_rate": 3.5337664957796587e-05,
"loss": 0.0323,
"step": 473
},
{
"epoch": 0.2456276719782355,
"grad_norm": 0.336458295583725,
"learning_rate": 3.531609089153493e-05,
"loss": 0.03,
"step": 474
},
{
"epoch": 0.24614587381785205,
"grad_norm": 0.21604609489440918,
"learning_rate": 3.52944736435467e-05,
"loss": 0.0146,
"step": 475
},
{
"epoch": 0.2466640756574686,
"grad_norm": 0.1912679821252823,
"learning_rate": 3.5272813274778885e-05,
"loss": 0.0217,
"step": 476
},
{
"epoch": 0.2471822774970851,
"grad_norm": 0.3697705864906311,
"learning_rate": 3.52511098463001e-05,
"loss": 0.0235,
"step": 477
},
{
"epoch": 0.24770047933670164,
"grad_norm": 0.21411247551441193,
"learning_rate": 3.522936341930033e-05,
"loss": 0.0203,
"step": 478
},
{
"epoch": 0.24821868117631818,
"grad_norm": 0.23912426829338074,
"learning_rate": 3.5207574055090786e-05,
"loss": 0.0196,
"step": 479
},
{
"epoch": 0.24873688301593472,
"grad_norm": 0.4297958016395569,
"learning_rate": 3.518574181510377e-05,
"loss": 0.0248,
"step": 480
},
{
"epoch": 0.24925508485555123,
"grad_norm": 0.34526073932647705,
"learning_rate": 3.516386676089243e-05,
"loss": 0.0293,
"step": 481
},
{
"epoch": 0.24977328669516777,
"grad_norm": 0.26696211099624634,
"learning_rate": 3.514194895413065e-05,
"loss": 0.0289,
"step": 482
},
{
"epoch": 0.2502914885347843,
"grad_norm": 0.22925925254821777,
"learning_rate": 3.511998845661282e-05,
"loss": 0.0169,
"step": 483
},
{
"epoch": 0.2508096903744008,
"grad_norm": 0.19786757230758667,
"learning_rate": 3.5097985330253715e-05,
"loss": 0.0167,
"step": 484
},
{
"epoch": 0.25132789221401736,
"grad_norm": 0.12485197931528091,
"learning_rate": 3.507593963708828e-05,
"loss": 0.0111,
"step": 485
},
{
"epoch": 0.2518460940536339,
"grad_norm": 0.21498580276966095,
"learning_rate": 3.50538514392715e-05,
"loss": 0.0303,
"step": 486
},
{
"epoch": 0.25236429589325043,
"grad_norm": 0.29731622338294983,
"learning_rate": 3.503172079907816e-05,
"loss": 0.0254,
"step": 487
},
{
"epoch": 0.252882497732867,
"grad_norm": 0.32734227180480957,
"learning_rate": 3.500954777890272e-05,
"loss": 0.0284,
"step": 488
},
{
"epoch": 0.25340069957248346,
"grad_norm": 0.18852336704730988,
"learning_rate": 3.4987332441259134e-05,
"loss": 0.0176,
"step": 489
},
{
"epoch": 0.2539189014121,
"grad_norm": 0.4650251269340515,
"learning_rate": 3.4965074848780665e-05,
"loss": 0.0395,
"step": 490
},
{
"epoch": 0.25443710325171653,
"grad_norm": 0.19879503548145294,
"learning_rate": 3.494277506421968e-05,
"loss": 0.0179,
"step": 491
},
{
"epoch": 0.2549553050913331,
"grad_norm": 0.23383589088916779,
"learning_rate": 3.492043315044754e-05,
"loss": 0.0175,
"step": 492
},
{
"epoch": 0.2554735069309496,
"grad_norm": 0.12329400330781937,
"learning_rate": 3.489804917045436e-05,
"loss": 0.0091,
"step": 493
},
{
"epoch": 0.25599170877056615,
"grad_norm": 0.10637083649635315,
"learning_rate": 3.487562318734885e-05,
"loss": 0.0096,
"step": 494
},
{
"epoch": 0.2565099106101827,
"grad_norm": 0.14903591573238373,
"learning_rate": 3.485315526435818e-05,
"loss": 0.0185,
"step": 495
},
{
"epoch": 0.2570281124497992,
"grad_norm": 0.20365048944950104,
"learning_rate": 3.483064546482771e-05,
"loss": 0.0177,
"step": 496
},
{
"epoch": 0.2575463142894157,
"grad_norm": 0.22113251686096191,
"learning_rate": 3.4808093852220906e-05,
"loss": 0.0259,
"step": 497
},
{
"epoch": 0.25806451612903225,
"grad_norm": 0.2057502418756485,
"learning_rate": 3.478550049011911e-05,
"loss": 0.015,
"step": 498
},
{
"epoch": 0.2585827179686488,
"grad_norm": 0.1919933706521988,
"learning_rate": 3.476286544222135e-05,
"loss": 0.0208,
"step": 499
},
{
"epoch": 0.25910091980826533,
"grad_norm": 0.19814780354499817,
"learning_rate": 3.474018877234422e-05,
"loss": 0.0165,
"step": 500
},
{
"epoch": 0.25961912164788187,
"grad_norm": 0.23219355940818787,
"learning_rate": 3.471747054442164e-05,
"loss": 0.0243,
"step": 501
},
{
"epoch": 0.2601373234874984,
"grad_norm": 0.2818119525909424,
"learning_rate": 3.469471082250468e-05,
"loss": 0.0418,
"step": 502
},
{
"epoch": 0.2606555253271149,
"grad_norm": 0.18665625154972076,
"learning_rate": 3.467190967076144e-05,
"loss": 0.0176,
"step": 503
},
{
"epoch": 0.26117372716673143,
"grad_norm": 0.1935970038175583,
"learning_rate": 3.464906715347679e-05,
"loss": 0.0188,
"step": 504
},
{
"epoch": 0.26169192900634797,
"grad_norm": 0.3840036392211914,
"learning_rate": 3.462618333505225e-05,
"loss": 0.0364,
"step": 505
},
{
"epoch": 0.2622101308459645,
"grad_norm": 0.12956850230693817,
"learning_rate": 3.4603258280005755e-05,
"loss": 0.0106,
"step": 506
},
{
"epoch": 0.26272833268558105,
"grad_norm": 0.2004137635231018,
"learning_rate": 3.458029205297151e-05,
"loss": 0.0218,
"step": 507
},
{
"epoch": 0.2632465345251976,
"grad_norm": 0.11050537973642349,
"learning_rate": 3.455728471869983e-05,
"loss": 0.0131,
"step": 508
},
{
"epoch": 0.26376473636481407,
"grad_norm": 0.23582395911216736,
"learning_rate": 3.4534236342056885e-05,
"loss": 0.0252,
"step": 509
},
{
"epoch": 0.2642829382044306,
"grad_norm": 0.19793295860290527,
"learning_rate": 3.4511146988024585e-05,
"loss": 0.0269,
"step": 510
},
{
"epoch": 0.26480114004404715,
"grad_norm": 0.18839754164218903,
"learning_rate": 3.448801672170035e-05,
"loss": 0.0193,
"step": 511
},
{
"epoch": 0.2653193418836637,
"grad_norm": 0.2332521229982376,
"learning_rate": 3.446484560829697e-05,
"loss": 0.0262,
"step": 512
},
{
"epoch": 0.2658375437232802,
"grad_norm": 0.17722775042057037,
"learning_rate": 3.444163371314238e-05,
"loss": 0.0143,
"step": 513
},
{
"epoch": 0.26635574556289676,
"grad_norm": 0.35751911997795105,
"learning_rate": 3.441838110167949e-05,
"loss": 0.0243,
"step": 514
},
{
"epoch": 0.2668739474025133,
"grad_norm": 0.19578954577445984,
"learning_rate": 3.439508783946604e-05,
"loss": 0.0264,
"step": 515
},
{
"epoch": 0.2673921492421298,
"grad_norm": 0.22078539431095123,
"learning_rate": 3.437175399217433e-05,
"loss": 0.0137,
"step": 516
},
{
"epoch": 0.2679103510817463,
"grad_norm": 0.12540170550346375,
"learning_rate": 3.434837962559112e-05,
"loss": 0.0158,
"step": 517
},
{
"epoch": 0.26842855292136286,
"grad_norm": 0.3015592694282532,
"learning_rate": 3.432496480561739e-05,
"loss": 0.03,
"step": 518
},
{
"epoch": 0.2689467547609794,
"grad_norm": 0.2720678150653839,
"learning_rate": 3.430150959826819e-05,
"loss": 0.0315,
"step": 519
},
{
"epoch": 0.26946495660059594,
"grad_norm": 0.14795976877212524,
"learning_rate": 3.427801406967242e-05,
"loss": 0.0212,
"step": 520
},
{
"epoch": 0.2699831584402125,
"grad_norm": 0.12961719930171967,
"learning_rate": 3.425447828607266e-05,
"loss": 0.0101,
"step": 521
},
{
"epoch": 0.270501360279829,
"grad_norm": 0.39186790585517883,
"learning_rate": 3.423090231382501e-05,
"loss": 0.0316,
"step": 522
},
{
"epoch": 0.2710195621194455,
"grad_norm": 0.1863209307193756,
"learning_rate": 3.420728621939884e-05,
"loss": 0.0182,
"step": 523
},
{
"epoch": 0.27153776395906204,
"grad_norm": 0.171253502368927,
"learning_rate": 3.418363006937668e-05,
"loss": 0.0165,
"step": 524
},
{
"epoch": 0.2720559657986786,
"grad_norm": 0.47378459572792053,
"learning_rate": 3.4159933930453944e-05,
"loss": 0.0442,
"step": 525
},
{
"epoch": 0.2725741676382951,
"grad_norm": 0.21674169600009918,
"learning_rate": 3.4136197869438824e-05,
"loss": 0.0237,
"step": 526
},
{
"epoch": 0.27309236947791166,
"grad_norm": 0.5017228126525879,
"learning_rate": 3.411242195325205e-05,
"loss": 0.0245,
"step": 527
},
{
"epoch": 0.2736105713175282,
"grad_norm": 0.21757693588733673,
"learning_rate": 3.408860624892675e-05,
"loss": 0.0177,
"step": 528
},
{
"epoch": 0.27412877315714473,
"grad_norm": 0.19579505920410156,
"learning_rate": 3.406475082360817e-05,
"loss": 0.0226,
"step": 529
},
{
"epoch": 0.2746469749967612,
"grad_norm": 0.2168528437614441,
"learning_rate": 3.4040855744553614e-05,
"loss": 0.0226,
"step": 530
},
{
"epoch": 0.27516517683637776,
"grad_norm": 0.14640694856643677,
"learning_rate": 3.401692107913213e-05,
"loss": 0.0155,
"step": 531
},
{
"epoch": 0.2756833786759943,
"grad_norm": 0.24729685485363007,
"learning_rate": 3.39929468948244e-05,
"loss": 0.0183,
"step": 532
},
{
"epoch": 0.27620158051561083,
"grad_norm": 0.16490469872951508,
"learning_rate": 3.396893325922251e-05,
"loss": 0.0149,
"step": 533
},
{
"epoch": 0.2767197823552274,
"grad_norm": 0.1809859275817871,
"learning_rate": 3.3944880240029804e-05,
"loss": 0.0151,
"step": 534
},
{
"epoch": 0.2772379841948439,
"grad_norm": 0.24216516315937042,
"learning_rate": 3.392078790506062e-05,
"loss": 0.0199,
"step": 535
},
{
"epoch": 0.2777561860344604,
"grad_norm": 0.12011023610830307,
"learning_rate": 3.3896656322240176e-05,
"loss": 0.0127,
"step": 536
},
{
"epoch": 0.27827438787407693,
"grad_norm": 0.1671677827835083,
"learning_rate": 3.387248555960433e-05,
"loss": 0.0276,
"step": 537
},
{
"epoch": 0.2787925897136935,
"grad_norm": 0.33307945728302,
"learning_rate": 3.38482756852994e-05,
"loss": 0.0236,
"step": 538
},
{
"epoch": 0.27931079155331,
"grad_norm": 0.24255521595478058,
"learning_rate": 3.382402676758198e-05,
"loss": 0.011,
"step": 539
},
{
"epoch": 0.27982899339292655,
"grad_norm": 0.3321969211101532,
"learning_rate": 3.379973887481874e-05,
"loss": 0.0303,
"step": 540
},
{
"epoch": 0.2803471952325431,
"grad_norm": 0.21098312735557556,
"learning_rate": 3.377541207548624e-05,
"loss": 0.0206,
"step": 541
},
{
"epoch": 0.28086539707215963,
"grad_norm": 0.7190823554992676,
"learning_rate": 3.375104643817073e-05,
"loss": 0.0247,
"step": 542
},
{
"epoch": 0.2813835989117761,
"grad_norm": 0.2180882841348648,
"learning_rate": 3.372664203156794e-05,
"loss": 0.0266,
"step": 543
},
{
"epoch": 0.28190180075139265,
"grad_norm": 0.20533742010593414,
"learning_rate": 3.3702198924482935e-05,
"loss": 0.016,
"step": 544
},
{
"epoch": 0.2824200025910092,
"grad_norm": 0.14692696928977966,
"learning_rate": 3.3677717185829865e-05,
"loss": 0.0237,
"step": 545
},
{
"epoch": 0.28293820443062573,
"grad_norm": 0.25798094272613525,
"learning_rate": 3.3653196884631826e-05,
"loss": 0.0374,
"step": 546
},
{
"epoch": 0.28345640627024227,
"grad_norm": 0.22581781446933746,
"learning_rate": 3.3628638090020604e-05,
"loss": 0.0248,
"step": 547
},
{
"epoch": 0.2839746081098588,
"grad_norm": 0.18555013835430145,
"learning_rate": 3.3604040871236536e-05,
"loss": 0.0172,
"step": 548
},
{
"epoch": 0.28449280994947534,
"grad_norm": 0.18623068928718567,
"learning_rate": 3.357940529762827e-05,
"loss": 0.0271,
"step": 549
},
{
"epoch": 0.28501101178909183,
"grad_norm": 0.38695311546325684,
"learning_rate": 3.355473143865261e-05,
"loss": 0.0254,
"step": 550
},
{
"epoch": 0.28552921362870837,
"grad_norm": 0.21091844141483307,
"learning_rate": 3.35300193638743e-05,
"loss": 0.0145,
"step": 551
},
{
"epoch": 0.2860474154683249,
"grad_norm": 0.284892737865448,
"learning_rate": 3.3505269142965807e-05,
"loss": 0.017,
"step": 552
},
{
"epoch": 0.28656561730794144,
"grad_norm": 0.1800401210784912,
"learning_rate": 3.348048084570717e-05,
"loss": 0.012,
"step": 553
},
{
"epoch": 0.287083819147558,
"grad_norm": 0.20696207880973816,
"learning_rate": 3.345565454198577e-05,
"loss": 0.0183,
"step": 554
},
{
"epoch": 0.2876020209871745,
"grad_norm": 0.21303308010101318,
"learning_rate": 3.343079030179613e-05,
"loss": 0.0216,
"step": 555
},
{
"epoch": 0.28812022282679106,
"grad_norm": 0.16014441847801208,
"learning_rate": 3.340588819523976e-05,
"loss": 0.0205,
"step": 556
},
{
"epoch": 0.28863842466640754,
"grad_norm": 0.3861747682094574,
"learning_rate": 3.338094829252489e-05,
"loss": 0.0283,
"step": 557
},
{
"epoch": 0.2891566265060241,
"grad_norm": 0.31511127948760986,
"learning_rate": 3.335597066396635e-05,
"loss": 0.0188,
"step": 558
},
{
"epoch": 0.2896748283456406,
"grad_norm": 0.11825576424598694,
"learning_rate": 3.3330955379985316e-05,
"loss": 0.0176,
"step": 559
},
{
"epoch": 0.29019303018525716,
"grad_norm": 0.2394377589225769,
"learning_rate": 3.3305902511109114e-05,
"loss": 0.0318,
"step": 560
},
{
"epoch": 0.2907112320248737,
"grad_norm": 0.22725820541381836,
"learning_rate": 3.328081212797107e-05,
"loss": 0.0228,
"step": 561
},
{
"epoch": 0.29122943386449024,
"grad_norm": 0.3189086616039276,
"learning_rate": 3.325568430131023e-05,
"loss": 0.0374,
"step": 562
},
{
"epoch": 0.2917476357041067,
"grad_norm": 0.18351779878139496,
"learning_rate": 3.323051910197125e-05,
"loss": 0.0132,
"step": 563
},
{
"epoch": 0.29226583754372326,
"grad_norm": 0.2662956118583679,
"learning_rate": 3.320531660090414e-05,
"loss": 0.0254,
"step": 564
},
{
"epoch": 0.2927840393833398,
"grad_norm": 0.23380038142204285,
"learning_rate": 3.318007686916408e-05,
"loss": 0.0204,
"step": 565
},
{
"epoch": 0.29330224122295634,
"grad_norm": 0.21842540800571442,
"learning_rate": 3.315479997791119e-05,
"loss": 0.0168,
"step": 566
},
{
"epoch": 0.2938204430625729,
"grad_norm": 0.2316482663154602,
"learning_rate": 3.312948599841039e-05,
"loss": 0.0345,
"step": 567
},
{
"epoch": 0.2943386449021894,
"grad_norm": 0.10342805832624435,
"learning_rate": 3.3104135002031154e-05,
"loss": 0.0098,
"step": 568
},
{
"epoch": 0.29485684674180596,
"grad_norm": 0.6484712958335876,
"learning_rate": 3.307874706024732e-05,
"loss": 0.0248,
"step": 569
},
{
"epoch": 0.29537504858142244,
"grad_norm": 0.24838627874851227,
"learning_rate": 3.305332224463689e-05,
"loss": 0.0207,
"step": 570
},
{
"epoch": 0.295893250421039,
"grad_norm": 0.27338987588882446,
"learning_rate": 3.302786062688182e-05,
"loss": 0.0187,
"step": 571
},
{
"epoch": 0.2964114522606555,
"grad_norm": 0.17893199622631073,
"learning_rate": 3.300236227876784e-05,
"loss": 0.0141,
"step": 572
},
{
"epoch": 0.29692965410027206,
"grad_norm": 0.11885008960962296,
"learning_rate": 3.29768272721842e-05,
"loss": 0.0151,
"step": 573
},
{
"epoch": 0.2974478559398886,
"grad_norm": 0.12600629031658173,
"learning_rate": 3.295125567912356e-05,
"loss": 0.0152,
"step": 574
},
{
"epoch": 0.29796605777950513,
"grad_norm": 0.25072181224823,
"learning_rate": 3.2925647571681697e-05,
"loss": 0.0301,
"step": 575
},
{
"epoch": 0.2984842596191217,
"grad_norm": 0.14617246389389038,
"learning_rate": 3.290000302205733e-05,
"loss": 0.0148,
"step": 576
},
{
"epoch": 0.29900246145873816,
"grad_norm": 0.11175688356161118,
"learning_rate": 3.287432210255194e-05,
"loss": 0.0131,
"step": 577
},
{
"epoch": 0.2995206632983547,
"grad_norm": 0.16760458052158356,
"learning_rate": 3.284860488556953e-05,
"loss": 0.0225,
"step": 578
},
{
"epoch": 0.30003886513797123,
"grad_norm": 0.21576490998268127,
"learning_rate": 3.282285144361646e-05,
"loss": 0.0189,
"step": 579
},
{
"epoch": 0.3005570669775878,
"grad_norm": 0.29469025135040283,
"learning_rate": 3.2797061849301205e-05,
"loss": 0.03,
"step": 580
},
{
"epoch": 0.3010752688172043,
"grad_norm": 0.16392472386360168,
"learning_rate": 3.2771236175334174e-05,
"loss": 0.0122,
"step": 581
},
{
"epoch": 0.30159347065682085,
"grad_norm": 0.19024868309497833,
"learning_rate": 3.274537449452749e-05,
"loss": 0.0191,
"step": 582
},
{
"epoch": 0.3021116724964374,
"grad_norm": 0.1978103071451187,
"learning_rate": 3.2719476879794804e-05,
"loss": 0.0147,
"step": 583
},
{
"epoch": 0.3026298743360539,
"grad_norm": 0.17215609550476074,
"learning_rate": 3.2693543404151084e-05,
"loss": 0.0304,
"step": 584
},
{
"epoch": 0.3031480761756704,
"grad_norm": 0.3572210967540741,
"learning_rate": 3.266757414071238e-05,
"loss": 0.0287,
"step": 585
},
{
"epoch": 0.30366627801528695,
"grad_norm": 0.24866849184036255,
"learning_rate": 3.2641569162695654e-05,
"loss": 0.0189,
"step": 586
},
{
"epoch": 0.3041844798549035,
"grad_norm": 0.24234169721603394,
"learning_rate": 3.2615528543418564e-05,
"loss": 0.0272,
"step": 587
},
{
"epoch": 0.30470268169452003,
"grad_norm": 0.15545207262039185,
"learning_rate": 3.258945235629924e-05,
"loss": 0.0154,
"step": 588
},
{
"epoch": 0.30522088353413657,
"grad_norm": 0.1727120280265808,
"learning_rate": 3.256334067485611e-05,
"loss": 0.019,
"step": 589
},
{
"epoch": 0.30573908537375305,
"grad_norm": 0.33689263463020325,
"learning_rate": 3.2537193572707665e-05,
"loss": 0.0228,
"step": 590
},
{
"epoch": 0.3062572872133696,
"grad_norm": 0.16162997484207153,
"learning_rate": 3.251101112357225e-05,
"loss": 0.0091,
"step": 591
},
{
"epoch": 0.30677548905298613,
"grad_norm": 0.1706080287694931,
"learning_rate": 3.248479340126789e-05,
"loss": 0.0138,
"step": 592
},
{
"epoch": 0.30729369089260267,
"grad_norm": 0.20587295293807983,
"learning_rate": 3.245854047971204e-05,
"loss": 0.0232,
"step": 593
},
{
"epoch": 0.3078118927322192,
"grad_norm": 0.22837670147418976,
"learning_rate": 3.2432252432921385e-05,
"loss": 0.0221,
"step": 594
},
{
"epoch": 0.30833009457183574,
"grad_norm": 0.20966672897338867,
"learning_rate": 3.240592933501166e-05,
"loss": 0.0133,
"step": 595
},
{
"epoch": 0.3088482964114523,
"grad_norm": 0.2917436361312866,
"learning_rate": 3.2379571260197434e-05,
"loss": 0.0237,
"step": 596
},
{
"epoch": 0.30936649825106877,
"grad_norm": 0.27110305428504944,
"learning_rate": 3.235317828279185e-05,
"loss": 0.025,
"step": 597
},
{
"epoch": 0.3098847000906853,
"grad_norm": 0.31341224908828735,
"learning_rate": 3.2326750477206484e-05,
"loss": 0.0301,
"step": 598
},
{
"epoch": 0.31040290193030184,
"grad_norm": 0.2728060185909271,
"learning_rate": 3.23002879179511e-05,
"loss": 0.0326,
"step": 599
},
{
"epoch": 0.3109211037699184,
"grad_norm": 0.20057561993598938,
"learning_rate": 3.227379067963343e-05,
"loss": 0.0232,
"step": 600
},
{
"epoch": 0.3114393056095349,
"grad_norm": 0.21433892846107483,
"learning_rate": 3.224725883695901e-05,
"loss": 0.0249,
"step": 601
},
{
"epoch": 0.31195750744915146,
"grad_norm": 0.20961421728134155,
"learning_rate": 3.222069246473091e-05,
"loss": 0.0198,
"step": 602
},
{
"epoch": 0.312475709288768,
"grad_norm": 0.14495067298412323,
"learning_rate": 3.2194091637849554e-05,
"loss": 0.0125,
"step": 603
},
{
"epoch": 0.3129939111283845,
"grad_norm": 0.26637324690818787,
"learning_rate": 3.216745643131254e-05,
"loss": 0.0204,
"step": 604
},
{
"epoch": 0.313512112968001,
"grad_norm": 0.2717486023902893,
"learning_rate": 3.214078692021434e-05,
"loss": 0.0188,
"step": 605
},
{
"epoch": 0.31403031480761756,
"grad_norm": 0.21144232153892517,
"learning_rate": 3.211408317974619e-05,
"loss": 0.0202,
"step": 606
},
{
"epoch": 0.3145485166472341,
"grad_norm": 0.1410185843706131,
"learning_rate": 3.208734528519581e-05,
"loss": 0.0149,
"step": 607
},
{
"epoch": 0.31506671848685064,
"grad_norm": 0.2241893708705902,
"learning_rate": 3.2060573311947214e-05,
"loss": 0.0169,
"step": 608
},
{
"epoch": 0.3155849203264672,
"grad_norm": 0.18555960059165955,
"learning_rate": 3.20337673354805e-05,
"loss": 0.016,
"step": 609
},
{
"epoch": 0.3161031221660837,
"grad_norm": 0.09408245235681534,
"learning_rate": 3.200692743137163e-05,
"loss": 0.0087,
"step": 610
},
{
"epoch": 0.3166213240057002,
"grad_norm": 0.24068722128868103,
"learning_rate": 3.1980053675292234e-05,
"loss": 0.0179,
"step": 611
},
{
"epoch": 0.31713952584531674,
"grad_norm": 0.20078520476818085,
"learning_rate": 3.1953146143009367e-05,
"loss": 0.0127,
"step": 612
},
{
"epoch": 0.3176577276849333,
"grad_norm": 0.478952556848526,
"learning_rate": 3.192620491038531e-05,
"loss": 0.0241,
"step": 613
},
{
"epoch": 0.3181759295245498,
"grad_norm": 0.18030868470668793,
"learning_rate": 3.189923005337737e-05,
"loss": 0.0131,
"step": 614
},
{
"epoch": 0.31869413136416636,
"grad_norm": 0.2694755494594574,
"learning_rate": 3.1872221648037654e-05,
"loss": 0.0179,
"step": 615
},
{
"epoch": 0.3192123332037829,
"grad_norm": 0.18882977962493896,
"learning_rate": 3.1845179770512845e-05,
"loss": 0.026,
"step": 616
},
{
"epoch": 0.3197305350433994,
"grad_norm": 0.29215094447135925,
"learning_rate": 3.181810449704399e-05,
"loss": 0.0164,
"step": 617
},
{
"epoch": 0.3202487368830159,
"grad_norm": 0.2687048017978668,
"learning_rate": 3.179099590396632e-05,
"loss": 0.0223,
"step": 618
},
{
"epoch": 0.32076693872263246,
"grad_norm": 0.15531520545482635,
"learning_rate": 3.1763854067708985e-05,
"loss": 0.0072,
"step": 619
},
{
"epoch": 0.321285140562249,
"grad_norm": 0.2893719971179962,
"learning_rate": 3.173667906479485e-05,
"loss": 0.0174,
"step": 620
},
{
"epoch": 0.32180334240186553,
"grad_norm": 0.23916292190551758,
"learning_rate": 3.170947097184031e-05,
"loss": 0.0231,
"step": 621
},
{
"epoch": 0.32232154424148207,
"grad_norm": 0.17515330016613007,
"learning_rate": 3.168222986555504e-05,
"loss": 0.0141,
"step": 622
},
{
"epoch": 0.3228397460810986,
"grad_norm": 0.2012537568807602,
"learning_rate": 3.16549558227418e-05,
"loss": 0.0181,
"step": 623
},
{
"epoch": 0.3233579479207151,
"grad_norm": 0.13551956415176392,
"learning_rate": 3.162764892029622e-05,
"loss": 0.0127,
"step": 624
},
{
"epoch": 0.32387614976033163,
"grad_norm": 0.10975822061300278,
"learning_rate": 3.1600309235206536e-05,
"loss": 0.0098,
"step": 625
},
{
"epoch": 0.32439435159994817,
"grad_norm": 0.3241739273071289,
"learning_rate": 3.1572936844553454e-05,
"loss": 0.0321,
"step": 626
},
{
"epoch": 0.3249125534395647,
"grad_norm": 0.31218957901000977,
"learning_rate": 3.154553182550986e-05,
"loss": 0.0239,
"step": 627
},
{
"epoch": 0.32543075527918125,
"grad_norm": 0.2881108522415161,
"learning_rate": 3.151809425534065e-05,
"loss": 0.0287,
"step": 628
},
{
"epoch": 0.3259489571187978,
"grad_norm": 0.20379316806793213,
"learning_rate": 3.1490624211402475e-05,
"loss": 0.0171,
"step": 629
},
{
"epoch": 0.3264671589584143,
"grad_norm": 0.21332648396492004,
"learning_rate": 3.1463121771143563e-05,
"loss": 0.0266,
"step": 630
},
{
"epoch": 0.3269853607980308,
"grad_norm": 0.2312246561050415,
"learning_rate": 3.143558701210345e-05,
"loss": 0.0268,
"step": 631
},
{
"epoch": 0.32750356263764735,
"grad_norm": 0.2009822130203247,
"learning_rate": 3.140802001191283e-05,
"loss": 0.027,
"step": 632
},
{
"epoch": 0.3280217644772639,
"grad_norm": 0.2293207049369812,
"learning_rate": 3.138042084829327e-05,
"loss": 0.0331,
"step": 633
},
{
"epoch": 0.3285399663168804,
"grad_norm": 0.14847518503665924,
"learning_rate": 3.1352789599057004e-05,
"loss": 0.0077,
"step": 634
},
{
"epoch": 0.32905816815649697,
"grad_norm": 0.26872965693473816,
"learning_rate": 3.132512634210676e-05,
"loss": 0.0272,
"step": 635
},
{
"epoch": 0.3295763699961135,
"grad_norm": 0.08970358967781067,
"learning_rate": 3.129743115543551e-05,
"loss": 0.0135,
"step": 636
},
{
"epoch": 0.33009457183573004,
"grad_norm": 0.15737628936767578,
"learning_rate": 3.1269704117126206e-05,
"loss": 0.0174,
"step": 637
},
{
"epoch": 0.3306127736753465,
"grad_norm": 0.24024097621440887,
"learning_rate": 3.1241945305351637e-05,
"loss": 0.0171,
"step": 638
},
{
"epoch": 0.33113097551496307,
"grad_norm": 0.17292901873588562,
"learning_rate": 3.121415479837416e-05,
"loss": 0.0252,
"step": 639
},
{
"epoch": 0.3316491773545796,
"grad_norm": 0.16394537687301636,
"learning_rate": 3.11863326745455e-05,
"loss": 0.0171,
"step": 640
},
{
"epoch": 0.33216737919419614,
"grad_norm": 0.17823824286460876,
"learning_rate": 3.115847901230652e-05,
"loss": 0.0201,
"step": 641
},
{
"epoch": 0.3326855810338127,
"grad_norm": 0.24929647147655487,
"learning_rate": 3.113059389018699e-05,
"loss": 0.0287,
"step": 642
},
{
"epoch": 0.3332037828734292,
"grad_norm": 0.1503756195306778,
"learning_rate": 3.110267738680539e-05,
"loss": 0.0141,
"step": 643
},
{
"epoch": 0.3337219847130457,
"grad_norm": 0.13402046263217926,
"learning_rate": 3.107472958086867e-05,
"loss": 0.0145,
"step": 644
},
{
"epoch": 0.33424018655266224,
"grad_norm": 0.2699407935142517,
"learning_rate": 3.104675055117203e-05,
"loss": 0.0183,
"step": 645
},
{
"epoch": 0.3347583883922788,
"grad_norm": 0.24766863882541656,
"learning_rate": 3.101874037659872e-05,
"loss": 0.0308,
"step": 646
},
{
"epoch": 0.3352765902318953,
"grad_norm": 0.10173456370830536,
"learning_rate": 3.099069913611977e-05,
"loss": 0.014,
"step": 647
},
{
"epoch": 0.33579479207151186,
"grad_norm": 0.13719849288463593,
"learning_rate": 3.0962626908793824e-05,
"loss": 0.014,
"step": 648
},
{
"epoch": 0.3363129939111284,
"grad_norm": 0.30098456144332886,
"learning_rate": 3.0934523773766864e-05,
"loss": 0.0194,
"step": 649
},
{
"epoch": 0.33683119575074494,
"grad_norm": 0.25823748111724854,
"learning_rate": 3.090638981027203e-05,
"loss": 0.0368,
"step": 650
},
{
"epoch": 0.3373493975903614,
"grad_norm": 0.13479436933994293,
"learning_rate": 3.087822509762938e-05,
"loss": 0.0114,
"step": 651
},
{
"epoch": 0.33786759942997796,
"grad_norm": 0.23411042988300323,
"learning_rate": 3.085002971524564e-05,
"loss": 0.0219,
"step": 652
},
{
"epoch": 0.3383858012695945,
"grad_norm": 0.361751914024353,
"learning_rate": 3.0821803742614044e-05,
"loss": 0.0487,
"step": 653
},
{
"epoch": 0.33890400310921104,
"grad_norm": 0.2769923508167267,
"learning_rate": 3.079354725931405e-05,
"loss": 0.0244,
"step": 654
},
{
"epoch": 0.3394222049488276,
"grad_norm": 0.28610527515411377,
"learning_rate": 3.076526034501113e-05,
"loss": 0.0416,
"step": 655
},
{
"epoch": 0.3399404067884441,
"grad_norm": 0.12687546014785767,
