swegym_self_merged_scaled_44K / trainer_state.json
sedrickkeh's picture
End of training
f424798 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 1730,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.002890173410404624,
"grad_norm": 7.090742589535237,
"learning_rate": 1.1560693641618497e-07,
"loss": 1.1688,
"step": 1
},
{
"epoch": 0.005780346820809248,
"grad_norm": 6.961285906182364,
"learning_rate": 2.3121387283236994e-07,
"loss": 1.152,
"step": 2
},
{
"epoch": 0.008670520231213872,
"grad_norm": 6.803426809209352,
"learning_rate": 3.468208092485549e-07,
"loss": 1.1394,
"step": 3
},
{
"epoch": 0.011560693641618497,
"grad_norm": 6.908189824429038,
"learning_rate": 4.624277456647399e-07,
"loss": 1.1456,
"step": 4
},
{
"epoch": 0.014450867052023121,
"grad_norm": 6.896200809170114,
"learning_rate": 5.780346820809249e-07,
"loss": 1.1535,
"step": 5
},
{
"epoch": 0.017341040462427744,
"grad_norm": 6.805578082000228,
"learning_rate": 6.936416184971098e-07,
"loss": 1.1366,
"step": 6
},
{
"epoch": 0.02023121387283237,
"grad_norm": 6.703973957994694,
"learning_rate": 8.092485549132949e-07,
"loss": 1.1601,
"step": 7
},
{
"epoch": 0.023121387283236993,
"grad_norm": 6.208768024958665,
"learning_rate": 9.248554913294798e-07,
"loss": 1.0901,
"step": 8
},
{
"epoch": 0.02601156069364162,
"grad_norm": 6.0765892308895255,
"learning_rate": 1.040462427745665e-06,
"loss": 1.0838,
"step": 9
},
{
"epoch": 0.028901734104046242,
"grad_norm": 4.955277796409471,
"learning_rate": 1.1560693641618499e-06,
"loss": 1.0948,
"step": 10
},
{
"epoch": 0.031791907514450865,
"grad_norm": 4.694651793751496,
"learning_rate": 1.2716763005780348e-06,
"loss": 1.0693,
"step": 11
},
{
"epoch": 0.03468208092485549,
"grad_norm": 4.519565070186095,
"learning_rate": 1.3872832369942195e-06,
"loss": 1.0612,
"step": 12
},
{
"epoch": 0.03757225433526012,
"grad_norm": 4.197708724581753,
"learning_rate": 1.502890173410405e-06,
"loss": 1.071,
"step": 13
},
{
"epoch": 0.04046242774566474,
"grad_norm": 2.892414618232367,
"learning_rate": 1.6184971098265898e-06,
"loss": 1.0341,
"step": 14
},
{
"epoch": 0.04335260115606936,
"grad_norm": 2.8567945230902487,
"learning_rate": 1.7341040462427746e-06,
"loss": 1.0408,
"step": 15
},
{
"epoch": 0.046242774566473986,
"grad_norm": 2.728165100469004,
"learning_rate": 1.8497109826589595e-06,
"loss": 1.0101,
"step": 16
},
{
"epoch": 0.049132947976878616,
"grad_norm": 2.6771874838000933,
"learning_rate": 1.965317919075145e-06,
"loss": 0.9826,
"step": 17
},
{
"epoch": 0.05202312138728324,
"grad_norm": 2.1602714711793687,
"learning_rate": 2.08092485549133e-06,
"loss": 0.9661,
"step": 18
},
{
"epoch": 0.05491329479768786,
"grad_norm": 2.566414332647808,
"learning_rate": 2.1965317919075148e-06,
"loss": 0.9596,
"step": 19
},
{
"epoch": 0.057803468208092484,
"grad_norm": 2.551741054940349,
"learning_rate": 2.3121387283236997e-06,
"loss": 0.9482,
"step": 20
},
{
"epoch": 0.06069364161849711,
"grad_norm": 2.5386956864281904,
"learning_rate": 2.4277456647398847e-06,
"loss": 0.9411,
"step": 21
},
{
"epoch": 0.06358381502890173,
"grad_norm": 2.1389476290630047,
"learning_rate": 2.5433526011560696e-06,
"loss": 0.9384,
"step": 22
},
{
"epoch": 0.06647398843930635,
"grad_norm": 1.8798822954453593,
"learning_rate": 2.658959537572254e-06,
"loss": 0.9385,
"step": 23
},
{
"epoch": 0.06936416184971098,
"grad_norm": 1.6443784984730219,
"learning_rate": 2.774566473988439e-06,
"loss": 0.9346,
"step": 24
},
{
"epoch": 0.07225433526011561,
"grad_norm": 1.8190270220824762,
"learning_rate": 2.890173410404625e-06,
"loss": 0.8848,
"step": 25
},
{
"epoch": 0.07514450867052024,
"grad_norm": 2.0826997260801092,
"learning_rate": 3.00578034682081e-06,
"loss": 0.8725,
"step": 26
},
{
"epoch": 0.07803468208092486,
"grad_norm": 1.7656621345001091,
"learning_rate": 3.1213872832369948e-06,
"loss": 0.8621,
"step": 27
},
{
"epoch": 0.08092485549132948,
"grad_norm": 1.4044326746181084,
"learning_rate": 3.2369942196531797e-06,
"loss": 0.8399,
"step": 28
},
{
"epoch": 0.0838150289017341,
"grad_norm": 1.1511152676326295,
"learning_rate": 3.3526011560693642e-06,
"loss": 0.8481,
"step": 29
},
{
"epoch": 0.08670520231213873,
"grad_norm": 1.1197545053925908,
"learning_rate": 3.468208092485549e-06,
"loss": 0.8354,
"step": 30
},
{
"epoch": 0.08959537572254335,
"grad_norm": 1.254685408280926,
"learning_rate": 3.583815028901734e-06,
"loss": 0.8387,
"step": 31
},
{
"epoch": 0.09248554913294797,
"grad_norm": 1.3419102903202926,
"learning_rate": 3.699421965317919e-06,
"loss": 0.8525,
"step": 32
},
{
"epoch": 0.0953757225433526,
"grad_norm": 1.2271588730403455,
"learning_rate": 3.815028901734104e-06,
"loss": 0.851,
"step": 33
},
{
"epoch": 0.09826589595375723,
"grad_norm": 1.0577363822256058,
"learning_rate": 3.93063583815029e-06,
"loss": 0.8317,
"step": 34
},
{
"epoch": 0.10115606936416185,
"grad_norm": 1.0440985673514709,
"learning_rate": 4.046242774566474e-06,
"loss": 0.8313,
"step": 35
},
{
"epoch": 0.10404624277456648,
"grad_norm": 1.0098807937064038,
"learning_rate": 4.16184971098266e-06,
"loss": 0.8044,
"step": 36
},
{
"epoch": 0.1069364161849711,
"grad_norm": 0.8580054683034375,
"learning_rate": 4.277456647398844e-06,
"loss": 0.8237,
"step": 37
},
{
"epoch": 0.10982658959537572,
"grad_norm": 0.9796431941968685,
"learning_rate": 4.3930635838150296e-06,
"loss": 0.806,
"step": 38
},
{
"epoch": 0.11271676300578035,
"grad_norm": 0.9439424055584376,
"learning_rate": 4.508670520231214e-06,
"loss": 0.8056,
"step": 39
},
{
"epoch": 0.11560693641618497,
"grad_norm": 0.9203463890270979,
"learning_rate": 4.6242774566473994e-06,
"loss": 0.7844,
"step": 40
},
{
"epoch": 0.11849710982658959,
"grad_norm": 0.8780398374830074,
"learning_rate": 4.739884393063584e-06,
"loss": 0.8,
"step": 41
},
{
"epoch": 0.12138728323699421,
"grad_norm": 0.7825324337099361,
"learning_rate": 4.855491329479769e-06,
"loss": 0.7943,
"step": 42
},
{
"epoch": 0.12427745664739884,
"grad_norm": 0.8066246593788152,
"learning_rate": 4.971098265895954e-06,
"loss": 0.789,
"step": 43
},
{
"epoch": 0.12716763005780346,
"grad_norm": 0.8289688925666006,
"learning_rate": 5.086705202312139e-06,
"loss": 0.7821,
"step": 44
},
{
"epoch": 0.13005780346820808,
"grad_norm": 0.8808962397795624,
"learning_rate": 5.202312138728324e-06,
"loss": 0.7866,
"step": 45
},
{
"epoch": 0.1329479768786127,
"grad_norm": 0.79942280517295,
"learning_rate": 5.317919075144508e-06,
"loss": 0.7666,
"step": 46
},
{
"epoch": 0.13583815028901733,
"grad_norm": 0.724258148574199,
"learning_rate": 5.433526011560694e-06,
"loss": 0.7692,
"step": 47
},
{
"epoch": 0.13872832369942195,
"grad_norm": 0.7437020529632945,
"learning_rate": 5.549132947976878e-06,
"loss": 0.7589,
"step": 48
},
{
"epoch": 0.1416184971098266,
"grad_norm": 0.7700612704319088,
"learning_rate": 5.664739884393064e-06,
"loss": 0.7591,
"step": 49
},
{
"epoch": 0.14450867052023122,
"grad_norm": 0.7923035177832577,
"learning_rate": 5.78034682080925e-06,
"loss": 0.7683,
"step": 50
},
{
"epoch": 0.14739884393063585,
"grad_norm": 0.7088889817945515,
"learning_rate": 5.895953757225434e-06,
"loss": 0.7353,
"step": 51
},
{
"epoch": 0.15028901734104047,
"grad_norm": 0.676374698276162,
"learning_rate": 6.01156069364162e-06,
"loss": 0.7644,
"step": 52
},
{
"epoch": 0.1531791907514451,
"grad_norm": 0.6924940688568527,
"learning_rate": 6.127167630057804e-06,
"loss": 0.752,
"step": 53
},
{
"epoch": 0.15606936416184972,
"grad_norm": 0.8527665295088717,
"learning_rate": 6.2427745664739895e-06,
"loss": 0.7639,
"step": 54
},
{
"epoch": 0.15895953757225434,
"grad_norm": 0.6888695783169247,
"learning_rate": 6.358381502890174e-06,
"loss": 0.7509,
"step": 55
},
{
"epoch": 0.16184971098265896,
"grad_norm": 0.7490728100821497,
"learning_rate": 6.473988439306359e-06,
"loss": 0.7484,
"step": 56
},
{
"epoch": 0.16473988439306358,
"grad_norm": 0.9157288705931512,
"learning_rate": 6.589595375722544e-06,
"loss": 0.7408,
"step": 57
},
{
"epoch": 0.1676300578034682,
"grad_norm": 0.700876792326669,
"learning_rate": 6.7052023121387284e-06,
"loss": 0.7599,
"step": 58
},
{
"epoch": 0.17052023121387283,
"grad_norm": 0.7075983505256961,
"learning_rate": 6.820809248554914e-06,
"loss": 0.7253,
"step": 59
},
{
"epoch": 0.17341040462427745,
"grad_norm": 0.7662259009582316,
"learning_rate": 6.936416184971098e-06,
"loss": 0.7493,
"step": 60
},
{
"epoch": 0.17630057803468208,
"grad_norm": 0.7106648098423958,
"learning_rate": 7.052023121387284e-06,
"loss": 0.7491,
"step": 61
},
{
"epoch": 0.1791907514450867,
"grad_norm": 0.8234290971439716,
"learning_rate": 7.167630057803468e-06,
"loss": 0.7437,
"step": 62
},
{
"epoch": 0.18208092485549132,
"grad_norm": 0.7893881123491794,
"learning_rate": 7.283236994219654e-06,
"loss": 0.7516,
"step": 63
},
{
"epoch": 0.18497109826589594,
"grad_norm": 0.7820983140830451,
"learning_rate": 7.398843930635838e-06,
"loss": 0.7288,
"step": 64
},
{
"epoch": 0.18786127167630057,
"grad_norm": 0.8347449630731727,
"learning_rate": 7.5144508670520235e-06,
"loss": 0.7576,
"step": 65
},
{
"epoch": 0.1907514450867052,
"grad_norm": 0.6638382496430837,
"learning_rate": 7.630057803468209e-06,
"loss": 0.7289,
"step": 66
},
{
"epoch": 0.1936416184971098,
"grad_norm": 0.8106026549694,
"learning_rate": 7.745664739884393e-06,
"loss": 0.7267,
"step": 67
},
{
"epoch": 0.19653179190751446,
"grad_norm": 0.7500535606813568,
"learning_rate": 7.86127167630058e-06,
"loss": 0.7516,
"step": 68
},
{
"epoch": 0.1994219653179191,
"grad_norm": 0.669136491901921,
"learning_rate": 7.976878612716764e-06,
"loss": 0.7149,
"step": 69
},
{
"epoch": 0.2023121387283237,
"grad_norm": 0.7266271821143816,
"learning_rate": 8.092485549132949e-06,
"loss": 0.7229,
"step": 70
},
{
"epoch": 0.20520231213872833,
"grad_norm": 0.8328723885762708,
"learning_rate": 8.208092485549133e-06,
"loss": 0.7367,
"step": 71
},
{
"epoch": 0.20809248554913296,
"grad_norm": 0.6611812919714526,
"learning_rate": 8.32369942196532e-06,
"loss": 0.7394,
"step": 72
},
{
"epoch": 0.21098265895953758,
"grad_norm": 0.7232692335331032,
"learning_rate": 8.439306358381504e-06,
"loss": 0.7383,
"step": 73
},
{
"epoch": 0.2138728323699422,
"grad_norm": 0.7028412695126045,
"learning_rate": 8.554913294797688e-06,
"loss": 0.7222,
"step": 74
},
{
"epoch": 0.21676300578034682,
"grad_norm": 0.7274403497707242,
"learning_rate": 8.670520231213873e-06,
"loss": 0.7311,
"step": 75
},
{
"epoch": 0.21965317919075145,
"grad_norm": 0.6871391487835555,
"learning_rate": 8.786127167630059e-06,
"loss": 0.7182,
"step": 76
},
{
"epoch": 0.22254335260115607,
"grad_norm": 0.7418967633087836,
"learning_rate": 8.901734104046244e-06,
"loss": 0.7141,
"step": 77
},
{
"epoch": 0.2254335260115607,
"grad_norm": 0.7513431838317323,
"learning_rate": 9.017341040462428e-06,
"loss": 0.7138,
"step": 78
},
{
"epoch": 0.22832369942196531,
"grad_norm": 0.6956743141202641,
"learning_rate": 9.132947976878613e-06,
"loss": 0.7006,
"step": 79
},
{
"epoch": 0.23121387283236994,
"grad_norm": 0.6747351175950522,
"learning_rate": 9.248554913294799e-06,
"loss": 0.7094,
"step": 80
},
{
"epoch": 0.23410404624277456,
"grad_norm": 0.6938212711390582,
"learning_rate": 9.364161849710983e-06,
"loss": 0.7294,
"step": 81
},
{
"epoch": 0.23699421965317918,
"grad_norm": 0.7336859571024911,
"learning_rate": 9.479768786127168e-06,
"loss": 0.734,
"step": 82
},
{
"epoch": 0.2398843930635838,
"grad_norm": 0.6857197575458358,
"learning_rate": 9.595375722543352e-06,
"loss": 0.7166,
"step": 83
},
{
"epoch": 0.24277456647398843,
"grad_norm": 0.7287371485546202,
"learning_rate": 9.710982658959539e-06,
"loss": 0.7245,
"step": 84
},
{
"epoch": 0.24566473988439305,
"grad_norm": 0.7055914697217203,
"learning_rate": 9.826589595375723e-06,
"loss": 0.7045,
"step": 85
},
{
"epoch": 0.24855491329479767,
"grad_norm": 0.8230354709821103,
"learning_rate": 9.942196531791908e-06,
"loss": 0.7053,
"step": 86
},
{
"epoch": 0.2514450867052023,
"grad_norm": 0.7267067340623982,
"learning_rate": 1.0057803468208094e-05,
"loss": 0.6994,
"step": 87
},
{
"epoch": 0.2543352601156069,
"grad_norm": 0.7721448103371933,
"learning_rate": 1.0173410404624278e-05,
"loss": 0.694,
"step": 88
},
{
"epoch": 0.25722543352601157,
"grad_norm": 0.7325214157840806,
"learning_rate": 1.0289017341040463e-05,
"loss": 0.7096,
"step": 89
},
{
"epoch": 0.26011560693641617,
"grad_norm": 0.6829001987547941,
"learning_rate": 1.0404624277456647e-05,
"loss": 0.7114,
"step": 90
},
{
"epoch": 0.2630057803468208,
"grad_norm": 0.8080043630636083,
"learning_rate": 1.0520231213872834e-05,
"loss": 0.7306,
"step": 91
},
{
"epoch": 0.2658959537572254,
"grad_norm": 0.7590424295029534,
"learning_rate": 1.0635838150289017e-05,
"loss": 0.6958,
"step": 92
},
{
"epoch": 0.26878612716763006,
"grad_norm": 0.8001743955786633,
"learning_rate": 1.0751445086705203e-05,
"loss": 0.7386,
"step": 93
},
{
"epoch": 0.27167630057803466,
"grad_norm": 0.8387588767102132,
"learning_rate": 1.0867052023121387e-05,
"loss": 0.7083,
"step": 94
},
{
"epoch": 0.2745664739884393,
"grad_norm": 0.8688553902234525,
"learning_rate": 1.0982658959537573e-05,
"loss": 0.7364,
"step": 95
},
{
"epoch": 0.2774566473988439,
"grad_norm": 0.7464318068601791,
"learning_rate": 1.1098265895953756e-05,
"loss": 0.7149,
"step": 96
},
{
"epoch": 0.28034682080924855,
"grad_norm": 0.757546152793497,
"learning_rate": 1.1213872832369943e-05,
"loss": 0.6983,
"step": 97
},
{
"epoch": 0.2832369942196532,
"grad_norm": 0.7438008526685018,
"learning_rate": 1.1329479768786129e-05,
"loss": 0.7057,
"step": 98
},
{
"epoch": 0.2861271676300578,
"grad_norm": 0.8578457525033802,
"learning_rate": 1.1445086705202313e-05,
"loss": 0.7249,
"step": 99
},
{
"epoch": 0.28901734104046245,
"grad_norm": 0.746719183950257,
"learning_rate": 1.15606936416185e-05,
"loss": 0.711,
"step": 100
},
{
"epoch": 0.29190751445086704,
"grad_norm": 1.0078208261088115,
"learning_rate": 1.1676300578034682e-05,
"loss": 0.7094,
"step": 101
},
{
"epoch": 0.2947976878612717,
"grad_norm": 0.7362091784291579,
"learning_rate": 1.1791907514450869e-05,
"loss": 0.696,
"step": 102
},
{
"epoch": 0.2976878612716763,
"grad_norm": 0.9770080398344241,
"learning_rate": 1.1907514450867053e-05,
"loss": 0.7214,
"step": 103
},
{
"epoch": 0.30057803468208094,
"grad_norm": 0.7751464484631613,
"learning_rate": 1.202312138728324e-05,
"loss": 0.712,
"step": 104
},
{
"epoch": 0.30346820809248554,
"grad_norm": 0.9674784175128227,
"learning_rate": 1.2138728323699422e-05,
"loss": 0.7091,
"step": 105
},
{
"epoch": 0.3063583815028902,
"grad_norm": 0.8537449224492079,
"learning_rate": 1.2254335260115608e-05,
"loss": 0.6907,
"step": 106
},
{
"epoch": 0.3092485549132948,
"grad_norm": 0.9801860723729289,
"learning_rate": 1.2369942196531793e-05,
"loss": 0.7041,
"step": 107
},
{
"epoch": 0.31213872832369943,
"grad_norm": 0.9229373895628625,
"learning_rate": 1.2485549132947979e-05,
"loss": 0.7204,
"step": 108
},
{
"epoch": 0.315028901734104,
"grad_norm": 0.8638908853520371,
"learning_rate": 1.2601156069364162e-05,
"loss": 0.7009,
"step": 109
},
{
"epoch": 0.3179190751445087,
"grad_norm": 0.7773844197376821,
"learning_rate": 1.2716763005780348e-05,
"loss": 0.6945,
"step": 110
},
{
"epoch": 0.3208092485549133,
"grad_norm": 0.8125551782431788,
"learning_rate": 1.2832369942196533e-05,
"loss": 0.7291,
"step": 111
},
{
"epoch": 0.3236994219653179,
"grad_norm": 0.7481929293692252,
"learning_rate": 1.2947976878612719e-05,
"loss": 0.6884,
"step": 112
},
{
"epoch": 0.3265895953757225,
"grad_norm": 0.8853702590472872,
"learning_rate": 1.3063583815028902e-05,
"loss": 0.7021,
"step": 113
},
{
"epoch": 0.32947976878612717,
"grad_norm": 0.7260205643390303,
"learning_rate": 1.3179190751445088e-05,
"loss": 0.6719,
"step": 114
},
{
"epoch": 0.33236994219653176,
"grad_norm": 0.8978278445772442,
"learning_rate": 1.3294797687861272e-05,
"loss": 0.7254,
"step": 115
},
{
"epoch": 0.3352601156069364,
"grad_norm": 0.7958403914693805,
"learning_rate": 1.3410404624277457e-05,
"loss": 0.7031,
"step": 116
},
{
"epoch": 0.33815028901734107,
"grad_norm": 0.7737575346947156,
"learning_rate": 1.3526011560693643e-05,
"loss": 0.6987,
"step": 117
},
{
"epoch": 0.34104046242774566,
"grad_norm": 1.127441618739487,
"learning_rate": 1.3641618497109828e-05,
"loss": 0.6682,
"step": 118
},
{
"epoch": 0.3439306358381503,
"grad_norm": 0.8463337023953159,
"learning_rate": 1.3757225433526014e-05,
"loss": 0.6923,
"step": 119
},
{
"epoch": 0.3468208092485549,
"grad_norm": 0.9189231681321044,
"learning_rate": 1.3872832369942197e-05,
"loss": 0.7202,
"step": 120
},
{
"epoch": 0.34971098265895956,
"grad_norm": 0.7824948205964652,
"learning_rate": 1.3988439306358383e-05,
"loss": 0.6936,
"step": 121
},
{
"epoch": 0.35260115606936415,
"grad_norm": 1.0322221560229676,
"learning_rate": 1.4104046242774567e-05,
"loss": 0.6995,
"step": 122
},
{
"epoch": 0.3554913294797688,
"grad_norm": 0.8141648711576799,
"learning_rate": 1.4219653179190754e-05,
"loss": 0.6927,
"step": 123
},
{
"epoch": 0.3583815028901734,
"grad_norm": 0.8449673387537935,
"learning_rate": 1.4335260115606936e-05,
"loss": 0.7048,
"step": 124
},
{
"epoch": 0.36127167630057805,
"grad_norm": 0.9002017571970037,
"learning_rate": 1.4450867052023123e-05,
"loss": 0.6681,
"step": 125
},
{
"epoch": 0.36416184971098264,
"grad_norm": 0.7809488544709244,
"learning_rate": 1.4566473988439307e-05,
"loss": 0.707,
"step": 126
},
{
"epoch": 0.3670520231213873,
"grad_norm": 0.9600878348909891,
"learning_rate": 1.4682080924855493e-05,
"loss": 0.6886,
"step": 127
},
{
"epoch": 0.3699421965317919,
"grad_norm": 0.8728269186167633,
"learning_rate": 1.4797687861271676e-05,
"loss": 0.7056,
"step": 128
},
{
"epoch": 0.37283236994219654,
"grad_norm": 0.896265984559966,
"learning_rate": 1.4913294797687862e-05,
"loss": 0.691,
"step": 129
},
{
"epoch": 0.37572254335260113,
"grad_norm": 0.7956220341021001,
"learning_rate": 1.5028901734104047e-05,
"loss": 0.6827,
"step": 130
},
{
"epoch": 0.3786127167630058,
"grad_norm": 0.8059535895529943,
"learning_rate": 1.5144508670520233e-05,
"loss": 0.6998,
"step": 131
},
{
"epoch": 0.3815028901734104,
"grad_norm": 0.8563980216140917,
"learning_rate": 1.5260115606936418e-05,
"loss": 0.705,
"step": 132
},
{
"epoch": 0.38439306358381503,
"grad_norm": 0.9640073892946409,
"learning_rate": 1.5375722543352604e-05,
"loss": 0.7163,
"step": 133
},
{
"epoch": 0.3872832369942196,
"grad_norm": 0.9671494807729399,
"learning_rate": 1.5491329479768787e-05,
"loss": 0.6882,
"step": 134
},
{
"epoch": 0.3901734104046243,
"grad_norm": 0.9263949710988968,
"learning_rate": 1.5606936416184973e-05,
"loss": 0.7018,
"step": 135
},
{
"epoch": 0.3930635838150289,
"grad_norm": 0.8605012636577827,
"learning_rate": 1.572254335260116e-05,
"loss": 0.697,
"step": 136
},
{
"epoch": 0.3959537572254335,
"grad_norm": 0.967358301859874,
"learning_rate": 1.5838150289017342e-05,
"loss": 0.6914,
"step": 137
},
{
"epoch": 0.3988439306358382,
"grad_norm": 0.9342870144851555,
"learning_rate": 1.5953757225433528e-05,
"loss": 0.706,
"step": 138
},
{
"epoch": 0.40173410404624277,
"grad_norm": 0.8328745505321132,
"learning_rate": 1.606936416184971e-05,
"loss": 0.6902,
"step": 139
},
{
"epoch": 0.4046242774566474,
"grad_norm": 0.8241167968873063,
"learning_rate": 1.6184971098265897e-05,
"loss": 0.6893,
"step": 140
},
{
"epoch": 0.407514450867052,
"grad_norm": 0.7412040936941583,
"learning_rate": 1.6300578034682083e-05,
"loss": 0.6977,
"step": 141
},
{
"epoch": 0.41040462427745666,
"grad_norm": 0.8819984014049386,
"learning_rate": 1.6416184971098266e-05,
"loss": 0.6668,
"step": 142
},
{
"epoch": 0.41329479768786126,
"grad_norm": 0.8385796939119565,
"learning_rate": 1.6531791907514452e-05,
"loss": 0.7113,
"step": 143
},
{
"epoch": 0.4161849710982659,
"grad_norm": 0.8604832658397593,
"learning_rate": 1.664739884393064e-05,
"loss": 0.7026,
"step": 144
},
{
"epoch": 0.4190751445086705,
"grad_norm": 1.0461593299403098,
"learning_rate": 1.676300578034682e-05,
"loss": 0.6878,
"step": 145
},
{
"epoch": 0.42196531791907516,
"grad_norm": 0.811865171671389,
"learning_rate": 1.6878612716763008e-05,
"loss": 0.6796,
"step": 146
},
{
"epoch": 0.42485549132947975,
"grad_norm": 0.9481614894881139,
"learning_rate": 1.699421965317919e-05,
"loss": 0.7025,
"step": 147
},
{
"epoch": 0.4277456647398844,
"grad_norm": 0.8343305893997477,
"learning_rate": 1.7109826589595377e-05,
"loss": 0.6874,
"step": 148
},
{
"epoch": 0.430635838150289,
"grad_norm": 0.9421367720906404,
"learning_rate": 1.722543352601156e-05,
"loss": 0.6868,
"step": 149
},
{
"epoch": 0.43352601156069365,
"grad_norm": 0.8390255587190718,
"learning_rate": 1.7341040462427746e-05,
"loss": 0.6938,
"step": 150
},
{
"epoch": 0.43641618497109824,
"grad_norm": 0.8951037803089976,
"learning_rate": 1.7456647398843932e-05,
"loss": 0.6899,
"step": 151
},
{
"epoch": 0.4393063583815029,
"grad_norm": 0.7867290840778911,
"learning_rate": 1.7572254335260118e-05,
"loss": 0.7028,
"step": 152
},
{
"epoch": 0.4421965317919075,
"grad_norm": 0.995998766689234,
"learning_rate": 1.76878612716763e-05,
"loss": 0.7028,
"step": 153
},
{
"epoch": 0.44508670520231214,
"grad_norm": 0.9880944969572394,
"learning_rate": 1.7803468208092487e-05,
"loss": 0.6727,
"step": 154
},
{
"epoch": 0.4479768786127168,
"grad_norm": 1.284513930041935,
"learning_rate": 1.7919075144508673e-05,
"loss": 0.6922,
"step": 155
},
{
"epoch": 0.4508670520231214,
"grad_norm": 0.9511258625683117,
"learning_rate": 1.8034682080924856e-05,
"loss": 0.6903,
"step": 156
},
{
"epoch": 0.45375722543352603,
"grad_norm": 1.7388405496528023,
"learning_rate": 1.8150289017341043e-05,
"loss": 0.6819,
"step": 157
},
{
"epoch": 0.45664739884393063,
"grad_norm": 1.3452366087747083,
"learning_rate": 1.8265895953757225e-05,
"loss": 0.7148,
"step": 158
},
{
"epoch": 0.4595375722543353,
"grad_norm": 1.271970881953204,
"learning_rate": 1.838150289017341e-05,
"loss": 0.6929,
"step": 159
},
{
"epoch": 0.4624277456647399,
"grad_norm": 1.0994489504483136,
"learning_rate": 1.8497109826589598e-05,
"loss": 0.6782,
"step": 160
},
{
"epoch": 0.4653179190751445,
"grad_norm": 1.1657326994628787,
"learning_rate": 1.8612716763005784e-05,
"loss": 0.6887,
"step": 161
},
{
"epoch": 0.4682080924855491,
"grad_norm": 0.9854583993441037,
"learning_rate": 1.8728323699421967e-05,
"loss": 0.6839,
"step": 162
},
{
"epoch": 0.47109826589595377,
"grad_norm": 1.3286703766678492,
"learning_rate": 1.8843930635838153e-05,
"loss": 0.6842,
"step": 163
},
{
"epoch": 0.47398843930635837,
"grad_norm": 1.0185723644859102,
"learning_rate": 1.8959537572254336e-05,
"loss": 0.6988,
"step": 164
},
{
"epoch": 0.476878612716763,
"grad_norm": 1.1241175262812633,
"learning_rate": 1.9075144508670522e-05,
"loss": 0.6887,
"step": 165
},
{
"epoch": 0.4797687861271676,
"grad_norm": 0.9824433207331327,
"learning_rate": 1.9190751445086705e-05,
"loss": 0.6952,
"step": 166
},
{
"epoch": 0.48265895953757226,
"grad_norm": 1.1152282227331158,
"learning_rate": 1.930635838150289e-05,
"loss": 0.6602,
"step": 167
},
{
"epoch": 0.48554913294797686,
"grad_norm": 1.034043682955398,
"learning_rate": 1.9421965317919077e-05,
"loss": 0.696,
"step": 168
},
{
"epoch": 0.4884393063583815,
"grad_norm": 0.8508152573776165,
"learning_rate": 1.9537572254335264e-05,
"loss": 0.6959,
"step": 169
},
{
"epoch": 0.4913294797687861,
"grad_norm": 1.0495255954356906,
"learning_rate": 1.9653179190751446e-05,
"loss": 0.6751,
"step": 170
},
{
"epoch": 0.49421965317919075,
"grad_norm": 0.9582542481428932,
"learning_rate": 1.9768786127167633e-05,
"loss": 0.696,
"step": 171
},
{
"epoch": 0.49710982658959535,
"grad_norm": 1.0332331094000415,
"learning_rate": 1.9884393063583815e-05,
"loss": 0.7099,
"step": 172
},
{
"epoch": 0.5,
"grad_norm": 0.883896628507405,
"learning_rate": 2e-05,
"loss": 0.6689,
"step": 173
},
{
"epoch": 0.5028901734104047,
"grad_norm": 0.8240741765805385,
"learning_rate": 1.9999979644003047e-05,
"loss": 0.666,
"step": 174
},
{
"epoch": 0.5057803468208093,
"grad_norm": 0.9906783783216854,
"learning_rate": 1.9999918576095053e-05,
"loss": 0.7095,
"step": 175
},
{
"epoch": 0.5086705202312138,
"grad_norm": 0.8757602478752999,
"learning_rate": 1.9999816796524642e-05,
"loss": 0.6636,
"step": 176
},
{
"epoch": 0.5115606936416185,
"grad_norm": 0.9816660715471465,
"learning_rate": 1.9999674305706178e-05,
"loss": 0.6844,
"step": 177
},
{
"epoch": 0.5144508670520231,
"grad_norm": 0.9646033715264206,
"learning_rate": 1.999949110421977e-05,
"loss": 0.708,
"step": 178
},
{
"epoch": 0.5173410404624278,
"grad_norm": 1.054095791885763,
"learning_rate": 1.999926719281127e-05,
"loss": 0.7004,
"step": 179
},
{
"epoch": 0.5202312138728323,
"grad_norm": 1.1074464249197975,
"learning_rate": 1.9999002572392253e-05,
"loss": 0.6768,
"step": 180
},
{
"epoch": 0.523121387283237,
"grad_norm": 1.1267450710175495,
"learning_rate": 1.9998697244040062e-05,
"loss": 0.6909,
"step": 181
},
{
"epoch": 0.5260115606936416,
"grad_norm": 1.2160503421027011,
"learning_rate": 1.9998351208997734e-05,
"loss": 0.686,
"step": 182
},
{
"epoch": 0.5289017341040463,
"grad_norm": 1.2058488485981815,
"learning_rate": 1.9997964468674055e-05,
"loss": 0.6785,
"step": 183
},
{
"epoch": 0.5317919075144508,
"grad_norm": 1.4265640706372753,
"learning_rate": 1.999753702464352e-05,
"loss": 0.6852,
"step": 184
},
{
"epoch": 0.5346820809248555,
"grad_norm": 1.3345764329808851,
"learning_rate": 1.9997068878646332e-05,
"loss": 0.686,
"step": 185
},
{
"epoch": 0.5375722543352601,
"grad_norm": 1.119659843212298,
"learning_rate": 1.999656003258842e-05,
"loss": 0.6994,
"step": 186
},
{
"epoch": 0.5404624277456648,
"grad_norm": 1.4079713400525928,
"learning_rate": 1.9996010488541384e-05,
"loss": 0.6895,
"step": 187
},
{
"epoch": 0.5433526011560693,
"grad_norm": 1.066543075627955,
"learning_rate": 1.9995420248742534e-05,
"loss": 0.6929,
"step": 188
},
{
"epoch": 0.546242774566474,
"grad_norm": 1.1769931562935132,
"learning_rate": 1.9994789315594854e-05,
"loss": 0.6971,
"step": 189
},
{
"epoch": 0.5491329479768786,
"grad_norm": 1.0303545526204303,
"learning_rate": 1.9994117691667e-05,
"loss": 0.6864,
"step": 190
},
{
"epoch": 0.5520231213872833,
"grad_norm": 1.1897764422644492,
"learning_rate": 1.999340537969329e-05,
"loss": 0.6997,
"step": 191
},
{
"epoch": 0.5549132947976878,
"grad_norm": 1.0634156313689362,
"learning_rate": 1.9992652382573677e-05,
"loss": 0.6858,
"step": 192
},
{
"epoch": 0.5578034682080925,
"grad_norm": 1.2297353763416312,
"learning_rate": 1.999185870337377e-05,
"loss": 0.7016,
"step": 193
},
{
"epoch": 0.5606936416184971,
"grad_norm": 0.9673042975439411,
"learning_rate": 1.9991024345324793e-05,
"loss": 0.6755,
"step": 194
},
{
"epoch": 0.5635838150289018,
"grad_norm": 1.3338585851272673,
"learning_rate": 1.9990149311823586e-05,
"loss": 0.6709,
"step": 195
},
{
"epoch": 0.5664739884393064,
"grad_norm": 1.293864196693847,
"learning_rate": 1.9989233606432587e-05,
"loss": 0.6916,
"step": 196
},
{
"epoch": 0.569364161849711,
"grad_norm": 1.0355424413471204,
"learning_rate": 1.9988277232879806e-05,
"loss": 0.6607,
"step": 197
},
{
"epoch": 0.5722543352601156,
"grad_norm": 1.113803363521732,
"learning_rate": 1.998728019505884e-05,
"loss": 0.6693,
"step": 198
},
{
"epoch": 0.5751445086705202,
"grad_norm": 1.1763066495609689,
"learning_rate": 1.9986242497028823e-05,
"loss": 0.6801,
"step": 199
},
{
"epoch": 0.5780346820809249,
"grad_norm": 1.112264303614586,
"learning_rate": 1.9985164143014433e-05,
"loss": 0.6726,
"step": 200
},
{
"epoch": 0.5809248554913294,
"grad_norm": 1.3725438934255987,
"learning_rate": 1.9984045137405864e-05,
"loss": 0.6865,
"step": 201
},
{
"epoch": 0.5838150289017341,
"grad_norm": 1.0494739786278526,
"learning_rate": 1.9982885484758813e-05,
"loss": 0.6453,
"step": 202
},
{
"epoch": 0.5867052023121387,
"grad_norm": 0.9977419788918668,
"learning_rate": 1.9981685189794457e-05,
"loss": 0.6843,
"step": 203
},
{
"epoch": 0.5895953757225434,
"grad_norm": 1.0197936450519736,
"learning_rate": 1.9980444257399428e-05,
"loss": 0.6981,
"step": 204
},
{
"epoch": 0.5924855491329479,
"grad_norm": 0.8579217853259621,
"learning_rate": 1.9979162692625817e-05,
"loss": 0.6855,
"step": 205
},
{
"epoch": 0.5953757225433526,
"grad_norm": 1.0477044603584735,
"learning_rate": 1.9977840500691133e-05,
"loss": 0.6869,
"step": 206
},
{
"epoch": 0.5982658959537572,
"grad_norm": 0.9027615423442625,
"learning_rate": 1.9976477686978274e-05,
"loss": 0.6627,
"step": 207
},
{
"epoch": 0.6011560693641619,
"grad_norm": 1.1728694608856245,
"learning_rate": 1.9975074257035533e-05,
"loss": 0.6831,
"step": 208
},
{
"epoch": 0.6040462427745664,
"grad_norm": 0.830419992847188,
"learning_rate": 1.9973630216576547e-05,
"loss": 0.6689,
"step": 209
},
{
"epoch": 0.6069364161849711,
"grad_norm": 1.0672356976307946,
"learning_rate": 1.9972145571480296e-05,
"loss": 0.6975,
"step": 210
},
{
"epoch": 0.6098265895953757,
"grad_norm": 0.9508554177191132,
"learning_rate": 1.997062032779107e-05,
"loss": 0.6528,
"step": 211
},
{
"epoch": 0.6127167630057804,
"grad_norm": 0.7800609305235056,
"learning_rate": 1.9969054491718437e-05,
"loss": 0.6849,
"step": 212
},
{
"epoch": 0.615606936416185,
"grad_norm": 1.0095266619422791,
"learning_rate": 1.9967448069637224e-05,
"loss": 0.6746,
"step": 213
},
{
"epoch": 0.6184971098265896,
"grad_norm": 0.8623507582546789,
"learning_rate": 1.9965801068087504e-05,
"loss": 0.6556,
"step": 214
},
{
"epoch": 0.6213872832369942,
"grad_norm": 0.8697052589621063,
"learning_rate": 1.996411349377454e-05,
"loss": 0.6805,
"step": 215
},
{
"epoch": 0.6242774566473989,
"grad_norm": 0.9633918268862216,
