cutelemonlili's picture
Add files using upload-large-folder tool
ccc5ab8 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 200,
"global_step": 668,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0029940119760479044,
"grad_norm": 2.8150834679093633,
"learning_rate": 9.999944704978835e-06,
"loss": 0.4278,
"step": 1
},
{
"epoch": 0.005988023952095809,
"grad_norm": 2.693132917479612,
"learning_rate": 9.999778821138357e-06,
"loss": 0.38,
"step": 2
},
{
"epoch": 0.008982035928143712,
"grad_norm": 2.3983437843858186,
"learning_rate": 9.999502352147583e-06,
"loss": 0.4083,
"step": 3
},
{
"epoch": 0.011976047904191617,
"grad_norm": 2.575839803104308,
"learning_rate": 9.999115304121459e-06,
"loss": 0.3763,
"step": 4
},
{
"epoch": 0.014970059880239521,
"grad_norm": 2.294929768951224,
"learning_rate": 9.998617685620715e-06,
"loss": 0.3286,
"step": 5
},
{
"epoch": 0.017964071856287425,
"grad_norm": 2.460127426054556,
"learning_rate": 9.998009507651683e-06,
"loss": 0.3068,
"step": 6
},
{
"epoch": 0.020958083832335328,
"grad_norm": 2.060737765218325,
"learning_rate": 9.997290783666048e-06,
"loss": 0.3299,
"step": 7
},
{
"epoch": 0.023952095808383235,
"grad_norm": 1.9339918722553782,
"learning_rate": 9.996461529560553e-06,
"loss": 0.3036,
"step": 8
},
{
"epoch": 0.02694610778443114,
"grad_norm": 1.9152499116881965,
"learning_rate": 9.995521763676645e-06,
"loss": 0.3097,
"step": 9
},
{
"epoch": 0.029940119760479042,
"grad_norm": 1.797680315451637,
"learning_rate": 9.994471506800078e-06,
"loss": 0.2547,
"step": 10
},
{
"epoch": 0.03293413173652695,
"grad_norm": 1.6028740343587435,
"learning_rate": 9.993310782160439e-06,
"loss": 0.2716,
"step": 11
},
{
"epoch": 0.03592814371257485,
"grad_norm": 1.6624923116034278,
"learning_rate": 9.992039615430648e-06,
"loss": 0.2457,
"step": 12
},
{
"epoch": 0.038922155688622756,
"grad_norm": 1.4851806876357339,
"learning_rate": 9.99065803472638e-06,
"loss": 0.2793,
"step": 13
},
{
"epoch": 0.041916167664670656,
"grad_norm": 1.6286442122708549,
"learning_rate": 9.989166070605447e-06,
"loss": 0.2457,
"step": 14
},
{
"epoch": 0.04491017964071856,
"grad_norm": 1.5939500051962867,
"learning_rate": 9.98756375606713e-06,
"loss": 0.2367,
"step": 15
},
{
"epoch": 0.04790419161676647,
"grad_norm": 1.420509428470971,
"learning_rate": 9.985851126551428e-06,
"loss": 0.2316,
"step": 16
},
{
"epoch": 0.05089820359281437,
"grad_norm": 1.160800465042313,
"learning_rate": 9.9840282199383e-06,
"loss": 0.1907,
"step": 17
},
{
"epoch": 0.05389221556886228,
"grad_norm": 1.4078790976938533,
"learning_rate": 9.982095076546806e-06,
"loss": 0.24,
"step": 18
},
{
"epoch": 0.05688622754491018,
"grad_norm": 2.0052537477976338,
"learning_rate": 9.980051739134235e-06,
"loss": 0.2362,
"step": 19
},
{
"epoch": 0.059880239520958084,
"grad_norm": 1.3889491746165774,
"learning_rate": 9.977898252895133e-06,
"loss": 0.2344,
"step": 20
},
{
"epoch": 0.06287425149700598,
"grad_norm": 1.3855874020137384,
"learning_rate": 9.975634665460333e-06,
"loss": 0.2236,
"step": 21
},
{
"epoch": 0.0658682634730539,
"grad_norm": 1.478688493729278,
"learning_rate": 9.973261026895878e-06,
"loss": 0.2335,
"step": 22
},
{
"epoch": 0.0688622754491018,
"grad_norm": 1.4309781016816272,
"learning_rate": 9.970777389701927e-06,
"loss": 0.2236,
"step": 23
},
{
"epoch": 0.0718562874251497,
"grad_norm": 1.199586528434631,
"learning_rate": 9.968183808811586e-06,
"loss": 0.1931,
"step": 24
},
{
"epoch": 0.0748502994011976,
"grad_norm": 1.4638579009118322,
"learning_rate": 9.965480341589702e-06,
"loss": 0.2246,
"step": 25
},
{
"epoch": 0.07784431137724551,
"grad_norm": 1.3973759708079496,
"learning_rate": 9.962667047831585e-06,
"loss": 0.2322,
"step": 26
},
{
"epoch": 0.08083832335329341,
"grad_norm": 1.213999348003484,
"learning_rate": 9.95974398976169e-06,
"loss": 0.2413,
"step": 27
},
{
"epoch": 0.08383233532934131,
"grad_norm": 1.1601348155377271,
"learning_rate": 9.95671123203224e-06,
"loss": 0.2236,
"step": 28
},
{
"epoch": 0.08682634730538923,
"grad_norm": 1.3714427838591285,
"learning_rate": 9.953568841721796e-06,
"loss": 0.2189,
"step": 29
},
{
"epoch": 0.08982035928143713,
"grad_norm": 1.1399197161600574,
"learning_rate": 9.950316888333775e-06,
"loss": 0.2128,
"step": 30
},
{
"epoch": 0.09281437125748503,
"grad_norm": 1.0774535593197172,
"learning_rate": 9.946955443794908e-06,
"loss": 0.2184,
"step": 31
},
{
"epoch": 0.09580838323353294,
"grad_norm": 1.2855868662382721,
"learning_rate": 9.943484582453653e-06,
"loss": 0.2266,
"step": 32
},
{
"epoch": 0.09880239520958084,
"grad_norm": 1.1367533635019165,
"learning_rate": 9.939904381078553e-06,
"loss": 0.1704,
"step": 33
},
{
"epoch": 0.10179640718562874,
"grad_norm": 1.2868725184024667,
"learning_rate": 9.93621491885653e-06,
"loss": 0.2486,
"step": 34
},
{
"epoch": 0.10479041916167664,
"grad_norm": 1.0529560896183001,
"learning_rate": 9.932416277391144e-06,
"loss": 0.1864,
"step": 35
},
{
"epoch": 0.10778443113772455,
"grad_norm": 1.060688039144272,
"learning_rate": 9.928508540700775e-06,
"loss": 0.2179,
"step": 36
},
{
"epoch": 0.11077844311377245,
"grad_norm": 1.4480578036381377,
"learning_rate": 9.924491795216777e-06,
"loss": 0.2496,
"step": 37
},
{
"epoch": 0.11377245508982035,
"grad_norm": 1.229310035254144,
"learning_rate": 9.920366129781564e-06,
"loss": 0.2104,
"step": 38
},
{
"epoch": 0.11676646706586827,
"grad_norm": 0.9752973514244784,
"learning_rate": 9.916131635646635e-06,
"loss": 0.1918,
"step": 39
},
{
"epoch": 0.11976047904191617,
"grad_norm": 1.215316633457128,
"learning_rate": 9.91178840647057e-06,
"loss": 0.2218,
"step": 40
},
{
"epoch": 0.12275449101796407,
"grad_norm": 1.0592819366757202,
"learning_rate": 9.907336538316946e-06,
"loss": 0.2157,
"step": 41
},
{
"epoch": 0.12574850299401197,
"grad_norm": 0.8952284692726973,
"learning_rate": 9.902776129652223e-06,
"loss": 0.1693,
"step": 42
},
{
"epoch": 0.12874251497005987,
"grad_norm": 1.1371161725074697,
"learning_rate": 9.898107281343557e-06,
"loss": 0.1968,
"step": 43
},
{
"epoch": 0.1317365269461078,
"grad_norm": 0.9761401052396169,
"learning_rate": 9.893330096656576e-06,
"loss": 0.168,
"step": 44
},
{
"epoch": 0.1347305389221557,
"grad_norm": 1.15744506860821,
"learning_rate": 9.888444681253087e-06,
"loss": 0.1902,
"step": 45
},
{
"epoch": 0.1377245508982036,
"grad_norm": 1.2298147136548452,
"learning_rate": 9.883451143188753e-06,
"loss": 0.2145,
"step": 46
},
{
"epoch": 0.1407185628742515,
"grad_norm": 1.0247940024843436,
"learning_rate": 9.878349592910694e-06,
"loss": 0.1818,
"step": 47
},
{
"epoch": 0.1437125748502994,
"grad_norm": 1.3186315420705188,
"learning_rate": 9.873140143255035e-06,
"loss": 0.2297,
"step": 48
},
{
"epoch": 0.1467065868263473,
"grad_norm": 1.199731238582795,
"learning_rate": 9.867822909444435e-06,
"loss": 0.2385,
"step": 49
},
{
"epoch": 0.1497005988023952,
"grad_norm": 1.1839482106123274,
"learning_rate": 9.862398009085511e-06,
"loss": 0.1749,
"step": 50
},
{
"epoch": 0.15269461077844312,
"grad_norm": 1.118825264862539,
"learning_rate": 9.856865562166256e-06,
"loss": 0.2094,
"step": 51
},
{
"epoch": 0.15568862275449102,
"grad_norm": 1.0444452183788182,
"learning_rate": 9.851225691053382e-06,
"loss": 0.1961,
"step": 52
},
{
"epoch": 0.15868263473053892,
"grad_norm": 1.1143406572520593,
"learning_rate": 9.8454785204896e-06,
"loss": 0.2277,
"step": 53
},
{
"epoch": 0.16167664670658682,
"grad_norm": 1.0368685510929592,
"learning_rate": 9.83962417759088e-06,
"loss": 0.1714,
"step": 54
},
{
"epoch": 0.16467065868263472,
"grad_norm": 1.1071257917006954,
"learning_rate": 9.833662791843628e-06,
"loss": 0.1941,
"step": 55
},
{
"epoch": 0.16766467065868262,
"grad_norm": 1.0631141307184537,
"learning_rate": 9.827594495101824e-06,
"loss": 0.2228,
"step": 56
},
{
"epoch": 0.17065868263473055,
"grad_norm": 1.185788150615812,
"learning_rate": 9.821419421584108e-06,
"loss": 0.2166,
"step": 57
},
{
"epoch": 0.17365269461077845,
"grad_norm": 1.2021004373140303,
"learning_rate": 9.815137707870806e-06,
"loss": 0.2842,
"step": 58
},
{
"epoch": 0.17664670658682635,
"grad_norm": 1.103199652486915,
"learning_rate": 9.808749492900917e-06,
"loss": 0.2075,
"step": 59
},
{
"epoch": 0.17964071856287425,
"grad_norm": 1.1735054068180313,
"learning_rate": 9.802254917969033e-06,
"loss": 0.2145,
"step": 60
},
{
"epoch": 0.18263473053892215,
"grad_norm": 1.3570777234997957,
"learning_rate": 9.795654126722218e-06,
"loss": 0.2257,
"step": 61
},
{
"epoch": 0.18562874251497005,
"grad_norm": 1.0669997217318379,
"learning_rate": 9.788947265156828e-06,
"loss": 0.1954,
"step": 62
},
{
"epoch": 0.18862275449101795,
"grad_norm": 1.0485872804775587,
"learning_rate": 9.782134481615282e-06,
"loss": 0.1548,
"step": 63
},
{
"epoch": 0.19161676646706588,
"grad_norm": 1.0698550472909447,
"learning_rate": 9.775215926782788e-06,
"loss": 0.2155,
"step": 64
},
{
"epoch": 0.19461077844311378,
"grad_norm": 1.2121835682976234,
"learning_rate": 9.768191753683997e-06,
"loss": 0.1991,
"step": 65
},
{
"epoch": 0.19760479041916168,
"grad_norm": 0.990626116535723,
"learning_rate": 9.761062117679632e-06,
"loss": 0.1928,
"step": 66
},
{
"epoch": 0.20059880239520958,
"grad_norm": 0.9413154074429412,
"learning_rate": 9.75382717646304e-06,
"loss": 0.1964,
"step": 67
},
{
"epoch": 0.20359281437125748,
"grad_norm": 1.057457657437376,
"learning_rate": 9.746487090056712e-06,
"loss": 0.1799,
"step": 68
},
{
"epoch": 0.20658682634730538,
"grad_norm": 0.9689737280070622,
"learning_rate": 9.739042020808746e-06,
"loss": 0.1857,
"step": 69
},
{
"epoch": 0.20958083832335328,
"grad_norm": 1.0643677107244633,
"learning_rate": 9.73149213338924e-06,
"loss": 0.1927,
"step": 70
},
{
"epoch": 0.2125748502994012,
"grad_norm": 1.1952426709269954,
"learning_rate": 9.72383759478667e-06,
"loss": 0.2369,
"step": 71
},
{
"epoch": 0.2155688622754491,
"grad_norm": 1.0762914524312819,
"learning_rate": 9.71607857430419e-06,
"loss": 0.1552,
"step": 72
},
{
"epoch": 0.218562874251497,
"grad_norm": 1.2159219254240334,
"learning_rate": 9.708215243555875e-06,
"loss": 0.2276,
"step": 73
},
{
"epoch": 0.2215568862275449,
"grad_norm": 1.1420392051285961,
"learning_rate": 9.700247776462944e-06,
"loss": 0.1778,
"step": 74
},
{
"epoch": 0.2245508982035928,
"grad_norm": 1.1809805195363519,
"learning_rate": 9.6921763492499e-06,
"loss": 0.2132,
"step": 75
},
{
"epoch": 0.2275449101796407,
"grad_norm": 0.8737455329843555,
"learning_rate": 9.68400114044064e-06,
"loss": 0.1532,
"step": 76
},
{
"epoch": 0.23053892215568864,
"grad_norm": 1.1026307657703391,
"learning_rate": 9.6757223308545e-06,
"loss": 0.1758,
"step": 77
},
{
"epoch": 0.23353293413173654,
"grad_norm": 1.110272194059896,
"learning_rate": 9.667340103602263e-06,
"loss": 0.2155,
"step": 78
},
{
"epoch": 0.23652694610778444,
"grad_norm": 1.0593925940813294,
"learning_rate": 9.658854644082099e-06,
"loss": 0.1938,
"step": 79
},
{
"epoch": 0.23952095808383234,
"grad_norm": 1.4648975096637658,
"learning_rate": 9.650266139975474e-06,
"loss": 0.2533,
"step": 80
},
{
"epoch": 0.24251497005988024,
"grad_norm": 1.0073488999654043,
"learning_rate": 9.641574781242999e-06,
"loss": 0.1875,
"step": 81
},
{
"epoch": 0.24550898203592814,
"grad_norm": 1.0018511990029086,
"learning_rate": 9.632780760120217e-06,
"loss": 0.1822,
"step": 82
},
{
"epoch": 0.24850299401197604,
"grad_norm": 1.0676938755908862,
"learning_rate": 9.62388427111336e-06,
"loss": 0.1829,
"step": 83
},
{
"epoch": 0.25149700598802394,