"learning_rate": 3.0736943079456554e-05,
"loss": 0.0164,
"step": 656
},
{
"epoch": 0.34045860862806065,
"grad_norm": 0.10315964370965958,
"learning_rate": 3.070859554248719e-05,
"loss": 0.0155,
"step": 657
},
{
"epoch": 0.34097681046767714,
"grad_norm": 0.1738579124212265,
"learning_rate": 3.0680217814025216e-05,
"loss": 0.0211,
"step": 658
},
{
"epoch": 0.3414950123072937,
"grad_norm": 0.13814416527748108,
"learning_rate": 3.065180997407795e-05,
"loss": 0.0136,
"step": 659
},
{
"epoch": 0.3420132141469102,
"grad_norm": 0.20698578655719757,
"learning_rate": 3.062337210273761e-05,
"loss": 0.0288,
"step": 660
},
{
"epoch": 0.34253141598652675,
"grad_norm": 0.14585328102111816,
"learning_rate": 3.059490428018106e-05,
"loss": 0.0189,
"step": 661
},
{
"epoch": 0.3430496178261433,
"grad_norm": 0.15746049582958221,
"learning_rate": 3.056640658666965e-05,
"loss": 0.0222,
"step": 662
},
{
"epoch": 0.34356781966575983,
"grad_norm": 0.14348691701889038,
"learning_rate": 3.0537879102548904e-05,
"loss": 0.0173,
"step": 663
},
{
"epoch": 0.34408602150537637,
"grad_norm": 0.14485906064510345,
"learning_rate": 3.0509321908248347e-05,
"loss": 0.0167,
"step": 664
},
{
"epoch": 0.34460422334499285,
"grad_norm": 0.21492420136928558,
"learning_rate": 3.0480735084281286e-05,
"loss": 0.0217,
"step": 665
},
{
"epoch": 0.3451224251846094,
"grad_norm": 0.1390220671892166,
"learning_rate": 3.0452118711244555e-05,
"loss": 0.0155,
"step": 666
},
{
"epoch": 0.34564062702422593,
"grad_norm": 0.38242873549461365,
"learning_rate": 3.042347286981829e-05,
"loss": 0.0359,
"step": 667
},
{
"epoch": 0.34615882886384247,
"grad_norm": 0.1517501026391983,
"learning_rate": 3.039479764076572e-05,
"loss": 0.0181,
"step": 668
},
{
"epoch": 0.346677030703459,
"grad_norm": 0.30889183282852173,
"learning_rate": 3.036609310493292e-05,
"loss": 0.036,
"step": 669
},
{
"epoch": 0.34719523254307555,
"grad_norm": 0.24580207467079163,
"learning_rate": 3.0337359343248606e-05,
"loss": 0.034,
"step": 670
},
{
"epoch": 0.34771343438269203,
"grad_norm": 0.24991121888160706,
"learning_rate": 3.030859643672387e-05,
"loss": 0.0227,
"step": 671
},
{
"epoch": 0.34823163622230857,
"grad_norm": 0.13242225348949432,
"learning_rate": 3.0279804466451996e-05,
"loss": 0.0125,
"step": 672
},
{
"epoch": 0.3487498380619251,
"grad_norm": 0.12930886447429657,
"learning_rate": 3.0250983513608198e-05,
"loss": 0.0211,
"step": 673
},
{
"epoch": 0.34926803990154165,
"grad_norm": 0.14146579802036285,
"learning_rate": 3.0222133659449404e-05,
"loss": 0.0117,
"step": 674
},
{
"epoch": 0.3497862417411582,
"grad_norm": 0.1282532513141632,
"learning_rate": 3.0193254985314015e-05,
"loss": 0.0169,
"step": 675
},
{
"epoch": 0.3503044435807747,
"grad_norm": 0.17600007355213165,
"learning_rate": 3.0164347572621717e-05,
"loss": 0.0196,
"step": 676
},
{
"epoch": 0.35082264542039127,
"grad_norm": 0.13536961376667023,
"learning_rate": 3.0135411502873188e-05,
"loss": 0.0131,
"step": 677
},
{
"epoch": 0.35134084726000775,
"grad_norm": 0.33690956234931946,
"learning_rate": 3.0106446857649922e-05,
"loss": 0.0243,
"step": 678
},
{
"epoch": 0.3518590490996243,
"grad_norm": 0.3075578808784485,
"learning_rate": 3.007745371861396e-05,
"loss": 0.0241,
"step": 679
},
{
"epoch": 0.3523772509392408,
"grad_norm": 0.1268237829208374,
"learning_rate": 3.00484321675077e-05,
"loss": 0.0104,
"step": 680
},
{
"epoch": 0.35289545277885737,
"grad_norm": 0.15431272983551025,
"learning_rate": 3.0019382286153625e-05,
"loss": 0.0139,
"step": 681
},
{
"epoch": 0.3534136546184739,
"grad_norm": 0.1659412533044815,
"learning_rate": 2.9990304156454088e-05,
"loss": 0.0148,
"step": 682
},
{
"epoch": 0.35393185645809044,
"grad_norm": 0.258486270904541,
"learning_rate": 2.996119786039111e-05,
"loss": 0.0226,
"step": 683
},
{
"epoch": 0.354450058297707,
"grad_norm": 0.12462588399648666,
"learning_rate": 2.99320634800261e-05,
"loss": 0.0106,
"step": 684
},
{
"epoch": 0.35496826013732347,
"grad_norm": 0.09379307180643082,
"learning_rate": 2.9902901097499656e-05,
"loss": 0.0057,
"step": 685
},
{
"epoch": 0.35548646197694,
"grad_norm": 0.25576797127723694,
"learning_rate": 2.9873710795031324e-05,
"loss": 0.018,
"step": 686
},
{
"epoch": 0.35600466381655654,
"grad_norm": 0.14848865568637848,
"learning_rate": 2.984449265491937e-05,
"loss": 0.0104,
"step": 687
},
{
"epoch": 0.3565228656561731,
"grad_norm": 0.2981567680835724,
"learning_rate": 2.9815246759540523e-05,
"loss": 0.0229,
"step": 688
},
{
"epoch": 0.3570410674957896,
"grad_norm": 0.15566019713878632,
"learning_rate": 2.9785973191349807e-05,
"loss": 0.0138,
"step": 689
},
{
"epoch": 0.35755926933540616,
"grad_norm": 0.3972475826740265,
"learning_rate": 2.9756672032880226e-05,
"loss": 0.0202,
"step": 690
},
{
"epoch": 0.3580774711750227,
"grad_norm": 0.18364618718624115,
"learning_rate": 2.9727343366742593e-05,
"loss": 0.0143,
"step": 691
},
{
"epoch": 0.3585956730146392,
"grad_norm": 0.6787497401237488,
"learning_rate": 2.969798727562526e-05,
"loss": 0.0212,
"step": 692
},
{
"epoch": 0.3591138748542557,
"grad_norm": 0.19514262676239014,
"learning_rate": 2.9668603842293914e-05,
"loss": 0.0151,
"step": 693
},
{
"epoch": 0.35963207669387226,
"grad_norm": 0.17772996425628662,
"learning_rate": 2.9639193149591317e-05,
"loss": 0.0171,
"step": 694
},
{
"epoch": 0.3601502785334888,
"grad_norm": 0.1942756623029709,
"learning_rate": 2.9609755280437107e-05,
"loss": 0.0191,
"step": 695
},
{
"epoch": 0.36066848037310534,
"grad_norm": 0.1511867791414261,
"learning_rate": 2.9580290317827512e-05,
"loss": 0.004,
"step": 696
},
{
"epoch": 0.3611866822127219,
"grad_norm": 0.35033461451530457,
"learning_rate": 2.9550798344835176e-05,
"loss": 0.0239,
"step": 697
},
{
"epoch": 0.36170488405233836,
"grad_norm": 0.2694636881351471,
"learning_rate": 2.9521279444608875e-05,
"loss": 0.0276,
"step": 698
},
{
"epoch": 0.3622230858919549,
"grad_norm": 0.2565995752811432,
"learning_rate": 2.9491733700373307e-05,
"loss": 0.0313,
"step": 699
},
{
"epoch": 0.36274128773157144,
"grad_norm": 0.42613688111305237,
"learning_rate": 2.9462161195428862e-05,
"loss": 0.0224,
"step": 700
},
{
"epoch": 0.363259489571188,
"grad_norm": 0.2020256370306015,
"learning_rate": 2.9432562013151372e-05,
"loss": 0.0222,
"step": 701
},
{
"epoch": 0.3637776914108045,
"grad_norm": 0.35377201437950134,
"learning_rate": 2.940293623699187e-05,
"loss": 0.0265,
"step": 702
},
{
"epoch": 0.36429589325042105,
"grad_norm": 0.42614316940307617,
"learning_rate": 2.93732839504764e-05,
"loss": 0.0334,
"step": 703
},
{
"epoch": 0.3648140950900376,
"grad_norm": 0.26751890778541565,
"learning_rate": 2.934360523720571e-05,
"loss": 0.0112,
"step": 704
},
{
"epoch": 0.3653322969296541,
"grad_norm": 0.1458798050880432,
"learning_rate": 2.9313900180855087e-05,
"loss": 0.0113,
"step": 705
},
{
"epoch": 0.3658504987692706,
"grad_norm": 0.20325633883476257,
"learning_rate": 2.9284168865174058e-05,
"loss": 0.0162,
"step": 706
},
{
"epoch": 0.36636870060888715,
"grad_norm": 0.3195553123950958,
"learning_rate": 2.9254411373986218e-05,
"loss": 0.018,
"step": 707
},
{
"epoch": 0.3668869024485037,
"grad_norm": 0.14034762978553772,
"learning_rate": 2.922462779118894e-05,
"loss": 0.0124,
"step": 708
},
{
"epoch": 0.36740510428812023,
"grad_norm": 0.12469898164272308,
"learning_rate": 2.9194818200753157e-05,
"loss": 0.0122,
"step": 709
},
{
"epoch": 0.36792330612773677,
"grad_norm": 0.15942120552062988,
"learning_rate": 2.916498268672313e-05,
"loss": 0.0159,
"step": 710
},
{
"epoch": 0.3684415079673533,
"grad_norm": 0.33030030131340027,
"learning_rate": 2.9135121333216218e-05,
"loss": 0.0164,
"step": 711
},
{
"epoch": 0.3689597098069698,
"grad_norm": 0.4804418087005615,
"learning_rate": 2.910523422442262e-05,
"loss": 0.0457,
"step": 712
},
{
"epoch": 0.36947791164658633,
"grad_norm": 0.14172638952732086,
"learning_rate": 2.907532144460516e-05,
"loss": 0.0131,
"step": 713
},
{
"epoch": 0.36999611348620287,
"grad_norm": 0.14897310733795166,
"learning_rate": 2.9045383078099018e-05,
"loss": 0.0091,
"step": 714
},
{
"epoch": 0.3705143153258194,
"grad_norm": 0.3274158239364624,
"learning_rate": 2.901541920931154e-05,
"loss": 0.0211,
"step": 715
},
{
"epoch": 0.37103251716543595,
"grad_norm": 0.367117702960968,
"learning_rate": 2.8985429922721953e-05,
"loss": 0.0283,
"step": 716
},
{
"epoch": 0.3715507190050525,
"grad_norm": 0.2538760304450989,
"learning_rate": 2.895541530288115e-05,
"loss": 0.0162,
"step": 717
},
{
"epoch": 0.37206892084466897,
"grad_norm": 0.16373780369758606,
"learning_rate": 2.892537543441144e-05,
"loss": 0.0129,
"step": 718
},
{
"epoch": 0.3725871226842855,
"grad_norm": 0.21004892885684967,
"learning_rate": 2.889531040200633e-05,
"loss": 0.0172,
"step": 719
},
{
"epoch": 0.37310532452390205,
"grad_norm": 0.15073063969612122,
"learning_rate": 2.8865220290430275e-05,
"loss": 0.018,
"step": 720
},
{
"epoch": 0.3736235263635186,
"grad_norm": 0.3549087345600128,
"learning_rate": 2.8835105184518438e-05,
"loss": 0.025,
"step": 721
},
{
"epoch": 0.3741417282031351,
"grad_norm": 0.3787144720554352,
"learning_rate": 2.880496516917642e-05,
"loss": 0.0116,
"step": 722
},
{
"epoch": 0.37465993004275167,
"grad_norm": 0.25196373462677,
"learning_rate": 2.8774800329380103e-05,
"loss": 0.014,
"step": 723
},
{
"epoch": 0.3751781318823682,
"grad_norm": 0.24565114080905914,
"learning_rate": 2.874461075017531e-05,
"loss": 0.0183,
"step": 724
},
{
"epoch": 0.3756963337219847,
"grad_norm": 0.11303957551717758,
"learning_rate": 2.871439651667764e-05,
"loss": 0.0088,
"step": 725
},
{
"epoch": 0.3762145355616012,
"grad_norm": 0.21302036941051483,
"learning_rate": 2.86841577140722e-05,
"loss": 0.0176,
"step": 726
},
{
"epoch": 0.37673273740121777,
"grad_norm": 0.31156080961227417,
"learning_rate": 2.865389442761336e-05,
"loss": 0.0213,
"step": 727
},
{
"epoch": 0.3772509392408343,
"grad_norm": 0.1727733612060547,
"learning_rate": 2.8623606742624513e-05,
"loss": 0.0161,
"step": 728
},
{
"epoch": 0.37776914108045084,
"grad_norm": 0.3240056037902832,
"learning_rate": 2.8593294744497857e-05,
"loss": 0.0202,
"step": 729
},
{
"epoch": 0.3782873429200674,
"grad_norm": 0.2706719636917114,
"learning_rate": 2.8562958518694123e-05,
"loss": 0.0228,
"step": 730
},
{
"epoch": 0.3788055447596839,
"grad_norm": 0.22442671656608582,
"learning_rate": 2.8532598150742364e-05,
"loss": 0.029,
"step": 731
},
{
"epoch": 0.3793237465993004,
"grad_norm": 0.15026988089084625,
"learning_rate": 2.8502213726239678e-05,
"loss": 0.0143,
"step": 732
},
{
"epoch": 0.37984194843891694,
"grad_norm": 0.3866124749183655,
"learning_rate": 2.8471805330851006e-05,
"loss": 0.0396,
"step": 733
},
{
"epoch": 0.3803601502785335,
"grad_norm": 0.1631077378988266,
"learning_rate": 2.844137305030886e-05,
"loss": 0.0144,
"step": 734
},
{
"epoch": 0.38087835211815,
"grad_norm": 0.258539080619812,
"learning_rate": 2.84109169704131e-05,
"loss": 0.0221,
"step": 735
},
{
"epoch": 0.38139655395776656,
"grad_norm": 0.09007499366998672,
"learning_rate": 2.8380437177030687e-05,
"loss": 0.0068,
"step": 736
},
{
"epoch": 0.3819147557973831,
"grad_norm": 0.1124981939792633,
"learning_rate": 2.8349933756095427e-05,
"loss": 0.0086,
"step": 737
},
{
"epoch": 0.38243295763699964,
"grad_norm": 0.21551914513111115,
"learning_rate": 2.8319406793607752e-05,
"loss": 0.0114,
"step": 738
},
{
"epoch": 0.3829511594766161,
"grad_norm": 0.1445673406124115,
"learning_rate": 2.828885637563447e-05,
"loss": 0.0079,
"step": 739
},
{
"epoch": 0.38346936131623266,
"grad_norm": 0.25108802318573,
"learning_rate": 2.82582825883085e-05,
"loss": 0.0225,
"step": 740
},
{
"epoch": 0.3839875631558492,
"grad_norm": 0.1586422175168991,
"learning_rate": 2.8227685517828662e-05,
"loss": 0.0131,
"step": 741
},
{
"epoch": 0.38450576499546574,
"grad_norm": 0.1375976800918579,
"learning_rate": 2.8197065250459422e-05,
"loss": 0.0133,
"step": 742
},
{
"epoch": 0.3850239668350823,
"grad_norm": 0.08903449028730392,
"learning_rate": 2.8166421872530647e-05,
"loss": 0.008,
"step": 743
},
{
"epoch": 0.3855421686746988,
"grad_norm": 0.21467937529087067,
"learning_rate": 2.813575547043734e-05,
"loss": 0.0219,
"step": 744
},
{
"epoch": 0.3860603705143153,
"grad_norm": 0.2640451192855835,
"learning_rate": 2.8105066130639457e-05,
"loss": 0.0155,
"step": 745
},
{
"epoch": 0.38657857235393184,
"grad_norm": 0.11930115520954132,
"learning_rate": 2.8074353939661575e-05,
"loss": 0.008,
"step": 746
},
{
"epoch": 0.3870967741935484,
"grad_norm": 0.21493662893772125,
"learning_rate": 2.804361898409274e-05,
"loss": 0.0091,
"step": 747
},
{
"epoch": 0.3876149760331649,
"grad_norm": 0.18691658973693848,
"learning_rate": 2.8012861350586155e-05,
"loss": 0.0147,
"step": 748
},
{
"epoch": 0.38813317787278145,
"grad_norm": 0.19181039929389954,
"learning_rate": 2.798208112585897e-05,
"loss": 0.027,
"step": 749
},
{
"epoch": 0.388651379712398,
"grad_norm": 0.2299458533525467,
"learning_rate": 2.795127839669203e-05,
"loss": 0.0079,
"step": 750
},
{
"epoch": 0.38916958155201453,
"grad_norm": 0.1063414141535759,
"learning_rate": 2.792045324992962e-05,
"loss": 0.004,
"step": 751
},
{
"epoch": 0.389687783391631,
"grad_norm": 0.22242259979248047,
"learning_rate": 2.7889605772479233e-05,
"loss": 0.0254,
"step": 752
},
{
"epoch": 0.39020598523124755,
"grad_norm": 0.25379589200019836,
"learning_rate": 2.7858736051311325e-05,
"loss": 0.0221,
"step": 753
},
{
"epoch": 0.3907241870708641,
"grad_norm": 0.0877365842461586,
"learning_rate": 2.782784417345905e-05,
"loss": 0.0034,
"step": 754
},
{
"epoch": 0.39124238891048063,
"grad_norm": 0.3404882848262787,
"learning_rate": 2.779693022601805e-05,
"loss": 0.0268,
"step": 755
},
{
"epoch": 0.39176059075009717,
"grad_norm": 0.39878952503204346,
"learning_rate": 2.7765994296146176e-05,
"loss": 0.0248,
"step": 756
},
{
"epoch": 0.3922787925897137,
"grad_norm": 0.6084442138671875,
"learning_rate": 2.7735036471063265e-05,
"loss": 0.0392,
"step": 757
},
{
"epoch": 0.39279699442933025,
"grad_norm": 0.35335928201675415,
"learning_rate": 2.770405683805087e-05,
"loss": 0.0192,
"step": 758
},
{
"epoch": 0.39331519626894673,
"grad_norm": 0.34677132964134216,
"learning_rate": 2.7673055484452042e-05,
"loss": 0.0205,
"step": 759
},
{
"epoch": 0.39383339810856327,
"grad_norm": 0.3506069481372833,
"learning_rate": 2.7642032497671065e-05,
"loss": 0.0216,
"step": 760
},
{
"epoch": 0.3943515999481798,
"grad_norm": 0.0815180316567421,
"learning_rate": 2.761098796517322e-05,
"loss": 0.0061,
"step": 761
},
{
"epoch": 0.39486980178779635,
"grad_norm": 0.15629258751869202,
"learning_rate": 2.7579921974484532e-05,
"loss": 0.0132,
"step": 762
},
{
"epoch": 0.3953880036274129,
"grad_norm": 0.06827951967716217,
"learning_rate": 2.754883461319151e-05,
"loss": 0.0065,
"step": 763
},
{
"epoch": 0.3959062054670294,
"grad_norm": 0.23429635167121887,
"learning_rate": 2.7517725968940935e-05,
"loss": 0.015,
"step": 764
},
{
"epoch": 0.39642440730664597,
"grad_norm": 0.40686559677124023,
"learning_rate": 2.748659612943958e-05,
"loss": 0.0325,
"step": 765
},
{
"epoch": 0.39694260914626245,
"grad_norm": 0.34594210982322693,
"learning_rate": 2.745544518245398e-05,
"loss": 0.0323,
"step": 766
},
{
"epoch": 0.397460810985879,
"grad_norm": 0.19717127084732056,
"learning_rate": 2.7424273215810192e-05,
"loss": 0.0093,
"step": 767
},
{
"epoch": 0.3979790128254955,
"grad_norm": 0.2936684191226959,
"learning_rate": 2.73930803173935e-05,
"loss": 0.0187,
"step": 768
},
{
"epoch": 0.39849721466511207,
"grad_norm": 0.25458282232284546,
"learning_rate": 2.7361866575148243e-05,
"loss": 0.03,
"step": 769
},
{
"epoch": 0.3990154165047286,
"grad_norm": 0.19077961146831512,
"learning_rate": 2.7330632077077497e-05,
"loss": 0.0164,
"step": 770
},
{
"epoch": 0.39953361834434514,
"grad_norm": 0.1585199385881424,
"learning_rate": 2.729937691124288e-05,
"loss": 0.014,
"step": 771
},
{
"epoch": 0.4000518201839616,
"grad_norm": 0.0987677276134491,
"learning_rate": 2.7268101165764256e-05,
"loss": 0.0085,
"step": 772
},
{
"epoch": 0.40057002202357817,
"grad_norm": 0.13913258910179138,
"learning_rate": 2.7236804928819525e-05,
"loss": 0.0128,
"step": 773
},
{
"epoch": 0.4010882238631947,
"grad_norm": 0.1244295984506607,
"learning_rate": 2.720548828864436e-05,
"loss": 0.0092,
"step": 774
},
{
"epoch": 0.40160642570281124,
"grad_norm": 0.461353063583374,
"learning_rate": 2.7174151333531953e-05,
"loss": 0.0264,
"step": 775
},
{
"epoch": 0.4021246275424278,
"grad_norm": 0.23609624803066254,
"learning_rate": 2.714279415183277e-05,
"loss": 0.0269,
"step": 776
},
{
"epoch": 0.4026428293820443,
"grad_norm": 0.1776999682188034,
"learning_rate": 2.7111416831954324e-05,
"loss": 0.0173,
"step": 777
},
{
"epoch": 0.40316103122166086,
"grad_norm": 0.15312516689300537,
"learning_rate": 2.708001946236087e-05,
"loss": 0.012,
"step": 778
},
{
"epoch": 0.40367923306127734,
"grad_norm": 0.2107289731502533,
"learning_rate": 2.7048602131573215e-05,
"loss": 0.0114,
"step": 779
},
{
"epoch": 0.4041974349008939,
"grad_norm": 0.34796831011772156,
"learning_rate": 2.7017164928168442e-05,
"loss": 0.0275,
"step": 780
},
{
"epoch": 0.4047156367405104,
"grad_norm": 0.22923240065574646,
"learning_rate": 2.698570794077965e-05,
"loss": 0.0077,
"step": 781
},
{
"epoch": 0.40523383858012696,
"grad_norm": 0.15492555499076843,
"learning_rate": 2.6954231258095728e-05,
"loss": 0.0237,
"step": 782
},
{
"epoch": 0.4057520404197435,
"grad_norm": 0.20873203873634338,
"learning_rate": 2.6922734968861095e-05,
"loss": 0.0115,
"step": 783
},
{
"epoch": 0.40627024225936004,
"grad_norm": 0.3824748396873474,
"learning_rate": 2.689121916187543e-05,
"loss": 0.0291,
"step": 784
},
{
"epoch": 0.4067884440989766,
"grad_norm": 0.2510291337966919,
"learning_rate": 2.6859683925993463e-05,
"loss": 0.0294,
"step": 785
},
{
"epoch": 0.40730664593859306,
"grad_norm": 0.14085540175437927,
"learning_rate": 2.6828129350124688e-05,
"loss": 0.0114,
"step": 786
},
{
"epoch": 0.4078248477782096,
"grad_norm": 0.17253312468528748,
"learning_rate": 2.679655552323313e-05,
"loss": 0.0104,
"step": 787
},
{
"epoch": 0.40834304961782614,
"grad_norm": 0.21511492133140564,
"learning_rate": 2.6764962534337082e-05,
"loss": 0.0165,
"step": 788
},
{
"epoch": 0.4088612514574427,
"grad_norm": 0.26060453057289124,
"learning_rate": 2.673335047250888e-05,
"loss": 0.0173,
"step": 789
},
{
"epoch": 0.4093794532970592,
"grad_norm": 0.07147006690502167,
"learning_rate": 2.670171942687461e-05,
"loss": 0.0041,
"step": 790
},
{
"epoch": 0.40989765513667575,
"grad_norm": 0.12942592799663544,
"learning_rate": 2.66700694866139e-05,
"loss": 0.0106,
"step": 791
},
{
"epoch": 0.4104158569762923,
"grad_norm": 0.1451411247253418,
"learning_rate": 2.663840074095963e-05,
"loss": 0.0105,
"step": 792
},
{
"epoch": 0.4109340588159088,
"grad_norm": 0.18926863372325897,
"learning_rate": 2.6606713279197722e-05,
"loss": 0.0227,
"step": 793
},
{
"epoch": 0.4114522606555253,
"grad_norm": 0.4707081913948059,
"learning_rate": 2.6575007190666843e-05,
"loss": 0.0452,
"step": 794
},
{
"epoch": 0.41197046249514185,
"grad_norm": 0.21161231398582458,
"learning_rate": 2.65432825647582e-05,
"loss": 0.016,
"step": 795
},
{
"epoch": 0.4124886643347584,
"grad_norm": 0.1153176873922348,
"learning_rate": 2.6511539490915233e-05,
"loss": 0.0049,
"step": 796
},
{
"epoch": 0.41300686617437493,
"grad_norm": 0.15783904492855072,
"learning_rate": 2.6479778058633426e-05,
"loss": 0.0132,
"step": 797
},
{
"epoch": 0.41352506801399147,
"grad_norm": 0.46011635661125183,
"learning_rate": 2.6447998357460006e-05,
"loss": 0.025,
"step": 798
},
{
"epoch": 0.41404326985360795,
"grad_norm": 0.31267356872558594,
"learning_rate": 2.6416200476993702e-05,
"loss": 0.0225,
"step": 799
},
{
"epoch": 0.4145614716932245,
"grad_norm": 0.17100736498832703,
"learning_rate": 2.6384384506884496e-05,
"loss": 0.0093,
"step": 800
},
{
"epoch": 0.41507967353284103,
"grad_norm": 0.48631441593170166,
"learning_rate": 2.6352550536833397e-05,
"loss": 0.0264,
"step": 801
},
{
"epoch": 0.41559787537245757,
"grad_norm": 0.1943615972995758,
"learning_rate": 2.6320698656592126e-05,
"loss": 0.0157,
"step": 802
},
{
"epoch": 0.4161160772120741,
"grad_norm": 0.2109791487455368,
"learning_rate": 2.6288828955962934e-05,
"loss": 0.0202,
"step": 803
},
{
"epoch": 0.41663427905169065,
"grad_norm": 0.2340904027223587,
"learning_rate": 2.6256941524798278e-05,
"loss": 0.028,
"step": 804
},
{
"epoch": 0.4171524808913072,
"grad_norm": 0.2932374179363251,
"learning_rate": 2.6225036453000643e-05,
"loss": 0.0189,
"step": 805
},
{
"epoch": 0.41767068273092367,
"grad_norm": 0.1333383321762085,
"learning_rate": 2.619311383052222e-05,
"loss": 0.0089,
"step": 806
},
{
"epoch": 0.4181888845705402,
"grad_norm": 0.1482066810131073,
"learning_rate": 2.6161173747364694e-05,
"loss": 0.0073,
"step": 807
},
{
"epoch": 0.41870708641015675,
"grad_norm": 0.16888538002967834,
"learning_rate": 2.6129216293578978e-05,
"loss": 0.0133,
"step": 808
},
{
"epoch": 0.4192252882497733,
"grad_norm": 0.32865652441978455,
"learning_rate": 2.609724155926495e-05,
"loss": 0.0335,
"step": 809
},
{
"epoch": 0.4197434900893898,
"grad_norm": 0.18831047415733337,
"learning_rate": 2.6065249634571215e-05,
"loss": 0.0128,
"step": 810
},
{
"epoch": 0.42026169192900636,
"grad_norm": 0.1461619734764099,
"learning_rate": 2.6033240609694864e-05,
"loss": 0.0159,
"step": 811
},
{
"epoch": 0.4207798937686229,
"grad_norm": 0.13002753257751465,
"learning_rate": 2.600121457488116e-05,
"loss": 0.0153,
"step": 812
},
{
"epoch": 0.4212980956082394,
"grad_norm": 0.25124236941337585,
"learning_rate": 2.596917162042335e-05,
"loss": 0.0219,
"step": 813
},
{
"epoch": 0.4218162974478559,
"grad_norm": 0.12394192814826965,
"learning_rate": 2.593711183666237e-05,
"loss": 0.0144,
"step": 814
},
{
"epoch": 0.42233449928747246,
"grad_norm": 0.3708241283893585,
"learning_rate": 2.5905035313986622e-05,
"loss": 0.0425,
"step": 815
},
{
"epoch": 0.422852701127089,
"grad_norm": 0.2006586492061615,
"learning_rate": 2.5872942142831692e-05,
"loss": 0.0299,
"step": 816
},
{
"epoch": 0.42337090296670554,
"grad_norm": 0.2943853735923767,
"learning_rate": 2.5840832413680096e-05,
"loss": 0.0456,
"step": 817
},
{
"epoch": 0.4238891048063221,
"grad_norm": 0.1277809590101242,
"learning_rate": 2.5808706217061038e-05,
"loss": 0.0115,
"step": 818
},
{
"epoch": 0.4244073066459386,
"grad_norm": 0.16021910309791565,
"learning_rate": 2.5776563643550153e-05,
"loss": 0.0186,
"step": 819
},
{
"epoch": 0.4249255084855551,
"grad_norm": 0.13433189690113068,
"learning_rate": 2.574440478376925e-05,
"loss": 0.0206,
"step": 820
},
{
"epoch": 0.42544371032517164,
"grad_norm": 0.19457033276557922,
"learning_rate": 2.5712229728386054e-05,
"loss": 0.0106,
"step": 821
},
{
"epoch": 0.4259619121647882,
"grad_norm": 0.3658684194087982,
"learning_rate": 2.5680038568113944e-05,
"loss": 0.03,
"step": 822
},
{
"epoch": 0.4264801140044047,
"grad_norm": 0.23988570272922516,
"learning_rate": 2.564783139371172e-05,
"loss": 0.0288,
"step": 823
},
{
"epoch": 0.42699831584402126,
"grad_norm": 0.16135792434215546,
"learning_rate": 2.5615608295983308e-05,
"loss": 0.0181,
"step": 824
},
{
"epoch": 0.4275165176836378,
"grad_norm": 0.22570829093456268,
"learning_rate": 2.5583369365777556e-05,
"loss": 0.0112,
"step": 825
},
{
"epoch": 0.4280347195232543,
"grad_norm": 0.18531742691993713,
"learning_rate": 2.555111469398793e-05,
"loss": 0.0145,
"step": 826
},
{
"epoch": 0.4285529213628708,
"grad_norm": 0.13606177270412445,
"learning_rate": 2.551884437155228e-05,
"loss": 0.0085,
"step": 827
},
{
"epoch": 0.42907112320248736,
"grad_norm": 0.10950173437595367,
"learning_rate": 2.548655848945258e-05,
"loss": 0.0094,
"step": 828
},
{
"epoch": 0.4295893250421039,
"grad_norm": 0.2311939150094986,
"learning_rate": 2.5454257138714686e-05,
"loss": 0.0131,
"step": 829
},
{
"epoch": 0.43010752688172044,
"grad_norm": 0.14860382676124573,
"learning_rate": 2.5421940410408047e-05,
"loss": 0.017,
"step": 830
},
{
"epoch": 0.430625728721337,
"grad_norm": 0.10067587345838547,
"learning_rate": 2.5389608395645485e-05,
"loss": 0.0129,
"step": 831
},
{
"epoch": 0.4311439305609535,
"grad_norm": 0.210439532995224,
"learning_rate": 2.5357261185582905e-05,
"loss": 0.0206,
"step": 832
},
{
"epoch": 0.43166213240057,
"grad_norm": 0.161096453666687,
"learning_rate": 2.532489887141906e-05,
"loss": 0.0152,
"step": 833
},
{
"epoch": 0.43218033424018654,
"grad_norm": 0.3122105896472931,
"learning_rate": 2.5292521544395295e-05,
"loss": 0.0086,
"step": 834
},
{
"epoch": 0.4326985360798031,
"grad_norm": 0.13776956498622894,
"learning_rate": 2.5260129295795257e-05,
"loss": 0.0183,
"step": 835
},
{
"epoch": 0.4332167379194196,
"grad_norm": 0.13521309196949005,
"learning_rate": 2.522772221694469e-05,
"loss": 0.0128,
"step": 836
},
{
"epoch": 0.43373493975903615,
"grad_norm": 0.14316008985042572,
"learning_rate": 2.5195300399211137e-05,
"loss": 0.0108,
"step": 837
},
{
"epoch": 0.4342531415986527,
"grad_norm": 0.2449924498796463,
"learning_rate": 2.5162863934003693e-05,
"loss": 0.0181,
"step": 838
},
{
"epoch": 0.43477134343826923,
"grad_norm": 0.263105571269989,