"learning_rate": 1.9962385353568787e-05,
"loss": 0.6675,
"step": 216
},
{
"epoch": 0.6271676300578035,
"grad_norm": 1.0957605776766808,
"learning_rate": 1.9960616654505852e-05,
"loss": 0.6406,
"step": 217
},
{
"epoch": 0.630057803468208,
"grad_norm": 0.9991877306193689,
"learning_rate": 1.9958807403786452e-05,
"loss": 0.6661,
"step": 218
},
{
"epoch": 0.6329479768786127,
"grad_norm": 0.8868819421389161,
"learning_rate": 1.995695760877642e-05,
"loss": 0.6556,
"step": 219
},
{
"epoch": 0.6358381502890174,
"grad_norm": 0.8810851372838674,
"learning_rate": 1.9955067277006633e-05,
"loss": 0.6687,
"step": 220
},
{
"epoch": 0.638728323699422,
"grad_norm": 0.8373803462646919,
"learning_rate": 1.9953136416173005e-05,
"loss": 0.6872,
"step": 221
},
{
"epoch": 0.6416184971098265,
"grad_norm": 0.8240195012502505,
"learning_rate": 1.9951165034136465e-05,
"loss": 0.6606,
"step": 222
},
{
"epoch": 0.6445086705202312,
"grad_norm": 0.8497327904586488,
"learning_rate": 1.9949153138922897e-05,
"loss": 0.6727,
"step": 223
},
{
"epoch": 0.6473988439306358,
"grad_norm": 0.9574664025504449,
"learning_rate": 1.9947100738723126e-05,
"loss": 0.6903,
"step": 224
},
{
"epoch": 0.6502890173410405,
"grad_norm": 0.7959314545389815,
"learning_rate": 1.9945007841892886e-05,
"loss": 0.648,
"step": 225
},
{
"epoch": 0.653179190751445,
"grad_norm": 0.8327210631747088,
"learning_rate": 1.994287445695277e-05,
"loss": 0.6723,
"step": 226
},
{
"epoch": 0.6560693641618497,
"grad_norm": 0.8491750263082852,
"learning_rate": 1.9940700592588228e-05,
"loss": 0.6951,
"step": 227
},
{
"epoch": 0.6589595375722543,
"grad_norm": 0.9783543443859294,
"learning_rate": 1.9938486257649483e-05,
"loss": 0.6521,
"step": 228
},
{
"epoch": 0.661849710982659,
"grad_norm": 0.8848987100308845,
"learning_rate": 1.9936231461151535e-05,
"loss": 0.6786,
"step": 229
},
{
"epoch": 0.6647398843930635,
"grad_norm": 0.880197695167373,
"learning_rate": 1.9933936212274115e-05,
"loss": 0.6619,
"step": 230
},
{
"epoch": 0.6676300578034682,
"grad_norm": 1.0154012991829307,
"learning_rate": 1.9931600520361637e-05,
"loss": 0.6549,
"step": 231
},
{
"epoch": 0.6705202312138728,
"grad_norm": 0.9283345521485898,
"learning_rate": 1.9929224394923167e-05,
"loss": 0.6781,
"step": 232
},
{
"epoch": 0.6734104046242775,
"grad_norm": 1.015308580501105,
"learning_rate": 1.9926807845632385e-05,
"loss": 0.6814,
"step": 233
},
{
"epoch": 0.6763005780346821,
"grad_norm": 0.9736049842517275,
"learning_rate": 1.9924350882327545e-05,
"loss": 0.6834,
"step": 234
},
{
"epoch": 0.6791907514450867,
"grad_norm": 0.9968601538853321,
"learning_rate": 1.992185351501144e-05,
"loss": 0.6735,
"step": 235
},
{
"epoch": 0.6820809248554913,
"grad_norm": 0.9532926457310584,
"learning_rate": 1.9919315753851343e-05,
"loss": 0.6593,
"step": 236
},
{
"epoch": 0.684971098265896,
"grad_norm": 1.2870145989902115,
"learning_rate": 1.991673760917899e-05,
"loss": 0.67,
"step": 237
},
{
"epoch": 0.6878612716763006,
"grad_norm": 1.1226345328773792,
"learning_rate": 1.991411909149052e-05,
"loss": 0.6785,
"step": 238
},
{
"epoch": 0.6907514450867052,
"grad_norm": 1.204681411549356,
"learning_rate": 1.9911460211446445e-05,
"loss": 0.6795,
"step": 239
},
{
"epoch": 0.6936416184971098,
"grad_norm": 1.1282608884854723,
"learning_rate": 1.990876097987159e-05,
"loss": 0.6641,
"step": 240
},
{
"epoch": 0.6965317919075145,
"grad_norm": 1.2759929528296277,
"learning_rate": 1.9906021407755064e-05,
"loss": 0.6738,
"step": 241
},
{
"epoch": 0.6994219653179191,
"grad_norm": 1.0879712951989373,
"learning_rate": 1.990324150625022e-05,
"loss": 0.6688,
"step": 242
},
{
"epoch": 0.7023121387283237,
"grad_norm": 1.4403079444759717,
"learning_rate": 1.9900421286674582e-05,
"loss": 0.6769,
"step": 243
},
{
"epoch": 0.7052023121387283,
"grad_norm": 1.191228603950231,
"learning_rate": 1.9897560760509834e-05,
"loss": 0.667,
"step": 244
},
{
"epoch": 0.708092485549133,
"grad_norm": 1.275787843674223,
"learning_rate": 1.989465993940174e-05,
"loss": 0.6734,
"step": 245
},
{
"epoch": 0.7109826589595376,
"grad_norm": 1.1457956849938589,
"learning_rate": 1.9891718835160128e-05,
"loss": 0.662,
"step": 246
},
{
"epoch": 0.7138728323699421,
"grad_norm": 1.2858432685195738,
"learning_rate": 1.9888737459758817e-05,
"loss": 0.6875,
"step": 247
},
{
"epoch": 0.7167630057803468,
"grad_norm": 0.9398046224404962,
"learning_rate": 1.9885715825335584e-05,
"loss": 0.6548,
"step": 248
},
{
"epoch": 0.7196531791907514,
"grad_norm": 1.6358172197686662,
"learning_rate": 1.9882653944192095e-05,
"loss": 0.6954,
"step": 249
},
{
"epoch": 0.7225433526011561,
"grad_norm": 0.9961704138959488,
"learning_rate": 1.9879551828793893e-05,
"loss": 0.6615,
"step": 250
},
{
"epoch": 0.7254335260115607,
"grad_norm": 2.0842179603061424,
"learning_rate": 1.98764094917703e-05,
"loss": 0.6884,
"step": 251
},
{
"epoch": 0.7283236994219653,
"grad_norm": 2.330063826601232,
"learning_rate": 1.9873226945914394e-05,
"loss": 0.6705,
"step": 252
},
{
"epoch": 0.7312138728323699,
"grad_norm": 0.8889229347289838,
"learning_rate": 1.9870004204182962e-05,
"loss": 0.6749,
"step": 253
},
{
"epoch": 0.7341040462427746,
"grad_norm": 2.0009447904173965,
"learning_rate": 1.986674127969642e-05,
"loss": 0.6542,
"step": 254
},
{
"epoch": 0.7369942196531792,
"grad_norm": 1.8777338758058673,
"learning_rate": 1.986343818573879e-05,
"loss": 0.6766,
"step": 255
},
{
"epoch": 0.7398843930635838,
"grad_norm": 1.2106551936340486,
"learning_rate": 1.986009493575763e-05,
"loss": 0.6674,
"step": 256
},
{
"epoch": 0.7427745664739884,
"grad_norm": 1.576538210921238,
"learning_rate": 1.985671154336396e-05,
"loss": 0.6727,
"step": 257
},
{
"epoch": 0.7456647398843931,
"grad_norm": 0.8573792365014181,
"learning_rate": 1.9853288022332267e-05,
"loss": 0.669,
"step": 258
},
{
"epoch": 0.7485549132947977,
"grad_norm": 1.8092997648187275,
"learning_rate": 1.9849824386600372e-05,
"loss": 0.6675,
"step": 259
},
{
"epoch": 0.7514450867052023,
"grad_norm": 1.5944489420897012,
"learning_rate": 1.984632065026943e-05,
"loss": 0.675,
"step": 260
},
{
"epoch": 0.7543352601156069,
"grad_norm": 0.9900937681751832,
"learning_rate": 1.9842776827603858e-05,
"loss": 0.6704,
"step": 261
},
{
"epoch": 0.7572254335260116,
"grad_norm": 1.5977627858005359,
"learning_rate": 1.9839192933031253e-05,
"loss": 0.6922,
"step": 262
},
{
"epoch": 0.7601156069364162,
"grad_norm": 1.273943078477883,
"learning_rate": 1.9835568981142376e-05,
"loss": 0.6667,
"step": 263
},
{
"epoch": 0.7630057803468208,
"grad_norm": 0.951797646828311,
"learning_rate": 1.9831904986691045e-05,
"loss": 0.6702,
"step": 264
},
{
"epoch": 0.7658959537572254,
"grad_norm": 1.4628752552787214,
"learning_rate": 1.9828200964594125e-05,
"loss": 0.6732,
"step": 265
},
{
"epoch": 0.7687861271676301,
"grad_norm": 1.0303268490984354,
"learning_rate": 1.9824456929931417e-05,
"loss": 0.6823,
"step": 266
},
{
"epoch": 0.7716763005780347,
"grad_norm": 1.0447939758438316,
"learning_rate": 1.982067289794564e-05,
"loss": 0.6739,
"step": 267
},
{
"epoch": 0.7745664739884393,
"grad_norm": 0.9399400681706109,
"learning_rate": 1.9816848884042342e-05,
"loss": 0.6619,
"step": 268
},
{
"epoch": 0.7774566473988439,
"grad_norm": 1.001440688633843,
"learning_rate": 1.9812984903789845e-05,
"loss": 0.6618,
"step": 269
},
{
"epoch": 0.7803468208092486,
"grad_norm": 0.8766732359528921,
"learning_rate": 1.9809080972919182e-05,
"loss": 0.6831,
"step": 270
},
{
"epoch": 0.7832369942196532,
"grad_norm": 1.059517462713445,
"learning_rate": 1.9805137107324034e-05,
"loss": 0.6717,
"step": 271
},
{
"epoch": 0.7861271676300579,
"grad_norm": 0.9174226778841795,
"learning_rate": 1.9801153323060667e-05,
"loss": 0.6666,
"step": 272
},
{
"epoch": 0.7890173410404624,
"grad_norm": 0.9500149856474327,
"learning_rate": 1.9797129636347857e-05,
"loss": 0.6756,
"step": 273
},
{
"epoch": 0.791907514450867,
"grad_norm": 1.1277366658505596,
"learning_rate": 1.979306606356684e-05,
"loss": 0.6678,
"step": 274
},
{
"epoch": 0.7947976878612717,
"grad_norm": 0.8597685475695553,
"learning_rate": 1.9788962621261224e-05,
"loss": 0.6746,
"step": 275
},
{
"epoch": 0.7976878612716763,
"grad_norm": 0.9175327448204571,
"learning_rate": 1.9784819326136947e-05,
"loss": 0.6663,
"step": 276
},
{
"epoch": 0.8005780346820809,
"grad_norm": 0.8783811136618952,
"learning_rate": 1.978063619506219e-05,
"loss": 0.6849,
"step": 277
},
{
"epoch": 0.8034682080924855,
"grad_norm": 0.9725568607677347,
"learning_rate": 1.9776413245067308e-05,
"loss": 0.6833,
"step": 278
},
{
"epoch": 0.8063583815028902,
"grad_norm": 1.0920362225909976,
"learning_rate": 1.9772150493344776e-05,
"loss": 0.6777,
"step": 279
},
{
"epoch": 0.8092485549132948,
"grad_norm": 0.8283527251152283,
"learning_rate": 1.9767847957249107e-05,
"loss": 0.6652,
"step": 280
},
{
"epoch": 0.8121387283236994,
"grad_norm": 0.9586230775456167,
"learning_rate": 1.9763505654296782e-05,
"loss": 0.6551,
"step": 281
},
{
"epoch": 0.815028901734104,
"grad_norm": 0.9276724961475561,
"learning_rate": 1.9759123602166183e-05,
"loss": 0.6629,
"step": 282
},
{
"epoch": 0.8179190751445087,
"grad_norm": 0.8478213794038515,
"learning_rate": 1.9754701818697515e-05,
"loss": 0.6754,
"step": 283
},
{
"epoch": 0.8208092485549133,
"grad_norm": 0.8916244711990654,
"learning_rate": 1.9750240321892743e-05,
"loss": 0.6653,
"step": 284
},
{
"epoch": 0.8236994219653179,
"grad_norm": 1.0196060923330346,
"learning_rate": 1.9745739129915508e-05,
"loss": 0.658,
"step": 285
},
{
"epoch": 0.8265895953757225,
"grad_norm": 0.7769568302916775,
"learning_rate": 1.9741198261091067e-05,
"loss": 0.6569,
"step": 286
},
{
"epoch": 0.8294797687861272,
"grad_norm": 1.0029214919025133,
"learning_rate": 1.973661773390619e-05,
"loss": 0.6749,
"step": 287
},
{
"epoch": 0.8323699421965318,
"grad_norm": 0.8327215361003341,
"learning_rate": 1.9731997567009125e-05,
"loss": 0.6593,
"step": 288
},
{
"epoch": 0.8352601156069365,
"grad_norm": 0.9901656491825068,
"learning_rate": 1.9727337779209496e-05,
"loss": 0.6544,
"step": 289
},
{
"epoch": 0.838150289017341,
"grad_norm": 0.9449653589975971,
"learning_rate": 1.9722638389478218e-05,
"loss": 0.6785,
"step": 290
},
{
"epoch": 0.8410404624277457,
"grad_norm": 0.9008910405778384,
"learning_rate": 1.9717899416947452e-05,
"loss": 0.6581,
"step": 291
},
{
"epoch": 0.8439306358381503,
"grad_norm": 0.9359751850845522,
"learning_rate": 1.9713120880910495e-05,
"loss": 0.6596,
"step": 292
},
{
"epoch": 0.846820809248555,
"grad_norm": 1.019479157630927,
"learning_rate": 1.9708302800821724e-05,
"loss": 0.6785,
"step": 293
},
{
"epoch": 0.8497109826589595,
"grad_norm": 0.807717813745512,
"learning_rate": 1.97034451962965e-05,
"loss": 0.6578,
"step": 294
},
{
"epoch": 0.8526011560693642,
"grad_norm": 0.914205212195178,
"learning_rate": 1.9698548087111104e-05,
"loss": 0.6594,
"step": 295
},
{
"epoch": 0.8554913294797688,
"grad_norm": 0.8701390512807908,
"learning_rate": 1.9693611493202637e-05,
"loss": 0.6449,
"step": 296
},
{
"epoch": 0.8583815028901735,
"grad_norm": 0.941351078554254,
"learning_rate": 1.9688635434668962e-05,
"loss": 0.6832,
"step": 297
},
{
"epoch": 0.861271676300578,
"grad_norm": 0.8193353541415879,
"learning_rate": 1.9683619931768606e-05,
"loss": 0.657,
"step": 298
},
{
"epoch": 0.8641618497109826,
"grad_norm": 0.8451216158980733,
"learning_rate": 1.967856500492068e-05,
"loss": 0.6463,
"step": 299
},
{
"epoch": 0.8670520231213873,
"grad_norm": 1.014043705928693,
"learning_rate": 1.96734706747048e-05,
"loss": 0.6678,
"step": 300
},
{
"epoch": 0.869942196531792,
"grad_norm": 0.9180712848357285,
"learning_rate": 1.9668336961861004e-05,
"loss": 0.6571,
"step": 301
},
{
"epoch": 0.8728323699421965,
"grad_norm": 0.9771471615649878,
"learning_rate": 1.966316388728965e-05,
"loss": 0.6841,
"step": 302
},
{
"epoch": 0.8757225433526011,
"grad_norm": 0.9172877654678109,
"learning_rate": 1.9657951472051363e-05,
"loss": 0.6673,
"step": 303
},
{
"epoch": 0.8786127167630058,
"grad_norm": 0.7838927377379438,
"learning_rate": 1.9652699737366924e-05,
"loss": 0.6799,
"step": 304
},
{
"epoch": 0.8815028901734104,
"grad_norm": 1.032296971095877,
"learning_rate": 1.964740870461719e-05,
"loss": 0.656,
"step": 305
},
{
"epoch": 0.884393063583815,
"grad_norm": 0.9338887493109845,
"learning_rate": 1.9642078395343015e-05,
"loss": 0.6778,
"step": 306
},
{
"epoch": 0.8872832369942196,
"grad_norm": 1.0366456810470164,
"learning_rate": 1.9636708831245147e-05,
"loss": 0.6687,
"step": 307
},
{
"epoch": 0.8901734104046243,
"grad_norm": 1.0352798236975347,
"learning_rate": 1.9631300034184155e-05,
"loss": 0.6614,
"step": 308
},
{
"epoch": 0.8930635838150289,
"grad_norm": 1.0719560333820592,
"learning_rate": 1.9625852026180325e-05,
"loss": 0.6511,
"step": 309
},
{
"epoch": 0.8959537572254336,
"grad_norm": 1.3788790895580383,
"learning_rate": 1.962036482941359e-05,
"loss": 0.6278,
"step": 310
},
{
"epoch": 0.8988439306358381,
"grad_norm": 1.2490418926438716,
"learning_rate": 1.9614838466223415e-05,
"loss": 0.6527,
"step": 311
},
{
"epoch": 0.9017341040462428,
"grad_norm": 1.0890036074128466,
"learning_rate": 1.9609272959108737e-05,
"loss": 0.6713,
"step": 312
},
{
"epoch": 0.9046242774566474,
"grad_norm": 1.2685369636239598,
"learning_rate": 1.9603668330727835e-05,
"loss": 0.6687,
"step": 313
},
{
"epoch": 0.9075144508670521,
"grad_norm": 0.9375781230616366,
"learning_rate": 1.9598024603898272e-05,
"loss": 0.6601,
"step": 314
},
{
"epoch": 0.9104046242774566,
"grad_norm": 0.9805983483939692,
"learning_rate": 1.959234180159679e-05,
"loss": 0.6636,
"step": 315
},
{
"epoch": 0.9132947976878613,
"grad_norm": 0.8345215644291492,
"learning_rate": 1.95866199469592e-05,
"loss": 0.6662,
"step": 316
},
{
"epoch": 0.9161849710982659,
"grad_norm": 1.1419645380118422,
"learning_rate": 1.9580859063280326e-05,
"loss": 0.6721,
"step": 317
},
{
"epoch": 0.9190751445086706,
"grad_norm": 0.8518284778478276,
"learning_rate": 1.9575059174013865e-05,
"loss": 0.6452,
"step": 318
},
{
"epoch": 0.9219653179190751,
"grad_norm": 1.0259765421703095,
"learning_rate": 1.9569220302772324e-05,
"loss": 0.68,
"step": 319
},
{
"epoch": 0.9248554913294798,
"grad_norm": 0.9338132523901228,
"learning_rate": 1.9563342473326915e-05,
"loss": 0.6669,
"step": 320
},
{
"epoch": 0.9277456647398844,
"grad_norm": 0.780487742190609,
"learning_rate": 1.955742570960745e-05,
"loss": 0.6569,
"step": 321
},
{
"epoch": 0.930635838150289,
"grad_norm": 1.063738587393258,
"learning_rate": 1.9551470035702254e-05,
"loss": 0.6612,
"step": 322
},
{
"epoch": 0.9335260115606936,
"grad_norm": 0.9243223458852142,
"learning_rate": 1.9545475475858068e-05,
"loss": 0.6542,
"step": 323
},
{
"epoch": 0.9364161849710982,
"grad_norm": 0.8371252573118421,
"learning_rate": 1.9539442054479932e-05,
"loss": 0.6642,
"step": 324
},
{
"epoch": 0.9393063583815029,
"grad_norm": 1.0489130551960444,
"learning_rate": 1.953336979613112e-05,
"loss": 0.6582,
"step": 325
},
{
"epoch": 0.9421965317919075,
"grad_norm": 0.8634370818517902,
"learning_rate": 1.952725872553299e-05,
"loss": 0.6517,
"step": 326
},
{
"epoch": 0.9450867052023122,
"grad_norm": 0.996549774465454,
"learning_rate": 1.9521108867564945e-05,
"loss": 0.6567,
"step": 327
},
{
"epoch": 0.9479768786127167,
"grad_norm": 0.8744840085025122,
"learning_rate": 1.9514920247264272e-05,
"loss": 0.6778,
"step": 328
},
{
"epoch": 0.9508670520231214,
"grad_norm": 0.7679718534975754,
"learning_rate": 1.9508692889826085e-05,
"loss": 0.641,
"step": 329
},
{
"epoch": 0.953757225433526,
"grad_norm": 1.0185045543013254,
"learning_rate": 1.9502426820603194e-05,
"loss": 0.6538,
"step": 330
},
{
"epoch": 0.9566473988439307,
"grad_norm": 0.7892516767280732,
"learning_rate": 1.9496122065106017e-05,
"loss": 0.6689,
"step": 331
},
{
"epoch": 0.9595375722543352,
"grad_norm": 0.9977508723060716,
"learning_rate": 1.948977864900247e-05,
"loss": 0.687,
"step": 332
},
{
"epoch": 0.9624277456647399,
"grad_norm": 0.8319535102491462,
"learning_rate": 1.9483396598117865e-05,
"loss": 0.6688,
"step": 333
},
{
"epoch": 0.9653179190751445,
"grad_norm": 0.799972427515052,
"learning_rate": 1.9476975938434807e-05,
"loss": 0.6519,
"step": 334
},
{
"epoch": 0.9682080924855492,
"grad_norm": 0.8287106660579454,
"learning_rate": 1.9470516696093075e-05,
"loss": 0.6804,
"step": 335
},
{
"epoch": 0.9710982658959537,
"grad_norm": 0.8408671077437165,
"learning_rate": 1.946401889738954e-05,
"loss": 0.6735,
"step": 336
},
{
"epoch": 0.9739884393063584,
"grad_norm": 0.778788680444376,
"learning_rate": 1.945748256877803e-05,
"loss": 0.6424,
"step": 337
},
{
"epoch": 0.976878612716763,
"grad_norm": 0.8202129621633388,
"learning_rate": 1.9450907736869244e-05,
"loss": 0.6616,
"step": 338
},
{
"epoch": 0.9797687861271677,
"grad_norm": 0.7097697012893457,
"learning_rate": 1.9444294428430637e-05,
"loss": 0.6468,
"step": 339
},
{
"epoch": 0.9826589595375722,
"grad_norm": 0.8208343688250482,
"learning_rate": 1.9437642670386303e-05,
"loss": 0.6354,
"step": 340
},
{
"epoch": 0.9855491329479769,
"grad_norm": 0.8621583549435662,
"learning_rate": 1.9430952489816876e-05,
"loss": 0.667,
"step": 341
},
{
"epoch": 0.9884393063583815,
"grad_norm": 0.7289222420622824,
"learning_rate": 1.9424223913959415e-05,
"loss": 0.6473,
"step": 342
},
{
"epoch": 0.9913294797687862,
"grad_norm": 0.7902494174858328,
"learning_rate": 1.9417456970207294e-05,
"loss": 0.6506,
"step": 343
},
{
"epoch": 0.9942196531791907,
"grad_norm": 0.8836340418645006,
"learning_rate": 1.941065168611009e-05,
"loss": 0.6588,
"step": 344
},
{
"epoch": 0.9971098265895953,
"grad_norm": 0.8029794711667372,
"learning_rate": 1.9403808089373473e-05,
"loss": 0.6514,
"step": 345
},
{
"epoch": 1.0,
"grad_norm": 0.7974385031738896,
"learning_rate": 1.9396926207859085e-05,
"loss": 0.6676,
"step": 346
},
{
"epoch": 1.0028901734104045,
"grad_norm": 0.759112688070958,
"learning_rate": 1.9390006069584443e-05,
"loss": 0.6119,
"step": 347
},
{
"epoch": 1.0057803468208093,
"grad_norm": 0.8862274468565814,
"learning_rate": 1.938304770272281e-05,
"loss": 0.6021,
"step": 348
},
{
"epoch": 1.0086705202312138,
"grad_norm": 0.7814352524501207,
"learning_rate": 1.9376051135603082e-05,
"loss": 0.6044,
"step": 349
},
{
"epoch": 1.0115606936416186,
"grad_norm": 0.9449437554187625,
"learning_rate": 1.936901639670968e-05,
"loss": 0.579,
"step": 350
},
{
"epoch": 1.0144508670520231,
"grad_norm": 0.8063769589686405,
"learning_rate": 1.9361943514682426e-05,
"loss": 0.6396,
"step": 351
},
{
"epoch": 1.0173410404624277,
"grad_norm": 0.854490910174394,
"learning_rate": 1.935483251831644e-05,
"loss": 0.5835,
"step": 352
},
{
"epoch": 1.0202312138728324,
"grad_norm": 0.8605749465708546,
"learning_rate": 1.9347683436562e-05,
"loss": 0.5972,
"step": 353
},
{
"epoch": 1.023121387283237,
"grad_norm": 0.8097614763630925,
"learning_rate": 1.9340496298524444e-05,
"loss": 0.5751,
"step": 354
},
{
"epoch": 1.0260115606936415,
"grad_norm": 0.8683551593034882,
"learning_rate": 1.9333271133464048e-05,
"loss": 0.5863,
"step": 355
},
{
"epoch": 1.0289017341040463,
"grad_norm": 0.7888873961905672,
"learning_rate": 1.932600797079589e-05,
"loss": 0.5917,
"step": 356
},
{
"epoch": 1.0317919075144508,
"grad_norm": 0.8782926408362381,
"learning_rate": 1.9318706840089766e-05,
"loss": 0.5962,
"step": 357
},
{
"epoch": 1.0346820809248556,
"grad_norm": 0.8178114356790833,
"learning_rate": 1.9311367771070025e-05,
"loss": 0.5864,
"step": 358
},
{
"epoch": 1.0375722543352601,
"grad_norm": 0.8629398360113137,
"learning_rate": 1.9303990793615485e-05,
"loss": 0.5984,
"step": 359
},
{
"epoch": 1.0404624277456647,
"grad_norm": 0.6814483238738642,
"learning_rate": 1.9296575937759293e-05,
"loss": 0.5891,
"step": 360
},
{
"epoch": 1.0433526011560694,
"grad_norm": 0.7624866277749663,
"learning_rate": 1.92891232336888e-05,
"loss": 0.5957,
"step": 361
},
{
"epoch": 1.046242774566474,
"grad_norm": 0.6997601771348588,
"learning_rate": 1.928163271174546e-05,
"loss": 0.5858,
"step": 362
},
{
"epoch": 1.0491329479768785,
"grad_norm": 0.6875497438405674,
"learning_rate": 1.9274104402424672e-05,
"loss": 0.5929,
"step": 363
},
{
"epoch": 1.0520231213872833,
"grad_norm": 0.7298576901663,
"learning_rate": 1.926653833637569e-05,
"loss": 0.5528,
"step": 364
},
{
"epoch": 1.0549132947976878,
"grad_norm": 0.7597749551660024,
"learning_rate": 1.9258934544401474e-05,
"loss": 0.5807,
"step": 365
},
{
"epoch": 1.0578034682080926,
"grad_norm": 0.8341419746232556,
"learning_rate": 1.925129305745858e-05,
"loss": 0.6007,
"step": 366
},
{
"epoch": 1.060693641618497,
"grad_norm": 0.8363319211825032,
"learning_rate": 1.924361390665702e-05,
"loss": 0.5866,
"step": 367
},
{
"epoch": 1.0635838150289016,
"grad_norm": 0.8341959472463382,
"learning_rate": 1.9235897123260155e-05,
"loss": 0.5955,
"step": 368
},
{
"epoch": 1.0664739884393064,
"grad_norm": 0.8133892730497435,
"learning_rate": 1.9228142738684546e-05,
"loss": 0.5784,
"step": 369
},
{
"epoch": 1.069364161849711,
"grad_norm": 1.2777260488694913,
"learning_rate": 1.922035078449984e-05,
"loss": 0.5847,
"step": 370
},
{
"epoch": 1.0722543352601157,
"grad_norm": 0.8653900558337858,
"learning_rate": 1.921252129242863e-05,
"loss": 0.5924,
"step": 371
},
{
"epoch": 1.0751445086705202,
"grad_norm": 0.9385954594567523,
"learning_rate": 1.9204654294346345e-05,
"loss": 0.6041,
"step": 372
},
{
"epoch": 1.0780346820809248,
"grad_norm": 0.87509462682995,
"learning_rate": 1.91967498222811e-05,
"loss": 0.5982,
"step": 373
},
{
"epoch": 1.0809248554913296,
"grad_norm": 0.94107834956112,
"learning_rate": 1.918880790841358e-05,
"loss": 0.5929,
"step": 374
},
{
"epoch": 1.083815028901734,
"grad_norm": 1.039879945508619,
"learning_rate": 1.91808285850769e-05,
"loss": 0.5743,
"step": 375
},
{
"epoch": 1.0867052023121386,
"grad_norm": 0.9683102748368639,
"learning_rate": 1.917281188475647e-05,
"loss": 0.5854,
"step": 376
},
{
"epoch": 1.0895953757225434,
"grad_norm": 0.9833933562897522,
"learning_rate": 1.9164757840089888e-05,
"loss": 0.5907,
"step": 377
},
{
"epoch": 1.092485549132948,
"grad_norm": 0.8360949807989344,
"learning_rate": 1.9156666483866764e-05,
"loss": 0.5704,
"step": 378
},
{
"epoch": 1.0953757225433527,
"grad_norm": 0.8195838450969974,
"learning_rate": 1.9148537849028624e-05,
"loss": 0.5908,
"step": 379
},
{
"epoch": 1.0982658959537572,
"grad_norm": 0.8210293235148685,
"learning_rate": 1.9140371968668767e-05,
"loss": 0.5777,
"step": 380
},
{
"epoch": 1.1011560693641618,
"grad_norm": 0.8082878324842184,
"learning_rate": 1.9132168876032115e-05,
"loss": 0.5913,
"step": 381
},
{
"epoch": 1.1040462427745665,
"grad_norm": 0.7053348622324761,
"learning_rate": 1.9123928604515093e-05,
"loss": 0.5811,
"step": 382
},
{
"epoch": 1.106936416184971,
"grad_norm": 0.8393957205841531,
"learning_rate": 1.9115651187665496e-05,
"loss": 0.5947,
"step": 383
},
{
"epoch": 1.1098265895953756,
"grad_norm": 0.8570047895468481,
"learning_rate": 1.9107336659182335e-05,
"loss": 0.5966,
"step": 384
},
{
"epoch": 1.1127167630057804,
"grad_norm": 0.7561662655121975,
"learning_rate": 1.909898505291571e-05,
"loss": 0.5923,
"step": 385
},
{
"epoch": 1.115606936416185,
"grad_norm": 0.8914708422505475,
"learning_rate": 1.909059640286668e-05,
"loss": 0.5947,
"step": 386
},
{
"epoch": 1.1184971098265897,
"grad_norm": 0.7647763521555886,
"learning_rate": 1.9082170743187107e-05,
"loss": 0.5834,
"step": 387
},
{
"epoch": 1.1213872832369942,
"grad_norm": 0.7433559499879944,
"learning_rate": 1.9073708108179536e-05,
"loss": 0.6127,
"step": 388
},
{
"epoch": 1.1242774566473988,
"grad_norm": 0.7292753284766555,
"learning_rate": 1.9065208532297043e-05,
"loss": 0.5756,
"step": 389
},
{
"epoch": 1.1271676300578035,
"grad_norm": 0.6464374837535348,
"learning_rate": 1.9056672050143087e-05,
"loss": 0.5833,
"step": 390
},
{
"epoch": 1.130057803468208,
"grad_norm": 0.7753730546050381,
"learning_rate": 1.90480986964714e-05,
"loss": 0.5922,
"step": 391
},
{
"epoch": 1.1329479768786128,
"grad_norm": 0.6435645880482299,
"learning_rate": 1.903948850618581e-05,
"loss": 0.6059,
"step": 392
},
{
"epoch": 1.1358381502890174,
"grad_norm": 0.7434435721364454,
"learning_rate": 1.9030841514340116e-05,
"loss": 0.602,
"step": 393
},
{
"epoch": 1.138728323699422,
"grad_norm": 60.08083679049716,
"learning_rate": 1.9022157756137948e-05,
"loss": 0.5944,
"step": 394
},
{
"epoch": 1.1416184971098267,
"grad_norm": 37.44031875571649,
"learning_rate": 1.9013437266932616e-05,
"loss": 0.6008,
"step": 395
},
{
"epoch": 1.1445086705202312,
"grad_norm": 1.6749714439835397,
"learning_rate": 1.900468008222697e-05,
"loss": 0.5846,
"step": 396
},
{
"epoch": 1.147398843930636,
"grad_norm": 0.9344473190434486,
"learning_rate": 1.8995886237673254e-05,
"loss": 0.598,
"step": 397
},
{
"epoch": 1.1502890173410405,
"grad_norm": 1.0836943912001913,
"learning_rate": 1.8987055769072973e-05,
"loss": 0.5762,
"step": 398
},
{
"epoch": 1.153179190751445,
"grad_norm": 1.2528088342761161,
"learning_rate": 1.897818871237671e-05,
"loss": 0.5925,
"step": 399
},
{
"epoch": 1.1560693641618498,
"grad_norm": 0.9917984973637564,
"learning_rate": 1.8969285103684033e-05,
"loss": 0.5671,
"step": 400
},
{
"epoch": 1.1589595375722543,
"grad_norm": 1.5567160801823279,
"learning_rate": 1.89603449792433e-05,
"loss": 0.5696,
"step": 401
},
{
"epoch": 1.1618497109826589,
"grad_norm": 1.4218814005716265,
"learning_rate": 1.8951368375451547e-05,
"loss": 0.5835,
"step": 402
},
{
"epoch": 1.1647398843930636,
"grad_norm": 1.073972437970269,
"learning_rate": 1.894235532885431e-05,
"loss": 0.579,
"step": 403
},
{
"epoch": 1.1676300578034682,
"grad_norm": 1.404716962457925,
"learning_rate": 1.8933305876145508e-05,
"loss": 0.5946,
"step": 404
},
{
"epoch": 1.1705202312138727,
"grad_norm": 0.7867748017486164,
"learning_rate": 1.8924220054167258e-05,
"loss": 0.6072,
"step": 405
},
{
"epoch": 1.1734104046242775,
"grad_norm": 1.2268275967968227,
"learning_rate": 1.891509789990976e-05,
"loss": 0.6026,
"step": 406
},
{
"epoch": 1.176300578034682,
"grad_norm": 0.782129228131316,
"learning_rate": 1.8905939450511117e-05,
"loss": 0.6027,
"step": 407
},
{
"epoch": 1.1791907514450868,
"grad_norm": 1.0957712387505731,
"learning_rate": 1.889674474325721e-05,
"loss": 0.5833,
"step": 408
},
{
"epoch": 1.1820809248554913,
"grad_norm": 0.7457776282138502,
"learning_rate": 1.888751381558152e-05,
"loss": 0.5905,
"step": 409
},
{
"epoch": 1.1849710982658959,
"grad_norm": 0.9771507342260098,
"learning_rate": 1.8878246705064995e-05,
"loss": 0.592,
"step": 410
},
{
"epoch": 1.1878612716763006,
"grad_norm": 0.7874916056588455,
"learning_rate": 1.886894344943589e-05,
"loss": 0.6106,
"step": 411
},
{
"epoch": 1.1907514450867052,
"grad_norm": 0.8386954062604738,
"learning_rate": 1.8859604086569616e-05,
"loss": 0.5819,
"step": 412
},
{
"epoch": 1.19364161849711,
"grad_norm": 0.6800465330568464,
"learning_rate": 1.885022865448858e-05,
"loss": 0.579,
"step": 413
},
{
"epoch": 1.1965317919075145,
"grad_norm": 0.84906552256597,
"learning_rate": 1.8840817191362033e-05,
"loss": 0.5783,
"step": 414
},
{
"epoch": 1.199421965317919,
"grad_norm": 0.7608413013366039,
"learning_rate": 1.883136973550592e-05,
"loss": 0.5881,
"step": 415
},
{
"epoch": 1.2023121387283238,
"grad_norm": 0.7850338345722294,
"learning_rate": 1.8821886325382718e-05,
"loss": 0.5625,
"step": 416
},
{
"epoch": 1.2052023121387283,
"grad_norm": 0.7620713234702672,
"learning_rate": 1.881236699960128e-05,
"loss": 0.6098,
"step": 417
},
{
"epoch": 1.208092485549133,
"grad_norm": 0.673183012859137,
"learning_rate": 1.8802811796916677e-05,
"loss": 0.5877,
"step": 418
},
{
"epoch": 1.2109826589595376,
"grad_norm": 0.725146027963119,
"learning_rate": 1.879322075623005e-05,
"loss": 0.5781,
"step": 419
},
{
"epoch": 1.2138728323699421,
"grad_norm": 0.8111014658303358,
"learning_rate": 1.878359391658843e-05,
"loss": 0.5909,
"step": 420
},
{
"epoch": 1.216763005780347,
"grad_norm": 0.7229365039244683,
"learning_rate": 1.8773931317184607e-05,
"loss": 0.581,
"step": 421
},
{
"epoch": 1.2196531791907514,
"grad_norm": 0.7844203333806608,
"learning_rate": 1.876423299735695e-05,
"loss": 0.602,
"step": 422
},
{
"epoch": 1.222543352601156,
"grad_norm": 0.7203643454120146,
"learning_rate": 1.875449899658925e-05,
"loss": 0.5783,
"step": 423
},
{
"epoch": 1.2254335260115607,
"grad_norm": 0.7232973902101298,
"learning_rate": 1.8744729354510566e-05,
"loss": 0.5731,
"step": 424
},
{
"epoch": 1.2283236994219653,
"grad_norm": 0.8824647184156448,
"learning_rate": 1.8734924110895056e-05,
"loss": 0.5805,
"step": 425
},
{
"epoch": 1.2312138728323698,
"grad_norm": 0.7228862642569523,
"learning_rate": 1.8725083305661825e-05,
"loss": 0.5799,
"step": 426
},
{
"epoch": 1.2341040462427746,
"grad_norm": 0.8780482452876094,
"learning_rate": 1.8715206978874753e-05,
"loss": 0.5855,
"step": 427
},
{
"epoch": 1.2369942196531791,
"grad_norm": 0.9453380989200908,
"learning_rate": 1.870529517074234e-05,
"loss": 0.5794,
"step": 428
},
{
"epoch": 1.239884393063584,
"grad_norm": 0.78977409227397,
"learning_rate": 1.869534792161752e-05,
"loss": 0.5858,
"step": 429
},
{
"epoch": 1.2427745664739884,
"grad_norm": 0.8341640471700158,
"learning_rate": 1.8685365271997543e-05,
"loss": 0.5889,
"step": 430
},
{
"epoch": 1.245664739884393,
"grad_norm": 0.8883116241060414,
"learning_rate": 1.8675347262523756e-05,
"loss": 0.5744,
"step": 431
},
{
"epoch": 1.2485549132947977,
"grad_norm": 0.822861993276753,
"learning_rate": 1.866529393398148e-05,
"loss": 0.5805,
"step": 432
},
{
"epoch": 1.2514450867052023,
"grad_norm": 0.8668081481118199,