"grad_norm": 1.0111211660013757,
"learning_rate": 9.614885510995047e-06,
"loss": 0.1728,
"step": 84
},
{
"epoch": 0.25449101796407186,
"grad_norm": 0.99572317889892,
"learning_rate": 9.605784678799934e-06,
"loss": 0.1515,
"step": 85
},
{
"epoch": 0.25748502994011974,
"grad_norm": 1.1793435972382196,
"learning_rate": 9.596581975820304e-06,
"loss": 0.1858,
"step": 86
},
{
"epoch": 0.26047904191616766,
"grad_norm": 1.0934326939261134,
"learning_rate": 9.587277605601617e-06,
"loss": 0.1872,
"step": 87
},
{
"epoch": 0.2634730538922156,
"grad_norm": 1.2761921354928731,
"learning_rate": 9.577871773938013e-06,
"loss": 0.1907,
"step": 88
},
{
"epoch": 0.26646706586826346,
"grad_norm": 1.2526409248001873,
"learning_rate": 9.568364688867757e-06,
"loss": 0.2524,
"step": 89
},
{
"epoch": 0.2694610778443114,
"grad_norm": 0.9915448190020771,
"learning_rate": 9.558756560668637e-06,
"loss": 0.1817,
"step": 90
},
{
"epoch": 0.27245508982035926,
"grad_norm": 1.1594028850612275,
"learning_rate": 9.549047601853313e-06,
"loss": 0.2027,
"step": 91
},
{
"epoch": 0.2754491017964072,
"grad_norm": 1.0538986353077433,
"learning_rate": 9.539238027164618e-06,
"loss": 0.1947,
"step": 92
},
{
"epoch": 0.27844311377245506,
"grad_norm": 1.0433563023163,
"learning_rate": 9.52932805357081e-06,
"loss": 0.1657,
"step": 93
},
{
"epoch": 0.281437125748503,
"grad_norm": 1.0865001262271157,
"learning_rate": 9.519317900260769e-06,
"loss": 0.2005,
"step": 94
},
{
"epoch": 0.2844311377245509,
"grad_norm": 0.9695362510496279,
"learning_rate": 9.509207788639148e-06,
"loss": 0.18,
"step": 95
},
{
"epoch": 0.2874251497005988,
"grad_norm": 1.0273556363429448,
"learning_rate": 9.498997942321484e-06,
"loss": 0.2081,
"step": 96
},
{
"epoch": 0.2904191616766467,
"grad_norm": 0.9320382980640325,
"learning_rate": 9.488688587129243e-06,
"loss": 0.1745,
"step": 97
},
{
"epoch": 0.2934131736526946,
"grad_norm": 0.9988731898033898,
"learning_rate": 9.47827995108483e-06,
"loss": 0.1581,
"step": 98
},
{
"epoch": 0.2964071856287425,
"grad_norm": 1.1961886622595943,
"learning_rate": 9.467772264406545e-06,
"loss": 0.2097,
"step": 99
},
{
"epoch": 0.2994011976047904,
"grad_norm": 0.9969878235469215,
"learning_rate": 9.457165759503492e-06,
"loss": 0.1801,
"step": 100
},
{
"epoch": 0.3023952095808383,
"grad_norm": 1.1620618710802573,
"learning_rate": 9.446460670970436e-06,
"loss": 0.1571,
"step": 101
},
{
"epoch": 0.30538922155688625,
"grad_norm": 1.0089954841085507,
"learning_rate": 9.435657235582616e-06,
"loss": 0.1513,
"step": 102
},
{
"epoch": 0.3083832335329341,
"grad_norm": 1.1549135896257823,
"learning_rate": 9.424755692290507e-06,
"loss": 0.1852,
"step": 103
},
{
"epoch": 0.31137724550898205,
"grad_norm": 1.1383534220672937,
"learning_rate": 9.413756282214538e-06,
"loss": 0.1886,
"step": 104
},
{
"epoch": 0.3143712574850299,
"grad_norm": 0.9937878106325314,
"learning_rate": 9.402659248639749e-06,
"loss": 0.1883,
"step": 105
},
{
"epoch": 0.31736526946107785,
"grad_norm": 1.1129420316441838,
"learning_rate": 9.391464837010428e-06,
"loss": 0.1897,
"step": 106
},
{
"epoch": 0.3203592814371258,
"grad_norm": 1.2313930833907614,
"learning_rate": 9.380173294924661e-06,
"loss": 0.2063,
"step": 107
},
{
"epoch": 0.32335329341317365,
"grad_norm": 1.0846586731662082,
"learning_rate": 9.368784872128877e-06,
"loss": 0.2012,
"step": 108
},
{
"epoch": 0.3263473053892216,
"grad_norm": 1.1590821798409205,
"learning_rate": 9.357299820512305e-06,
"loss": 0.1915,
"step": 109
},
{
"epoch": 0.32934131736526945,
"grad_norm": 1.1839350360721475,
"learning_rate": 9.345718394101412e-06,
"loss": 0.2046,
"step": 110
},
{
"epoch": 0.3323353293413174,
"grad_norm": 1.2382980854013712,
"learning_rate": 9.334040849054288e-06,
"loss": 0.2339,
"step": 111
},
{
"epoch": 0.33532934131736525,
"grad_norm": 1.0455267795927479,
"learning_rate": 9.322267443654974e-06,
"loss": 0.1843,
"step": 112
},
{
"epoch": 0.3383233532934132,
"grad_norm": 1.0978070870220789,
"learning_rate": 9.310398438307747e-06,
"loss": 0.2345,
"step": 113
},
{
"epoch": 0.3413173652694611,
"grad_norm": 1.0980445092751268,
"learning_rate": 9.29843409553137e-06,
"loss": 0.1826,
"step": 114
},
{
"epoch": 0.344311377245509,
"grad_norm": 1.0244932918108935,
"learning_rate": 9.286374679953278e-06,
"loss": 0.1827,
"step": 115
},
{
"epoch": 0.3473053892215569,
"grad_norm": 0.9766314584557123,
"learning_rate": 9.274220458303727e-06,
"loss": 0.1805,
"step": 116
},
{
"epoch": 0.3502994011976048,
"grad_norm": 1.245078231054336,
"learning_rate": 9.261971699409893e-06,
"loss": 0.231,
"step": 117
},
{
"epoch": 0.3532934131736527,
"grad_norm": 1.2500627086462586,
"learning_rate": 9.249628674189928e-06,
"loss": 0.188,
"step": 118
},
{
"epoch": 0.3562874251497006,
"grad_norm": 1.07615268950645,
"learning_rate": 9.237191655646972e-06,
"loss": 0.209,
"step": 119
},
{
"epoch": 0.3592814371257485,
"grad_norm": 1.1465286146661815,
"learning_rate": 9.224660918863104e-06,
"loss": 0.1808,
"step": 120
},
{
"epoch": 0.36227544910179643,
"grad_norm": 1.0810184295472054,
"learning_rate": 9.212036740993265e-06,
"loss": 0.1913,
"step": 121
},
{
"epoch": 0.3652694610778443,
"grad_norm": 1.1621044310258672,
"learning_rate": 9.199319401259132e-06,
"loss": 0.1837,
"step": 122
},
{
"epoch": 0.36826347305389223,
"grad_norm": 1.075496064430152,
"learning_rate": 9.186509180942928e-06,
"loss": 0.1489,
"step": 123
},
{
"epoch": 0.3712574850299401,
"grad_norm": 1.0314250180698594,
"learning_rate": 9.173606363381218e-06,
"loss": 0.1866,
"step": 124
},
{
"epoch": 0.37425149700598803,
"grad_norm": 1.0390982033675518,
"learning_rate": 9.16061123395863e-06,
"loss": 0.1547,
"step": 125
},
{
"epoch": 0.3772455089820359,
"grad_norm": 1.140332641631481,
"learning_rate": 9.147524080101543e-06,
"loss": 0.1733,
"step": 126
},
{
"epoch": 0.38023952095808383,
"grad_norm": 1.1973509089045273,
"learning_rate": 9.134345191271742e-06,
"loss": 0.201,
"step": 127
},
{
"epoch": 0.38323353293413176,
"grad_norm": 1.0892699993328228,
"learning_rate": 9.121074858959997e-06,
"loss": 0.2129,
"step": 128
},
{
"epoch": 0.38622754491017963,
"grad_norm": 1.1514737930635182,
"learning_rate": 9.107713376679634e-06,
"loss": 0.2064,
"step": 129
},
{
"epoch": 0.38922155688622756,
"grad_norm": 1.171003282755829,
"learning_rate": 9.094261039960028e-06,
"loss": 0.1882,
"step": 130
},
{
"epoch": 0.39221556886227543,
"grad_norm": 1.2210388190830823,
"learning_rate": 9.08071814634008e-06,
"loss": 0.1964,
"step": 131
},
{
"epoch": 0.39520958083832336,
"grad_norm": 1.0914451882016283,
"learning_rate": 9.067084995361623e-06,
"loss": 0.1796,
"step": 132
},
{
"epoch": 0.39820359281437123,
"grad_norm": 1.1433256612502316,
"learning_rate": 9.053361888562807e-06,
"loss": 0.2261,
"step": 133
},
{
"epoch": 0.40119760479041916,
"grad_norm": 1.0793771196116007,
"learning_rate": 9.039549129471423e-06,
"loss": 0.1829,
"step": 134
},
{
"epoch": 0.4041916167664671,
"grad_norm": 0.9313032725107899,
"learning_rate": 9.025647023598196e-06,
"loss": 0.1643,
"step": 135
},
{
"epoch": 0.40718562874251496,
"grad_norm": 1.0040398241493609,
"learning_rate": 9.011655878430018e-06,
"loss": 0.1655,
"step": 136
},
{
"epoch": 0.4101796407185629,
"grad_norm": 1.3144471830859998,
"learning_rate": 8.99757600342316e-06,
"loss": 0.2331,
"step": 137
},
{
"epoch": 0.41317365269461076,
"grad_norm": 1.090095813401022,
"learning_rate": 8.983407709996415e-06,
"loss": 0.1679,
"step": 138
},
{
"epoch": 0.4161676646706587,
"grad_norm": 0.9877496552102605,
"learning_rate": 8.969151311524215e-06,
"loss": 0.1448,
"step": 139
},
{
"epoch": 0.41916167664670656,
"grad_norm": 1.1391089805041432,
"learning_rate": 8.954807123329703e-06,
"loss": 0.1906,
"step": 140
},
{
"epoch": 0.4221556886227545,
"grad_norm": 0.9816847446658088,
"learning_rate": 8.940375462677758e-06,
"loss": 0.1825,
"step": 141
},
{
"epoch": 0.4251497005988024,
"grad_norm": 1.0667631129499833,
"learning_rate": 8.92585664876797e-06,
"loss": 0.1935,
"step": 142
},
{
"epoch": 0.4281437125748503,
"grad_norm": 0.9777959867455948,
"learning_rate": 8.911251002727588e-06,
"loss": 0.1788,
"step": 143
},
{
"epoch": 0.4311377245508982,
"grad_norm": 0.9882304170689388,
"learning_rate": 8.896558847604414e-06,
"loss": 0.177,
"step": 144
},
{
"epoch": 0.4341317365269461,
"grad_norm": 1.0643024255768412,
"learning_rate": 8.881780508359661e-06,
"loss": 0.2003,
"step": 145
},
{
"epoch": 0.437125748502994,
"grad_norm": 1.2972319681523306,
"learning_rate": 8.86691631186076e-06,
"loss": 0.2014,
"step": 146
},
{
"epoch": 0.44011976047904194,
"grad_norm": 0.9431046604921747,
"learning_rate": 8.851966586874138e-06,
"loss": 0.187,
"step": 147
},
{
"epoch": 0.4431137724550898,
"grad_norm": 1.1755087369044228,
"learning_rate": 8.836931664057935e-06,
"loss": 0.2079,
"step": 148
},
{
"epoch": 0.44610778443113774,
"grad_norm": 1.1231916460137625,
"learning_rate": 8.821811875954705e-06,
"loss": 0.1713,
"step": 149
},
{
"epoch": 0.4491017964071856,
"grad_norm": 1.035580875400447,
"learning_rate": 8.806607556984045e-06,
"loss": 0.1987,
"step": 150
},
{
"epoch": 0.45209580838323354,
"grad_norm": 1.0730797448130465,
"learning_rate": 8.791319043435213e-06,
"loss": 0.2031,
"step": 151
},
{
"epoch": 0.4550898203592814,
"grad_norm": 0.9908673117276247,
"learning_rate": 8.775946673459682e-06,
"loss": 0.1904,
"step": 152
},
{
"epoch": 0.45808383233532934,
"grad_norm": 0.9976482344657122,
"learning_rate": 8.76049078706366e-06,
"loss": 0.1655,
"step": 153
},
{
"epoch": 0.46107784431137727,
"grad_norm": 1.0065854192111559,
"learning_rate": 8.744951726100572e-06,
"loss": 0.1605,
"step": 154
},
{
"epoch": 0.46407185628742514,
"grad_norm": 1.004113977597953,
"learning_rate": 8.729329834263503e-06,
"loss": 0.1469,
"step": 155
},
{
"epoch": 0.46706586826347307,
"grad_norm": 1.1429652944231066,
"learning_rate": 8.713625457077585e-06,
"loss": 0.2128,
"step": 156
},
{
"epoch": 0.47005988023952094,
"grad_norm": 1.195064268054406,
"learning_rate": 8.697838941892371e-06,
"loss": 0.1858,
"step": 157
},
{
"epoch": 0.47305389221556887,
"grad_norm": 0.9971041614220809,
"learning_rate": 8.681970637874131e-06,
"loss": 0.2029,
"step": 158
},
{
"epoch": 0.47604790419161674,
"grad_norm": 1.143426130959058,
"learning_rate": 8.666020895998154e-06,
"loss": 0.2117,
"step": 159
},
{
"epoch": 0.47904191616766467,
"grad_norm": 1.1074731970652962,
"learning_rate": 8.64999006904096e-06,
"loss": 0.1765,
"step": 160
},
{
"epoch": 0.4820359281437126,
"grad_norm": 1.2156539019753092,
"learning_rate": 8.63387851157252e-06,
"loss": 0.1898,
"step": 161
},
{
"epoch": 0.48502994011976047,
"grad_norm": 1.0241349873882812,
"learning_rate": 8.617686579948396e-06,
"loss": 0.1866,
"step": 162
},
{
"epoch": 0.4880239520958084,
"grad_norm": 1.1729033825229422,
"learning_rate": 8.60141463230187e-06,
"loss": 0.1866,
"step": 163
},
{
"epoch": 0.49101796407185627,
"grad_norm": 1.2064360329146644,
"learning_rate": 8.585063028536015e-06,
"loss": 0.1897,
"step": 164
},
{
"epoch": 0.4940119760479042,
"grad_norm": 1.1998211725378871,
"learning_rate": 8.568632130315747e-06,
"loss": 0.1987,
"step": 165
},
{
"epoch": 0.49700598802395207,
"grad_norm": 1.0515498405842385,
"learning_rate": 8.552122301059807e-06,
"loss": 0.1659,
"step": 166
},
{
"epoch": 0.5,
"grad_norm": 1.092845201970978,
"learning_rate": 8.535533905932739e-06,
"loss": 0.1751,
"step": 167
},
{
"epoch": 0.5029940119760479,
"grad_norm": 0.9800625655478471,
"learning_rate": 8.518867311836808e-06,
"loss": 0.2257,