"learning_rate": 2.5130412912772762e-05,
"loss": 0.03,
"step": 839
},
{
"epoch": 0.4352895452778857,
"grad_norm": 0.33116382360458374,
"learning_rate": 2.5097947427009767e-05,
"loss": 0.0178,
"step": 840
},
{
"epoch": 0.43580774711750225,
"grad_norm": 0.24194487929344177,
"learning_rate": 2.506546756824694e-05,
"loss": 0.0125,
"step": 841
},
{
"epoch": 0.4363259489571188,
"grad_norm": 0.3121300935745239,
"learning_rate": 2.5032973428057e-05,
"loss": 0.0128,
"step": 842
},
{
"epoch": 0.43684415079673533,
"grad_norm": 0.2334538698196411,
"learning_rate": 2.500046509805296e-05,
"loss": 0.0167,
"step": 843
},
{
"epoch": 0.43736235263635187,
"grad_norm": 0.1214141920208931,
"learning_rate": 2.496794266988783e-05,
"loss": 0.0083,
"step": 844
},
{
"epoch": 0.4378805544759684,
"grad_norm": 0.4726041555404663,
"learning_rate": 2.4935406235254367e-05,
"loss": 0.0291,
"step": 845
},
{
"epoch": 0.43839875631558495,
"grad_norm": 0.10701345652341843,
"learning_rate": 2.490285588588481e-05,
"loss": 0.0064,
"step": 846
},
{
"epoch": 0.43891695815520143,
"grad_norm": 0.4372049868106842,
"learning_rate": 2.4870291713550648e-05,
"loss": 0.0211,
"step": 847
},
{
"epoch": 0.43943515999481797,
"grad_norm": 0.1732262223958969,
"learning_rate": 2.4837713810062317e-05,
"loss": 0.0149,
"step": 848
},
{
"epoch": 0.4399533618344345,
"grad_norm": 0.32579711079597473,
"learning_rate": 2.480512226726899e-05,
"loss": 0.0154,
"step": 849
},
{
"epoch": 0.44047156367405105,
"grad_norm": 0.2965090572834015,
"learning_rate": 2.477251717705828e-05,
"loss": 0.0325,
"step": 850
},
{
"epoch": 0.4409897655136676,
"grad_norm": 0.418514609336853,
"learning_rate": 2.4739898631355995e-05,
"loss": 0.0164,
"step": 851
},
{
"epoch": 0.4415079673532841,
"grad_norm": 0.2681904733181,
"learning_rate": 2.4707266722125888e-05,
"loss": 0.0115,
"step": 852
},
{
"epoch": 0.4420261691929006,
"grad_norm": 0.40225011110305786,
"learning_rate": 2.4674621541369392e-05,
"loss": 0.0143,
"step": 853
},
{
"epoch": 0.44254437103251715,
"grad_norm": 0.24150243401527405,
"learning_rate": 2.4641963181125332e-05,
"loss": 0.021,
"step": 854
},
{
"epoch": 0.4430625728721337,
"grad_norm": 0.23854626715183258,
"learning_rate": 2.460929173346972e-05,
"loss": 0.0231,
"step": 855
},
{
"epoch": 0.4435807747117502,
"grad_norm": 0.18777628242969513,
"learning_rate": 2.4576607290515445e-05,
"loss": 0.0114,
"step": 856
},
{
"epoch": 0.44409897655136676,
"grad_norm": 0.12525376677513123,
"learning_rate": 2.4543909944412048e-05,
"loss": 0.0095,
"step": 857
},
{
"epoch": 0.4446171783909833,
"grad_norm": 0.24290180206298828,
"learning_rate": 2.4511199787345447e-05,
"loss": 0.0189,
"step": 858
},
{
"epoch": 0.44513538023059984,
"grad_norm": 0.26251229643821716,
"learning_rate": 2.447847691153767e-05,
"loss": 0.0202,
"step": 859
},
{
"epoch": 0.4456535820702163,
"grad_norm": 0.35255348682403564,
"learning_rate": 2.4445741409246616e-05,
"loss": 0.0187,
"step": 860
},
{
"epoch": 0.44617178390983286,
"grad_norm": 0.15728411078453064,
"learning_rate": 2.4412993372765782e-05,
"loss": 0.0095,
"step": 861
},
{
"epoch": 0.4466899857494494,
"grad_norm": 0.33080241084098816,
"learning_rate": 2.438023289442399e-05,
"loss": 0.0203,
"step": 862
},
{
"epoch": 0.44720818758906594,
"grad_norm": 0.3215973973274231,
"learning_rate": 2.4347460066585146e-05,
"loss": 0.037,
"step": 863
},
{
"epoch": 0.4477263894286825,
"grad_norm": 0.25812244415283203,
"learning_rate": 2.4314674981647982e-05,
"loss": 0.0169,
"step": 864
},
{
"epoch": 0.448244591268299,
"grad_norm": 0.21332046389579773,
"learning_rate": 2.4281877732045778e-05,
"loss": 0.0195,
"step": 865
},
{
"epoch": 0.44876279310791556,
"grad_norm": 0.42107102274894714,
"learning_rate": 2.4249068410246114e-05,
"loss": 0.0157,
"step": 866
},
{
"epoch": 0.44928099494753204,
"grad_norm": 0.1260032057762146,
"learning_rate": 2.4216247108750613e-05,
"loss": 0.0156,
"step": 867
},
{
"epoch": 0.4497991967871486,
"grad_norm": 0.19122031331062317,
"learning_rate": 2.4183413920094657e-05,
"loss": 0.0083,
"step": 868
},
{
"epoch": 0.4503173986267651,
"grad_norm": 0.18674705922603607,
"learning_rate": 2.4150568936847166e-05,
"loss": 0.023,
"step": 869
},
{
"epoch": 0.45083560046638166,
"grad_norm": 0.33427202701568604,
"learning_rate": 2.4117712251610287e-05,
"loss": 0.026,
"step": 870
},
{
"epoch": 0.4513538023059982,
"grad_norm": 0.2598010003566742,
"learning_rate": 2.4084843957019174e-05,
"loss": 0.0188,
"step": 871
},
{
"epoch": 0.45187200414561474,
"grad_norm": 0.21907593309879303,
"learning_rate": 2.405196414574171e-05,
"loss": 0.0165,
"step": 872
},
{
"epoch": 0.4523902059852313,
"grad_norm": 0.20734968781471252,
"learning_rate": 2.4019072910478248e-05,
"loss": 0.0133,
"step": 873
},
{
"epoch": 0.45290840782484776,
"grad_norm": 0.18882878124713898,
"learning_rate": 2.3986170343961342e-05,
"loss": 0.0115,
"step": 874
},
{
"epoch": 0.4534266096644643,
"grad_norm": 0.2494608759880066,
"learning_rate": 2.3953256538955517e-05,
"loss": 0.0146,
"step": 875
},
{
"epoch": 0.45394481150408084,
"grad_norm": 0.22158919274806976,
"learning_rate": 2.3920331588256945e-05,
"loss": 0.0138,
"step": 876
},
{
"epoch": 0.4544630133436974,
"grad_norm": 0.3354347050189972,
"learning_rate": 2.3887395584693258e-05,
"loss": 0.0276,
"step": 877
},
{
"epoch": 0.4549812151833139,
"grad_norm": 0.20476201176643372,
"learning_rate": 2.3854448621123227e-05,
"loss": 0.0121,
"step": 878
},
{
"epoch": 0.45549941702293045,
"grad_norm": 0.36594030261039734,
"learning_rate": 2.382149079043654e-05,
"loss": 0.0271,
"step": 879
},
{
"epoch": 0.45601761886254694,
"grad_norm": 0.1319293975830078,
"learning_rate": 2.3788522185553505e-05,
"loss": 0.0088,
"step": 880
},
{
"epoch": 0.4565358207021635,
"grad_norm": 0.2355639934539795,
"learning_rate": 2.375554289942483e-05,
"loss": 0.0196,
"step": 881
},
{
"epoch": 0.45705402254178,
"grad_norm": 0.09457190334796906,
"learning_rate": 2.3722553025031304e-05,
"loss": 0.0071,
"step": 882
},
{
"epoch": 0.45757222438139655,
"grad_norm": 0.2910573184490204,
"learning_rate": 2.36895526553836e-05,
"loss": 0.0191,
"step": 883
},
{
"epoch": 0.4580904262210131,
"grad_norm": 0.36989885568618774,
"learning_rate": 2.3656541883521963e-05,
"loss": 0.0255,
"step": 884
},
{
"epoch": 0.45860862806062963,
"grad_norm": 0.29597464203834534,
"learning_rate": 2.362352080251598e-05,
"loss": 0.0198,
"step": 885
},
{
"epoch": 0.45912682990024617,
"grad_norm": 0.32694950699806213,
"learning_rate": 2.359048950546429e-05,
"loss": 0.019,
"step": 886
},
{
"epoch": 0.45964503173986265,
"grad_norm": 0.2195930778980255,
"learning_rate": 2.3557448085494343e-05,
"loss": 0.0279,
"step": 887
},
{
"epoch": 0.4601632335794792,
"grad_norm": 0.20476271212100983,
"learning_rate": 2.352439663576212e-05,
"loss": 0.0137,
"step": 888
},
{
"epoch": 0.46068143541909573,
"grad_norm": 0.3710532784461975,
"learning_rate": 2.349133524945189e-05,
"loss": 0.0216,
"step": 889
},
{
"epoch": 0.46119963725871227,
"grad_norm": 0.15248271822929382,
"learning_rate": 2.3458264019775933e-05,
"loss": 0.0161,
"step": 890
},
{
"epoch": 0.4617178390983288,
"grad_norm": 0.2434094101190567,
"learning_rate": 2.3425183039974284e-05,
"loss": 0.0175,
"step": 891
},
{
"epoch": 0.46223604093794535,
"grad_norm": 0.33260586857795715,
"learning_rate": 2.3392092403314447e-05,
"loss": 0.0206,
"step": 892
},
{
"epoch": 0.4627542427775619,
"grad_norm": 0.10504089295864105,
"learning_rate": 2.3358992203091185e-05,
"loss": 0.0059,
"step": 893
},
{
"epoch": 0.46327244461717837,
"grad_norm": 0.08952392637729645,
"learning_rate": 2.3325882532626203e-05,
"loss": 0.011,
"step": 894
},
{
"epoch": 0.4637906464567949,
"grad_norm": 0.1848842054605484,
"learning_rate": 2.3292763485267917e-05,
"loss": 0.015,
"step": 895
},
{
"epoch": 0.46430884829641145,
"grad_norm": 0.1017458513379097,
"learning_rate": 2.325963515439116e-05,
"loss": 0.0131,
"step": 896
},
{
"epoch": 0.464827050136028,
"grad_norm": 0.1866416186094284,
"learning_rate": 2.3226497633396978e-05,
"loss": 0.0142,
"step": 897
},
{
"epoch": 0.4653452519756445,
"grad_norm": 0.19123072922229767,
"learning_rate": 2.3193351015712278e-05,
"loss": 0.0155,
"step": 898
},
{
"epoch": 0.46586345381526106,
"grad_norm": 0.1844363510608673,
"learning_rate": 2.3160195394789648e-05,
"loss": 0.0142,
"step": 899
},
{
"epoch": 0.46638165565487755,
"grad_norm": 0.12927821278572083,
"learning_rate": 2.3127030864107044e-05,
"loss": 0.008,
"step": 900
},
{
"epoch": 0.4668998574944941,
"grad_norm": 0.2421107441186905,
"learning_rate": 2.3093857517167555e-05,
"loss": 0.0189,
"step": 901
},
{
"epoch": 0.4674180593341106,
"grad_norm": 0.12359347939491272,
"learning_rate": 2.3060675447499116e-05,
"loss": 0.0081,
"step": 902
},
{
"epoch": 0.46793626117372716,
"grad_norm": 0.13112476468086243,
"learning_rate": 2.302748474865426e-05,
"loss": 0.0166,
"step": 903
},
{
"epoch": 0.4684544630133437,
"grad_norm": 0.2681474983692169,
"learning_rate": 2.299428551420984e-05,
"loss": 0.0217,
"step": 904
},
{
"epoch": 0.46897266485296024,
"grad_norm": 0.12790656089782715,
"learning_rate": 2.296107783776679e-05,
"loss": 0.0084,
"step": 905
},
{
"epoch": 0.4694908666925768,
"grad_norm": 0.21175391972064972,
"learning_rate": 2.2927861812949823e-05,
"loss": 0.0123,
"step": 906
},
{
"epoch": 0.47000906853219326,
"grad_norm": 0.14393571019172668,
"learning_rate": 2.2894637533407212e-05,
"loss": 0.0097,
"step": 907
},
{
"epoch": 0.4705272703718098,
"grad_norm": 0.21133379638195038,
"learning_rate": 2.2861405092810492e-05,
"loss": 0.0175,
"step": 908
},
{
"epoch": 0.47104547221142634,
"grad_norm": 0.12299802154302597,
"learning_rate": 2.2828164584854204e-05,
"loss": 0.0091,
"step": 909
},
{
"epoch": 0.4715636740510429,
"grad_norm": 0.29604965448379517,
"learning_rate": 2.2794916103255632e-05,
"loss": 0.0153,
"step": 910
},
{
"epoch": 0.4720818758906594,
"grad_norm": 0.3092305362224579,
"learning_rate": 2.2761659741754562e-05,
"loss": 0.0142,
"step": 911
},
{
"epoch": 0.47260007773027596,
"grad_norm": 0.3267214000225067,
"learning_rate": 2.2728395594112965e-05,
"loss": 0.036,
"step": 912
},
{
"epoch": 0.4731182795698925,
"grad_norm": 0.13050074875354767,
"learning_rate": 2.269512375411479e-05,
"loss": 0.0096,
"step": 913
},
{
"epoch": 0.473636481409509,
"grad_norm": 0.38647061586380005,
"learning_rate": 2.266184431556566e-05,
"loss": 0.0173,
"step": 914
},
{
"epoch": 0.4741546832491255,
"grad_norm": 0.1383042186498642,
"learning_rate": 2.2628557372292618e-05,
"loss": 0.0133,
"step": 915
},
{
"epoch": 0.47467288508874206,
"grad_norm": 0.19705870747566223,
"learning_rate": 2.2595263018143885e-05,
"loss": 0.0126,
"step": 916
},
{
"epoch": 0.4751910869283586,
"grad_norm": 0.19639675319194794,
"learning_rate": 2.2561961346988553e-05,
"loss": 0.0104,
"step": 917
},
{
"epoch": 0.47570928876797514,
"grad_norm": 0.28665390610694885,
"learning_rate": 2.2528652452716354e-05,
"loss": 0.0236,
"step": 918
},
{
"epoch": 0.4762274906075917,
"grad_norm": 0.15427164733409882,
"learning_rate": 2.2495336429237388e-05,
"loss": 0.0151,
"step": 919
},
{
"epoch": 0.4767456924472082,
"grad_norm": 0.36349889636039734,
"learning_rate": 2.246201337048185e-05,
"loss": 0.0209,
"step": 920
},
{
"epoch": 0.4772638942868247,
"grad_norm": 0.21488307416439056,
"learning_rate": 2.2428683370399767e-05,
"loss": 0.0197,
"step": 921
},
{
"epoch": 0.47778209612644124,
"grad_norm": 0.27596259117126465,
"learning_rate": 2.2395346522960742e-05,
"loss": 0.0231,
"step": 922
},
{
"epoch": 0.4783002979660578,
"grad_norm": 0.21963416039943695,
"learning_rate": 2.2362002922153686e-05,
"loss": 0.0193,
"step": 923
},
{
"epoch": 0.4788184998056743,
"grad_norm": 0.25122183561325073,
"learning_rate": 2.232865266198654e-05,
"loss": 0.0182,
"step": 924
},
{
"epoch": 0.47933670164529085,
"grad_norm": 0.2015954703092575,
"learning_rate": 2.2295295836486028e-05,
"loss": 0.0127,
"step": 925
},
{
"epoch": 0.4798549034849074,
"grad_norm": 0.1243843361735344,
"learning_rate": 2.226193253969738e-05,
"loss": 0.0078,
"step": 926
},
{
"epoch": 0.4803731053245239,
"grad_norm": 0.5754460096359253,
"learning_rate": 2.222856286568408e-05,
"loss": 0.0329,
"step": 927
},
{
"epoch": 0.4808913071641404,
"grad_norm": 0.1487659215927124,
"learning_rate": 2.219518690852757e-05,
"loss": 0.0069,
"step": 928
},
{
"epoch": 0.48140950900375695,
"grad_norm": 0.1283818930387497,
"learning_rate": 2.2161804762327036e-05,
"loss": 0.0036,
"step": 929
},
{
"epoch": 0.4819277108433735,
"grad_norm": 0.26554226875305176,
"learning_rate": 2.2128416521199085e-05,
"loss": 0.0109,
"step": 930
},
{
"epoch": 0.48244591268299003,
"grad_norm": 0.2600436508655548,
"learning_rate": 2.2095022279277535e-05,
"loss": 0.0229,
"step": 931
},
{
"epoch": 0.48296411452260657,
"grad_norm": 0.5942787528038025,
"learning_rate": 2.2061622130713097e-05,
"loss": 0.0356,
"step": 932
},
{
"epoch": 0.4834823163622231,
"grad_norm": 0.30313023924827576,
"learning_rate": 2.2028216169673155e-05,
"loss": 0.0229,
"step": 933
},
{
"epoch": 0.4840005182018396,
"grad_norm": 0.06324416399002075,
"learning_rate": 2.1994804490341467e-05,
"loss": 0.0019,
"step": 934
},
{
"epoch": 0.48451872004145613,
"grad_norm": 0.3620297312736511,
"learning_rate": 2.1961387186917926e-05,
"loss": 0.0128,
"step": 935
},
{
"epoch": 0.48503692188107267,
"grad_norm": 0.5341509580612183,
"learning_rate": 2.1927964353618253e-05,
"loss": 0.0253,
"step": 936
},
{
"epoch": 0.4855551237206892,
"grad_norm": 0.1990864872932434,
"learning_rate": 2.18945360846738e-05,
"loss": 0.0103,
"step": 937
},
{
"epoch": 0.48607332556030575,
"grad_norm": 0.05690270662307739,
"learning_rate": 2.186110247433122e-05,
"loss": 0.0058,
"step": 938
},
{
"epoch": 0.4865915273999223,
"grad_norm": 0.2440994381904602,
"learning_rate": 2.1827663616852228e-05,
"loss": 0.0094,
"step": 939
},
{
"epoch": 0.4871097292395388,
"grad_norm": 0.16593553125858307,
"learning_rate": 2.1794219606513335e-05,
"loss": 0.0097,
"step": 940
},
{
"epoch": 0.4876279310791553,
"grad_norm": 0.24518586695194244,
"learning_rate": 2.176077053760558e-05,
"loss": 0.0313,
"step": 941
},
{
"epoch": 0.48814613291877185,
"grad_norm": 0.19542941451072693,
"learning_rate": 2.172731650443425e-05,
"loss": 0.023,
"step": 942
},
{
"epoch": 0.4886643347583884,
"grad_norm": 0.2354109138250351,
"learning_rate": 2.1693857601318665e-05,
"loss": 0.0114,
"step": 943
},
{
"epoch": 0.4891825365980049,
"grad_norm": 0.3304251730442047,
"learning_rate": 2.166039392259184e-05,
"loss": 0.0118,
"step": 944
},
{
"epoch": 0.48970073843762146,
"grad_norm": 0.21442767977714539,
"learning_rate": 2.1626925562600266e-05,
"loss": 0.0143,
"step": 945
},
{
"epoch": 0.490218940277238,
"grad_norm": 0.1758863776922226,
"learning_rate": 2.159345261570362e-05,
"loss": 0.0145,
"step": 946
},
{
"epoch": 0.49073714211685454,
"grad_norm": 0.3084407150745392,
"learning_rate": 2.1559975176274545e-05,
"loss": 0.015,
"step": 947
},
{
"epoch": 0.491255343956471,
"grad_norm": 0.26537203788757324,
"learning_rate": 2.152649333869831e-05,
"loss": 0.0112,
"step": 948
},
{
"epoch": 0.49177354579608756,
"grad_norm": 0.21452540159225464,
"learning_rate": 2.1493007197372614e-05,
"loss": 0.0125,
"step": 949
},
{
"epoch": 0.4922917476357041,
"grad_norm": 0.23216988146305084,
"learning_rate": 2.1459516846707267e-05,
"loss": 0.0198,
"step": 950
},
{
"epoch": 0.49280994947532064,
"grad_norm": 0.2911728024482727,
"learning_rate": 2.1426022381123967e-05,
"loss": 0.0165,
"step": 951
},
{
"epoch": 0.4933281513149372,
"grad_norm": 0.25865858793258667,
"learning_rate": 2.1392523895056e-05,
"loss": 0.0177,
"step": 952
},
{
"epoch": 0.4938463531545537,
"grad_norm": 0.2825429439544678,
"learning_rate": 2.1359021482947996e-05,
"loss": 0.0276,
"step": 953
},
{
"epoch": 0.4943645549941702,
"grad_norm": 0.09475862979888916,
"learning_rate": 2.1325515239255642e-05,
"loss": 0.0062,
"step": 954
},
{
"epoch": 0.49488275683378674,
"grad_norm": 0.29453566670417786,
"learning_rate": 2.1292005258445442e-05,
"loss": 0.0216,
"step": 955
},
{
"epoch": 0.4954009586734033,
"grad_norm": 0.15916626155376434,
"learning_rate": 2.1258491634994422e-05,
"loss": 0.0087,
"step": 956
},
{
"epoch": 0.4959191605130198,
"grad_norm": 0.29559406638145447,
"learning_rate": 2.1224974463389892e-05,
"loss": 0.0111,
"step": 957
},
{
"epoch": 0.49643736235263636,
"grad_norm": 0.19165553152561188,
"learning_rate": 2.119145383812915e-05,
"loss": 0.0189,
"step": 958
},
{
"epoch": 0.4969555641922529,
"grad_norm": 0.2623184025287628,
"learning_rate": 2.1157929853719255e-05,
"loss": 0.0228,
"step": 959
},
{
"epoch": 0.49747376603186944,
"grad_norm": 0.6063542366027832,
"learning_rate": 2.11244026046767e-05,
"loss": 0.0241,
"step": 960
},
{
"epoch": 0.4979919678714859,
"grad_norm": 0.6173031330108643,
"learning_rate": 2.109087218552723e-05,
"loss": 0.0243,
"step": 961
},
{
"epoch": 0.49851016971110246,
"grad_norm": 0.22902703285217285,
"learning_rate": 2.1057338690805485e-05,
"loss": 0.0218,
"step": 962
},
{
"epoch": 0.499028371550719,
"grad_norm": 0.2686314582824707,
"learning_rate": 2.1023802215054797e-05,
"loss": 0.022,
"step": 963
},
{
"epoch": 0.49954657339033554,
"grad_norm": 0.1550043672323227,
"learning_rate": 2.099026285282689e-05,
"loss": 0.0093,
"step": 964
},
{
"epoch": 0.5000647752299521,
"grad_norm": 0.0828891471028328,
"learning_rate": 2.095672069868165e-05,
"loss": 0.0054,
"step": 965
},
{
"epoch": 0.5005829770695686,
"grad_norm": 0.2590828239917755,
"learning_rate": 2.0923175847186797e-05,
"loss": 0.0138,
"step": 966
},
{
"epoch": 0.5011011789091852,
"grad_norm": 0.26930099725723267,
"learning_rate": 2.08896283929177e-05,
"loss": 0.0233,
"step": 967
},
{
"epoch": 0.5016193807488016,
"grad_norm": 0.23394693434238434,
"learning_rate": 2.0856078430457033e-05,
"loss": 0.0151,
"step": 968
},
{
"epoch": 0.5021375825884182,
"grad_norm": 0.18417996168136597,
"learning_rate": 2.0822526054394562e-05,
"loss": 0.013,
"step": 969
},
{
"epoch": 0.5026557844280347,
"grad_norm": 0.16068002581596375,
"learning_rate": 2.0788971359326827e-05,
"loss": 0.0128,
"step": 970
},
{
"epoch": 0.5031739862676512,
"grad_norm": 0.29491886496543884,
"learning_rate": 2.0755414439856948e-05,
"loss": 0.0224,
"step": 971
},
{
"epoch": 0.5036921881072678,
"grad_norm": 0.16440480947494507,
"learning_rate": 2.0721855390594294e-05,
"loss": 0.0135,
"step": 972
},
{
"epoch": 0.5042103899468843,
"grad_norm": 0.1847657710313797,
"learning_rate": 2.068829430615423e-05,
"loss": 0.0121,
"step": 973
},
{
"epoch": 0.5047285917865009,
"grad_norm": 0.31765875220298767,
"learning_rate": 2.0654731281157874e-05,
"loss": 0.0332,
"step": 974
},
{
"epoch": 0.5052467936261174,
"grad_norm": 0.27984386682510376,
"learning_rate": 2.062116641023181e-05,
"loss": 0.0122,
"step": 975
},
{
"epoch": 0.505764995465734,
"grad_norm": 0.24907967448234558,
"learning_rate": 2.0587599788007825e-05,
"loss": 0.018,
"step": 976
},
{
"epoch": 0.5062831973053504,
"grad_norm": 0.18288204073905945,
"learning_rate": 2.0554031509122656e-05,
"loss": 0.0122,
"step": 977
},
{
"epoch": 0.5068013991449669,
"grad_norm": 0.13284114003181458,
"learning_rate": 2.0520461668217683e-05,
"loss": 0.0083,
"step": 978
},
{
"epoch": 0.5073196009845835,
"grad_norm": 0.20933015644550323,
"learning_rate": 2.0486890359938724e-05,
"loss": 0.0162,
"step": 979
},
{
"epoch": 0.5078378028242,
"grad_norm": 0.23051418364048004,
"learning_rate": 2.045331767893571e-05,
"loss": 0.0186,
"step": 980
},
{
"epoch": 0.5083560046638166,
"grad_norm": 0.23583345115184784,
"learning_rate": 2.0419743719862447e-05,
"loss": 0.0188,
"step": 981
},
{
"epoch": 0.5088742065034331,
"grad_norm": 0.2092517614364624,
"learning_rate": 2.0386168577376346e-05,
"loss": 0.0202,
"step": 982
},
{
"epoch": 0.5093924083430497,
"grad_norm": 0.2943207025527954,
"learning_rate": 2.035259234613816e-05,
"loss": 0.0287,
"step": 983
},
{
"epoch": 0.5099106101826661,
"grad_norm": 0.18729962408542633,
"learning_rate": 2.03190151208117e-05,
"loss": 0.0166,
"step": 984
},
{
"epoch": 0.5104288120222826,
"grad_norm": 0.19974197447299957,
"learning_rate": 2.0285436996063596e-05,
"loss": 0.0148,
"step": 985
},
{
"epoch": 0.5109470138618992,
"grad_norm": 0.09993212670087814,
"learning_rate": 2.0251858066562997e-05,
"loss": 0.0042,
"step": 986
},
{
"epoch": 0.5114652157015157,
"grad_norm": 0.17130297422409058,
"learning_rate": 2.0218278426981332e-05,
"loss": 0.0117,
"step": 987
},
{
"epoch": 0.5119834175411323,
"grad_norm": 0.0700158104300499,
"learning_rate": 2.0184698171992023e-05,
"loss": 0.0046,
"step": 988
},
{
"epoch": 0.5125016193807488,
"grad_norm": 0.24379920959472656,
"learning_rate": 2.0151117396270243e-05,
"loss": 0.0132,
"step": 989
},
{
"epoch": 0.5130198212203654,
"grad_norm": 0.1190122663974762,
"learning_rate": 2.0117536194492616e-05,
"loss": 0.0107,
"step": 990
},
{
"epoch": 0.5135380230599819,
"grad_norm": 0.3585830628871918,
"learning_rate": 2.008395466133697e-05,
"loss": 0.022,
"step": 991
},
{
"epoch": 0.5140562248995983,
"grad_norm": 0.1896945983171463,
"learning_rate": 2.0050372891482065e-05,
"loss": 0.0169,
"step": 992
},
{
"epoch": 0.5145744267392149,
"grad_norm": 0.27251240611076355,
"learning_rate": 2.0016790979607353e-05,
"loss": 0.0197,
"step": 993
},
{
"epoch": 0.5150926285788314,
"grad_norm": 0.2470308244228363,
"learning_rate": 1.9983209020392657e-05,
"loss": 0.0107,
"step": 994
},
{
"epoch": 0.515610830418448,
"grad_norm": 0.1589849293231964,
"learning_rate": 1.9949627108517938e-05,
"loss": 0.0075,
"step": 995
},
{
"epoch": 0.5161290322580645,
"grad_norm": 0.283698171377182,
"learning_rate": 1.991604533866304e-05,
"loss": 0.0148,
"step": 996
},
{
"epoch": 0.5166472340976811,
"grad_norm": 0.2342384308576584,
"learning_rate": 1.988246380550739e-05,
"loss": 0.0427,
"step": 997
},
{
"epoch": 0.5171654359372976,
"grad_norm": 0.30012941360473633,
"learning_rate": 1.9848882603729763e-05,
"loss": 0.0139,
"step": 998
},
{
"epoch": 0.5176836377769141,
"grad_norm": 0.1278998851776123,
"learning_rate": 1.981530182800798e-05,
"loss": 0.0117,
"step": 999
},
{
"epoch": 0.5182018396165307,
"grad_norm": 0.31228339672088623,
"learning_rate": 1.978172157301867e-05,
"loss": 0.0191,
"step": 1000
},
{
"epoch": 0.5187200414561471,
"grad_norm": 0.13579922914505005,
"learning_rate": 1.9748141933437003e-05,
"loss": 0.0052,
"step": 1001
},
{
"epoch": 0.5192382432957637,
"grad_norm": 0.29426661133766174,
"learning_rate": 1.9714563003936414e-05,
"loss": 0.0255,
"step": 1002
},
{
"epoch": 0.5197564451353802,
"grad_norm": 0.15366186201572418,
"learning_rate": 1.9680984879188306e-05,
"loss": 0.0109,
"step": 1003
},
{
"epoch": 0.5202746469749968,
"grad_norm": 0.2718539834022522,
"learning_rate": 1.9647407653861846e-05,
"loss": 0.0136,
"step": 1004
},
{
"epoch": 0.5207928488146133,
"grad_norm": 0.307129830121994,
"learning_rate": 1.9613831422623657e-05,
"loss": 0.0193,
"step": 1005
},
{
"epoch": 0.5213110506542298,
"grad_norm": 0.21338942646980286,
"learning_rate": 1.9580256280137563e-05,
"loss": 0.0152,
"step": 1006
},
{
"epoch": 0.5218292524938464,
"grad_norm": 0.3313189446926117,
"learning_rate": 1.9546682321064296e-05,
"loss": 0.0202,
"step": 1007
},
{
"epoch": 0.5223474543334629,
"grad_norm": 0.17557553946971893,
"learning_rate": 1.951310964006128e-05,
"loss": 0.0094,
"step": 1008
},
{
"epoch": 0.5228656561730795,
"grad_norm": 0.22019444406032562,
"learning_rate": 1.9479538331782317e-05,
"loss": 0.0097,
"step": 1009
},
{
"epoch": 0.5233838580126959,
"grad_norm": 0.15170864760875702,
"learning_rate": 1.9445968490877357e-05,
"loss": 0.0075,
"step": 1010
},
{
"epoch": 0.5239020598523125,
"grad_norm": 0.5065960884094238,
"learning_rate": 1.9412400211992178e-05,
"loss": 0.0285,
"step": 1011
},
{
"epoch": 0.524420261691929,
"grad_norm": 0.14924491941928864,
"learning_rate": 1.9378833589768196e-05,
"loss": 0.0158,
"step": 1012
},
{
"epoch": 0.5249384635315455,
"grad_norm": 0.2105136215686798,
"learning_rate": 1.934526871884213e-05,
"loss": 0.0107,
"step": 1013
},
{
"epoch": 0.5254566653711621,
"grad_norm": 0.21951298415660858,
"learning_rate": 1.931170569384578e-05,
"loss": 0.0125,
"step": 1014
},
{
"epoch": 0.5259748672107786,
"grad_norm": 0.31380099058151245,
"learning_rate": 1.9278144609405712e-05,
"loss": 0.0218,
"step": 1015
},
{
"epoch": 0.5264930690503952,
"grad_norm": 0.10603537410497665,
"learning_rate": 1.9244585560143055e-05,
"loss": 0.007,
"step": 1016
},
{
"epoch": 0.5270112708900117,
"grad_norm": 0.2262280434370041,
"learning_rate": 1.9211028640673173e-05,
"loss": 0.0117,
"step": 1017
},
{
"epoch": 0.5275294727296281,
"grad_norm": 0.1582176387310028,
"learning_rate": 1.917747394560545e-05,
"loss": 0.0051,
"step": 1018
},
{
"epoch": 0.5280476745692447,
"grad_norm": 0.17075197398662567,
"learning_rate": 1.914392156954297e-05,
"loss": 0.0147,
"step": 1019
},
{
"epoch": 0.5285658764088612,
"grad_norm": 0.3450043499469757,
"learning_rate": 1.9110371607082303e-05,
"loss": 0.0201,
"step": 1020
},
{
"epoch": 0.5290840782484778,
"grad_norm": 0.11497087776660919,
"learning_rate": 1.9076824152813203e-05,
"loss": 0.0048,
"step": 1021
},
{
