| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 200, | |
| "global_step": 812, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0012315270935960591, | |
| "grad_norm": 57.29781321672375, | |
| "learning_rate": 1.2195121951219514e-07, | |
| "loss": 4.0819, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006157635467980296, | |
| "grad_norm": 52.431237738982844, | |
| "learning_rate": 6.097560975609757e-07, | |
| "loss": 3.9153, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.012315270935960592, | |
| "grad_norm": 54.37135835417464, | |
| "learning_rate": 1.2195121951219514e-06, | |
| "loss": 3.9031, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01847290640394089, | |
| "grad_norm": 56.234490942529554, | |
| "learning_rate": 1.8292682926829268e-06, | |
| "loss": 3.8372, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.024630541871921183, | |
| "grad_norm": 46.67322399671909, | |
| "learning_rate": 2.4390243902439027e-06, | |
| "loss": 3.5462, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03078817733990148, | |
| "grad_norm": 28.51851019738398, | |
| "learning_rate": 3.0487804878048782e-06, | |
| "loss": 3.1669, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.03694581280788178, | |
| "grad_norm": 24.78506618043469, | |
| "learning_rate": 3.6585365853658537e-06, | |
| "loss": 2.7541, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.04310344827586207, | |
| "grad_norm": 12.098718231013407, | |
| "learning_rate": 4.268292682926829e-06, | |
| "loss": 2.3912, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.04926108374384237, | |
| "grad_norm": 11.07208120873305, | |
| "learning_rate": 4.8780487804878055e-06, | |
| "loss": 2.0806, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.05541871921182266, | |
| "grad_norm": 8.230990914967798, | |
| "learning_rate": 4.999667943970417e-06, | |
| "loss": 1.9354, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.06157635467980296, | |
| "grad_norm": 6.398154701368842, | |
| "learning_rate": 4.9983191175272635e-06, | |
| "loss": 1.7661, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.06773399014778325, | |
| "grad_norm": 4.8094047139439775, | |
| "learning_rate": 4.995933326588439e-06, | |
| "loss": 1.6504, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.07389162561576355, | |
| "grad_norm": 4.2116211664516845, | |
| "learning_rate": 4.992511561411944e-06, | |
| "loss": 1.5804, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.08004926108374384, | |
| "grad_norm": 3.8696635220995197, | |
| "learning_rate": 4.988055242252294e-06, | |
| "loss": 1.5485, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.08620689655172414, | |
| "grad_norm": 3.490496212045865, | |
| "learning_rate": 4.982566218771011e-06, | |
| "loss": 1.4986, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.09236453201970443, | |
| "grad_norm": 3.227704361032009, | |
| "learning_rate": 4.976046769268905e-06, | |
| "loss": 1.4551, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.09852216748768473, | |
| "grad_norm": 2.9568158470090333, | |
| "learning_rate": 4.968499599740427e-06, | |
| "loss": 1.4365, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.10467980295566502, | |
| "grad_norm": 2.6489029135193696, | |
| "learning_rate": 4.959927842750501e-06, | |
| "loss": 1.3971, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.11083743842364532, | |
| "grad_norm": 2.695449139827939, | |
| "learning_rate": 4.9503350561343165e-06, | |
| "loss": 1.362, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.11699507389162561, | |
| "grad_norm": 2.475139119509685, | |
| "learning_rate": 4.939725221520586e-06, | |
| "loss": 1.358, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.12315270935960591, | |
| "grad_norm": 2.271449895430399, | |
| "learning_rate": 4.928102742678918e-06, | |
| "loss": 1.3248, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.12931034482758622, | |
| "grad_norm": 2.278223643874042, | |
| "learning_rate": 4.915472443691962e-06, | |
| "loss": 1.3154, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.1354679802955665, | |
| "grad_norm": 2.050985273196281, | |
| "learning_rate": 4.901839566953105e-06, | |
| "loss": 1.3065, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1416256157635468, | |