"step": 168
},
{
"epoch": 0.5059880239520959,
"grad_norm": 1.0029352788262598,
"learning_rate": 8.502122887403882e-06,
"loss": 0.1553,
"step": 169
},
{
"epoch": 0.5089820359281437,
"grad_norm": 1.0575758579748156,
"learning_rate": 8.485301002987285e-06,
"loss": 0.1617,
"step": 170
},
{
"epoch": 0.5119760479041916,
"grad_norm": 1.0675440692743716,
"learning_rate": 8.468402030653598e-06,
"loss": 0.2154,
"step": 171
},
{
"epoch": 0.5149700598802395,
"grad_norm": 0.9937563224263936,
"learning_rate": 8.451426344174433e-06,
"loss": 0.1936,
"step": 172
},
{
"epoch": 0.5179640718562875,
"grad_norm": 1.1564319817528537,
"learning_rate": 8.434374319018165e-06,
"loss": 0.1815,
"step": 173
},
{
"epoch": 0.5209580838323353,
"grad_norm": 1.095629825119387,
"learning_rate": 8.417246332341638e-06,
"loss": 0.1649,
"step": 174
},
{
"epoch": 0.5239520958083832,
"grad_norm": 1.1356867336530758,
"learning_rate": 8.4000427629818e-06,
"loss": 0.1994,
"step": 175
},
{
"epoch": 0.5269461077844312,
"grad_norm": 0.9898184362299568,
"learning_rate": 8.382763991447344e-06,
"loss": 0.1596,
"step": 176
},
{
"epoch": 0.5299401197604791,
"grad_norm": 1.1721276063600639,
"learning_rate": 8.365410399910287e-06,
"loss": 0.1993,
"step": 177
},
{
"epoch": 0.5329341317365269,
"grad_norm": 1.207382801140143,
"learning_rate": 8.347982372197515e-06,
"loss": 0.1781,
"step": 178
},
{
"epoch": 0.5359281437125748,
"grad_norm": 1.1124925532166055,
"learning_rate": 8.33048029378229e-06,
"loss": 0.1899,
"step": 179
},
{
"epoch": 0.5389221556886228,
"grad_norm": 1.0702701369576466,
"learning_rate": 8.312904551775731e-06,
"loss": 0.2228,
"step": 180
},
{
"epoch": 0.5419161676646707,
"grad_norm": 1.167191638850714,
"learning_rate": 8.295255534918249e-06,
"loss": 0.2334,
"step": 181
},
{
"epoch": 0.5449101796407185,
"grad_norm": 1.1287161009521436,
"learning_rate": 8.277533633570948e-06,
"loss": 0.1907,
"step": 182
},
{
"epoch": 0.5479041916167665,
"grad_norm": 1.1484304526883422,
"learning_rate": 8.25973923970699e-06,
"loss": 0.1792,
"step": 183
},
{
"epoch": 0.5508982035928144,
"grad_norm": 0.9814088592391755,
"learning_rate": 8.241872746902934e-06,
"loss": 0.1972,
"step": 184
},
{
"epoch": 0.5538922155688623,
"grad_norm": 1.073621134321228,
"learning_rate": 8.223934550330015e-06,
"loss": 0.1956,
"step": 185
},
{
"epoch": 0.5568862275449101,
"grad_norm": 1.1166531899346925,
"learning_rate": 8.20592504674542e-06,
"loss": 0.2134,
"step": 186
},
{
"epoch": 0.5598802395209581,
"grad_norm": 1.3767248433689827,
"learning_rate": 8.187844634483495e-06,
"loss": 0.2311,
"step": 187
},
{
"epoch": 0.562874251497006,
"grad_norm": 1.2216941427182566,
"learning_rate": 8.16969371344696e-06,
"loss": 0.2009,
"step": 188
},
{
"epoch": 0.5658682634730539,
"grad_norm": 1.1694713719173795,
"learning_rate": 8.151472685098037e-06,
"loss": 0.2173,
"step": 189
},
{
"epoch": 0.5688622754491018,
"grad_norm": 0.9509186227831858,
"learning_rate": 8.13318195244958e-06,
"loss": 0.1848,
"step": 190
},
{
"epoch": 0.5718562874251497,
"grad_norm": 1.0353492930952461,
"learning_rate": 8.114821920056177e-06,
"loss": 0.1733,
"step": 191
},
{
"epoch": 0.5748502994011976,
"grad_norm": 1.1729661160272977,
"learning_rate": 8.096392994005177e-06,
"loss": 0.2247,
"step": 192
},
{
"epoch": 0.5778443113772455,
"grad_norm": 0.9935556814181943,
"learning_rate": 8.077895581907719e-06,
"loss": 0.1925,
"step": 193
},
{
"epoch": 0.5808383233532934,
"grad_norm": 0.999276530710793,
"learning_rate": 8.059330092889724e-06,
"loss": 0.1795,
"step": 194
},
{
"epoch": 0.5838323353293413,
"grad_norm": 1.1664539704139143,
"learning_rate": 8.040696937582833e-06,
"loss": 0.2262,
"step": 195
},
{
"epoch": 0.5868263473053892,
"grad_norm": 0.9761615189299239,
"learning_rate": 8.021996528115335e-06,
"loss": 0.1606,
"step": 196
},
{
"epoch": 0.5898203592814372,
"grad_norm": 1.107893538938559,
"learning_rate": 8.003229278103044e-06,
"loss": 0.1639,
"step": 197
},
{
"epoch": 0.592814371257485,
"grad_norm": 1.1857966631635803,
"learning_rate": 7.984395602640153e-06,
"loss": 0.1857,
"step": 198
},
{
"epoch": 0.5958083832335329,
"grad_norm": 1.0762086947496277,
"learning_rate": 7.96549591829006e-06,
"loss": 0.1698,
"step": 199
},
{
"epoch": 0.5988023952095808,
"grad_norm": 1.0119650005461605,
"learning_rate": 7.946530643076138e-06,
"loss": 0.1906,
"step": 200
},
{
"epoch": 0.5988023952095808,
"eval_loss": 0.19403506815433502,
"eval_runtime": 2.0264,
"eval_samples_per_second": 26.648,
"eval_steps_per_second": 6.909,
"step": 200
},
{
"epoch": 0.6017964071856288,
"grad_norm": 0.8670867619750887,
"learning_rate": 7.927500196472506e-06,
"loss": 0.1807,
"step": 201
},
{
"epoch": 0.6047904191616766,
"grad_norm": 1.1080056255964856,
"learning_rate": 7.908404999394747e-06,
"loss": 0.1632,
"step": 202
},
{
"epoch": 0.6077844311377245,
"grad_norm": 1.2794566598448773,
"learning_rate": 7.889245474190588e-06,
"loss": 0.2055,
"step": 203
},
{
"epoch": 0.6107784431137725,
"grad_norm": 1.402113148727178,
"learning_rate": 7.870022044630569e-06,
"loss": 0.2274,
"step": 204
},
{
"epoch": 0.6137724550898204,
"grad_norm": 0.9876013024237756,
"learning_rate": 7.85073513589867e-06,
"loss": 0.1808,
"step": 205
},
{
"epoch": 0.6167664670658682,
"grad_norm": 1.0775534299204246,
"learning_rate": 7.831385174582901e-06,
"loss": 0.1788,
"step": 206
},
{
"epoch": 0.6197604790419161,
"grad_norm": 1.027795331227106,
"learning_rate": 7.81197258866587e-06,
"loss": 0.1698,
"step": 207
},
{
"epoch": 0.6227544910179641,
"grad_norm": 1.0129694339445037,
"learning_rate": 7.792497807515317e-06,
"loss": 0.1748,
"step": 208
},
{
"epoch": 0.625748502994012,
"grad_norm": 0.964664757829955,
"learning_rate": 7.772961261874615e-06,
"loss": 0.1681,
"step": 209
},
{
"epoch": 0.6287425149700598,
"grad_norm": 1.1256327178304772,
"learning_rate": 7.75336338385325e-06,
"loss": 0.1809,
"step": 210
},
{
"epoch": 0.6317365269461078,
"grad_norm": 1.117439139811535,
"learning_rate": 7.733704606917248e-06,
"loss": 0.2235,
"step": 211
},
{
"epoch": 0.6347305389221557,
"grad_norm": 1.119314811289219,
"learning_rate": 7.713985365879607e-06,
"loss": 0.1779,
"step": 212
},
{
"epoch": 0.6377245508982036,
"grad_norm": 1.2003383198858455,
"learning_rate": 7.694206096890667e-06,
"loss": 0.2386,
"step": 213
},
{
"epoch": 0.6407185628742516,
"grad_norm": 0.9921842649589848,
"learning_rate": 7.674367237428467e-06,
"loss": 0.1661,
"step": 214
},
{
"epoch": 0.6437125748502994,
"grad_norm": 1.0187256824731497,
"learning_rate": 7.654469226289068e-06,
"loss": 0.1723,
"step": 215
},
{
"epoch": 0.6467065868263473,
"grad_norm": 1.177744437744105,
"learning_rate": 7.63451250357685e-06,
"loss": 0.2055,
"step": 216
},
{
"epoch": 0.6497005988023952,
"grad_norm": 1.077919096367905,
"learning_rate": 7.614497510694774e-06,
"loss": 0.1809,
"step": 217
},
{
"epoch": 0.6526946107784432,
"grad_norm": 1.0849470179916494,
"learning_rate": 7.5944246903346204e-06,
"loss": 0.1874,
"step": 218
},
{
"epoch": 0.655688622754491,
"grad_norm": 0.9380239148428942,
"learning_rate": 7.574294486467204e-06,
"loss": 0.1465,
"step": 219
},
{
"epoch": 0.6586826347305389,
"grad_norm": 0.9481603069565345,
"learning_rate": 7.55410734433254e-06,
"loss": 0.1698,
"step": 220
},
{
"epoch": 0.6616766467065869,
"grad_norm": 1.1591476508876961,
"learning_rate": 7.533863710430011e-06,
"loss": 0.2035,
"step": 221
},
{
"epoch": 0.6646706586826348,
"grad_norm": 0.8029977142306275,
"learning_rate": 7.513564032508484e-06,
"loss": 0.142,
"step": 222
},
{
"epoch": 0.6676646706586826,
"grad_norm": 1.2722893791722965,
"learning_rate": 7.493208759556406e-06,
"loss": 0.2029,
"step": 223
},
{
"epoch": 0.6706586826347305,
"grad_norm": 1.1546451766759405,
"learning_rate": 7.472798341791877e-06,
"loss": 0.2023,
"step": 224
},
{
"epoch": 0.6736526946107785,
"grad_norm": 0.9124110935847347,
"learning_rate": 7.452333230652688e-06,
"loss": 0.1447,
"step": 225
},
{
"epoch": 0.6766467065868264,
"grad_norm": 1.0799524658830582,
"learning_rate": 7.431813878786343e-06,
"loss": 0.2169,
"step": 226
},
{
"epoch": 0.6796407185628742,
"grad_norm": 0.9714681663748792,
"learning_rate": 7.4112407400400395e-06,
"loss": 0.1315,
"step": 227
},
{
"epoch": 0.6826347305389222,
"grad_norm": 0.9318144186610402,
"learning_rate": 7.390614269450633e-06,
"loss": 0.1473,
"step": 228
},
{
"epoch": 0.6856287425149701,
"grad_norm": 1.0929794856582224,
"learning_rate": 7.369934923234577e-06,
"loss": 0.1562,
"step": 229
},
{
"epoch": 0.688622754491018,
"grad_norm": 1.1914365885408456,
"learning_rate": 7.349203158777826e-06,
"loss": 0.1762,
"step": 230
},
{
"epoch": 0.6916167664670658,
"grad_norm": 1.0051400812972897,
"learning_rate": 7.32841943462572e-06,
"loss": 0.1712,
"step": 231
},
{
"epoch": 0.6946107784431138,
"grad_norm": 1.1589015439483208,
"learning_rate": 7.3075842104728445e-06,
"loss": 0.1975,
"step": 232
},
{
"epoch": 0.6976047904191617,
"grad_norm": 1.072493554886191,
"learning_rate": 7.286697947152868e-06,
"loss": 0.1837,
"step": 233
},
{
"epoch": 0.7005988023952096,
"grad_norm": 0.9231193845094225,
"learning_rate": 7.265761106628338e-06,
"loss": 0.1567,
"step": 234
},
{
"epoch": 0.7035928143712575,
"grad_norm": 1.0095281552713928,
"learning_rate": 7.244774151980466e-06,
"loss": 0.172,
"step": 235
},
{
"epoch": 0.7065868263473054,
"grad_norm": 1.0912402302664788,
"learning_rate": 7.223737547398898e-06,
"loss": 0.1724,
"step": 236
},
{
"epoch": 0.7095808383233533,
"grad_norm": 1.0978372072121994,
"learning_rate": 7.20265175817143e-06,
"loss": 0.1893,
"step": 237
},
{
"epoch": 0.7125748502994012,
"grad_norm": 1.1675630640333539,
"learning_rate": 7.181517250673729e-06,
"loss": 0.2329,
"step": 238
},
{
"epoch": 0.7155688622754491,
"grad_norm": 1.0759703207878364,
"learning_rate": 7.1603344923590065e-06,
"loss": 0.1884,
"step": 239
},
{
"epoch": 0.718562874251497,
"grad_norm": 1.2202588582638902,
"learning_rate": 7.139103951747694e-06,
"loss": 0.2321,
"step": 240
},
{
"epoch": 0.7215568862275449,
"grad_norm": 1.053396310889173,
"learning_rate": 7.1178260984170675e-06,
"loss": 0.1796,
"step": 241
},
{
"epoch": 0.7245508982035929,
"grad_norm": 1.0224065131035902,
"learning_rate": 7.0965014029908654e-06,
"loss": 0.1604,
"step": 242
},
{
"epoch": 0.7275449101796407,
"grad_norm": 1.0259365033253793,
"learning_rate": 7.075130337128883e-06,
"loss": 0.2209,
"step": 243
},
{
"epoch": 0.7305389221556886,
"grad_norm": 1.1111222431261176,
"learning_rate": 7.053713373516538e-06,
"loss": 0.2229,
"step": 244
},
{
"epoch": 0.7335329341317365,
"grad_norm": 1.181540415327289,
"learning_rate": 7.03225098585441e-06,
"loss": 0.2264,
"step": 245
},
{
"epoch": 0.7365269461077845,
"grad_norm": 1.2341726734203058,
"learning_rate": 7.0107436488477694e-06,
"loss": 0.2308,
"step": 246
},
{
"epoch": 0.7395209580838323,
"grad_norm": 1.2542546597215263,
"learning_rate": 6.989191838196083e-06,
"loss": 0.1947,
"step": 247
},
{
"epoch": 0.7425149700598802,
"grad_norm": 0.9539702608475984,
"learning_rate": 6.9675960305824785e-06,
"loss": 0.1585,
"step": 248
},
{
"epoch": 0.7455089820359282,
"grad_norm": 1.1480458397573547,
"learning_rate": 6.945956703663212e-06,
"loss": 0.1857,
"step": 249
},
{
"epoch": 0.7485029940119761,
"grad_norm": 1.033070273816392,
"learning_rate": 6.9242743360570985e-06,
"loss": 0.1921,
"step": 250
},
{
"epoch": 0.7514970059880239,
"grad_norm": 1.091042140596752,
"learning_rate": 6.9025494073349284e-06,
"loss": 0.1948,
"step": 251
},
{