"epoch": 0.5296022800880943,
"grad_norm": 0.25621140003204346,
"learning_rate": 1.9043279301318365e-05,
"loss": 0.0117,
"step": 1022
},
{
"epoch": 0.5301204819277109,
"grad_norm": 0.18441203236579895,
"learning_rate": 1.9009737147173116e-05,
"loss": 0.0094,
"step": 1023
},
{
"epoch": 0.5306386837673274,
"grad_norm": 0.18254609405994415,
"learning_rate": 1.897619778494521e-05,
"loss": 0.0115,
"step": 1024
},
{
"epoch": 0.5311568856069439,
"grad_norm": 0.336505264043808,
"learning_rate": 1.894266130919452e-05,
"loss": 0.0287,
"step": 1025
},
{
"epoch": 0.5316750874465604,
"grad_norm": 0.2124987095594406,
"learning_rate": 1.8909127814472776e-05,
"loss": 0.0139,
"step": 1026
},
{
"epoch": 0.5321932892861769,
"grad_norm": 0.30894720554351807,
"learning_rate": 1.88755973953233e-05,
"loss": 0.0175,
"step": 1027
},
{
"epoch": 0.5327114911257935,
"grad_norm": 0.16775032877922058,
"learning_rate": 1.8842070146280752e-05,
"loss": 0.0101,
"step": 1028
},
{
"epoch": 0.53322969296541,
"grad_norm": 0.2674119472503662,
"learning_rate": 1.880854616187085e-05,
"loss": 0.0136,
"step": 1029
},
{
"epoch": 0.5337478948050266,
"grad_norm": 0.31328970193862915,
"learning_rate": 1.8775025536610118e-05,
"loss": 0.0193,
"step": 1030
},
{
"epoch": 0.5342660966446431,
"grad_norm": 0.18065671622753143,
"learning_rate": 1.8741508365005584e-05,
"loss": 0.0079,
"step": 1031
},
{
"epoch": 0.5347842984842596,
"grad_norm": 0.1782873570919037,
"learning_rate": 1.8707994741554565e-05,
"loss": 0.0073,
"step": 1032
},
{
"epoch": 0.5353025003238762,
"grad_norm": 0.20292030274868011,
"learning_rate": 1.867448476074436e-05,
"loss": 0.0115,
"step": 1033
},
{
"epoch": 0.5358207021634926,
"grad_norm": 0.19517368078231812,
"learning_rate": 1.8640978517052014e-05,
"loss": 0.0187,
"step": 1034
},
{
"epoch": 0.5363389040031092,
"grad_norm": 0.20960001647472382,
"learning_rate": 1.8607476104944006e-05,
"loss": 0.0142,
"step": 1035
},
{
"epoch": 0.5368571058427257,
"grad_norm": 0.5607621073722839,
"learning_rate": 1.8573977618876036e-05,
"loss": 0.0207,
"step": 1036
},
{
"epoch": 0.5373753076823423,
"grad_norm": 0.21276597678661346,
"learning_rate": 1.8540483153292736e-05,
"loss": 0.0137,
"step": 1037
},
{
"epoch": 0.5378935095219588,
"grad_norm": 0.11740380525588989,
"learning_rate": 1.85069928026274e-05,
"loss": 0.0082,
"step": 1038
},
{
"epoch": 0.5384117113615753,
"grad_norm": 0.33550262451171875,
"learning_rate": 1.8473506661301697e-05,
"loss": 0.0194,
"step": 1039
},
{
"epoch": 0.5389299132011919,
"grad_norm": 0.33623600006103516,
"learning_rate": 1.8440024823725462e-05,
"loss": 0.0336,
"step": 1040
},
{
"epoch": 0.5394481150408084,
"grad_norm": 0.2196187973022461,
"learning_rate": 1.8406547384296377e-05,
"loss": 0.0172,
"step": 1041
},
{
"epoch": 0.539966316880425,
"grad_norm": 0.17992262542247772,
"learning_rate": 1.8373074437399744e-05,
"loss": 0.0115,
"step": 1042
},
{
"epoch": 0.5404845187200414,
"grad_norm": 0.19862842559814453,
"learning_rate": 1.8339606077408165e-05,
"loss": 0.0238,
"step": 1043
},
{
"epoch": 0.541002720559658,
"grad_norm": 0.30180686712265015,
"learning_rate": 1.8306142398681338e-05,
"loss": 0.0115,
"step": 1044
},
{
"epoch": 0.5415209223992745,
"grad_norm": 0.24334149062633514,
"learning_rate": 1.827268349556575e-05,
"loss": 0.0172,
"step": 1045
},
{
"epoch": 0.542039124238891,
"grad_norm": 0.17345716059207916,
"learning_rate": 1.8239229462394435e-05,
"loss": 0.0109,
"step": 1046
},
{
"epoch": 0.5425573260785076,
"grad_norm": 0.25446006655693054,
"learning_rate": 1.8205780393486675e-05,
"loss": 0.0113,
"step": 1047
},
{
"epoch": 0.5430755279181241,
"grad_norm": 0.319873571395874,
"learning_rate": 1.817233638314778e-05,
"loss": 0.0232,
"step": 1048
},
{
"epoch": 0.5435937297577407,
"grad_norm": 0.42063799500465393,
"learning_rate": 1.813889752566878e-05,
"loss": 0.026,
"step": 1049
},
{
"epoch": 0.5441119315973572,
"grad_norm": 0.33134227991104126,
"learning_rate": 1.8105463915326205e-05,
"loss": 0.0196,
"step": 1050
},
{
"epoch": 0.5446301334369738,
"grad_norm": 0.26480934023857117,
"learning_rate": 1.807203564638175e-05,
"loss": 0.0267,
"step": 1051
},
{
"epoch": 0.5451483352765902,
"grad_norm": 0.18234863877296448,
"learning_rate": 1.8038612813082084e-05,
"loss": 0.0137,
"step": 1052
},
{
"epoch": 0.5456665371162067,
"grad_norm": 0.15523044764995575,
"learning_rate": 1.8005195509658536e-05,
"loss": 0.0111,
"step": 1053
},
{
"epoch": 0.5461847389558233,
"grad_norm": 0.3344533145427704,
"learning_rate": 1.797178383032685e-05,
"loss": 0.0168,
"step": 1054
},
{
"epoch": 0.5467029407954398,
"grad_norm": 0.12319575250148773,
"learning_rate": 1.793837786928691e-05,
"loss": 0.0053,
"step": 1055
},
{
"epoch": 0.5472211426350564,
"grad_norm": 0.13881735503673553,
"learning_rate": 1.790497772072247e-05,
"loss": 0.0114,
"step": 1056
},
{
"epoch": 0.5477393444746729,
"grad_norm": 0.19909988343715668,
"learning_rate": 1.7871583478800915e-05,
"loss": 0.0138,
"step": 1057
},
{
"epoch": 0.5482575463142895,
"grad_norm": 0.08285260200500488,
"learning_rate": 1.7838195237672974e-05,
"loss": 0.0034,
"step": 1058
},
{
"epoch": 0.548775748153906,
"grad_norm": 0.5260575413703918,
"learning_rate": 1.7804813091472435e-05,
"loss": 0.0355,
"step": 1059
},
{
"epoch": 0.5492939499935224,
"grad_norm": 0.13995790481567383,
"learning_rate": 1.7771437134315927e-05,
"loss": 0.0102,
"step": 1060
},
{
"epoch": 0.549812151833139,
"grad_norm": 0.3183762729167938,
"learning_rate": 1.7738067460302625e-05,
"loss": 0.0168,
"step": 1061
},
{
"epoch": 0.5503303536727555,
"grad_norm": 0.19481916725635529,
"learning_rate": 1.770470416351398e-05,
"loss": 0.0129,
"step": 1062
},
{
"epoch": 0.5508485555123721,
"grad_norm": 0.18306854367256165,
"learning_rate": 1.7671347338013468e-05,
"loss": 0.0224,
"step": 1063
},
{
"epoch": 0.5513667573519886,
"grad_norm": 0.22059708833694458,
"learning_rate": 1.763799707784632e-05,
"loss": 0.0117,
"step": 1064
},
{
"epoch": 0.5518849591916052,
"grad_norm": 0.4268891513347626,
"learning_rate": 1.7604653477039258e-05,
"loss": 0.0138,
"step": 1065
},
{
"epoch": 0.5524031610312217,
"grad_norm": 0.10974422097206116,
"learning_rate": 1.7571316629600243e-05,
"loss": 0.0057,
"step": 1066
},
{
"epoch": 0.5529213628708382,
"grad_norm": 0.09162997454404831,
"learning_rate": 1.753798662951816e-05,
"loss": 0.0061,
"step": 1067
},
{
"epoch": 0.5534395647104547,
"grad_norm": 0.12637154757976532,
"learning_rate": 1.7504663570762615e-05,
"loss": 0.0085,
"step": 1068
},
{
"epoch": 0.5539577665500712,
"grad_norm": 0.1373872607946396,
"learning_rate": 1.747134754728365e-05,
"loss": 0.0097,
"step": 1069
},
{
"epoch": 0.5544759683896878,
"grad_norm": 0.23044104874134064,
"learning_rate": 1.7438038653011457e-05,
"loss": 0.0148,
"step": 1070
},
{
"epoch": 0.5549941702293043,
"grad_norm": 0.22062622010707855,
"learning_rate": 1.740473698185612e-05,
"loss": 0.0199,
"step": 1071
},
{
"epoch": 0.5555123720689208,
"grad_norm": 0.32974135875701904,
"learning_rate": 1.7371442627707385e-05,
"loss": 0.0099,
"step": 1072
},
{
"epoch": 0.5560305739085374,
"grad_norm": 0.18324632942676544,
"learning_rate": 1.7338155684434344e-05,
"loss": 0.0103,
"step": 1073
},
{
"epoch": 0.5565487757481539,
"grad_norm": 0.1241641715168953,
"learning_rate": 1.730487624588522e-05,
"loss": 0.0079,
"step": 1074
},
{
"epoch": 0.5570669775877705,
"grad_norm": 0.22878427803516388,
"learning_rate": 1.727160440588704e-05,
"loss": 0.0148,
"step": 1075
},
{
"epoch": 0.557585179427387,
"grad_norm": 0.549785315990448,
"learning_rate": 1.723834025824544e-05,
"loss": 0.0457,
"step": 1076
},
{
"epoch": 0.5581033812670035,
"grad_norm": 0.5593258142471313,
"learning_rate": 1.7205083896744365e-05,
"loss": 0.0259,
"step": 1077
},
{
"epoch": 0.55862158310662,
"grad_norm": 0.1253921538591385,
"learning_rate": 1.7171835415145806e-05,
"loss": 0.0117,
"step": 1078
},
{
"epoch": 0.5591397849462365,
"grad_norm": 0.15490829944610596,
"learning_rate": 1.713859490718951e-05,
"loss": 0.0103,
"step": 1079
},
{
"epoch": 0.5596579867858531,
"grad_norm": 0.19711671769618988,
"learning_rate": 1.710536246659279e-05,
"loss": 0.0094,
"step": 1080
},
{
"epoch": 0.5601761886254696,
"grad_norm": 0.24218136072158813,
"learning_rate": 1.7072138187050177e-05,
"loss": 0.0223,
"step": 1081
},
{
"epoch": 0.5606943904650862,
"grad_norm": 0.43247753381729126,
"learning_rate": 1.7038922162233224e-05,
"loss": 0.0437,
"step": 1082
},
{
"epoch": 0.5612125923047027,
"grad_norm": 0.3048844039440155,
"learning_rate": 1.7005714485790167e-05,
"loss": 0.0164,
"step": 1083
},
{
"epoch": 0.5617307941443193,
"grad_norm": 0.269682377576828,
"learning_rate": 1.6972515251345746e-05,
"loss": 0.0153,
"step": 1084
},
{
"epoch": 0.5622489959839357,
"grad_norm": 0.22269891202449799,
"learning_rate": 1.6939324552500887e-05,
"loss": 0.0158,
"step": 1085
},
{
"epoch": 0.5627671978235522,
"grad_norm": 0.11371760815382004,
"learning_rate": 1.690614248283245e-05,
"loss": 0.0072,
"step": 1086
},
{
"epoch": 0.5632853996631688,
"grad_norm": 0.2383551448583603,
"learning_rate": 1.6872969135892963e-05,
"loss": 0.0207,
"step": 1087
},
{
"epoch": 0.5638036015027853,
"grad_norm": 0.27557238936424255,
"learning_rate": 1.683980460521036e-05,
"loss": 0.0137,
"step": 1088
},
{
"epoch": 0.5643218033424019,
"grad_norm": 0.16742661595344543,
"learning_rate": 1.680664898428773e-05,
"loss": 0.0085,
"step": 1089
},
{
"epoch": 0.5648400051820184,
"grad_norm": 0.17626698315143585,
"learning_rate": 1.6773502366603032e-05,
"loss": 0.0079,
"step": 1090
},
{
"epoch": 0.565358207021635,
"grad_norm": 0.27058324217796326,
"learning_rate": 1.6740364845608842e-05,
"loss": 0.0169,
"step": 1091
},
{
"epoch": 0.5658764088612515,
"grad_norm": 0.2151581048965454,
"learning_rate": 1.670723651473209e-05,
"loss": 0.0185,
"step": 1092
},
{
"epoch": 0.5663946107008679,
"grad_norm": 0.20677979290485382,
"learning_rate": 1.6674117467373797e-05,
"loss": 0.0114,
"step": 1093
},
{
"epoch": 0.5669128125404845,
"grad_norm": 0.23222865164279938,
"learning_rate": 1.6641007796908822e-05,
"loss": 0.0212,
"step": 1094
},
{
"epoch": 0.567431014380101,
"grad_norm": 0.4040347635746002,
"learning_rate": 1.6607907596685556e-05,
"loss": 0.0174,
"step": 1095
},
{
"epoch": 0.5679492162197176,
"grad_norm": 0.15651348233222961,
"learning_rate": 1.6574816960025722e-05,
"loss": 0.0096,
"step": 1096
},
{
"epoch": 0.5684674180593341,
"grad_norm": 0.2294890284538269,
"learning_rate": 1.654173598022407e-05,
"loss": 0.0184,
"step": 1097
},
{
"epoch": 0.5689856198989507,
"grad_norm": 0.23169976472854614,
"learning_rate": 1.6508664750548114e-05,
"loss": 0.0101,
"step": 1098
},
{
"epoch": 0.5695038217385672,
"grad_norm": 0.16020269691944122,
"learning_rate": 1.6475603364237886e-05,
"loss": 0.0108,
"step": 1099
},
{
"epoch": 0.5700220235781837,
"grad_norm": 0.27258041501045227,
"learning_rate": 1.644255191450566e-05,
"loss": 0.0172,
"step": 1100
},
{
"epoch": 0.5705402254178003,
"grad_norm": 0.25970420241355896,
"learning_rate": 1.640951049453571e-05,
"loss": 0.015,
"step": 1101
},
{
"epoch": 0.5710584272574167,
"grad_norm": 0.15470710396766663,
"learning_rate": 1.6376479197484028e-05,
"loss": 0.0108,
"step": 1102
},
{
"epoch": 0.5715766290970333,
"grad_norm": 0.15609830617904663,
"learning_rate": 1.634345811647804e-05,
"loss": 0.0179,
"step": 1103
},
{
"epoch": 0.5720948309366498,
"grad_norm": 0.24201931059360504,
"learning_rate": 1.6310447344616406e-05,
"loss": 0.0152,
"step": 1104
},
{
"epoch": 0.5726130327762664,
"grad_norm": 0.2512819468975067,
"learning_rate": 1.62774469749687e-05,
"loss": 0.018,
"step": 1105
},
{
"epoch": 0.5731312346158829,
"grad_norm": 0.13237565755844116,
"learning_rate": 1.6244457100575184e-05,
"loss": 0.0102,
"step": 1106
},
{
"epoch": 0.5736494364554994,
"grad_norm": 0.17929814755916595,
"learning_rate": 1.6211477814446498e-05,
"loss": 0.0147,
"step": 1107
},
{
"epoch": 0.574167638295116,
"grad_norm": 0.26788702607154846,
"learning_rate": 1.6178509209563465e-05,
"loss": 0.0326,
"step": 1108
},
{
"epoch": 0.5746858401347325,
"grad_norm": 0.3604764938354492,
"learning_rate": 1.6145551378876773e-05,
"loss": 0.0134,
"step": 1109
},
{
"epoch": 0.575204041974349,
"grad_norm": 0.1976727545261383,
"learning_rate": 1.611260441530675e-05,
"loss": 0.0138,
"step": 1110
},
{
"epoch": 0.5757222438139655,
"grad_norm": 0.24070017039775848,
"learning_rate": 1.6079668411743058e-05,
"loss": 0.0204,
"step": 1111
},
{
"epoch": 0.5762404456535821,
"grad_norm": 0.3317483365535736,
"learning_rate": 1.604674346104449e-05,
"loss": 0.0183,
"step": 1112
},
{
"epoch": 0.5767586474931986,
"grad_norm": 0.18072566390037537,
"learning_rate": 1.6013829656038654e-05,
"loss": 0.0181,
"step": 1113
},
{
"epoch": 0.5772768493328151,
"grad_norm": 0.21476292610168457,
"learning_rate": 1.5980927089521762e-05,
"loss": 0.0124,
"step": 1114
},
{
"epoch": 0.5777950511724317,
"grad_norm": 0.1312938779592514,
"learning_rate": 1.5948035854258294e-05,
"loss": 0.0071,
"step": 1115
},
{
"epoch": 0.5783132530120482,
"grad_norm": 0.08147620409727097,
"learning_rate": 1.5915156042980833e-05,
"loss": 0.0028,
"step": 1116
},
{
"epoch": 0.5788314548516648,
"grad_norm": 0.06438157707452774,
"learning_rate": 1.5882287748389716e-05,
"loss": 0.0031,
"step": 1117
},
{
"epoch": 0.5793496566912812,
"grad_norm": 0.23561817407608032,
"learning_rate": 1.584943106315284e-05,
"loss": 0.0308,
"step": 1118
},
{
"epoch": 0.5798678585308978,
"grad_norm": 0.13540908694267273,
"learning_rate": 1.5816586079905346e-05,
"loss": 0.0135,
"step": 1119
},
{
"epoch": 0.5803860603705143,
"grad_norm": 0.10627519339323044,
"learning_rate": 1.578375289124939e-05,
"loss": 0.0075,
"step": 1120
},
{
"epoch": 0.5809042622101308,
"grad_norm": 0.15574534237384796,
"learning_rate": 1.575093158975389e-05,
"loss": 0.0073,
"step": 1121
},
{
"epoch": 0.5814224640497474,
"grad_norm": 0.22351405024528503,
"learning_rate": 1.5718122267954232e-05,
"loss": 0.0161,
"step": 1122
},
{
"epoch": 0.5819406658893639,
"grad_norm": 0.3184528648853302,
"learning_rate": 1.5685325018352028e-05,
"loss": 0.0326,
"step": 1123
},
{
"epoch": 0.5824588677289805,
"grad_norm": 0.1077670082449913,
"learning_rate": 1.5652539933414858e-05,
"loss": 0.0037,
"step": 1124
},
{
"epoch": 0.582977069568597,
"grad_norm": 0.16742312908172607,
"learning_rate": 1.561976710557602e-05,
"loss": 0.0118,
"step": 1125
},
{
"epoch": 0.5834952714082134,
"grad_norm": 0.17053693532943726,
"learning_rate": 1.5587006627234225e-05,
"loss": 0.0151,
"step": 1126
},
{
"epoch": 0.58401347324783,
"grad_norm": 0.26882749795913696,
"learning_rate": 1.5554258590753388e-05,
"loss": 0.0153,
"step": 1127
},
{
"epoch": 0.5845316750874465,
"grad_norm": 0.17190419137477875,
"learning_rate": 1.5521523088462332e-05,
"loss": 0.0104,
"step": 1128
},
{
"epoch": 0.5850498769270631,
"grad_norm": 0.2322155237197876,
"learning_rate": 1.5488800212654557e-05,
"loss": 0.0248,
"step": 1129
},
{
"epoch": 0.5855680787666796,
"grad_norm": 0.3169136047363281,
"learning_rate": 1.545609005558796e-05,
"loss": 0.0321,
"step": 1130
},
{
"epoch": 0.5860862806062962,
"grad_norm": 0.35867586731910706,
"learning_rate": 1.5423392709484565e-05,
"loss": 0.0089,
"step": 1131
},
{
"epoch": 0.5866044824459127,
"grad_norm": 0.17669306695461273,
"learning_rate": 1.5390708266530288e-05,
"loss": 0.0166,
"step": 1132
},
{
"epoch": 0.5871226842855292,
"grad_norm": 0.3532368540763855,
"learning_rate": 1.5358036818874668e-05,
"loss": 0.0289,
"step": 1133
},
{
"epoch": 0.5876408861251458,
"grad_norm": 0.16300423443317413,
"learning_rate": 1.5325378458630615e-05,
"loss": 0.0084,
"step": 1134
},
{
"epoch": 0.5881590879647622,
"grad_norm": 0.29001057147979736,
"learning_rate": 1.5292733277874116e-05,
"loss": 0.0244,
"step": 1135
},
{
"epoch": 0.5886772898043788,
"grad_norm": 0.1256478726863861,
"learning_rate": 1.5260101368644008e-05,
"loss": 0.0075,
"step": 1136
},
{
"epoch": 0.5891954916439953,
"grad_norm": 0.19914460182189941,
"learning_rate": 1.5227482822941725e-05,
"loss": 0.0143,
"step": 1137
},
{
"epoch": 0.5897136934836119,
"grad_norm": 0.2361249178647995,
"learning_rate": 1.519487773273102e-05,
"loss": 0.0169,
"step": 1138
},
{
"epoch": 0.5902318953232284,
"grad_norm": 0.19691789150238037,
"learning_rate": 1.516228618993769e-05,
"loss": 0.0108,
"step": 1139
},
{
"epoch": 0.5907500971628449,
"grad_norm": 0.3483859896659851,
"learning_rate": 1.5129708286449359e-05,
"loss": 0.0218,
"step": 1140
},
{
"epoch": 0.5912682990024615,
"grad_norm": 0.3023143708705902,
"learning_rate": 1.509714411411519e-05,
"loss": 0.0143,
"step": 1141
},
{
"epoch": 0.591786500842078,
"grad_norm": 0.17124785482883453,
"learning_rate": 1.506459376474564e-05,
"loss": 0.0095,
"step": 1142
},
{
"epoch": 0.5923047026816946,
"grad_norm": 0.19885291159152985,
"learning_rate": 1.5032057330112174e-05,
"loss": 0.0091,
"step": 1143
},
{
"epoch": 0.592822904521311,
"grad_norm": 0.6884151697158813,
"learning_rate": 1.4999534901947046e-05,
"loss": 0.0288,
"step": 1144
},
{
"epoch": 0.5933411063609276,
"grad_norm": 0.3492492139339447,
"learning_rate": 1.4967026571943004e-05,
"loss": 0.0219,
"step": 1145
},
{
"epoch": 0.5938593082005441,
"grad_norm": 0.2615423798561096,
"learning_rate": 1.4934532431753073e-05,
"loss": 0.021,
"step": 1146
},
{
"epoch": 0.5943775100401606,
"grad_norm": 0.23267151415348053,
"learning_rate": 1.4902052572990236e-05,
"loss": 0.0166,
"step": 1147
},
{
"epoch": 0.5948957118797772,
"grad_norm": 0.20014171302318573,
"learning_rate": 1.4869587087227241e-05,
"loss": 0.0098,
"step": 1148
},
{
"epoch": 0.5954139137193937,
"grad_norm": 0.13207559287548065,
"learning_rate": 1.4837136065996307e-05,
"loss": 0.0077,
"step": 1149
},
{
"epoch": 0.5959321155590103,
"grad_norm": 0.36822882294654846,
"learning_rate": 1.4804699600788872e-05,
"loss": 0.0262,
"step": 1150
},
{
"epoch": 0.5964503173986268,
"grad_norm": 0.1576855331659317,
"learning_rate": 1.4772277783055318e-05,
"loss": 0.0077,
"step": 1151
},
{
"epoch": 0.5969685192382433,
"grad_norm": 0.056045811623334885,
"learning_rate": 1.4739870704204746e-05,
"loss": 0.0034,
"step": 1152
},
{
"epoch": 0.5974867210778598,
"grad_norm": 0.164437398314476,
"learning_rate": 1.4707478455604713e-05,
"loss": 0.0047,
"step": 1153
},
{
"epoch": 0.5980049229174763,
"grad_norm": 0.1738058626651764,
"learning_rate": 1.4675101128580944e-05,
"loss": 0.0106,
"step": 1154
},
{
"epoch": 0.5985231247570929,
"grad_norm": 0.27060168981552124,
"learning_rate": 1.4642738814417104e-05,
"loss": 0.0233,
"step": 1155
},
{
"epoch": 0.5990413265967094,
"grad_norm": 0.3752515912055969,
"learning_rate": 1.4610391604354522e-05,
"loss": 0.0152,
"step": 1156
},
{
"epoch": 0.599559528436326,
"grad_norm": 0.32041823863983154,
"learning_rate": 1.4578059589591953e-05,
"loss": 0.0149,
"step": 1157
},
{
"epoch": 0.6000777302759425,
"grad_norm": 0.0778937041759491,
"learning_rate": 1.4545742861285325e-05,
"loss": 0.006,
"step": 1158
},
{
"epoch": 0.6005959321155591,
"grad_norm": 0.22142964601516724,
"learning_rate": 1.4513441510547428e-05,
"loss": 0.0119,
"step": 1159
},
{
"epoch": 0.6011141339551755,
"grad_norm": 0.3265921473503113,
"learning_rate": 1.4481155628447728e-05,
"loss": 0.0286,
"step": 1160
},
{
"epoch": 0.601632335794792,
"grad_norm": 0.22287465631961823,
"learning_rate": 1.4448885306012078e-05,
"loss": 0.0118,
"step": 1161
},
{
"epoch": 0.6021505376344086,
"grad_norm": 0.14569589495658875,
"learning_rate": 1.4416630634222449e-05,
"loss": 0.0106,
"step": 1162
},
{
"epoch": 0.6026687394740251,
"grad_norm": 0.1785486936569214,
"learning_rate": 1.4384391704016695e-05,
"loss": 0.0106,
"step": 1163
},
{
"epoch": 0.6031869413136417,
"grad_norm": 0.22368593513965607,
"learning_rate": 1.4352168606288285e-05,
"loss": 0.0145,
"step": 1164
},
{
"epoch": 0.6037051431532582,
"grad_norm": 0.18533925712108612,
"learning_rate": 1.4319961431886054e-05,
"loss": 0.0178,
"step": 1165
},
{
"epoch": 0.6042233449928748,
"grad_norm": 0.22897221148014069,
"learning_rate": 1.4287770271613952e-05,
"loss": 0.0156,
"step": 1166
},
{
"epoch": 0.6047415468324913,
"grad_norm": 0.28994596004486084,
"learning_rate": 1.4255595216230753e-05,
"loss": 0.0277,
"step": 1167
},
{
"epoch": 0.6052597486721077,
"grad_norm": 0.17252403497695923,
"learning_rate": 1.4223436356449852e-05,
"loss": 0.0094,
"step": 1168
},
{
"epoch": 0.6057779505117243,
"grad_norm": 0.17040708661079407,
"learning_rate": 1.4191293782938967e-05,
"loss": 0.0117,
"step": 1169
},
{
"epoch": 0.6062961523513408,
"grad_norm": 0.47751736640930176,
"learning_rate": 1.4159167586319916e-05,
"loss": 0.0296,
"step": 1170
},
{
"epoch": 0.6068143541909574,
"grad_norm": 0.19108998775482178,
"learning_rate": 1.4127057857168316e-05,
"loss": 0.0138,
"step": 1171
},
{
"epoch": 0.6073325560305739,
"grad_norm": 0.235890194773674,
"learning_rate": 1.4094964686013381e-05,
"loss": 0.0107,
"step": 1172
},
{
"epoch": 0.6078507578701905,
"grad_norm": 0.12942662835121155,
"learning_rate": 1.4062888163337633e-05,
"loss": 0.0059,
"step": 1173
},
{
"epoch": 0.608368959709807,
"grad_norm": 0.4324270486831665,
"learning_rate": 1.4030828379576664e-05,
"loss": 0.0296,
"step": 1174
},
{
"epoch": 0.6088871615494235,
"grad_norm": 0.11905232071876526,
"learning_rate": 1.3998785425118851e-05,
"loss": 0.012,
"step": 1175
},
{
"epoch": 0.6094053633890401,
"grad_norm": 0.20753410458564758,
"learning_rate": 1.3966759390305143e-05,
"loss": 0.0133,
"step": 1176
},
{
"epoch": 0.6099235652286565,
"grad_norm": 0.2876964211463928,
"learning_rate": 1.3934750365428783e-05,
"loss": 0.0315,
"step": 1177
},
{
"epoch": 0.6104417670682731,
"grad_norm": 0.20571279525756836,
"learning_rate": 1.3902758440735058e-05,
"loss": 0.0189,
"step": 1178
},
{
"epoch": 0.6109599689078896,
"grad_norm": 0.16015420854091644,
"learning_rate": 1.3870783706421032e-05,
"loss": 0.0191,
"step": 1179
},
{
"epoch": 0.6114781707475061,
"grad_norm": 0.23814105987548828,
"learning_rate": 1.3838826252635313e-05,
"loss": 0.033,
"step": 1180
},
{
"epoch": 0.6119963725871227,
"grad_norm": 0.2253902405500412,
"learning_rate": 1.3806886169477784e-05,
"loss": 0.0109,
"step": 1181
},
{
"epoch": 0.6125145744267392,
"grad_norm": 0.17431442439556122,
"learning_rate": 1.3774963546999364e-05,
"loss": 0.0113,
"step": 1182
},
{
"epoch": 0.6130327762663558,
"grad_norm": 0.24366681277751923,
"learning_rate": 1.3743058475201727e-05,
"loss": 0.019,
"step": 1183
},
{
"epoch": 0.6135509781059723,
"grad_norm": 0.20095647871494293,
"learning_rate": 1.3711171044037074e-05,
"loss": 0.0145,
"step": 1184
},
{
"epoch": 0.6140691799455888,
"grad_norm": 0.37862977385520935,
"learning_rate": 1.3679301343407874e-05,
"loss": 0.0226,
"step": 1185
},
{
"epoch": 0.6145873817852053,
"grad_norm": 0.23509663343429565,
"learning_rate": 1.3647449463166616e-05,
"loss": 0.0125,
"step": 1186
},
{
"epoch": 0.6151055836248218,
"grad_norm": 0.1528037041425705,
"learning_rate": 1.3615615493115509e-05,
"loss": 0.0136,
"step": 1187
},
{
"epoch": 0.6156237854644384,
"grad_norm": 0.24438060820102692,
"learning_rate": 1.3583799523006307e-05,
"loss": 0.0134,
"step": 1188
},
{
"epoch": 0.6161419873040549,
"grad_norm": 0.1713854968547821,
"learning_rate": 1.3552001642540003e-05,
"loss": 0.0085,
"step": 1189
},
{
"epoch": 0.6166601891436715,
"grad_norm": 0.4027296006679535,
"learning_rate": 1.3520221941366579e-05,
"loss": 0.0288,
"step": 1190
},
{
"epoch": 0.617178390983288,
"grad_norm": 0.25324851274490356,
"learning_rate": 1.3488460509084774e-05,
"loss": 0.0355,
"step": 1191
},
{
"epoch": 0.6176965928229046,
"grad_norm": 0.15679526329040527,
"learning_rate": 1.3456717435241808e-05,
"loss": 0.0141,
"step": 1192
},
{
"epoch": 0.618214794662521,
"grad_norm": 0.3524555563926697,
"learning_rate": 1.342499280933316e-05,
"loss": 0.0229,
"step": 1193
},
{
"epoch": 0.6187329965021375,
"grad_norm": 0.12044969946146011,
"learning_rate": 1.339328672080229e-05,
"loss": 0.01,
"step": 1194
},
{
"epoch": 0.6192511983417541,
"grad_norm": 0.20489737391471863,
"learning_rate": 1.3361599259040375e-05,
"loss": 0.0128,
"step": 1195
},
{
"epoch": 0.6197694001813706,
"grad_norm": 0.30765610933303833,
"learning_rate": 1.3329930513386105e-05,
"loss": 0.0166,
"step": 1196
},
{
"epoch": 0.6202876020209872,
"grad_norm": 0.15374019742012024,
"learning_rate": 1.3298280573125392e-05,
"loss": 0.0099,
"step": 1197
},
{
"epoch": 0.6208058038606037,
"grad_norm": 0.12795308232307434,
"learning_rate": 1.3266649527491127e-05,
"loss": 0.0037,
"step": 1198
},
{
"epoch": 0.6213240057002203,
"grad_norm": 0.0716242715716362,
"learning_rate": 1.3235037465662921e-05,
"loss": 0.0051,
"step": 1199
},
{
"epoch": 0.6218422075398368,
"grad_norm": 0.18337176740169525,
"learning_rate": 1.3203444476766874e-05,
"loss": 0.0113,
"step": 1200
},
{
"epoch": 0.6223604093794532,
"grad_norm": 0.25055044889450073,
"learning_rate": 1.3171870649875313e-05,
"loss": 0.0111,
"step": 1201
},
{
"epoch": 0.6228786112190698,
"grad_norm": 0.28508034348487854,
"learning_rate": 1.3140316074006545e-05,
"loss": 0.0113,
"step": 1202
},
{
"epoch": 0.6233968130586863,
"grad_norm": 0.1235652044415474,
"learning_rate": 1.3108780838124577e-05,