"learning_rate": 1.8655205327299813e-05,
"loss": 0.5986,
"step": 433
},
{
"epoch": 1.254335260115607,
"grad_norm": 0.7150804000976962,
"learning_rate": 1.864508148355149e-05,
"loss": 0.5978,
"step": 434
},
{
"epoch": 1.2572254335260116,
"grad_norm": 0.9364709689963131,
"learning_rate": 1.8634922443952692e-05,
"loss": 0.5956,
"step": 435
},
{
"epoch": 1.260115606936416,
"grad_norm": 0.73444038368573,
"learning_rate": 1.86247282498629e-05,
"loss": 0.5738,
"step": 436
},
{
"epoch": 1.2630057803468209,
"grad_norm": 0.7427485980050539,
"learning_rate": 1.861449894278471e-05,
"loss": 0.5847,
"step": 437
},
{
"epoch": 1.2658959537572254,
"grad_norm": 0.7263094084914291,
"learning_rate": 1.8604234564363667e-05,
"loss": 0.5709,
"step": 438
},
{
"epoch": 1.2687861271676302,
"grad_norm": 0.6408013443679268,
"learning_rate": 1.8593935156388104e-05,
"loss": 0.5758,
"step": 439
},
{
"epoch": 1.2716763005780347,
"grad_norm": 0.7502681132648505,
"learning_rate": 1.8583600760788967e-05,
"loss": 0.5942,
"step": 440
},
{
"epoch": 1.2745664739884393,
"grad_norm": 0.745513544179955,
"learning_rate": 1.857323141963964e-05,
"loss": 0.6025,
"step": 441
},
{
"epoch": 1.2774566473988438,
"grad_norm": 0.6942252533539747,
"learning_rate": 1.8562827175155777e-05,
"loss": 0.6167,
"step": 442
},
{
"epoch": 1.2803468208092486,
"grad_norm": 0.718098643343022,
"learning_rate": 1.855238806969513e-05,
"loss": 0.5633,
"step": 443
},
{
"epoch": 1.2832369942196533,
"grad_norm": 0.7839064869449918,
"learning_rate": 1.8541914145757383e-05,
"loss": 0.5997,
"step": 444
},
{
"epoch": 1.2861271676300579,
"grad_norm": 0.7956894509101192,
"learning_rate": 1.853140544598397e-05,
"loss": 0.5964,
"step": 445
},
{
"epoch": 1.2890173410404624,
"grad_norm": 0.7422825022983778,
"learning_rate": 1.8520862013157898e-05,
"loss": 0.5919,
"step": 446
},
{
"epoch": 1.291907514450867,
"grad_norm": 0.7641838189194473,
"learning_rate": 1.8510283890203585e-05,
"loss": 0.587,
"step": 447
},
{
"epoch": 1.2947976878612717,
"grad_norm": 0.7606612268729219,
"learning_rate": 1.8499671120186683e-05,
"loss": 0.5858,
"step": 448
},
{
"epoch": 1.2976878612716762,
"grad_norm": 0.7372439820132116,
"learning_rate": 1.848902374631389e-05,
"loss": 0.5837,
"step": 449
},
{
"epoch": 1.300578034682081,
"grad_norm": 0.967164565722023,
"learning_rate": 1.847834181193279e-05,
"loss": 0.6018,
"step": 450
},
{
"epoch": 1.3034682080924855,
"grad_norm": 0.8255799111031907,
"learning_rate": 1.846762536053167e-05,
"loss": 0.6087,
"step": 451
},
{
"epoch": 1.30635838150289,
"grad_norm": 0.663848287941645,
"learning_rate": 1.8456874435739337e-05,
"loss": 0.5849,
"step": 452
},
{
"epoch": 1.3092485549132948,
"grad_norm": 0.8432857541018854,
"learning_rate": 1.8446089081324947e-05,
"loss": 0.5681,
"step": 453
},
{
"epoch": 1.3121387283236994,
"grad_norm": 0.6729630606749915,
"learning_rate": 1.8435269341197836e-05,
"loss": 0.5841,
"step": 454
},
{
"epoch": 1.3150289017341041,
"grad_norm": 0.7541439660338949,
"learning_rate": 1.8424415259407317e-05,
"loss": 0.5722,
"step": 455
},
{
"epoch": 1.3179190751445087,
"grad_norm": 0.8533593226136065,
"learning_rate": 1.8413526880142524e-05,
"loss": 0.5863,
"step": 456
},
{
"epoch": 1.3208092485549132,
"grad_norm": 0.7611155456663428,
"learning_rate": 1.8402604247732224e-05,
"loss": 0.6055,
"step": 457
},
{
"epoch": 1.323699421965318,
"grad_norm": 0.8690834992386883,
"learning_rate": 1.839164740664462e-05,
"loss": 0.5986,
"step": 458
},
{
"epoch": 1.3265895953757225,
"grad_norm": 0.8169643185444966,
"learning_rate": 1.8380656401487208e-05,
"loss": 0.6068,
"step": 459
},
{
"epoch": 1.3294797687861273,
"grad_norm": 0.8979789289219052,
"learning_rate": 1.8369631277006556e-05,
"loss": 0.6058,
"step": 460
},
{
"epoch": 1.3323699421965318,
"grad_norm": 0.681886344899434,
"learning_rate": 1.8358572078088144e-05,
"loss": 0.5847,
"step": 461
},
{
"epoch": 1.3352601156069364,
"grad_norm": 0.9141104041541918,
"learning_rate": 1.8347478849756182e-05,
"loss": 0.564,
"step": 462
},
{
"epoch": 1.3381502890173411,
"grad_norm": 0.6669758128756574,
"learning_rate": 1.8336351637173405e-05,
"loss": 0.5694,
"step": 463
},
{
"epoch": 1.3410404624277457,
"grad_norm": 0.9401071775838993,
"learning_rate": 1.8325190485640924e-05,
"loss": 0.5824,
"step": 464
},
{
"epoch": 1.3439306358381504,
"grad_norm": 0.7492669020477855,
"learning_rate": 1.8313995440598002e-05,
"loss": 0.5916,
"step": 465
},
{
"epoch": 1.346820809248555,
"grad_norm": 0.7722595161498734,
"learning_rate": 1.830276654762191e-05,
"loss": 0.5841,
"step": 466
},
{
"epoch": 1.3497109826589595,
"grad_norm": 0.770036690413647,
"learning_rate": 1.8291503852427698e-05,
"loss": 0.5941,
"step": 467
},
{
"epoch": 1.352601156069364,
"grad_norm": 0.822537024839394,
"learning_rate": 1.8280207400868057e-05,
"loss": 0.6011,
"step": 468
},
{
"epoch": 1.3554913294797688,
"grad_norm": 0.7717763755818814,
"learning_rate": 1.8268877238933084e-05,
"loss": 0.5704,
"step": 469
},
{
"epoch": 1.3583815028901733,
"grad_norm": 0.6273504834608445,
"learning_rate": 1.825751341275013e-05,
"loss": 0.573,
"step": 470
},
{
"epoch": 1.361271676300578,
"grad_norm": 0.7204175643195626,
"learning_rate": 1.8246115968583597e-05,
"loss": 0.5836,
"step": 471
},
{
"epoch": 1.3641618497109826,
"grad_norm": 0.7107257229105907,
"learning_rate": 1.8234684952834758e-05,
"loss": 0.585,
"step": 472
},
{
"epoch": 1.3670520231213872,
"grad_norm": 0.7589385988495859,
"learning_rate": 1.8223220412041552e-05,
"loss": 0.587,
"step": 473
},
{
"epoch": 1.369942196531792,
"grad_norm": 0.8390223582562657,
"learning_rate": 1.821172239287841e-05,
"loss": 0.5989,
"step": 474
},
{
"epoch": 1.3728323699421965,
"grad_norm": 0.7671563822376801,
"learning_rate": 1.8200190942156063e-05,
"loss": 0.5843,
"step": 475
},
{
"epoch": 1.3757225433526012,
"grad_norm": 0.8165672732561593,
"learning_rate": 1.8188626106821346e-05,
"loss": 0.5922,
"step": 476
},
{
"epoch": 1.3786127167630058,
"grad_norm": 0.7548591808890495,
"learning_rate": 1.817702793395701e-05,
"loss": 0.5714,
"step": 477
},
{
"epoch": 1.3815028901734103,
"grad_norm": 0.7895371681175447,
"learning_rate": 1.816539647078153e-05,
"loss": 0.5876,
"step": 478
},
{
"epoch": 1.384393063583815,
"grad_norm": 0.7362383319580541,
"learning_rate": 1.8153731764648907e-05,
"loss": 0.5896,
"step": 479
},
{
"epoch": 1.3872832369942196,
"grad_norm": 0.8682834177952186,
"learning_rate": 1.8142033863048487e-05,
"loss": 0.5916,
"step": 480
},
{
"epoch": 1.3901734104046244,
"grad_norm": 0.843342860143743,
"learning_rate": 1.8130302813604762e-05,
"loss": 0.5968,
"step": 481
},
{
"epoch": 1.393063583815029,
"grad_norm": 0.9357147713641858,
"learning_rate": 1.8118538664077175e-05,
"loss": 0.5776,
"step": 482
},
{
"epoch": 1.3959537572254335,
"grad_norm": 0.726516144204578,
"learning_rate": 1.810674146235992e-05,
"loss": 0.5837,
"step": 483
},
{
"epoch": 1.3988439306358382,
"grad_norm": 0.7231249100124149,
"learning_rate": 1.8094911256481765e-05,
"loss": 0.5676,
"step": 484
},
{
"epoch": 1.4017341040462428,
"grad_norm": 0.8236757722841949,
"learning_rate": 1.8083048094605826e-05,
"loss": 0.5874,
"step": 485
},
{
"epoch": 1.4046242774566475,
"grad_norm": 0.9619067429859167,
"learning_rate": 1.807115202502941e-05,
"loss": 0.5861,
"step": 486
},
{
"epoch": 1.407514450867052,
"grad_norm": 0.7062417807318491,
"learning_rate": 1.805922309618378e-05,
"loss": 0.5912,
"step": 487
},
{
"epoch": 1.4104046242774566,
"grad_norm": 0.7900913898127147,
"learning_rate": 1.804726135663399e-05,
"loss": 0.5937,
"step": 488
},
{
"epoch": 1.4132947976878611,
"grad_norm": 0.876647311651091,
"learning_rate": 1.8035266855078663e-05,
"loss": 0.6093,
"step": 489
},
{
"epoch": 1.416184971098266,
"grad_norm": 0.795569177139578,
"learning_rate": 1.8023239640349814e-05,
"loss": 0.5921,
"step": 490
},
{
"epoch": 1.4190751445086704,
"grad_norm": 0.8846891720315945,
"learning_rate": 1.801117976141262e-05,
"loss": 0.58,
"step": 491
},
{
"epoch": 1.4219653179190752,
"grad_norm": 0.7553556873287146,
"learning_rate": 1.7999087267365265e-05,
"loss": 0.5704,
"step": 492
},
{
"epoch": 1.4248554913294798,
"grad_norm": 0.7580506512265982,
"learning_rate": 1.7986962207438692e-05,
"loss": 0.5928,
"step": 493
},
{
"epoch": 1.4277456647398843,
"grad_norm": 0.8692821453087939,
"learning_rate": 1.797480463099645e-05,
"loss": 0.577,
"step": 494
},
{
"epoch": 1.430635838150289,
"grad_norm": 0.681599198571664,
"learning_rate": 1.7962614587534448e-05,
"loss": 0.5753,
"step": 495
},
{
"epoch": 1.4335260115606936,
"grad_norm": 0.8477803298084209,
"learning_rate": 1.795039212668078e-05,
"loss": 0.5764,
"step": 496
},
{
"epoch": 1.4364161849710984,
"grad_norm": 0.7899636423682991,
"learning_rate": 1.793813729819553e-05,
"loss": 0.5931,
"step": 497
},
{
"epoch": 1.439306358381503,
"grad_norm": 0.7751880599800394,
"learning_rate": 1.792585015197055e-05,
"loss": 0.6065,
"step": 498
},
{
"epoch": 1.4421965317919074,
"grad_norm": 0.7819584492038231,
"learning_rate": 1.7913530738029252e-05,
"loss": 0.5922,
"step": 499
},
{
"epoch": 1.4450867052023122,
"grad_norm": 0.7410757363702914,
"learning_rate": 1.7901179106526438e-05,
"loss": 0.5901,
"step": 500
},
{
"epoch": 1.4479768786127167,
"grad_norm": 0.8294915410142112,
"learning_rate": 1.788879530774805e-05,
"loss": 0.6094,
"step": 501
},
{
"epoch": 1.4508670520231215,
"grad_norm": 0.825270638932142,
"learning_rate": 1.7876379392111012e-05,
"loss": 0.5941,
"step": 502
},
{
"epoch": 1.453757225433526,
"grad_norm": 0.8695263847370847,
"learning_rate": 1.7863931410162987e-05,
"loss": 0.5728,
"step": 503
},
{
"epoch": 1.4566473988439306,
"grad_norm": 0.7280349111620275,
"learning_rate": 1.7851451412582196e-05,
"loss": 0.5911,
"step": 504
},
{
"epoch": 1.4595375722543353,
"grad_norm": 1.0251637552257442,
"learning_rate": 1.783893945017719e-05,
"loss": 0.5721,
"step": 505
},
{
"epoch": 1.4624277456647399,
"grad_norm": 0.7591697317876812,
"learning_rate": 1.782639557388667e-05,
"loss": 0.5854,
"step": 506
},
{
"epoch": 1.4653179190751446,
"grad_norm": 0.9795699818825268,
"learning_rate": 1.7813819834779258e-05,
"loss": 0.5845,
"step": 507
},
{
"epoch": 1.4682080924855492,
"grad_norm": 0.8818773497711714,
"learning_rate": 1.780121228405329e-05,
"loss": 0.5863,
"step": 508
},
{
"epoch": 1.4710982658959537,
"grad_norm": 0.6655630594025235,
"learning_rate": 1.778857297303662e-05,
"loss": 0.5792,
"step": 509
},
{
"epoch": 1.4739884393063583,
"grad_norm": 0.8556541902196599,
"learning_rate": 1.777590195318641e-05,
"loss": 0.5746,
"step": 510
},
{
"epoch": 1.476878612716763,
"grad_norm": 0.6319453690933835,
"learning_rate": 1.77631992760889e-05,
"loss": 0.5774,
"step": 511
},
{
"epoch": 1.4797687861271676,
"grad_norm": 0.6634365810897271,
"learning_rate": 1.775046499345922e-05,
"loss": 0.6016,
"step": 512
},
{
"epoch": 1.4826589595375723,
"grad_norm": 0.7453817549418363,
"learning_rate": 1.773769915714118e-05,
"loss": 0.5818,
"step": 513
},
{
"epoch": 1.4855491329479769,
"grad_norm": 0.7329107259848197,
"learning_rate": 1.7724901819107047e-05,
"loss": 0.5789,
"step": 514
},
{
"epoch": 1.4884393063583814,
"grad_norm": 0.7397082974840387,
"learning_rate": 1.7712073031457332e-05,
"loss": 0.595,
"step": 515
},
{
"epoch": 1.4913294797687862,
"grad_norm": 0.7751550546024663,
"learning_rate": 1.769921284642058e-05,
"loss": 0.5936,
"step": 516
},
{
"epoch": 1.4942196531791907,
"grad_norm": 0.6978582914470572,
"learning_rate": 1.768632131635318e-05,
"loss": 0.5873,
"step": 517
},
{
"epoch": 1.4971098265895955,
"grad_norm": 0.7745359745262627,
"learning_rate": 1.767339849373912e-05,
"loss": 0.5946,
"step": 518
},
{
"epoch": 1.5,
"grad_norm": 0.7287263926354944,
"learning_rate": 1.766044443118978e-05,
"loss": 0.5682,
"step": 519
},
{
"epoch": 1.5028901734104045,
"grad_norm": 0.7722236362673658,
"learning_rate": 1.764745918144374e-05,
"loss": 0.5556,
"step": 520
},
{
"epoch": 1.5057803468208093,
"grad_norm": 0.7299445800710093,
"learning_rate": 1.7634442797366537e-05,
"loss": 0.5786,
"step": 521
},
{
"epoch": 1.5086705202312138,
"grad_norm": 0.6751944887927009,
"learning_rate": 1.7621395331950464e-05,
"loss": 0.5804,
"step": 522
},
{
"epoch": 1.5115606936416186,
"grad_norm": 0.6793007632896605,
"learning_rate": 1.7608316838314355e-05,
"loss": 0.5649,
"step": 523
},
{
"epoch": 1.5144508670520231,
"grad_norm": 0.6875547584975547,
"learning_rate": 1.759520736970337e-05,
"loss": 0.6058,
"step": 524
},
{
"epoch": 1.5173410404624277,
"grad_norm": 0.7035985997297632,
"learning_rate": 1.7582066979488764e-05,
"loss": 0.5863,
"step": 525
},
{
"epoch": 1.5202312138728322,
"grad_norm": 0.6824434177269557,
"learning_rate": 1.756889572116769e-05,
"loss": 0.5831,
"step": 526
},
{
"epoch": 1.523121387283237,
"grad_norm": 0.7919452976996221,
"learning_rate": 1.755569364836296e-05,
"loss": 0.581,
"step": 527
},
{
"epoch": 1.5260115606936417,
"grad_norm": 0.6716502550573372,
"learning_rate": 1.7542460814822853e-05,
"loss": 0.5966,
"step": 528
},
{
"epoch": 1.5289017341040463,
"grad_norm": 0.7018593765581507,
"learning_rate": 1.752919727442087e-05,
"loss": 0.5976,
"step": 529
},
{
"epoch": 1.5317919075144508,
"grad_norm": 0.7171067506633475,
"learning_rate": 1.7515903081155525e-05,
"loss": 0.5766,
"step": 530
},
{
"epoch": 1.5346820809248554,
"grad_norm": 0.7058408714374818,
"learning_rate": 1.7502578289150137e-05,
"loss": 0.5859,
"step": 531
},
{
"epoch": 1.5375722543352601,
"grad_norm": 0.7707834529909837,
"learning_rate": 1.748922295265258e-05,
"loss": 0.5989,
"step": 532
},
{
"epoch": 1.5404624277456649,
"grad_norm": 0.730864740486449,
"learning_rate": 1.7475837126035105e-05,
"loss": 0.5797,
"step": 533
},
{
"epoch": 1.5433526011560694,
"grad_norm": 0.7443397273692416,
"learning_rate": 1.7462420863794075e-05,
"loss": 0.5942,
"step": 534
},
{
"epoch": 1.546242774566474,
"grad_norm": 0.6699753066639157,
"learning_rate": 1.7448974220549765e-05,
"loss": 0.5843,
"step": 535
},
{
"epoch": 1.5491329479768785,
"grad_norm": 0.7834712901051348,
"learning_rate": 1.743549725104614e-05,
"loss": 0.5854,
"step": 536
},
{
"epoch": 1.5520231213872833,
"grad_norm": 0.7375236904395911,
"learning_rate": 1.742199001015064e-05,
"loss": 0.5873,
"step": 537
},
{
"epoch": 1.5549132947976878,
"grad_norm": 0.7520797737041237,
"learning_rate": 1.740845255285393e-05,
"loss": 0.5737,
"step": 538
},
{
"epoch": 1.5578034682080926,
"grad_norm": 0.750510184354196,
"learning_rate": 1.7394884934269695e-05,
"loss": 0.564,
"step": 539
},
{
"epoch": 1.560693641618497,
"grad_norm": 0.7504237975136158,
"learning_rate": 1.7381287209634417e-05,
"loss": 0.6057,
"step": 540
},
{
"epoch": 1.5635838150289016,
"grad_norm": 0.6607096876637601,
"learning_rate": 1.736765943430715e-05,
"loss": 0.5931,
"step": 541
},
{
"epoch": 1.5664739884393064,
"grad_norm": 0.7692482037901299,
"learning_rate": 1.7354001663769278e-05,
"loss": 0.5989,
"step": 542
},
{
"epoch": 1.569364161849711,
"grad_norm": 0.6528108961728156,
"learning_rate": 1.734031395362431e-05,
"loss": 0.5681,
"step": 543
},
{
"epoch": 1.5722543352601157,
"grad_norm": 0.8314040661899244,
"learning_rate": 1.7326596359597646e-05,
"loss": 0.5808,
"step": 544
},
{
"epoch": 1.5751445086705202,
"grad_norm": 0.6551143635361942,
"learning_rate": 1.731284893753634e-05,
"loss": 0.5824,
"step": 545
},
{
"epoch": 1.5780346820809248,
"grad_norm": 0.7504473441299307,
"learning_rate": 1.7299071743408894e-05,
"loss": 0.6188,
"step": 546
},
{
"epoch": 1.5809248554913293,
"grad_norm": 0.7011203491881448,
"learning_rate": 1.728526483330501e-05,
"loss": 0.5731,
"step": 547
},
{
"epoch": 1.583815028901734,
"grad_norm": 0.7685686045630253,
"learning_rate": 1.7271428263435373e-05,
"loss": 0.5714,
"step": 548
},
{
"epoch": 1.5867052023121389,
"grad_norm": 0.7044883437721119,
"learning_rate": 1.7257562090131422e-05,
"loss": 0.573,
"step": 549
},
{
"epoch": 1.5895953757225434,
"grad_norm": 0.8203622929135783,
"learning_rate": 1.7243666369845104e-05,
"loss": 0.5845,
"step": 550
},
{
"epoch": 1.592485549132948,
"grad_norm": 0.6889783456040496,
"learning_rate": 1.7229741159148676e-05,
"loss": 0.5933,
"step": 551
},
{
"epoch": 1.5953757225433525,
"grad_norm": 0.9765254946088467,
"learning_rate": 1.7215786514734438e-05,
"loss": 0.5665,
"step": 552
},
{
"epoch": 1.5982658959537572,
"grad_norm": 0.6955017748996175,
"learning_rate": 1.7201802493414538e-05,
"loss": 0.5571,
"step": 553
},
{
"epoch": 1.601156069364162,
"grad_norm": 0.9455064828292016,
"learning_rate": 1.7187789152120712e-05,
"loss": 0.5948,
"step": 554
},
{
"epoch": 1.6040462427745665,
"grad_norm": 0.8420994974130152,
"learning_rate": 1.7173746547904065e-05,
"loss": 0.5988,
"step": 555
},
{
"epoch": 1.606936416184971,
"grad_norm": 0.7954238132099828,
"learning_rate": 1.7159674737934843e-05,
"loss": 0.5639,
"step": 556
},
{
"epoch": 1.6098265895953756,
"grad_norm": 0.894451573841629,
"learning_rate": 1.7145573779502185e-05,
"loss": 0.596,
"step": 557
},
{
"epoch": 1.6127167630057804,
"grad_norm": 0.7857493523002436,
"learning_rate": 1.7131443730013907e-05,
"loss": 0.5838,
"step": 558
},
{
"epoch": 1.6156069364161851,
"grad_norm": 0.9094958240641396,
"learning_rate": 1.7117284646996256e-05,
"loss": 0.577,
"step": 559
},
{
"epoch": 1.6184971098265897,
"grad_norm": 0.7426411606349894,
"learning_rate": 1.7103096588093686e-05,
"loss": 0.5964,
"step": 560
},
{
"epoch": 1.6213872832369942,
"grad_norm": 0.7420940463823611,
"learning_rate": 1.708887961106861e-05,
"loss": 0.554,
"step": 561
},
{
"epoch": 1.6242774566473988,
"grad_norm": 0.8081581100315017,
"learning_rate": 1.707463377380118e-05,
"loss": 0.605,
"step": 562
},
{
"epoch": 1.6271676300578035,
"grad_norm": 0.6331916502112901,
"learning_rate": 1.706035913428904e-05,
"loss": 0.5921,
"step": 563
},
{
"epoch": 1.630057803468208,
"grad_norm": 0.7310464650272858,
"learning_rate": 1.704605575064709e-05,
"loss": 0.5707,
"step": 564
},
{
"epoch": 1.6329479768786128,
"grad_norm": 0.7523392612827752,
"learning_rate": 1.7031723681107254e-05,
"loss": 0.595,
"step": 565
},
{
"epoch": 1.6358381502890174,
"grad_norm": 0.7449756861349839,
"learning_rate": 1.7017362984018256e-05,
"loss": 0.5841,
"step": 566
},
{
"epoch": 1.638728323699422,
"grad_norm": 0.831280655903283,
"learning_rate": 1.7002973717845345e-05,
"loss": 0.5902,
"step": 567
},
{
"epoch": 1.6416184971098264,
"grad_norm": 0.83598321136025,
"learning_rate": 1.69885559411701e-05,
"loss": 0.5823,
"step": 568
},
{
"epoch": 1.6445086705202312,
"grad_norm": 0.7138345328389332,
"learning_rate": 1.6974109712690163e-05,
"loss": 0.5889,
"step": 569
},
{
"epoch": 1.647398843930636,
"grad_norm": 0.7110975012072633,
"learning_rate": 1.695963509121901e-05,
"loss": 0.5686,
"step": 570
},
{
"epoch": 1.6502890173410405,
"grad_norm": 0.8042107045528893,
"learning_rate": 1.6945132135685712e-05,
"loss": 0.5947,
"step": 571
},
{
"epoch": 1.653179190751445,
"grad_norm": 0.7013763940024369,
"learning_rate": 1.6930600905134688e-05,
"loss": 0.5604,
"step": 572
},
{
"epoch": 1.6560693641618496,
"grad_norm": 0.7785765624779681,
"learning_rate": 1.6916041458725486e-05,
"loss": 0.5963,
"step": 573
},
{
"epoch": 1.6589595375722543,
"grad_norm": 0.7821908290786825,
"learning_rate": 1.6901453855732503e-05,
"loss": 0.5621,
"step": 574
},
{
"epoch": 1.661849710982659,
"grad_norm": 0.7536723428025114,
"learning_rate": 1.6886838155544786e-05,
"loss": 0.6013,
"step": 575
},
{
"epoch": 1.6647398843930636,
"grad_norm": 0.7614681747533972,
"learning_rate": 1.6872194417665767e-05,
"loss": 0.5888,
"step": 576
},
{
"epoch": 1.6676300578034682,
"grad_norm": 0.7311883818551561,
"learning_rate": 1.6857522701713016e-05,
"loss": 0.5788,
"step": 577
},
{
"epoch": 1.6705202312138727,
"grad_norm": 0.7241599297519454,
"learning_rate": 1.684282306741802e-05,
"loss": 0.5602,
"step": 578
},
{
"epoch": 1.6734104046242775,
"grad_norm": 0.7858570405102655,
"learning_rate": 1.6828095574625917e-05,
"loss": 0.5684,
"step": 579
},
{
"epoch": 1.6763005780346822,
"grad_norm": 0.7306983679762703,
"learning_rate": 1.6813340283295265e-05,
"loss": 0.5733,
"step": 580
},
{
"epoch": 1.6791907514450868,
"grad_norm": 0.8176853531742918,
"learning_rate": 1.6798557253497802e-05,
"loss": 0.5927,
"step": 581
},
{
"epoch": 1.6820809248554913,
"grad_norm": 0.7464641313209691,
"learning_rate": 1.678374654541819e-05,
"loss": 0.5862,
"step": 582
},
{
"epoch": 1.6849710982658959,
"grad_norm": 0.7413436012267695,
"learning_rate": 1.6768908219353774e-05,
"loss": 0.5804,
"step": 583
},
{
"epoch": 1.6878612716763006,
"grad_norm": 0.7126278020137226,
"learning_rate": 1.6754042335714333e-05,
"loss": 0.5759,
"step": 584
},
{
"epoch": 1.6907514450867052,
"grad_norm": 0.7809699652694174,
"learning_rate": 1.6739148955021854e-05,
"loss": 0.5877,
"step": 585
},
{
"epoch": 1.69364161849711,
"grad_norm": 0.6640250271493856,
"learning_rate": 1.6724228137910247e-05,
"loss": 0.5947,
"step": 586
},
{
"epoch": 1.6965317919075145,
"grad_norm": 0.7176063205379599,
"learning_rate": 1.670927994512514e-05,
"loss": 0.6073,
"step": 587
},
{
"epoch": 1.699421965317919,
"grad_norm": 0.7913883631540534,
"learning_rate": 1.6694304437523608e-05,
"loss": 0.5699,
"step": 588
},
{
"epoch": 1.7023121387283235,
"grad_norm": 0.6369673178626757,
"learning_rate": 1.6679301676073923e-05,
"loss": 0.5606,
"step": 589
},
{
"epoch": 1.7052023121387283,
"grad_norm": 0.7416646166687233,
"learning_rate": 1.6664271721855325e-05,
"loss": 0.5834,
"step": 590
},
{
"epoch": 1.708092485549133,
"grad_norm": 0.6721518206723244,
"learning_rate": 1.664921463605775e-05,
"loss": 0.579,
"step": 591
},
{
"epoch": 1.7109826589595376,
"grad_norm": 0.6573326596850098,
"learning_rate": 1.6634130479981597e-05,
"loss": 0.5853,
"step": 592
},
{
"epoch": 1.7138728323699421,
"grad_norm": 0.7290689810748527,
"learning_rate": 1.6619019315037472e-05,
"loss": 0.5886,
"step": 593
},
{
"epoch": 1.7167630057803467,
"grad_norm": 0.8025709845059936,
"learning_rate": 1.6603881202745942e-05,
"loss": 0.5613,
"step": 594
},
{
"epoch": 1.7196531791907514,
"grad_norm": 0.7682312752649727,
"learning_rate": 1.6588716204737282e-05,
"loss": 0.5921,
"step": 595
},
{
"epoch": 1.7225433526011562,
"grad_norm": 0.6035407151326911,
"learning_rate": 1.657352438275122e-05,
"loss": 0.5655,
"step": 596
},
{
"epoch": 1.7254335260115607,
"grad_norm": 0.7977173862284962,
"learning_rate": 1.6558305798636697e-05,
"loss": 0.5904,
"step": 597
},
{
"epoch": 1.7283236994219653,
"grad_norm": 0.6749869989363424,
"learning_rate": 1.6543060514351594e-05,
"loss": 0.5812,
"step": 598
},
{
"epoch": 1.7312138728323698,
"grad_norm": 0.7134556633614937,
"learning_rate": 1.6527788591962516e-05,
"loss": 0.6047,
"step": 599
},
{
"epoch": 1.7341040462427746,
"grad_norm": 0.8293158717228061,
"learning_rate": 1.6512490093644494e-05,
"loss": 0.5879,
"step": 600
},
{
"epoch": 1.7369942196531793,
"grad_norm": 0.6811278610703695,
"learning_rate": 1.6497165081680766e-05,
"loss": 0.5921,
"step": 601
},
{
"epoch": 1.739884393063584,
"grad_norm": 0.7337901858628713,
"learning_rate": 1.6481813618462513e-05,
"loss": 0.5894,
"step": 602
},
{
"epoch": 1.7427745664739884,
"grad_norm": 0.689549851951907,
"learning_rate": 1.646643576648861e-05,
"loss": 0.5843,
"step": 603
},
{
"epoch": 1.745664739884393,
"grad_norm": 0.7939807044818971,
"learning_rate": 1.6451031588365345e-05,
"loss": 0.577,
"step": 604
},
{
"epoch": 1.7485549132947977,
"grad_norm": 0.7021558092806988,
"learning_rate": 1.643560114680621e-05,
"loss": 0.5683,
"step": 605
},
{
"epoch": 1.7514450867052023,
"grad_norm": 0.7212482485557895,
"learning_rate": 1.6420144504631604e-05,
"loss": 0.5791,
"step": 606
},
{
"epoch": 1.754335260115607,
"grad_norm": 0.6706672483451787,
"learning_rate": 1.6404661724768595e-05,
"loss": 0.6124,
"step": 607
},
{
"epoch": 1.7572254335260116,
"grad_norm": 0.7432156095439864,
"learning_rate": 1.6389152870250677e-05,
"loss": 0.6018,
"step": 608
},
{
"epoch": 1.760115606936416,
"grad_norm": 0.6406322039091135,
"learning_rate": 1.6373618004217483e-05,
"loss": 0.6029,
"step": 609
},
{
"epoch": 1.7630057803468207,
"grad_norm": 0.7828798092411525,
"learning_rate": 1.6358057189914552e-05,
"loss": 0.5625,
"step": 610
},
{
"epoch": 1.7658959537572254,
"grad_norm": 0.643002344168019,
"learning_rate": 1.634247049069306e-05,
"loss": 0.5912,
"step": 611
},
{
"epoch": 1.7687861271676302,
"grad_norm": 0.8961535556561148,
"learning_rate": 1.6326857970009568e-05,
"loss": 0.5687,
"step": 612
},
{
"epoch": 1.7716763005780347,
"grad_norm": 0.6866826745827112,
"learning_rate": 1.6311219691425764e-05,
"loss": 0.6037,
"step": 613
},
{
"epoch": 1.7745664739884393,
"grad_norm": 0.7766521938446624,
"learning_rate": 1.629555571860819e-05,
"loss": 0.5952,
"step": 614
},
{
"epoch": 1.7774566473988438,
"grad_norm": 0.7822299785872244,
"learning_rate": 1.6279866115328015e-05,
"loss": 0.5882,
"step": 615
},
{
"epoch": 1.7803468208092486,
"grad_norm": 0.6732758382850704,
"learning_rate": 1.6264150945460726e-05,
"loss": 0.5907,
"step": 616
},
{
"epoch": 1.7832369942196533,
"grad_norm": 0.6998293133649915,
"learning_rate": 1.6248410272985928e-05,
"loss": 0.595,
"step": 617
},
{
"epoch": 1.7861271676300579,
"grad_norm": 0.6427786891334882,
"learning_rate": 1.623264416198703e-05,
"loss": 0.5647,
"step": 618
},
{
"epoch": 1.7890173410404624,
"grad_norm": 0.769900518742028,
"learning_rate": 1.6216852676651016e-05,
"loss": 0.5719,
"step": 619
},
{
"epoch": 1.791907514450867,
"grad_norm": 0.6494905375694605,
"learning_rate": 1.6201035881268168e-05,
"loss": 0.6067,
"step": 620
},
{
"epoch": 1.7947976878612717,
"grad_norm": 0.7126382140280971,
"learning_rate": 1.6185193840231818e-05,
"loss": 0.5755,
"step": 621
},
{
"epoch": 1.7976878612716765,
"grad_norm": 0.7579906096389109,
"learning_rate": 1.6169326618038067e-05,
"loss": 0.5847,
"step": 622
},
{
"epoch": 1.800578034682081,
"grad_norm": 0.7487652525196908,
"learning_rate": 1.615343427928555e-05,
"loss": 0.5767,
"step": 623
},
{
"epoch": 1.8034682080924855,
"grad_norm": 0.7661053667162049,
"learning_rate": 1.6137516888675143e-05,
"loss": 0.5772,
"step": 624
},
{
"epoch": 1.80635838150289,
"grad_norm": 0.7960982048126151,
"learning_rate": 1.6121574511009712e-05,
"loss": 0.5838,
"step": 625
},
{
"epoch": 1.8092485549132948,
"grad_norm": 0.7049898412611506,
"learning_rate": 1.6105607211193862e-05,
"loss": 0.5893,
"step": 626
},
{
"epoch": 1.8121387283236994,
"grad_norm": 0.8326584145081675,
"learning_rate": 1.608961505423365e-05,
"loss": 0.5952,
"step": 627
},
{
"epoch": 1.8150289017341041,
"grad_norm": 0.8354389404361381,
"learning_rate": 1.6073598105236338e-05,
"loss": 0.5819,
"step": 628
},
{
"epoch": 1.8179190751445087,
"grad_norm": 0.6945106150715511,
"learning_rate": 1.605755642941012e-05,
"loss": 0.5849,
"step": 629
},
{
"epoch": 1.8208092485549132,
"grad_norm": 0.7338790569970648,
"learning_rate": 1.6041490092063853e-05,
"loss": 0.5963,
"step": 630
},
{
"epoch": 1.8236994219653178,
"grad_norm": 0.6655937405423681,
"learning_rate": 1.60253991586068e-05,
"loss": 0.5774,
"step": 631
},
{
"epoch": 1.8265895953757225,
"grad_norm": 0.6996809292538112,
"learning_rate": 1.6009283694548365e-05,
"loss": 0.6022,
"step": 632
},
{
"epoch": 1.8294797687861273,
"grad_norm": 0.6925368246008983,
"learning_rate": 1.5993143765497812e-05,
"loss": 0.5755,
"step": 633
},
{
"epoch": 1.8323699421965318,
"grad_norm": 0.6529851507970346,
"learning_rate": 1.597697943716401e-05,
"loss": 0.5908,
"step": 634
},
{
"epoch": 1.8352601156069364,
"grad_norm": 0.7321812506216079,
"learning_rate": 1.596079077535516e-05,
"loss": 0.5692,
"step": 635
},
{
"epoch": 1.838150289017341,
"grad_norm": 0.6191201485424263,
"learning_rate": 1.5944577845978534e-05,
"loss": 0.5988,
"step": 636
},
{
"epoch": 1.8410404624277457,
"grad_norm": 0.646374783015124,
"learning_rate": 1.5928340715040204e-05,
"loss": 0.5767,
"step": 637
},
{
"epoch": 1.8439306358381504,
"grad_norm": 0.6819560688437241,
"learning_rate": 1.5912079448644765e-05,
"loss": 0.5937,
"step": 638
},
{
"epoch": 1.846820809248555,
"grad_norm": 0.7164886573033494,
"learning_rate": 1.5895794112995074e-05,
"loss": 0.5966,
"step": 639
},
{
"epoch": 1.8497109826589595,
"grad_norm": 0.6644650088840722,
"learning_rate": 1.587948477439198e-05,
"loss": 0.5557,
"step": 640
},
{
"epoch": 1.852601156069364,
"grad_norm": 0.7455243137484561,
"learning_rate": 1.5863151499234053e-05,
"loss": 0.5913,
"step": 641
},
{
"epoch": 1.8554913294797688,
"grad_norm": 0.6655201268971009,
"learning_rate": 1.584679435401731e-05,
"loss": 0.5881,
"step": 642
},
{
"epoch": 1.8583815028901736,
"grad_norm": 0.6620269723317904,
"learning_rate": 1.5830413405334954e-05,
"loss": 0.5953,
"step": 643
},
{
"epoch": 1.861271676300578,
"grad_norm": 0.6763021820636497,
"learning_rate": 1.5814008719877096e-05,
"loss": 0.5629,
"step": 644
},
{
"epoch": 1.8641618497109826,
"grad_norm": 0.6941229988108782,
"learning_rate": 1.5797580364430475e-05,
"loss": 0.5896,
"step": 645
},
{
"epoch": 1.8670520231213872,
"grad_norm": 0.6665773655498453,
"learning_rate": 1.5781128405878202e-05,
"loss": 0.5801,
"step": 646
},
{
"epoch": 1.869942196531792,
"grad_norm": 0.7132340160866797,
"learning_rate": 1.5764652911199488e-05,
"loss": 0.5945,
"step": 647
},
{
"epoch": 1.8728323699421965,
"grad_norm": 0.6665820793295506,
"learning_rate": 1.5748153947469348e-05,
"loss": 0.5814,
"step": 648
},
{
"epoch": 1.8757225433526012,
"grad_norm": 0.6871939300797527,
"learning_rate": 1.573163158185836e-05,
"loss": 0.594,
"step": 649
},
{
"epoch": 1.8786127167630058,
"grad_norm": 0.6388515167925531,