"epoch": 0.7544910179640718,
"grad_norm": 1.0955721668309388,
"learning_rate": 6.880782398008862e-06,
"loss": 0.2001,
"step": 252
},
{
"epoch": 0.7574850299401198,
"grad_norm": 1.0269047023230375,
"learning_rate": 6.858973789521792e-06,
"loss": 0.1753,
"step": 253
},
{
"epoch": 0.7604790419161677,
"grad_norm": 1.0635602199930048,
"learning_rate": 6.837124064236709e-06,
"loss": 0.1818,
"step": 254
},
{
"epoch": 0.7634730538922155,
"grad_norm": 0.9376369057249547,
"learning_rate": 6.815233705426019e-06,
"loss": 0.1748,
"step": 255
},
{
"epoch": 0.7664670658682635,
"grad_norm": 1.0152678397768258,
"learning_rate": 6.7933031972608644e-06,
"loss": 0.1713,
"step": 256
},
{
"epoch": 0.7694610778443114,
"grad_norm": 1.2431046391202505,
"learning_rate": 6.771333024800411e-06,
"loss": 0.2037,
"step": 257
},
{
"epoch": 0.7724550898203593,
"grad_norm": 1.2062361698466657,
"learning_rate": 6.74932367398112e-06,
"loss": 0.1955,
"step": 258
},
{
"epoch": 0.7754491017964071,
"grad_norm": 0.9354637760329599,
"learning_rate": 6.727275631605996e-06,
"loss": 0.1617,
"step": 259
},
{
"epoch": 0.7784431137724551,
"grad_norm": 1.077882112471302,
"learning_rate": 6.70518938533383e-06,
"loss": 0.1926,
"step": 260
},
{
"epoch": 0.781437125748503,
"grad_norm": 0.9618519239521814,
"learning_rate": 6.683065423668403e-06,
"loss": 0.1546,
"step": 261
},
{
"epoch": 0.7844311377245509,
"grad_norm": 1.0912517889369722,
"learning_rate": 6.660904235947687e-06,
"loss": 0.1807,
"step": 262
},
{
"epoch": 0.7874251497005988,
"grad_norm": 1.0114813437674908,
"learning_rate": 6.638706312333018e-06,
"loss": 0.1977,
"step": 263
},
{
"epoch": 0.7904191616766467,
"grad_norm": 1.1680767210070957,
"learning_rate": 6.61647214379826e-06,
"loss": 0.192,
"step": 264
},
{
"epoch": 0.7934131736526946,
"grad_norm": 1.3484758693142584,
"learning_rate": 6.594202222118941e-06,
"loss": 0.2184,
"step": 265
},
{
"epoch": 0.7964071856287425,
"grad_norm": 1.0128583818056613,
"learning_rate": 6.571897039861377e-06,
"loss": 0.1554,
"step": 266
},
{
"epoch": 0.7994011976047904,
"grad_norm": 1.088325853654402,
"learning_rate": 6.549557090371775e-06,
"loss": 0.199,
"step": 267
},
{
"epoch": 0.8023952095808383,
"grad_norm": 0.8625973408955091,
"learning_rate": 6.527182867765333e-06,
"loss": 0.1393,
"step": 268
},
{
"epoch": 0.8053892215568862,
"grad_norm": 1.3368710004340063,
"learning_rate": 6.504774866915291e-06,
"loss": 0.2379,
"step": 269
},
{
"epoch": 0.8083832335329342,
"grad_norm": 0.9432207335415682,
"learning_rate": 6.482333583442002e-06,
"loss": 0.155,
"step": 270
},
{
"epoch": 0.811377245508982,
"grad_norm": 1.1686539000933664,
"learning_rate": 6.459859513701967e-06,
"loss": 0.2048,
"step": 271
},
{
"epoch": 0.8143712574850299,
"grad_norm": 1.2955360948182033,
"learning_rate": 6.437353154776848e-06,
"loss": 0.1808,
"step": 272
},
{
"epoch": 0.8173652694610778,
"grad_norm": 1.042662804377474,
"learning_rate": 6.414815004462483e-06,
"loss": 0.1595,
"step": 273
},
{
"epoch": 0.8203592814371258,
"grad_norm": 1.0972171545457146,
"learning_rate": 6.3922455612578715e-06,
"loss": 0.2072,
"step": 274
},
{
"epoch": 0.8233532934131736,
"grad_norm": 1.0336059399738913,
"learning_rate": 6.369645324354149e-06,
"loss": 0.1577,
"step": 275
},
{
"epoch": 0.8263473053892215,
"grad_norm": 1.1505157826965169,
"learning_rate": 6.3470147936235485e-06,
"loss": 0.215,
"step": 276
},
{
"epoch": 0.8293413173652695,
"grad_norm": 0.9053127874700678,
"learning_rate": 6.3243544696083355e-06,
"loss": 0.1685,
"step": 277
},
{
"epoch": 0.8323353293413174,
"grad_norm": 1.2467394707770185,
"learning_rate": 6.301664853509755e-06,
"loss": 0.2052,
"step": 278
},
{
"epoch": 0.8353293413173652,
"grad_norm": 0.9830203378623781,
"learning_rate": 6.278946447176924e-06,
"loss": 0.1518,
"step": 279
},
{
"epoch": 0.8383233532934131,
"grad_norm": 1.0922515106879667,
"learning_rate": 6.256199753095745e-06,
"loss": 0.1834,
"step": 280
},
{
"epoch": 0.8413173652694611,
"grad_norm": 0.9723357985812227,
"learning_rate": 6.233425274377793e-06,
"loss": 0.1756,
"step": 281
},
{
"epoch": 0.844311377245509,
"grad_norm": 1.1144716905522072,
"learning_rate": 6.21062351474918e-06,
"loss": 0.1693,
"step": 282
},
{
"epoch": 0.8473053892215568,
"grad_norm": 1.0423737610688735,
"learning_rate": 6.18779497853942e-06,
"loss": 0.1788,
"step": 283
},
{
"epoch": 0.8502994011976048,
"grad_norm": 1.117885550779959,
"learning_rate": 6.164940170670266e-06,
"loss": 0.2091,
"step": 284
},
{
"epoch": 0.8532934131736527,
"grad_norm": 1.0298529967496155,
"learning_rate": 6.142059596644557e-06,
"loss": 0.1683,
"step": 285
},
{
"epoch": 0.8562874251497006,
"grad_norm": 0.9756291101598146,
"learning_rate": 6.11915376253502e-06,
"loss": 0.1511,
"step": 286
},
{
"epoch": 0.8592814371257484,
"grad_norm": 1.0042673700910687,
"learning_rate": 6.096223174973091e-06,
"loss": 0.1699,
"step": 287
},
{
"epoch": 0.8622754491017964,
"grad_norm": 0.9851715951325666,
"learning_rate": 6.073268341137694e-06,
"loss": 0.1408,
"step": 288
},
{
"epoch": 0.8652694610778443,
"grad_norm": 1.027112654035205,
"learning_rate": 6.050289768744042e-06,
"loss": 0.1679,
"step": 289
},
{
"epoch": 0.8682634730538922,
"grad_norm": 1.1510171033640384,
"learning_rate": 6.0272879660323936e-06,
"loss": 0.1946,
"step": 290
},
{
"epoch": 0.8712574850299402,
"grad_norm": 1.1451698297220647,
"learning_rate": 6.004263441756815e-06,
"loss": 0.1786,
"step": 291
},
{
"epoch": 0.874251497005988,
"grad_norm": 1.0997704939265416,
"learning_rate": 5.98121670517393e-06,
"loss": 0.1937,
"step": 292
},
{
"epoch": 0.8772455089820359,
"grad_norm": 0.8984124360537784,
"learning_rate": 5.958148266031654e-06,
"loss": 0.179,
"step": 293
},
{
"epoch": 0.8802395209580839,
"grad_norm": 0.9425256228444931,
"learning_rate": 5.935058634557917e-06,
"loss": 0.1677,
"step": 294
},
{
"epoch": 0.8832335329341318,
"grad_norm": 0.9608841729768236,
"learning_rate": 5.911948321449384e-06,
"loss": 0.1683,
"step": 295
},
{
"epoch": 0.8862275449101796,
"grad_norm": 1.117096522756304,
"learning_rate": 5.8888178378601565e-06,
"loss": 0.2047,
"step": 296
},
{
"epoch": 0.8892215568862275,
"grad_norm": 1.0714465323416391,
"learning_rate": 5.865667695390468e-06,
"loss": 0.1854,
"step": 297
},
{
"epoch": 0.8922155688622755,
"grad_norm": 0.9843579210903848,
"learning_rate": 5.842498406075363e-06,
"loss": 0.1576,
"step": 298
},
{
"epoch": 0.8952095808383234,
"grad_norm": 0.9267318266745215,
"learning_rate": 5.819310482373381e-06,
"loss": 0.1288,
"step": 299
},
{
"epoch": 0.8982035928143712,
"grad_norm": 1.0525981747008615,
"learning_rate": 5.796104437155213e-06,
"loss": 0.1875,
"step": 300
},
{
"epoch": 0.9011976047904192,
"grad_norm": 1.005727753806926,
"learning_rate": 5.772880783692363e-06,
"loss": 0.16,
"step": 301
},
{
"epoch": 0.9041916167664671,
"grad_norm": 1.0706420333824367,
"learning_rate": 5.749640035645798e-06,
"loss": 0.1848,
"step": 302
},
{
"epoch": 0.907185628742515,
"grad_norm": 1.0592052097995426,
"learning_rate": 5.726382707054578e-06,
"loss": 0.1465,
"step": 303
},
{
"epoch": 0.9101796407185628,
"grad_norm": 0.993031808407336,
"learning_rate": 5.703109312324493e-06,
"loss": 0.1736,
"step": 304
},
{
"epoch": 0.9131736526946108,
"grad_norm": 1.1590297579759776,
"learning_rate": 5.679820366216684e-06,
"loss": 0.196,
"step": 305
},
{
"epoch": 0.9161676646706587,
"grad_norm": 0.8536906647586928,
"learning_rate": 5.656516383836263e-06,
"loss": 0.1429,
"step": 306
},
{
"epoch": 0.9191616766467066,
"grad_norm": 1.0743142128668395,
"learning_rate": 5.6331978806209044e-06,
"loss": 0.2057,
"step": 307
},
{
"epoch": 0.9221556886227545,
"grad_norm": 1.0688403919841896,
"learning_rate": 5.609865372329461e-06,
"loss": 0.167,
"step": 308
},
{
"epoch": 0.9251497005988024,
"grad_norm": 1.121449299134721,
"learning_rate": 5.586519375030549e-06,
"loss": 0.1951,
"step": 309
},
{
"epoch": 0.9281437125748503,
"grad_norm": 1.005148878729297,
"learning_rate": 5.5631604050911354e-06,
"loss": 0.1773,
"step": 310
},
{
"epoch": 0.9311377245508982,
"grad_norm": 1.0849081343374762,
"learning_rate": 5.539788979165115e-06,
"loss": 0.1777,
"step": 311
},
{
"epoch": 0.9341317365269461,
"grad_norm": 0.8628134241266023,
"learning_rate": 5.516405614181883e-06,
"loss": 0.1669,
"step": 312
},
{
"epoch": 0.937125748502994,
"grad_norm": 0.9573284961930177,
"learning_rate": 5.4930108273349034e-06,
"loss": 0.1734,
"step": 313
},
{
"epoch": 0.9401197604790419,
"grad_norm": 0.9683824661704818,
"learning_rate": 5.4696051360702725e-06,
"loss": 0.1761,
"step": 314
},
{
"epoch": 0.9431137724550899,
"grad_norm": 1.2979049922982,
"learning_rate": 5.446189058075265e-06,
"loss": 0.2027,
"step": 315
},
{
"epoch": 0.9461077844311377,
"grad_norm": 1.1799048018466403,
"learning_rate": 5.4227631112668955e-06,
"loss": 0.2161,
"step": 316
},
{
"epoch": 0.9491017964071856,
"grad_norm": 1.1476919638388374,
"learning_rate": 5.39932781378045e-06,
"loss": 0.1687,
"step": 317
},
{
"epoch": 0.9520958083832335,
"grad_norm": 1.1877302474355855,
"learning_rate": 5.375883683958041e-06,
"loss": 0.2272,
"step": 318
},
{
"epoch": 0.9550898203592815,
"grad_norm": 0.9237104555803209,
"learning_rate": 5.3524312403371255e-06,
"loss": 0.1574,
"step": 319
},
{
"epoch": 0.9580838323353293,
"grad_norm": 1.1070080049213966,
"learning_rate": 5.328971001639054e-06,
"loss": 0.1902,
"step": 320
},
{
"epoch": 0.9610778443113772,
"grad_norm": 1.152762554365944,
"learning_rate": 5.3055034867575825e-06,
"loss": 0.2038,
"step": 321
},
{
"epoch": 0.9640718562874252,
"grad_norm": 1.1517767502650393,
"learning_rate": 5.282029214747404e-06,
"loss": 0.1811,
"step": 322
},
{
"epoch": 0.9670658682634731,
"grad_norm": 1.1626349166784884,
"learning_rate": 5.258548704812667e-06,
"loss": 0.1813,
"step": 323
},
{
"epoch": 0.9700598802395209,
"grad_norm": 1.178599392963165,
"learning_rate": 5.235062476295488e-06,
"loss": 0.2103,
"step": 324
},
{
"epoch": 0.9730538922155688,
"grad_norm": 1.1922317101600366,
"learning_rate": 5.211571048664469e-06,
"loss": 0.1818,
"step": 325
},
{
"epoch": 0.9760479041916168,
"grad_norm": 1.047045246847648,
"learning_rate": 5.188074941503203e-06,
"loss": 0.1748,
"step": 326
},
{
"epoch": 0.9790419161676647,
"grad_norm": 1.1908363272944777,
"learning_rate": 5.164574674498788e-06,
"loss": 0.1909,
"step": 327
},
{
"epoch": 0.9820359281437125,
"grad_norm": 1.0780471295286016,
"learning_rate": 5.141070767430331e-06,
"loss": 0.1702,
"step": 328
},
{
"epoch": 0.9850299401197605,
"grad_norm": 1.0762915319611592,
"learning_rate": 5.117563740157444e-06,
"loss": 0.1777,
"step": 329
},
{
"epoch": 0.9880239520958084,
"grad_norm": 1.2072192662885768,
"learning_rate": 5.094054112608758e-06,
"loss": 0.1833,
"step": 330
},
{
"epoch": 0.9910179640718563,
"grad_norm": 1.079654890325274,
"learning_rate": 5.070542404770413e-06,
"loss": 0.1845,
"step": 331
},
{
"epoch": 0.9940119760479041,
"grad_norm": 0.898002399587854,
"learning_rate": 5.047029136674563e-06,
"loss": 0.1575,
"step": 332
},
{
"epoch": 0.9970059880239521,
"grad_norm": 1.2704491837034548,
"learning_rate": 5.023514828387868e-06,
"loss": 0.2231,
"step": 333
},
{
"epoch": 1.0,
"grad_norm": 1.0268037499480638,
"learning_rate": 5e-06,
"loss": 0.1779,
"step": 334
},
{
"epoch": 1.0029940119760479,
"grad_norm": 1.0319220998665655,
"learning_rate": 4.976485171612134e-06,
"loss": 0.1374,
"step": 335
},
{
"epoch": 1.0059880239520957,
"grad_norm": 1.102202524357858,
"learning_rate": 4.95297086332544e-06,
"loss": 0.1375,