"loss": 0.0101,
"step": 1203
},
{
"epoch": 0.6239150148983029,
"grad_norm": 0.18104441463947296,
"learning_rate": 1.3077265031138913e-05,
"loss": 0.0068,
"step": 1204
},
{
"epoch": 0.6244332167379194,
"grad_norm": 0.1736707240343094,
"learning_rate": 1.3045768741904272e-05,
"loss": 0.0255,
"step": 1205
},
{
"epoch": 0.624951418577536,
"grad_norm": 0.31340596079826355,
"learning_rate": 1.3014292059220357e-05,
"loss": 0.0154,
"step": 1206
},
{
"epoch": 0.6254696204171525,
"grad_norm": 0.3333686292171478,
"learning_rate": 1.2982835071831565e-05,
"loss": 0.0221,
"step": 1207
},
{
"epoch": 0.625987822256769,
"grad_norm": 0.09395372867584229,
"learning_rate": 1.295139786842679e-05,
"loss": 0.0028,
"step": 1208
},
{
"epoch": 0.6265060240963856,
"grad_norm": 0.19452796876430511,
"learning_rate": 1.2919980537639134e-05,
"loss": 0.0173,
"step": 1209
},
{
"epoch": 0.627024225936002,
"grad_norm": 0.16185300052165985,
"learning_rate": 1.2888583168045688e-05,
"loss": 0.0104,
"step": 1210
},
{
"epoch": 0.6275424277756186,
"grad_norm": 0.3420233726501465,
"learning_rate": 1.2857205848167231e-05,
"loss": 0.0224,
"step": 1211
},
{
"epoch": 0.6280606296152351,
"grad_norm": 0.18805178999900818,
"learning_rate": 1.2825848666468052e-05,
"loss": 0.0125,
"step": 1212
},
{
"epoch": 0.6285788314548517,
"grad_norm": 0.32541659474372864,
"learning_rate": 1.2794511711355644e-05,
"loss": 0.0225,
"step": 1213
},
{
"epoch": 0.6290970332944682,
"grad_norm": 0.6056791543960571,
"learning_rate": 1.2763195071180483e-05,
"loss": 0.029,
"step": 1214
},
{
"epoch": 0.6296152351340847,
"grad_norm": 0.06527632474899292,
"learning_rate": 1.2731898834235752e-05,
"loss": 0.0057,
"step": 1215
},
{
"epoch": 0.6301334369737013,
"grad_norm": 0.1526113748550415,
"learning_rate": 1.270062308875713e-05,
"loss": 0.0073,
"step": 1216
},
{
"epoch": 0.6306516388133178,
"grad_norm": 0.14983856678009033,
"learning_rate": 1.2669367922922504e-05,
"loss": 0.0051,
"step": 1217
},
{
"epoch": 0.6311698406529344,
"grad_norm": 0.17253729701042175,
"learning_rate": 1.2638133424851766e-05,
"loss": 0.0093,
"step": 1218
},
{
"epoch": 0.6316880424925508,
"grad_norm": 0.42797988653182983,
"learning_rate": 1.2606919682606505e-05,
"loss": 0.0314,
"step": 1219
},
{
"epoch": 0.6322062443321674,
"grad_norm": 0.22709913551807404,
"learning_rate": 1.2575726784189816e-05,
"loss": 0.0295,
"step": 1220
},
{
"epoch": 0.6327244461717839,
"grad_norm": 0.16602033376693726,
"learning_rate": 1.2544554817546017e-05,
"loss": 0.0139,
"step": 1221
},
{
"epoch": 0.6332426480114004,
"grad_norm": 0.31228166818618774,
"learning_rate": 1.2513403870560429e-05,
"loss": 0.0191,
"step": 1222
},
{
"epoch": 0.633760849851017,
"grad_norm": 0.1140090823173523,
"learning_rate": 1.2482274031059073e-05,
"loss": 0.0102,
"step": 1223
},
{
"epoch": 0.6342790516906335,
"grad_norm": 0.37553247809410095,
"learning_rate": 1.2451165386808496e-05,
"loss": 0.0115,
"step": 1224
},
{
"epoch": 0.6347972535302501,
"grad_norm": 0.16082100570201874,
"learning_rate": 1.2420078025515476e-05,
"loss": 0.0138,
"step": 1225
},
{
"epoch": 0.6353154553698666,
"grad_norm": 0.1942625790834427,
"learning_rate": 1.2389012034826787e-05,
"loss": 0.0114,
"step": 1226
},
{
"epoch": 0.635833657209483,
"grad_norm": 0.0908832773566246,
"learning_rate": 1.2357967502328943e-05,
"loss": 0.004,
"step": 1227
},
{
"epoch": 0.6363518590490996,
"grad_norm": 0.32251858711242676,
"learning_rate": 1.2326944515547963e-05,
"loss": 0.0133,
"step": 1228
},
{
"epoch": 0.6368700608887161,
"grad_norm": 0.42536473274230957,
"learning_rate": 1.2295943161949135e-05,
"loss": 0.0407,
"step": 1229
},
{
"epoch": 0.6373882627283327,
"grad_norm": 0.2966007888317108,
"learning_rate": 1.2264963528936747e-05,
"loss": 0.0168,
"step": 1230
},
{
"epoch": 0.6379064645679492,
"grad_norm": 0.30448687076568604,
"learning_rate": 1.2234005703853829e-05,
"loss": 0.0304,
"step": 1231
},
{
"epoch": 0.6384246664075658,
"grad_norm": 0.22583232820034027,
"learning_rate": 1.2203069773981953e-05,
"loss": 0.0199,
"step": 1232
},
{
"epoch": 0.6389428682471823,
"grad_norm": 0.21759001910686493,
"learning_rate": 1.2172155826540952e-05,
"loss": 0.0123,
"step": 1233
},
{
"epoch": 0.6394610700867988,
"grad_norm": 0.2320861667394638,
"learning_rate": 1.2141263948688687e-05,
"loss": 0.0127,
"step": 1234
},
{
"epoch": 0.6399792719264153,
"grad_norm": 0.3261067271232605,
"learning_rate": 1.211039422752077e-05,
"loss": 0.0145,
"step": 1235
},
{
"epoch": 0.6404974737660318,
"grad_norm": 0.15837307274341583,
"learning_rate": 1.2079546750070385e-05,
"loss": 0.0073,
"step": 1236
},
{
"epoch": 0.6410156756056484,
"grad_norm": 0.13467198610305786,
"learning_rate": 1.2048721603307971e-05,
"loss": 0.0073,
"step": 1237
},
{
"epoch": 0.6415338774452649,
"grad_norm": 0.10030899196863174,
"learning_rate": 1.2017918874141037e-05,
"loss": 0.0066,
"step": 1238
},
{
"epoch": 0.6420520792848815,
"grad_norm": 0.1476227343082428,
"learning_rate": 1.198713864941385e-05,
"loss": 0.0082,
"step": 1239
},
{
"epoch": 0.642570281124498,
"grad_norm": 0.15087570250034332,
"learning_rate": 1.1956381015907266e-05,
"loss": 0.0115,
"step": 1240
},
{
"epoch": 0.6430884829641145,
"grad_norm": 0.271362841129303,
"learning_rate": 1.1925646060338427e-05,
"loss": 0.0212,
"step": 1241
},
{
"epoch": 0.6436066848037311,
"grad_norm": 0.22935399413108826,
"learning_rate": 1.1894933869360555e-05,
"loss": 0.014,
"step": 1242
},
{
"epoch": 0.6441248866433475,
"grad_norm": 0.12292534857988358,
"learning_rate": 1.186424452956266e-05,
"loss": 0.0108,
"step": 1243
},
{
"epoch": 0.6446430884829641,
"grad_norm": 0.0863785669207573,
"learning_rate": 1.1833578127469361e-05,
"loss": 0.0026,
"step": 1244
},
{
"epoch": 0.6451612903225806,
"grad_norm": 0.4094049334526062,
"learning_rate": 1.1802934749540578e-05,
"loss": 0.0201,
"step": 1245
},
{
"epoch": 0.6456794921621972,
"grad_norm": 0.1941591501235962,
"learning_rate": 1.1772314482171346e-05,
"loss": 0.0103,
"step": 1246
},
{
"epoch": 0.6461976940018137,
"grad_norm": 0.05845485255122185,
"learning_rate": 1.1741717411691509e-05,
"loss": 0.0027,
"step": 1247
},
{
"epoch": 0.6467158958414302,
"grad_norm": 0.22471730411052704,
"learning_rate": 1.1711143624365537e-05,
"loss": 0.0181,
"step": 1248
},
{
"epoch": 0.6472340976810468,
"grad_norm": 0.22589747607707977,
"learning_rate": 1.1680593206392248e-05,
"loss": 0.016,
"step": 1249
},
{
"epoch": 0.6477522995206633,
"grad_norm": 0.1910380721092224,
"learning_rate": 1.1650066243904585e-05,
"loss": 0.0026,
"step": 1250
},
{
"epoch": 0.6482705013602799,
"grad_norm": 0.4072948396205902,
"learning_rate": 1.1619562822969323e-05,
"loss": 0.0121,
"step": 1251
},
{
"epoch": 0.6487887031998963,
"grad_norm": 0.1868102103471756,
"learning_rate": 1.1589083029586906e-05,
"loss": 0.0045,
"step": 1252
},
{
"epoch": 0.6493069050395129,
"grad_norm": 0.33840611577033997,
"learning_rate": 1.1558626949691142e-05,
"loss": 0.0317,
"step": 1253
},
{
"epoch": 0.6498251068791294,
"grad_norm": 0.1256544589996338,
"learning_rate": 1.1528194669149006e-05,
"loss": 0.0108,
"step": 1254
},
{
"epoch": 0.6503433087187459,
"grad_norm": 0.1379401683807373,
"learning_rate": 1.1497786273760329e-05,
"loss": 0.0073,
"step": 1255
},
{
"epoch": 0.6508615105583625,
"grad_norm": 0.21062274277210236,
"learning_rate": 1.1467401849257643e-05,
"loss": 0.0141,
"step": 1256
},
{
"epoch": 0.651379712397979,
"grad_norm": 0.20837582647800446,
"learning_rate": 1.1437041481305877e-05,
"loss": 0.0132,
"step": 1257
},
{
"epoch": 0.6518979142375956,
"grad_norm": 0.3153347074985504,
"learning_rate": 1.1406705255502151e-05,
"loss": 0.0095,
"step": 1258
},
{
"epoch": 0.6524161160772121,
"grad_norm": 0.18465188145637512,
"learning_rate": 1.137639325737549e-05,
"loss": 0.0068,
"step": 1259
},
{
"epoch": 0.6529343179168287,
"grad_norm": 0.1789904236793518,
"learning_rate": 1.1346105572386646e-05,
"loss": 0.0091,
"step": 1260
},
{
"epoch": 0.6534525197564451,
"grad_norm": 0.10834602266550064,
"learning_rate": 1.13158422859278e-05,
"loss": 0.0048,
"step": 1261
},
{
"epoch": 0.6539707215960616,
"grad_norm": 0.1637331247329712,
"learning_rate": 1.1285603483322362e-05,
"loss": 0.0077,
"step": 1262
},
{
"epoch": 0.6544889234356782,
"grad_norm": 0.3413477838039398,
"learning_rate": 1.1255389249824696e-05,
"loss": 0.0208,
"step": 1263
},
{
"epoch": 0.6550071252752947,
"grad_norm": 0.1726699322462082,
"learning_rate": 1.1225199670619899e-05,
"loss": 0.0051,
"step": 1264
},
{
"epoch": 0.6555253271149113,
"grad_norm": 0.38212427496910095,
"learning_rate": 1.1195034830823583e-05,
"loss": 0.0242,
"step": 1265
},
{
"epoch": 0.6560435289545278,
"grad_norm": 0.2110847383737564,
"learning_rate": 1.116489481548157e-05,
"loss": 0.0113,
"step": 1266
},
{
"epoch": 0.6565617307941444,
"grad_norm": 0.27684441208839417,
"learning_rate": 1.1134779709569732e-05,
"loss": 0.0153,
"step": 1267
},
{
"epoch": 0.6570799326337609,
"grad_norm": 0.7447668313980103,
"learning_rate": 1.1104689597993677e-05,
"loss": 0.0215,
"step": 1268
},
{
"epoch": 0.6575981344733773,
"grad_norm": 0.281093955039978,
"learning_rate": 1.107462456558857e-05,
"loss": 0.0203,
"step": 1269
},
{
"epoch": 0.6581163363129939,
"grad_norm": 0.39335325360298157,
"learning_rate": 1.1044584697118868e-05,
"loss": 0.0326,
"step": 1270
},
{
"epoch": 0.6586345381526104,
"grad_norm": 0.2588825523853302,
"learning_rate": 1.1014570077278057e-05,
"loss": 0.0261,
"step": 1271
},
{
"epoch": 0.659152739992227,
"grad_norm": 0.19893081486225128,
"learning_rate": 1.0984580790688463e-05,
"loss": 0.0097,
"step": 1272
},
{
"epoch": 0.6596709418318435,
"grad_norm": 0.19710324704647064,
"learning_rate": 1.0954616921900982e-05,
"loss": 0.0202,
"step": 1273
},
{
"epoch": 0.6601891436714601,
"grad_norm": 0.5276082158088684,
"learning_rate": 1.0924678555394852e-05,
"loss": 0.0363,
"step": 1274
},
{
"epoch": 0.6607073455110766,
"grad_norm": 0.11454619467258453,
"learning_rate": 1.0894765775577385e-05,
"loss": 0.0064,
"step": 1275
},
{
"epoch": 0.661225547350693,
"grad_norm": 0.20593585073947906,
"learning_rate": 1.0864878666783789e-05,
"loss": 0.009,
"step": 1276
},
{
"epoch": 0.6617437491903096,
"grad_norm": 0.23478110134601593,
"learning_rate": 1.0835017313276874e-05,
"loss": 0.0177,
"step": 1277
},
{
"epoch": 0.6622619510299261,
"grad_norm": 0.4277084171772003,
"learning_rate": 1.0805181799246856e-05,
"loss": 0.0237,
"step": 1278
},
{
"epoch": 0.6627801528695427,
"grad_norm": 0.3500283658504486,
"learning_rate": 1.0775372208811067e-05,
"loss": 0.0214,
"step": 1279
},
{
"epoch": 0.6632983547091592,
"grad_norm": 0.23808035254478455,
"learning_rate": 1.074558862601378e-05,
"loss": 0.024,
"step": 1280
},
{
"epoch": 0.6638165565487757,
"grad_norm": 0.2519163489341736,
"learning_rate": 1.0715831134825936e-05,
"loss": 0.0121,
"step": 1281
},
{
"epoch": 0.6643347583883923,
"grad_norm": 0.2995316982269287,
"learning_rate": 1.068609981914492e-05,
"loss": 0.0145,
"step": 1282
},
{
"epoch": 0.6648529602280088,
"grad_norm": 0.3232516050338745,
"learning_rate": 1.065639476279429e-05,
"loss": 0.0092,
"step": 1283
},
{
"epoch": 0.6653711620676254,
"grad_norm": 0.3077804744243622,
"learning_rate": 1.06267160495236e-05,
"loss": 0.0111,
"step": 1284
},
{
"epoch": 0.6658893639072418,
"grad_norm": 0.3714086413383484,
"learning_rate": 1.059706376300813e-05,
"loss": 0.0085,
"step": 1285
},
{
"epoch": 0.6664075657468584,
"grad_norm": 0.17870378494262695,
"learning_rate": 1.0567437986848635e-05,
"loss": 0.0105,
"step": 1286
},
{
"epoch": 0.6669257675864749,
"grad_norm": 0.15498405694961548,
"learning_rate": 1.0537838804571145e-05,
"loss": 0.0071,
"step": 1287
},
{
"epoch": 0.6674439694260914,
"grad_norm": 0.2603953778743744,
"learning_rate": 1.05082662996267e-05,
"loss": 0.0155,
"step": 1288
},
{
"epoch": 0.667962171265708,
"grad_norm": 0.2905248999595642,
"learning_rate": 1.0478720555391131e-05,
"loss": 0.0151,
"step": 1289
},
{
"epoch": 0.6684803731053245,
"grad_norm": 0.2210564762353897,
"learning_rate": 1.0449201655164836e-05,
"loss": 0.0097,
"step": 1290
},
{
"epoch": 0.6689985749449411,
"grad_norm": 0.2719863951206207,
"learning_rate": 1.0419709682172495e-05,
"loss": 0.0161,
"step": 1291
},
{
"epoch": 0.6695167767845576,
"grad_norm": 0.2112817019224167,
"learning_rate": 1.0390244719562901e-05,
"loss": 0.0126,
"step": 1292
},
{
"epoch": 0.6700349786241742,
"grad_norm": 0.12838172912597656,
"learning_rate": 1.0360806850408684e-05,
"loss": 0.0069,
"step": 1293
},
{
"epoch": 0.6705531804637906,
"grad_norm": 0.1306982785463333,
"learning_rate": 1.0331396157706098e-05,
"loss": 0.0044,
"step": 1294
},
{
"epoch": 0.6710713823034071,
"grad_norm": 0.25365006923675537,
"learning_rate": 1.0302012724374748e-05,
"loss": 0.0144,
"step": 1295
},
{
"epoch": 0.6715895841430237,
"grad_norm": 0.18223069608211517,
"learning_rate": 1.0272656633257412e-05,
"loss": 0.0105,
"step": 1296
},
{
"epoch": 0.6721077859826402,
"grad_norm": 0.3096601068973541,
"learning_rate": 1.0243327967119772e-05,
"loss": 0.0104,
"step": 1297
},
{
"epoch": 0.6726259878222568,
"grad_norm": 0.10256693512201309,
"learning_rate": 1.0214026808650197e-05,
"loss": 0.0048,
"step": 1298
},
{
"epoch": 0.6731441896618733,
"grad_norm": 0.2325192242860794,
"learning_rate": 1.0184753240459475e-05,
"loss": 0.02,
"step": 1299
},
{
"epoch": 0.6736623915014899,
"grad_norm": 0.1683931052684784,
"learning_rate": 1.0155507345080635e-05,
"loss": 0.0054,
"step": 1300
},
{
"epoch": 0.6741805933411064,
"grad_norm": 0.14104929566383362,
"learning_rate": 1.0126289204968673e-05,
"loss": 0.0085,
"step": 1301
},
{
"epoch": 0.6746987951807228,
"grad_norm": 0.18550588190555573,
"learning_rate": 1.0097098902500346e-05,
"loss": 0.0135,
"step": 1302
},
{
"epoch": 0.6752169970203394,
"grad_norm": 0.1749538630247116,
"learning_rate": 1.0067936519973901e-05,
"loss": 0.0151,
"step": 1303
},
{
"epoch": 0.6757351988599559,
"grad_norm": 0.12418388575315475,
"learning_rate": 1.0038802139608897e-05,
"loss": 0.0039,
"step": 1304
},
{
"epoch": 0.6762534006995725,
"grad_norm": 0.3040088713169098,
"learning_rate": 1.0009695843545914e-05,
"loss": 0.0169,
"step": 1305
},
{
"epoch": 0.676771602539189,
"grad_norm": 0.1392335146665573,
"learning_rate": 9.980617713846382e-06,
"loss": 0.0053,
"step": 1306
},
{
"epoch": 0.6772898043788056,
"grad_norm": 0.19270426034927368,
"learning_rate": 9.951567832492308e-06,
"loss": 0.0063,
"step": 1307
},
{
"epoch": 0.6778080062184221,
"grad_norm": 1.204883098602295,
"learning_rate": 9.922546281386044e-06,
"loss": 0.0185,
"step": 1308
},
{
"epoch": 0.6783262080580386,
"grad_norm": 0.30739647150039673,
"learning_rate": 9.893553142350083e-06,
"loss": 0.0091,
"step": 1309
},
{
"epoch": 0.6788444098976552,
"grad_norm": 0.32659727334976196,
"learning_rate": 9.864588497126812e-06,
"loss": 0.0201,
"step": 1310
},
{
"epoch": 0.6793626117372716,
"grad_norm": 0.4216388165950775,
"learning_rate": 9.83565242737829e-06,
"loss": 0.0188,
"step": 1311
},
{
"epoch": 0.6798808135768882,
"grad_norm": 0.4315127730369568,
"learning_rate": 9.80674501468599e-06,
"loss": 0.0079,
"step": 1312
},
{
"epoch": 0.6803990154165047,
"grad_norm": 0.7377837300300598,
"learning_rate": 9.777866340550606e-06,
"loss": 0.024,
"step": 1313
},
{
"epoch": 0.6809172172561213,
"grad_norm": 0.10035476833581924,
"learning_rate": 9.749016486391805e-06,
"loss": 0.0041,
"step": 1314
},
{
"epoch": 0.6814354190957378,
"grad_norm": 0.18348383903503418,
"learning_rate": 9.720195533548012e-06,
"loss": 0.0143,
"step": 1315
},
{
"epoch": 0.6819536209353543,
"grad_norm": 0.274054616689682,
"learning_rate": 9.691403563276132e-06,
"loss": 0.0084,
"step": 1316
},
{
"epoch": 0.6824718227749709,
"grad_norm": 0.3879346251487732,
"learning_rate": 9.662640656751396e-06,
"loss": 0.0154,
"step": 1317
},
{
"epoch": 0.6829900246145874,
"grad_norm": 0.6137412190437317,
"learning_rate": 9.633906895067075e-06,
"loss": 0.0083,
"step": 1318
},
{
"epoch": 0.683508226454204,
"grad_norm": 0.2440262883901596,
"learning_rate": 9.605202359234284e-06,
"loss": 0.0166,
"step": 1319
},
{
"epoch": 0.6840264282938204,
"grad_norm": 0.1609223335981369,
"learning_rate": 9.576527130181713e-06,
"loss": 0.0113,
"step": 1320
},
{
"epoch": 0.684544630133437,
"grad_norm": 0.37860408425331116,
"learning_rate": 9.547881288755455e-06,
"loss": 0.0137,
"step": 1321
},
{
"epoch": 0.6850628319730535,
"grad_norm": 0.0857042595744133,
"learning_rate": 9.519264915718717e-06,
"loss": 0.0083,
"step": 1322
},
{
"epoch": 0.68558103381267,
"grad_norm": 0.24749597907066345,
"learning_rate": 9.490678091751665e-06,
"loss": 0.0295,
"step": 1323
},
{
"epoch": 0.6860992356522866,
"grad_norm": 0.22808784246444702,
"learning_rate": 9.462120897451108e-06,
"loss": 0.0137,
"step": 1324
},
{
"epoch": 0.6866174374919031,
"grad_norm": 0.46637922525405884,
"learning_rate": 9.433593413330355e-06,
"loss": 0.0244,
"step": 1325
},
{
"epoch": 0.6871356393315197,
"grad_norm": 0.349376916885376,
"learning_rate": 9.405095719818938e-06,
"loss": 0.0121,
"step": 1326
},
{
"epoch": 0.6876538411711361,
"grad_norm": 0.5376272797584534,
"learning_rate": 9.3766278972624e-06,
"loss": 0.0246,
"step": 1327
},
{
"epoch": 0.6881720430107527,
"grad_norm": 0.47698912024497986,
"learning_rate": 9.348190025922055e-06,
"loss": 0.0269,
"step": 1328
},
{
"epoch": 0.6886902448503692,
"grad_norm": 0.24001258611679077,
"learning_rate": 9.31978218597479e-06,
"loss": 0.008,
"step": 1329
},
{
"epoch": 0.6892084466899857,
"grad_norm": 0.14646369218826294,
"learning_rate": 9.291404457512813e-06,
"loss": 0.0119,
"step": 1330
},
{
"epoch": 0.6897266485296023,
"grad_norm": 0.5990729331970215,
"learning_rate": 9.26305692054345e-06,
"loss": 0.0223,
"step": 1331
},
{
"epoch": 0.6902448503692188,
"grad_norm": 1.1851770877838135,
"learning_rate": 9.234739654988879e-06,
"loss": 0.0298,
"step": 1332
},
{
"epoch": 0.6907630522088354,
"grad_norm": 0.5354000329971313,
"learning_rate": 9.206452740685954e-06,
"loss": 0.0333,
"step": 1333
},
{
"epoch": 0.6912812540484519,
"grad_norm": 0.6018664240837097,
"learning_rate": 9.178196257385952e-06,
"loss": 0.0234,
"step": 1334
},
{
"epoch": 0.6917994558880683,
"grad_norm": 0.23644517362117767,
"learning_rate": 9.14997028475436e-06,
"loss": 0.0076,
"step": 1335
},
{
"epoch": 0.6923176577276849,
"grad_norm": 0.14124254882335663,
"learning_rate": 9.121774902370628e-06,
"loss": 0.0104,
"step": 1336
},
{
"epoch": 0.6928358595673014,
"grad_norm": 0.37195298075675964,
"learning_rate": 9.09361018972797e-06,
"loss": 0.0303,
"step": 1337
},
{
"epoch": 0.693354061406918,
"grad_norm": 0.24142932891845703,
"learning_rate": 9.065476226233143e-06,
"loss": 0.0097,
"step": 1338
},
{
"epoch": 0.6938722632465345,
"grad_norm": 0.17429199814796448,
"learning_rate": 9.037373091206182e-06,
"loss": 0.0134,
"step": 1339
},
{
"epoch": 0.6943904650861511,
"grad_norm": 0.17789122462272644,
"learning_rate": 9.009300863880236e-06,
"loss": 0.0131,
"step": 1340
},
{
"epoch": 0.6949086669257676,
"grad_norm": 0.1673128455877304,
"learning_rate": 8.981259623401288e-06,
"loss": 0.0164,
"step": 1341
},
{
"epoch": 0.6954268687653841,
"grad_norm": 0.20929870009422302,
"learning_rate": 8.953249448827971e-06,
"loss": 0.0098,
"step": 1342
},
{
"epoch": 0.6959450706050007,
"grad_norm": 0.1541186422109604,
"learning_rate": 8.925270419131342e-06,
"loss": 0.0134,
"step": 1343
},
{
"epoch": 0.6964632724446171,
"grad_norm": 0.2542741596698761,
"learning_rate": 8.89732261319462e-06,
"loss": 0.0255,
"step": 1344
},
{
"epoch": 0.6969814742842337,
"grad_norm": 0.20983968675136566,
"learning_rate": 8.869406109813017e-06,
"loss": 0.0146,
"step": 1345
},
{
"epoch": 0.6974996761238502,
"grad_norm": 0.3096301555633545,
"learning_rate": 8.841520987693486e-06,
"loss": 0.014,
"step": 1346
},
{
"epoch": 0.6980178779634668,
"grad_norm": 0.20430825650691986,
"learning_rate": 8.813667325454507e-06,
"loss": 0.0156,
"step": 1347
},
{
"epoch": 0.6985360798030833,
"grad_norm": 0.14121589064598083,
"learning_rate": 8.785845201625844e-06,
"loss": 0.0077,
"step": 1348
},
{
"epoch": 0.6990542816426998,
"grad_norm": 0.11843699216842651,
"learning_rate": 8.758054694648368e-06,
"loss": 0.0126,
"step": 1349
},
{
"epoch": 0.6995724834823164,
"grad_norm": 0.32668405771255493,
"learning_rate": 8.730295882873798e-06,
"loss": 0.012,
"step": 1350
},
{
"epoch": 0.7000906853219329,
"grad_norm": 0.2675720751285553,
"learning_rate": 8.702568844564498e-06,
"loss": 0.0244,
"step": 1351
},
{
"epoch": 0.7006088871615495,
"grad_norm": 0.15671458840370178,
"learning_rate": 8.674873657893236e-06,
"loss": 0.01,
"step": 1352
},
{
"epoch": 0.7011270890011659,
"grad_norm": 0.10088402777910233,
"learning_rate": 8.647210400942998e-06,
"loss": 0.0052,
"step": 1353
},
{
"epoch": 0.7016452908407825,
"grad_norm": 0.5230119824409485,
"learning_rate": 8.619579151706735e-06,
"loss": 0.0263,
"step": 1354
},
{
"epoch": 0.702163492680399,
"grad_norm": 0.1367189586162567,
"learning_rate": 8.591979988087173e-06,
"loss": 0.0047,
"step": 1355
},
{
"epoch": 0.7026816945200155,
"grad_norm": 0.2677091658115387,
"learning_rate": 8.564412987896549e-06,
"loss": 0.0257,
"step": 1356
},
{
"epoch": 0.7031998963596321,
"grad_norm": 0.5703830122947693,
"learning_rate": 8.536878228856447e-06,
"loss": 0.0145,
"step": 1357
},
{
"epoch": 0.7037180981992486,
"grad_norm": 0.2206091731786728,
"learning_rate": 8.509375788597528e-06,
"loss": 0.0072,
"step": 1358
},
{
"epoch": 0.7042363000388652,
"grad_norm": 0.19818182289600372,
"learning_rate": 8.48190574465936e-06,
"loss": 0.0111,
"step": 1359
},
{
"epoch": 0.7047545018784817,
"grad_norm": 0.10023718327283859,
"learning_rate": 8.454468174490145e-06,
"loss": 0.0115,
"step": 1360
},
{
"epoch": 0.7052727037180982,
"grad_norm": 0.14288294315338135,
"learning_rate": 8.42706315544655e-06,
"loss": 0.0099,
"step": 1361
},
{
"epoch": 0.7057909055577147,
"grad_norm": 0.2115078568458557,
"learning_rate": 8.399690764793464e-06,
"loss": 0.0116,
"step": 1362
},
{
"epoch": 0.7063091073973312,
"grad_norm": 0.17733272910118103,
"learning_rate": 8.372351079703788e-06,
"loss": 0.0157,
"step": 1363
},
{
"epoch": 0.7068273092369478,
"grad_norm": 0.08629205077886581,
"learning_rate": 8.3450441772582e-06,
"loss": 0.0033,
"step": 1364
},
{
"epoch": 0.7073455110765643,
"grad_norm": 0.1436304897069931,
"learning_rate": 8.317770134444962e-06,
"loss": 0.018,
"step": 1365
},
{
"epoch": 0.7078637129161809,
"grad_norm": 0.27016061544418335,
"learning_rate": 8.290529028159696e-06,
"loss": 0.0254,
"step": 1366
},
{
"epoch": 0.7083819147557974,
"grad_norm": 0.2512115240097046,
"learning_rate": 8.26332093520516e-06,
"loss": 0.0224,
"step": 1367
},
{
"epoch": 0.708900116595414,
"grad_norm": 0.057625047862529755,
"learning_rate": 8.236145932291022e-06,
"loss": 0.0022,
"step": 1368
},
{
"epoch": 0.7094183184350304,
"grad_norm": 0.10072803497314453,
"learning_rate": 8.20900409603368e-06,
"loss": 0.0097,
"step": 1369
},
{
"epoch": 0.7099365202746469,
"grad_norm": 0.1774268001317978,
"learning_rate": 8.181895502956006e-06,
"loss": 0.0207,
"step": 1370
},
{
"epoch": 0.7104547221142635,
"grad_norm": 0.2120877206325531,
"learning_rate": 8.154820229487163e-06,
"loss": 0.0182,
"step": 1371
},
{
"epoch": 0.71097292395388,
"grad_norm": 0.12090042233467102,
"learning_rate": 8.127778351962349e-06,
"loss": 0.0047,
"step": 1372
},
{
"epoch": 0.7114911257934966,
"grad_norm": 0.2732641100883484,
"learning_rate": 8.100769946622632e-06,
"loss": 0.0178,
"step": 1373
},
{
"epoch": 0.7120093276331131,
"grad_norm": 0.12582547962665558,
"learning_rate": 8.0737950896147e-06,
"loss": 0.0053,
"step": 1374
},
{
"epoch": 0.7125275294727297,
"grad_norm": 0.10372617095708847,
"learning_rate": 8.046853856990643e-06,
"loss": 0.0083,
"step": 1375
},
{
"epoch": 0.7130457313123462,
"grad_norm": 0.1782783567905426,
"learning_rate": 8.019946324707776e-06,
"loss": 0.0167,
"step": 1376
},
{
"epoch": 0.7135639331519626,
"grad_norm": 0.2992905378341675,
"learning_rate": 7.993072568628378e-06,
"loss": 0.0197,
"step": 1377
},
{
"epoch": 0.7140821349915792,
"grad_norm": 0.3926917016506195,
"learning_rate": 7.966232664519508e-06,
"loss": 0.0256,
"step": 1378
},
{
"epoch": 0.7146003368311957,
"grad_norm": 0.1238306313753128,
"learning_rate": 7.9394266880528e-06,
"loss": 0.0075,
"step": 1379
},
{
"epoch": 0.7151185386708123,
"grad_norm": 0.1858736127614975,
"learning_rate": 7.912654714804201e-06,
"loss": 0.0112,
"step": 1380
},
{
"epoch": 0.7156367405104288,
"grad_norm": 0.08168400824069977,
"learning_rate": 7.885916820253818e-06,
"loss": 0.005,
"step": 1381
},
{
"epoch": 0.7161549423500454,
"grad_norm": 0.1177031546831131,
"learning_rate": 7.859213079785665e-06,
"loss": 0.0069,
"step": 1382
},
{
"epoch": 0.7166731441896619,
"grad_norm": 0.2270517796278,
"learning_rate": 7.832543568687474e-06,
"loss": 0.0127,
"step": 1383
},
{
"epoch": 0.7171913460292784,
"grad_norm": 0.12286113202571869,
"learning_rate": 7.805908362150449e-06,
"loss": 0.007,
"step": 1384
},
{
"epoch": 0.717709547868895,
"grad_norm": 0.2528562843799591,
"learning_rate": 7.779307535269099e-06,
"loss": 0.0126,
"step": 1385
},
{