"learning_rate": 1.5715085881632366e-05,
"loss": 0.5777,
"step": 650
},
{
"epoch": 1.8815028901734103,
"grad_norm": 0.7241320262561728,
"learning_rate": 1.569851691415221e-05,
"loss": 0.5658,
"step": 651
},
{
"epoch": 1.8843930635838149,
"grad_norm": 0.6666561142088108,
"learning_rate": 1.568192474687346e-05,
"loss": 0.583,
"step": 652
},
{
"epoch": 1.8872832369942196,
"grad_norm": 0.787208876465865,
"learning_rate": 1.5665309447346145e-05,
"loss": 0.5604,
"step": 653
},
{
"epoch": 1.8901734104046244,
"grad_norm": 0.7039870339303322,
"learning_rate": 1.5648671083214454e-05,
"loss": 0.5824,
"step": 654
},
{
"epoch": 1.893063583815029,
"grad_norm": 0.7121898583798462,
"learning_rate": 1.5632009722216493e-05,
"loss": 0.5903,
"step": 655
},
{
"epoch": 1.8959537572254335,
"grad_norm": 0.7757465203962726,
"learning_rate": 1.5615325432183977e-05,
"loss": 0.5795,
"step": 656
},
{
"epoch": 1.898843930635838,
"grad_norm": 0.6402622736455414,
"learning_rate": 1.559861828104198e-05,
"loss": 0.581,
"step": 657
},
{
"epoch": 1.9017341040462428,
"grad_norm": 0.7782034708328404,
"learning_rate": 1.558188833680865e-05,
"loss": 0.5742,
"step": 658
},
{
"epoch": 1.9046242774566475,
"grad_norm": 0.6324314813341162,
"learning_rate": 1.5565135667594916e-05,
"loss": 0.5714,
"step": 659
},
{
"epoch": 1.907514450867052,
"grad_norm": 0.7762351957222875,
"learning_rate": 1.5548360341604246e-05,
"loss": 0.5917,
"step": 660
},
{
"epoch": 1.9104046242774566,
"grad_norm": 0.8247409287869959,
"learning_rate": 1.5531562427132327e-05,
"loss": 0.566,
"step": 661
},
{
"epoch": 1.9132947976878611,
"grad_norm": 0.6626264291918831,
"learning_rate": 1.5514741992566826e-05,
"loss": 0.5767,
"step": 662
},
{
"epoch": 1.916184971098266,
"grad_norm": 0.9294772053397637,
"learning_rate": 1.549789910638708e-05,
"loss": 0.576,
"step": 663
},
{
"epoch": 1.9190751445086707,
"grad_norm": 0.6981208084909047,
"learning_rate": 1.5481033837163844e-05,
"loss": 0.58,
"step": 664
},
{
"epoch": 1.9219653179190752,
"grad_norm": 0.7557466097080571,
"learning_rate": 1.5464146253558987e-05,
"loss": 0.5751,
"step": 665
},
{
"epoch": 1.9248554913294798,
"grad_norm": 0.8115938243522869,
"learning_rate": 1.5447236424325226e-05,
"loss": 0.6041,
"step": 666
},
{
"epoch": 1.9277456647398843,
"grad_norm": 0.7042039351280144,
"learning_rate": 1.5430304418305853e-05,
"loss": 0.5815,
"step": 667
},
{
"epoch": 1.930635838150289,
"grad_norm": 0.8255191008863216,
"learning_rate": 1.541335030443444e-05,
"loss": 0.5897,
"step": 668
},
{
"epoch": 1.9335260115606936,
"grad_norm": 0.6302561443016468,
"learning_rate": 1.5396374151734564e-05,
"loss": 0.5841,
"step": 669
},
{
"epoch": 1.9364161849710984,
"grad_norm": 0.6704642962297728,
"learning_rate": 1.5379376029319525e-05,
"loss": 0.5891,
"step": 670
},
{
"epoch": 1.939306358381503,
"grad_norm": 0.7314361115096814,
"learning_rate": 1.5362356006392073e-05,
"loss": 0.591,
"step": 671
},
{
"epoch": 1.9421965317919074,
"grad_norm": 0.7058800141993237,
"learning_rate": 1.534531415224411e-05,
"loss": 0.578,
"step": 672
},
{
"epoch": 1.9450867052023122,
"grad_norm": 0.8166135309836842,
"learning_rate": 1.532825053625643e-05,
"loss": 0.5933,
"step": 673
},
{
"epoch": 1.9479768786127167,
"grad_norm": 0.752724434644058,
"learning_rate": 1.5311165227898407e-05,
"loss": 0.5771,
"step": 674
},
{
"epoch": 1.9508670520231215,
"grad_norm": 0.6785478290130587,
"learning_rate": 1.529405829672775e-05,
"loss": 0.5889,
"step": 675
},
{
"epoch": 1.953757225433526,
"grad_norm": 0.7944689622366047,
"learning_rate": 1.527692981239017e-05,
"loss": 0.5785,
"step": 676
},
{
"epoch": 1.9566473988439306,
"grad_norm": 0.7205575606177707,
"learning_rate": 1.5259779844619152e-05,
"loss": 0.5993,
"step": 677
},
{
"epoch": 1.9595375722543351,
"grad_norm": 0.7506990852727408,
"learning_rate": 1.5242608463235638e-05,
"loss": 0.5646,
"step": 678
},
{
"epoch": 1.9624277456647399,
"grad_norm": 0.7726824362277099,
"learning_rate": 1.5225415738147735e-05,
"loss": 0.5786,
"step": 679
},
{
"epoch": 1.9653179190751446,
"grad_norm": 0.7716448033850631,
"learning_rate": 1.520820173935046e-05,
"loss": 0.5803,
"step": 680
},
{
"epoch": 1.9682080924855492,
"grad_norm": 0.7829292348981737,
"learning_rate": 1.5190966536925445e-05,
"loss": 0.5425,
"step": 681
},
{
"epoch": 1.9710982658959537,
"grad_norm": 0.7948589901415222,
"learning_rate": 1.5173710201040615e-05,
"loss": 0.5952,
"step": 682
},
{
"epoch": 1.9739884393063583,
"grad_norm": 0.7215578037880686,
"learning_rate": 1.5156432801949971e-05,
"loss": 0.5704,
"step": 683
},
{
"epoch": 1.976878612716763,
"grad_norm": 0.7284053994843306,
"learning_rate": 1.5139134409993239e-05,
"loss": 0.5972,
"step": 684
},
{
"epoch": 1.9797687861271678,
"grad_norm": 0.7013744195894402,
"learning_rate": 1.5121815095595631e-05,
"loss": 0.6006,
"step": 685
},
{
"epoch": 1.9826589595375723,
"grad_norm": 0.6538513212308937,
"learning_rate": 1.510447492926752e-05,
"loss": 0.5731,
"step": 686
},
{
"epoch": 1.9855491329479769,
"grad_norm": 0.62634361862159,
"learning_rate": 1.508711398160419e-05,
"loss": 0.5587,
"step": 687
},
{
"epoch": 1.9884393063583814,
"grad_norm": 0.6596385884566939,
"learning_rate": 1.5069732323285511e-05,
"loss": 0.5756,
"step": 688
},
{
"epoch": 1.9913294797687862,
"grad_norm": 0.6799960306298413,
"learning_rate": 1.5052330025075687e-05,
"loss": 0.5736,
"step": 689
},
{
"epoch": 1.9942196531791907,
"grad_norm": 0.7728857944499702,
"learning_rate": 1.5034907157822943e-05,
"loss": 0.5829,
"step": 690
},
{
"epoch": 1.9971098265895955,
"grad_norm": 0.7361233644197462,
"learning_rate": 1.501746379245924e-05,
"loss": 0.5655,
"step": 691
},
{
"epoch": 2.0,
"grad_norm": 0.7892457565262699,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.5565,
"step": 692
},
{
"epoch": 2.0028901734104045,
"grad_norm": 0.8917557233017886,
"learning_rate": 1.4982515851543808e-05,
"loss": 0.49,
"step": 693
},
{
"epoch": 2.005780346820809,
"grad_norm": 0.731826726927124,
"learning_rate": 1.4965011418272115e-05,
"loss": 0.5039,
"step": 694
},
{
"epoch": 2.008670520231214,
"grad_norm": 0.9584484699434676,
"learning_rate": 1.4947486771448955e-05,
"loss": 0.492,
"step": 695
},
{
"epoch": 2.0115606936416186,
"grad_norm": 0.7924464564666959,
"learning_rate": 1.4929941982420665e-05,
"loss": 0.4976,
"step": 696
},
{
"epoch": 2.014450867052023,
"grad_norm": 0.7694228147087654,
"learning_rate": 1.4912377122615582e-05,
"loss": 0.4684,
"step": 697
},
{
"epoch": 2.0173410404624277,
"grad_norm": 0.8337400822313988,
"learning_rate": 1.4894792263543744e-05,
"loss": 0.4821,
"step": 698
},
{
"epoch": 2.020231213872832,
"grad_norm": 0.81377528853195,
"learning_rate": 1.4877187476796626e-05,
"loss": 0.4809,
"step": 699
},
{
"epoch": 2.023121387283237,
"grad_norm": 0.8880995758760973,
"learning_rate": 1.4859562834046821e-05,
"loss": 0.4959,
"step": 700
},
{
"epoch": 2.0260115606936417,
"grad_norm": 0.8842076932638161,
"learning_rate": 1.4841918407047766e-05,
"loss": 0.4838,
"step": 701
},
{
"epoch": 2.0289017341040463,
"grad_norm": 0.8021024993393874,
"learning_rate": 1.482425426763344e-05,
"loss": 0.4743,
"step": 702
},
{
"epoch": 2.031791907514451,
"grad_norm": 0.8562302068077103,
"learning_rate": 1.4806570487718076e-05,
"loss": 0.4999,
"step": 703
},
{
"epoch": 2.0346820809248554,
"grad_norm": 0.6671176985974866,
"learning_rate": 1.478886713929587e-05,
"loss": 0.475,
"step": 704
},
{
"epoch": 2.03757225433526,
"grad_norm": 0.8076830275058707,
"learning_rate": 1.4771144294440682e-05,
"loss": 0.4763,
"step": 705
},
{
"epoch": 2.040462427745665,
"grad_norm": 0.6776188308044887,
"learning_rate": 1.475340202530575e-05,
"loss": 0.4797,
"step": 706
},
{
"epoch": 2.0433526011560694,
"grad_norm": 0.8562085572335955,
"learning_rate": 1.4735640404123384e-05,
"loss": 0.4905,
"step": 707
},
{
"epoch": 2.046242774566474,
"grad_norm": 0.666760122889654,
"learning_rate": 1.4717859503204693e-05,
"loss": 0.4993,
"step": 708
},
{
"epoch": 2.0491329479768785,
"grad_norm": 0.8158052894657422,
"learning_rate": 1.4700059394939259e-05,
"loss": 0.4724,
"step": 709
},
{
"epoch": 2.052023121387283,
"grad_norm": 0.7725444120997191,
"learning_rate": 1.4682240151794881e-05,
"loss": 0.499,
"step": 710
},
{
"epoch": 2.054913294797688,
"grad_norm": 0.7224644944519482,
"learning_rate": 1.4664401846317246e-05,
"loss": 0.4895,
"step": 711
},
{
"epoch": 2.0578034682080926,
"grad_norm": 0.6375736963910287,
"learning_rate": 1.4646544551129658e-05,
"loss": 0.471,
"step": 712
},
{
"epoch": 2.060693641618497,
"grad_norm": 0.6785909939945065,
"learning_rate": 1.4628668338932721e-05,
"loss": 0.476,
"step": 713
},
{
"epoch": 2.0635838150289016,
"grad_norm": 0.6480648234855784,
"learning_rate": 1.461077328250406e-05,
"loss": 0.4755,
"step": 714
},
{
"epoch": 2.066473988439306,
"grad_norm": 0.5973861172716439,
"learning_rate": 1.4592859454698021e-05,
"loss": 0.4741,
"step": 715
},
{
"epoch": 2.069364161849711,
"grad_norm": 0.6341756244691673,
"learning_rate": 1.4574926928445366e-05,
"loss": 0.4755,
"step": 716
},
{
"epoch": 2.0722543352601157,
"grad_norm": 0.6404979607956962,
"learning_rate": 1.4556975776752987e-05,
"loss": 0.4772,
"step": 717
},
{
"epoch": 2.0751445086705202,
"grad_norm": 0.6990360916968367,
"learning_rate": 1.4539006072703596e-05,
"loss": 0.4997,
"step": 718
},
{
"epoch": 2.078034682080925,
"grad_norm": 0.587308996267915,
"learning_rate": 1.4521017889455449e-05,
"loss": 0.481,
"step": 719
},
{
"epoch": 2.0809248554913293,
"grad_norm": 0.6861567135110433,
"learning_rate": 1.4503011300242023e-05,
"loss": 0.4837,
"step": 720
},
{
"epoch": 2.0838150289017343,
"grad_norm": 0.5852818001000318,
"learning_rate": 1.4484986378371733e-05,
"loss": 0.5066,
"step": 721
},
{
"epoch": 2.086705202312139,
"grad_norm": 0.613823797692949,
"learning_rate": 1.446694319722763e-05,
"loss": 0.4941,
"step": 722
},
{
"epoch": 2.0895953757225434,
"grad_norm": 0.6226013333663201,
"learning_rate": 1.44488818302671e-05,
"loss": 0.487,
"step": 723
},
{
"epoch": 2.092485549132948,
"grad_norm": 0.6031779435865413,
"learning_rate": 1.4430802351021576e-05,
"loss": 0.4789,
"step": 724
},
{
"epoch": 2.0953757225433525,
"grad_norm": 0.667404269101217,
"learning_rate": 1.4412704833096215e-05,
"loss": 0.4956,
"step": 725
},
{
"epoch": 2.098265895953757,
"grad_norm": 0.6511613836239132,
"learning_rate": 1.4394589350169628e-05,
"loss": 0.4893,
"step": 726
},
{
"epoch": 2.101156069364162,
"grad_norm": 0.7171807632032887,
"learning_rate": 1.4376455975993553e-05,
"loss": 0.4953,
"step": 727
},
{
"epoch": 2.1040462427745665,
"grad_norm": 0.651541501340034,
"learning_rate": 1.4358304784392569e-05,
"loss": 0.4794,
"step": 728
},
{
"epoch": 2.106936416184971,
"grad_norm": 0.6388598045623446,
"learning_rate": 1.4340135849263808e-05,
"loss": 0.4831,
"step": 729
},
{
"epoch": 2.1098265895953756,
"grad_norm": 0.5985686832414758,
"learning_rate": 1.4321949244576616e-05,
"loss": 0.4699,
"step": 730
},
{
"epoch": 2.11271676300578,
"grad_norm": 0.6741219856319731,
"learning_rate": 1.4303745044372293e-05,
"loss": 0.4803,
"step": 731
},
{
"epoch": 2.115606936416185,
"grad_norm": 0.6241553007984232,
"learning_rate": 1.4285523322763763e-05,
"loss": 0.4901,
"step": 732
},
{
"epoch": 2.1184971098265897,
"grad_norm": 0.6342238231670317,
"learning_rate": 1.4267284153935295e-05,
"loss": 0.4857,
"step": 733
},
{
"epoch": 2.121387283236994,
"grad_norm": 0.6615212499200348,
"learning_rate": 1.4249027612142174e-05,
"loss": 0.494,
"step": 734
},
{
"epoch": 2.1242774566473988,
"grad_norm": 0.6287401379522876,
"learning_rate": 1.4230753771710428e-05,
"loss": 0.4641,
"step": 735
},
{
"epoch": 2.1271676300578033,
"grad_norm": 0.6156860490644589,
"learning_rate": 1.4212462707036501e-05,
"loss": 0.4755,
"step": 736
},
{
"epoch": 2.1300578034682083,
"grad_norm": 0.8425617665798458,
"learning_rate": 1.4194154492586967e-05,
"loss": 0.4899,
"step": 737
},
{
"epoch": 2.132947976878613,
"grad_norm": 0.610199754195904,
"learning_rate": 1.4175829202898216e-05,
"loss": 0.475,
"step": 738
},
{
"epoch": 2.1358381502890174,
"grad_norm": 0.6664221509260175,
"learning_rate": 1.415748691257616e-05,
"loss": 0.4632,
"step": 739
},
{
"epoch": 2.138728323699422,
"grad_norm": 0.6198598713665869,
"learning_rate": 1.4139127696295913e-05,
"loss": 0.492,
"step": 740
},
{
"epoch": 2.1416184971098264,
"grad_norm": 0.6627895786844057,
"learning_rate": 1.4120751628801512e-05,
"loss": 0.4887,
"step": 741
},
{
"epoch": 2.1445086705202314,
"grad_norm": 0.6174039399372138,
"learning_rate": 1.4102358784905587e-05,
"loss": 0.4965,
"step": 742
},
{
"epoch": 2.147398843930636,
"grad_norm": 0.6139937983480757,
"learning_rate": 1.4083949239489077e-05,
"loss": 0.474,
"step": 743
},
{
"epoch": 2.1502890173410405,
"grad_norm": 0.6250851147746078,
"learning_rate": 1.4065523067500905e-05,
"loss": 0.4651,
"step": 744
},
{
"epoch": 2.153179190751445,
"grad_norm": 0.6141885023959468,
"learning_rate": 1.4047080343957701e-05,
"loss": 0.4856,
"step": 745
},
{
"epoch": 2.1560693641618496,
"grad_norm": 0.6398962354405308,
"learning_rate": 1.4028621143943463e-05,
"loss": 0.4617,
"step": 746
},
{
"epoch": 2.1589595375722546,
"grad_norm": 0.5634682725662835,
"learning_rate": 1.4010145542609277e-05,
"loss": 0.469,
"step": 747
},
{
"epoch": 2.161849710982659,
"grad_norm": 0.6086661562226492,
"learning_rate": 1.3991653615173002e-05,
"loss": 0.4704,
"step": 748
},
{
"epoch": 2.1647398843930636,
"grad_norm": 0.6557665168728067,
"learning_rate": 1.3973145436918957e-05,
"loss": 0.4605,
"step": 749
},
{
"epoch": 2.167630057803468,
"grad_norm": 0.5541968389544825,
"learning_rate": 1.3954621083197629e-05,
"loss": 0.4871,
"step": 750
},
{
"epoch": 2.1705202312138727,
"grad_norm": 0.6466833656315703,
"learning_rate": 1.3936080629425356e-05,
"loss": 0.4775,
"step": 751
},
{
"epoch": 2.1734104046242773,
"grad_norm": 0.6256229255740271,
"learning_rate": 1.3917524151084019e-05,
"loss": 0.5013,
"step": 752
},
{
"epoch": 2.1763005780346822,
"grad_norm": 0.5926214989262687,
"learning_rate": 1.3898951723720743e-05,
"loss": 0.4802,
"step": 753
},
{
"epoch": 2.179190751445087,
"grad_norm": 0.7398296579967744,
"learning_rate": 1.388036342294759e-05,
"loss": 0.472,
"step": 754
},
{
"epoch": 2.1820809248554913,
"grad_norm": 0.6205573840232363,
"learning_rate": 1.3861759324441225e-05,
"loss": 0.4813,
"step": 755
},
{
"epoch": 2.184971098265896,
"grad_norm": 0.7578188848578328,
"learning_rate": 1.3843139503942653e-05,
"loss": 0.4745,
"step": 756
},
{
"epoch": 2.1878612716763004,
"grad_norm": 0.6298054483595766,
"learning_rate": 1.3824504037256872e-05,
"loss": 0.5086,
"step": 757
},
{
"epoch": 2.1907514450867054,
"grad_norm": 0.6325488679123493,
"learning_rate": 1.3805853000252584e-05,
"loss": 0.4926,
"step": 758
},
{
"epoch": 2.19364161849711,
"grad_norm": 0.6047656801596065,
"learning_rate": 1.3787186468861875e-05,
"loss": 0.4763,
"step": 759
},
{
"epoch": 2.1965317919075145,
"grad_norm": 0.7132919959963034,
"learning_rate": 1.3768504519079923e-05,
"loss": 0.5122,
"step": 760
},
{
"epoch": 2.199421965317919,
"grad_norm": 0.6238046703872717,
"learning_rate": 1.374980722696467e-05,
"loss": 0.4685,
"step": 761
},
{
"epoch": 2.2023121387283235,
"grad_norm": 0.6243010760642064,
"learning_rate": 1.3731094668636516e-05,
"loss": 0.4916,
"step": 762
},
{
"epoch": 2.2052023121387285,
"grad_norm": 0.6843853662389136,
"learning_rate": 1.3712366920278017e-05,
"loss": 0.477,
"step": 763
},
{
"epoch": 2.208092485549133,
"grad_norm": 0.5871073798328166,
"learning_rate": 1.3693624058133575e-05,
"loss": 0.4677,
"step": 764
},
{
"epoch": 2.2109826589595376,
"grad_norm": 0.7228739280407132,
"learning_rate": 1.3674866158509116e-05,
"loss": 0.487,
"step": 765
},
{
"epoch": 2.213872832369942,
"grad_norm": 0.6275057686003038,
"learning_rate": 1.365609329777179e-05,
"loss": 0.488,
"step": 766
},
{
"epoch": 2.2167630057803467,
"grad_norm": 0.6804954453923439,
"learning_rate": 1.3637305552349656e-05,
"loss": 0.4942,
"step": 767
},
{
"epoch": 2.2196531791907512,
"grad_norm": 0.6318708333825124,
"learning_rate": 1.3618502998731372e-05,
"loss": 0.4957,
"step": 768
},
{
"epoch": 2.222543352601156,
"grad_norm": 0.6122840766568456,
"learning_rate": 1.3599685713465882e-05,
"loss": 0.472,
"step": 769
},
{
"epoch": 2.2254335260115607,
"grad_norm": 0.6265270578425641,
"learning_rate": 1.358085377316211e-05,
"loss": 0.4866,
"step": 770
},
{
"epoch": 2.2283236994219653,
"grad_norm": 0.6832285523902126,
"learning_rate": 1.3562007254488634e-05,
"loss": 0.477,
"step": 771
},
{
"epoch": 2.23121387283237,
"grad_norm": 0.5864628346946127,
"learning_rate": 1.3543146234173395e-05,
"loss": 0.4772,
"step": 772
},
{
"epoch": 2.2341040462427744,
"grad_norm": 0.6951998969486958,
"learning_rate": 1.3524270789003362e-05,
"loss": 0.4999,
"step": 773
},
{
"epoch": 2.2369942196531793,
"grad_norm": 0.5765080186621816,
"learning_rate": 1.350538099582424e-05,
"loss": 0.4799,
"step": 774
},
{
"epoch": 2.239884393063584,
"grad_norm": 0.6479445268775318,
"learning_rate": 1.3486476931540145e-05,
"loss": 0.4889,
"step": 775
},
{
"epoch": 2.2427745664739884,
"grad_norm": 0.6266173892843901,
"learning_rate": 1.3467558673113286e-05,
"loss": 0.4754,
"step": 776
},
{
"epoch": 2.245664739884393,
"grad_norm": 0.5691738042569816,
"learning_rate": 1.344862629756367e-05,
"loss": 0.4905,
"step": 777
},
{
"epoch": 2.2485549132947975,
"grad_norm": 0.6745684340820571,
"learning_rate": 1.342967988196877e-05,
"loss": 0.4878,
"step": 778
},
{
"epoch": 2.2514450867052025,
"grad_norm": 0.5991121170038285,
"learning_rate": 1.3410719503463225e-05,
"loss": 0.5049,
"step": 779
},
{
"epoch": 2.254335260115607,
"grad_norm": 0.6662560641610981,
"learning_rate": 1.3391745239238509e-05,
"loss": 0.4745,
"step": 780
},
{
"epoch": 2.2572254335260116,
"grad_norm": 0.7675180134150684,
"learning_rate": 1.3372757166542647e-05,
"loss": 0.4982,
"step": 781
},
{
"epoch": 2.260115606936416,
"grad_norm": 0.5630850126003241,
"learning_rate": 1.3353755362679856e-05,
"loss": 0.4794,
"step": 782
},
{
"epoch": 2.2630057803468207,
"grad_norm": 0.7541638366537354,
"learning_rate": 1.333473990501028e-05,
"loss": 0.4716,
"step": 783
},
{
"epoch": 2.2658959537572256,
"grad_norm": 0.6226190873894291,
"learning_rate": 1.3315710870949632e-05,
"loss": 0.4982,
"step": 784
},
{
"epoch": 2.26878612716763,
"grad_norm": 0.6071189139845049,
"learning_rate": 1.3296668337968904e-05,
"loss": 0.4954,
"step": 785
},
{
"epoch": 2.2716763005780347,
"grad_norm": 0.6478422528890034,
"learning_rate": 1.3277612383594045e-05,
"loss": 0.4836,
"step": 786
},
{
"epoch": 2.2745664739884393,
"grad_norm": 0.5724767544283145,
"learning_rate": 1.3258543085405649e-05,
"loss": 0.4822,
"step": 787
},
{
"epoch": 2.277456647398844,
"grad_norm": 0.6411210199107057,
"learning_rate": 1.3239460521038626e-05,
"loss": 0.4791,
"step": 788
},
{
"epoch": 2.2803468208092488,
"grad_norm": 0.58890510940359,
"learning_rate": 1.3220364768181901e-05,
"loss": 0.4837,
"step": 789
},
{
"epoch": 2.2832369942196533,
"grad_norm": 0.6087119937815791,
"learning_rate": 1.3201255904578094e-05,
"loss": 0.4758,
"step": 790
},
{
"epoch": 2.286127167630058,
"grad_norm": 0.580194410281497,
"learning_rate": 1.3182134008023198e-05,
"loss": 0.4655,
"step": 791
},
{
"epoch": 2.2890173410404624,
"grad_norm": 0.6231249120490117,
"learning_rate": 1.3162999156366265e-05,
"loss": 0.4724,
"step": 792
},
{
"epoch": 2.291907514450867,
"grad_norm": 0.5683956777923486,
"learning_rate": 1.3143851427509093e-05,
"loss": 0.4944,
"step": 793
},
{
"epoch": 2.294797687861272,
"grad_norm": 0.5813223298187833,
"learning_rate": 1.3124690899405903e-05,
"loss": 0.4716,
"step": 794
},
{
"epoch": 2.2976878612716765,
"grad_norm": 0.5910053696331238,
"learning_rate": 1.3105517650063026e-05,
"loss": 0.48,
"step": 795
},
{
"epoch": 2.300578034682081,
"grad_norm": 0.5717961169940149,
"learning_rate": 1.3086331757538582e-05,
"loss": 0.4843,
"step": 796
},
{
"epoch": 2.3034682080924855,
"grad_norm": 0.5610354183402453,
"learning_rate": 1.3067133299942172e-05,
"loss": 0.4785,
"step": 797
},
{
"epoch": 2.30635838150289,
"grad_norm": 0.6113141495689187,
"learning_rate": 1.3047922355434533e-05,
"loss": 0.4818,
"step": 798
},
{
"epoch": 2.3092485549132946,
"grad_norm": 0.5861183126949195,
"learning_rate": 1.3028699002227256e-05,
"loss": 0.4963,
"step": 799
},
{
"epoch": 2.3121387283236996,
"grad_norm": 0.6521803246079821,
"learning_rate": 1.3009463318582448e-05,
"loss": 0.4724,
"step": 800
},
{
"epoch": 2.315028901734104,
"grad_norm": 0.6034099548310038,
"learning_rate": 1.2990215382812408e-05,
"loss": 0.4703,
"step": 801
},
{
"epoch": 2.3179190751445087,
"grad_norm": 0.6878646284940723,
"learning_rate": 1.2970955273279324e-05,
"loss": 0.4742,
"step": 802
},
{
"epoch": 2.320809248554913,
"grad_norm": 0.6163785271184989,
"learning_rate": 1.2951683068394941e-05,
"loss": 0.4877,
"step": 803
},
{
"epoch": 2.3236994219653178,
"grad_norm": 0.7017702766508197,
"learning_rate": 1.2932398846620244e-05,
"loss": 0.4985,
"step": 804
},
{
"epoch": 2.3265895953757223,
"grad_norm": 0.6514042902173288,
"learning_rate": 1.291310268646515e-05,
"loss": 0.4869,
"step": 805
},
{
"epoch": 2.3294797687861273,
"grad_norm": 0.6582908602483947,
"learning_rate": 1.2893794666488175e-05,
"loss": 0.4681,
"step": 806
},
{
"epoch": 2.332369942196532,
"grad_norm": 0.9008026884763592,
"learning_rate": 1.2874474865296112e-05,
"loss": 0.4951,
"step": 807
},
{
"epoch": 2.3352601156069364,
"grad_norm": 0.7098416272580851,
"learning_rate": 1.285514336154373e-05,
"loss": 0.4928,
"step": 808
},
{
"epoch": 2.338150289017341,
"grad_norm": 0.7763595007460438,
"learning_rate": 1.283580023393343e-05,
"loss": 0.4944,
"step": 809
},
{
"epoch": 2.3410404624277454,
"grad_norm": 0.6153457741221146,
"learning_rate": 1.2816445561214947e-05,
"loss": 0.5017,
"step": 810
},
{
"epoch": 2.3439306358381504,
"grad_norm": 0.6256664364781374,
"learning_rate": 1.2797079422185013e-05,
"loss": 0.5045,
"step": 811
},
{
"epoch": 2.346820809248555,
"grad_norm": 0.6221766375998093,
"learning_rate": 1.2777701895687034e-05,
"loss": 0.4917,
"step": 812
},
{
"epoch": 2.3497109826589595,
"grad_norm": 0.576313700159756,
"learning_rate": 1.275831306061079e-05,
"loss": 0.489,
"step": 813
},
{
"epoch": 2.352601156069364,
"grad_norm": 0.6741249523223151,
"learning_rate": 1.2738912995892095e-05,
"loss": 0.4737,
"step": 814
},
{
"epoch": 2.3554913294797686,
"grad_norm": 0.6292357151471923,
"learning_rate": 1.2719501780512475e-05,
"loss": 0.4832,
"step": 815
},
{
"epoch": 2.3583815028901736,
"grad_norm": 0.5467682217417565,
"learning_rate": 1.2700079493498864e-05,
"loss": 0.4887,
"step": 816
},
{
"epoch": 2.361271676300578,
"grad_norm": 0.6152550291647984,
"learning_rate": 1.2680646213923264e-05,
"loss": 0.4892,
"step": 817
},
{
"epoch": 2.3641618497109826,
"grad_norm": 0.6023318829337962,
"learning_rate": 1.2661202020902432e-05,
"loss": 0.4617,
"step": 818
},
{
"epoch": 2.367052023121387,
"grad_norm": 0.6204224802741845,
"learning_rate": 1.2641746993597547e-05,
"loss": 0.4812,
"step": 819
},
{
"epoch": 2.3699421965317917,
"grad_norm": 0.5972288325305012,
"learning_rate": 1.2622281211213916e-05,
"loss": 0.4897,
"step": 820
},
{
"epoch": 2.3728323699421967,
"grad_norm": 0.6005242709543999,
"learning_rate": 1.2602804753000611e-05,
"loss": 0.4961,
"step": 821
},
{
"epoch": 2.3757225433526012,
"grad_norm": 0.6075305884093503,
"learning_rate": 1.2583317698250184e-05,
"loss": 0.496,
"step": 822
},
{
"epoch": 2.378612716763006,
"grad_norm": 0.6080984427620649,
"learning_rate": 1.2563820126298313e-05,
"loss": 0.5099,
"step": 823
},
{
"epoch": 2.3815028901734103,
"grad_norm": 0.6183012204818146,
"learning_rate": 1.2544312116523508e-05,
"loss": 0.5213,
"step": 824
},
{
"epoch": 2.384393063583815,
"grad_norm": 0.6124206661136334,
"learning_rate": 1.252479374834676e-05,
"loss": 0.482,
"step": 825
},
{
"epoch": 2.38728323699422,
"grad_norm": 0.6338531257566175,
"learning_rate": 1.2505265101231243e-05,
"loss": 0.4884,
"step": 826
},
{
"epoch": 2.3901734104046244,
"grad_norm": 0.582075543595997,
"learning_rate": 1.2485726254681971e-05,
"loss": 0.5121,
"step": 827
},
{
"epoch": 2.393063583815029,
"grad_norm": 0.6033739430944867,
"learning_rate": 1.2466177288245486e-05,
"loss": 0.4869,
"step": 828
},
{
"epoch": 2.3959537572254335,
"grad_norm": 0.5995179007325128,
"learning_rate": 1.2446618281509526e-05,
"loss": 0.4728,
"step": 829
},
{
"epoch": 2.398843930635838,
"grad_norm": 0.5561904176522109,
"learning_rate": 1.2427049314102708e-05,
"loss": 0.5027,
"step": 830
},
{
"epoch": 2.401734104046243,
"grad_norm": 0.5803956118986667,
"learning_rate": 1.24074704656942e-05,
"loss": 0.4874,
"step": 831
},
{
"epoch": 2.4046242774566475,
"grad_norm": 0.6133935938899633,
"learning_rate": 1.2387881815993396e-05,
"loss": 0.472,
"step": 832
},
{
"epoch": 2.407514450867052,
"grad_norm": 0.5570740395566359,
"learning_rate": 1.2368283444749602e-05,
"loss": 0.4774,
"step": 833
},
{
"epoch": 2.4104046242774566,
"grad_norm": 0.6263728298267713,
"learning_rate": 1.2348675431751685e-05,
"loss": 0.4907,
"step": 834
},
{
"epoch": 2.413294797687861,
"grad_norm": 0.659842661942265,
"learning_rate": 1.2329057856827781e-05,
"loss": 0.4961,
"step": 835
},
{
"epoch": 2.416184971098266,
"grad_norm": 0.6564033064344027,
"learning_rate": 1.230943079984495e-05,
"loss": 0.4588,
"step": 836
},
{
"epoch": 2.4190751445086707,
"grad_norm": 0.6595577915231365,
"learning_rate": 1.2289794340708855e-05,
"loss": 0.4797,
"step": 837
},
{
"epoch": 2.421965317919075,
"grad_norm": 0.6674280365531452,
"learning_rate": 1.2270148559363429e-05,
"loss": 0.4848,
"step": 838
},
{
"epoch": 2.4248554913294798,
"grad_norm": 0.6890891118583778,
"learning_rate": 1.2250493535790574e-05,
"loss": 0.4943,
"step": 839
},
{
"epoch": 2.4277456647398843,
"grad_norm": 0.6128280313168251,
"learning_rate": 1.2230829350009805e-05,
"loss": 0.4873,
"step": 840
},
{
"epoch": 2.430635838150289,
"grad_norm": 0.7196046761480148,
"learning_rate": 1.2211156082077945e-05,
"loss": 0.4816,
"step": 841
},
{
"epoch": 2.433526011560694,
"grad_norm": 0.5682730567737198,
"learning_rate": 1.219147381208879e-05,
"loss": 0.4851,
"step": 842
},
{
"epoch": 2.4364161849710984,
"grad_norm": 0.7132152076959606,
"learning_rate": 1.2171782620172787e-05,
"loss": 0.4909,
"step": 843
},
{
"epoch": 2.439306358381503,
"grad_norm": 0.6805602468690977,
"learning_rate": 1.2152082586496701e-05,
"loss": 0.487,
"step": 844
},
{
"epoch": 2.4421965317919074,
"grad_norm": 0.6756276813728793,
"learning_rate": 1.21323737912633e-05,
"loss": 0.4845,
"step": 845
},
{
"epoch": 2.445086705202312,
"grad_norm": 0.6410611739001532,
"learning_rate": 1.2112656314711017e-05,
"loss": 0.4819,
"step": 846
},
{
"epoch": 2.447976878612717,
"grad_norm": 0.6613685466896656,
"learning_rate": 1.2092930237113637e-05,
"loss": 0.5023,
"step": 847
},
{
"epoch": 2.4508670520231215,
"grad_norm": 0.6028188017759273,
"learning_rate": 1.2073195638779944e-05,
"loss": 0.5049,
"step": 848
},
{
"epoch": 2.453757225433526,
"grad_norm": 0.5882497100696249,
"learning_rate": 1.2053452600053435e-05,
"loss": 0.4728,
"step": 849
},
{
"epoch": 2.4566473988439306,
"grad_norm": 0.6097538357238959,
"learning_rate": 1.2033701201311945e-05,
"loss": 0.4936,
"step": 850
},
{
"epoch": 2.459537572254335,
"grad_norm": 0.6036063170459649,
"learning_rate": 1.2013941522967365e-05,
"loss": 0.4718,
"step": 851
},
{
"epoch": 2.4624277456647397,
"grad_norm": 0.5859989981371178,
"learning_rate": 1.1994173645465283e-05,
"loss": 0.4912,
"step": 852
},
{
"epoch": 2.4653179190751446,
"grad_norm": 0.5423461898525301,
"learning_rate": 1.197439764928467e-05,
"loss": 0.4851,
"step": 853
},
{
"epoch": 2.468208092485549,
"grad_norm": 0.6108838211505494,
"learning_rate": 1.1954613614937549e-05,
"loss": 0.4621,
"step": 854
},
{
"epoch": 2.4710982658959537,
"grad_norm": 0.5689087773053834,
"learning_rate": 1.193482162296867e-05,
"loss": 0.4786,
"step": 855
},
{
"epoch": 2.4739884393063583,
"grad_norm": 0.7301557323716112,
"learning_rate": 1.1915021753955178e-05,
"loss": 0.4852,
"step": 856
},
{
"epoch": 2.476878612716763,
"grad_norm": 0.5652516257706838,
"learning_rate": 1.1895214088506284e-05,
"loss": 0.4906,
"step": 857
},
{
"epoch": 2.479768786127168,
"grad_norm": 0.5913465429932605,
"learning_rate": 1.187539870726295e-05,
"loss": 0.464,
"step": 858
},
{
"epoch": 2.4826589595375723,
"grad_norm": 0.6112116690553486,
"learning_rate": 1.185557569089754e-05,
"loss": 0.5007,
"step": 859
},
{
"epoch": 2.485549132947977,
"grad_norm": 0.544535409192471,
"learning_rate": 1.183574512011351e-05,
"loss": 0.4896,
"step": 860
},
{
"epoch": 2.4884393063583814,
"grad_norm": 0.5999298855888491,
"learning_rate": 1.181590707564506e-05,
"loss": 0.4689,
"step": 861
},
{
"epoch": 2.491329479768786,
"grad_norm": 0.63230318229943,
"learning_rate": 1.1796061638256831e-05,
"loss": 0.4984,
"step": 862
},
{
"epoch": 2.494219653179191,
"grad_norm": 0.5875879683407528,
"learning_rate": 1.1776208888743554e-05,
"loss": 0.4876,
"step": 863
},
{
"epoch": 2.4971098265895955,
"grad_norm": 0.5740009661133425,
"learning_rate": 1.1756348907929733e-05,
"loss": 0.4996,
"step": 864
},
{
"epoch": 2.5,
"grad_norm": 0.6467715034441753,
"learning_rate": 1.1736481776669307e-05,
"loss": 0.4899,
"step": 865
},
{
"epoch": 2.5028901734104045,
"grad_norm": 0.6658254556399299,
"learning_rate": 1.1716607575845327e-05,
"loss": 0.4849,
"step": 866
},
{
"epoch": 2.505780346820809,
"grad_norm": 0.6304554371559444,
"learning_rate": 1.1696726386369627e-05,