"step": 336
},
{
"epoch": 1.0089820359281436,
"grad_norm": 0.9157940542057146,
"learning_rate": 4.9294575952295896e-06,
"loss": 0.1277,
"step": 337
},
{
"epoch": 1.0119760479041917,
"grad_norm": 1.0724140942663891,
"learning_rate": 4.905945887391242e-06,
"loss": 0.1566,
"step": 338
},
{
"epoch": 1.0149700598802396,
"grad_norm": 0.876666631070817,
"learning_rate": 4.882436259842556e-06,
"loss": 0.1061,
"step": 339
},
{
"epoch": 1.0179640718562875,
"grad_norm": 0.9162402243151477,
"learning_rate": 4.858929232569671e-06,
"loss": 0.1261,
"step": 340
},
{
"epoch": 1.0209580838323353,
"grad_norm": 1.0543131283325944,
"learning_rate": 4.835425325501214e-06,
"loss": 0.161,
"step": 341
},
{
"epoch": 1.0239520958083832,
"grad_norm": 0.8400230970072093,
"learning_rate": 4.811925058496799e-06,
"loss": 0.1416,
"step": 342
},
{
"epoch": 1.026946107784431,
"grad_norm": 0.9829924332640012,
"learning_rate": 4.788428951335534e-06,
"loss": 0.1403,
"step": 343
},
{
"epoch": 1.029940119760479,
"grad_norm": 0.8699526492060706,
"learning_rate": 4.7649375237045135e-06,
"loss": 0.1575,
"step": 344
},
{
"epoch": 1.032934131736527,
"grad_norm": 0.7687640256714353,
"learning_rate": 4.741451295187333e-06,
"loss": 0.1141,
"step": 345
},
{
"epoch": 1.035928143712575,
"grad_norm": 0.9038256776179985,
"learning_rate": 4.717970785252596e-06,
"loss": 0.1463,
"step": 346
},
{
"epoch": 1.0389221556886228,
"grad_norm": 0.9477966708303615,
"learning_rate": 4.694496513242418e-06,
"loss": 0.1634,
"step": 347
},
{
"epoch": 1.0419161676646707,
"grad_norm": 0.8460829147104356,
"learning_rate": 4.671028998360947e-06,
"loss": 0.1081,
"step": 348
},
{
"epoch": 1.0449101796407185,
"grad_norm": 0.8332112973617621,
"learning_rate": 4.647568759662876e-06,
"loss": 0.1288,
"step": 349
},
{
"epoch": 1.0479041916167664,
"grad_norm": 0.9263063648543007,
"learning_rate": 4.624116316041962e-06,
"loss": 0.1508,
"step": 350
},
{
"epoch": 1.0508982035928143,
"grad_norm": 0.8830156295508476,
"learning_rate": 4.600672186219551e-06,
"loss": 0.1392,
"step": 351
},
{
"epoch": 1.0538922155688624,
"grad_norm": 1.1432309658692876,
"learning_rate": 4.5772368887331044e-06,
"loss": 0.1304,
"step": 352
},
{
"epoch": 1.0568862275449102,
"grad_norm": 1.1738063244656172,
"learning_rate": 4.553810941924735e-06,
"loss": 0.1975,
"step": 353
},
{
"epoch": 1.0598802395209581,
"grad_norm": 0.9041205937072112,
"learning_rate": 4.530394863929728e-06,
"loss": 0.1514,
"step": 354
},
{
"epoch": 1.062874251497006,
"grad_norm": 0.9186698905326833,
"learning_rate": 4.506989172665097e-06,
"loss": 0.1612,
"step": 355
},
{
"epoch": 1.0658682634730539,
"grad_norm": 0.8648113253450141,
"learning_rate": 4.483594385818119e-06,
"loss": 0.1058,
"step": 356
},
{
"epoch": 1.0688622754491017,
"grad_norm": 1.0799570850287572,
"learning_rate": 4.460211020834887e-06,
"loss": 0.1453,
"step": 357
},
{
"epoch": 1.0718562874251496,
"grad_norm": 0.9907906529336925,
"learning_rate": 4.436839594908866e-06,
"loss": 0.1703,
"step": 358
},
{
"epoch": 1.0748502994011977,
"grad_norm": 0.9238171937983599,
"learning_rate": 4.4134806249694514e-06,
"loss": 0.1409,
"step": 359
},
{
"epoch": 1.0778443113772456,
"grad_norm": 1.1976462413655093,
"learning_rate": 4.39013462767054e-06,
"loss": 0.2147,
"step": 360
},
{
"epoch": 1.0808383233532934,
"grad_norm": 0.844379209933977,
"learning_rate": 4.366802119379098e-06,
"loss": 0.1258,
"step": 361
},
{
"epoch": 1.0838323353293413,
"grad_norm": 0.9814017973361157,
"learning_rate": 4.34348361616374e-06,
"loss": 0.1526,
"step": 362
},
{
"epoch": 1.0868263473053892,
"grad_norm": 1.0683220931878272,
"learning_rate": 4.3201796337833165e-06,
"loss": 0.1367,
"step": 363
},
{
"epoch": 1.089820359281437,
"grad_norm": 1.0219893023672373,
"learning_rate": 4.29689068767551e-06,
"loss": 0.1283,
"step": 364
},
{
"epoch": 1.092814371257485,
"grad_norm": 0.8955146565557671,
"learning_rate": 4.273617292945425e-06,
"loss": 0.1414,
"step": 365
},
{
"epoch": 1.095808383233533,
"grad_norm": 0.883401633858211,
"learning_rate": 4.250359964354203e-06,
"loss": 0.1294,
"step": 366
},
{
"epoch": 1.098802395209581,
"grad_norm": 0.8660420575857484,
"learning_rate": 4.227119216307637e-06,
"loss": 0.1207,
"step": 367
},
{
"epoch": 1.1017964071856288,
"grad_norm": 0.9156496693325334,
"learning_rate": 4.203895562844789e-06,
"loss": 0.1387,
"step": 368
},
{
"epoch": 1.1047904191616766,
"grad_norm": 0.9674918812464737,
"learning_rate": 4.18068951762662e-06,
"loss": 0.1378,
"step": 369
},
{
"epoch": 1.1077844311377245,
"grad_norm": 0.8592997789835183,
"learning_rate": 4.157501593924638e-06,
"loss": 0.1352,
"step": 370
},
{
"epoch": 1.1107784431137724,
"grad_norm": 1.0182328766381579,
"learning_rate": 4.134332304609533e-06,
"loss": 0.1845,
"step": 371
},
{
"epoch": 1.1137724550898203,
"grad_norm": 0.8497981844959756,
"learning_rate": 4.111182162139844e-06,
"loss": 0.1298,
"step": 372
},
{
"epoch": 1.1167664670658684,
"grad_norm": 0.8832443389536418,
"learning_rate": 4.088051678550617e-06,
"loss": 0.1142,
"step": 373
},
{
"epoch": 1.1197604790419162,
"grad_norm": 0.876236726893519,
"learning_rate": 4.064941365442084e-06,
"loss": 0.1397,
"step": 374
},
{
"epoch": 1.122754491017964,
"grad_norm": 1.044940461950853,
"learning_rate": 4.041851733968348e-06,
"loss": 0.1581,
"step": 375
},
{
"epoch": 1.125748502994012,
"grad_norm": 1.1118936393642198,
"learning_rate": 4.018783294826071e-06,
"loss": 0.1436,
"step": 376
},
{
"epoch": 1.1287425149700598,
"grad_norm": 1.1388063623052478,
"learning_rate": 3.995736558243186e-06,
"loss": 0.1294,
"step": 377
},
{
"epoch": 1.1317365269461077,
"grad_norm": 1.0801467704154517,
"learning_rate": 3.972712033967608e-06,
"loss": 0.1923,
"step": 378
},
{
"epoch": 1.1347305389221556,
"grad_norm": 0.8523632905220382,
"learning_rate": 3.949710231255961e-06,
"loss": 0.171,
"step": 379
},
{
"epoch": 1.1377245508982037,
"grad_norm": 1.0338939052079452,
"learning_rate": 3.926731658862307e-06,
"loss": 0.1477,
"step": 380
},
{
"epoch": 1.1407185628742516,
"grad_norm": 0.8416712903160954,
"learning_rate": 3.903776825026912e-06,
"loss": 0.1518,
"step": 381
},
{
"epoch": 1.1437125748502994,
"grad_norm": 0.8897392386034587,
"learning_rate": 3.8808462374649805e-06,
"loss": 0.1622,
"step": 382
},
{
"epoch": 1.1467065868263473,
"grad_norm": 0.8834715663977217,
"learning_rate": 3.857940403355444e-06,
"loss": 0.1323,
"step": 383
},
{
"epoch": 1.1497005988023952,
"grad_norm": 0.9869457235753144,
"learning_rate": 3.8350598293297345e-06,
"loss": 0.1884,
"step": 384
},
{
"epoch": 1.152694610778443,
"grad_norm": 1.023564737107474,
"learning_rate": 3.8122050214605822e-06,
"loss": 0.1572,
"step": 385
},
{
"epoch": 1.1556886227544911,
"grad_norm": 1.09593993637393,
"learning_rate": 3.7893764852508207e-06,
"loss": 0.1976,
"step": 386
},
{
"epoch": 1.158682634730539,
"grad_norm": 0.8367385226849327,
"learning_rate": 3.766574725622208e-06,
"loss": 0.1188,
"step": 387
},
{
"epoch": 1.1616766467065869,
"grad_norm": 0.9628427700146952,
"learning_rate": 3.7438002469042567e-06,
"loss": 0.1387,
"step": 388
},
{
"epoch": 1.1646706586826348,
"grad_norm": 1.0525030518761778,
"learning_rate": 3.721053552823078e-06,
"loss": 0.17,
"step": 389
},
{
"epoch": 1.1676646706586826,
"grad_norm": 0.9140459323847402,
"learning_rate": 3.698335146490246e-06,
"loss": 0.1383,
"step": 390
},
{
"epoch": 1.1706586826347305,
"grad_norm": 0.9468999096238121,
"learning_rate": 3.675645530391665e-06,
"loss": 0.1265,
"step": 391
},
{
"epoch": 1.1736526946107784,
"grad_norm": 0.9508965162103243,
"learning_rate": 3.652985206376455e-06,
"loss": 0.1435,
"step": 392
},
{
"epoch": 1.1766467065868262,
"grad_norm": 0.9083768420099183,
"learning_rate": 3.630354675645853e-06,
"loss": 0.1496,
"step": 393
},
{
"epoch": 1.1796407185628743,
"grad_norm": 0.9325125863887872,
"learning_rate": 3.6077544387421293e-06,
"loss": 0.1305,
"step": 394
},
{
"epoch": 1.1826347305389222,
"grad_norm": 0.8146851998801456,
"learning_rate": 3.5851849955375177e-06,
"loss": 0.1117,
"step": 395
},
{
"epoch": 1.18562874251497,
"grad_norm": 0.8658716514573122,
"learning_rate": 3.5626468452231534e-06,
"loss": 0.1401,
"step": 396
},
{
"epoch": 1.188622754491018,
"grad_norm": 0.8505641982515318,
"learning_rate": 3.540140486298035e-06,
"loss": 0.1316,
"step": 397
},
{
"epoch": 1.1916167664670658,
"grad_norm": 0.9521633193671332,
"learning_rate": 3.517666416557999e-06,
"loss": 0.1713,
"step": 398
},
{
"epoch": 1.1946107784431137,
"grad_norm": 0.9342440052646195,
"learning_rate": 3.495225133084712e-06,
"loss": 0.1305,
"step": 399
},
{
"epoch": 1.1976047904191618,
"grad_norm": 0.812420007817787,
"learning_rate": 3.472817132234669e-06,
"loss": 0.1362,
"step": 400
},
{
"epoch": 1.1976047904191618,
"eval_loss": 0.18990673124790192,
"eval_runtime": 2.0102,
"eval_samples_per_second": 26.862,
"eval_steps_per_second": 6.964,
"step": 400
},
{
"epoch": 1.2005988023952097,
"grad_norm": 0.9036526843804629,
"learning_rate": 3.4504429096282246e-06,
"loss": 0.1597,
"step": 401
},
{
"epoch": 1.2035928143712575,
"grad_norm": 0.8784455144407392,
"learning_rate": 3.428102960138625e-06,
"loss": 0.1511,
"step": 402
},
{
"epoch": 1.2065868263473054,
"grad_norm": 1.028429520772402,
"learning_rate": 3.405797777881059e-06,
"loss": 0.1604,
"step": 403
},
{
"epoch": 1.2095808383233533,
"grad_norm": 0.855369863530219,
"learning_rate": 3.3835278562017405e-06,
"loss": 0.1295,
"step": 404
},
{
"epoch": 1.2125748502994012,
"grad_norm": 0.8008225949523208,
"learning_rate": 3.3612936876669834e-06,
"loss": 0.1348,
"step": 405
},
{
"epoch": 1.215568862275449,
"grad_norm": 0.9043922805777009,
"learning_rate": 3.3390957640523147e-06,
"loss": 0.1652,
"step": 406
},
{
"epoch": 1.218562874251497,
"grad_norm": 0.8626058635148262,
"learning_rate": 3.3169345763315986e-06,
"loss": 0.1316,
"step": 407
},
{
"epoch": 1.221556886227545,
"grad_norm": 0.904751440459687,
"learning_rate": 3.29481061466617e-06,
"loss": 0.1558,
"step": 408
},
{
"epoch": 1.2245508982035929,
"grad_norm": 0.9973356317264691,
"learning_rate": 3.2727243683940045e-06,
"loss": 0.1537,
"step": 409
},
{
"epoch": 1.2275449101796407,
"grad_norm": 0.8921887769203182,
"learning_rate": 3.2506763260188824e-06,
"loss": 0.1369,
"step": 410
},
{
"epoch": 1.2305389221556886,
"grad_norm": 0.9067647769697026,
"learning_rate": 3.2286669751995905e-06,
"loss": 0.1274,
"step": 411
},
{
"epoch": 1.2335329341317365,
"grad_norm": 1.0265791747283641,
"learning_rate": 3.2066968027391377e-06,
"loss": 0.1354,
"step": 412
},
{
"epoch": 1.2365269461077844,
"grad_norm": 0.8390241757533995,
"learning_rate": 3.1847662945739833e-06,
"loss": 0.1354,
"step": 413
},
{
"epoch": 1.2395209580838324,
"grad_norm": 0.8909516496612402,
"learning_rate": 3.1628759357632943e-06,
"loss": 0.1368,
"step": 414
},
{
"epoch": 1.2425149700598803,
"grad_norm": 0.8642916153882426,
"learning_rate": 3.1410262104782086e-06,
"loss": 0.1208,
"step": 415
},
{
"epoch": 1.2455089820359282,
"grad_norm": 1.0032850850047146,
"learning_rate": 3.119217601991139e-06,
"loss": 0.1687,
"step": 416
},
{
"epoch": 1.248502994011976,
"grad_norm": 0.9310586092515343,
"learning_rate": 3.0974505926650724e-06,
"loss": 0.1253,
"step": 417
},
{
"epoch": 1.251497005988024,
"grad_norm": 0.8457067715777292,
"learning_rate": 3.0757256639429027e-06,
"loss": 0.1402,
"step": 418
},
{
"epoch": 1.2544910179640718,
"grad_norm": 0.9109407739477747,
"learning_rate": 3.0540432963367907e-06,
"loss": 0.1499,
"step": 419