"epoch": 0.7182277497085114,
"grad_norm": 0.289711594581604,
"learning_rate": 7.752741163040996e-06,
"loss": 0.0229,
"step": 1386
},
{
"epoch": 0.718745951548128,
"grad_norm": 0.24089190363883972,
"learning_rate": 7.726209320366575e-06,
"loss": 0.009,
"step": 1387
},
{
"epoch": 0.7192641533877445,
"grad_norm": 0.24581708014011383,
"learning_rate": 7.699712082048908e-06,
"loss": 0.013,
"step": 1388
},
{
"epoch": 0.719782355227361,
"grad_norm": 0.19005197286605835,
"learning_rate": 7.67324952279352e-06,
"loss": 0.0153,
"step": 1389
},
{
"epoch": 0.7203005570669776,
"grad_norm": 0.19025154411792755,
"learning_rate": 7.646821717208153e-06,
"loss": 0.0094,
"step": 1390
},
{
"epoch": 0.7208187589065941,
"grad_norm": 0.325736403465271,
"learning_rate": 7.620428739802575e-06,
"loss": 0.0228,
"step": 1391
},
{
"epoch": 0.7213369607462107,
"grad_norm": 0.1527811884880066,
"learning_rate": 7.594070664988338e-06,
"loss": 0.0101,
"step": 1392
},
{
"epoch": 0.7218551625858272,
"grad_norm": 0.2867499589920044,
"learning_rate": 7.567747567078625e-06,
"loss": 0.0218,
"step": 1393
},
{
"epoch": 0.7223733644254438,
"grad_norm": 0.09495671093463898,
"learning_rate": 7.54145952028797e-06,
"loss": 0.0051,
"step": 1394
},
{
"epoch": 0.7228915662650602,
"grad_norm": 0.376874178647995,
"learning_rate": 7.515206598732119e-06,
"loss": 0.0132,
"step": 1395
},
{
"epoch": 0.7234097681046767,
"grad_norm": 0.3421163558959961,
"learning_rate": 7.488988876427754e-06,
"loss": 0.0243,
"step": 1396
},
{
"epoch": 0.7239279699442933,
"grad_norm": 0.27136725187301636,
"learning_rate": 7.462806427292342e-06,
"loss": 0.0154,
"step": 1397
},
{
"epoch": 0.7244461717839098,
"grad_norm": 0.4345359802246094,
"learning_rate": 7.436659325143893e-06,
"loss": 0.0247,
"step": 1398
},
{
"epoch": 0.7249643736235264,
"grad_norm": 0.13415028154850006,
"learning_rate": 7.41054764370077e-06,
"loss": 0.0079,
"step": 1399
},
{
"epoch": 0.7254825754631429,
"grad_norm": 0.3060521185398102,
"learning_rate": 7.384471456581448e-06,
"loss": 0.0308,
"step": 1400
},
{
"epoch": 0.7260007773027595,
"grad_norm": 0.13133402168750763,
"learning_rate": 7.358430837304353e-06,
"loss": 0.0083,
"step": 1401
},
{
"epoch": 0.726518979142376,
"grad_norm": 0.168381467461586,
"learning_rate": 7.332425859287624e-06,
"loss": 0.0075,
"step": 1402
},
{
"epoch": 0.7270371809819924,
"grad_norm": 0.12512297928333282,
"learning_rate": 7.306456595848923e-06,
"loss": 0.0066,
"step": 1403
},
{
"epoch": 0.727555382821609,
"grad_norm": 0.1732812225818634,
"learning_rate": 7.280523120205196e-06,
"loss": 0.0146,
"step": 1404
},
{
"epoch": 0.7280735846612255,
"grad_norm": 0.27395644783973694,
"learning_rate": 7.254625505472513e-06,
"loss": 0.0174,
"step": 1405
},
{
"epoch": 0.7285917865008421,
"grad_norm": 0.24181146919727325,
"learning_rate": 7.228763824665832e-06,
"loss": 0.0203,
"step": 1406
},
{
"epoch": 0.7291099883404586,
"grad_norm": 0.32806146144866943,
"learning_rate": 7.202938150698804e-06,
"loss": 0.0181,
"step": 1407
},
{
"epoch": 0.7296281901800752,
"grad_norm": 0.12564238905906677,
"learning_rate": 7.1771485563835465e-06,
"loss": 0.0094,
"step": 1408
},
{
"epoch": 0.7301463920196917,
"grad_norm": 0.3716842532157898,
"learning_rate": 7.151395114430473e-06,
"loss": 0.025,
"step": 1409
},
{
"epoch": 0.7306645938593082,
"grad_norm": 0.06230960413813591,
"learning_rate": 7.1256778974480686e-06,
"loss": 0.0041,
"step": 1410
},
{
"epoch": 0.7311827956989247,
"grad_norm": 0.32549139857292175,
"learning_rate": 7.099996977942673e-06,
"loss": 0.0137,
"step": 1411
},
{
"epoch": 0.7317009975385412,
"grad_norm": 0.13841503858566284,
"learning_rate": 7.074352428318312e-06,
"loss": 0.0106,
"step": 1412
},
{
"epoch": 0.7322191993781578,
"grad_norm": 0.12878160178661346,
"learning_rate": 7.048744320876444e-06,
"loss": 0.0053,
"step": 1413
},
{
"epoch": 0.7327374012177743,
"grad_norm": 0.14967595040798187,
"learning_rate": 7.0231727278158035e-06,
"loss": 0.0051,
"step": 1414
},
{
"epoch": 0.7332556030573909,
"grad_norm": 0.16011378169059753,
"learning_rate": 6.99763772123218e-06,
"loss": 0.0055,
"step": 1415
},
{
"epoch": 0.7337738048970074,
"grad_norm": 0.29788485169410706,
"learning_rate": 6.97213937311819e-06,
"loss": 0.0221,
"step": 1416
},
{
"epoch": 0.7342920067366239,
"grad_norm": 0.33558785915374756,
"learning_rate": 6.946677755363116e-06,
"loss": 0.0087,
"step": 1417
},
{
"epoch": 0.7348102085762405,
"grad_norm": 0.22806207835674286,
"learning_rate": 6.921252939752681e-06,
"loss": 0.0257,
"step": 1418
},
{
"epoch": 0.735328410415857,
"grad_norm": 0.24824737012386322,
"learning_rate": 6.895864997968851e-06,
"loss": 0.0143,
"step": 1419
},
{
"epoch": 0.7358466122554735,
"grad_norm": 0.1711535006761551,
"learning_rate": 6.8705140015896145e-06,
"loss": 0.0099,
"step": 1420
},
{
"epoch": 0.73636481409509,
"grad_norm": 0.19238120317459106,
"learning_rate": 6.8452000220888135e-06,
"loss": 0.0104,
"step": 1421
},
{
"epoch": 0.7368830159347066,
"grad_norm": 0.21341420710086823,
"learning_rate": 6.819923130835926e-06,
"loss": 0.0266,
"step": 1422
},
{
"epoch": 0.7374012177743231,
"grad_norm": 0.23991352319717407,
"learning_rate": 6.79468339909586e-06,
"loss": 0.0159,
"step": 1423
},
{
"epoch": 0.7379194196139396,
"grad_norm": 0.15172316133975983,
"learning_rate": 6.769480898028751e-06,
"loss": 0.0123,
"step": 1424
},
{
"epoch": 0.7384376214535562,
"grad_norm": 0.11418157070875168,
"learning_rate": 6.744315698689774e-06,
"loss": 0.0089,
"step": 1425
},
{
"epoch": 0.7389558232931727,
"grad_norm": 0.1747230738401413,
"learning_rate": 6.719187872028938e-06,
"loss": 0.0148,
"step": 1426
},
{
"epoch": 0.7394740251327893,
"grad_norm": 0.6717311143875122,
"learning_rate": 6.694097488890889e-06,
"loss": 0.0244,
"step": 1427
},
{
"epoch": 0.7399922269724057,
"grad_norm": 0.27945640683174133,
"learning_rate": 6.669044620014691e-06,
"loss": 0.0115,
"step": 1428
},
{
"epoch": 0.7405104288120223,
"grad_norm": 0.13768765330314636,
"learning_rate": 6.644029336033657e-06,
"loss": 0.0123,
"step": 1429
},
{
"epoch": 0.7410286306516388,
"grad_norm": 0.19821152091026306,
"learning_rate": 6.6190517074751145e-06,
"loss": 0.0162,
"step": 1430
},
{
"epoch": 0.7415468324912553,
"grad_norm": 0.2635100781917572,
"learning_rate": 6.594111804760257e-06,
"loss": 0.0164,
"step": 1431
},
{
"epoch": 0.7420650343308719,
"grad_norm": 0.16833581030368805,
"learning_rate": 6.56920969820388e-06,
"loss": 0.0117,
"step": 1432
},
{
"epoch": 0.7425832361704884,
"grad_norm": 0.12360576540231705,
"learning_rate": 6.5443454580142425e-06,
"loss": 0.0069,
"step": 1433
},
{
"epoch": 0.743101438010105,
"grad_norm": 0.1418430656194687,
"learning_rate": 6.519519154292837e-06,
"loss": 0.0077,
"step": 1434
},
{
"epoch": 0.7436196398497215,
"grad_norm": 0.21706590056419373,
"learning_rate": 6.494730857034202e-06,
"loss": 0.0152,
"step": 1435
},
{
"epoch": 0.7441378416893379,
"grad_norm": 0.2138921618461609,
"learning_rate": 6.46998063612571e-06,
"loss": 0.0142,
"step": 1436
},
{
"epoch": 0.7446560435289545,
"grad_norm": 0.37199369072914124,
"learning_rate": 6.445268561347393e-06,
"loss": 0.0298,
"step": 1437
},
{
"epoch": 0.745174245368571,
"grad_norm": 0.13695186376571655,
"learning_rate": 6.4205947023717344e-06,
"loss": 0.0057,
"step": 1438
},
{
"epoch": 0.7456924472081876,
"grad_norm": 0.19427013397216797,
"learning_rate": 6.395959128763476e-06,
"loss": 0.0125,
"step": 1439
},
{
"epoch": 0.7462106490478041,
"grad_norm": 0.3446883261203766,
"learning_rate": 6.371361909979401e-06,
"loss": 0.0246,
"step": 1440
},
{
"epoch": 0.7467288508874207,
"grad_norm": 0.14575010538101196,
"learning_rate": 6.3468031153681785e-06,
"loss": 0.0164,
"step": 1441
},
{
"epoch": 0.7472470527270372,
"grad_norm": 0.4180688261985779,
"learning_rate": 6.3222828141701335e-06,
"loss": 0.0192,
"step": 1442
},
{
"epoch": 0.7477652545666537,
"grad_norm": 0.3175865709781647,
"learning_rate": 6.297801075517074e-06,
"loss": 0.0189,
"step": 1443
},
{
"epoch": 0.7482834564062703,
"grad_norm": 0.16741381585597992,
"learning_rate": 6.2733579684320655e-06,
"loss": 0.0105,
"step": 1444
},
{
"epoch": 0.7488016582458867,
"grad_norm": 0.344389408826828,
"learning_rate": 6.248953561829276e-06,
"loss": 0.0225,
"step": 1445
},
{
"epoch": 0.7493198600855033,
"grad_norm": 0.31048911809921265,
"learning_rate": 6.224587924513763e-06,
"loss": 0.0107,
"step": 1446
},
{
"epoch": 0.7498380619251198,
"grad_norm": 0.3725675344467163,
"learning_rate": 6.200261125181262e-06,
"loss": 0.023,
"step": 1447
},
{
"epoch": 0.7503562637647364,
"grad_norm": 0.19193294644355774,
"learning_rate": 6.175973232418029e-06,
"loss": 0.0183,
"step": 1448
},
{
"epoch": 0.7508744656043529,
"grad_norm": 0.28560853004455566,
"learning_rate": 6.151724314700609e-06,
"loss": 0.0095,
"step": 1449
},
{
"epoch": 0.7513926674439694,
"grad_norm": 0.18966034054756165,
"learning_rate": 6.127514440395676e-06,
"loss": 0.0159,
"step": 1450
},
{
"epoch": 0.751910869283586,
"grad_norm": 0.13293716311454773,
"learning_rate": 6.103343677759834e-06,
"loss": 0.005,
"step": 1451
},
{
"epoch": 0.7524290711232025,
"grad_norm": 0.47814297676086426,
"learning_rate": 6.0792120949393885e-06,
"loss": 0.0251,
"step": 1452
},
{
"epoch": 0.752947272962819,
"grad_norm": 0.2816180884838104,
"learning_rate": 6.055119759970205e-06,
"loss": 0.0161,
"step": 1453
},
{
"epoch": 0.7534654748024355,
"grad_norm": 0.45219457149505615,
"learning_rate": 6.031066740777491e-06,
"loss": 0.0177,
"step": 1454
},
{
"epoch": 0.7539836766420521,
"grad_norm": 0.14777222275733948,
"learning_rate": 6.007053105175613e-06,
"loss": 0.0073,
"step": 1455
},
{
"epoch": 0.7545018784816686,
"grad_norm": 0.21083438396453857,
"learning_rate": 5.983078920867879e-06,
"loss": 0.0141,
"step": 1456
},
{
"epoch": 0.7550200803212851,
"grad_norm": 0.22181129455566406,
"learning_rate": 5.959144255446392e-06,
"loss": 0.013,
"step": 1457
},
{
"epoch": 0.7555382821609017,
"grad_norm": 0.11170242726802826,
"learning_rate": 5.935249176391828e-06,
"loss": 0.0052,
"step": 1458
},
{
"epoch": 0.7560564840005182,
"grad_norm": 0.21665945649147034,
"learning_rate": 5.911393751073262e-06,
"loss": 0.0115,
"step": 1459
},
{
"epoch": 0.7565746858401348,
"grad_norm": 0.22410617768764496,
"learning_rate": 5.8875780467479525e-06,
"loss": 0.0074,
"step": 1460
},
{
"epoch": 0.7570928876797512,
"grad_norm": 0.16454528272151947,
"learning_rate": 5.863802130561183e-06,
"loss": 0.0078,
"step": 1461
},
{
"epoch": 0.7576110895193678,
"grad_norm": 0.1663055717945099,
"learning_rate": 5.840066069546062e-06,
"loss": 0.0119,
"step": 1462
},
{
"epoch": 0.7581292913589843,
"grad_norm": 0.3612779378890991,
"learning_rate": 5.816369930623329e-06,
"loss": 0.0238,
"step": 1463
},
{
"epoch": 0.7586474931986008,
"grad_norm": 0.17329257726669312,
"learning_rate": 5.792713780601158e-06,
"loss": 0.0107,
"step": 1464
},
{
"epoch": 0.7591656950382174,
"grad_norm": 0.201755091547966,
"learning_rate": 5.769097686174989e-06,
"loss": 0.0097,
"step": 1465
},
{
"epoch": 0.7596838968778339,
"grad_norm": 0.26469019055366516,
"learning_rate": 5.74552171392734e-06,
"loss": 0.0111,
"step": 1466
},
{
"epoch": 0.7602020987174505,
"grad_norm": 0.08461549133062363,
"learning_rate": 5.721985930327585e-06,
"loss": 0.0038,
"step": 1467
},
{
"epoch": 0.760720300557067,
"grad_norm": 0.09042751044034958,
"learning_rate": 5.698490401731818e-06,
"loss": 0.0037,
"step": 1468
},
{
"epoch": 0.7612385023966836,
"grad_norm": 0.2408987283706665,
"learning_rate": 5.675035194382614e-06,
"loss": 0.0135,
"step": 1469
},
{
"epoch": 0.7617567042363,
"grad_norm": 0.1702120453119278,
"learning_rate": 5.651620374408886e-06,
"loss": 0.0184,
"step": 1470
},
{
"epoch": 0.7622749060759165,
"grad_norm": 0.25853922963142395,
"learning_rate": 5.6282460078256795e-06,
"loss": 0.0135,
"step": 1471
},
{
"epoch": 0.7627931079155331,
"grad_norm": 0.40889886021614075,
"learning_rate": 5.604912160533968e-06,
"loss": 0.0143,
"step": 1472
},
{
"epoch": 0.7633113097551496,
"grad_norm": 0.35814812779426575,
"learning_rate": 5.581618898320511e-06,
"loss": 0.0228,
"step": 1473
},
{
"epoch": 0.7638295115947662,
"grad_norm": 0.15147720277309418,
"learning_rate": 5.558366286857628e-06,
"loss": 0.0106,
"step": 1474
},
{
"epoch": 0.7643477134343827,
"grad_norm": 0.35429468750953674,
"learning_rate": 5.535154391703039e-06,
"loss": 0.0231,
"step": 1475
},
{
"epoch": 0.7648659152739993,
"grad_norm": 0.28859689831733704,
"learning_rate": 5.511983278299655e-06,
"loss": 0.011,
"step": 1476
},
{
"epoch": 0.7653841171136158,
"grad_norm": 0.22457551956176758,
"learning_rate": 5.488853011975421e-06,
"loss": 0.0123,
"step": 1477
},
{
"epoch": 0.7659023189532322,
"grad_norm": 0.29618382453918457,
"learning_rate": 5.465763657943115e-06,
"loss": 0.0175,
"step": 1478
},
{
"epoch": 0.7664205207928488,
"grad_norm": 0.28490161895751953,
"learning_rate": 5.442715281300175e-06,
"loss": 0.0155,
"step": 1479
},
{
"epoch": 0.7669387226324653,
"grad_norm": 0.2284500151872635,
"learning_rate": 5.41970794702849e-06,
"loss": 0.0145,
"step": 1480
},
{
"epoch": 0.7674569244720819,
"grad_norm": 0.25994613766670227,
"learning_rate": 5.396741719994252e-06,
"loss": 0.0146,
"step": 1481
},
{
"epoch": 0.7679751263116984,
"grad_norm": 0.1707499772310257,
"learning_rate": 5.373816664947753e-06,
"loss": 0.0062,
"step": 1482
},
{
"epoch": 0.768493328151315,
"grad_norm": 0.17810523509979248,
"learning_rate": 5.3509328465232095e-06,
"loss": 0.0183,
"step": 1483
},
{
"epoch": 0.7690115299909315,
"grad_norm": 0.1567215621471405,
"learning_rate": 5.32809032923856e-06,
"loss": 0.0086,
"step": 1484
},
{
"epoch": 0.769529731830548,
"grad_norm": 0.11671585589647293,
"learning_rate": 5.305289177495323e-06,
"loss": 0.0065,
"step": 1485
},
{
"epoch": 0.7700479336701646,
"grad_norm": 0.16693814098834991,
"learning_rate": 5.282529455578369e-06,
"loss": 0.0067,
"step": 1486
},
{
"epoch": 0.770566135509781,
"grad_norm": 0.15044854581356049,
"learning_rate": 5.259811227655789e-06,
"loss": 0.0061,
"step": 1487
},
{
"epoch": 0.7710843373493976,
"grad_norm": 0.2685668170452118,
"learning_rate": 5.2371345577786555e-06,
"loss": 0.0145,
"step": 1488
},
{
"epoch": 0.7716025391890141,
"grad_norm": 0.24585674703121185,
"learning_rate": 5.214499509880901e-06,
"loss": 0.009,
"step": 1489
},
{
"epoch": 0.7721207410286306,
"grad_norm": 0.48471522331237793,
"learning_rate": 5.191906147779098e-06,
"loss": 0.0177,
"step": 1490
},
{
"epoch": 0.7726389428682472,
"grad_norm": 0.27829957008361816,
"learning_rate": 5.1693545351722974e-06,
"loss": 0.0313,
"step": 1491
},
{
"epoch": 0.7731571447078637,
"grad_norm": 0.1682707667350769,
"learning_rate": 5.146844735641827e-06,
"loss": 0.0137,
"step": 1492
},
{
"epoch": 0.7736753465474803,
"grad_norm": 0.12730175256729126,
"learning_rate": 5.124376812651149e-06,
"loss": 0.0079,
"step": 1493
},
{
"epoch": 0.7741935483870968,
"grad_norm": 0.2443925142288208,
"learning_rate": 5.101950829545646e-06,
"loss": 0.0138,
"step": 1494
},
{
"epoch": 0.7747117502267133,
"grad_norm": 0.2247922420501709,
"learning_rate": 5.079566849552466e-06,
"loss": 0.0103,
"step": 1495
},
{
"epoch": 0.7752299520663298,
"grad_norm": 0.38228839635849,
"learning_rate": 5.057224935780325e-06,
"loss": 0.0147,
"step": 1496
},
{
"epoch": 0.7757481539059463,
"grad_norm": 0.425642192363739,
"learning_rate": 5.034925151219343e-06,
"loss": 0.0204,
"step": 1497
},
{
"epoch": 0.7762663557455629,
"grad_norm": 0.16054239869117737,
"learning_rate": 5.012667558740865e-06,
"loss": 0.007,
"step": 1498
},
{
"epoch": 0.7767845575851794,
"grad_norm": 0.13907340168952942,
"learning_rate": 4.9904522210972836e-06,
"loss": 0.0033,
"step": 1499
},
{
"epoch": 0.777302759424796,
"grad_norm": 0.32811591029167175,
"learning_rate": 4.968279200921844e-06,
"loss": 0.01,
"step": 1500
},
{
"epoch": 0.7778209612644125,
"grad_norm": 0.31322336196899414,
"learning_rate": 4.946148560728501e-06,
"loss": 0.0197,
"step": 1501
},
{
"epoch": 0.7783391631040291,
"grad_norm": 0.25639650225639343,
"learning_rate": 4.9240603629117175e-06,
"loss": 0.0124,
"step": 1502
},
{
"epoch": 0.7788573649436455,
"grad_norm": 0.31371498107910156,
"learning_rate": 4.902014669746286e-06,
"loss": 0.0152,
"step": 1503
},
{
"epoch": 0.779375566783262,
"grad_norm": 0.15151162445545197,
"learning_rate": 4.880011543387186e-06,
"loss": 0.011,
"step": 1504
},
{
"epoch": 0.7798937686228786,
"grad_norm": 0.2916856110095978,
"learning_rate": 4.858051045869354e-06,
"loss": 0.0136,
"step": 1505
},
{
"epoch": 0.7804119704624951,
"grad_norm": 0.25480547547340393,
"learning_rate": 4.836133239107567e-06,
"loss": 0.0234,
"step": 1506
},
{
"epoch": 0.7809301723021117,
"grad_norm": 0.18069693446159363,
"learning_rate": 4.814258184896234e-06,
"loss": 0.0098,
"step": 1507
},
{
"epoch": 0.7814483741417282,
"grad_norm": 0.1400531381368637,
"learning_rate": 4.792425944909218e-06,
"loss": 0.0077,
"step": 1508
},
{
"epoch": 0.7819665759813448,
"grad_norm": 0.27684223651885986,
"learning_rate": 4.770636580699681e-06,
"loss": 0.0176,
"step": 1509
},
{
"epoch": 0.7824847778209613,
"grad_norm": 0.12374227494001389,
"learning_rate": 4.748890153699907e-06,
"loss": 0.009,
"step": 1510
},
{
"epoch": 0.7830029796605777,
"grad_norm": 0.06070549041032791,
"learning_rate": 4.7271867252211224e-06,
"loss": 0.0047,
"step": 1511
},
{
"epoch": 0.7835211815001943,
"grad_norm": 0.2304593175649643,
"learning_rate": 4.7055263564533096e-06,
"loss": 0.0131,
"step": 1512
},
{
"epoch": 0.7840393833398108,
"grad_norm": 0.24694843590259552,
"learning_rate": 4.683909108465068e-06,
"loss": 0.0104,
"step": 1513
},
{
"epoch": 0.7845575851794274,
"grad_norm": 0.22078731656074524,
"learning_rate": 4.662335042203417e-06,
"loss": 0.0157,
"step": 1514
},
{
"epoch": 0.7850757870190439,
"grad_norm": 0.13630007207393646,
"learning_rate": 4.640804218493638e-06,
"loss": 0.0028,
"step": 1515
},
{
"epoch": 0.7855939888586605,
"grad_norm": 0.5278586149215698,
"learning_rate": 4.6193166980390755e-06,
"loss": 0.0158,
"step": 1516
},
{
"epoch": 0.786112190698277,
"grad_norm": 0.08716629445552826,
"learning_rate": 4.597872541421007e-06,
"loss": 0.0089,
"step": 1517
},
{
"epoch": 0.7866303925378935,
"grad_norm": 0.18373537063598633,
"learning_rate": 4.5764718090984395e-06,
"loss": 0.0065,
"step": 1518
},
{
"epoch": 0.7871485943775101,
"grad_norm": 0.3253873586654663,
"learning_rate": 4.555114561407965e-06,
"loss": 0.0167,
"step": 1519
},
{
"epoch": 0.7876667962171265,
"grad_norm": 0.22763191163539886,
"learning_rate": 4.533800858563553e-06,
"loss": 0.0191,
"step": 1520
},
{
"epoch": 0.7881849980567431,
"grad_norm": 0.09696512669324875,
"learning_rate": 4.512530760656429e-06,
"loss": 0.007,
"step": 1521
},
{
"epoch": 0.7887031998963596,
"grad_norm": 0.2907656133174896,
"learning_rate": 4.491304327654855e-06,
"loss": 0.0091,
"step": 1522
},
{
"epoch": 0.7892214017359762,
"grad_norm": 0.3057742714881897,
"learning_rate": 4.4701216194040105e-06,
"loss": 0.0132,
"step": 1523
},
{
"epoch": 0.7897396035755927,
"grad_norm": 0.374776154756546,
"learning_rate": 4.448982695625778e-06,
"loss": 0.0206,
"step": 1524
},
{
"epoch": 0.7902578054152092,
"grad_norm": 0.14000919461250305,
"learning_rate": 4.4278876159186064e-06,
"loss": 0.013,
"step": 1525
},
{
"epoch": 0.7907760072548258,
"grad_norm": 0.2944836914539337,
"learning_rate": 4.4068364397573335e-06,
"loss": 0.0087,
"step": 1526
},
{
"epoch": 0.7912942090944423,
"grad_norm": 0.5514625310897827,
"learning_rate": 4.385829226493015e-06,
"loss": 0.0183,
"step": 1527
},
{
"epoch": 0.7918124109340589,
"grad_norm": 0.29109495878219604,
"learning_rate": 4.364866035352749e-06,
"loss": 0.0127,
"step": 1528
},
{
"epoch": 0.7923306127736753,
"grad_norm": 0.5156393051147461,
"learning_rate": 4.343946925439529e-06,
"loss": 0.0207,
"step": 1529
},
{
"epoch": 0.7928488146132919,
"grad_norm": 0.2541627883911133,
"learning_rate": 4.323071955732065e-06,
"loss": 0.0173,
"step": 1530
},
{
"epoch": 0.7933670164529084,
"grad_norm": 0.2753004729747772,
"learning_rate": 4.30224118508463e-06,
"loss": 0.018,
"step": 1531
},
{
"epoch": 0.7938852182925249,
"grad_norm": 0.15237921476364136,
"learning_rate": 4.2814546722268595e-06,
"loss": 0.0178,
"step": 1532
},
{
"epoch": 0.7944034201321415,
"grad_norm": 0.0733996033668518,
"learning_rate": 4.260712475763627e-06,
"loss": 0.0069,
"step": 1533
},
{
"epoch": 0.794921621971758,
"grad_norm": 0.159640833735466,
"learning_rate": 4.240014654174857e-06,
"loss": 0.0123,
"step": 1534
},
{
"epoch": 0.7954398238113746,
"grad_norm": 0.26743224263191223,
"learning_rate": 4.219361265815378e-06,
"loss": 0.0073,
"step": 1535
},
{
"epoch": 0.795958025650991,
"grad_norm": 0.15401487052440643,
"learning_rate": 4.1987523689147155e-06,
"loss": 0.0079,
"step": 1536
},
{
"epoch": 0.7964762274906076,
"grad_norm": 0.22780843079090118,
"learning_rate": 4.178188021576983e-06,
"loss": 0.0079,
"step": 1537
},
{
"epoch": 0.7969944293302241,
"grad_norm": 0.3683816194534302,
"learning_rate": 4.157668281780689e-06,
"loss": 0.0155,
"step": 1538
},
{
"epoch": 0.7975126311698406,
"grad_norm": 0.27427056431770325,
"learning_rate": 4.137193207378561e-06,
"loss": 0.0213,
"step": 1539
},
{
"epoch": 0.7980308330094572,
"grad_norm": 0.1827021986246109,
"learning_rate": 4.116762856097416e-06,
"loss": 0.0092,
"step": 1540
},
{
"epoch": 0.7985490348490737,
"grad_norm": 0.08515940606594086,
"learning_rate": 4.096377285537965e-06,
"loss": 0.0036,
"step": 1541
},
{
"epoch": 0.7990672366886903,
"grad_norm": 0.23422442376613617,
"learning_rate": 4.076036553174678e-06,
"loss": 0.0271,
"step": 1542
},
{
"epoch": 0.7995854385283068,
"grad_norm": 0.21451188623905182,
"learning_rate": 4.05574071635561e-06,
"loss": 0.0212,
"step": 1543
},
{
"epoch": 0.8001036403679233,
"grad_norm": 0.08211307972669601,
"learning_rate": 4.035489832302224e-06,
"loss": 0.0045,
"step": 1544
},
{
"epoch": 0.8006218422075398,
"grad_norm": 0.2552872896194458,
"learning_rate": 4.01528395810926e-06,
"loss": 0.0147,
"step": 1545
},
{
"epoch": 0.8011400440471563,
"grad_norm": 0.2829188406467438,
"learning_rate": 3.995123150744548e-06,
"loss": 0.021,
"step": 1546
},
{
"epoch": 0.8016582458867729,
"grad_norm": 0.33412739634513855,
"learning_rate": 3.975007467048875e-06,
"loss": 0.0249,
"step": 1547
},
{
"epoch": 0.8021764477263894,
"grad_norm": 0.05663935840129852,
"learning_rate": 3.954936963735778e-06,
"loss": 0.0023,
"step": 1548
},
{
"epoch": 0.802694649566006,
"grad_norm": 0.15037386119365692,
"learning_rate": 3.934911697391442e-06,
"loss": 0.0047,
"step": 1549
},
{
"epoch": 0.8032128514056225,
"grad_norm": 0.09823647886514664,
"learning_rate": 3.9149317244745e-06,
"loss": 0.0043,
"step": 1550
},
{
"epoch": 0.803731053245239,
"grad_norm": 0.1772301197052002,
"learning_rate": 3.8949971013158894e-06,
"loss": 0.0197,
"step": 1551
},
{
"epoch": 0.8042492550848556,
"grad_norm": 0.2572917342185974,
"learning_rate": 3.875107884118681e-06,
"loss": 0.0237,
"step": 1552
},
{
"epoch": 0.804767456924472,
"grad_norm": 0.1350460946559906,
"learning_rate": 3.8552641289579405e-06,
"loss": 0.0099,
"step": 1553
},
{
"epoch": 0.8052856587640886,
"grad_norm": 0.2017035335302353,
"learning_rate": 3.835465891780554e-06,
"loss": 0.0105,
"step": 1554
},
{
"epoch": 0.8058038606037051,
"grad_norm": 0.19643430411815643,
"learning_rate": 3.815713228405085e-06,
"loss": 0.0115,
"step": 1555
},
{
"epoch": 0.8063220624433217,
"grad_norm": 0.1383439600467682,
"learning_rate": 3.7960061945215863e-06,
"loss": 0.0089,
"step": 1556
},
{
"epoch": 0.8068402642829382,
"grad_norm": 0.16496983170509338,
"learning_rate": 3.7763448456914886e-06,
"loss": 0.0076,
"step": 1557
},
{
"epoch": 0.8073584661225547,
"grad_norm": 0.27516958117485046,
"learning_rate": 3.756729237347396e-06,
"loss": 0.0111,
"step": 1558
},
{
"epoch": 0.8078766679621713,
"grad_norm": 0.7920728325843811,
"learning_rate": 3.7371594247929776e-06,
"loss": 0.0204,
"step": 1559
},
{
"epoch": 0.8083948698017878,
"grad_norm": 0.21858440339565277,
"learning_rate": 3.7176354632027668e-06,
"loss": 0.019,
"step": 1560
},
{
"epoch": 0.8089130716414044,
"grad_norm": 0.3572098910808563,
"learning_rate": 3.698157407622034e-06,
"loss": 0.0384,
"step": 1561
},
{
"epoch": 0.8094312734810208,
"grad_norm": 0.23645535111427307,
"learning_rate": 3.678725312966629e-06,
"loss": 0.01,
"step": 1562
},
{
"epoch": 0.8099494753206374,
"grad_norm": 0.222391739487648,
"learning_rate": 3.6593392340228183e-06,
"loss": 0.0173,
"step": 1563
},
{
"epoch": 0.8104676771602539,
"grad_norm": 0.1789429634809494,
"learning_rate": 3.6399992254471215e-06,
"loss": 0.0097,
"step": 1564
},
{
"epoch": 0.8109858789998704,
"grad_norm": 0.37951716780662537,
"learning_rate": 3.620705341766184e-06,
"loss": 0.0233,
"step": 1565
},
{
"epoch": 0.811504080839487,
"grad_norm": 0.17974528670310974,
"learning_rate": 3.6014576373765973e-06,
"loss": 0.013,
"step": 1566
},
{
"epoch": 0.8120222826791035,
"grad_norm": 0.12391126900911331,
"learning_rate": 3.582256166544773e-06,
"loss": 0.0084,
"step": 1567
},
{
"epoch": 0.8125404845187201,
"grad_norm": 0.0891820415854454,
"learning_rate": 3.563100983406744e-06,