"loss": 0.4749,
"step": 867
},
{
"epoch": 2.508670520231214,
"grad_norm": 0.553572005776321,
"learning_rate": 1.1676838289182502e-05,
"loss": 0.4692,
"step": 868
},
{
"epoch": 2.5115606936416186,
"grad_norm": 0.5989811250733211,
"learning_rate": 1.1656943365252351e-05,
"loss": 0.491,
"step": 869
},
{
"epoch": 2.514450867052023,
"grad_norm": 0.5750522619417978,
"learning_rate": 1.1637041695575383e-05,
"loss": 0.4971,
"step": 870
},
{
"epoch": 2.5173410404624277,
"grad_norm": 0.6037262595078183,
"learning_rate": 1.1617133361175257e-05,
"loss": 0.472,
"step": 871
},
{
"epoch": 2.520231213872832,
"grad_norm": 0.6374098762461646,
"learning_rate": 1.159721844310278e-05,
"loss": 0.5009,
"step": 872
},
{
"epoch": 2.523121387283237,
"grad_norm": 0.5574371831910762,
"learning_rate": 1.1577297022435548e-05,
"loss": 0.4644,
"step": 873
},
{
"epoch": 2.5260115606936417,
"grad_norm": 0.6285846684044717,
"learning_rate": 1.155736918027764e-05,
"loss": 0.4657,
"step": 874
},
{
"epoch": 2.5289017341040463,
"grad_norm": 0.5680264691810969,
"learning_rate": 1.153743499775927e-05,
"loss": 0.5,
"step": 875
},
{
"epoch": 2.531791907514451,
"grad_norm": 0.6392558071023373,
"learning_rate": 1.1517494556036477e-05,
"loss": 0.5144,
"step": 876
},
{
"epoch": 2.5346820809248554,
"grad_norm": 0.5635030096595876,
"learning_rate": 1.149754793629077e-05,
"loss": 0.4854,
"step": 877
},
{
"epoch": 2.5375722543352603,
"grad_norm": 0.5188679966546239,
"learning_rate": 1.1477595219728817e-05,
"loss": 0.4893,
"step": 878
},
{
"epoch": 2.540462427745665,
"grad_norm": 0.5485891984650414,
"learning_rate": 1.1457636487582104e-05,
"loss": 0.4676,
"step": 879
},
{
"epoch": 2.5433526011560694,
"grad_norm": 0.5917527574622287,
"learning_rate": 1.143767182110661e-05,
"loss": 0.5032,
"step": 880
},
{
"epoch": 2.546242774566474,
"grad_norm": 0.5373968873975472,
"learning_rate": 1.1417701301582476e-05,
"loss": 0.4932,
"step": 881
},
{
"epoch": 2.5491329479768785,
"grad_norm": 0.5661442996520238,
"learning_rate": 1.1397725010313666e-05,
"loss": 0.4959,
"step": 882
},
{
"epoch": 2.5520231213872835,
"grad_norm": 0.581891800537539,
"learning_rate": 1.137774302862764e-05,
"loss": 0.499,
"step": 883
},
{
"epoch": 2.5549132947976876,
"grad_norm": 0.5480548231275786,
"learning_rate": 1.135775543787504e-05,
"loss": 0.483,
"step": 884
},
{
"epoch": 2.5578034682080926,
"grad_norm": 0.5677370412015945,
"learning_rate": 1.1337762319429326e-05,
"loss": 0.4628,
"step": 885
},
{
"epoch": 2.560693641618497,
"grad_norm": 0.5586053485925068,
"learning_rate": 1.1317763754686474e-05,
"loss": 0.4548,
"step": 886
},
{
"epoch": 2.5635838150289016,
"grad_norm": 0.5578888374471789,
"learning_rate": 1.1297759825064624e-05,
"loss": 0.5068,
"step": 887
},
{
"epoch": 2.5664739884393066,
"grad_norm": 0.5930380142370744,
"learning_rate": 1.1277750612003767e-05,
"loss": 0.4723,
"step": 888
},
{
"epoch": 2.5693641618497107,
"grad_norm": 0.5963766508065118,
"learning_rate": 1.1257736196965392e-05,
"loss": 0.4845,
"step": 889
},
{
"epoch": 2.5722543352601157,
"grad_norm": 0.5901899876827236,
"learning_rate": 1.123771666143218e-05,
"loss": 0.4797,
"step": 890
},
{
"epoch": 2.5751445086705202,
"grad_norm": 0.605878081712083,
"learning_rate": 1.1217692086907653e-05,
"loss": 0.4768,
"step": 891
},
{
"epoch": 2.578034682080925,
"grad_norm": 0.5815366679890774,
"learning_rate": 1.1197662554915836e-05,
"loss": 0.4831,
"step": 892
},
{
"epoch": 2.5809248554913293,
"grad_norm": 0.5410869254991583,
"learning_rate": 1.1177628147000961e-05,
"loss": 0.4843,
"step": 893
},
{
"epoch": 2.583815028901734,
"grad_norm": 0.5840458561279575,
"learning_rate": 1.1157588944727087e-05,
"loss": 0.5055,
"step": 894
},
{
"epoch": 2.586705202312139,
"grad_norm": 0.5641649806183132,
"learning_rate": 1.113754502967781e-05,
"loss": 0.4994,
"step": 895
},
{
"epoch": 2.5895953757225434,
"grad_norm": 0.5514001572280435,
"learning_rate": 1.1117496483455898e-05,
"loss": 0.4731,
"step": 896
},
{
"epoch": 2.592485549132948,
"grad_norm": 0.5887720885509165,
"learning_rate": 1.1097443387682984e-05,
"loss": 0.4717,
"step": 897
},
{
"epoch": 2.5953757225433525,
"grad_norm": 0.6310085081831622,
"learning_rate": 1.1077385823999218e-05,
"loss": 0.4806,
"step": 898
},
{
"epoch": 2.598265895953757,
"grad_norm": 0.6074205452371207,
"learning_rate": 1.1057323874062942e-05,
"loss": 0.4841,
"step": 899
},
{
"epoch": 2.601156069364162,
"grad_norm": 0.5338180979357099,
"learning_rate": 1.1037257619550353e-05,
"loss": 0.4805,
"step": 900
},
{
"epoch": 2.6040462427745665,
"grad_norm": 0.5762991464150397,
"learning_rate": 1.1017187142155174e-05,
"loss": 0.4719,
"step": 901
},
{
"epoch": 2.606936416184971,
"grad_norm": 0.6045308731955183,
"learning_rate": 1.0997112523588322e-05,
"loss": 0.4924,
"step": 902
},
{
"epoch": 2.6098265895953756,
"grad_norm": 0.6210687094964608,
"learning_rate": 1.0977033845577574e-05,
"loss": 0.4912,
"step": 903
},
{
"epoch": 2.61271676300578,
"grad_norm": 0.5971492004599623,
"learning_rate": 1.0956951189867224e-05,
"loss": 0.4893,
"step": 904
},
{
"epoch": 2.615606936416185,
"grad_norm": 0.5160192763229764,
"learning_rate": 1.0936864638217777e-05,
"loss": 0.4858,
"step": 905
},
{
"epoch": 2.6184971098265897,
"grad_norm": 0.6613318429592981,
"learning_rate": 1.0916774272405581e-05,
"loss": 0.4604,
"step": 906
},
{
"epoch": 2.621387283236994,
"grad_norm": 0.6634111921366653,
"learning_rate": 1.089668017422253e-05,
"loss": 0.4689,
"step": 907
},
{
"epoch": 2.6242774566473988,
"grad_norm": 0.582676908856468,
"learning_rate": 1.0876582425475694e-05,
"loss": 0.4932,
"step": 908
},
{
"epoch": 2.6271676300578033,
"grad_norm": 0.5759991723829018,
"learning_rate": 1.0856481107987024e-05,
"loss": 0.4909,
"step": 909
},
{
"epoch": 2.6300578034682083,
"grad_norm": 0.6602743078348354,
"learning_rate": 1.0836376303592991e-05,
"loss": 0.4854,
"step": 910
},
{
"epoch": 2.632947976878613,
"grad_norm": 0.6242047289159153,
"learning_rate": 1.0816268094144257e-05,
"loss": 0.4885,
"step": 911
},
{
"epoch": 2.6358381502890174,
"grad_norm": 0.6142487967868995,
"learning_rate": 1.0796156561505355e-05,
"loss": 0.4852,
"step": 912
},
{
"epoch": 2.638728323699422,
"grad_norm": 0.6562842223395773,
"learning_rate": 1.0776041787554348e-05,
"loss": 0.4965,
"step": 913
},
{
"epoch": 2.6416184971098264,
"grad_norm": 0.6288923772138604,
"learning_rate": 1.0755923854182484e-05,
"loss": 0.4812,
"step": 914
},
{
"epoch": 2.6445086705202314,
"grad_norm": 0.593632937537067,
"learning_rate": 1.0735802843293888e-05,
"loss": 0.4802,
"step": 915
},
{
"epoch": 2.647398843930636,
"grad_norm": 0.5534510402258848,
"learning_rate": 1.0715678836805204e-05,
"loss": 0.4793,
"step": 916
},
{
"epoch": 2.6502890173410405,
"grad_norm": 0.5820026877478321,
"learning_rate": 1.0695551916645273e-05,
"loss": 0.4899,
"step": 917
},
{
"epoch": 2.653179190751445,
"grad_norm": 0.5938224966700754,
"learning_rate": 1.0675422164754808e-05,
"loss": 0.4823,
"step": 918
},
{
"epoch": 2.6560693641618496,
"grad_norm": 0.5738073249813049,
"learning_rate": 1.065528966308603e-05,
"loss": 0.4726,
"step": 919
},
{
"epoch": 2.6589595375722546,
"grad_norm": 0.5912461023132076,
"learning_rate": 1.063515449360238e-05,
"loss": 0.4905,
"step": 920
},
{
"epoch": 2.661849710982659,
"grad_norm": 0.5913815785583798,
"learning_rate": 1.0615016738278138e-05,
"loss": 0.4794,
"step": 921
},
{
"epoch": 2.6647398843930636,
"grad_norm": 0.5272010239980836,
"learning_rate": 1.059487647909813e-05,
"loss": 0.4904,
"step": 922
},
{
"epoch": 2.667630057803468,
"grad_norm": 0.6053997555307846,
"learning_rate": 1.0574733798057359e-05,
"loss": 0.5063,
"step": 923
},
{
"epoch": 2.6705202312138727,
"grad_norm": 0.6350430971708139,
"learning_rate": 1.05545887771607e-05,
"loss": 0.5098,
"step": 924
},
{
"epoch": 2.6734104046242777,
"grad_norm": 0.5647027666397387,
"learning_rate": 1.0534441498422552e-05,
"loss": 0.4815,
"step": 925
},
{
"epoch": 2.6763005780346822,
"grad_norm": 0.5956593825863362,
"learning_rate": 1.0514292043866498e-05,
"loss": 0.4849,
"step": 926
},
{
"epoch": 2.679190751445087,
"grad_norm": 0.5922804270276485,
"learning_rate": 1.0494140495524989e-05,
"loss": 0.4805,
"step": 927
},
{
"epoch": 2.6820809248554913,
"grad_norm": 0.6253790789940853,
"learning_rate": 1.0473986935438997e-05,
"loss": 0.4874,
"step": 928
},
{
"epoch": 2.684971098265896,
"grad_norm": 0.5946787261865738,
"learning_rate": 1.045383144565768e-05,
"loss": 0.4961,
"step": 929
},
{
"epoch": 2.687861271676301,
"grad_norm": 0.6003433405556385,
"learning_rate": 1.0433674108238058e-05,
"loss": 0.4952,
"step": 930
},
{
"epoch": 2.690751445086705,
"grad_norm": 0.5866411248720449,
"learning_rate": 1.0413515005244673e-05,
"loss": 0.4741,
"step": 931
},
{
"epoch": 2.69364161849711,
"grad_norm": 0.6495719140271753,
"learning_rate": 1.0393354218749247e-05,
"loss": 0.4786,
"step": 932
},
{
"epoch": 2.6965317919075145,
"grad_norm": 0.6131357075260606,
"learning_rate": 1.0373191830830367e-05,
"loss": 0.4779,
"step": 933
},
{
"epoch": 2.699421965317919,
"grad_norm": 0.569930378251785,
"learning_rate": 1.0353027923573134e-05,
"loss": 0.4615,
"step": 934
},
{
"epoch": 2.7023121387283235,
"grad_norm": 0.6128028358275525,
"learning_rate": 1.033286257906883e-05,
"loss": 0.4796,
"step": 935
},
{
"epoch": 2.705202312138728,
"grad_norm": 0.6255418069337046,
"learning_rate": 1.0312695879414598e-05,
"loss": 0.4729,
"step": 936
},
{
"epoch": 2.708092485549133,
"grad_norm": 0.5709860292638275,
"learning_rate": 1.0292527906713092e-05,
"loss": 0.4711,
"step": 937
},
{
"epoch": 2.7109826589595376,
"grad_norm": 0.5902086869559793,
"learning_rate": 1.0272358743072152e-05,
"loss": 0.4703,
"step": 938
},
{
"epoch": 2.713872832369942,
"grad_norm": 0.5640946249724696,
"learning_rate": 1.0252188470604462e-05,
"loss": 0.4892,
"step": 939
},
{
"epoch": 2.7167630057803467,
"grad_norm": 0.5894758987503583,
"learning_rate": 1.0232017171427225e-05,
"loss": 0.48,
"step": 940
},
{
"epoch": 2.7196531791907512,
"grad_norm": 0.5615067004061257,
"learning_rate": 1.021184492766182e-05,
"loss": 0.4987,
"step": 941
},
{
"epoch": 2.722543352601156,
"grad_norm": 0.5453863485675157,
"learning_rate": 1.0191671821433474e-05,
"loss": 0.4639,
"step": 942
},
{
"epoch": 2.7254335260115607,
"grad_norm": 0.6129944896000652,
"learning_rate": 1.0171497934870927e-05,
"loss": 0.4963,
"step": 943
},
{
"epoch": 2.7283236994219653,
"grad_norm": 0.5875588066436983,
"learning_rate": 1.0151323350106088e-05,
"loss": 0.4985,
"step": 944
},
{
"epoch": 2.73121387283237,
"grad_norm": 0.5679917028336635,
"learning_rate": 1.0131148149273724e-05,
"loss": 0.4777,
"step": 945
},
{
"epoch": 2.7341040462427744,
"grad_norm": 0.5778596279001901,
"learning_rate": 1.0110972414511089e-05,
"loss": 0.4752,
"step": 946
},
{
"epoch": 2.7369942196531793,
"grad_norm": 0.5596499004879297,
"learning_rate": 1.0090796227957633e-05,
"loss": 0.4583,
"step": 947
},
{
"epoch": 2.739884393063584,
"grad_norm": 0.6132924162564561,
"learning_rate": 1.0070619671754624e-05,
"loss": 0.486,
"step": 948
},
{
"epoch": 2.7427745664739884,
"grad_norm": 0.5653037104976039,
"learning_rate": 1.0050442828044853e-05,
"loss": 0.4966,
"step": 949
},
{
"epoch": 2.745664739884393,
"grad_norm": 0.5883371672044092,
"learning_rate": 1.003026577897227e-05,
"loss": 0.48,
"step": 950
},
{
"epoch": 2.7485549132947975,
"grad_norm": 0.6451986399775324,
"learning_rate": 1.0010088606681666e-05,
"loss": 0.4679,
"step": 951
},
{
"epoch": 2.7514450867052025,
"grad_norm": 0.5665043387089622,
"learning_rate": 9.989911393318334e-06,
"loss": 0.4627,
"step": 952
},
{
"epoch": 2.754335260115607,
"grad_norm": 0.6220616186244058,
"learning_rate": 9.969734221027732e-06,
"loss": 0.496,
"step": 953
},
{
"epoch": 2.7572254335260116,
"grad_norm": 0.6152265052786257,
"learning_rate": 9.94955717195515e-06,
"loss": 0.4848,
"step": 954
},
{
"epoch": 2.760115606936416,
"grad_norm": 0.5661554513001946,
"learning_rate": 9.929380328245378e-06,
"loss": 0.4872,
"step": 955
},
{
"epoch": 2.7630057803468207,
"grad_norm": 0.6052397841615009,
"learning_rate": 9.909203772042369e-06,
"loss": 0.4748,
"step": 956
},
{
"epoch": 2.7658959537572256,
"grad_norm": 0.591030265881203,
"learning_rate": 9.889027585488914e-06,
"loss": 0.4935,
"step": 957
},
{
"epoch": 2.76878612716763,
"grad_norm": 0.6729411575926139,
"learning_rate": 9.868851850726281e-06,
"loss": 0.4791,
"step": 958
},
{
"epoch": 2.7716763005780347,
"grad_norm": 0.5739839955739133,
"learning_rate": 9.848676649893912e-06,
"loss": 0.4984,
"step": 959
},
{
"epoch": 2.7745664739884393,
"grad_norm": 0.6792832962777284,
"learning_rate": 9.828502065129076e-06,
"loss": 0.4687,
"step": 960
},
{
"epoch": 2.777456647398844,
"grad_norm": 0.6192342735329593,
"learning_rate": 9.808328178566531e-06,
"loss": 0.4772,
"step": 961
},
{
"epoch": 2.7803468208092488,
"grad_norm": 0.5352041062460219,
"learning_rate": 9.788155072338185e-06,
"loss": 0.4875,
"step": 962
},
{
"epoch": 2.7832369942196533,
"grad_norm": 0.6095338051981292,
"learning_rate": 9.767982828572778e-06,
"loss": 0.4924,
"step": 963
},
{
"epoch": 2.786127167630058,
"grad_norm": 0.6221863600074393,
"learning_rate": 9.74781152939554e-06,
"loss": 0.4967,
"step": 964
},
{
"epoch": 2.7890173410404624,
"grad_norm": 0.5988117572685457,
"learning_rate": 9.72764125692785e-06,
"loss": 0.4866,
"step": 965
},
{
"epoch": 2.791907514450867,
"grad_norm": 0.5678508843431816,
"learning_rate": 9.707472093286906e-06,
"loss": 0.4703,
"step": 966
},
{
"epoch": 2.794797687861272,
"grad_norm": 0.7097948148937187,
"learning_rate": 9.687304120585405e-06,
"loss": 0.5112,
"step": 967
},
{
"epoch": 2.7976878612716765,
"grad_norm": 0.6722679540742882,
"learning_rate": 9.667137420931174e-06,
"loss": 0.4931,
"step": 968
},
{
"epoch": 2.800578034682081,
"grad_norm": 0.5801107851006374,
"learning_rate": 9.64697207642687e-06,
"loss": 0.4782,
"step": 969
},
{
"epoch": 2.8034682080924855,
"grad_norm": 0.598003373613563,
"learning_rate": 9.626808169169635e-06,
"loss": 0.4734,
"step": 970
},
{
"epoch": 2.80635838150289,
"grad_norm": 0.6432491971726029,
"learning_rate": 9.606645781250754e-06,
"loss": 0.5032,
"step": 971
},
{
"epoch": 2.809248554913295,
"grad_norm": 0.5811311713045815,
"learning_rate": 9.586484994755332e-06,
"loss": 0.4731,
"step": 972
},
{
"epoch": 2.812138728323699,
"grad_norm": 0.5536057155729461,
"learning_rate": 9.566325891761944e-06,
"loss": 0.4922,
"step": 973
},
{
"epoch": 2.815028901734104,
"grad_norm": 0.5537928759336738,
"learning_rate": 9.546168554342323e-06,
"loss": 0.4901,
"step": 974
},
{
"epoch": 2.8179190751445087,
"grad_norm": 0.5930344405673645,
"learning_rate": 9.526013064561006e-06,
"loss": 0.4901,
"step": 975
},
{
"epoch": 2.820809248554913,
"grad_norm": 0.5781843724894278,
"learning_rate": 9.505859504475015e-06,
"loss": 0.4737,
"step": 976
},
{
"epoch": 2.8236994219653178,
"grad_norm": 0.5517901321295087,
"learning_rate": 9.485707956133502e-06,
"loss": 0.4852,
"step": 977
},
{
"epoch": 2.8265895953757223,
"grad_norm": 0.5720443943430279,
"learning_rate": 9.465558501577451e-06,
"loss": 0.4988,
"step": 978
},
{
"epoch": 2.8294797687861273,
"grad_norm": 0.6164460011881768,
"learning_rate": 9.445411222839301e-06,
"loss": 0.4622,
"step": 979
},
{
"epoch": 2.832369942196532,
"grad_norm": 0.6438301207439332,
"learning_rate": 9.425266201942646e-06,
"loss": 0.4874,
"step": 980
},
{
"epoch": 2.8352601156069364,
"grad_norm": 0.5870456593275332,
"learning_rate": 9.405123520901871e-06,
"loss": 0.4748,
"step": 981
},
{
"epoch": 2.838150289017341,
"grad_norm": 0.6667924069896521,
"learning_rate": 9.384983261721865e-06,
"loss": 0.4682,
"step": 982
},
{
"epoch": 2.8410404624277454,
"grad_norm": 0.5473846946457717,
"learning_rate": 9.364845506397625e-06,
"loss": 0.4857,
"step": 983
},
{
"epoch": 2.8439306358381504,
"grad_norm": 0.5383036248842854,
"learning_rate": 9.344710336913971e-06,
"loss": 0.4799,
"step": 984
},
{
"epoch": 2.846820809248555,
"grad_norm": 0.5949748460882582,
"learning_rate": 9.324577835245197e-06,
"loss": 0.4979,
"step": 985
},
{
"epoch": 2.8497109826589595,
"grad_norm": 0.6178058328203948,
"learning_rate": 9.30444808335473e-06,
"loss": 0.4975,
"step": 986
},
{
"epoch": 2.852601156069364,
"grad_norm": 0.630418096260648,
"learning_rate": 9.284321163194801e-06,
"loss": 0.471,
"step": 987
},
{
"epoch": 2.8554913294797686,
"grad_norm": 0.5495926281924244,
"learning_rate": 9.264197156706115e-06,
"loss": 0.4809,
"step": 988
},
{
"epoch": 2.8583815028901736,
"grad_norm": 0.7086518063216776,
"learning_rate": 9.244076145817518e-06,
"loss": 0.4614,
"step": 989
},
{
"epoch": 2.861271676300578,
"grad_norm": 0.6924691302106323,
"learning_rate": 9.223958212445657e-06,
"loss": 0.5023,
"step": 990
},
{
"epoch": 2.8641618497109826,
"grad_norm": 0.5755074319539322,
"learning_rate": 9.203843438494644e-06,
"loss": 0.5028,
"step": 991
},
{
"epoch": 2.867052023121387,
"grad_norm": 0.7339069507333836,
"learning_rate": 9.183731905855746e-06,
"loss": 0.481,
"step": 992
},
{
"epoch": 2.8699421965317917,
"grad_norm": 0.5573775780123593,
"learning_rate": 9.163623696407014e-06,
"loss": 0.4907,
"step": 993
},
{
"epoch": 2.8728323699421967,
"grad_norm": 0.6062346515497278,
"learning_rate": 9.14351889201298e-06,
"loss": 0.4749,
"step": 994
},
{
"epoch": 2.8757225433526012,
"grad_norm": 0.631590029057518,
"learning_rate": 9.123417574524308e-06,
"loss": 0.4883,
"step": 995
},
{
"epoch": 2.878612716763006,
"grad_norm": 0.5722689231967004,
"learning_rate": 9.103319825777475e-06,
"loss": 0.4915,
"step": 996
},
{
"epoch": 2.8815028901734103,
"grad_norm": 0.6002470468969344,
"learning_rate": 9.083225727594422e-06,
"loss": 0.4914,
"step": 997
},
{
"epoch": 2.884393063583815,
"grad_norm": 0.5315745683150092,
"learning_rate": 9.063135361782228e-06,
"loss": 0.4761,
"step": 998
},
{
"epoch": 2.88728323699422,
"grad_norm": 0.5901949821752347,
"learning_rate": 9.043048810132778e-06,
"loss": 0.4927,
"step": 999
},
{
"epoch": 2.8901734104046244,
"grad_norm": 0.696770939108106,
"learning_rate": 9.02296615442243e-06,
"loss": 0.4979,
"step": 1000
},
{
"epoch": 2.893063583815029,
"grad_norm": 0.5335722900416059,
"learning_rate": 9.002887476411681e-06,
"loss": 0.4506,
"step": 1001
},
{
"epoch": 2.8959537572254335,
"grad_norm": 0.5448959769913916,
"learning_rate": 8.982812857844826e-06,
"loss": 0.4768,
"step": 1002
},
{
"epoch": 2.898843930635838,
"grad_norm": 0.6178055948176863,
"learning_rate": 8.96274238044965e-06,
"loss": 0.4867,
"step": 1003
},
{
"epoch": 2.901734104046243,
"grad_norm": 0.6091101645045837,
"learning_rate": 8.942676125937063e-06,
"loss": 0.4854,
"step": 1004
},
{
"epoch": 2.9046242774566475,
"grad_norm": 0.57752939009839,
"learning_rate": 8.922614176000783e-06,
"loss": 0.47,
"step": 1005
},
{
"epoch": 2.907514450867052,
"grad_norm": 0.532063897992589,
"learning_rate": 8.902556612317017e-06,
"loss": 0.5041,
"step": 1006
},
{
"epoch": 2.9104046242774566,
"grad_norm": 0.6190519199948481,
"learning_rate": 8.882503516544105e-06,
"loss": 0.4674,
"step": 1007
},
{
"epoch": 2.913294797687861,
"grad_norm": 0.6052048247524936,
"learning_rate": 8.862454970322194e-06,
"loss": 0.4713,
"step": 1008
},
{
"epoch": 2.916184971098266,
"grad_norm": 0.6017438939459462,
"learning_rate": 8.842411055272913e-06,
"loss": 0.478,
"step": 1009
},
{
"epoch": 2.9190751445086707,
"grad_norm": 0.5768744304740853,
"learning_rate": 8.82237185299904e-06,
"loss": 0.4875,
"step": 1010
},
{
"epoch": 2.921965317919075,
"grad_norm": 0.5951970872451218,
"learning_rate": 8.802337445084165e-06,
"loss": 0.4814,
"step": 1011
},
{
"epoch": 2.9248554913294798,
"grad_norm": 0.5608710672407204,
"learning_rate": 8.782307913092354e-06,
"loss": 0.4752,
"step": 1012
},
{
"epoch": 2.9277456647398843,
"grad_norm": 0.5509512795540918,
"learning_rate": 8.762283338567822e-06,
"loss": 0.492,
"step": 1013
},
{
"epoch": 2.9306358381502893,
"grad_norm": 0.5975138590995895,
"learning_rate": 8.74226380303461e-06,
"loss": 0.4828,
"step": 1014
},
{
"epoch": 2.9335260115606934,
"grad_norm": 0.5864321790518813,
"learning_rate": 8.722249387996238e-06,
"loss": 0.4859,
"step": 1015
},
{
"epoch": 2.9364161849710984,
"grad_norm": 0.5555533284809581,
"learning_rate": 8.702240174935377e-06,
"loss": 0.5051,
"step": 1016
},
{
"epoch": 2.939306358381503,
"grad_norm": 0.5650980172734773,
"learning_rate": 8.68223624531353e-06,
"loss": 0.4781,
"step": 1017
},
{
"epoch": 2.9421965317919074,
"grad_norm": 0.5991051719573498,
"learning_rate": 8.662237680570676e-06,
"loss": 0.4972,
"step": 1018
},
{
"epoch": 2.9450867052023124,
"grad_norm": 0.6209924930788805,
"learning_rate": 8.642244562124962e-06,
"loss": 0.4771,
"step": 1019
},
{
"epoch": 2.9479768786127165,
"grad_norm": 0.5411586227886513,
"learning_rate": 8.622256971372361e-06,
"loss": 0.4847,
"step": 1020
},
{
"epoch": 2.9508670520231215,
"grad_norm": 0.5668517081503266,
"learning_rate": 8.602274989686338e-06,
"loss": 0.4848,
"step": 1021
},
{
"epoch": 2.953757225433526,
"grad_norm": 0.5759877651439558,
"learning_rate": 8.582298698417529e-06,
"loss": 0.4884,
"step": 1022
},
{
"epoch": 2.9566473988439306,
"grad_norm": 0.5862302492886513,
"learning_rate": 8.562328178893392e-06,
"loss": 0.4932,
"step": 1023
},
{
"epoch": 2.959537572254335,
"grad_norm": 0.5481516742462423,
"learning_rate": 8.542363512417898e-06,
"loss": 0.5047,
"step": 1024
},
{
"epoch": 2.9624277456647397,
"grad_norm": 0.513093768869479,
"learning_rate": 8.522404780271187e-06,
"loss": 0.4664,
"step": 1025
},
{
"epoch": 2.9653179190751446,
"grad_norm": 0.5575656175840487,
"learning_rate": 8.502452063709235e-06,
"loss": 0.4606,
"step": 1026
},
{
"epoch": 2.968208092485549,
"grad_norm": 0.5556502229251714,
"learning_rate": 8.482505443963526e-06,
"loss": 0.4775,
"step": 1027
},
{
"epoch": 2.9710982658959537,
"grad_norm": 0.5733747154067709,
"learning_rate": 8.462565002240733e-06,
"loss": 0.4666,
"step": 1028
},
{
"epoch": 2.9739884393063583,
"grad_norm": 0.5612472015255541,
"learning_rate": 8.442630819722365e-06,
"loss": 0.4949,
"step": 1029
},
{
"epoch": 2.976878612716763,
"grad_norm": 0.6413581942492151,
"learning_rate": 8.422702977564453e-06,
"loss": 0.4987,
"step": 1030
},
{
"epoch": 2.979768786127168,
"grad_norm": 0.5763965026676654,
"learning_rate": 8.402781556897222e-06,
"loss": 0.4829,
"step": 1031
},
{
"epoch": 2.9826589595375723,
"grad_norm": 0.5273947424887864,
"learning_rate": 8.382866638824745e-06,
"loss": 0.4574,
"step": 1032
},
{
"epoch": 2.985549132947977,
"grad_norm": 0.7908791681969232,
"learning_rate": 8.362958304424622e-06,
"loss": 0.4724,
"step": 1033
},
{
"epoch": 2.9884393063583814,
"grad_norm": 0.6609996660422045,
"learning_rate": 8.34305663474765e-06,
"loss": 0.481,
"step": 1034
},
{
"epoch": 2.991329479768786,
"grad_norm": 0.6358348872654778,
"learning_rate": 8.323161710817502e-06,
"loss": 0.4858,
"step": 1035
},
{
"epoch": 2.994219653179191,
"grad_norm": 0.6714732526625259,
"learning_rate": 8.303273613630376e-06,
"loss": 0.5107,
"step": 1036
},
{
"epoch": 2.9971098265895955,
"grad_norm": 0.5450824548582315,
"learning_rate": 8.28339242415468e-06,
"loss": 0.4952,
"step": 1037
},
{
"epoch": 3.0,
"grad_norm": 0.6874759980679123,
"learning_rate": 8.263518223330698e-06,
"loss": 0.4575,
"step": 1038
},
{
"epoch": 3.0028901734104045,
"grad_norm": 0.9527898393580617,
"learning_rate": 8.24365109207027e-06,
"loss": 0.3802,
"step": 1039
},
{
"epoch": 3.005780346820809,
"grad_norm": 0.7089585065956318,
"learning_rate": 8.223791111256447e-06,
"loss": 0.3941,
"step": 1040
},
{
"epoch": 3.008670520231214,
"grad_norm": 1.1904284295357626,
"learning_rate": 8.203938361743169e-06,
"loss": 0.3994,
"step": 1041
},
{
"epoch": 3.0115606936416186,
"grad_norm": 0.6023200450037057,
"learning_rate": 8.184092924354943e-06,
"loss": 0.3836,
"step": 1042
},
{
"epoch": 3.014450867052023,
"grad_norm": 0.7823993760566921,
"learning_rate": 8.164254879886496e-06,
"loss": 0.4116,
"step": 1043
},
{
"epoch": 3.0173410404624277,
"grad_norm": 0.7376436181873413,
"learning_rate": 8.144424309102463e-06,
"loss": 0.4203,
"step": 1044
},
{
"epoch": 3.020231213872832,
"grad_norm": 0.614281631401755,
"learning_rate": 8.124601292737051e-06,
"loss": 0.3856,
"step": 1045
},
{
"epoch": 3.023121387283237,
"grad_norm": 0.6160206160107256,
"learning_rate": 8.10478591149372e-06,
"loss": 0.4013,
"step": 1046
},
{
"epoch": 3.0260115606936417,
"grad_norm": 0.6319733094936663,
"learning_rate": 8.084978246044827e-06,
"loss": 0.3874,
"step": 1047
},
{
"epoch": 3.0289017341040463,
"grad_norm": 0.5472529492540473,
"learning_rate": 8.065178377031333e-06,
"loss": 0.3672,
"step": 1048
},
{
"epoch": 3.031791907514451,
"grad_norm": 0.6252569965995163,
"learning_rate": 8.045386385062453e-06,
"loss": 0.4023,
"step": 1049
},
{
"epoch": 3.0346820809248554,
"grad_norm": 0.5822538115356047,
"learning_rate": 8.025602350715332e-06,
"loss": 0.3942,
"step": 1050
},
{
"epoch": 3.03757225433526,
"grad_norm": 0.567501237074041,
"learning_rate": 8.005826354534722e-06,
"loss": 0.3783,
"step": 1051
},
{
"epoch": 3.040462427745665,
"grad_norm": 0.5400206076485143,
"learning_rate": 7.986058477032638e-06,
"loss": 0.3871,
"step": 1052
},
{
"epoch": 3.0433526011560694,
"grad_norm": 0.5846560249755006,
"learning_rate": 7.966298798688059e-06,
"loss": 0.3816,
"step": 1053
},
{
"epoch": 3.046242774566474,
"grad_norm": 0.5676434874248799,
"learning_rate": 7.94654739994657e-06,
"loss": 0.3924,
"step": 1054
},
{
"epoch": 3.0491329479768785,
"grad_norm": 0.5333952954526797,
"learning_rate": 7.926804361220056e-06,
"loss": 0.3878,
"step": 1055
},
{
"epoch": 3.052023121387283,
"grad_norm": 0.5851971305918012,
"learning_rate": 7.907069762886366e-06,
"loss": 0.3824,
"step": 1056
},
{
"epoch": 3.054913294797688,
"grad_norm": 0.578493351433017,
"learning_rate": 7.887343685288985e-06,
"loss": 0.3799,
"step": 1057
},
{
"epoch": 3.0578034682080926,
"grad_norm": 0.5889989245232254,
"learning_rate": 7.867626208736703e-06,
"loss": 0.3819,
"step": 1058
},
{
"epoch": 3.060693641618497,
"grad_norm": 0.5723245049646746,
"learning_rate": 7.8479174135033e-06,
"loss": 0.3722,
"step": 1059
},
{
"epoch": 3.0635838150289016,
"grad_norm": 0.6215894667582502,
"learning_rate": 7.828217379827216e-06,
"loss": 0.3899,
"step": 1060
},
{
"epoch": 3.066473988439306,
"grad_norm": 0.5938247793105494,
"learning_rate": 7.808526187911214e-06,
"loss": 0.431,
"step": 1061
},
{
"epoch": 3.069364161849711,
"grad_norm": 0.5228702312038274,
"learning_rate": 7.788843917922056e-06,
"loss": 0.3647,
"step": 1062
},
{
"epoch": 3.0722543352601157,
"grad_norm": 0.6091134386113599,
"learning_rate": 7.769170649990198e-06,
"loss": 0.3816,
"step": 1063
},
{
"epoch": 3.0751445086705202,
"grad_norm": 0.5642938672189238,
"learning_rate": 7.749506464209428e-06,
"loss": 0.3939,
"step": 1064
},
{
"epoch": 3.078034682080925,
"grad_norm": 0.5421265675643979,
"learning_rate": 7.729851440636576e-06,
"loss": 0.3886,
"step": 1065
},
{
"epoch": 3.0809248554913293,
"grad_norm": 0.5902121269790457,
"learning_rate": 7.710205659291147e-06,
"loss": 0.372,
"step": 1066
},
{
"epoch": 3.0838150289017343,
"grad_norm": 0.5532348889649435,
"learning_rate": 7.690569200155052e-06,
"loss": 0.3899,
"step": 1067
},
{
"epoch": 3.086705202312139,
"grad_norm": 0.5288570224088542,
"learning_rate": 7.67094214317222e-06,
"loss": 0.3981,
"step": 1068
},
{
"epoch": 3.0895953757225434,
"grad_norm": 0.5529303979607351,
"learning_rate": 7.651324568248319e-06,
"loss": 0.3915,
"step": 1069
},
{
"epoch": 3.092485549132948,
"grad_norm": 0.574796902074762,
"learning_rate": 7.631716555250401e-06,
"loss": 0.3738,
"step": 1070
},
{
"epoch": 3.0953757225433525,
"grad_norm": 0.5997982476864651,
"learning_rate": 7.612118184006606e-06,
"loss": 0.4025,
"step": 1071
},
{
"epoch": 3.098265895953757,
"grad_norm": 0.5586255763175132,
"learning_rate": 7.592529534305805e-06,
"loss": 0.3829,
"step": 1072
},
{
"epoch": 3.101156069364162,
"grad_norm": 0.5662255742739997,
"learning_rate": 7.572950685897295e-06,
"loss": 0.3796,
"step": 1073
},
{
"epoch": 3.1040462427745665,
"grad_norm": 0.5455105617146326,
"learning_rate": 7.5533817184904765e-06,
"loss": 0.3812,
"step": 1074
},
{
"epoch": 3.106936416184971,
"grad_norm": 0.5541786996420895,
"learning_rate": 7.533822711754516e-06,
"loss": 0.3903,
"step": 1075
},
{
"epoch": 3.1098265895953756,
"grad_norm": 0.4957281719052488,
"learning_rate": 7.514273745318032e-06,
"loss": 0.4175,
"step": 1076
},
{
"epoch": 3.11271676300578,
"grad_norm": 0.5194029043856342,
"learning_rate": 7.494734898768758e-06,
"loss": 0.3985,
"step": 1077
},
{
"epoch": 3.115606936416185,
"grad_norm": 0.5467787010237095,
"learning_rate": 7.475206251653242e-06,
"loss": 0.3798,
"step": 1078
},
{
"epoch": 3.1184971098265897,
"grad_norm": 0.4946840290814573,
"learning_rate": 7.4556878834764965e-06,
"loss": 0.3862,
"step": 1079
},
{
"epoch": 3.121387283236994,
"grad_norm": 0.5675664027874467,
"learning_rate": 7.436179873701687e-06,
"loss": 0.404,
"step": 1080
},
{
"epoch": 3.1242774566473988,
"grad_norm": 0.5758919601256814,
"learning_rate": 7.416682301749819e-06,
"loss": 0.3864,
"step": 1081
},
{
"epoch": 3.1271676300578033,
"grad_norm": 0.4944933437081969,
"learning_rate": 7.397195246999391e-06,
"loss": 0.3918,
"step": 1082
},
{
"epoch": 3.1300578034682083,
"grad_norm": 0.5370826053866485,
"learning_rate": 7.377718788786088e-06,
"loss": 0.3908,
"step": 1083
},
{
"epoch": 3.132947976878613,
"grad_norm": 0.5074787765859067,
"learning_rate": 7.358253006402454e-06,
"loss": 0.3911,
"step": 1084
},
{