},
{
"epoch": 1.2574850299401197,
"grad_norm": 0.8596469212810119,
"learning_rate": 3.032403969417523e-06,
"loss": 0.1215,
"step": 420
},
{
"epoch": 1.2604790419161676,
"grad_norm": 0.7373811256512488,
"learning_rate": 3.010808161803917e-06,
"loss": 0.1368,
"step": 421
},
{
"epoch": 1.2634730538922156,
"grad_norm": 0.993658813416458,
"learning_rate": 2.9892563511522305e-06,
"loss": 0.1699,
"step": 422
},
{
"epoch": 1.2664670658682635,
"grad_norm": 0.9201549320501152,
"learning_rate": 2.9677490141455915e-06,
"loss": 0.1476,
"step": 423
},
{
"epoch": 1.2694610778443114,
"grad_norm": 1.0571479202805087,
"learning_rate": 2.946286626483463e-06,
"loss": 0.1903,
"step": 424
},
{
"epoch": 1.2724550898203593,
"grad_norm": 0.9519915245178604,
"learning_rate": 2.924869662871117e-06,
"loss": 0.1399,
"step": 425
},
{
"epoch": 1.2754491017964071,
"grad_norm": 0.9079781213033773,
"learning_rate": 2.903498597009136e-06,
"loss": 0.1591,
"step": 426
},
{
"epoch": 1.278443113772455,
"grad_norm": 0.9246737433477424,
"learning_rate": 2.8821739015829338e-06,
"loss": 0.133,
"step": 427
},
{
"epoch": 1.281437125748503,
"grad_norm": 0.7363248463941603,
"learning_rate": 2.8608960482523058e-06,
"loss": 0.1124,
"step": 428
},
{
"epoch": 1.284431137724551,
"grad_norm": 0.8737016446025797,
"learning_rate": 2.839665507640992e-06,
"loss": 0.1272,
"step": 429
},
{
"epoch": 1.2874251497005988,
"grad_norm": 0.9073164248412204,
"learning_rate": 2.818482749326272e-06,
"loss": 0.1351,
"step": 430
},
{
"epoch": 1.2904191616766467,
"grad_norm": 0.8684495772464258,
"learning_rate": 2.797348241828569e-06,
"loss": 0.1302,
"step": 431
},
{
"epoch": 1.2934131736526946,
"grad_norm": 0.8538642008415324,
"learning_rate": 2.776262452601104e-06,
"loss": 0.1465,
"step": 432
},
{
"epoch": 1.2964071856287425,
"grad_norm": 0.8448195193131449,
"learning_rate": 2.7552258480195348e-06,
"loss": 0.125,
"step": 433
},
{
"epoch": 1.2994011976047903,
"grad_norm": 0.8447034301042033,
"learning_rate": 2.734238893371667e-06,
"loss": 0.1239,
"step": 434
},
{
"epoch": 1.3023952095808382,
"grad_norm": 1.0097602680081321,
"learning_rate": 2.7133020528471322e-06,
"loss": 0.1491,
"step": 435
},
{
"epoch": 1.3053892215568863,
"grad_norm": 0.8889458463846382,
"learning_rate": 2.6924157895271563e-06,
"loss": 0.1447,
"step": 436
},
{
"epoch": 1.3083832335329342,
"grad_norm": 0.9066059341360602,
"learning_rate": 2.671580565374282e-06,
"loss": 0.1501,
"step": 437
},
{
"epoch": 1.311377245508982,
"grad_norm": 0.9117154834445957,
"learning_rate": 2.6507968412221763e-06,
"loss": 0.1266,
"step": 438
},
{
"epoch": 1.31437125748503,
"grad_norm": 0.9720832980163194,
"learning_rate": 2.6300650767654234e-06,
"loss": 0.145,
"step": 439
},
{
"epoch": 1.3173652694610778,
"grad_norm": 0.9622333749513002,
"learning_rate": 2.6093857305493666e-06,
"loss": 0.1384,
"step": 440
},
{
"epoch": 1.3203592814371259,
"grad_norm": 0.9740785035569228,
"learning_rate": 2.588759259959962e-06,
"loss": 0.1441,
"step": 441
},
{
"epoch": 1.3233532934131738,
"grad_norm": 0.8739974395467459,
"learning_rate": 2.568186121213658e-06,
"loss": 0.1441,
"step": 442
},
{
"epoch": 1.3263473053892216,
"grad_norm": 0.9233098797099563,
"learning_rate": 2.547666769347312e-06,
"loss": 0.1659,
"step": 443
},
{
"epoch": 1.3293413173652695,
"grad_norm": 1.0411078104986013,
"learning_rate": 2.5272016582081236e-06,
"loss": 0.1574,
"step": 444
},
{
"epoch": 1.3323353293413174,
"grad_norm": 0.9114250858380806,
"learning_rate": 2.5067912404435952e-06,
"loss": 0.1516,
"step": 445
},
{
"epoch": 1.3353293413173652,
"grad_norm": 0.9173880915766272,
"learning_rate": 2.486435967491516e-06,
"loss": 0.1397,
"step": 446
},
{
"epoch": 1.3383233532934131,
"grad_norm": 0.9090437228226232,
"learning_rate": 2.4661362895699903e-06,
"loss": 0.1549,
"step": 447
},
{
"epoch": 1.341317365269461,
"grad_norm": 1.1087775532758153,
"learning_rate": 2.445892655667462e-06,
"loss": 0.2068,
"step": 448
},
{
"epoch": 1.3443113772455089,
"grad_norm": 1.0192756372847187,
"learning_rate": 2.425705513532798e-06,
"loss": 0.1839,
"step": 449
},
{
"epoch": 1.347305389221557,
"grad_norm": 0.9578542967453554,
"learning_rate": 2.4055753096653795e-06,
"loss": 0.1539,
"step": 450
},
{
"epoch": 1.3502994011976048,
"grad_norm": 1.0125811517800805,
"learning_rate": 2.3855024893052286e-06,
"loss": 0.1421,
"step": 451
},
{
"epoch": 1.3532934131736527,
"grad_norm": 0.9357901533850678,
"learning_rate": 2.365487496423152e-06,
"loss": 0.1534,
"step": 452
},
{
"epoch": 1.3562874251497006,
"grad_norm": 0.8562121683243458,
"learning_rate": 2.3455307737109338e-06,
"loss": 0.129,
"step": 453
},
{
"epoch": 1.3592814371257484,
"grad_norm": 0.8472835502962511,
"learning_rate": 2.3256327625715345e-06,
"loss": 0.1365,
"step": 454
},
{
"epoch": 1.3622754491017965,
"grad_norm": 0.8562839203189296,
"learning_rate": 2.3057939031093346e-06,
"loss": 0.1177,
"step": 455
},
{
"epoch": 1.3652694610778444,
"grad_norm": 0.7767264502752892,
"learning_rate": 2.2860146341203936e-06,
"loss": 0.1228,
"step": 456
},
{
"epoch": 1.3682634730538923,
"grad_norm": 0.9433584734149557,
"learning_rate": 2.2662953930827546e-06,
"loss": 0.1627,
"step": 457
},
{
"epoch": 1.3712574850299402,
"grad_norm": 0.8926549461701841,
"learning_rate": 2.2466366161467528e-06,
"loss": 0.1249,
"step": 458
},
{
"epoch": 1.374251497005988,
"grad_norm": 0.9473466818324262,
"learning_rate": 2.227038738125385e-06,
"loss": 0.1169,
"step": 459
},
{
"epoch": 1.377245508982036,
"grad_norm": 1.0828071363920466,
"learning_rate": 2.207502192484685e-06,
"loss": 0.148,
"step": 460
},
{
"epoch": 1.3802395209580838,
"grad_norm": 0.859522314927212,
"learning_rate": 2.188027411334131e-06,
"loss": 0.1323,
"step": 461
},
{
"epoch": 1.3832335329341316,
"grad_norm": 0.9639874215918843,
"learning_rate": 2.1686148254171012e-06,
"loss": 0.1385,
"step": 462
},
{
"epoch": 1.3862275449101795,
"grad_norm": 0.9345939401340367,
"learning_rate": 2.1492648641013305e-06,
"loss": 0.1178,
"step": 463
},
{
"epoch": 1.3892215568862276,
"grad_norm": 0.7921932148009695,
"learning_rate": 2.1299779553694323e-06,
"loss": 0.1171,
"step": 464
},
{
"epoch": 1.3922155688622755,
"grad_norm": 1.0039646804861526,
"learning_rate": 2.1107545258094135e-06,
"loss": 0.1467,
"step": 465
},
{
"epoch": 1.3952095808383234,
"grad_norm": 0.8809020178269484,
"learning_rate": 2.0915950006052555e-06,
"loss": 0.1215,
"step": 466
},
{
"epoch": 1.3982035928143712,
"grad_norm": 0.8809452454727839,
"learning_rate": 2.0724998035274947e-06,
"loss": 0.1333,
"step": 467
},
{
"epoch": 1.401197604790419,
"grad_norm": 0.7797325243843375,
"learning_rate": 2.053469356923865e-06,
"loss": 0.1246,
"step": 468
},
{
"epoch": 1.4041916167664672,
"grad_norm": 0.927565519591365,
"learning_rate": 2.0345040817099433e-06,
"loss": 0.1434,
"step": 469
},
{
"epoch": 1.407185628742515,
"grad_norm": 0.784259516906681,
"learning_rate": 2.0156043973598475e-06,
"loss": 0.1469,
"step": 470
},
{
"epoch": 1.410179640718563,
"grad_norm": 1.0748175357596759,
"learning_rate": 1.996770721896957e-06,
"loss": 0.1462,
"step": 471
},
{
"epoch": 1.4131736526946108,
"grad_norm": 0.8253260762972392,
"learning_rate": 1.9780034718846653e-06,
"loss": 0.1112,
"step": 472
},
{
"epoch": 1.4161676646706587,
"grad_norm": 1.040360283072913,
"learning_rate": 1.9593030624171683e-06,
"loss": 0.1679,
"step": 473
},
{
"epoch": 1.4191616766467066,
"grad_norm": 0.927960553923833,
"learning_rate": 1.9406699071102774e-06,
"loss": 0.1389,
"step": 474
},
{
"epoch": 1.4221556886227544,
"grad_norm": 0.9207738510423578,
"learning_rate": 1.9221044180922833e-06,
"loss": 0.1305,
"step": 475
},
{
"epoch": 1.4251497005988023,
"grad_norm": 0.9244736619519355,
"learning_rate": 1.9036070059948253e-06,
"loss": 0.1499,
"step": 476
},
{
"epoch": 1.4281437125748502,
"grad_norm": 1.1643798267780687,
"learning_rate": 1.885178079943823e-06,
"loss": 0.153,
"step": 477
},
{
"epoch": 1.4311377245508983,
"grad_norm": 1.0152074632299224,
"learning_rate": 1.866818047550419e-06,
"loss": 0.1836,
"step": 478
},
{
"epoch": 1.4341317365269461,
"grad_norm": 0.8694819347979745,
"learning_rate": 1.8485273149019655e-06,
"loss": 0.1427,
"step": 479
},
{
"epoch": 1.437125748502994,
"grad_norm": 0.8368507029042666,
"learning_rate": 1.8303062865530407e-06,
"loss": 0.1141,
"step": 480
},
{
"epoch": 1.4401197604790419,
"grad_norm": 0.8762331129302771,
"learning_rate": 1.8121553655165058e-06,
"loss": 0.116,
"step": 481
},
{
"epoch": 1.4431137724550898,
"grad_norm": 1.3190796774306477,
"learning_rate": 1.7940749532545832e-06,
"loss": 0.1546,
"step": 482
},
{
"epoch": 1.4461077844311379,
"grad_norm": 0.9869523118055206,
"learning_rate": 1.7760654496699876e-06,
"loss": 0.1406,
"step": 483
},
{
"epoch": 1.4491017964071857,
"grad_norm": 1.0006141912867437,
"learning_rate": 1.7581272530970666e-06,
"loss": 0.1507,
"step": 484
},
{
"epoch": 1.4520958083832336,
"grad_norm": 0.8557120591249597,
"learning_rate": 1.7402607602930106e-06,
"loss": 0.1289,
"step": 485
},
{
"epoch": 1.4550898203592815,
"grad_norm": 0.9304840741542314,
"learning_rate": 1.7224663664290537e-06,
"loss": 0.1396,
"step": 486
},
{
"epoch": 1.4580838323353293,
"grad_norm": 0.913596669946821,
"learning_rate": 1.7047444650817518e-06,
"loss": 0.1478,
"step": 487
},
{
"epoch": 1.4610778443113772,
"grad_norm": 0.8624677964042632,
"learning_rate": 1.6870954482242707e-06,
"loss": 0.1326,
"step": 488
},
{
"epoch": 1.464071856287425,
"grad_norm": 0.9034506197857363,
"learning_rate": 1.669519706217711e-06,
"loss": 0.1376,
"step": 489
},
{
"epoch": 1.467065868263473,
"grad_norm": 1.0420363818885627,
"learning_rate": 1.652017627802487e-06,
"loss": 0.1558,
"step": 490
},
{
"epoch": 1.4700598802395208,
"grad_norm": 0.9398992963413748,
"learning_rate": 1.6345896000897122e-06,
"loss": 0.143,
"step": 491
},
{
"epoch": 1.473053892215569,
"grad_norm": 0.8976180093625737,
"learning_rate": 1.6172360085526567e-06,
"loss": 0.1462,
"step": 492
},
{
"epoch": 1.4760479041916168,
"grad_norm": 0.9285802987980569,
"learning_rate": 1.5999572370182016e-06,
"loss": 0.1719,
"step": 493
},
{
"epoch": 1.4790419161676647,
"grad_norm": 0.9788042388747751,
"learning_rate": 1.5827536676583643e-06,
"loss": 0.1431,
"step": 494
},
{
"epoch": 1.4820359281437125,
"grad_norm": 0.9439418038609391,
"learning_rate": 1.5656256809818343e-06,
"loss": 0.1525,
"step": 495
},
{
"epoch": 1.4850299401197604,
"grad_norm": 0.9644567735451506,
"learning_rate": 1.54857365582557e-06,
"loss": 0.1391,
"step": 496
},
{
"epoch": 1.4880239520958085,
"grad_norm": 1.0099126681219934,
"learning_rate": 1.5315979693464039e-06,
"loss": 0.164,
"step": 497
},
{
"epoch": 1.4910179640718564,
"grad_norm": 0.992965295723431,
"learning_rate": 1.5146989970127158e-06,
"loss": 0.1385,
"step": 498
},
{
"epoch": 1.4940119760479043,
"grad_norm": 1.083717947030965,
"learning_rate": 1.4978771125961177e-06,
"loss": 0.1662,
"step": 499
},
{
"epoch": 1.4970059880239521,
"grad_norm": 0.8583481848212289,
"learning_rate": 1.4811326881631937e-06,
"loss": 0.1436,
"step": 500
},
{
"epoch": 1.5,
"grad_norm": 0.9217710894059078,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.1566,
"step": 501
},
{
"epoch": 1.5029940119760479,
"grad_norm": 1.0266018415213711,
"learning_rate": 1.4478776989401949e-06,
"loss": 0.1546,
"step": 502
},
{
"epoch": 1.5059880239520957,
"grad_norm": 1.1038168478543808,
"learning_rate": 1.4313678696842559e-06,
"loss": 0.1665,
"step": 503
},
{
"epoch": 1.5089820359281436,