"loss": 0.0033,
"step": 1568
},
{
"epoch": 0.8130586863583366,
"grad_norm": 0.06003464758396149,
"learning_rate": 3.543992141968069e-06,
"loss": 0.0019,
"step": 1569
},
{
"epoch": 0.8135768881979532,
"grad_norm": 0.21977417171001434,
"learning_rate": 3.5249296961036384e-06,
"loss": 0.0143,
"step": 1570
},
{
"epoch": 0.8140950900375696,
"grad_norm": 0.12480836361646652,
"learning_rate": 3.5059136995575438e-06,
"loss": 0.0115,
"step": 1571
},
{
"epoch": 0.8146132918771861,
"grad_norm": 0.09526639431715012,
"learning_rate": 3.4869442059429084e-06,
"loss": 0.0101,
"step": 1572
},
{
"epoch": 0.8151314937168027,
"grad_norm": 0.20937477052211761,
"learning_rate": 3.4680212687417524e-06,
"loss": 0.0083,
"step": 1573
},
{
"epoch": 0.8156496955564192,
"grad_norm": 0.16887736320495605,
"learning_rate": 3.449144941304847e-06,
"loss": 0.0073,
"step": 1574
},
{
"epoch": 0.8161678973960358,
"grad_norm": 0.35441237688064575,
"learning_rate": 3.43031527685153e-06,
"loss": 0.0281,
"step": 1575
},
{
"epoch": 0.8166860992356523,
"grad_norm": 0.07251258194446564,
"learning_rate": 3.411532328469602e-06,
"loss": 0.007,
"step": 1576
},
{
"epoch": 0.8172043010752689,
"grad_norm": 0.1627056747674942,
"learning_rate": 3.392796149115132e-06,
"loss": 0.0053,
"step": 1577
},
{
"epoch": 0.8177225029148854,
"grad_norm": 0.19298546016216278,
"learning_rate": 3.3741067916123504e-06,
"loss": 0.0122,
"step": 1578
},
{
"epoch": 0.8182407047545018,
"grad_norm": 0.1412855088710785,
"learning_rate": 3.355464308653471e-06,
"loss": 0.0062,
"step": 1579
},
{
"epoch": 0.8187589065941184,
"grad_norm": 0.461653470993042,
"learning_rate": 3.336868752798541e-06,
"loss": 0.0097,
"step": 1580
},
{
"epoch": 0.8192771084337349,
"grad_norm": 0.22594031691551208,
"learning_rate": 3.31832017647532e-06,
"loss": 0.0142,
"step": 1581
},
{
"epoch": 0.8197953102733515,
"grad_norm": 0.23084904253482819,
"learning_rate": 3.2998186319791037e-06,
"loss": 0.0169,
"step": 1582
},
{
"epoch": 0.820313512112968,
"grad_norm": 0.30051642656326294,
"learning_rate": 3.2813641714725963e-06,
"loss": 0.0183,
"step": 1583
},
{
"epoch": 0.8208317139525846,
"grad_norm": 0.184076189994812,
"learning_rate": 3.262956846985741e-06,
"loss": 0.0139,
"step": 1584
},
{
"epoch": 0.8213499157922011,
"grad_norm": 0.14037007093429565,
"learning_rate": 3.2445967104156018e-06,
"loss": 0.0055,
"step": 1585
},
{
"epoch": 0.8218681176318176,
"grad_norm": 0.1952822506427765,
"learning_rate": 3.226283813526192e-06,
"loss": 0.0115,
"step": 1586
},
{
"epoch": 0.8223863194714341,
"grad_norm": 0.2174348533153534,
"learning_rate": 3.2080182079483492e-06,
"loss": 0.0116,
"step": 1587
},
{
"epoch": 0.8229045213110506,
"grad_norm": 0.4994555711746216,
"learning_rate": 3.189799945179566e-06,
"loss": 0.0226,
"step": 1588
},
{
"epoch": 0.8234227231506672,
"grad_norm": 0.30759531259536743,
"learning_rate": 3.1716290765838664e-06,
"loss": 0.0173,
"step": 1589
},
{
"epoch": 0.8239409249902837,
"grad_norm": 0.22034572064876556,
"learning_rate": 3.1535056533916554e-06,
"loss": 0.0061,
"step": 1590
},
{
"epoch": 0.8244591268299002,
"grad_norm": 0.3770354390144348,
"learning_rate": 3.13542972669957e-06,
"loss": 0.0193,
"step": 1591
},
{
"epoch": 0.8249773286695168,
"grad_norm": 0.16963061690330505,
"learning_rate": 3.1174013474703255e-06,
"loss": 0.0065,
"step": 1592
},
{
"epoch": 0.8254955305091333,
"grad_norm": 0.16006392240524292,
"learning_rate": 3.0994205665326048e-06,
"loss": 0.0169,
"step": 1593
},
{
"epoch": 0.8260137323487499,
"grad_norm": 0.2839926481246948,
"learning_rate": 3.0814874345808677e-06,
"loss": 0.0151,
"step": 1594
},
{
"epoch": 0.8265319341883663,
"grad_norm": 0.24182650446891785,
"learning_rate": 3.063602002175261e-06,
"loss": 0.0152,
"step": 1595
},
{
"epoch": 0.8270501360279829,
"grad_norm": 0.291611909866333,
"learning_rate": 3.0457643197414244e-06,
"loss": 0.0181,
"step": 1596
},
{
"epoch": 0.8275683378675994,
"grad_norm": 0.1036018580198288,
"learning_rate": 3.027974437570389e-06,
"loss": 0.012,
"step": 1597
},
{
"epoch": 0.8280865397072159,
"grad_norm": 0.06551776081323624,
"learning_rate": 3.010232405818414e-06,
"loss": 0.002,
"step": 1598
},
{
"epoch": 0.8286047415468325,
"grad_norm": 0.5260184407234192,
"learning_rate": 2.9925382745068576e-06,
"loss": 0.026,
"step": 1599
},
{
"epoch": 0.829122943386449,
"grad_norm": 0.29389244318008423,
"learning_rate": 2.9748920935220083e-06,
"loss": 0.0122,
"step": 1600
},
{
"epoch": 0.8296411452260656,
"grad_norm": 0.2470783144235611,
"learning_rate": 2.957293912614987e-06,
"loss": 0.0063,
"step": 1601
},
{
"epoch": 0.8301593470656821,
"grad_norm": 0.2586616277694702,
"learning_rate": 2.939743781401576e-06,
"loss": 0.0073,
"step": 1602
},
{
"epoch": 0.8306775489052987,
"grad_norm": 0.17473287880420685,
"learning_rate": 2.9222417493620912e-06,
"loss": 0.0159,
"step": 1603
},
{
"epoch": 0.8311957507449151,
"grad_norm": 0.10318975895643234,
"learning_rate": 2.9047878658412274e-06,
"loss": 0.0047,
"step": 1604
},
{
"epoch": 0.8317139525845316,
"grad_norm": 0.1832498013973236,
"learning_rate": 2.887382180047946e-06,
"loss": 0.0069,
"step": 1605
},
{
"epoch": 0.8322321544241482,
"grad_norm": 0.5013145208358765,
"learning_rate": 2.8700247410553107e-06,
"loss": 0.0176,
"step": 1606
},
{
"epoch": 0.8327503562637647,
"grad_norm": 0.20341339707374573,
"learning_rate": 2.852715597800373e-06,
"loss": 0.0095,
"step": 1607
},
{
"epoch": 0.8332685581033813,
"grad_norm": 0.15497015416622162,
"learning_rate": 2.8354547990839963e-06,
"loss": 0.0035,
"step": 1608
},
{
"epoch": 0.8337867599429978,
"grad_norm": 0.19624440371990204,
"learning_rate": 2.8182423935707692e-06,
"loss": 0.0147,
"step": 1609
},
{
"epoch": 0.8343049617826144,
"grad_norm": 0.1570410281419754,
"learning_rate": 2.8010784297888305e-06,
"loss": 0.0096,
"step": 1610
},
{
"epoch": 0.8348231636222309,
"grad_norm": 0.34960222244262695,
"learning_rate": 2.7839629561297377e-06,
"loss": 0.0252,
"step": 1611
},
{
"epoch": 0.8353413654618473,
"grad_norm": 0.1798889935016632,
"learning_rate": 2.766896020848351e-06,
"loss": 0.0065,
"step": 1612
},
{
"epoch": 0.8358595673014639,
"grad_norm": 0.17661862075328827,
"learning_rate": 2.749877672062664e-06,
"loss": 0.0098,
"step": 1613
},
{
"epoch": 0.8363777691410804,
"grad_norm": 0.45938044786453247,
"learning_rate": 2.732907957753708e-06,
"loss": 0.0106,
"step": 1614
},
{
"epoch": 0.836895970980697,
"grad_norm": 0.18319405615329742,
"learning_rate": 2.715986925765386e-06,
"loss": 0.0092,
"step": 1615
},
{
"epoch": 0.8374141728203135,
"grad_norm": 0.2548009157180786,
"learning_rate": 2.69911462380434e-06,
"loss": 0.0113,
"step": 1616
},
{
"epoch": 0.8379323746599301,
"grad_norm": 0.27619224786758423,
"learning_rate": 2.6822910994398334e-06,
"loss": 0.0246,
"step": 1617
},
{
"epoch": 0.8384505764995466,
"grad_norm": 0.2859005331993103,
"learning_rate": 2.6655164001036073e-06,
"loss": 0.0123,
"step": 1618
},
{
"epoch": 0.8389687783391631,
"grad_norm": 0.28470805287361145,
"learning_rate": 2.6487905730897524e-06,
"loss": 0.0129,
"step": 1619
},
{
"epoch": 0.8394869801787797,
"grad_norm": 0.3091945946216583,
"learning_rate": 2.6321136655545475e-06,
"loss": 0.0112,
"step": 1620
},
{
"epoch": 0.8400051820183961,
"grad_norm": 0.26233717799186707,
"learning_rate": 2.6154857245163755e-06,
"loss": 0.014,
"step": 1621
},
{
"epoch": 0.8405233838580127,
"grad_norm": 0.255073606967926,
"learning_rate": 2.5989067968555514e-06,
"loss": 0.0133,
"step": 1622
},
{
"epoch": 0.8410415856976292,
"grad_norm": 0.07703150063753128,
"learning_rate": 2.5823769293142074e-06,
"loss": 0.0047,
"step": 1623
},
{
"epoch": 0.8415597875372458,
"grad_norm": 0.23450684547424316,
"learning_rate": 2.5658961684961537e-06,
"loss": 0.009,
"step": 1624
},
{
"epoch": 0.8420779893768623,
"grad_norm": 0.4751213788986206,
"learning_rate": 2.5494645608667503e-06,
"loss": 0.0219,
"step": 1625
},
{
"epoch": 0.8425961912164788,
"grad_norm": 0.2951497733592987,
"learning_rate": 2.5330821527527793e-06,
"loss": 0.0145,
"step": 1626
},
{
"epoch": 0.8431143930560954,
"grad_norm": 0.20706944167613983,
"learning_rate": 2.516748990342317e-06,
"loss": 0.0064,
"step": 1627
},
{
"epoch": 0.8436325948957119,
"grad_norm": 0.3090721666812897,
"learning_rate": 2.5004651196845786e-06,
"loss": 0.0122,
"step": 1628
},
{
"epoch": 0.8441507967353284,
"grad_norm": 0.5408398509025574,
"learning_rate": 2.484230586689833e-06,
"loss": 0.0273,
"step": 1629
},
{
"epoch": 0.8446689985749449,
"grad_norm": 0.12723809480667114,
"learning_rate": 2.468045437129223e-06,
"loss": 0.0088,
"step": 1630
},
{
"epoch": 0.8451872004145615,
"grad_norm": 0.18674789369106293,
"learning_rate": 2.4519097166346816e-06,
"loss": 0.0092,
"step": 1631
},
{
"epoch": 0.845705402254178,
"grad_norm": 0.2017349749803543,
"learning_rate": 2.435823470698768e-06,
"loss": 0.0109,
"step": 1632
},
{
"epoch": 0.8462236040937945,
"grad_norm": 0.17251279950141907,
"learning_rate": 2.419786744674564e-06,
"loss": 0.0195,
"step": 1633
},
{
"epoch": 0.8467418059334111,
"grad_norm": 0.28121671080589294,
"learning_rate": 2.4037995837755324e-06,
"loss": 0.0079,
"step": 1634
},
{
"epoch": 0.8472600077730276,
"grad_norm": 0.403934508562088,
"learning_rate": 2.387862033075399e-06,
"loss": 0.0136,
"step": 1635
},
{
"epoch": 0.8477782096126442,
"grad_norm": 0.28479158878326416,
"learning_rate": 2.3719741375080084e-06,
"loss": 0.0114,
"step": 1636
},
{
"epoch": 0.8482964114522606,
"grad_norm": 0.4967851936817169,
"learning_rate": 2.356135941867217e-06,
"loss": 0.0238,
"step": 1637
},
{
"epoch": 0.8488146132918772,
"grad_norm": 0.10027792304754257,
"learning_rate": 2.3403474908067624e-06,
"loss": 0.0028,
"step": 1638
},
{
"epoch": 0.8493328151314937,
"grad_norm": 0.20121395587921143,
"learning_rate": 2.3246088288401313e-06,
"loss": 0.0112,
"step": 1639
},
{
"epoch": 0.8498510169711102,
"grad_norm": 0.2108752280473709,
"learning_rate": 2.3089200003404265e-06,
"loss": 0.0071,
"step": 1640
},
{
"epoch": 0.8503692188107268,
"grad_norm": 0.1123914048075676,
"learning_rate": 2.2932810495402612e-06,
"loss": 0.0064,
"step": 1641
},
{
"epoch": 0.8508874206503433,
"grad_norm": 0.24224601686000824,
"learning_rate": 2.2776920205316276e-06,
"loss": 0.0244,
"step": 1642
},
{
"epoch": 0.8514056224899599,
"grad_norm": 0.40744614601135254,
"learning_rate": 2.2621529572657685e-06,
"loss": 0.0208,
"step": 1643
},
{
"epoch": 0.8519238243295764,
"grad_norm": 0.28944340348243713,
"learning_rate": 2.2466639035530435e-06,
"loss": 0.0105,
"step": 1644
},
{
"epoch": 0.8524420261691928,
"grad_norm": 0.19422157108783722,
"learning_rate": 2.2312249030628297e-06,
"loss": 0.0066,
"step": 1645
},
{
"epoch": 0.8529602280088094,
"grad_norm": 0.09577490389347076,
"learning_rate": 2.2158359993233814e-06,
"loss": 0.0051,
"step": 1646
},
{
"epoch": 0.8534784298484259,
"grad_norm": 0.16629934310913086,
"learning_rate": 2.2004972357217146e-06,
"loss": 0.0044,
"step": 1647
},
{
"epoch": 0.8539966316880425,
"grad_norm": 0.1245030090212822,
"learning_rate": 2.1852086555034747e-06,
"loss": 0.0055,
"step": 1648
},
{
"epoch": 0.854514833527659,
"grad_norm": 0.35298582911491394,
"learning_rate": 2.1699703017728237e-06,
"loss": 0.0174,
"step": 1649
},
{
"epoch": 0.8550330353672756,
"grad_norm": 0.17353323101997375,
"learning_rate": 2.1547822174923196e-06,
"loss": 0.0065,
"step": 1650
},
{
"epoch": 0.8555512372068921,
"grad_norm": 0.2672171890735626,
"learning_rate": 2.1396444454827956e-06,
"loss": 0.0131,
"step": 1651
},
{
"epoch": 0.8560694390465086,
"grad_norm": 0.1073671206831932,
"learning_rate": 2.1245570284232263e-06,
"loss": 0.0041,
"step": 1652
},
{
"epoch": 0.8565876408861252,
"grad_norm": 0.11136187613010406,
"learning_rate": 2.1095200088506227e-06,
"loss": 0.0143,
"step": 1653
},
{
"epoch": 0.8571058427257416,
"grad_norm": 0.15751762688159943,
"learning_rate": 2.0945334291599084e-06,
"loss": 0.0062,
"step": 1654
},
{
"epoch": 0.8576240445653582,
"grad_norm": 0.5879334211349487,
"learning_rate": 2.079597331603798e-06,
"loss": 0.0172,
"step": 1655
},
{
"epoch": 0.8581422464049747,
"grad_norm": 0.24367937445640564,
"learning_rate": 2.064711758292668e-06,
"loss": 0.0142,
"step": 1656
},
{
"epoch": 0.8586604482445913,
"grad_norm": 0.14427150785923004,
"learning_rate": 2.049876751194464e-06,
"loss": 0.0114,
"step": 1657
},
{
"epoch": 0.8591786500842078,
"grad_norm": 0.355831116437912,
"learning_rate": 2.035092352134558e-06,
"loss": 0.0218,
"step": 1658
},
{
"epoch": 0.8596968519238243,
"grad_norm": 0.24205715954303741,
"learning_rate": 2.0203586027956423e-06,
"loss": 0.0144,
"step": 1659
},
{
"epoch": 0.8602150537634409,
"grad_norm": 0.16865667700767517,
"learning_rate": 2.0056755447176003e-06,
"loss": 0.008,
"step": 1660
},
{
"epoch": 0.8607332556030574,
"grad_norm": 0.35851263999938965,
"learning_rate": 1.9910432192974105e-06,
"loss": 0.0191,
"step": 1661
},
{
"epoch": 0.861251457442674,
"grad_norm": 0.3120400905609131,
"learning_rate": 1.976461667789009e-06,
"loss": 0.0169,
"step": 1662
},
{
"epoch": 0.8617696592822904,
"grad_norm": 0.23192231357097626,
"learning_rate": 1.961930931303191e-06,
"loss": 0.013,
"step": 1663
},
{
"epoch": 0.862287861121907,
"grad_norm": 0.2518763840198517,
"learning_rate": 1.9474510508074694e-06,
"loss": 0.0118,
"step": 1664
},
{
"epoch": 0.8628060629615235,
"grad_norm": 0.17504754662513733,
"learning_rate": 1.933022067125987e-06,
"loss": 0.0097,
"step": 1665
},
{
"epoch": 0.86332426480114,
"grad_norm": 0.2840370535850525,
"learning_rate": 1.918644020939393e-06,
"loss": 0.0141,
"step": 1666
},
{
"epoch": 0.8638424666407566,
"grad_norm": 0.4161594808101654,
"learning_rate": 1.9043169527847106e-06,
"loss": 0.0228,
"step": 1667
},
{
"epoch": 0.8643606684803731,
"grad_norm": 0.216422900557518,
"learning_rate": 1.8900409030552502e-06,
"loss": 0.0195,
"step": 1668
},
{
"epoch": 0.8648788703199897,
"grad_norm": 0.27762821316719055,
"learning_rate": 1.8758159120004737e-06,
"loss": 0.0069,
"step": 1669
},
{
"epoch": 0.8653970721596062,
"grad_norm": 0.17240680754184723,
"learning_rate": 1.8616420197258977e-06,
"loss": 0.0051,
"step": 1670
},
{
"epoch": 0.8659152739992227,
"grad_norm": 0.23954899609088898,
"learning_rate": 1.8475192661929676e-06,
"loss": 0.0126,
"step": 1671
},
{
"epoch": 0.8664334758388392,
"grad_norm": 0.4306492507457733,
"learning_rate": 1.833447691218948e-06,
"loss": 0.0106,
"step": 1672
},
{
"epoch": 0.8669516776784557,
"grad_norm": 0.31067901849746704,
"learning_rate": 1.8194273344768153e-06,
"loss": 0.006,
"step": 1673
},
{
"epoch": 0.8674698795180723,
"grad_norm": 0.25931859016418457,
"learning_rate": 1.8054582354951434e-06,
"loss": 0.0171,
"step": 1674
},
{
"epoch": 0.8679880813576888,
"grad_norm": 0.27569490671157837,
"learning_rate": 1.7915404336579923e-06,
"loss": 0.0106,
"step": 1675
},
{
"epoch": 0.8685062831973054,
"grad_norm": 0.15181864798069,
"learning_rate": 1.7776739682047894e-06,
"loss": 0.0108,
"step": 1676
},
{
"epoch": 0.8690244850369219,
"grad_norm": 0.1656786948442459,
"learning_rate": 1.7638588782302336e-06,
"loss": 0.0075,
"step": 1677
},
{
"epoch": 0.8695426868765385,
"grad_norm": 0.2415611892938614,
"learning_rate": 1.7500952026841766e-06,
"loss": 0.0173,
"step": 1678
},
{
"epoch": 0.870060888716155,
"grad_norm": 0.13132847845554352,
"learning_rate": 1.7363829803715115e-06,
"loss": 0.0061,
"step": 1679
},
{
"epoch": 0.8705790905557714,
"grad_norm": 0.2118501365184784,
"learning_rate": 1.7227222499520601e-06,
"loss": 0.008,
"step": 1680
},
{
"epoch": 0.871097292395388,
"grad_norm": 0.25103017687797546,
"learning_rate": 1.7091130499404807e-06,
"loss": 0.0086,
"step": 1681
},
{
"epoch": 0.8716154942350045,
"grad_norm": 0.08915001153945923,
"learning_rate": 1.6955554187061406e-06,
"loss": 0.004,
"step": 1682
},
{
"epoch": 0.8721336960746211,
"grad_norm": 0.08841594308614731,
"learning_rate": 1.6820493944730266e-06,
"loss": 0.0032,
"step": 1683
},
{
"epoch": 0.8726518979142376,
"grad_norm": 0.40854230523109436,
"learning_rate": 1.668595015319603e-06,
"loss": 0.0191,
"step": 1684
},
{
"epoch": 0.8731700997538542,
"grad_norm": 0.27266111969947815,
"learning_rate": 1.6551923191787555e-06,
"loss": 0.0121,
"step": 1685
},
{
"epoch": 0.8736883015934707,
"grad_norm": 0.2558603584766388,
"learning_rate": 1.6418413438376335e-06,
"loss": 0.0098,
"step": 1686
},
{
"epoch": 0.8742065034330871,
"grad_norm": 0.14733348786830902,
"learning_rate": 1.6285421269375823e-06,
"loss": 0.0102,
"step": 1687
},
{
"epoch": 0.8747247052727037,
"grad_norm": 0.29257291555404663,
"learning_rate": 1.6152947059740087e-06,
"loss": 0.0175,
"step": 1688
},
{
"epoch": 0.8752429071123202,
"grad_norm": 0.40963825583457947,
"learning_rate": 1.6020991182962964e-06,
"loss": 0.0256,
"step": 1689
},
{
"epoch": 0.8757611089519368,
"grad_norm": 0.2824581265449524,
"learning_rate": 1.5889554011076901e-06,
"loss": 0.0146,
"step": 1690
},
{
"epoch": 0.8762793107915533,
"grad_norm": 0.28801703453063965,
"learning_rate": 1.5758635914651922e-06,
"loss": 0.016,
"step": 1691
},
{
"epoch": 0.8767975126311699,
"grad_norm": 0.17242257297039032,
"learning_rate": 1.5628237262794544e-06,
"loss": 0.0104,
"step": 1692
},
{
"epoch": 0.8773157144707864,
"grad_norm": 0.28166109323501587,
"learning_rate": 1.5498358423146797e-06,
"loss": 0.0292,
"step": 1693
},
{
"epoch": 0.8778339163104029,
"grad_norm": 0.19362136721611023,
"learning_rate": 1.5368999761885218e-06,
"loss": 0.0158,
"step": 1694
},
{
"epoch": 0.8783521181500195,
"grad_norm": 0.19558751583099365,
"learning_rate": 1.5240161643719753e-06,
"loss": 0.0064,
"step": 1695
},
{
"epoch": 0.8788703199896359,
"grad_norm": 0.14374202489852905,
"learning_rate": 1.5111844431892643e-06,
"loss": 0.0131,
"step": 1696
},
{
"epoch": 0.8793885218292525,
"grad_norm": 0.24548299610614777,
"learning_rate": 1.4984048488177628e-06,
"loss": 0.0218,
"step": 1697
},
{
"epoch": 0.879906723668869,
"grad_norm": 0.2641260325908661,
"learning_rate": 1.485677417287874e-06,
"loss": 0.0085,
"step": 1698
},
{
"epoch": 0.8804249255084855,
"grad_norm": 0.24745555222034454,
"learning_rate": 1.4730021844829411e-06,
"loss": 0.0098,
"step": 1699
},
{
"epoch": 0.8809431273481021,
"grad_norm": 0.3807303309440613,
"learning_rate": 1.460379186139127e-06,
"loss": 0.0263,
"step": 1700
},
{
"epoch": 0.8814613291877186,
"grad_norm": 0.3894922733306885,
"learning_rate": 1.4478084578453389e-06,
"loss": 0.0162,
"step": 1701
},
{
"epoch": 0.8819795310273352,
"grad_norm": 0.2722943127155304,
"learning_rate": 1.435290035043111e-06,
"loss": 0.0172,
"step": 1702
},
{
"epoch": 0.8824977328669517,
"grad_norm": 0.23959621787071228,
"learning_rate": 1.4228239530265042e-06,
"loss": 0.0105,
"step": 1703
},
{
"epoch": 0.8830159347065683,
"grad_norm": 0.20592434704303741,
"learning_rate": 1.4104102469420222e-06,
"loss": 0.0077,
"step": 1704
},
{
"epoch": 0.8835341365461847,
"grad_norm": 0.37758317589759827,
"learning_rate": 1.3980489517884865e-06,
"loss": 0.0138,
"step": 1705
},
{
"epoch": 0.8840523383858012,
"grad_norm": 0.047654855996370316,
"learning_rate": 1.3857401024169682e-06,
"loss": 0.0022,
"step": 1706
},
{
"epoch": 0.8845705402254178,
"grad_norm": 0.16144023835659027,
"learning_rate": 1.373483733530665e-06,
"loss": 0.0121,
"step": 1707
},
{
"epoch": 0.8850887420650343,
"grad_norm": 0.059519317001104355,
"learning_rate": 1.361279879684816e-06,
"loss": 0.0018,
"step": 1708
},
{
"epoch": 0.8856069439046509,
"grad_norm": 0.36740997433662415,
"learning_rate": 1.3491285752865956e-06,
"loss": 0.021,
"step": 1709
},
{
"epoch": 0.8861251457442674,
"grad_norm": 0.0914311334490776,
"learning_rate": 1.3370298545950334e-06,
"loss": 0.003,
"step": 1710
},
{
"epoch": 0.886643347583884,
"grad_norm": 0.5390928387641907,
"learning_rate": 1.3249837517208986e-06,
"loss": 0.0099,
"step": 1711
},
{
"epoch": 0.8871615494235005,
"grad_norm": 0.1984144002199173,
"learning_rate": 1.3129903006266065e-06,
"loss": 0.0057,
"step": 1712
},
{
"epoch": 0.8876797512631169,
"grad_norm": 0.19295454025268555,
"learning_rate": 1.3010495351261333e-06,
"loss": 0.0177,
"step": 1713
},
{
"epoch": 0.8881979531027335,
"grad_norm": 0.2037198692560196,
"learning_rate": 1.2891614888849202e-06,
"loss": 0.013,
"step": 1714
},
{
"epoch": 0.88871615494235,
"grad_norm": 0.2238919585943222,
"learning_rate": 1.2773261954197657e-06,
"loss": 0.0143,
"step": 1715
},
{
"epoch": 0.8892343567819666,
"grad_norm": 0.6479920148849487,
"learning_rate": 1.265543688098736e-06,
"loss": 0.0361,
"step": 1716
},
{
"epoch": 0.8897525586215831,
"grad_norm": 0.5114318132400513,
"learning_rate": 1.2538140001410826e-06,
"loss": 0.024,
"step": 1717
},
{
"epoch": 0.8902707604611997,
"grad_norm": 0.132893905043602,
"learning_rate": 1.2421371646171343e-06,
"loss": 0.0051,
"step": 1718
},
{
"epoch": 0.8907889623008162,
"grad_norm": 0.18512026965618134,
"learning_rate": 1.2305132144482123e-06,
"loss": 0.0117,
"step": 1719
},
{
"epoch": 0.8913071641404327,
"grad_norm": 0.20125305652618408,
"learning_rate": 1.2189421824065284e-06,
"loss": 0.0136,
"step": 1720
},
{
"epoch": 0.8918253659800492,
"grad_norm": 0.6059772372245789,
"learning_rate": 1.207424101115109e-06,
"loss": 0.0167,
"step": 1721
},
{
"epoch": 0.8923435678196657,
"grad_norm": 0.4089493155479431,
"learning_rate": 1.1959590030476798e-06,
"loss": 0.0134,
"step": 1722
},
{
"epoch": 0.8928617696592823,
"grad_norm": 0.31541603803634644,
"learning_rate": 1.1845469205285998e-06,
"loss": 0.0299,
"step": 1723
},
{
"epoch": 0.8933799714988988,
"grad_norm": 0.5524953603744507,
"learning_rate": 1.173187885732745e-06,
"loss": 0.0274,
"step": 1724
},
{
"epoch": 0.8938981733385154,
"grad_norm": 0.09349332749843597,
"learning_rate": 1.1618819306854402e-06,
"loss": 0.0079,
"step": 1725
},
{
"epoch": 0.8944163751781319,
"grad_norm": 0.25360822677612305,
"learning_rate": 1.150629087262356e-06,
"loss": 0.0072,
"step": 1726
},
{
"epoch": 0.8949345770177484,
"grad_norm": 0.36719512939453125,
"learning_rate": 1.1394293871894258e-06,
"loss": 0.0104,
"step": 1727
},
{
"epoch": 0.895452778857365,
"grad_norm": 0.17874731123447418,
"learning_rate": 1.1282828620427378e-06,
"loss": 0.0091,
"step": 1728
},
{
"epoch": 0.8959709806969814,
"grad_norm": 0.29093101620674133,
"learning_rate": 1.1171895432484758e-06,
"loss": 0.0199,
"step": 1729
},
{
"epoch": 0.896489182536598,
"grad_norm": 0.35000985860824585,
"learning_rate": 1.1061494620828084e-06,
"loss": 0.017,
"step": 1730
},
{
"epoch": 0.8970073843762145,
"grad_norm": 0.4346601068973541,
"learning_rate": 1.0951626496718126e-06,
"loss": 0.0314,
"step": 1731
},
{
"epoch": 0.8975255862158311,
"grad_norm": 0.21371477842330933,
"learning_rate": 1.0842291369913705e-06,
"loss": 0.0105,
"step": 1732
},
{
"epoch": 0.8980437880554476,
"grad_norm": 0.11501491814851761,
"learning_rate": 1.073348954867104e-06,
"loss": 0.0101,
"step": 1733
},
{
"epoch": 0.8985619898950641,
"grad_norm": 0.23166029155254364,
"learning_rate": 1.0625221339742708e-06,
"loss": 0.0103,
"step": 1734
},
{
"epoch": 0.8990801917346807,
"grad_norm": 0.3538880944252014,
"learning_rate": 1.0517487048376852e-06,
"loss": 0.0238,
"step": 1735
},
{
"epoch": 0.8995983935742972,
"grad_norm": 0.21843011677265167,
"learning_rate": 1.0410286978316276e-06,
"loss": 0.0203,
"step": 1736
},
{
"epoch": 0.9001165954139138,
"grad_norm": 0.14948545396327972,
"learning_rate": 1.0303621431797638e-06,
"loss": 0.0044,
"step": 1737
},
{
"epoch": 0.9006347972535302,
"grad_norm": 0.2392454892396927,
"learning_rate": 1.0197490709550606e-06,
"loss": 0.0094,
"step": 1738
},
{
"epoch": 0.9011529990931468,
"grad_norm": 0.06702300161123276,
"learning_rate": 1.0091895110796935e-06,
"loss": 0.0031,
"step": 1739
},
{
"epoch": 0.9016712009327633,
"grad_norm": 0.13145937025547028,
"learning_rate": 9.986834933249678e-07,
"loss": 0.01,
"step": 1740
},
{
"epoch": 0.9021894027723798,
"grad_norm": 0.237742617726326,
"learning_rate": 9.882310473112323e-07,
"loss": 0.0243,
"step": 1741
},
{
"epoch": 0.9027076046119964,
"grad_norm": 0.0943235382437706,
"learning_rate": 9.778322025078025e-07,
"loss": 0.0043,
"step": 1742
},
{
"epoch": 0.9032258064516129,
"grad_norm": 0.2087947428226471,
"learning_rate": 9.674869882328708e-07,
"loss": 0.0065,
"step": 1743
},
{
"epoch": 0.9037440082912295,
"grad_norm": 0.22540369629859924,
"learning_rate": 9.571954336534194e-07,
"loss": 0.007,
"step": 1744
},
{
"epoch": 0.904262210130846,
"grad_norm": 0.7248985171318054,
"learning_rate": 9.469575677851473e-07,
"loss": 0.0261,
"step": 1745
},
{
"epoch": 0.9047804119704626,
"grad_norm": 0.14132840931415558,
"learning_rate": 9.367734194923895e-07,
"loss": 0.008,
"step": 1746
},
{
"epoch": 0.905298613810079,
"grad_norm": 0.2634159028530121,
"learning_rate": 9.266430174880314e-07,
"loss": 0.0166,
"step": 1747
},
{
"epoch": 0.9058168156496955,
"grad_norm": 0.3727540373802185,
"learning_rate": 9.165663903334132e-07,
"loss": 0.0325,
"step": 1748
},
{
"epoch": 0.9063350174893121,
"grad_norm": 0.406827449798584,
"learning_rate": 9.065435664382827e-07,
"loss": 0.0264,
"step": 1749
},
{
"epoch": 0.9068532193289286,
"grad_norm": 0.2695000469684601,
"learning_rate": 8.965745740606868e-07,