"epoch": 3.1358381502890174,
"grad_norm": 0.5337743052057817,
"learning_rate": 7.338797979097572e-06,
"loss": 0.3871,
"step": 1085
},
{
"epoch": 3.138728323699422,
"grad_norm": 0.5266563755134518,
"learning_rate": 7.319353786076739e-06,
"loss": 0.3826,
"step": 1086
},
{
"epoch": 3.1416184971098264,
"grad_norm": 0.6054998252332005,
"learning_rate": 7.299920506501135e-06,
"loss": 0.3885,
"step": 1087
},
{
"epoch": 3.1445086705202314,
"grad_norm": 0.5609546262425186,
"learning_rate": 7.280498219487526e-06,
"loss": 0.4117,
"step": 1088
},
{
"epoch": 3.147398843930636,
"grad_norm": 0.5517930019790442,
"learning_rate": 7.261087004107909e-06,
"loss": 0.4003,
"step": 1089
},
{
"epoch": 3.1502890173410405,
"grad_norm": 0.518951997928602,
"learning_rate": 7.241686939389214e-06,
"loss": 0.4002,
"step": 1090
},
{
"epoch": 3.153179190751445,
"grad_norm": 0.5541234205878507,
"learning_rate": 7.222298104312966e-06,
"loss": 0.3917,
"step": 1091
},
{
"epoch": 3.1560693641618496,
"grad_norm": 0.6050039791749773,
"learning_rate": 7.202920577814991e-06,
"loss": 0.4036,
"step": 1092
},
{
"epoch": 3.1589595375722546,
"grad_norm": 0.48174559599735306,
"learning_rate": 7.183554438785054e-06,
"loss": 0.3919,
"step": 1093
},
{
"epoch": 3.161849710982659,
"grad_norm": 0.5735781752986523,
"learning_rate": 7.164199766066571e-06,
"loss": 0.3865,
"step": 1094
},
{
"epoch": 3.1647398843930636,
"grad_norm": 0.54707725193784,
"learning_rate": 7.144856638456271e-06,
"loss": 0.3848,
"step": 1095
},
{
"epoch": 3.167630057803468,
"grad_norm": 0.5156168830536783,
"learning_rate": 7.125525134703891e-06,
"loss": 0.3659,
"step": 1096
},
{
"epoch": 3.1705202312138727,
"grad_norm": 0.5268935444771893,
"learning_rate": 7.106205333511829e-06,
"loss": 0.3818,
"step": 1097
},
{
"epoch": 3.1734104046242773,
"grad_norm": 0.47788719205051533,
"learning_rate": 7.086897313534851e-06,
"loss": 0.4042,
"step": 1098
},
{
"epoch": 3.1763005780346822,
"grad_norm": 0.5259915318997668,
"learning_rate": 7.067601153379757e-06,
"loss": 0.4052,
"step": 1099
},
{
"epoch": 3.179190751445087,
"grad_norm": 0.5283480560745818,
"learning_rate": 7.048316931605062e-06,
"loss": 0.4134,
"step": 1100
},
{
"epoch": 3.1820809248554913,
"grad_norm": 0.4586208673070595,
"learning_rate": 7.029044726720681e-06,
"loss": 0.4114,
"step": 1101
},
{
"epoch": 3.184971098265896,
"grad_norm": 0.49357031094595183,
"learning_rate": 7.009784617187594e-06,
"loss": 0.4041,
"step": 1102
},
{
"epoch": 3.1878612716763004,
"grad_norm": 0.5037877175085824,
"learning_rate": 6.990536681417555e-06,
"loss": 0.4198,
"step": 1103
},
{
"epoch": 3.1907514450867054,
"grad_norm": 0.5017313490764773,
"learning_rate": 6.971300997772746e-06,
"loss": 0.3801,
"step": 1104
},
{
"epoch": 3.19364161849711,
"grad_norm": 0.4879258565266107,
"learning_rate": 6.9520776445654696e-06,
"loss": 0.4155,
"step": 1105
},
{
"epoch": 3.1965317919075145,
"grad_norm": 0.5095618152215767,
"learning_rate": 6.932866700057832e-06,
"loss": 0.396,
"step": 1106
},
{
"epoch": 3.199421965317919,
"grad_norm": 0.5221572691725388,
"learning_rate": 6.91366824246142e-06,
"loss": 0.3836,
"step": 1107
},
{
"epoch": 3.2023121387283235,
"grad_norm": 0.5089182169867535,
"learning_rate": 6.894482349936977e-06,
"loss": 0.3892,
"step": 1108
},
{
"epoch": 3.2052023121387285,
"grad_norm": 0.4800548236411057,
"learning_rate": 6.875309100594098e-06,
"loss": 0.3998,
"step": 1109
},
{
"epoch": 3.208092485549133,
"grad_norm": 0.5125112881492649,
"learning_rate": 6.8561485724909105e-06,
"loss": 0.4056,
"step": 1110
},
{
"epoch": 3.2109826589595376,
"grad_norm": 0.5757448225017041,
"learning_rate": 6.83700084363374e-06,
"loss": 0.4033,
"step": 1111
},
{
"epoch": 3.213872832369942,
"grad_norm": 0.5119217186783843,
"learning_rate": 6.8178659919768056e-06,
"loss": 0.3985,
"step": 1112
},
{
"epoch": 3.2167630057803467,
"grad_norm": 0.5094018785467254,
"learning_rate": 6.798744095421909e-06,
"loss": 0.3727,
"step": 1113
},
{
"epoch": 3.2196531791907512,
"grad_norm": 0.5371490562135045,
"learning_rate": 6.7796352318181015e-06,
"loss": 0.4222,
"step": 1114
},
{
"epoch": 3.222543352601156,
"grad_norm": 0.530597597226532,
"learning_rate": 6.76053947896138e-06,
"loss": 0.4138,
"step": 1115
},
{
"epoch": 3.2254335260115607,
"grad_norm": 0.5360287546084145,
"learning_rate": 6.741456914594353e-06,
"loss": 0.406,
"step": 1116
},
{
"epoch": 3.2283236994219653,
"grad_norm": 0.5254677842306964,
"learning_rate": 6.722387616405957e-06,
"loss": 0.3964,
"step": 1117
},
{
"epoch": 3.23121387283237,
"grad_norm": 0.5118935706321023,
"learning_rate": 6.7033316620310985e-06,
"loss": 0.3854,
"step": 1118
},
{
"epoch": 3.2341040462427744,
"grad_norm": 0.542111399954812,
"learning_rate": 6.68428912905037e-06,
"loss": 0.3889,
"step": 1119
},
{
"epoch": 3.2369942196531793,
"grad_norm": 0.5231724529368768,
"learning_rate": 6.6652600949897206e-06,
"loss": 0.4199,
"step": 1120
},
{
"epoch": 3.239884393063584,
"grad_norm": 0.4928163479094274,
"learning_rate": 6.646244637320145e-06,
"loss": 0.3981,
"step": 1121
},
{
"epoch": 3.2427745664739884,
"grad_norm": 0.5345331053208331,
"learning_rate": 6.627242833457357e-06,
"loss": 0.3847,
"step": 1122
},
{
"epoch": 3.245664739884393,
"grad_norm": 0.559335727163023,
"learning_rate": 6.60825476076149e-06,
"loss": 0.402,
"step": 1123
},
{
"epoch": 3.2485549132947975,
"grad_norm": 0.5029209372564127,
"learning_rate": 6.5892804965367785e-06,
"loss": 0.3916,
"step": 1124
},
{
"epoch": 3.2514450867052025,
"grad_norm": 0.49343284170126384,
"learning_rate": 6.570320118031233e-06,
"loss": 0.4022,
"step": 1125
},
{
"epoch": 3.254335260115607,
"grad_norm": 0.5401891502634477,
"learning_rate": 6.551373702436335e-06,
"loss": 0.3948,
"step": 1126
},
{
"epoch": 3.2572254335260116,
"grad_norm": 0.49439119896843026,
"learning_rate": 6.532441326886716e-06,
"loss": 0.3892,
"step": 1127
},
{
"epoch": 3.260115606936416,
"grad_norm": 0.4974225213503987,
"learning_rate": 6.513523068459859e-06,
"loss": 0.4086,
"step": 1128
},
{
"epoch": 3.2630057803468207,
"grad_norm": 0.5162026470577603,
"learning_rate": 6.494619004175761e-06,
"loss": 0.403,
"step": 1129
},
{
"epoch": 3.2658959537572256,
"grad_norm": 0.524460346825612,
"learning_rate": 6.475729210996637e-06,
"loss": 0.3998,
"step": 1130
},
{
"epoch": 3.26878612716763,
"grad_norm": 0.5513128758798284,
"learning_rate": 6.456853765826609e-06,
"loss": 0.3894,
"step": 1131
},
{
"epoch": 3.2716763005780347,
"grad_norm": 0.48701380796712024,
"learning_rate": 6.43799274551137e-06,
"loss": 0.4033,
"step": 1132
},
{
"epoch": 3.2745664739884393,
"grad_norm": 0.4970848753735955,
"learning_rate": 6.419146226837894e-06,
"loss": 0.4109,
"step": 1133
},
{
"epoch": 3.277456647398844,
"grad_norm": 0.5093548978079553,
"learning_rate": 6.400314286534119e-06,
"loss": 0.391,
"step": 1134
},
{
"epoch": 3.2803468208092488,
"grad_norm": 0.5429732351329991,
"learning_rate": 6.381497001268631e-06,
"loss": 0.3989,
"step": 1135
},
{
"epoch": 3.2832369942196533,
"grad_norm": 0.5457996431220703,
"learning_rate": 6.3626944476503485e-06,
"loss": 0.3822,
"step": 1136
},
{
"epoch": 3.286127167630058,
"grad_norm": 0.5885691843952315,
"learning_rate": 6.343906702228213e-06,
"loss": 0.4026,
"step": 1137
},
{
"epoch": 3.2890173410404624,
"grad_norm": 0.675751707823907,
"learning_rate": 6.325133841490887e-06,
"loss": 0.3897,
"step": 1138
},
{
"epoch": 3.291907514450867,
"grad_norm": 0.544504709401962,
"learning_rate": 6.306375941866428e-06,
"loss": 0.3808,
"step": 1139
},
{
"epoch": 3.294797687861272,
"grad_norm": 0.5233663308147422,
"learning_rate": 6.287633079721986e-06,
"loss": 0.3775,
"step": 1140
},
{
"epoch": 3.2976878612716765,
"grad_norm": 0.5962514875461025,
"learning_rate": 6.268905331363487e-06,
"loss": 0.4067,
"step": 1141
},
{
"epoch": 3.300578034682081,
"grad_norm": 0.7022260424825366,
"learning_rate": 6.250192773035333e-06,
"loss": 0.4003,
"step": 1142
},
{
"epoch": 3.3034682080924855,
"grad_norm": 0.489621823505066,
"learning_rate": 6.231495480920078e-06,
"loss": 0.4192,
"step": 1143
},
{
"epoch": 3.30635838150289,
"grad_norm": 0.5934308896956955,
"learning_rate": 6.212813531138124e-06,
"loss": 0.3865,
"step": 1144
},
{
"epoch": 3.3092485549132946,
"grad_norm": 0.5852026626951878,
"learning_rate": 6.194146999747419e-06,
"loss": 0.4077,
"step": 1145
},
{
"epoch": 3.3121387283236996,
"grad_norm": 0.5034245128179445,
"learning_rate": 6.1754959627431325e-06,
"loss": 0.4108,
"step": 1146
},
{
"epoch": 3.315028901734104,
"grad_norm": 0.48420407726634407,
"learning_rate": 6.156860496057351e-06,
"loss": 0.4003,
"step": 1147
},
{
"epoch": 3.3179190751445087,
"grad_norm": 0.4906099299312336,
"learning_rate": 6.138240675558778e-06,
"loss": 0.4012,
"step": 1148
},
{
"epoch": 3.320809248554913,
"grad_norm": 0.5112625863289267,
"learning_rate": 6.119636577052414e-06,
"loss": 0.3972,
"step": 1149
},
{
"epoch": 3.3236994219653178,
"grad_norm": 0.4758612139780586,
"learning_rate": 6.101048276279259e-06,
"loss": 0.3867,
"step": 1150
},
{
"epoch": 3.3265895953757223,
"grad_norm": 0.5099399487696842,
"learning_rate": 6.082475848915985e-06,
"loss": 0.4004,
"step": 1151
},
{
"epoch": 3.3294797687861273,
"grad_norm": 0.49659635472420044,
"learning_rate": 6.063919370574647e-06,
"loss": 0.4203,
"step": 1152
},
{
"epoch": 3.332369942196532,
"grad_norm": 0.5246518362455117,
"learning_rate": 6.045378916802374e-06,
"loss": 0.3774,
"step": 1153
},
{
"epoch": 3.3352601156069364,
"grad_norm": 0.505469636516935,
"learning_rate": 6.026854563081046e-06,
"loss": 0.4151,
"step": 1154
},
{
"epoch": 3.338150289017341,
"grad_norm": 0.515524111669419,
"learning_rate": 6.008346384827e-06,
"loss": 0.3923,
"step": 1155
},
{
"epoch": 3.3410404624277454,
"grad_norm": 0.5294276547027134,
"learning_rate": 5.989854457390725e-06,
"loss": 0.3772,
"step": 1156
},
{
"epoch": 3.3439306358381504,
"grad_norm": 0.5162545524791332,
"learning_rate": 5.97137885605654e-06,
"loss": 0.4161,
"step": 1157
},
{
"epoch": 3.346820809248555,
"grad_norm": 0.4833823444894276,
"learning_rate": 5.952919656042302e-06,
"loss": 0.4192,
"step": 1158
},
{
"epoch": 3.3497109826589595,
"grad_norm": 0.5098142017762833,
"learning_rate": 5.934476932499096e-06,
"loss": 0.4095,
"step": 1159
},
{
"epoch": 3.352601156069364,
"grad_norm": 0.49912178138612384,
"learning_rate": 5.916050760510928e-06,
"loss": 0.4193,
"step": 1160
},
{
"epoch": 3.3554913294797686,
"grad_norm": 0.48744060348108537,
"learning_rate": 5.897641215094417e-06,
"loss": 0.4107,
"step": 1161
},
{
"epoch": 3.3583815028901736,
"grad_norm": 0.5061412741063811,
"learning_rate": 5.879248371198492e-06,
"loss": 0.379,
"step": 1162
},
{
"epoch": 3.361271676300578,
"grad_norm": 0.4994719238375096,
"learning_rate": 5.8608723037040894e-06,
"loss": 0.4095,
"step": 1163
},
{
"epoch": 3.3641618497109826,
"grad_norm": 0.5358940583528377,
"learning_rate": 5.842513087423844e-06,
"loss": 0.3814,
"step": 1164
},
{
"epoch": 3.367052023121387,
"grad_norm": 0.5235124787983995,
"learning_rate": 5.824170797101787e-06,
"loss": 0.3933,
"step": 1165
},
{
"epoch": 3.3699421965317917,
"grad_norm": 0.5039212115911841,
"learning_rate": 5.805845507413032e-06,
"loss": 0.3985,
"step": 1166
},
{
"epoch": 3.3728323699421967,
"grad_norm": 0.5060043732907776,
"learning_rate": 5.7875372929635e-06,
"loss": 0.3974,
"step": 1167
},
{
"epoch": 3.3757225433526012,
"grad_norm": 0.5581048525929695,
"learning_rate": 5.769246228289578e-06,
"loss": 0.3877,
"step": 1168
},
{
"epoch": 3.378612716763006,
"grad_norm": 0.5495580859568931,
"learning_rate": 5.750972387857827e-06,
"loss": 0.3819,
"step": 1169
},
{
"epoch": 3.3815028901734103,
"grad_norm": 0.5067255290694974,
"learning_rate": 5.732715846064707e-06,
"loss": 0.4018,
"step": 1170
},
{
"epoch": 3.384393063583815,
"grad_norm": 0.5077609535944343,
"learning_rate": 5.714476677236238e-06,
"loss": 0.395,
"step": 1171
},
{
"epoch": 3.38728323699422,
"grad_norm": 0.5579418064478606,
"learning_rate": 5.6962549556277134e-06,
"loss": 0.3984,
"step": 1172
},
{
"epoch": 3.3901734104046244,
"grad_norm": 0.49258839842495805,
"learning_rate": 5.678050755423385e-06,
"loss": 0.4233,
"step": 1173
},
{
"epoch": 3.393063583815029,
"grad_norm": 0.5423286857222281,
"learning_rate": 5.659864150736194e-06,
"loss": 0.4053,
"step": 1174
},
{
"epoch": 3.3959537572254335,
"grad_norm": 0.5232668352102138,
"learning_rate": 5.64169521560743e-06,
"loss": 0.3846,
"step": 1175
},
{
"epoch": 3.398843930635838,
"grad_norm": 0.491353654719555,
"learning_rate": 5.623544024006451e-06,
"loss": 0.3775,
"step": 1176
},
{
"epoch": 3.401734104046243,
"grad_norm": 0.51410989795105,
"learning_rate": 5.605410649830375e-06,
"loss": 0.3864,
"step": 1177
},
{
"epoch": 3.4046242774566475,
"grad_norm": 0.5219888675404522,
"learning_rate": 5.587295166903785e-06,
"loss": 0.3727,
"step": 1178
},
{
"epoch": 3.407514450867052,
"grad_norm": 0.5458655990227211,
"learning_rate": 5.569197648978427e-06,
"loss": 0.3958,
"step": 1179
},
{
"epoch": 3.4104046242774566,
"grad_norm": 0.4953302314125495,
"learning_rate": 5.551118169732901e-06,
"loss": 0.3569,
"step": 1180
},
{
"epoch": 3.413294797687861,
"grad_norm": 0.5476231027492036,
"learning_rate": 5.533056802772374e-06,
"loss": 0.3827,
"step": 1181
},
{
"epoch": 3.416184971098266,
"grad_norm": 0.5230969438602984,
"learning_rate": 5.515013621628271e-06,
"loss": 0.3857,
"step": 1182
},
{
"epoch": 3.4190751445086707,
"grad_norm": 0.5335942107594615,
"learning_rate": 5.496988699757982e-06,
"loss": 0.4121,
"step": 1183
},
{
"epoch": 3.421965317919075,
"grad_norm": 0.5220805134558412,
"learning_rate": 5.478982110544555e-06,
"loss": 0.3929,
"step": 1184
},
{
"epoch": 3.4248554913294798,
"grad_norm": 0.481695491417452,
"learning_rate": 5.460993927296408e-06,
"loss": 0.4079,
"step": 1185
},
{
"epoch": 3.4277456647398843,
"grad_norm": 0.5091817629646063,
"learning_rate": 5.44302422324702e-06,
"loss": 0.4012,
"step": 1186
},
{
"epoch": 3.430635838150289,
"grad_norm": 0.5603932046574709,
"learning_rate": 5.425073071554634e-06,
"loss": 0.3784,
"step": 1187
},
{
"epoch": 3.433526011560694,
"grad_norm": 0.5131330214838938,
"learning_rate": 5.4071405453019786e-06,
"loss": 0.3929,
"step": 1188
},
{
"epoch": 3.4364161849710984,
"grad_norm": 0.5512023006176935,
"learning_rate": 5.389226717495943e-06,
"loss": 0.4071,
"step": 1189
},
{
"epoch": 3.439306358381503,
"grad_norm": 0.5472275221249205,
"learning_rate": 5.371331661067284e-06,
"loss": 0.4022,
"step": 1190
},
{
"epoch": 3.4421965317919074,
"grad_norm": 0.46058321008388775,
"learning_rate": 5.353455448870344e-06,
"loss": 0.414,
"step": 1191
},
{
"epoch": 3.445086705202312,
"grad_norm": 0.5247383099681343,
"learning_rate": 5.335598153682754e-06,
"loss": 0.3908,
"step": 1192
},
{
"epoch": 3.447976878612717,
"grad_norm": 0.5232079526618546,
"learning_rate": 5.317759848205124e-06,
"loss": 0.3983,
"step": 1193
},
{
"epoch": 3.4508670520231215,
"grad_norm": 0.4903680174032517,
"learning_rate": 5.299940605060743e-06,
"loss": 0.4113,
"step": 1194
},
{
"epoch": 3.453757225433526,
"grad_norm": 0.5314891850338158,
"learning_rate": 5.2821404967953116e-06,
"loss": 0.3715,
"step": 1195
},
{
"epoch": 3.4566473988439306,
"grad_norm": 0.5397537307529937,
"learning_rate": 5.264359595876618e-06,
"loss": 0.4043,
"step": 1196
},
{
"epoch": 3.459537572254335,
"grad_norm": 0.47328749106274787,
"learning_rate": 5.2465979746942565e-06,
"loss": 0.4137,
"step": 1197
},
{
"epoch": 3.4624277456647397,
"grad_norm": 0.5362144233759804,
"learning_rate": 5.228855705559319e-06,
"loss": 0.3914,
"step": 1198
},
{
"epoch": 3.4653179190751446,
"grad_norm": 0.5422695568332802,
"learning_rate": 5.211132860704131e-06,
"loss": 0.4199,
"step": 1199
},
{
"epoch": 3.468208092485549,
"grad_norm": 0.4718961202791377,
"learning_rate": 5.193429512281925e-06,
"loss": 0.4026,
"step": 1200
},
{
"epoch": 3.4710982658959537,
"grad_norm": 0.4876987449009726,
"learning_rate": 5.1757457323665625e-06,
"loss": 0.3996,
"step": 1201
},
{
"epoch": 3.4739884393063583,
"grad_norm": 0.5155793556037093,
"learning_rate": 5.158081592952237e-06,
"loss": 0.3901,
"step": 1202
},
{
"epoch": 3.476878612716763,
"grad_norm": 0.5272401689022481,
"learning_rate": 5.1404371659531806e-06,
"loss": 0.4087,
"step": 1203
},
{
"epoch": 3.479768786127168,
"grad_norm": 0.48712333585359746,
"learning_rate": 5.122812523203376e-06,
"loss": 0.3904,
"step": 1204
},
{
"epoch": 3.4826589595375723,
"grad_norm": 0.5184846661575929,
"learning_rate": 5.105207736456257e-06,
"loss": 0.4176,
"step": 1205
},
{
"epoch": 3.485549132947977,
"grad_norm": 0.4786955304589319,
"learning_rate": 5.087622877384422e-06,
"loss": 0.4181,
"step": 1206
},
{
"epoch": 3.4884393063583814,
"grad_norm": 0.5042504732169668,
"learning_rate": 5.070058017579335e-06,
"loss": 0.3857,
"step": 1207
},
{
"epoch": 3.491329479768786,
"grad_norm": 0.4799858953189233,
"learning_rate": 5.052513228551048e-06,
"loss": 0.3914,
"step": 1208
},
{
"epoch": 3.494219653179191,
"grad_norm": 0.4658279526771844,
"learning_rate": 5.034988581727891e-06,
"loss": 0.3988,
"step": 1209
},
{
"epoch": 3.4971098265895955,
"grad_norm": 0.5108447742628904,
"learning_rate": 5.017484148456195e-06,
"loss": 0.4215,
"step": 1210
},
{
"epoch": 3.5,
"grad_norm": 0.5402881254642645,
"learning_rate": 5.000000000000003e-06,
"loss": 0.3946,
"step": 1211
},
{
"epoch": 3.5028901734104045,
"grad_norm": 0.51161404379665,
"learning_rate": 4.98253620754076e-06,
"loss": 0.4165,
"step": 1212
},
{
"epoch": 3.505780346820809,
"grad_norm": 0.4974311157637146,
"learning_rate": 4.965092842177063e-06,
"loss": 0.393,
"step": 1213
},
{
"epoch": 3.508670520231214,
"grad_norm": 0.50083094507991,
"learning_rate": 4.947669974924318e-06,
"loss": 0.3925,
"step": 1214
},
{
"epoch": 3.5115606936416186,
"grad_norm": 0.5084864388305871,
"learning_rate": 4.930267676714493e-06,
"loss": 0.4202,
"step": 1215
},
{
"epoch": 3.514450867052023,
"grad_norm": 0.48637370325861295,
"learning_rate": 4.912886018395812e-06,
"loss": 0.3939,
"step": 1216
},
{
"epoch": 3.5173410404624277,
"grad_norm": 0.4827838362996411,
"learning_rate": 4.89552507073248e-06,
"loss": 0.4036,
"step": 1217
},
{
"epoch": 3.520231213872832,
"grad_norm": 0.5081515501494188,
"learning_rate": 4.878184904404375e-06,
"loss": 0.4024,
"step": 1218
},
{
"epoch": 3.523121387283237,
"grad_norm": 0.509383375704138,
"learning_rate": 4.86086559000676e-06,
"loss": 0.4007,
"step": 1219
},
{
"epoch": 3.5260115606936417,
"grad_norm": 0.5044115703037381,
"learning_rate": 4.843567198050031e-06,
"loss": 0.4243,
"step": 1220
},
{
"epoch": 3.5289017341040463,
"grad_norm": 0.5019649447142344,
"learning_rate": 4.8262897989593846e-06,
"loss": 0.3872,
"step": 1221
},
{
"epoch": 3.531791907514451,
"grad_norm": 0.4958060052336904,
"learning_rate": 4.809033463074564e-06,
"loss": 0.3898,
"step": 1222
},
{
"epoch": 3.5346820809248554,
"grad_norm": 0.4656358996491728,
"learning_rate": 4.791798260649538e-06,
"loss": 0.382,
"step": 1223
},
{
"epoch": 3.5375722543352603,
"grad_norm": 0.5046209644376203,
"learning_rate": 4.774584261852266e-06,
"loss": 0.3973,
"step": 1224
},
{
"epoch": 3.540462427745665,
"grad_norm": 0.473802279248715,
"learning_rate": 4.757391536764366e-06,
"loss": 0.3669,
"step": 1225
},
{
"epoch": 3.5433526011560694,
"grad_norm": 0.47987460911133734,
"learning_rate": 4.74022015538085e-06,
"loss": 0.4229,
"step": 1226
},
{
"epoch": 3.546242774566474,
"grad_norm": 0.47898689908086095,
"learning_rate": 4.723070187609834e-06,
"loss": 0.3934,
"step": 1227
},
{
"epoch": 3.5491329479768785,
"grad_norm": 0.5192852542397911,
"learning_rate": 4.705941703272256e-06,
"loss": 0.3769,
"step": 1228
},
{
"epoch": 3.5520231213872835,
"grad_norm": 0.4773762683627811,
"learning_rate": 4.688834772101593e-06,
"loss": 0.4012,
"step": 1229
},
{
"epoch": 3.5549132947976876,
"grad_norm": 0.4845549925124719,
"learning_rate": 4.671749463743572e-06,
"loss": 0.3907,
"step": 1230
},
{
"epoch": 3.5578034682080926,
"grad_norm": 0.48089857297464916,
"learning_rate": 4.65468584775589e-06,
"loss": 0.384,
"step": 1231
},
{
"epoch": 3.560693641618497,
"grad_norm": 0.5174166036846797,
"learning_rate": 4.637643993607928e-06,
"loss": 0.3983,
"step": 1232
},
{
"epoch": 3.5635838150289016,
"grad_norm": 0.5121238560523732,
"learning_rate": 4.620623970680473e-06,
"loss": 0.3893,
"step": 1233
},
{
"epoch": 3.5664739884393066,
"grad_norm": 0.516055886227455,
"learning_rate": 4.60362584826544e-06,
"loss": 0.3895,
"step": 1234
},
{
"epoch": 3.5693641618497107,
"grad_norm": 0.4991463204559961,
"learning_rate": 4.586649695565563e-06,
"loss": 0.3996,
"step": 1235
},
{
"epoch": 3.5722543352601157,
"grad_norm": 0.4823769438999723,
"learning_rate": 4.569695581694149e-06,
"loss": 0.399,
"step": 1236
},
{
"epoch": 3.5751445086705202,
"grad_norm": 0.547610891529702,
"learning_rate": 4.552763575674772e-06,
"loss": 0.3975,
"step": 1237
},
{
"epoch": 3.578034682080925,
"grad_norm": 0.49499114644780257,
"learning_rate": 4.535853746441018e-06,
"loss": 0.3927,
"step": 1238
},
{
"epoch": 3.5809248554913293,
"grad_norm": 0.48678386713561067,
"learning_rate": 4.518966162836159e-06,
"loss": 0.3871,
"step": 1239
},
{
"epoch": 3.583815028901734,
"grad_norm": 0.5034769399989679,
"learning_rate": 4.502100893612922e-06,
"loss": 0.416,
"step": 1240
},
{
"epoch": 3.586705202312139,
"grad_norm": 0.5066444239071285,
"learning_rate": 4.485258007433175e-06,
"loss": 0.4096,
"step": 1241
},
{
"epoch": 3.5895953757225434,
"grad_norm": 0.5016451005013188,
"learning_rate": 4.468437572867673e-06,
"loss": 0.4118,
"step": 1242
},
{
"epoch": 3.592485549132948,
"grad_norm": 0.5150332892840315,
"learning_rate": 4.451639658395761e-06,
"loss": 0.3969,
"step": 1243
},
{
"epoch": 3.5953757225433525,
"grad_norm": 0.5549645396194977,
"learning_rate": 4.434864332405085e-06,
"loss": 0.4178,
"step": 1244
},
{
"epoch": 3.598265895953757,
"grad_norm": 0.565355807978924,
"learning_rate": 4.418111663191355e-06,
"loss": 0.3883,
"step": 1245
},
{
"epoch": 3.601156069364162,
"grad_norm": 0.5472283526926208,
"learning_rate": 4.401381718958022e-06,
"loss": 0.4079,
"step": 1246
},
{
"epoch": 3.6040462427745665,
"grad_norm": 0.5124990525593747,
"learning_rate": 4.384674567816029e-06,
"loss": 0.4001,
"step": 1247
},
{
"epoch": 3.606936416184971,
"grad_norm": 0.4889182687602902,
"learning_rate": 4.36799027778351e-06,
"loss": 0.3888,
"step": 1248
},
{
"epoch": 3.6098265895953756,
"grad_norm": 0.48849681036045084,
"learning_rate": 4.3513289167855464e-06,
"loss": 0.3973,
"step": 1249
},
{
"epoch": 3.61271676300578,
"grad_norm": 0.5325960144702762,
"learning_rate": 4.3346905526538576e-06,
"loss": 0.4119,
"step": 1250
},
{
"epoch": 3.615606936416185,
"grad_norm": 0.5252422646593361,
"learning_rate": 4.31807525312654e-06,
"loss": 0.3844,
"step": 1251
},
{
"epoch": 3.6184971098265897,
"grad_norm": 0.515346473715291,
"learning_rate": 4.301483085847793e-06,
"loss": 0.406,
"step": 1252
},
{
"epoch": 3.621387283236994,
"grad_norm": 0.5590943402054944,
"learning_rate": 4.284914118367637e-06,
"loss": 0.3971,
"step": 1253
},
{
"epoch": 3.6242774566473988,
"grad_norm": 0.5178331984088388,
"learning_rate": 4.268368418141642e-06,
"loss": 0.4122,
"step": 1254
},
{
"epoch": 3.6271676300578033,
"grad_norm": 0.49667709804077503,
"learning_rate": 4.251846052530653e-06,
"loss": 0.3802,
"step": 1255
},
{
"epoch": 3.6300578034682083,
"grad_norm": 0.5084603762461885,
"learning_rate": 4.235347088800515e-06,
"loss": 0.3916,
"step": 1256
},
{
"epoch": 3.632947976878613,
"grad_norm": 0.5694214487049762,
"learning_rate": 4.2188715941217996e-06,
"loss": 0.4046,
"step": 1257
},
{
"epoch": 3.6358381502890174,
"grad_norm": 0.5171477728373481,
"learning_rate": 4.2024196355695254e-06,
"loss": 0.3905,
"step": 1258
},
{
"epoch": 3.638728323699422,
"grad_norm": 0.47173687115678026,
"learning_rate": 4.185991280122909e-06,
"loss": 0.3884,
"step": 1259
},
{
"epoch": 3.6416184971098264,
"grad_norm": 0.5208263137887216,
"learning_rate": 4.169586594665048e-06,
"loss": 0.3914,
"step": 1260
},
{
"epoch": 3.6445086705202314,
"grad_norm": 0.5659791491973168,
"learning_rate": 4.153205645982693e-06,
"loss": 0.3951,
"step": 1261
},
{
"epoch": 3.647398843930636,
"grad_norm": 0.6075893121263456,
"learning_rate": 4.136848500765948e-06,
"loss": 0.3949,
"step": 1262
},
{
"epoch": 3.6502890173410405,
"grad_norm": 0.512881571716333,
"learning_rate": 4.120515225608025e-06,
"loss": 0.3849,
"step": 1263
},
{
"epoch": 3.653179190751445,
"grad_norm": 0.5075832985581248,
"learning_rate": 4.104205887004931e-06,
"loss": 0.3824,
"step": 1264
},
{
"epoch": 3.6560693641618496,
"grad_norm": 0.5076914828789318,
"learning_rate": 4.08792055135524e-06,
"loss": 0.4022,
"step": 1265
},
{
"epoch": 3.6589595375722546,
"grad_norm": 0.5025241926533359,
"learning_rate": 4.0716592849597955e-06,
"loss": 0.3905,
"step": 1266
},
{
"epoch": 3.661849710982659,
"grad_norm": 0.496222438413654,
"learning_rate": 4.055422154021465e-06,
"loss": 0.3892,
"step": 1267
},
{
"epoch": 3.6647398843930636,
"grad_norm": 0.48534032089531876,
"learning_rate": 4.039209224644845e-06,
"loss": 0.3886,
"step": 1268
},
{
"epoch": 3.667630057803468,
"grad_norm": 0.5418497860771618,
"learning_rate": 4.0230205628359945e-06,
"loss": 0.3982,
"step": 1269
},
{
"epoch": 3.6705202312138727,
"grad_norm": 0.5147052856265083,
"learning_rate": 4.006856234502191e-06,
"loss": 0.39,
"step": 1270
},
{
"epoch": 3.6734104046242777,
"grad_norm": 0.494017251636487,
"learning_rate": 3.990716305451636e-06,
"loss": 0.4083,
"step": 1271
},
{
"epoch": 3.6763005780346822,
"grad_norm": 0.5717362220657668,
"learning_rate": 3.974600841393204e-06,
"loss": 0.3824,
"step": 1272
},
{
"epoch": 3.679190751445087,
"grad_norm": 0.5763029540574294,
"learning_rate": 3.958509907936149e-06,
"loss": 0.3934,
"step": 1273
},
{
"epoch": 3.6820809248554913,
"grad_norm": 0.47695488593530383,
"learning_rate": 3.942443570589882e-06,
"loss": 0.4102,
"step": 1274
},
{
"epoch": 3.684971098265896,
"grad_norm": 0.46522466243173516,
"learning_rate": 3.9264018947636626e-06,
"loss": 0.4064,
"step": 1275
},
{
"epoch": 3.687861271676301,
"grad_norm": 0.4901388621576566,
"learning_rate": 3.910384945766351e-06,
"loss": 0.4064,
"step": 1276
},
{
"epoch": 3.690751445086705,
"grad_norm": 0.5319476053231882,
"learning_rate": 3.894392788806141e-06,
"loss": 0.3584,
"step": 1277
},
{
"epoch": 3.69364161849711,
"grad_norm": 0.50332879249549,
"learning_rate": 3.878425488990289e-06,
"loss": 0.4003,
"step": 1278
},
{
"epoch": 3.6965317919075145,
"grad_norm": 0.4718359464324147,
"learning_rate": 3.862483111324861e-06,
"loss": 0.4125,
"step": 1279
},
{
"epoch": 3.699421965317919,
"grad_norm": 0.5047398998313649,
"learning_rate": 3.846565720714451e-06,
"loss": 0.395,
"step": 1280
},
{
"epoch": 3.7023121387283235,
"grad_norm": 0.5229923048712791,
"learning_rate": 3.830673381961933e-06,
"loss": 0.3812,
"step": 1281
},
{
"epoch": 3.705202312138728,
"grad_norm": 0.5775482777590354,
"learning_rate": 3.814806159768187e-06,
"loss": 0.3894,
"step": 1282
},
{
"epoch": 3.708092485549133,
"grad_norm": 0.49362257902570184,
"learning_rate": 3.7989641187318327e-06,
"loss": 0.3914,
"step": 1283
},
{
"epoch": 3.7109826589595376,
"grad_norm": 0.4725314340419177,
"learning_rate": 3.783147323348989e-06,
"loss": 0.4157,
"step": 1284
},
{
"epoch": 3.713872832369942,
"grad_norm": 0.493051522080004,
"learning_rate": 3.767355838012974e-06,
"loss": 0.3932,
"step": 1285
},
{
"epoch": 3.7167630057803467,
"grad_norm": 0.4938182282472529,
"learning_rate": 3.7515897270140755e-06,
"loss": 0.4056,
"step": 1286
},
{
"epoch": 3.7196531791907512,
"grad_norm": 0.5050054128840156,
"learning_rate": 3.7358490545392734e-06,
"loss": 0.3993,
"step": 1287
},
{
"epoch": 3.722543352601156,
"grad_norm": 0.4855493281052087,
"learning_rate": 3.7201338846719926e-06,
"loss": 0.397,
"step": 1288
},
{
"epoch": 3.7254335260115607,
"grad_norm": 0.4948276672840782,
"learning_rate": 3.7044442813918125e-06,
"loss": 0.4146,
"step": 1289
},
{
"epoch": 3.7283236994219653,
"grad_norm": 0.465286865663395,
"learning_rate": 3.688780308574238e-06,
"loss": 0.4193,
"step": 1290
},
{
"epoch": 3.73121387283237,
"grad_norm": 0.5040094181912368,
"learning_rate": 3.6731420299904317e-06,
"loss": 0.3888,
"step": 1291
},
{
"epoch": 3.7341040462427744,
"grad_norm": 0.484768931579548,
"learning_rate": 3.6575295093069397e-06,
"loss": 0.3908,
"step": 1292
},
{
"epoch": 3.7369942196531793,
"grad_norm": 0.49290459637271655,
"learning_rate": 3.6419428100854515e-06,
"loss": 0.3783,
"step": 1293
},
{
"epoch": 3.739884393063584,
"grad_norm": 0.5173781080231862,
"learning_rate": 3.6263819957825163e-06,
"loss": 0.4028,
"step": 1294
},
{
"epoch": 3.7427745664739884,
"grad_norm": 0.48333701877259566,
"learning_rate": 3.610847129749323e-06,
"loss": 0.3908,
"step": 1295
},
{
"epoch": 3.745664739884393,
"grad_norm": 0.46514377347373626,
"learning_rate": 3.5953382752314036e-06,
"loss": 0.4052,
"step": 1296
},
{
"epoch": 3.7485549132947975,
"grad_norm": 0.4611540713015221,
"learning_rate": 3.579855495368403e-06,
"loss": 0.4148,
"step": 1297
},
{
"epoch": 3.7514450867052025,
"grad_norm": 0.4918418237562193,
"learning_rate": 3.5643988531937923e-06,
"loss": 0.3999,
"step": 1298
},
{
"epoch": 3.754335260115607,
"grad_norm": 0.48102179878442763,
"learning_rate": 3.5489684116346556e-06,
"loss": 0.3781,
"step": 1299
},
{
"epoch": 3.7572254335260116,
"grad_norm": 0.5010976552573143,
"learning_rate": 3.533564233511394e-06,
"loss": 0.3775,
"step": 1300
},
{
"epoch": 3.760115606936416,
"grad_norm": 0.46729685486727424,