| "grad_norm": 2.0541869429643076, | |
| "learning_rate": 4.887209770990537e-06, | |
| "loss": 1.308, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.1477832512315271, | |
| "grad_norm": 1.970557118232297, | |
| "learning_rate": 4.871589128118593e-06, | |
| "loss": 1.2931, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1539408866995074, | |
| "grad_norm": 1.9219121016470442, | |
| "learning_rate": 4.854984121917349e-06, | |
| "loss": 1.2728, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.16009852216748768, | |
| "grad_norm": 2.02792404527473, | |
| "learning_rate": 4.83740164454152e-06, | |
| "loss": 1.2507, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.16625615763546797, | |
| "grad_norm": 1.864850210875446, | |
| "learning_rate": 4.8188489938597635e-06, | |
| "loss": 1.2556, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.1724137931034483, | |
| "grad_norm": 1.7725358951542842, | |
| "learning_rate": 4.799333870425594e-06, | |
| "loss": 1.2468, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.17857142857142858, | |
| "grad_norm": 1.8954908664106795, | |
| "learning_rate": 4.778864374281147e-06, | |
| "loss": 1.2515, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.18472906403940886, | |
| "grad_norm": 1.7533093986020087, | |
| "learning_rate": 4.757449001595149e-06, | |
| "loss": 1.2318, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.19088669950738915, | |
| "grad_norm": 1.674563355682731, | |
| "learning_rate": 4.735096641136443e-06, | |
| "loss": 1.2476, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.19704433497536947, | |
| "grad_norm": 1.6022339474949587, | |
| "learning_rate": 4.711816570584583e-06, | |
| "loss": 1.2405, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.20320197044334976, | |
| "grad_norm": 1.4858028100041947, | |
| "learning_rate": 4.687618452678986e-06, | |
| "loss": 1.2521, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.20935960591133004, | |
| "grad_norm": 1.5829619539731021, | |
| "learning_rate": 4.662512331208277e-06, | |
| "loss": 1.2319, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.21551724137931033, | |
| "grad_norm": 1.4585812804310208, | |
| "learning_rate": 4.636508626841458e-06, | |
| "loss": 1.2192, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.22167487684729065, | |
| "grad_norm": 1.4404840359187807, | |
| "learning_rate": 4.609618132802661e-06, | |
| "loss": 1.2405, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.22783251231527094, | |
| "grad_norm": 1.4822891953876243, | |
| "learning_rate": 4.581852010391253e-06, | |
| "loss": 1.2269, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.23399014778325122, | |
| "grad_norm": 1.442337203314052, | |
| "learning_rate": 4.55322178434918e-06, | |
| "loss": 1.205, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.24014778325123154, | |
| "grad_norm": 1.392426493644894, | |
| "learning_rate": 4.523739338077443e-06, | |
| "loss": 1.2183, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.24630541871921183, | |
| "grad_norm": 1.3520518479942292, | |
| "learning_rate": 4.493416908703725e-06, | |
| "loss": 1.2087, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.24630541871921183, | |
| "eval_loss": 1.2089186906814575, | |
| "eval_runtime": 145.5636, | |
| "eval_samples_per_second": 158.762, | |
| "eval_steps_per_second": 2.487, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2524630541871921, | |
| "grad_norm": 1.3364357978087034, | |
| "learning_rate": 4.462267082003183e-06, | |
| "loss": 1.1903, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.25862068965517243, | |
| "grad_norm": 1.3098633628715541, | |
| "learning_rate": 4.430302787174535e-06, | |
| "loss": 1.2175, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.2647783251231527, | |
| "grad_norm": 1.2146074250279708, | |
| "learning_rate": 4.397537291473607e-06, | |
| "loss": 1.1877, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.270935960591133, | |
| "grad_norm": 1.2919790646323188, | |
| "learning_rate": 4.36398419470656e-06, | |
| "loss": 1.196, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.2770935960591133, | |
| "grad_norm": 1.2530332956228611, | |
| "learning_rate": 4.329657423585088e-06, | |