"loss": 0.0103,
"step": 1750
},
{
"epoch": 0.9073714211685452,
"grad_norm": 0.3922761082649231,
"learning_rate": 8.866594413069074e-07,
"loss": 0.0165,
"step": 1751
},
{
"epoch": 0.9078896230081617,
"grad_norm": 0.23204711079597473,
"learning_rate": 8.767981961313632e-07,
"loss": 0.0069,
"step": 1752
},
{
"epoch": 0.9084078248477782,
"grad_norm": 0.2488652914762497,
"learning_rate": 8.669908663365567e-07,
"loss": 0.0122,
"step": 1753
},
{
"epoch": 0.9089260266873948,
"grad_norm": 0.13076753914356232,
"learning_rate": 8.57237479572981e-07,
"loss": 0.0096,
"step": 1754
},
{
"epoch": 0.9094442285270112,
"grad_norm": 0.2364853173494339,
"learning_rate": 8.475380633390417e-07,
"loss": 0.0125,
"step": 1755
},
{
"epoch": 0.9099624303666278,
"grad_norm": 0.23385170102119446,
"learning_rate": 8.378926449809799e-07,
"loss": 0.0139,
"step": 1756
},
{
"epoch": 0.9104806322062443,
"grad_norm": 0.18230576813220978,
"learning_rate": 8.283012516927979e-07,
"loss": 0.0054,
"step": 1757
},
{
"epoch": 0.9109988340458609,
"grad_norm": 0.2961260676383972,
"learning_rate": 8.187639105161805e-07,
"loss": 0.0204,
"step": 1758
},
{
"epoch": 0.9115170358854774,
"grad_norm": 0.2320140302181244,
"learning_rate": 8.09280648340427e-07,
"loss": 0.0103,
"step": 1759
},
{
"epoch": 0.9120352377250939,
"grad_norm": 0.3515540361404419,
"learning_rate": 7.998514919023547e-07,
"loss": 0.0222,
"step": 1760
},
{
"epoch": 0.9125534395647105,
"grad_norm": 0.460862398147583,
"learning_rate": 7.904764677862475e-07,
"loss": 0.0279,
"step": 1761
},
{
"epoch": 0.913071641404327,
"grad_norm": 0.3304995596408844,
"learning_rate": 7.811556024237644e-07,
"loss": 0.0221,
"step": 1762
},
{
"epoch": 0.9135898432439435,
"grad_norm": 0.49983087182044983,
"learning_rate": 7.718889220938796e-07,
"loss": 0.0214,
"step": 1763
},
{
"epoch": 0.91410804508356,
"grad_norm": 0.20241613686084747,
"learning_rate": 7.626764529227859e-07,
"loss": 0.0159,
"step": 1764
},
{
"epoch": 0.9146262469231766,
"grad_norm": 0.20853309333324432,
"learning_rate": 7.535182208838465e-07,
"loss": 0.0192,
"step": 1765
},
{
"epoch": 0.9151444487627931,
"grad_norm": 0.19687706232070923,
"learning_rate": 7.444142517975028e-07,
"loss": 0.0087,
"step": 1766
},
{
"epoch": 0.9156626506024096,
"grad_norm": 0.08084183931350708,
"learning_rate": 7.353645713312164e-07,
"loss": 0.0037,
"step": 1767
},
{
"epoch": 0.9161808524420262,
"grad_norm": 0.18536792695522308,
"learning_rate": 7.263692049993798e-07,
"loss": 0.0102,
"step": 1768
},
{
"epoch": 0.9166990542816427,
"grad_norm": 0.1274387687444687,
"learning_rate": 7.174281781632597e-07,
"loss": 0.0043,
"step": 1769
},
{
"epoch": 0.9172172561212593,
"grad_norm": 0.42621079087257385,
"learning_rate": 7.085415160309184e-07,
"loss": 0.0232,
"step": 1770
},
{
"epoch": 0.9177354579608757,
"grad_norm": 0.3298850357532501,
"learning_rate": 6.997092436571473e-07,
"loss": 0.013,
"step": 1771
},
{
"epoch": 0.9182536598004923,
"grad_norm": 0.2310313880443573,
"learning_rate": 6.909313859433874e-07,
"loss": 0.0125,
"step": 1772
},
{
"epoch": 0.9187718616401088,
"grad_norm": 0.36734509468078613,
"learning_rate": 6.822079676376647e-07,
"loss": 0.0155,
"step": 1773
},
{
"epoch": 0.9192900634797253,
"grad_norm": 0.15562260150909424,
"learning_rate": 6.735390133345299e-07,
"loss": 0.0063,
"step": 1774
},
{
"epoch": 0.9198082653193419,
"grad_norm": 0.11486683785915375,
"learning_rate": 6.649245474749632e-07,
"loss": 0.0081,
"step": 1775
},
{
"epoch": 0.9203264671589584,
"grad_norm": 0.13218870759010315,
"learning_rate": 6.563645943463392e-07,
"loss": 0.0078,
"step": 1776
},
{
"epoch": 0.920844668998575,
"grad_norm": 0.14428293704986572,
"learning_rate": 6.478591780823262e-07,
"loss": 0.008,
"step": 1777
},
{
"epoch": 0.9213628708381915,
"grad_norm": 0.48469075560569763,
"learning_rate": 6.394083226628422e-07,
"loss": 0.0217,
"step": 1778
},
{
"epoch": 0.9218810726778081,
"grad_norm": 0.2574191689491272,
"learning_rate": 6.310120519139773e-07,
"loss": 0.0155,
"step": 1779
},
{
"epoch": 0.9223992745174245,
"grad_norm": 0.09832003712654114,
"learning_rate": 6.226703895079201e-07,
"loss": 0.0041,
"step": 1780
},
{
"epoch": 0.922917476357041,
"grad_norm": 0.15627124905586243,
"learning_rate": 6.143833589629045e-07,
"loss": 0.0154,
"step": 1781
},
{
"epoch": 0.9234356781966576,
"grad_norm": 0.09480048716068268,
"learning_rate": 6.061509836431367e-07,
"loss": 0.0027,
"step": 1782
},
{
"epoch": 0.9239538800362741,
"grad_norm": 0.22857904434204102,
"learning_rate": 5.97973286758733e-07,
"loss": 0.01,
"step": 1783
},
{
"epoch": 0.9244720818758907,
"grad_norm": 0.1899561583995819,
"learning_rate": 5.898502913656368e-07,
"loss": 0.0068,
"step": 1784
},
{
"epoch": 0.9249902837155072,
"grad_norm": 0.1858498901128769,
"learning_rate": 5.817820203655822e-07,
"loss": 0.0084,
"step": 1785
},
{
"epoch": 0.9255084855551238,
"grad_norm": 0.32740896940231323,
"learning_rate": 5.737684965060109e-07,
"loss": 0.0177,
"step": 1786
},
{
"epoch": 0.9260266873947403,
"grad_norm": 0.12484689801931381,
"learning_rate": 5.658097423800124e-07,
"loss": 0.007,
"step": 1787
},
{
"epoch": 0.9265448892343567,
"grad_norm": 0.30967602133750916,
"learning_rate": 5.579057804262533e-07,
"loss": 0.0223,
"step": 1788
},
{
"epoch": 0.9270630910739733,
"grad_norm": 0.2253575623035431,
"learning_rate": 5.500566329289325e-07,
"loss": 0.0191,
"step": 1789
},
{
"epoch": 0.9275812929135898,
"grad_norm": 0.18831248581409454,
"learning_rate": 5.42262322017697e-07,
"loss": 0.0066,
"step": 1790
},
{
"epoch": 0.9280994947532064,
"grad_norm": 0.08481088280677795,
"learning_rate": 5.345228696675975e-07,
"loss": 0.0041,
"step": 1791
},
{
"epoch": 0.9286176965928229,
"grad_norm": 0.4872352182865143,
"learning_rate": 5.268382976990083e-07,
"loss": 0.0174,
"step": 1792
},
{
"epoch": 0.9291358984324395,
"grad_norm": 0.1179795041680336,
"learning_rate": 5.192086277775855e-07,
"loss": 0.0073,
"step": 1793
},
{
"epoch": 0.929654100272056,
"grad_norm": 0.14406712353229523,
"learning_rate": 5.116338814141842e-07,
"loss": 0.0062,
"step": 1794
},
{
"epoch": 0.9301723021116725,
"grad_norm": 0.33626288175582886,
"learning_rate": 5.041140799648237e-07,
"loss": 0.0171,
"step": 1795
},
{
"epoch": 0.930690503951289,
"grad_norm": 0.32739537954330444,
"learning_rate": 4.966492446306026e-07,
"loss": 0.0424,
"step": 1796
},
{
"epoch": 0.9312087057909055,
"grad_norm": 0.3170034885406494,
"learning_rate": 4.892393964576547e-07,
"loss": 0.0103,
"step": 1797
},
{
"epoch": 0.9317269076305221,
"grad_norm": 0.12314608693122864,
"learning_rate": 4.818845563370822e-07,
"loss": 0.0071,
"step": 1798
},
{
"epoch": 0.9322451094701386,
"grad_norm": 0.3412172794342041,
"learning_rate": 4.745847450049046e-07,
"loss": 0.0151,
"step": 1799
},
{
"epoch": 0.9327633113097551,
"grad_norm": 0.13909827172756195,
"learning_rate": 4.6733998304198334e-07,
"loss": 0.0133,
"step": 1800
},
{
"epoch": 0.9332815131493717,
"grad_norm": 0.19364029169082642,
"learning_rate": 4.6015029087398855e-07,
"loss": 0.013,
"step": 1801
},
{
"epoch": 0.9337997149889882,
"grad_norm": 0.42233768105506897,
"learning_rate": 4.5301568877132106e-07,
"loss": 0.0328,
"step": 1802
},
{
"epoch": 0.9343179168286048,
"grad_norm": 0.3611566126346588,
"learning_rate": 4.4593619684906387e-07,
"loss": 0.0189,
"step": 1803
},
{
"epoch": 0.9348361186682212,
"grad_norm": 0.18141718208789825,
"learning_rate": 4.389118350669241e-07,
"loss": 0.0082,
"step": 1804
},
{
"epoch": 0.9353543205078378,
"grad_norm": 0.1644567996263504,
"learning_rate": 4.3194262322917347e-07,
"loss": 0.0058,
"step": 1805
},
{
"epoch": 0.9358725223474543,
"grad_norm": 0.2701861262321472,
"learning_rate": 4.250285809846011e-07,
"loss": 0.0148,
"step": 1806
},
{
"epoch": 0.9363907241870708,
"grad_norm": 0.16967220604419708,
"learning_rate": 4.181697278264496e-07,
"loss": 0.0069,
"step": 1807
},
{
"epoch": 0.9369089260266874,
"grad_norm": 0.3081384003162384,
"learning_rate": 4.113660830923638e-07,
"loss": 0.014,
"step": 1808
},
{
"epoch": 0.9374271278663039,
"grad_norm": 0.23030611872673035,
"learning_rate": 4.046176659643353e-07,
"loss": 0.0144,
"step": 1809
},
{
"epoch": 0.9379453297059205,
"grad_norm": 0.3798378109931946,
"learning_rate": 3.9792449546864677e-07,
"loss": 0.0106,
"step": 1810
},
{
"epoch": 0.938463531545537,
"grad_norm": 0.15186521410942078,
"learning_rate": 3.912865904758256e-07,
"loss": 0.0064,
"step": 1811
},
{
"epoch": 0.9389817333851536,
"grad_norm": 0.2820347845554352,
"learning_rate": 3.8470396970057946e-07,
"loss": 0.0201,
"step": 1812
},
{
"epoch": 0.93949993522477,
"grad_norm": 0.1129571720957756,
"learning_rate": 3.781766517017493e-07,
"loss": 0.0036,
"step": 1813
},
{
"epoch": 0.9400181370643865,
"grad_norm": 0.16945257782936096,
"learning_rate": 3.717046548822634e-07,
"loss": 0.0122,
"step": 1814
},
{
"epoch": 0.9405363389040031,
"grad_norm": 0.2008204162120819,
"learning_rate": 3.652879974890766e-07,
"loss": 0.0114,
"step": 1815
},
{
"epoch": 0.9410545407436196,
"grad_norm": 0.11003681272268295,
"learning_rate": 3.589266976131134e-07,
"loss": 0.0025,
"step": 1816
},
{
"epoch": 0.9415727425832362,
"grad_norm": 0.19532270729541779,
"learning_rate": 3.5262077318923615e-07,
"loss": 0.0051,
"step": 1817
},
{
"epoch": 0.9420909444228527,
"grad_norm": 0.12835298478603363,
"learning_rate": 3.4637024199617896e-07,
"loss": 0.0047,
"step": 1818
},
{
"epoch": 0.9426091462624693,
"grad_norm": 0.19439566135406494,
"learning_rate": 3.401751216565008e-07,
"loss": 0.0064,
"step": 1819
},
{
"epoch": 0.9431273481020858,
"grad_norm": 0.2604013979434967,
"learning_rate": 3.3403542963653666e-07,
"loss": 0.0284,
"step": 1820
},
{
"epoch": 0.9436455499417022,
"grad_norm": 0.3678673207759857,
"learning_rate": 3.279511832463511e-07,
"loss": 0.0178,
"step": 1821
},
{
"epoch": 0.9441637517813188,
"grad_norm": 0.22065676748752594,
"learning_rate": 3.2192239963968477e-07,
"loss": 0.0092,
"step": 1822
},
{
"epoch": 0.9446819536209353,
"grad_norm": 0.16649556159973145,
"learning_rate": 3.159490958139122e-07,
"loss": 0.006,
"step": 1823
},
{
"epoch": 0.9452001554605519,
"grad_norm": 0.2621767222881317,
"learning_rate": 3.100312886099821e-07,
"loss": 0.0192,
"step": 1824
},
{
"epoch": 0.9457183573001684,
"grad_norm": 0.25440752506256104,
"learning_rate": 3.0416899471238606e-07,
"loss": 0.0088,
"step": 1825
},
{
"epoch": 0.946236559139785,
"grad_norm": 0.18357832729816437,
"learning_rate": 2.983622306490963e-07,
"loss": 0.015,
"step": 1826
},
{
"epoch": 0.9467547609794015,
"grad_norm": 0.2571721374988556,
"learning_rate": 2.926110127915327e-07,
"loss": 0.0101,
"step": 1827
},
{
"epoch": 0.947272962819018,
"grad_norm": 0.2212700992822647,
"learning_rate": 2.869153573545025e-07,
"loss": 0.0094,
"step": 1828
},
{
"epoch": 0.9477911646586346,
"grad_norm": 0.3824091851711273,
"learning_rate": 2.8127528039616935e-07,
"loss": 0.0184,
"step": 1829
},
{
"epoch": 0.948309366498251,
"grad_norm": 0.2520553171634674,
"learning_rate": 2.7569079781799124e-07,
"loss": 0.0191,
"step": 1830
},
{
"epoch": 0.9488275683378676,
"grad_norm": 0.07536312937736511,
"learning_rate": 2.701619253646959e-07,
"loss": 0.0022,
"step": 1831
},
{
"epoch": 0.9493457701774841,
"grad_norm": 0.17381742596626282,
"learning_rate": 2.6468867862421865e-07,
"loss": 0.0109,
"step": 1832
},
{
"epoch": 0.9498639720171007,
"grad_norm": 0.37904366850852966,
"learning_rate": 2.5927107302766483e-07,
"loss": 0.0234,
"step": 1833
},
{
"epoch": 0.9503821738567172,
"grad_norm": 0.23849669098854065,
"learning_rate": 2.539091238492719e-07,
"loss": 0.0166,
"step": 1834
},
{
"epoch": 0.9509003756963337,
"grad_norm": 0.256930410861969,
"learning_rate": 2.486028462063561e-07,
"loss": 0.0104,
"step": 1835
},
{
"epoch": 0.9514185775359503,
"grad_norm": 0.09233854711055756,
"learning_rate": 2.433522550592793e-07,
"loss": 0.0029,
"step": 1836
},
{
"epoch": 0.9519367793755668,
"grad_norm": 0.18194934725761414,
"learning_rate": 2.3815736521139775e-07,
"loss": 0.0089,
"step": 1837
},
{
"epoch": 0.9524549812151833,
"grad_norm": 0.28712162375450134,
"learning_rate": 2.3301819130903124e-07,
"loss": 0.0099,
"step": 1838
},
{
"epoch": 0.9529731830547998,
"grad_norm": 0.21452538669109344,
"learning_rate": 2.2793474784140957e-07,
"loss": 0.007,
"step": 1839
},
{
"epoch": 0.9534913848944164,
"grad_norm": 0.23924578726291656,
"learning_rate": 2.2290704914064155e-07,
"loss": 0.0217,
"step": 1840
},
{
"epoch": 0.9540095867340329,
"grad_norm": 0.3106108009815216,
"learning_rate": 2.179351093816706e-07,
"loss": 0.0186,
"step": 1841
},
{
"epoch": 0.9545277885736494,
"grad_norm": 0.5527723431587219,
"learning_rate": 2.1301894258223045e-07,
"loss": 0.0249,
"step": 1842
},
{
"epoch": 0.955045990413266,
"grad_norm": 0.1543256938457489,
"learning_rate": 2.081585626028204e-07,
"loss": 0.0097,
"step": 1843
},
{
"epoch": 0.9555641922528825,
"grad_norm": 0.27393466234207153,
"learning_rate": 2.0335398314664578e-07,
"loss": 0.0126,
"step": 1844
},
{
"epoch": 0.9560823940924991,
"grad_norm": 0.24850605428218842,
"learning_rate": 1.986052177595954e-07,
"loss": 0.0118,
"step": 1845
},
{
"epoch": 0.9566005959321155,
"grad_norm": 0.12293598055839539,
"learning_rate": 1.939122798301929e-07,
"loss": 0.0075,
"step": 1846
},
{
"epoch": 0.9571187977717321,
"grad_norm": 0.2503874599933624,
"learning_rate": 1.8927518258957e-07,
"loss": 0.0139,
"step": 1847
},
{
"epoch": 0.9576369996113486,
"grad_norm": 0.24323077499866486,
"learning_rate": 1.8469393911141553e-07,
"loss": 0.013,
"step": 1848
},
{
"epoch": 0.9581552014509651,
"grad_norm": 0.32238924503326416,
"learning_rate": 1.8016856231195533e-07,
"loss": 0.0116,
"step": 1849
},
{
"epoch": 0.9586734032905817,
"grad_norm": 0.3171192407608032,
"learning_rate": 1.7569906494989463e-07,
"loss": 0.011,
"step": 1850
},
{
"epoch": 0.9591916051301982,
"grad_norm": 0.2532232105731964,
"learning_rate": 1.7128545962640242e-07,
"loss": 0.0083,
"step": 1851
},
{
"epoch": 0.9597098069698148,
"grad_norm": 0.12934939563274384,
"learning_rate": 1.6692775878506705e-07,
"loss": 0.0061,
"step": 1852
},
{
"epoch": 0.9602280088094313,
"grad_norm": 0.5671677589416504,
"learning_rate": 1.6262597471185636e-07,
"loss": 0.0225,
"step": 1853
},
{
"epoch": 0.9607462106490477,
"grad_norm": 0.19938619434833527,
"learning_rate": 1.5838011953509313e-07,
"loss": 0.02,
"step": 1854
},
{
"epoch": 0.9612644124886643,
"grad_norm": 0.15334980189800262,
"learning_rate": 1.5419020522541295e-07,
"loss": 0.0237,
"step": 1855
},
{
"epoch": 0.9617826143282808,
"grad_norm": 0.15145401656627655,
"learning_rate": 1.5005624359573533e-07,
"loss": 0.0069,
"step": 1856
},
{
"epoch": 0.9623008161678974,
"grad_norm": 0.17450299859046936,
"learning_rate": 1.4597824630122604e-07,
"loss": 0.0127,
"step": 1857
},
{
"epoch": 0.9628190180075139,
"grad_norm": 0.2813020348548889,
"learning_rate": 1.4195622483926809e-07,
"loss": 0.0218,
"step": 1858
},
{
"epoch": 0.9633372198471305,
"grad_norm": 0.21633239090442657,
"learning_rate": 1.3799019054943073e-07,
"loss": 0.0058,
"step": 1859
},
{
"epoch": 0.963855421686747,
"grad_norm": 0.06448470056056976,
"learning_rate": 1.340801546134296e-07,
"loss": 0.003,
"step": 1860
},
{
"epoch": 0.9643736235263635,
"grad_norm": 0.1556619107723236,
"learning_rate": 1.3022612805510204e-07,
"loss": 0.0082,
"step": 1861
},
{
"epoch": 0.9648918253659801,
"grad_norm": 0.21246463060379028,
"learning_rate": 1.2642812174037621e-07,
"loss": 0.0071,
"step": 1862
},
{
"epoch": 0.9654100272055965,
"grad_norm": 0.24705225229263306,
"learning_rate": 1.2268614637723554e-07,
"loss": 0.0165,
"step": 1863
},
{
"epoch": 0.9659282290452131,
"grad_norm": 0.2733205258846283,
"learning_rate": 1.1900021251569415e-07,
"loss": 0.0194,
"step": 1864
},
{
"epoch": 0.9664464308848296,
"grad_norm": 0.07219294458627701,
"learning_rate": 1.153703305477638e-07,
"loss": 0.0054,
"step": 1865
},
{
"epoch": 0.9669646327244462,
"grad_norm": 0.08205874264240265,
"learning_rate": 1.117965107074248e-07,
"loss": 0.0068,
"step": 1866
},
{
"epoch": 0.9674828345640627,
"grad_norm": 0.2987622320652008,
"learning_rate": 1.0827876307059503e-07,
"loss": 0.0209,
"step": 1867
},
{
"epoch": 0.9680010364036792,
"grad_norm": 0.1716400384902954,
"learning_rate": 1.0481709755510994e-07,
"loss": 0.0062,
"step": 1868
},
{
"epoch": 0.9685192382432958,
"grad_norm": 0.2547082304954529,
"learning_rate": 1.0141152392068254e-07,
"loss": 0.0094,
"step": 1869
},
{
"epoch": 0.9690374400829123,
"grad_norm": 0.17818105220794678,
"learning_rate": 9.80620517688835e-08,
"loss": 0.0046,
"step": 1870
},
{
"epoch": 0.9695556419225289,
"grad_norm": 0.1513381004333496,
"learning_rate": 9.476869054311888e-08,
"loss": 0.0098,
"step": 1871
},
{
"epoch": 0.9700738437621453,
"grad_norm": 0.2405107617378235,
"learning_rate": 9.153144952859017e-08,
"loss": 0.0121,
"step": 1872
},
{
"epoch": 0.9705920456017619,
"grad_norm": 0.05375241860747337,
"learning_rate": 8.835033785227432e-08,
"loss": 0.0013,
"step": 1873
},
{
"epoch": 0.9711102474413784,
"grad_norm": 0.15731529891490936,
"learning_rate": 8.522536448290597e-08,
"loss": 0.0076,
"step": 1874
},
{
"epoch": 0.9716284492809949,
"grad_norm": 0.2520892322063446,
"learning_rate": 8.215653823093973e-08,
"loss": 0.0193,
"step": 1875
},
{
"epoch": 0.9721466511206115,
"grad_norm": 0.2123764306306839,
"learning_rate": 7.914386774853234e-08,
"loss": 0.0224,
"step": 1876
},
{
"epoch": 0.972664852960228,
"grad_norm": 0.5205470323562622,
"learning_rate": 7.618736152951611e-08,
"loss": 0.0177,
"step": 1877
},
{
"epoch": 0.9731830547998446,
"grad_norm": 0.45584774017333984,
"learning_rate": 7.32870279093767e-08,
"loss": 0.0305,
"step": 1878
},
{
"epoch": 0.973701256639461,
"grad_norm": 0.3030462861061096,
"learning_rate": 7.044287506522862e-08,
"loss": 0.0084,
"step": 1879
},
{
"epoch": 0.9742194584790776,
"grad_norm": 0.49225398898124695,
"learning_rate": 6.76549110157887e-08,
"loss": 0.012,
"step": 1880
},
{
"epoch": 0.9747376603186941,
"grad_norm": 0.40178224444389343,
"learning_rate": 6.49231436213582e-08,
"loss": 0.0227,
"step": 1881
},
{
"epoch": 0.9752558621583106,
"grad_norm": 0.30079546570777893,
"learning_rate": 6.224758058380298e-08,
"loss": 0.0146,
"step": 1882
},
{
"epoch": 0.9757740639979272,
"grad_norm": 0.15537235140800476,
"learning_rate": 5.962822944652224e-08,
"loss": 0.0087,
"step": 1883
},
{
"epoch": 0.9762922658375437,
"grad_norm": 0.27355560660362244,
"learning_rate": 5.706509759443979e-08,
"loss": 0.0107,
"step": 1884
},
{
"epoch": 0.9768104676771603,
"grad_norm": 0.1678086519241333,
"learning_rate": 5.455819225396841e-08,
"loss": 0.01,
"step": 1885
},
{
"epoch": 0.9773286695167768,
"grad_norm": 0.3235739767551422,
"learning_rate": 5.210752049300327e-08,
"loss": 0.0218,
"step": 1886
},
{
"epoch": 0.9778468713563934,
"grad_norm": 0.24950754642486572,
"learning_rate": 4.971308922089746e-08,
"loss": 0.0174,
"step": 1887
},
{
"epoch": 0.9783650731960098,
"grad_norm": 0.35669779777526855,
"learning_rate": 4.737490518843313e-08,
"loss": 0.0229,
"step": 1888
},
{
"epoch": 0.9788832750356263,
"grad_norm": 0.2362123280763626,
"learning_rate": 4.509297498781928e-08,
"loss": 0.014,
"step": 1889
},
{
"epoch": 0.9794014768752429,
"grad_norm": 0.11838959157466888,
"learning_rate": 4.28673050526629e-08,
"loss": 0.0057,
"step": 1890
},
{
"epoch": 0.9799196787148594,
"grad_norm": 0.21067118644714355,
"learning_rate": 4.069790165794896e-08,
"loss": 0.0083,
"step": 1891
},
{
"epoch": 0.980437880554476,
"grad_norm": 0.11189526319503784,
"learning_rate": 3.85847709200271e-08,
"loss": 0.0029,
"step": 1892
},
{
"epoch": 0.9809560823940925,
"grad_norm": 0.12419242411851883,
"learning_rate": 3.652791879659612e-08,
"loss": 0.0053,
"step": 1893
},
{
"epoch": 0.9814742842337091,
"grad_norm": 0.24143096804618835,
"learning_rate": 3.452735108668393e-08,
"loss": 0.011,
"step": 1894
},
{
"epoch": 0.9819924860733256,
"grad_norm": 0.24221214652061462,
"learning_rate": 3.258307343062983e-08,
"loss": 0.0096,
"step": 1895
},
{
"epoch": 0.982510687912942,
"grad_norm": 0.2091411054134369,
"learning_rate": 3.069509131007342e-08,
"loss": 0.0147,
"step": 1896
},
{
"epoch": 0.9830288897525586,
"grad_norm": 0.2547149062156677,
"learning_rate": 2.886341004793014e-08,
"loss": 0.0145,
"step": 1897
},
{
"epoch": 0.9835470915921751,
"grad_norm": 0.2162514328956604,
"learning_rate": 2.7088034808393503e-08,
"loss": 0.0198,
"step": 1898
},
{
"epoch": 0.9840652934317917,
"grad_norm": 0.4096549451351166,
"learning_rate": 2.536897059689958e-08,
"loss": 0.0157,
"step": 1899
},
{
"epoch": 0.9845834952714082,
"grad_norm": 0.14970317482948303,
"learning_rate": 2.3706222260126977e-08,
"loss": 0.0153,
"step": 1900
},
{
"epoch": 0.9851016971110248,
"grad_norm": 0.331853449344635,
"learning_rate": 2.2099794485976878e-08,
"loss": 0.0106,
"step": 1901
},
{
"epoch": 0.9856198989506413,
"grad_norm": 0.2141299843788147,
"learning_rate": 2.0549691803566362e-08,
"loss": 0.0243,
"step": 1902
},
{
"epoch": 0.9861381007902578,
"grad_norm": 0.1847657710313797,
"learning_rate": 1.905591858320621e-08,
"loss": 0.0043,
"step": 1903
},
{
"epoch": 0.9866563026298744,
"grad_norm": 0.2540552616119385,
"learning_rate": 1.7618479036396463e-08,
"loss": 0.0228,
"step": 1904
},
{
"epoch": 0.9871745044694908,
"grad_norm": 0.09786626696586609,
"learning_rate": 1.623737721581087e-08,
"loss": 0.0032,
"step": 1905
},
{
"epoch": 0.9876927063091074,
"grad_norm": 0.3893474042415619,
"learning_rate": 1.491261701528579e-08,
"loss": 0.0287,
"step": 1906
},
{
"epoch": 0.9882109081487239,
"grad_norm": 0.1743607223033905,
"learning_rate": 1.3644202169813547e-08,
"loss": 0.013,
"step": 1907
},
{
"epoch": 0.9887291099883404,
"grad_norm": 0.14568978548049927,
"learning_rate": 1.24321362555202e-08,
"loss": 0.0035,
"step": 1908
},
{
"epoch": 0.989247311827957,
"grad_norm": 0.24614860117435455,
"learning_rate": 1.1276422689672217e-08,
"loss": 0.0082,
"step": 1909
},
{
"epoch": 0.9897655136675735,
"grad_norm": 0.07558930665254593,
"learning_rate": 1.0177064730649832e-08,
"loss": 0.0029,
"step": 1910
},
{
"epoch": 0.9902837155071901,
"grad_norm": 0.30173760652542114,
"learning_rate": 9.134065477953701e-09,
"loss": 0.0146,
"step": 1911
},
{
"epoch": 0.9908019173468066,
"grad_norm": 0.13991938531398773,
"learning_rate": 8.147427872180481e-09,
"loss": 0.0077,
"step": 1912
},
{
"epoch": 0.9913201191864232,
"grad_norm": 0.15137632191181183,
"learning_rate": 7.217154695029482e-09,
"loss": 0.0076,
"step": 1913
},
{
"epoch": 0.9918383210260396,
"grad_norm": 0.3737245798110962,
"learning_rate": 6.343248569284921e-09,
"loss": 0.0188,
"step": 1914
},
{
"epoch": 0.9923565228656561,
"grad_norm": 0.1780516505241394,
"learning_rate": 5.525711958811464e-09,
"loss": 0.0216,
"step": 1915
},
{
"epoch": 0.9928747247052727,
"grad_norm": 0.1532387137413025,
"learning_rate": 4.7645471685453525e-09,
"loss": 0.0045,
"step": 1916
},
{
"epoch": 0.9933929265448892,
"grad_norm": 0.2650953233242035,
"learning_rate": 4.059756344492183e-09,
"loss": 0.0093,
"step": 1917
},
{
"epoch": 0.9939111283845058,
"grad_norm": 0.1893928050994873,
"learning_rate": 3.411341473720242e-09,
"loss": 0.0083,
"step": 1918
},
{
"epoch": 0.9944293302241223,
"grad_norm": 0.24139444530010223,
"learning_rate": 2.8193043843471878e-09,
"loss": 0.0169,
"step": 1919
},
{
"epoch": 0.9949475320637389,
"grad_norm": 0.36209264397621155,
"learning_rate": 2.2836467455489286e-09,
"loss": 0.0158,
"step": 1920
},
{
"epoch": 0.9954657339033554,
"grad_norm": 0.25236305594444275,
"learning_rate": 1.8043700675418607e-09,
"loss": 0.0163,
"step": 1921
},
{
"epoch": 0.9959839357429718,
"grad_norm": 0.1075766533613205,
"learning_rate": 1.381475701580648e-09,
"loss": 0.0045,
"step": 1922
},
{
"epoch": 0.9965021375825884,
"grad_norm": 0.062261324375867844,
"learning_rate": 1.0149648399648827e-09,
"loss": 0.0019,
"step": 1923
},
{
"epoch": 0.9970203394222049,
"grad_norm": 0.11437200754880905,
"learning_rate": 7.048385160213223e-10,
"loss": 0.0079,
"step": 1924
},
{
"epoch": 0.9975385412618215,
"grad_norm": 0.14774172008037567,
"learning_rate": 4.5109760411055084e-10,
"loss": 0.0091,
"step": 1925
},
{
"epoch": 0.998056743101438,
"grad_norm": 0.1296924203634262,
"learning_rate": 2.5374281962697866e-10,
"loss": 0.005,
"step": 1926
},
{
"epoch": 0.9985749449410546,
"grad_norm": 0.13076238334178925,
"learning_rate": 1.1277471898107905e-10,
"loss": 0.0075,
"step": 1927
},
{
"epoch": 0.9990931467806711,
"grad_norm": 0.2739886939525604,
"learning_rate": 2.8193699617151682e-11,
"loss": 0.012,
"step": 1928
},
{
"epoch": 0.9996113486202876,
"grad_norm": 0.28379353880882263,
"learning_rate": 0.0,
"loss": 0.0162,
"step": 1929
},
{
"epoch": 0.9996113486202876,
"step": 1929,
"total_flos": 2.4104116310493763e+21,
"train_loss": 0.02103021735445673,
"train_runtime": 5871.0379,
"train_samples_per_second": 5.259,
"train_steps_per_second": 0.329
}
],
"logging_steps": 1.0,
"max_steps": 1929,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 50,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2.4104116310493763e+21,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}