"learning_rate": 3.518186381537486e-06,
"loss": 0.373,
"step": 1301
},
{
"epoch": 3.7630057803468207,
"grad_norm": 0.48872167812753237,
"learning_rate": 3.5028349183192365e-06,
"loss": 0.3839,
"step": 1302
},
{
"epoch": 3.7658959537572256,
"grad_norm": 0.4833733217230953,
"learning_rate": 3.4875099063555105e-06,
"loss": 0.4164,
"step": 1303
},
{
"epoch": 3.76878612716763,
"grad_norm": 0.48933797219192665,
"learning_rate": 3.472211408037488e-06,
"loss": 0.4062,
"step": 1304
},
{
"epoch": 3.7716763005780347,
"grad_norm": 0.48178223947944576,
"learning_rate": 3.456939485648406e-06,
"loss": 0.3915,
"step": 1305
},
{
"epoch": 3.7745664739884393,
"grad_norm": 0.5220222059582195,
"learning_rate": 3.441694201363307e-06,
"loss": 0.3781,
"step": 1306
},
{
"epoch": 3.777456647398844,
"grad_norm": 0.5024321900572036,
"learning_rate": 3.4264756172487813e-06,
"loss": 0.3738,
"step": 1307
},
{
"epoch": 3.7803468208092488,
"grad_norm": 0.5223109129718773,
"learning_rate": 3.4112837952627175e-06,
"loss": 0.4036,
"step": 1308
},
{
"epoch": 3.7832369942196533,
"grad_norm": 0.4840530871394166,
"learning_rate": 3.3961187972540598e-06,
"loss": 0.3788,
"step": 1309
},
{
"epoch": 3.786127167630058,
"grad_norm": 0.5403498006249526,
"learning_rate": 3.3809806849625315e-06,
"loss": 0.3727,
"step": 1310
},
{
"epoch": 3.7890173410404624,
"grad_norm": 0.4959578369395937,
"learning_rate": 3.3658695200184075e-06,
"loss": 0.3714,
"step": 1311
},
{
"epoch": 3.791907514450867,
"grad_norm": 0.46367701058442384,
"learning_rate": 3.3507853639422514e-06,
"loss": 0.3886,
"step": 1312
},
{
"epoch": 3.794797687861272,
"grad_norm": 0.5315085676034551,
"learning_rate": 3.3357282781446787e-06,
"loss": 0.4034,
"step": 1313
},
{
"epoch": 3.7976878612716765,
"grad_norm": 0.5185261007954255,
"learning_rate": 3.320698323926079e-06,
"loss": 0.3988,
"step": 1314
},
{
"epoch": 3.800578034682081,
"grad_norm": 0.5194949353467634,
"learning_rate": 3.305695562476393e-06,
"loss": 0.3799,
"step": 1315
},
{
"epoch": 3.8034682080924855,
"grad_norm": 0.47443505146882026,
"learning_rate": 3.29072005487486e-06,
"loss": 0.4158,
"step": 1316
},
{
"epoch": 3.80635838150289,
"grad_norm": 0.4720956993326887,
"learning_rate": 3.2757718620897574e-06,
"loss": 0.4226,
"step": 1317
},
{
"epoch": 3.809248554913295,
"grad_norm": 0.5187600274566376,
"learning_rate": 3.2608510449781526e-06,
"loss": 0.4108,
"step": 1318
},
{
"epoch": 3.812138728323699,
"grad_norm": 0.48257622577269477,
"learning_rate": 3.245957664285666e-06,
"loss": 0.4029,
"step": 1319
},
{
"epoch": 3.815028901734104,
"grad_norm": 0.48180267094515367,
"learning_rate": 3.2310917806462274e-06,
"loss": 0.398,
"step": 1320
},
{
"epoch": 3.8179190751445087,
"grad_norm": 0.5990856236553259,
"learning_rate": 3.2162534545818104e-06,
"loss": 0.385,
"step": 1321
},
{
"epoch": 3.820809248554913,
"grad_norm": 0.5016426592485167,
"learning_rate": 3.2014427465022014e-06,
"loss": 0.3844,
"step": 1322
},
{
"epoch": 3.8236994219653178,
"grad_norm": 0.48857761234195757,
"learning_rate": 3.1866597167047364e-06,
"loss": 0.4028,
"step": 1323
},
{
"epoch": 3.8265895953757223,
"grad_norm": 0.47082788199594566,
"learning_rate": 3.1719044253740874e-06,
"loss": 0.4185,
"step": 1324
},
{
"epoch": 3.8294797687861273,
"grad_norm": 0.4839708323294498,
"learning_rate": 3.1571769325819834e-06,
"loss": 0.3902,
"step": 1325
},
{
"epoch": 3.832369942196532,
"grad_norm": 0.46860357381932183,
"learning_rate": 3.1424772982869866e-06,
"loss": 0.3887,
"step": 1326
},
{
"epoch": 3.8352601156069364,
"grad_norm": 0.4867042348950247,
"learning_rate": 3.127805582334236e-06,
"loss": 0.3944,
"step": 1327
},
{
"epoch": 3.838150289017341,
"grad_norm": 0.5089482015595527,
"learning_rate": 3.1131618444552148e-06,
"loss": 0.3823,
"step": 1328
},
{
"epoch": 3.8410404624277454,
"grad_norm": 0.4886105855662341,
"learning_rate": 3.098546144267499e-06,
"loss": 0.4218,
"step": 1329
},
{
"epoch": 3.8439306358381504,
"grad_norm": 0.5143763846812575,
"learning_rate": 3.0839585412745178e-06,
"loss": 0.3961,
"step": 1330
},
{
"epoch": 3.846820809248555,
"grad_norm": 0.47071190340225716,
"learning_rate": 3.069399094865312e-06,
"loss": 0.3972,
"step": 1331
},
{
"epoch": 3.8497109826589595,
"grad_norm": 0.49795601399511025,
"learning_rate": 3.0548678643142916e-06,
"loss": 0.3891,
"step": 1332
},
{
"epoch": 3.852601156069364,
"grad_norm": 0.5066155576475394,
"learning_rate": 3.0403649087809905e-06,
"loss": 0.4085,
"step": 1333
},
{
"epoch": 3.8554913294797686,
"grad_norm": 0.4720442115733795,
"learning_rate": 3.0258902873098406e-06,
"loss": 0.3927,
"step": 1334
},
{
"epoch": 3.8583815028901736,
"grad_norm": 0.5187384473204723,
"learning_rate": 3.011444058829903e-06,
"loss": 0.3834,
"step": 1335
},
{
"epoch": 3.861271676300578,
"grad_norm": 0.46952384983808076,
"learning_rate": 2.997026282154658e-06,
"loss": 0.4172,
"step": 1336
},
{
"epoch": 3.8641618497109826,
"grad_norm": 0.47076026733477655,
"learning_rate": 2.9826370159817465e-06,
"loss": 0.3982,
"step": 1337
},
{
"epoch": 3.867052023121387,
"grad_norm": 0.48222677969659716,
"learning_rate": 2.9682763188927476e-06,
"loss": 0.3895,
"step": 1338
},
{
"epoch": 3.8699421965317917,
"grad_norm": 0.4707754155595722,
"learning_rate": 2.953944249352916e-06,
"loss": 0.3982,
"step": 1339
},
{
"epoch": 3.8728323699421967,
"grad_norm": 0.48597477112642706,
"learning_rate": 2.939640865710961e-06,
"loss": 0.3828,
"step": 1340
},
{
"epoch": 3.8757225433526012,
"grad_norm": 0.4893204505305053,
"learning_rate": 2.925366226198819e-06,
"loss": 0.384,
"step": 1341
},
{
"epoch": 3.878612716763006,
"grad_norm": 0.4511534069636998,
"learning_rate": 2.911120388931392e-06,
"loss": 0.4129,
"step": 1342
},
{
"epoch": 3.8815028901734103,
"grad_norm": 0.4665021353659143,
"learning_rate": 2.8969034119063176e-06,
"loss": 0.3853,
"step": 1343
},
{
"epoch": 3.884393063583815,
"grad_norm": 0.46725384145430443,
"learning_rate": 2.8827153530037444e-06,
"loss": 0.3926,
"step": 1344
},
{
"epoch": 3.88728323699422,
"grad_norm": 0.4463766065379448,
"learning_rate": 2.868556269986096e-06,
"loss": 0.4163,
"step": 1345
},
{
"epoch": 3.8901734104046244,
"grad_norm": 0.4641928583572942,
"learning_rate": 2.8544262204978178e-06,
"loss": 0.379,
"step": 1346
},
{
"epoch": 3.893063583815029,
"grad_norm": 0.6233225476901676,
"learning_rate": 2.840325262065161e-06,
"loss": 0.3705,
"step": 1347
},
{
"epoch": 3.8959537572254335,
"grad_norm": 0.46924299510074996,
"learning_rate": 2.826253452095937e-06,
"loss": 0.4119,
"step": 1348
},
{
"epoch": 3.898843930635838,
"grad_norm": 0.4834134656368184,
"learning_rate": 2.812210847879291e-06,
"loss": 0.3861,
"step": 1349
},
{
"epoch": 3.901734104046243,
"grad_norm": 0.46251839056953825,
"learning_rate": 2.798197506585464e-06,
"loss": 0.3822,
"step": 1350
},
{
"epoch": 3.9046242774566475,
"grad_norm": 0.5271881209856408,
"learning_rate": 2.7842134852655635e-06,
"loss": 0.3849,
"step": 1351
},
{
"epoch": 3.907514450867052,
"grad_norm": 0.4798577219404323,
"learning_rate": 2.7702588408513276e-06,
"loss": 0.3879,
"step": 1352
},
{
"epoch": 3.9104046242774566,
"grad_norm": 0.47590125838213543,
"learning_rate": 2.756333630154897e-06,
"loss": 0.4067,
"step": 1353
},
{
"epoch": 3.913294797687861,
"grad_norm": 0.4909876690916432,
"learning_rate": 2.7424379098685806e-06,
"loss": 0.3943,
"step": 1354
},
{
"epoch": 3.916184971098266,
"grad_norm": 0.4989923470121859,
"learning_rate": 2.7285717365646258e-06,
"loss": 0.4009,
"step": 1355
},
{
"epoch": 3.9190751445086707,
"grad_norm": 0.47528546667752497,
"learning_rate": 2.7147351666949904e-06,
"loss": 0.3988,
"step": 1356
},
{
"epoch": 3.921965317919075,
"grad_norm": 0.43780206676111283,
"learning_rate": 2.7009282565911077e-06,
"loss": 0.3959,
"step": 1357
},
{
"epoch": 3.9248554913294798,
"grad_norm": 0.4679114324049943,
"learning_rate": 2.6871510624636586e-06,
"loss": 0.3951,
"step": 1358
},
{
"epoch": 3.9277456647398843,
"grad_norm": 0.4637620747382237,
"learning_rate": 2.673403640402359e-06,
"loss": 0.4088,
"step": 1359
},
{
"epoch": 3.9306358381502893,
"grad_norm": 0.45457986314823645,
"learning_rate": 2.6596860463756935e-06,
"loss": 0.3929,
"step": 1360
},
{
"epoch": 3.9335260115606934,
"grad_norm": 0.4761238450741067,
"learning_rate": 2.6459983362307263e-06,
"loss": 0.383,
"step": 1361
},
{
"epoch": 3.9364161849710984,
"grad_norm": 0.4927539770863447,
"learning_rate": 2.632340565692851e-06,
"loss": 0.3787,
"step": 1362
},
{
"epoch": 3.939306358381503,
"grad_norm": 0.4858307892706921,
"learning_rate": 2.618712790365585e-06,
"loss": 0.3924,
"step": 1363
},
{
"epoch": 3.9421965317919074,
"grad_norm": 0.46312831934840626,
"learning_rate": 2.6051150657303103e-06,
"loss": 0.4016,
"step": 1364
},
{
"epoch": 3.9450867052023124,
"grad_norm": 0.4777949934568993,
"learning_rate": 2.5915474471460733e-06,
"loss": 0.3923,
"step": 1365
},
{
"epoch": 3.9479768786127165,
"grad_norm": 0.506011577237325,
"learning_rate": 2.5780099898493617e-06,
"loss": 0.3773,
"step": 1366
},
{
"epoch": 3.9508670520231215,
"grad_norm": 0.49038674746439304,
"learning_rate": 2.5645027489538633e-06,
"loss": 0.3909,
"step": 1367
},
{
"epoch": 3.953757225433526,
"grad_norm": 0.47337641434225514,
"learning_rate": 2.5510257794502423e-06,
"loss": 0.3887,
"step": 1368
},
{
"epoch": 3.9566473988439306,
"grad_norm": 0.47126439390870567,
"learning_rate": 2.5375791362059287e-06,
"loss": 0.3946,
"step": 1369
},
{
"epoch": 3.959537572254335,
"grad_norm": 0.49310721717359374,
"learning_rate": 2.524162873964896e-06,
"loss": 0.4,
"step": 1370
},
{
"epoch": 3.9624277456647397,
"grad_norm": 0.5107504252689111,
"learning_rate": 2.510777047347419e-06,
"loss": 0.3738,
"step": 1371
},
{
"epoch": 3.9653179190751446,
"grad_norm": 0.48531406074206546,
"learning_rate": 2.4974217108498656e-06,
"loss": 0.3981,
"step": 1372
},
{
"epoch": 3.968208092485549,
"grad_norm": 0.49546193426834495,
"learning_rate": 2.4840969188444754e-06,
"loss": 0.3835,
"step": 1373
},
{
"epoch": 3.9710982658959537,
"grad_norm": 0.46548489662595355,
"learning_rate": 2.4708027255791323e-06,
"loss": 0.4155,
"step": 1374
},
{
"epoch": 3.9739884393063583,
"grad_norm": 0.44586516274034266,
"learning_rate": 2.457539185177148e-06,
"loss": 0.3964,
"step": 1375
},
{
"epoch": 3.976878612716763,
"grad_norm": 0.5251475706820591,
"learning_rate": 2.4443063516370415e-06,
"loss": 0.3769,
"step": 1376
},
{
"epoch": 3.979768786127168,
"grad_norm": 0.45717894373015,
"learning_rate": 2.431104278832315e-06,
"loss": 0.4089,
"step": 1377
},
{
"epoch": 3.9826589595375723,
"grad_norm": 0.4770020764557666,
"learning_rate": 2.4179330205112383e-06,
"loss": 0.3949,
"step": 1378
},
{
"epoch": 3.985549132947977,
"grad_norm": 0.48235033068032995,
"learning_rate": 2.404792630296633e-06,
"loss": 0.3924,
"step": 1379
},
{
"epoch": 3.9884393063583814,
"grad_norm": 0.4878128649867659,
"learning_rate": 2.3916831616856475e-06,
"loss": 0.4063,
"step": 1380
},
{
"epoch": 3.991329479768786,
"grad_norm": 0.46889111549403895,
"learning_rate": 2.37860466804954e-06,
"loss": 0.3921,
"step": 1381
},
{
"epoch": 3.994219653179191,
"grad_norm": 0.4566000580810184,
"learning_rate": 2.3655572026334674e-06,
"loss": 0.4113,
"step": 1382
},
{
"epoch": 3.9971098265895955,
"grad_norm": 0.5291952784963643,
"learning_rate": 2.3525408185562626e-06,
"loss": 0.3827,
"step": 1383
},
{
"epoch": 4.0,
"grad_norm": 0.5182380653301021,
"learning_rate": 2.339555568810221e-06,
"loss": 0.3698,
"step": 1384
},
{
"epoch": 4.002890173410405,
"grad_norm": 0.6698862793434772,
"learning_rate": 2.3266015062608837e-06,
"loss": 0.3456,
"step": 1385
},
{
"epoch": 4.005780346820809,
"grad_norm": 0.5774888203625725,
"learning_rate": 2.3136786836468204e-06,
"loss": 0.3233,
"step": 1386
},
{
"epoch": 4.008670520231214,
"grad_norm": 0.496192301374043,
"learning_rate": 2.300787153579418e-06,
"loss": 0.338,
"step": 1387
},
{
"epoch": 4.011560693641618,
"grad_norm": 0.6361731482559027,
"learning_rate": 2.2879269685426742e-06,
"loss": 0.3243,
"step": 1388
},
{
"epoch": 4.014450867052023,
"grad_norm": 0.7439747647099445,
"learning_rate": 2.2750981808929563e-06,
"loss": 0.349,
"step": 1389
},
{
"epoch": 4.017341040462428,
"grad_norm": 0.5589574759688413,
"learning_rate": 2.262300842858818e-06,
"loss": 0.3521,
"step": 1390
},
{
"epoch": 4.020231213872832,
"grad_norm": 0.4896141267394832,
"learning_rate": 2.2495350065407797e-06,
"loss": 0.3426,
"step": 1391
},
{
"epoch": 4.023121387283237,
"grad_norm": 0.5639827670600378,
"learning_rate": 2.236800723911107e-06,
"loss": 0.3395,
"step": 1392
},
{
"epoch": 4.026011560693641,
"grad_norm": 0.5752090316401731,
"learning_rate": 2.2240980468135962e-06,
"loss": 0.3644,
"step": 1393
},
{
"epoch": 4.028901734104046,
"grad_norm": 0.5788393068913655,
"learning_rate": 2.21142702696338e-06,
"loss": 0.3454,
"step": 1394
},
{
"epoch": 4.031791907514451,
"grad_norm": 0.5063205120505316,
"learning_rate": 2.198787715946712e-06,
"loss": 0.334,
"step": 1395
},
{
"epoch": 4.034682080924855,
"grad_norm": 0.46871942776478365,
"learning_rate": 2.1861801652207472e-06,
"loss": 0.352,
"step": 1396
},
{
"epoch": 4.03757225433526,
"grad_norm": 0.4814257774715175,
"learning_rate": 2.1736044261133305e-06,
"loss": 0.3537,
"step": 1397
},
{
"epoch": 4.040462427745664,
"grad_norm": 0.4944947737271302,
"learning_rate": 2.161060549822811e-06,
"loss": 0.3443,
"step": 1398
},
{
"epoch": 4.043352601156069,
"grad_norm": 0.5422000937512472,
"learning_rate": 2.148548587417808e-06,
"loss": 0.3629,
"step": 1399
},
{
"epoch": 4.046242774566474,
"grad_norm": 0.5033172587407463,
"learning_rate": 2.1360685898370147e-06,
"loss": 0.37,
"step": 1400
},
{
"epoch": 4.0491329479768785,
"grad_norm": 0.47142049139932146,
"learning_rate": 2.123620607888991e-06,
"loss": 0.3451,
"step": 1401
},
{
"epoch": 4.0520231213872835,
"grad_norm": 0.4592932152977094,
"learning_rate": 2.1112046922519524e-06,
"loss": 0.3365,
"step": 1402
},
{
"epoch": 4.054913294797688,
"grad_norm": 0.4677681799519137,
"learning_rate": 2.0988208934735664e-06,
"loss": 0.332,
"step": 1403
},
{
"epoch": 4.057803468208093,
"grad_norm": 0.47746092825207925,
"learning_rate": 2.0864692619707493e-06,
"loss": 0.3152,
"step": 1404
},
{
"epoch": 4.0606936416184976,
"grad_norm": 0.48786302102807205,
"learning_rate": 2.074149848029453e-06,
"loss": 0.3359,
"step": 1405
},
{
"epoch": 4.063583815028902,
"grad_norm": 0.4775605547438963,
"learning_rate": 2.06186270180447e-06,
"loss": 0.3126,
"step": 1406
},
{
"epoch": 4.066473988439307,
"grad_norm": 0.46623710415414105,
"learning_rate": 2.0496078733192216e-06,
"loss": 0.3415,
"step": 1407
},
{
"epoch": 4.069364161849711,
"grad_norm": 0.5176099022694585,
"learning_rate": 2.037385412465558e-06,
"loss": 0.3048,
"step": 1408
},
{
"epoch": 4.072254335260116,
"grad_norm": 0.4787733747514296,
"learning_rate": 2.0251953690035543e-06,
"loss": 0.3398,
"step": 1409
},
{
"epoch": 4.07514450867052,
"grad_norm": 0.47788782780581107,
"learning_rate": 2.01303779256131e-06,
"loss": 0.3416,
"step": 1410
},
{
"epoch": 4.078034682080925,
"grad_norm": 0.4601914046139569,
"learning_rate": 2.00091273263474e-06,
"loss": 0.3271,
"step": 1411
},
{
"epoch": 4.08092485549133,
"grad_norm": 0.4361807898527824,
"learning_rate": 1.988820238587379e-06,
"loss": 0.3391,
"step": 1412
},
{
"epoch": 4.083815028901734,
"grad_norm": 0.46670742374819324,
"learning_rate": 1.9767603596501915e-06,
"loss": 0.3339,
"step": 1413
},
{
"epoch": 4.086705202312139,
"grad_norm": 0.43632480253238454,
"learning_rate": 1.9647331449213393e-06,
"loss": 0.3686,
"step": 1414
},
{
"epoch": 4.089595375722543,
"grad_norm": 0.44832973060002657,
"learning_rate": 1.952738643366011e-06,
"loss": 0.3505,
"step": 1415
},
{
"epoch": 4.092485549132948,
"grad_norm": 0.45939235288075486,
"learning_rate": 1.9407769038162216e-06,
"loss": 0.3297,
"step": 1416
},
{
"epoch": 4.095375722543353,
"grad_norm": 0.4409722647472371,
"learning_rate": 1.9288479749705967e-06,
"loss": 0.3128,
"step": 1417
},
{
"epoch": 4.098265895953757,
"grad_norm": 0.47224181678349825,
"learning_rate": 1.9169519053941786e-06,
"loss": 0.3208,
"step": 1418
},
{
"epoch": 4.101156069364162,
"grad_norm": 0.47439726771277907,
"learning_rate": 1.9050887435182386e-06,
"loss": 0.3398,
"step": 1419
},
{
"epoch": 4.104046242774566,
"grad_norm": 0.4658176671090241,
"learning_rate": 1.8932585376400803e-06,
"loss": 0.3396,
"step": 1420
},
{
"epoch": 4.106936416184971,
"grad_norm": 0.45373368834737315,
"learning_rate": 1.8814613359228296e-06,
"loss": 0.3579,
"step": 1421
},
{
"epoch": 4.109826589595376,
"grad_norm": 0.4416328012965478,
"learning_rate": 1.8696971863952385e-06,
"loss": 0.3299,
"step": 1422
},
{
"epoch": 4.11271676300578,
"grad_norm": 0.48886940224883624,
"learning_rate": 1.8579661369515155e-06,
"loss": 0.3488,
"step": 1423
},
{
"epoch": 4.115606936416185,
"grad_norm": 0.4590041411360483,
"learning_rate": 1.8462682353510974e-06,
"loss": 0.331,
"step": 1424
},
{
"epoch": 4.118497109826589,
"grad_norm": 0.4495947318171395,
"learning_rate": 1.834603529218475e-06,
"loss": 0.3447,
"step": 1425
},
{
"epoch": 4.121387283236994,
"grad_norm": 0.5098247907708044,
"learning_rate": 1.8229720660429916e-06,
"loss": 0.3313,
"step": 1426
},
{
"epoch": 4.124277456647399,
"grad_norm": 0.45941639945045426,
"learning_rate": 1.8113738931786551e-06,
"loss": 0.3509,
"step": 1427
},
{
"epoch": 4.127167630057803,
"grad_norm": 0.46263196317597743,
"learning_rate": 1.7998090578439375e-06,
"loss": 0.3466,
"step": 1428
},
{
"epoch": 4.130057803468208,
"grad_norm": 0.4743888485808943,
"learning_rate": 1.7882776071215912e-06,
"loss": 0.3103,
"step": 1429
},
{
"epoch": 4.132947976878612,
"grad_norm": 0.4553318759212632,
"learning_rate": 1.7767795879584504e-06,
"loss": 0.352,
"step": 1430
},
{
"epoch": 4.135838150289017,
"grad_norm": 0.4535461955976109,
"learning_rate": 1.7653150471652435e-06,
"loss": 0.349,
"step": 1431
},
{
"epoch": 4.138728323699422,
"grad_norm": 0.43364595489226815,
"learning_rate": 1.753884031416403e-06,
"loss": 0.3393,
"step": 1432
},
{
"epoch": 4.141618497109826,
"grad_norm": 0.49226388211750033,
"learning_rate": 1.742486587249873e-06,
"loss": 0.349,
"step": 1433
},
{
"epoch": 4.144508670520231,
"grad_norm": 0.4588798495243272,
"learning_rate": 1.7311227610669202e-06,
"loss": 0.3455,
"step": 1434
},
{
"epoch": 4.1473988439306355,
"grad_norm": 0.4548756292362763,
"learning_rate": 1.7197925991319486e-06,
"loss": 0.3469,
"step": 1435
},
{
"epoch": 4.1502890173410405,
"grad_norm": 0.4582296941792938,
"learning_rate": 1.708496147572305e-06,
"loss": 0.3342,
"step": 1436
},
{
"epoch": 4.1531791907514455,
"grad_norm": 0.4708824572032611,
"learning_rate": 1.697233452378093e-06,
"loss": 0.3441,
"step": 1437
},
{
"epoch": 4.15606936416185,
"grad_norm": 0.45376778758330355,
"learning_rate": 1.6860045594020003e-06,
"loss": 0.3307,
"step": 1438
},
{
"epoch": 4.158959537572255,
"grad_norm": 0.4725339238588213,
"learning_rate": 1.6748095143590804e-06,
"loss": 0.3535,
"step": 1439
},
{
"epoch": 4.161849710982659,
"grad_norm": 0.4692515895943304,
"learning_rate": 1.6636483628265942e-06,
"loss": 0.3573,
"step": 1440
},
{
"epoch": 4.164739884393064,
"grad_norm": 0.4371382256412352,
"learning_rate": 1.6525211502438188e-06,
"loss": 0.3616,
"step": 1441
},
{
"epoch": 4.167630057803469,
"grad_norm": 0.4604484221618385,
"learning_rate": 1.6414279219118568e-06,
"loss": 0.354,
"step": 1442
},
{
"epoch": 4.170520231213873,
"grad_norm": 0.4431830293421041,
"learning_rate": 1.6303687229934461e-06,
"loss": 0.3476,
"step": 1443
},
{
"epoch": 4.173410404624278,
"grad_norm": 0.44939578224273397,
"learning_rate": 1.6193435985127926e-06,
"loss": 0.331,
"step": 1444
},
{
"epoch": 4.176300578034682,
"grad_norm": 0.4512301766319807,
"learning_rate": 1.60835259335538e-06,
"loss": 0.3471,
"step": 1445
},
{
"epoch": 4.179190751445087,
"grad_norm": 0.4476904324897404,
"learning_rate": 1.5973957522677818e-06,
"loss": 0.3319,
"step": 1446
},
{
"epoch": 4.182080924855491,
"grad_norm": 0.45894696212505476,
"learning_rate": 1.5864731198574768e-06,
"loss": 0.3534,
"step": 1447
},
{
"epoch": 4.184971098265896,
"grad_norm": 0.4431828943221232,
"learning_rate": 1.575584740592685e-06,
"loss": 0.3523,
"step": 1448
},
{
"epoch": 4.187861271676301,
"grad_norm": 0.4761638484119464,
"learning_rate": 1.5647306588021672e-06,
"loss": 0.3376,
"step": 1449
},
{
"epoch": 4.190751445086705,
"grad_norm": 0.4847200373219092,
"learning_rate": 1.5539109186750544e-06,
"loss": 0.3097,
"step": 1450
},
{
"epoch": 4.19364161849711,
"grad_norm": 0.46747500615396664,
"learning_rate": 1.543125564260668e-06,
"loss": 0.3339,
"step": 1451
},
{
"epoch": 4.196531791907514,
"grad_norm": 0.47459635670072037,
"learning_rate": 1.5323746394683348e-06,
"loss": 0.3398,
"step": 1452
},
{
"epoch": 4.199421965317919,
"grad_norm": 0.4544955022339449,
"learning_rate": 1.5216581880672122e-06,
"loss": 0.3489,
"step": 1453
},
{
"epoch": 4.202312138728324,
"grad_norm": 0.4726389666592722,
"learning_rate": 1.5109762536861127e-06,
"loss": 0.3438,
"step": 1454
},
{
"epoch": 4.205202312138728,
"grad_norm": 0.4311706192954136,
"learning_rate": 1.5003288798133197e-06,
"loss": 0.3499,
"step": 1455
},
{
"epoch": 4.208092485549133,
"grad_norm": 0.45368588849949304,
"learning_rate": 1.4897161097964164e-06,
"loss": 0.3261,
"step": 1456
},
{
"epoch": 4.210982658959537,
"grad_norm": 0.44332312547505837,
"learning_rate": 1.4791379868421052e-06,
"loss": 0.3467,
"step": 1457
},
{
"epoch": 4.213872832369942,
"grad_norm": 0.4530598562314223,
"learning_rate": 1.4685945540160328e-06,
"loss": 0.3245,
"step": 1458
},
{
"epoch": 4.216763005780347,
"grad_norm": 0.46322900429248803,
"learning_rate": 1.458085854242617e-06,
"loss": 0.3452,
"step": 1459
},
{
"epoch": 4.219653179190751,
"grad_norm": 0.45588154749357934,
"learning_rate": 1.4476119303048709e-06,
"loss": 0.3262,
"step": 1460
},
{
"epoch": 4.222543352601156,
"grad_norm": 0.44527165693981313,
"learning_rate": 1.437172824844224e-06,
"loss": 0.3373,
"step": 1461
},
{
"epoch": 4.22543352601156,
"grad_norm": 0.45716413877466067,
"learning_rate": 1.4267685803603615e-06,
"loss": 0.3352,
"step": 1462
},
{
"epoch": 4.228323699421965,
"grad_norm": 0.46836574889519467,
"learning_rate": 1.416399239211036e-06,
"loss": 0.3178,
"step": 1463
},
{
"epoch": 4.23121387283237,
"grad_norm": 0.4534593906103841,
"learning_rate": 1.4060648436118985e-06,
"loss": 0.3415,
"step": 1464
},
{
"epoch": 4.234104046242774,
"grad_norm": 0.45695949138372266,
"learning_rate": 1.395765435636335e-06,
"loss": 0.3449,
"step": 1465
},
{
"epoch": 4.236994219653179,
"grad_norm": 0.4651531043657846,
"learning_rate": 1.3855010572152927e-06,
"loss": 0.3426,
"step": 1466
},
{
"epoch": 4.2398843930635834,
"grad_norm": 0.4434611463546597,
"learning_rate": 1.3752717501371037e-06,
"loss": 0.3437,
"step": 1467
},
{
"epoch": 4.242774566473988,
"grad_norm": 0.4569031499337852,
"learning_rate": 1.3650775560473118e-06,
"loss": 0.3317,
"step": 1468
},
{
"epoch": 4.245664739884393,
"grad_norm": 0.4377009893024095,
"learning_rate": 1.3549185164485135e-06,
"loss": 0.3469,
"step": 1469
},
{
"epoch": 4.2485549132947975,
"grad_norm": 0.439428219289406,
"learning_rate": 1.3447946727001881e-06,
"loss": 0.3508,
"step": 1470
},
{
"epoch": 4.2514450867052025,
"grad_norm": 0.4240581035879888,
"learning_rate": 1.3347060660185251e-06,
"loss": 0.3337,
"step": 1471
},
{
"epoch": 4.254335260115607,
"grad_norm": 0.44821038073621283,
"learning_rate": 1.324652737476244e-06,
"loss": 0.3334,
"step": 1472
},
{
"epoch": 4.257225433526012,
"grad_norm": 0.4717586682934258,
"learning_rate": 1.3146347280024586e-06,
"loss": 0.3285,
"step": 1473
},
{
"epoch": 4.2601156069364166,
"grad_norm": 0.45970185661583424,
"learning_rate": 1.304652078382479e-06,
"loss": 0.3345,
"step": 1474
},
{
"epoch": 4.263005780346821,
"grad_norm": 0.4537595102615837,
"learning_rate": 1.2947048292576635e-06,
"loss": 0.3471,
"step": 1475
},
{
"epoch": 4.265895953757226,
"grad_norm": 0.4488199865437098,
"learning_rate": 1.284793021125247e-06,
"loss": 0.338,
"step": 1476
},
{
"epoch": 4.26878612716763,
"grad_norm": 0.4765959434746097,
"learning_rate": 1.2749166943381763e-06,
"loss": 0.3462,
"step": 1477
},
{
"epoch": 4.271676300578035,
"grad_norm": 0.45420852625737385,
"learning_rate": 1.2650758891049464e-06,
"loss": 0.3381,
"step": 1478
},
{
"epoch": 4.27456647398844,
"grad_norm": 0.4537304824757465,
"learning_rate": 1.255270645489438e-06,
"loss": 0.3455,
"step": 1479
},
{
"epoch": 4.277456647398844,
"grad_norm": 0.43717037502746525,
"learning_rate": 1.2455010034107529e-06,
"loss": 0.3358,
"step": 1480
},
{
"epoch": 4.280346820809249,
"grad_norm": 0.4507791761925461,
"learning_rate": 1.2357670026430524e-06,
"loss": 0.3448,
"step": 1481
},
{
"epoch": 4.283236994219653,
"grad_norm": 0.4511509118103023,
"learning_rate": 1.2260686828153934e-06,
"loss": 0.3451,
"step": 1482
},
{
"epoch": 4.286127167630058,
"grad_norm": 0.4734480889620212,
"learning_rate": 1.216406083411571e-06,
"loss": 0.3321,
"step": 1483
},
{
"epoch": 4.289017341040463,
"grad_norm": 0.4537221581801662,
"learning_rate": 1.2067792437699532e-06,
"loss": 0.3322,
"step": 1484
},
{
"epoch": 4.291907514450867,
"grad_norm": 0.4665588809965585,
"learning_rate": 1.1971882030833248e-06,
"loss": 0.3377,
"step": 1485
},
{
"epoch": 4.294797687861272,
"grad_norm": 0.42934685126372396,
"learning_rate": 1.1876330003987214e-06,
"loss": 0.348,
"step": 1486
},
{
"epoch": 4.297687861271676,
"grad_norm": 0.44276344174103444,
"learning_rate": 1.178113674617285e-06,
"loss": 0.34,
"step": 1487
},
{
"epoch": 4.300578034682081,
"grad_norm": 0.4370964807675226,
"learning_rate": 1.1686302644940817e-06,
"loss": 0.3395,
"step": 1488
},
{
"epoch": 4.303468208092486,
"grad_norm": 0.4473311511207275,
"learning_rate": 1.1591828086379697e-06,
"loss": 0.3225,
"step": 1489
},
{
"epoch": 4.30635838150289,
"grad_norm": 0.46150180784367567,
"learning_rate": 1.1497713455114212e-06,
"loss": 0.3517,
"step": 1490
},
{
"epoch": 4.309248554913295,
"grad_norm": 0.44794545147604187,
"learning_rate": 1.1403959134303832e-06,
"loss": 0.3388,
"step": 1491
},
{
"epoch": 4.312138728323699,
"grad_norm": 0.4453836249012523,
"learning_rate": 1.1310565505641114e-06,
"loss": 0.3366,
"step": 1492
},
{
"epoch": 4.315028901734104,
"grad_norm": 0.4639655297914257,
"learning_rate": 1.1217532949350075e-06,
"loss": 0.3541,
"step": 1493
},
{
"epoch": 4.317919075144509,
"grad_norm": 0.42738098566365257,
"learning_rate": 1.1124861844184809e-06,
"loss": 0.3328,
"step": 1494
},
{
"epoch": 4.320809248554913,
"grad_norm": 0.42359690830348806,
"learning_rate": 1.1032552567427911e-06,
"loss": 0.3401,
"step": 1495
},
{
"epoch": 4.323699421965318,
"grad_norm": 0.44323975529607346,
"learning_rate": 1.0940605494888856e-06,
"loss": 0.3309,
"step": 1496
},
{
"epoch": 4.326589595375722,
"grad_norm": 0.42295793743010124,
"learning_rate": 1.0849021000902427e-06,
"loss": 0.3474,
"step": 1497
},
{
"epoch": 4.329479768786127,
"grad_norm": 0.4531839136253995,
"learning_rate": 1.0757799458327445e-06,
"loss": 0.3272,
"step": 1498
},
{
"epoch": 4.332369942196531,
"grad_norm": 0.452886303528469,
"learning_rate": 1.0666941238544958e-06,
"loss": 0.3374,
"step": 1499
},
{
"epoch": 4.335260115606936,
"grad_norm": 0.4724238524003501,
"learning_rate": 1.0576446711456933e-06,
"loss": 0.3302,
"step": 1500
},
{
"epoch": 4.338150289017341,
"grad_norm": 0.42187648027212826,
"learning_rate": 1.0486316245484574e-06,
"loss": 0.3506,
"step": 1501
},
{
"epoch": 4.341040462427745,
"grad_norm": 0.4473258685192444,
"learning_rate": 1.0396550207567014e-06,
"loss": 0.3541,
"step": 1502
},
{
"epoch": 4.34393063583815,
"grad_norm": 0.4382796239703373,
"learning_rate": 1.0307148963159697e-06,
"loss": 0.341,
"step": 1503
},
{
"epoch": 4.3468208092485545,
"grad_norm": 0.46978904741617084,
"learning_rate": 1.021811287623291e-06,
"loss": 0.3073,
"step": 1504
},
{
"epoch": 4.3497109826589595,
"grad_norm": 0.4391751386270589,
"learning_rate": 1.012944230927031e-06,
"loss": 0.3378,
"step": 1505
},
{
"epoch": 4.3526011560693645,
"grad_norm": 0.43296798929341374,
"learning_rate": 1.0041137623267461e-06,
"loss": 0.3312,
"step": 1506
},
{
"epoch": 4.355491329479769,
"grad_norm": 0.44452658529893974,
"learning_rate": 9.953199177730334e-07,
"loss": 0.344,
"step": 1507
},
{
"epoch": 4.358381502890174,
"grad_norm": 0.44374319164635884,
"learning_rate": 9.86562733067389e-07,
"loss": 0.3248,
"step": 1508
},
{
"epoch": 4.361271676300578,
"grad_norm": 0.451766337725714,
"learning_rate": 9.778422438620572e-07,
"loss": 0.3158,
"step": 1509
},
{
"epoch": 4.364161849710983,
"grad_norm": 0.46955060860789766,
"learning_rate": 9.69158485659889e-07,
"loss": 0.3491,
"step": 1510
},
{
"epoch": 4.367052023121388,
"grad_norm": 0.47181813714230825,
"learning_rate": 9.605114938141935e-07,
"loss": 0.3294,
"step": 1511
},
{
"epoch": 4.369942196531792,
"grad_norm": 0.45297589475510724,
"learning_rate": 9.519013035286029e-07,
"loss": 0.3428,
"step": 1512
},
{
"epoch": 4.372832369942197,
"grad_norm": 0.42322621216412715,
"learning_rate": 9.433279498569147e-07,
"loss": 0.3656,
"step": 1513
},
{
"epoch": 4.375722543352601,
"grad_norm": 0.456838453411949,
"learning_rate": 9.347914677029624e-07,
"loss": 0.3144,
"step": 1514
},
{
"epoch": 4.378612716763006,
"grad_norm": 0.459132470959214,
"learning_rate": 9.262918918204644e-07,
"loss": 0.3401,
"step": 1515
},
{
"epoch": 4.381502890173411,
"grad_norm": 0.44484864784662287,
"learning_rate": 9.178292568128944e-07,
"loss": 0.343,
"step": 1516
},
{
"epoch": 4.384393063583815,