"grad_norm": 0.8829638226484359,
"learning_rate": 1.4149369714639856e-06,
"loss": 0.1221,
"step": 504
},
{
"epoch": 1.5119760479041915,
"grad_norm": 1.0565582201869181,
"learning_rate": 1.3985853676981316e-06,
"loss": 0.1635,
"step": 505
},
{
"epoch": 1.5149700598802394,
"grad_norm": 0.885100353857755,
"learning_rate": 1.3823134200516043e-06,
"loss": 0.1095,
"step": 506
},
{
"epoch": 1.5179640718562875,
"grad_norm": 0.8616172347072097,
"learning_rate": 1.366121488427481e-06,
"loss": 0.1343,
"step": 507
},
{
"epoch": 1.5209580838323353,
"grad_norm": 0.8915416546311651,
"learning_rate": 1.3500099309590397e-06,
"loss": 0.128,
"step": 508
},
{
"epoch": 1.5239520958083832,
"grad_norm": 0.8874014809598043,
"learning_rate": 1.3339791040018479e-06,
"loss": 0.1235,
"step": 509
},
{
"epoch": 1.5269461077844313,
"grad_norm": 0.9144473328770585,
"learning_rate": 1.3180293621258694e-06,
"loss": 0.1603,
"step": 510
},
{
"epoch": 1.5299401197604792,
"grad_norm": 0.8822056486714703,
"learning_rate": 1.3021610581076316e-06,
"loss": 0.1279,
"step": 511
},
{
"epoch": 1.532934131736527,
"grad_norm": 0.9091388048715009,
"learning_rate": 1.2863745429224145e-06,
"loss": 0.1163,
"step": 512
},
{
"epoch": 1.535928143712575,
"grad_norm": 0.8189063710243679,
"learning_rate": 1.270670165736499e-06,
"loss": 0.1235,
"step": 513
},
{
"epoch": 1.5389221556886228,
"grad_norm": 1.108896487433323,
"learning_rate": 1.2550482738994284e-06,
"loss": 0.1769,
"step": 514
},
{
"epoch": 1.5419161676646707,
"grad_norm": 0.9363999910581178,
"learning_rate": 1.239509212936343e-06,
"loss": 0.1555,
"step": 515
},
{
"epoch": 1.5449101796407185,
"grad_norm": 0.9367751591261786,
"learning_rate": 1.22405332654032e-06,
"loss": 0.1386,
"step": 516
},
{
"epoch": 1.5479041916167664,
"grad_norm": 0.9110865106945877,
"learning_rate": 1.2086809565647877e-06,
"loss": 0.1895,
"step": 517
},
{
"epoch": 1.5508982035928143,
"grad_norm": 0.8718949492749408,
"learning_rate": 1.1933924430159571e-06,
"loss": 0.1227,
"step": 518
},
{
"epoch": 1.5538922155688621,
"grad_norm": 0.9477945708244115,
"learning_rate": 1.1781881240452958e-06,
"loss": 0.1405,
"step": 519
},
{
"epoch": 1.55688622754491,
"grad_norm": 0.9504895992716209,
"learning_rate": 1.1630683359420653e-06,
"loss": 0.1441,
"step": 520
},
{
"epoch": 1.5598802395209581,
"grad_norm": 0.8177673304195049,
"learning_rate": 1.1480334131258626e-06,
"loss": 0.1223,
"step": 521
},
{
"epoch": 1.562874251497006,
"grad_norm": 0.9973484386670456,
"learning_rate": 1.1330836881392405e-06,
"loss": 0.1433,
"step": 522
},
{
"epoch": 1.5658682634730539,
"grad_norm": 0.8602521224063562,
"learning_rate": 1.11821949164034e-06,
"loss": 0.1373,
"step": 523
},
{
"epoch": 1.568862275449102,
"grad_norm": 0.836916410504276,
"learning_rate": 1.103441152395588e-06,
"loss": 0.1305,
"step": 524
},
{
"epoch": 1.5718562874251498,
"grad_norm": 0.9490968553667904,
"learning_rate": 1.088748997272414e-06,
"loss": 0.1359,
"step": 525
},
{
"epoch": 1.5748502994011977,
"grad_norm": 0.8982543962139155,
"learning_rate": 1.0741433512320316e-06,
"loss": 0.1175,
"step": 526
},
{
"epoch": 1.5778443113772456,
"grad_norm": 0.8654436838274447,
"learning_rate": 1.0596245373222424e-06,
"loss": 0.1111,
"step": 527
},
{
"epoch": 1.5808383233532934,
"grad_norm": 0.9386912551518253,
"learning_rate": 1.045192876670298e-06,
"loss": 0.141,
"step": 528
},
{
"epoch": 1.5838323353293413,
"grad_norm": 1.0950716057753231,
"learning_rate": 1.0308486884757868e-06,
"loss": 0.1646,
"step": 529
},
{
"epoch": 1.5868263473053892,
"grad_norm": 0.8996026041376449,
"learning_rate": 1.0165922900035886e-06,
"loss": 0.1357,
"step": 530
},
{
"epoch": 1.589820359281437,
"grad_norm": 0.8003223845109205,
"learning_rate": 1.0024239965768417e-06,
"loss": 0.1273,
"step": 531
},
{
"epoch": 1.592814371257485,
"grad_norm": 0.8311468183467691,
"learning_rate": 9.883441215699824e-07,
"loss": 0.1005,
"step": 532
},
{
"epoch": 1.5958083832335328,
"grad_norm": 1.1196062264777686,
"learning_rate": 9.74352976401805e-07,
"loss": 0.1649,
"step": 533
},
{
"epoch": 1.5988023952095807,
"grad_norm": 1.018680320296977,
"learning_rate": 9.604508705285765e-07,
"loss": 0.1482,
"step": 534
},
{
"epoch": 1.6017964071856288,
"grad_norm": 0.8978055051446752,
"learning_rate": 9.466381114371942e-07,
"loss": 0.1563,
"step": 535
},
{
"epoch": 1.6047904191616766,
"grad_norm": 1.007445640213005,
"learning_rate": 9.329150046383773e-07,
"loss": 0.1332,
"step": 536
},
{
"epoch": 1.6077844311377245,
"grad_norm": 0.8400024646244197,
"learning_rate": 9.192818536599213e-07,
"loss": 0.1492,
"step": 537
},
{
"epoch": 1.6107784431137726,
"grad_norm": 0.969414618915732,
"learning_rate": 9.057389600399719e-07,
"loss": 0.155,
"step": 538
},
{
"epoch": 1.6137724550898205,
"grad_norm": 0.9730156074624097,
"learning_rate": 8.922866233203681e-07,
"loss": 0.1611,
"step": 539
},
{
"epoch": 1.6167664670658684,
"grad_norm": 0.8248753397895976,
"learning_rate": 8.789251410400024e-07,
"loss": 0.1217,
"step": 540
},
{
"epoch": 1.6197604790419162,
"grad_norm": 0.8177957508782012,
"learning_rate": 8.65654808728259e-07,
"loss": 0.1656,
"step": 541
},
{
"epoch": 1.622754491017964,
"grad_norm": 0.9180206259631332,
"learning_rate": 8.524759198984567e-07,
"loss": 0.1227,
"step": 542
},
{
"epoch": 1.625748502994012,
"grad_norm": 1.0154922840197738,
"learning_rate": 8.393887660413719e-07,
"loss": 0.1573,
"step": 543
},
{
"epoch": 1.6287425149700598,
"grad_norm": 1.1505004357556825,
"learning_rate": 8.263936366187825e-07,
"loss": 0.1882,
"step": 544
},
{
"epoch": 1.6317365269461077,
"grad_norm": 0.9610299471026473,
"learning_rate": 8.134908190570723e-07,
"loss": 0.1579,
"step": 545
},
{
"epoch": 1.6347305389221556,
"grad_norm": 0.979356430095458,
"learning_rate": 8.006805987408705e-07,
"loss": 0.1544,
"step": 546
},
{
"epoch": 1.6377245508982035,
"grad_norm": 0.8469806565060088,
"learning_rate": 7.879632590067354e-07,
"loss": 0.1305,
"step": 547
},
{
"epoch": 1.6407185628742516,
"grad_norm": 0.9166835127825786,
"learning_rate": 7.753390811368972e-07,
"loss": 0.1427,
"step": 548
},
{
"epoch": 1.6437125748502994,
"grad_norm": 1.049703175355985,
"learning_rate": 7.628083443530287e-07,
"loss": 0.1721,
"step": 549
},
{
"epoch": 1.6467065868263473,
"grad_norm": 1.0190431397338302,
"learning_rate": 7.503713258100726e-07,
"loss": 0.1565,
"step": 550
},
{
"epoch": 1.6497005988023952,
"grad_norm": 1.0255564163331858,
"learning_rate": 7.380283005901084e-07,
"loss": 0.1659,
"step": 551
},
{
"epoch": 1.6526946107784433,
"grad_norm": 0.896660050977769,
"learning_rate": 7.257795416962754e-07,
"loss": 0.1329,
"step": 552
},
{
"epoch": 1.6556886227544911,
"grad_norm": 0.8177809025371469,
"learning_rate": 7.136253200467231e-07,
"loss": 0.0998,
"step": 553
},
{
"epoch": 1.658682634730539,
"grad_norm": 0.8839376270997219,
"learning_rate": 7.015659044686307e-07,
"loss": 0.1305,
"step": 554
},
{
"epoch": 1.6616766467065869,
"grad_norm": 0.9731086244835102,
"learning_rate": 6.896015616922535e-07,
"loss": 0.1442,
"step": 555
},
{
"epoch": 1.6646706586826348,
"grad_norm": 0.9435418608225515,
"learning_rate": 6.777325563450282e-07,
"loss": 0.1532,
"step": 556
},
{
"epoch": 1.6676646706586826,
"grad_norm": 1.0167015749155883,
"learning_rate": 6.659591509457125e-07,
"loss": 0.1679,
"step": 557
},
{
"epoch": 1.6706586826347305,
"grad_norm": 0.9154997366355084,
"learning_rate": 6.542816058985896e-07,
"loss": 0.139,
"step": 558
},
{
"epoch": 1.6736526946107784,
"grad_norm": 1.0074081226147986,
"learning_rate": 6.427001794876974e-07,
"loss": 0.1738,
"step": 559
},
{
"epoch": 1.6766467065868262,
"grad_norm": 0.8669489889728673,
"learning_rate": 6.312151278711237e-07,
"loss": 0.1173,
"step": 560
},
{
"epoch": 1.6796407185628741,
"grad_norm": 0.8177874520766348,
"learning_rate": 6.198267050753387e-07,
"loss": 0.1296,
"step": 561
},
{
"epoch": 1.6826347305389222,
"grad_norm": 0.9084594961113138,
"learning_rate": 6.085351629895736e-07,
"loss": 0.1465,
"step": 562
},
{
"epoch": 1.68562874251497,
"grad_norm": 0.9130732004115695,
"learning_rate": 5.973407513602514e-07,
"loss": 0.1166,
"step": 563
},
{
"epoch": 1.688622754491018,
"grad_norm": 1.0119910928321039,
"learning_rate": 5.862437177854629e-07,
"loss": 0.1728,
"step": 564
},
{
"epoch": 1.6916167664670658,
"grad_norm": 0.8616491415321033,
"learning_rate": 5.752443077094927e-07,
"loss": 0.1232,
"step": 565
},
{
"epoch": 1.694610778443114,
"grad_norm": 0.9826174711971898,
"learning_rate": 5.643427644173838e-07,
"loss": 0.1667,
"step": 566
},
{
"epoch": 1.6976047904191618,
"grad_norm": 0.9137857756605849,
"learning_rate": 5.535393290295643e-07,
"loss": 0.1591,
"step": 567
},
{
"epoch": 1.7005988023952097,
"grad_norm": 0.8460857943613574,
"learning_rate": 5.428342404965076e-07,
"loss": 0.1264,
"step": 568
},
{
"epoch": 1.7035928143712575,
"grad_norm": 0.982712212301185,
"learning_rate": 5.322277355934557e-07,
"loss": 0.1578,
"step": 569
},
{
"epoch": 1.7065868263473054,
"grad_norm": 0.8106622020757973,
"learning_rate": 5.217200489151714e-07,
"loss": 0.123,
"step": 570
},
{
"epoch": 1.7095808383233533,
"grad_norm": 0.852108383726623,
"learning_rate": 5.113114128707592e-07,
"loss": 0.1316,
"step": 571
},
{
"epoch": 1.7125748502994012,
"grad_norm": 0.9516643161630692,
"learning_rate": 5.010020576785174e-07,
"loss": 0.1548,
"step": 572
},
{
"epoch": 1.715568862275449,
"grad_norm": 0.999853143716961,
"learning_rate": 4.907922113608532e-07,
"loss": 0.1394,
"step": 573
},
{
"epoch": 1.718562874251497,
"grad_norm": 1.00274877571203,
"learning_rate": 4.806820997392325e-07,
"loss": 0.1686,
"step": 574
},
{
"epoch": 1.7215568862275448,
"grad_norm": 0.8171457486012267,
"learning_rate": 4.7067194642919036e-07,
"loss": 0.129,
"step": 575
},
{
"epoch": 1.7245508982035929,
"grad_norm": 0.9715852335556845,
"learning_rate": 4.607619728353818e-07,
"loss": 0.1384,
"step": 576
},
{
"epoch": 1.7275449101796407,
"grad_norm": 0.947655466236461,
"learning_rate": 4.50952398146689e-07,
"loss": 0.1464,
"step": 577
},
{
"epoch": 1.7305389221556886,
"grad_norm": 0.8561342460498808,
"learning_rate": 4.4124343933136525e-07,
"loss": 0.1296,
"step": 578
},
{
"epoch": 1.7335329341317365,
"grad_norm": 0.8784120388259629,
"learning_rate": 4.3163531113224466e-07,
"loss": 0.1538,
"step": 579
},
{
"epoch": 1.7365269461077846,
"grad_norm": 1.103021627958148,
"learning_rate": 4.221282260619891e-07,
"loss": 0.172,
"step": 580
},
{
"epoch": 1.7395209580838324,
"grad_norm": 0.8379354946212306,
"learning_rate": 4.127223943983849e-07,
"loss": 0.117,
"step": 581
},
{
"epoch": 1.7425149700598803,
"grad_norm": 0.8999170134308602,
"learning_rate": 4.03418024179697e-07,
"loss": 0.1373,
"step": 582
},
{
"epoch": 1.7455089820359282,
"grad_norm": 0.9311831171696595,
"learning_rate": 3.9421532120006544e-07,
"loss": 0.1688,
"step": 583
},
{
"epoch": 1.748502994011976,
"grad_norm": 0.9158945600479338,
"learning_rate": 3.851144890049535e-07,
"loss": 0.1468,
"step": 584
},
{
"epoch": 1.751497005988024,
"grad_norm": 0.8775057375830257,
"learning_rate": 3.761157288866418e-07,
"loss": 0.1507,
"step": 585
},
{
"epoch": 1.7544910179640718,
"grad_norm": 0.874731978333,
"learning_rate": 3.672192398797858e-07,
"loss": 0.1605,
"step": 586
},
{
"epoch": 1.7574850299401197,
"grad_norm": 1.0648099909558486,
"learning_rate": 3.58425218757002e-07,
"loss": 0.1628,
"step": 587
},
{
"epoch": 1.7604790419161676,
"grad_norm": 0.8752139917934191,
"learning_rate": 3.497338600245254e-07,