"grad_norm": 0.43728677906015745,
"learning_rate": 9.094035971333248e-07,
"loss": 0.3569,
"step": 1517
},
{
"epoch": 4.38728323699422,
"grad_norm": 0.45722288513195564,
"learning_rate": 9.010149470842933e-07,
"loss": 0.3682,
"step": 1518
},
{
"epoch": 4.390173410404624,
"grad_norm": 0.4468422842165502,
"learning_rate": 8.926633408176677e-07,
"loss": 0.3587,
"step": 1519
},
{
"epoch": 4.393063583815029,
"grad_norm": 0.4523020085308276,
"learning_rate": 8.843488123345045e-07,
"loss": 0.3693,
"step": 1520
},
{
"epoch": 4.395953757225434,
"grad_norm": 0.46504497106862763,
"learning_rate": 8.760713954849087e-07,
"loss": 0.348,
"step": 1521
},
{
"epoch": 4.398843930635838,
"grad_norm": 0.4572345065882788,
"learning_rate": 8.678311239678872e-07,
"loss": 0.3329,
"step": 1522
},
{
"epoch": 4.401734104046243,
"grad_norm": 0.45212729648907957,
"learning_rate": 8.596280313312355e-07,
"loss": 0.3667,
"step": 1523
},
{
"epoch": 4.404624277456647,
"grad_norm": 0.44257006814068606,
"learning_rate": 8.514621509713772e-07,
"loss": 0.3469,
"step": 1524
},
{
"epoch": 4.407514450867052,
"grad_norm": 0.4876093667366179,
"learning_rate": 8.433335161332412e-07,
"loss": 0.3238,
"step": 1525
},
{
"epoch": 4.410404624277457,
"grad_norm": 0.48545364939390045,
"learning_rate": 8.352421599101157e-07,
"loss": 0.3584,
"step": 1526
},
{
"epoch": 4.413294797687861,
"grad_norm": 0.4582224069454899,
"learning_rate": 8.27188115243529e-07,
"loss": 0.3395,
"step": 1527
},
{
"epoch": 4.416184971098266,
"grad_norm": 0.44141560547767483,
"learning_rate": 8.191714149231022e-07,
"loss": 0.3486,
"step": 1528
},
{
"epoch": 4.41907514450867,
"grad_norm": 0.4337129493212092,
"learning_rate": 8.111920915864213e-07,
"loss": 0.3648,
"step": 1529
},
{
"epoch": 4.421965317919075,
"grad_norm": 0.4485339769994593,
"learning_rate": 8.032501777189016e-07,
"loss": 0.3587,
"step": 1530
},
{
"epoch": 4.424855491329479,
"grad_norm": 0.43414525391667297,
"learning_rate": 7.953457056536596e-07,
"loss": 0.3399,
"step": 1531
},
{
"epoch": 4.427745664739884,
"grad_norm": 0.43376869094234677,
"learning_rate": 7.874787075713742e-07,
"loss": 0.3462,
"step": 1532
},
{
"epoch": 4.430635838150289,
"grad_norm": 0.45750126312142625,
"learning_rate": 7.796492155001656e-07,
"loss": 0.3271,
"step": 1533
},
{
"epoch": 4.433526011560693,
"grad_norm": 0.46580876925404385,
"learning_rate": 7.718572613154574e-07,
"loss": 0.3321,
"step": 1534
},
{
"epoch": 4.436416184971098,
"grad_norm": 0.4446063454528743,
"learning_rate": 7.641028767398473e-07,
"loss": 0.3328,
"step": 1535
},
{
"epoch": 4.4393063583815024,
"grad_norm": 0.47054539897600545,
"learning_rate": 7.563860933429789e-07,
"loss": 0.3315,
"step": 1536
},
{
"epoch": 4.442196531791907,
"grad_norm": 0.4673864472062917,
"learning_rate": 7.48706942541424e-07,
"loss": 0.3197,
"step": 1537
},
{
"epoch": 4.445086705202312,
"grad_norm": 0.44842080291220054,
"learning_rate": 7.410654555985287e-07,
"loss": 0.3515,
"step": 1538
},
{
"epoch": 4.4479768786127165,
"grad_norm": 0.41012475610075394,
"learning_rate": 7.33461663624313e-07,
"loss": 0.3612,
"step": 1539
},
{
"epoch": 4.4508670520231215,
"grad_norm": 0.45327829164620953,
"learning_rate": 7.258955975753279e-07,
"loss": 0.3518,
"step": 1540
},
{
"epoch": 4.453757225433526,
"grad_norm": 0.45932612475028833,
"learning_rate": 7.183672882545401e-07,
"loss": 0.3439,
"step": 1541
},
{
"epoch": 4.456647398843931,
"grad_norm": 0.4859816695971326,
"learning_rate": 7.108767663111993e-07,
"loss": 0.3407,
"step": 1542
},
{
"epoch": 4.459537572254336,
"grad_norm": 0.4323085447994151,
"learning_rate": 7.034240622407085e-07,
"loss": 0.3493,
"step": 1543
},
{
"epoch": 4.46242774566474,
"grad_norm": 0.4284267539030077,
"learning_rate": 6.960092063845148e-07,
"loss": 0.3516,
"step": 1544
},
{
"epoch": 4.465317919075145,
"grad_norm": 0.4418025417466207,
"learning_rate": 6.886322289299763e-07,
"loss": 0.3059,
"step": 1545
},
{
"epoch": 4.468208092485549,
"grad_norm": 0.429128506405581,
"learning_rate": 6.812931599102379e-07,
"loss": 0.3352,
"step": 1546
},
{
"epoch": 4.471098265895954,
"grad_norm": 0.43750017936727364,
"learning_rate": 6.739920292041102e-07,
"loss": 0.3456,
"step": 1547
},
{
"epoch": 4.473988439306359,
"grad_norm": 0.44953049761535996,
"learning_rate": 6.667288665359562e-07,
"loss": 0.3363,
"step": 1548
},
{
"epoch": 4.476878612716763,
"grad_norm": 0.4489565986882684,
"learning_rate": 6.595037014755578e-07,
"loss": 0.3266,
"step": 1549
},
{
"epoch": 4.479768786127168,
"grad_norm": 0.4309164502180863,
"learning_rate": 6.523165634380047e-07,
"loss": 0.3472,
"step": 1550
},
{
"epoch": 4.482658959537572,
"grad_norm": 0.42593312051489646,
"learning_rate": 6.451674816835618e-07,
"loss": 0.3605,
"step": 1551
},
{
"epoch": 4.485549132947977,
"grad_norm": 0.425417146538673,
"learning_rate": 6.380564853175741e-07,
"loss": 0.3358,
"step": 1552
},
{
"epoch": 4.488439306358382,
"grad_norm": 0.46193990483385394,
"learning_rate": 6.309836032903228e-07,
"loss": 0.2985,
"step": 1553
},
{
"epoch": 4.491329479768786,
"grad_norm": 0.43253506570848216,
"learning_rate": 6.239488643969205e-07,
"loss": 0.3226,
"step": 1554
},
{
"epoch": 4.494219653179191,
"grad_norm": 0.4433378890746248,
"learning_rate": 6.169522972771924e-07,
"loss": 0.3273,
"step": 1555
},
{
"epoch": 4.497109826589595,
"grad_norm": 0.42184643595929217,
"learning_rate": 6.099939304155577e-07,
"loss": 0.3202,
"step": 1556
},
{
"epoch": 4.5,
"grad_norm": 0.47640273912437076,
"learning_rate": 6.030737921409169e-07,
"loss": 0.2722,
"step": 1557
},
{
"epoch": 4.502890173410405,
"grad_norm": 0.45407141264243817,
"learning_rate": 5.961919106265313e-07,
"loss": 0.3439,
"step": 1558
},
{
"epoch": 4.505780346820809,
"grad_norm": 0.4376449567849091,
"learning_rate": 5.893483138899125e-07,
"loss": 0.3513,
"step": 1559
},
{
"epoch": 4.508670520231214,
"grad_norm": 0.46675708746409345,
"learning_rate": 5.825430297927093e-07,
"loss": 0.3303,
"step": 1560
},
{
"epoch": 4.511560693641618,
"grad_norm": 0.44146368613515313,
"learning_rate": 5.757760860405859e-07,
"loss": 0.332,
"step": 1561
},
{
"epoch": 4.514450867052023,
"grad_norm": 0.418864555410123,
"learning_rate": 5.690475101831261e-07,
"loss": 0.3472,
"step": 1562
},
{
"epoch": 4.517341040462428,
"grad_norm": 0.4586426558580989,
"learning_rate": 5.623573296136997e-07,
"loss": 0.3382,
"step": 1563
},
{
"epoch": 4.520231213872832,
"grad_norm": 0.42898089728523714,
"learning_rate": 5.557055715693649e-07,
"loss": 0.3324,
"step": 1564
},
{
"epoch": 4.523121387283237,
"grad_norm": 0.4676061900308173,
"learning_rate": 5.49092263130756e-07,
"loss": 0.3348,
"step": 1565
},
{
"epoch": 4.526011560693641,
"grad_norm": 0.4543700925597362,
"learning_rate": 5.425174312219739e-07,
"loss": 0.3113,
"step": 1566
},
{
"epoch": 4.528901734104046,
"grad_norm": 0.465108477478383,
"learning_rate": 5.359811026104645e-07,
"loss": 0.3391,
"step": 1567
},
{
"epoch": 4.531791907514451,
"grad_norm": 0.4808284435115195,
"learning_rate": 5.294833039069269e-07,
"loss": 0.3346,
"step": 1568
},
{
"epoch": 4.534682080924855,
"grad_norm": 0.44287503653196786,
"learning_rate": 5.230240615651972e-07,
"loss": 0.3357,
"step": 1569
},
{
"epoch": 4.53757225433526,
"grad_norm": 0.43946866441137844,
"learning_rate": 5.166034018821364e-07,
"loss": 0.3283,
"step": 1570
},
{
"epoch": 4.540462427745664,
"grad_norm": 0.436053347850471,
"learning_rate": 5.10221350997534e-07,
"loss": 0.3538,
"step": 1571
},
{
"epoch": 4.543352601156069,
"grad_norm": 0.4410619890844322,
"learning_rate": 5.03877934893986e-07,
"loss": 0.3456,
"step": 1572
},
{
"epoch": 4.546242774566474,
"grad_norm": 0.4711814603971082,
"learning_rate": 4.975731793968075e-07,
"loss": 0.3352,
"step": 1573
},
{
"epoch": 4.5491329479768785,
"grad_norm": 0.4545605967025349,
"learning_rate": 4.913071101739164e-07,
"loss": 0.3403,
"step": 1574
},
{
"epoch": 4.5520231213872835,
"grad_norm": 0.43235496189688866,
"learning_rate": 4.850797527357287e-07,
"loss": 0.3349,
"step": 1575
},
{
"epoch": 4.554913294797688,
"grad_norm": 0.47022676916855527,
"learning_rate": 4.788911324350564e-07,
"loss": 0.3256,
"step": 1576
},
{
"epoch": 4.557803468208093,
"grad_norm": 0.4452792759670324,
"learning_rate": 4.72741274467009e-07,
"loss": 0.3511,
"step": 1577
},
{
"epoch": 4.5606936416184976,
"grad_norm": 0.44372850033678257,
"learning_rate": 4.6663020386888416e-07,
"loss": 0.3181,
"step": 1578
},
{
"epoch": 4.563583815028902,
"grad_norm": 0.4398620339745168,
"learning_rate": 4.6055794552006817e-07,
"loss": 0.3326,
"step": 1579
},
{
"epoch": 4.566473988439307,
"grad_norm": 0.4662378155864558,
"learning_rate": 4.5452452414193495e-07,
"loss": 0.3237,
"step": 1580
},
{
"epoch": 4.569364161849711,
"grad_norm": 0.42225338968814385,
"learning_rate": 4.485299642977481e-07,
"loss": 0.3468,
"step": 1581
},
{
"epoch": 4.572254335260116,
"grad_norm": 0.44278019638014077,
"learning_rate": 4.425742903925534e-07,
"loss": 0.3204,
"step": 1582
},
{
"epoch": 4.575144508670521,
"grad_norm": 0.5016084179745222,
"learning_rate": 4.366575266730888e-07,
"loss": 0.3332,
"step": 1583
},
{
"epoch": 4.578034682080925,
"grad_norm": 0.469191932516797,
"learning_rate": 4.3077969722767897e-07,
"loss": 0.3471,
"step": 1584
},
{
"epoch": 4.58092485549133,
"grad_norm": 0.42826255111061495,
"learning_rate": 4.2494082598613875e-07,
"loss": 0.3682,
"step": 1585
},
{
"epoch": 4.583815028901734,
"grad_norm": 0.4597551582701016,
"learning_rate": 4.191409367196753e-07,
"loss": 0.3388,
"step": 1586
},
{
"epoch": 4.586705202312139,
"grad_norm": 0.4373614912267801,
"learning_rate": 4.133800530407994e-07,
"loss": 0.3203,
"step": 1587
},
{
"epoch": 4.589595375722544,
"grad_norm": 0.43968666610635687,
"learning_rate": 4.0765819840321353e-07,
"loss": 0.3402,
"step": 1588
},
{
"epoch": 4.592485549132948,
"grad_norm": 0.44351171734497535,
"learning_rate": 4.019753961017292e-07,
"loss": 0.3357,
"step": 1589
},
{
"epoch": 4.595375722543353,
"grad_norm": 0.4415848556722189,
"learning_rate": 3.9633166927216637e-07,
"loss": 0.3291,
"step": 1590
},
{
"epoch": 4.598265895953757,
"grad_norm": 0.4652210434796125,
"learning_rate": 3.9072704089126666e-07,
"loss": 0.3418,
"step": 1591
},
{
"epoch": 4.601156069364162,
"grad_norm": 0.4543224901650214,
"learning_rate": 3.8516153377658574e-07,
"loss": 0.3356,
"step": 1592
},
{
"epoch": 4.604046242774566,
"grad_norm": 0.4553390702328126,
"learning_rate": 3.7963517058641206e-07,
"loss": 0.3335,
"step": 1593
},
{
"epoch": 4.606936416184971,
"grad_norm": 0.4352888331046387,
"learning_rate": 3.741479738196763e-07,
"loss": 0.3502,
"step": 1594
},
{
"epoch": 4.609826589595376,
"grad_norm": 0.4374869535509039,
"learning_rate": 3.6869996581584746e-07,
"loss": 0.3477,
"step": 1595
},
{
"epoch": 4.61271676300578,
"grad_norm": 0.4348034826002002,
"learning_rate": 3.632911687548546e-07,
"loss": 0.3609,
"step": 1596
},
{
"epoch": 4.615606936416185,
"grad_norm": 0.448678536253477,
"learning_rate": 3.5792160465698555e-07,
"loss": 0.3503,
"step": 1597
},
{
"epoch": 4.618497109826589,
"grad_norm": 0.4574142336139396,
"learning_rate": 3.5259129538281034e-07,
"loss": 0.3355,
"step": 1598
},
{
"epoch": 4.621387283236994,
"grad_norm": 0.45364396763072645,
"learning_rate": 3.473002626330779e-07,
"loss": 0.3321,
"step": 1599
},
{
"epoch": 4.624277456647399,
"grad_norm": 0.4506428297942708,
"learning_rate": 3.4204852794863855e-07,
"loss": 0.3357,
"step": 1600
},
{
"epoch": 4.627167630057803,
"grad_norm": 0.4041242550173493,
"learning_rate": 3.3683611271035256e-07,
"loss": 0.3563,
"step": 1601
},
{
"epoch": 4.630057803468208,
"grad_norm": 0.44434168162177407,
"learning_rate": 3.316630381389996e-07,
"loss": 0.3306,
"step": 1602
},
{
"epoch": 4.632947976878612,
"grad_norm": 0.47932184626470214,
"learning_rate": 3.2652932529519843e-07,
"loss": 0.3447,
"step": 1603
},
{
"epoch": 4.635838150289017,
"grad_norm": 0.46908958631347786,
"learning_rate": 3.214349950793183e-07,
"loss": 0.3006,
"step": 1604
},
{
"epoch": 4.638728323699422,
"grad_norm": 0.4694171525413062,
"learning_rate": 3.163800682313933e-07,
"loss": 0.3245,
"step": 1605
},
{
"epoch": 4.641618497109826,
"grad_norm": 0.41468178348710893,
"learning_rate": 3.113645653310382e-07,
"loss": 0.3303,
"step": 1606
},
{
"epoch": 4.644508670520231,
"grad_norm": 0.4462666197450191,
"learning_rate": 3.0638850679736485e-07,
"loss": 0.3413,
"step": 1607
},
{
"epoch": 4.6473988439306355,
"grad_norm": 0.4469140886647421,
"learning_rate": 3.014519128888993e-07,
"loss": 0.3368,
"step": 1608
},
{
"epoch": 4.6502890173410405,
"grad_norm": 0.4272319628195244,
"learning_rate": 2.965548037035015e-07,
"loss": 0.3462,
"step": 1609
},
{
"epoch": 4.653179190751445,
"grad_norm": 0.4643364991741565,
"learning_rate": 2.91697199178278e-07,
"loss": 0.3222,
"step": 1610
},
{
"epoch": 4.65606936416185,
"grad_norm": 0.457141530509153,
"learning_rate": 2.868791190895048e-07,
"loss": 0.3429,
"step": 1611
},
{
"epoch": 4.658959537572255,
"grad_norm": 0.4700258494548005,
"learning_rate": 2.8210058305255026e-07,
"loss": 0.3539,
"step": 1612
},
{
"epoch": 4.661849710982659,
"grad_norm": 0.45216546599088037,
"learning_rate": 2.773616105217836e-07,
"loss": 0.3421,
"step": 1613
},
{
"epoch": 4.664739884393064,
"grad_norm": 0.44191986169167397,
"learning_rate": 2.7266222079050717e-07,
"loss": 0.3241,
"step": 1614
},
{
"epoch": 4.667630057803468,
"grad_norm": 0.4334486391581851,
"learning_rate": 2.680024329908737e-07,
"loss": 0.3407,
"step": 1615
},
{
"epoch": 4.670520231213873,
"grad_norm": 0.46200158841454086,
"learning_rate": 2.633822660938112e-07,
"loss": 0.3357,
"step": 1616
},
{
"epoch": 4.673410404624278,
"grad_norm": 0.4500995146006072,
"learning_rate": 2.588017389089381e-07,
"loss": 0.3321,
"step": 1617
},
{
"epoch": 4.676300578034682,
"grad_norm": 0.4541876298171543,
"learning_rate": 2.5426087008449173e-07,
"loss": 0.3335,
"step": 1618
},
{
"epoch": 4.679190751445087,
"grad_norm": 0.42614932613213036,
"learning_rate": 2.4975967810725865e-07,
"loss": 0.3252,
"step": 1619
},
{
"epoch": 4.682080924855491,
"grad_norm": 0.453308198867481,
"learning_rate": 2.452981813024868e-07,
"loss": 0.3569,
"step": 1620
},
{
"epoch": 4.684971098265896,
"grad_norm": 0.46066755061318243,
"learning_rate": 2.4087639783382133e-07,
"loss": 0.3318,
"step": 1621
},
{
"epoch": 4.687861271676301,
"grad_norm": 0.4253730592911683,
"learning_rate": 2.3649434570321984e-07,
"loss": 0.3318,
"step": 1622
},
{
"epoch": 4.690751445086705,
"grad_norm": 0.4170677149513208,
"learning_rate": 2.3215204275089388e-07,
"loss": 0.3237,
"step": 1623
},
{
"epoch": 4.69364161849711,
"grad_norm": 0.43218581173346604,
"learning_rate": 2.2784950665522443e-07,
"loss": 0.3507,
"step": 1624
},
{
"epoch": 4.696531791907514,
"grad_norm": 0.42990916259901707,
"learning_rate": 2.235867549326931e-07,
"loss": 0.3425,
"step": 1625
},
{
"epoch": 4.699421965317919,
"grad_norm": 0.43695936554688053,
"learning_rate": 2.1936380493781218e-07,
"loss": 0.3426,
"step": 1626
},
{
"epoch": 4.702312138728324,
"grad_norm": 0.46400517481162973,
"learning_rate": 2.151806738630524e-07,
"loss": 0.3274,
"step": 1627
},
{
"epoch": 4.705202312138728,
"grad_norm": 0.4490940433053954,
"learning_rate": 2.110373787387754e-07,
"loss": 0.335,
"step": 1628
},
{
"epoch": 4.708092485549133,
"grad_norm": 0.4252766476863728,
"learning_rate": 2.069339364331624e-07,
"loss": 0.3484,
"step": 1629
},
{
"epoch": 4.710982658959537,
"grad_norm": 0.4559621883794528,
"learning_rate": 2.028703636521434e-07,
"loss": 0.3285,
"step": 1630
},
{
"epoch": 4.713872832369942,
"grad_norm": 0.47333283731170095,
"learning_rate": 1.988466769393349e-07,
"loss": 0.3169,
"step": 1631
},
{
"epoch": 4.716763005780347,
"grad_norm": 0.44926712131599644,
"learning_rate": 1.948628926759666e-07,
"loss": 0.3577,
"step": 1632
},
{
"epoch": 4.719653179190751,
"grad_norm": 0.4406937465143532,
"learning_rate": 1.909190270808192e-07,
"loss": 0.3519,
"step": 1633
},
{
"epoch": 4.722543352601156,
"grad_norm": 0.43131627374523984,
"learning_rate": 1.870150962101569e-07,
"loss": 0.3606,
"step": 1634
},
{
"epoch": 4.72543352601156,
"grad_norm": 0.4328772609998375,
"learning_rate": 1.8315111595765933e-07,
"loss": 0.3271,
"step": 1635
},
{
"epoch": 4.728323699421965,
"grad_norm": 0.4351914089691055,
"learning_rate": 1.7932710205435966e-07,
"loss": 0.3429,
"step": 1636
},
{
"epoch": 4.73121387283237,
"grad_norm": 0.4500492231878512,
"learning_rate": 1.755430700685845e-07,
"loss": 0.3483,
"step": 1637
},
{
"epoch": 4.734104046242774,
"grad_norm": 0.4351826250212104,
"learning_rate": 1.7179903540587962e-07,
"loss": 0.3311,
"step": 1638
},
{
"epoch": 4.736994219653179,
"grad_norm": 0.4476955647395091,
"learning_rate": 1.680950133089565e-07,
"loss": 0.3359,
"step": 1639
},
{
"epoch": 4.7398843930635834,
"grad_norm": 0.4395449881892697,
"learning_rate": 1.6443101885762812e-07,
"loss": 0.3207,
"step": 1640
},
{
"epoch": 4.742774566473988,
"grad_norm": 0.425174851579623,
"learning_rate": 1.6080706696874893e-07,
"loss": 0.3344,
"step": 1641
},
{
"epoch": 4.745664739884393,
"grad_norm": 0.47236662938582863,
"learning_rate": 1.5722317239614592e-07,
"loss": 0.346,
"step": 1642
},
{
"epoch": 4.7485549132947975,
"grad_norm": 0.421518858724491,
"learning_rate": 1.5367934973056997e-07,
"loss": 0.3268,
"step": 1643
},
{
"epoch": 4.7514450867052025,
"grad_norm": 0.4184742989579567,
"learning_rate": 1.5017561339963015e-07,
"loss": 0.3537,
"step": 1644
},
{
"epoch": 4.754335260115607,
"grad_norm": 0.428901797065803,
"learning_rate": 1.4671197766773616e-07,
"loss": 0.3217,
"step": 1645
},
{
"epoch": 4.757225433526012,
"grad_norm": 0.3907897216934688,
"learning_rate": 1.432884566360393e-07,
"loss": 0.3367,
"step": 1646
},
{
"epoch": 4.7601156069364166,
"grad_norm": 0.4292623463312309,
"learning_rate": 1.3990506424237382e-07,
"loss": 0.3533,
"step": 1647
},
{
"epoch": 4.763005780346821,
"grad_norm": 0.44620500922357464,
"learning_rate": 1.3656181426120907e-07,
"loss": 0.3254,
"step": 1648
},
{
"epoch": 4.765895953757226,
"grad_norm": 0.45625749916755926,
"learning_rate": 1.3325872030357955e-07,
"loss": 0.3317,
"step": 1649
},
{
"epoch": 4.76878612716763,
"grad_norm": 0.47956732191703283,
"learning_rate": 1.2999579581703948e-07,
"loss": 0.3295,
"step": 1650
},
{
"epoch": 4.771676300578035,
"grad_norm": 0.4548429868833743,
"learning_rate": 1.2677305408560602e-07,
"loss": 0.3312,
"step": 1651
},
{
"epoch": 4.77456647398844,
"grad_norm": 0.43813428920113245,
"learning_rate": 1.2359050822970287e-07,
"loss": 0.348,
"step": 1652
},
{
"epoch": 4.777456647398844,
"grad_norm": 0.4354750741426635,
"learning_rate": 1.2044817120610896e-07,
"loss": 0.3516,
"step": 1653
},
{
"epoch": 4.780346820809249,
"grad_norm": 0.4304180249939501,
"learning_rate": 1.1734605580790426e-07,
"loss": 0.3552,
"step": 1654
},
{
"epoch": 4.783236994219653,
"grad_norm": 0.46194955068474436,
"learning_rate": 1.1428417466442077e-07,
"loss": 0.3351,
"step": 1655
},
{
"epoch": 4.786127167630058,
"grad_norm": 0.44976136379540815,
"learning_rate": 1.1126254024118488e-07,
"loss": 0.3256,
"step": 1656
},
{
"epoch": 4.789017341040463,
"grad_norm": 0.45376970823270474,
"learning_rate": 1.0828116483987405e-07,
"loss": 0.3103,
"step": 1657
},
{
"epoch": 4.791907514450867,
"grad_norm": 0.4252964208492344,
"learning_rate": 1.053400605982613e-07,
"loss": 0.3421,
"step": 1658
},
{
"epoch": 4.794797687861272,
"grad_norm": 0.4575525290978499,
"learning_rate": 1.0243923949016966e-07,
"loss": 0.3318,
"step": 1659
},
{
"epoch": 4.797687861271676,
"grad_norm": 0.4315729396314116,
"learning_rate": 9.957871332541891e-08,
"loss": 0.3333,
"step": 1660
},
{
"epoch": 4.800578034682081,
"grad_norm": 0.4632976454740385,
"learning_rate": 9.67584937497823e-08,
"loss": 0.3298,
"step": 1661
},
{
"epoch": 4.803468208092486,
"grad_norm": 0.4499425222875539,
"learning_rate": 9.397859224493656e-08,
"loss": 0.3576,
"step": 1662
},
{
"epoch": 4.80635838150289,
"grad_norm": 0.4396266921318992,
"learning_rate": 9.123902012841301e-08,
"loss": 0.3612,
"step": 1663
},
{
"epoch": 4.809248554913295,
"grad_norm": 0.44847121001326007,
"learning_rate": 8.853978855355883e-08,
"loss": 0.3393,
"step": 1664
},
{
"epoch": 4.812138728323699,
"grad_norm": 0.4644459108259271,
"learning_rate": 8.588090850948028e-08,
"loss": 0.3246,
"step": 1665
},
{
"epoch": 4.815028901734104,
"grad_norm": 0.4268024970743814,
"learning_rate": 8.326239082101173e-08,
"loss": 0.3414,
"step": 1666
},
{
"epoch": 4.817919075144509,
"grad_norm": 0.43070189303356615,
"learning_rate": 8.0684246148659e-08,
"loss": 0.3478,
"step": 1667
},
{
"epoch": 4.820809248554913,
"grad_norm": 0.421476341135529,
"learning_rate": 7.81464849885627e-08,
"loss": 0.357,
"step": 1668
},
{
"epoch": 4.823699421965318,
"grad_norm": 0.43518704104148476,
"learning_rate": 7.564911767245609e-08,
"loss": 0.345,
"step": 1669
},
{
"epoch": 4.826589595375722,
"grad_norm": 0.43385274927828715,
"learning_rate": 7.31921543676184e-08,
"loss": 0.3447,
"step": 1670
},
{
"epoch": 4.829479768786127,
"grad_norm": 0.45581456572701706,
"learning_rate": 7.077560507683712e-08,
"loss": 0.3302,
"step": 1671
},
{
"epoch": 4.832369942196532,
"grad_norm": 0.4283824645917809,
"learning_rate": 6.83994796383669e-08,
"loss": 0.3684,
"step": 1672
},
{
"epoch": 4.835260115606936,
"grad_norm": 0.43036551648477195,
"learning_rate": 6.60637877258874e-08,
"loss": 0.3647,
"step": 1673
},
{
"epoch": 4.838150289017341,
"grad_norm": 0.42690403732855864,
"learning_rate": 6.376853884846656e-08,
"loss": 0.3455,
"step": 1674
},
{
"epoch": 4.841040462427745,
"grad_norm": 0.4219758733193144,
"learning_rate": 6.151374235051966e-08,
"loss": 0.3672,
"step": 1675
},
{
"epoch": 4.84393063583815,
"grad_norm": 0.4379144455753565,
"learning_rate": 5.929940741177476e-08,
"loss": 0.3383,
"step": 1676
},
{
"epoch": 4.846820809248555,
"grad_norm": 0.4434027052467159,
"learning_rate": 5.7125543047228395e-08,
"loss": 0.321,
"step": 1677
},
{
"epoch": 4.8497109826589595,
"grad_norm": 0.44390976754330297,
"learning_rate": 5.4992158107116664e-08,
"loss": 0.3346,
"step": 1678
},
{
"epoch": 4.8526011560693645,
"grad_norm": 0.4330405056189658,
"learning_rate": 5.289926127687639e-08,
"loss": 0.3314,
"step": 1679
},
{
"epoch": 4.855491329479769,
"grad_norm": 0.45680791949365546,
"learning_rate": 5.084686107710512e-08,
"loss": 0.3207,
"step": 1680
},
{
"epoch": 4.858381502890174,
"grad_norm": 0.4448930539952292,
"learning_rate": 4.8834965863536755e-08,
"loss": 0.3454,
"step": 1681
},
{
"epoch": 4.861271676300578,
"grad_norm": 0.42026937097621103,
"learning_rate": 4.686358382699485e-08,
"loss": 0.3387,
"step": 1682
},
{
"epoch": 4.864161849710983,
"grad_norm": 0.4507339735950156,
"learning_rate": 4.493272299337048e-08,
"loss": 0.3391,
"step": 1683
},
{
"epoch": 4.867052023121388,
"grad_norm": 0.45447507822713396,
"learning_rate": 4.3042391223582226e-08,
"loss": 0.3041,
"step": 1684
},
{
"epoch": 4.869942196531792,
"grad_norm": 0.4432495734282729,
"learning_rate": 4.119259621354843e-08,
"loss": 0.3549,
"step": 1685
},
{
"epoch": 4.872832369942197,
"grad_norm": 0.4034187281207412,
"learning_rate": 3.938334549415168e-08,
"loss": 0.3489,
"step": 1686
},
{
"epoch": 4.875722543352601,
"grad_norm": 0.44807856188638867,
"learning_rate": 3.761464643121548e-08,
"loss": 0.3437,
"step": 1687
},
{
"epoch": 4.878612716763006,
"grad_norm": 0.42048520977505255,
"learning_rate": 3.5886506225463194e-08,
"loss": 0.3457,
"step": 1688
},
{
"epoch": 4.881502890173411,
"grad_norm": 0.43027452694512436,
"learning_rate": 3.419893191250023e-08,
"loss": 0.3163,
"step": 1689
},
{
"epoch": 4.884393063583815,
"grad_norm": 0.4308993314733086,
"learning_rate": 3.255193036277637e-08,
"loss": 0.3437,
"step": 1690
},
{
"epoch": 4.88728323699422,
"grad_norm": 0.431450612571325,
"learning_rate": 3.094550828156573e-08,
"loss": 0.3375,
"step": 1691
},
{
"epoch": 4.890173410404624,
"grad_norm": 0.4431635701323724,
"learning_rate": 2.937967220893123e-08,
"loss": 0.3318,
"step": 1692
},
{
"epoch": 4.893063583815029,
"grad_norm": 0.4330022821758298,
"learning_rate": 2.7854428519703546e-08,
"loss": 0.3431,
"step": 1693
},
{
"epoch": 4.895953757225434,
"grad_norm": 0.42453430258622377,
"learning_rate": 2.636978342345553e-08,
"loss": 0.3743,
"step": 1694
},
{
"epoch": 4.898843930635838,
"grad_norm": 0.44637278693106885,
"learning_rate": 2.4925742964471145e-08,
"loss": 0.3201,
"step": 1695
},
{
"epoch": 4.901734104046243,
"grad_norm": 0.43885017321147307,
"learning_rate": 2.3522313021728805e-08,
"loss": 0.3423,
"step": 1696
},
{
"epoch": 4.904624277456647,
"grad_norm": 0.43250678490930095,
"learning_rate": 2.215949930887029e-08,
"loss": 0.3514,
"step": 1697
},
{
"epoch": 4.907514450867052,
"grad_norm": 0.4196196401753552,
"learning_rate": 2.083730737418299e-08,
"loss": 0.3452,
"step": 1698
},
{
"epoch": 4.910404624277456,
"grad_norm": 0.4542871643674,
"learning_rate": 1.9555742600573247e-08,
"loss": 0.3408,
"step": 1699
},
{
"epoch": 4.913294797687861,
"grad_norm": 0.43417937342131485,
"learning_rate": 1.8314810205547485e-08,
"loss": 0.3175,
"step": 1700
},
{
"epoch": 4.916184971098266,
"grad_norm": 0.42866176133661305,
"learning_rate": 1.7114515241188902e-08,
"loss": 0.3496,
"step": 1701
},
{
"epoch": 4.91907514450867,
"grad_norm": 0.4331937602055982,
"learning_rate": 1.5954862594136367e-08,
"loss": 0.3539,
"step": 1702
},
{
"epoch": 4.921965317919075,
"grad_norm": 0.4546919927327096,
"learning_rate": 1.4835856985568887e-08,
"loss": 0.3437,
"step": 1703
},
{
"epoch": 4.924855491329479,
"grad_norm": 0.44616273282584096,
"learning_rate": 1.3757502971178948e-08,
"loss": 0.3352,
"step": 1704
},
{
"epoch": 4.927745664739884,
"grad_norm": 0.43042187879856225,
"learning_rate": 1.2719804941163649e-08,
"loss": 0.3277,
"step": 1705
},
{
"epoch": 4.930635838150289,
"grad_norm": 0.42419151713724373,
"learning_rate": 1.1722767120196932e-08,
"loss": 0.3515,
"step": 1706
},
{
"epoch": 4.933526011560693,
"grad_norm": 0.4330173444705608,
"learning_rate": 1.0766393567418487e-08,
"loss": 0.365,
"step": 1707
},
{
"epoch": 4.936416184971098,
"grad_norm": 0.4393369170187094,
"learning_rate": 9.85068817641599e-09,
"loss": 0.3393,
"step": 1708
},
{
"epoch": 4.9393063583815024,
"grad_norm": 0.4237121649080187,
"learning_rate": 8.975654675208445e-09,
"loss": 0.3511,
"step": 1709
},
{
"epoch": 4.942196531791907,
"grad_norm": 0.45994595122972776,
"learning_rate": 8.141296626231754e-09,
"loss": 0.3504,
"step": 1710
},
{
"epoch": 4.945086705202312,
"grad_norm": 0.43930885084169213,
"learning_rate": 7.347617426325393e-09,
"loss": 0.3481,
"step": 1711
},
{
"epoch": 4.9479768786127165,
"grad_norm": 0.4272366397651899,
"learning_rate": 6.5946203067135395e-09,
"loss": 0.3327,
"step": 1712
},
{
"epoch": 4.9508670520231215,
"grad_norm": 0.4559639493418854,
"learning_rate": 5.88230833299841e-09,
"loss": 0.3435,
"step": 1713
},
{
"epoch": 4.953757225433526,
"grad_norm": 0.47676301343666305,
"learning_rate": 5.210684405144717e-09,
"loss": 0.3166,
"step": 1714
},
{
"epoch": 4.956647398843931,
"grad_norm": 0.4231619739273129,
"learning_rate": 4.579751257466347e-09,
"loss": 0.3723,
"step": 1715
},
{
"epoch": 4.959537572254336,
"grad_norm": 0.45086590601295234,
"learning_rate": 3.989511458618589e-09,
"loss": 0.3431,
"step": 1716
},
{
"epoch": 4.96242774566474,
"grad_norm": 0.43706998519645374,
"learning_rate": 3.43996741158481e-09,
"loss": 0.3149,
"step": 1717
},
{
"epoch": 4.965317919075145,
"grad_norm": 0.42743087853849465,
"learning_rate": 2.9311213536686868e-09,
"loss": 0.3576,
"step": 1718
},
{
"epoch": 4.968208092485549,
"grad_norm": 0.4188172942844786,
"learning_rate": 2.4629753564842095e-09,
"loss": 0.3515,
"step": 1719
},
{
"epoch": 4.971098265895954,
"grad_norm": 0.4425146472622218,
"learning_rate": 2.0355313259468045e-09,
"loss": 0.336,
"step": 1720
},
{
"epoch": 4.973988439306359,
"grad_norm": 0.44582007168071364,
"learning_rate": 1.6487910022666698e-09,
"loss": 0.3322,
"step": 1721
},
{
"epoch": 4.976878612716763,
"grad_norm": 0.43565612155479977,
"learning_rate": 1.3027559599410044e-09,
"loss": 0.3699,
"step": 1722
},
{
"epoch": 4.979768786127168,
"grad_norm": 0.42224175219522175,
"learning_rate": 9.974276077462375e-10,
"loss": 0.3458,
"step": 1723
},
{
"epoch": 4.982658959537572,
"grad_norm": 0.43381161082142367,
"learning_rate": 7.328071887358068e-10,
"loss": 0.3569,
"step": 1724
},
{
"epoch": 4.985549132947977,
"grad_norm": 0.4595886172323835,
"learning_rate": 5.088957802323879e-10,
"loss": 0.3033,
"step": 1725
},
{
"epoch": 4.988439306358382,
"grad_norm": 0.45716642590282125,
"learning_rate": 3.256942938234531e-10,
"loss": 0.3148,
"step": 1726
},
{
"epoch": 4.991329479768786,
"grad_norm": 0.4488664794487837,
"learning_rate": 1.8320347536016082e-10,
"loss": 0.3591,
"step": 1727
},
{
"epoch": 4.994219653179191,
"grad_norm": 0.40480634643061625,
"learning_rate": 8.14239049484744e-11,
"loss": 0.3753,
"step": 1728
},
{
"epoch": 4.997109826589595,
"grad_norm": 0.4362498372754465,
"learning_rate": 2.035599695582313e-11,
"loss": 0.3142,
"step": 1729
},
{
"epoch": 5.0,
"grad_norm": 0.43942713509202896,
"learning_rate": 0.0,
"loss": 0.3327,
"step": 1730
},
{
"epoch": 5.0,
"step": 1730,
"total_flos": 5.076921768434729e+18,
"train_loss": 0.5051262063442628,
"train_runtime": 40991.9822,
"train_samples_per_second": 5.397,
"train_steps_per_second": 0.042
}
],
"logging_steps": 1.0,
"max_steps": 1730,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 5.076921768434729e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}