"loss": 0.1183,
"step": 588
},
{
"epoch": 1.7634730538922154,
"grad_norm": 0.8602950484402034,
"learning_rate": 3.4114535591790233e-07,
"loss": 0.134,
"step": 589
},
{
"epoch": 1.7664670658682635,
"grad_norm": 1.0967638836794942,
"learning_rate": 3.326598963977395e-07,
"loss": 0.1521,
"step": 590
},
{
"epoch": 1.7694610778443114,
"grad_norm": 0.943348956982322,
"learning_rate": 3.242776691455013e-07,
"loss": 0.1476,
"step": 591
},
{
"epoch": 1.7724550898203593,
"grad_norm": 0.8459155127797662,
"learning_rate": 3.159988595593616e-07,
"loss": 0.1298,
"step": 592
},
{
"epoch": 1.7754491017964071,
"grad_norm": 0.8461297235123652,
"learning_rate": 3.078236507501015e-07,
"loss": 0.1348,
"step": 593
},
{
"epoch": 1.7784431137724552,
"grad_norm": 0.872540575355707,
"learning_rate": 2.9975222353705757e-07,
"loss": 0.1428,
"step": 594
},
{
"epoch": 1.781437125748503,
"grad_norm": 1.0190835501483781,
"learning_rate": 2.917847564441256e-07,
"loss": 0.1666,
"step": 595
},
{
"epoch": 1.784431137724551,
"grad_norm": 0.9690923649053796,
"learning_rate": 2.839214256958106e-07,
"loss": 0.1487,
"step": 596
},
{
"epoch": 1.7874251497005988,
"grad_norm": 0.9335334042619156,
"learning_rate": 2.7616240521332884e-07,
"loss": 0.1485,
"step": 597
},
{
"epoch": 1.7904191616766467,
"grad_norm": 0.918173952003638,
"learning_rate": 2.6850786661076047e-07,
"loss": 0.1305,
"step": 598
},
{
"epoch": 1.7934131736526946,
"grad_norm": 0.8372973085552247,
"learning_rate": 2.6095797919125533e-07,
"loss": 0.1206,
"step": 599
},
{
"epoch": 1.7964071856287425,
"grad_norm": 0.9019470302628685,
"learning_rate": 2.5351290994328703e-07,
"loss": 0.1514,
"step": 600
},
{
"epoch": 1.7964071856287425,
"eval_loss": 0.18844835460186005,
"eval_runtime": 2.0195,
"eval_samples_per_second": 26.74,
"eval_steps_per_second": 6.933,
"step": 600
},
{
"epoch": 1.7994011976047903,
"grad_norm": 0.9738882849689182,
"learning_rate": 2.4617282353696093e-07,
"loss": 0.1445,
"step": 601
},
{
"epoch": 1.8023952095808382,
"grad_norm": 1.088337757765501,
"learning_rate": 2.3893788232036807e-07,
"loss": 0.1727,
"step": 602
},
{
"epoch": 1.805389221556886,
"grad_norm": 0.9413511508473933,
"learning_rate": 2.318082463160032e-07,
"loss": 0.154,
"step": 603
},
{
"epoch": 1.8083832335329342,
"grad_norm": 0.8906763833610244,
"learning_rate": 2.2478407321721295e-07,
"loss": 0.1478,
"step": 604
},
{
"epoch": 1.811377245508982,
"grad_norm": 0.9204262547416184,
"learning_rate": 2.1786551838471892e-07,
"loss": 0.1632,
"step": 605
},
{
"epoch": 1.81437125748503,
"grad_norm": 0.9158992750652618,
"learning_rate": 2.1105273484317402e-07,
"loss": 0.1103,
"step": 606
},
{
"epoch": 1.8173652694610778,
"grad_norm": 0.9058087350013028,
"learning_rate": 2.043458732777831e-07,
"loss": 0.1634,
"step": 607
},
{
"epoch": 1.8203592814371259,
"grad_norm": 1.1503340247085205,
"learning_rate": 1.9774508203096843e-07,
"loss": 0.1425,
"step": 608
},
{
"epoch": 1.8233532934131738,
"grad_norm": 0.8520863328185501,
"learning_rate": 1.9125050709908388e-07,
"loss": 0.1434,
"step": 609
},
{
"epoch": 1.8263473053892216,
"grad_norm": 0.8732656027981258,
"learning_rate": 1.8486229212919482e-07,
"loss": 0.1261,
"step": 610
},
{
"epoch": 1.8293413173652695,
"grad_norm": 0.802820428288992,
"learning_rate": 1.7858057841589281e-07,
"loss": 0.1221,
"step": 611
},
{
"epoch": 1.8323353293413174,
"grad_norm": 0.8884382160713526,
"learning_rate": 1.7240550489817652e-07,
"loss": 0.1397,
"step": 612
},
{
"epoch": 1.8353293413173652,
"grad_norm": 0.9617356681238824,
"learning_rate": 1.66337208156373e-07,
"loss": 0.1293,
"step": 613
},
{
"epoch": 1.8383233532934131,
"grad_norm": 1.0814305579849692,
"learning_rate": 1.6037582240912175e-07,
"loss": 0.1753,
"step": 614
},
{
"epoch": 1.841317365269461,
"grad_norm": 0.9515946250398718,
"learning_rate": 1.5452147951040165e-07,
"loss": 0.1438,
"step": 615
},
{
"epoch": 1.8443113772455089,
"grad_norm": 0.9390528079838442,
"learning_rate": 1.4877430894662037e-07,
"loss": 0.1796,
"step": 616
},
{
"epoch": 1.8473053892215567,
"grad_norm": 0.830632614199469,
"learning_rate": 1.4313443783374405e-07,
"loss": 0.1405,
"step": 617
},
{
"epoch": 1.8502994011976048,
"grad_norm": 0.8703299097909912,
"learning_rate": 1.3760199091449045e-07,
"loss": 0.1355,
"step": 618
},
{
"epoch": 1.8532934131736527,
"grad_norm": 0.8744113844922257,
"learning_rate": 1.3217709055556638e-07,
"loss": 0.133,
"step": 619
},
{
"epoch": 1.8562874251497006,
"grad_norm": 0.9679599458334553,
"learning_rate": 1.268598567449647e-07,
"loss": 0.1542,
"step": 620
},
{
"epoch": 1.8592814371257484,
"grad_norm": 0.8720483283482804,
"learning_rate": 1.2165040708930763e-07,
"loss": 0.1385,
"step": 621
},
{
"epoch": 1.8622754491017965,
"grad_norm": 0.7777818970181021,
"learning_rate": 1.1654885681124661e-07,
"loss": 0.1502,
"step": 622
},
{
"epoch": 1.8652694610778444,
"grad_norm": 0.9458453836857916,
"learning_rate": 1.1155531874691372e-07,
"loss": 0.1417,
"step": 623
},
{
"epoch": 1.8682634730538923,
"grad_norm": 0.8606182199588169,
"learning_rate": 1.0666990334342708e-07,
"loss": 0.1237,
"step": 624
},
{
"epoch": 1.8712574850299402,
"grad_norm": 0.9226841442710839,
"learning_rate": 1.0189271865644445e-07,
"loss": 0.138,
"step": 625
},
{
"epoch": 1.874251497005988,
"grad_norm": 1.108295450204729,
"learning_rate": 9.722387034777847e-08,
"loss": 0.1607,
"step": 626
},
{
"epoch": 1.877245508982036,
"grad_norm": 0.9248156176002923,
"learning_rate": 9.266346168305518e-08,
"loss": 0.1778,
"step": 627
},
{
"epoch": 1.8802395209580838,
"grad_norm": 0.8257927268500367,
"learning_rate": 8.821159352943142e-08,
"loss": 0.1308,
"step": 628
},
{
"epoch": 1.8832335329341316,
"grad_norm": 0.8367147942951034,
"learning_rate": 8.38683643533661e-08,
"loss": 0.1084,
"step": 629
},
{
"epoch": 1.8862275449101795,
"grad_norm": 0.8333139828901431,
"learning_rate": 7.963387021843683e-08,
"loss": 0.1083,
"step": 630
},
{
"epoch": 1.8892215568862274,
"grad_norm": 0.9759436671186427,
"learning_rate": 7.550820478322285e-08,
"loss": 0.1685,
"step": 631
},
{
"epoch": 1.8922155688622755,
"grad_norm": 0.9683133570488356,
"learning_rate": 7.149145929922607e-08,
"loss": 0.1608,
"step": 632
},
{
"epoch": 1.8952095808383234,
"grad_norm": 1.00452164187292,
"learning_rate": 6.758372260885714e-08,
"loss": 0.1591,
"step": 633
},
{
"epoch": 1.8982035928143712,
"grad_norm": 1.0882027297781565,
"learning_rate": 6.378508114346982e-08,
"loss": 0.1661,
"step": 634
},
{
"epoch": 1.9011976047904193,
"grad_norm": 0.9376919791225709,
"learning_rate": 6.009561892144744e-08,
"loss": 0.1264,
"step": 635
},
{
"epoch": 1.9041916167664672,
"grad_norm": 0.8718664063551013,
"learning_rate": 5.651541754634726e-08,
"loss": 0.116,
"step": 636
},
{
"epoch": 1.907185628742515,
"grad_norm": 1.0962547105719953,
"learning_rate": 5.304455620509297e-08,
"loss": 0.1658,
"step": 637
},
{
"epoch": 1.910179640718563,
"grad_norm": 0.8444331369083732,
"learning_rate": 4.968311166622553e-08,
"loss": 0.118,
"step": 638
},
{
"epoch": 1.9131736526946108,
"grad_norm": 0.8755255877061737,
"learning_rate": 4.643115827820399e-08,
"loss": 0.1389,
"step": 639
},
{
"epoch": 1.9161676646706587,
"grad_norm": 1.0401377351967207,
"learning_rate": 4.328876796776071e-08,
"loss": 0.1528,
"step": 640
},
{
"epoch": 1.9191616766467066,
"grad_norm": 0.8205343141147756,
"learning_rate": 4.0256010238310936e-08,
"loss": 0.1374,
"step": 641
},
{
"epoch": 1.9221556886227544,
"grad_norm": 0.9554420773559469,
"learning_rate": 3.733295216841626e-08,
"loss": 0.1458,
"step": 642
},
{
"epoch": 1.9251497005988023,
"grad_norm": 0.9617010988411111,
"learning_rate": 3.451965841029914e-08,
"loss": 0.1326,
"step": 643
},
{
"epoch": 1.9281437125748502,
"grad_norm": 0.885822154024949,
"learning_rate": 3.181619118841517e-08,
"loss": 0.1479,
"step": 644
},
{
"epoch": 1.931137724550898,
"grad_norm": 0.8517372654243078,
"learning_rate": 2.9222610298074717e-08,
"loss": 0.1124,
"step": 645
},
{
"epoch": 1.9341317365269461,
"grad_norm": 0.889581701325014,
"learning_rate": 2.673897310412288e-08,
"loss": 0.1463,
"step": 646
},
{
"epoch": 1.937125748502994,
"grad_norm": 0.8927223462662564,
"learning_rate": 2.4365334539667717e-08,
"loss": 0.1324,
"step": 647
},
{
"epoch": 1.9401197604790419,
"grad_norm": 0.8931371797655226,
"learning_rate": 2.210174710486679e-08,
"loss": 0.1577,
"step": 648
},
{
"epoch": 1.94311377245509,
"grad_norm": 0.7475654836588981,
"learning_rate": 1.99482608657664e-08,
"loss": 0.1323,
"step": 649
},
{
"epoch": 1.9461077844311379,
"grad_norm": 0.8566046172113009,
"learning_rate": 1.7904923453193056e-08,
"loss": 0.1265,
"step": 650
},
{
"epoch": 1.9491017964071857,
"grad_norm": 0.922914836921713,
"learning_rate": 1.5971780061701524e-08,
"loss": 0.1297,
"step": 651
},
{
"epoch": 1.9520958083832336,
"grad_norm": 0.7618184765695328,
"learning_rate": 1.4148873448573408e-08,
"loss": 0.1229,
"step": 652
},
{
"epoch": 1.9550898203592815,
"grad_norm": 0.9931476002175111,
"learning_rate": 1.2436243932872349e-08,
"loss": 0.1992,
"step": 653
},
{
"epoch": 1.9580838323353293,
"grad_norm": 0.9597919407272322,
"learning_rate": 1.0833929394552523e-08,
"loss": 0.1415,
"step": 654
},
{
"epoch": 1.9610778443113772,
"grad_norm": 0.8329589914520703,
"learning_rate": 9.341965273621522e-09,
"loss": 0.148,
"step": 655
},
{
"epoch": 1.964071856287425,
"grad_norm": 0.9882537553963469,
"learning_rate": 7.96038456935322e-09,
"loss": 0.1421,
"step": 656
},
{
"epoch": 1.967065868263473,
"grad_norm": 1.0770613946943581,
"learning_rate": 6.6892178395611125e-09,
"loss": 0.1629,
"step": 657
},
{
"epoch": 1.9700598802395208,
"grad_norm": 0.9365212966499,
"learning_rate": 5.528493199922769e-09,
"loss": 0.1432,
"step": 658
},
{
"epoch": 1.9730538922155687,
"grad_norm": 0.8106032378735075,
"learning_rate": 4.478236323355312e-09,
"loss": 0.1348,
"step": 659
},
{
"epoch": 1.9760479041916168,
"grad_norm": 1.0072791782756132,
"learning_rate": 3.538470439448105e-09,
"loss": 0.1613,
"step": 660
},
{
"epoch": 1.9790419161676647,
"grad_norm": 1.0764627832650429,
"learning_rate": 2.709216333952602e-09,
"loss": 0.1764,
"step": 661
},
{
"epoch": 1.9820359281437125,
"grad_norm": 0.8296902417003312,
"learning_rate": 1.9904923483171632e-09,
"loss": 0.1325,
"step": 662
},
{
"epoch": 1.9850299401197606,
"grad_norm": 1.1000758712159762,
"learning_rate": 1.3823143792851545e-09,
"loss": 0.1515,
"step": 663
},
{
"epoch": 1.9880239520958085,
"grad_norm": 1.0669799134614573,
"learning_rate": 8.846958785418969e-10,
"loss": 0.1571,
"step": 664
},
{
"epoch": 1.9910179640718564,
"grad_norm": 0.8749441806620147,
"learning_rate": 4.97647852417682e-10,
"loss": 0.1345,
"step": 665
},
{
"epoch": 1.9940119760479043,
"grad_norm": 0.8503476696496564,
"learning_rate": 2.2117886164407797e-10,
"loss": 0.1252,
"step": 666
},
{
"epoch": 1.9970059880239521,
"grad_norm": 0.8720864558941632,
"learning_rate": 5.529502116519148e-11,
"loss": 0.1487,
"step": 667
},
{
"epoch": 2.0,
"grad_norm": 0.8395153084458445,
"learning_rate": 0.0,
"loss": 0.1211,
"step": 668
},
{
"epoch": 2.0,
"step": 668,
"total_flos": 4854467788800.0,
"train_loss": 0.1698140936741929,
"train_runtime": 469.4132,
"train_samples_per_second": 22.705,
"train_steps_per_second": 1.423
}
],
"logging_steps": 1,
"max_steps": 668,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 5000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 4854467788800.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}