| "loss": 1.1847, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.2832512315270936, | |
| "grad_norm": 1.2233743401286648, | |
| "learning_rate": 4.2945712259459245e-06, | |
| "loss": 1.1782, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.2894088669950739, | |
| "grad_norm": 1.0821088244874544, | |
| "learning_rate": 4.258740164837079e-06, | |
| "loss": 1.1876, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.2955665024630542, | |
| "grad_norm": 1.1214966186233786, | |
| "learning_rate": 4.222179112473211e-06, | |
| "loss": 1.1752, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.3017241379310345, | |
| "grad_norm": 1.1405907082061064, | |
| "learning_rate": 4.184903244062709e-06, | |
| "loss": 1.1834, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.3078817733990148, | |
| "grad_norm": 1.1180113089928194, | |
| "learning_rate": 4.146928031508988e-06, | |
| "loss": 1.1766, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.31403940886699505, | |
| "grad_norm": 1.105523316105366, | |
| "learning_rate": 4.108269236988648e-06, | |
| "loss": 1.1948, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.32019704433497537, | |
| "grad_norm": 0.9550453095593314, | |
| "learning_rate": 4.068942906409148e-06, | |
| "loss": 1.1578, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.3263546798029557, | |
| "grad_norm": 0.9941221294224154, | |
| "learning_rate": 4.028965362748714e-06, | |
| "loss": 1.1718, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.33251231527093594, | |
| "grad_norm": 0.9622174118526656, | |
| "learning_rate": 3.988353199281242e-06, | |
| "loss": 1.1588, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.33866995073891626, | |
| "grad_norm": 0.9518007967855452, | |
| "learning_rate": 3.9471232726890185e-06, | |
| "loss": 1.1772, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.3448275862068966, | |
| "grad_norm": 1.0353051472296388, | |
| "learning_rate": 3.905292696066095e-06, | |
| "loss": 1.1431, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.35098522167487683, | |
| "grad_norm": 0.9833569587890238, | |
| "learning_rate": 3.862878831815253e-06, | |
| "loss": 1.1732, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 0.9725651600200993, | |
| "learning_rate": 3.8198992844414736e-06, | |
| "loss": 1.1227, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.3633004926108374, | |
| "grad_norm": 0.8918313638798159, | |
| "learning_rate": 3.7763718932449322e-06, | |
| "loss": 1.1715, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.3694581280788177, | |
| "grad_norm": 1.059809808553125, | |
| "learning_rate": 3.7323147249165255e-06, | |
| "loss": 1.1438, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.37561576354679804, | |
| "grad_norm": 0.8911037607615869, | |
| "learning_rate": 3.6877460660390257e-06, | |
| "loss": 1.1491, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.3817733990147783, | |
| "grad_norm": 0.8758678976318, | |
| "learning_rate": 3.6426844154969654e-06, | |
| "loss": 1.1351, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.3879310344827586, | |
| "grad_norm": 0.9319756004629981, | |
| "learning_rate": 3.597148476798393e-06, | |
| "loss": 1.1457, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.39408866995073893, | |
| "grad_norm": 0.8562011655041667, | |
| "learning_rate": 3.5511571503117125e-06, | |
| "loss": 1.153, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.4002463054187192, | |
| "grad_norm": 0.8631125371209072, | |
| "learning_rate": 3.504729525420798e-06, | |
| "loss": 1.1278, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.4064039408866995, | |
| "grad_norm": 0.8372726176442157, | |
| "learning_rate": 3.4578848726016646e-06, | |
| "loss": 1.1448, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.4125615763546798, | |
| "grad_norm": 0.8427591587304757, | |
| "learning_rate": 3.410642635423967e-06, | |
| "loss": 1.1498, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.4187192118226601, | |
| "grad_norm": 0.8284034797316332, | |
| "learning_rate": 3.3630224224806616e-06, | |
| "loss": 1.1347, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.4248768472906404, | |
| "grad_norm": 0.7969943838916709, | |
| "learning_rate": 3.3150439992491667e-06, | |
| "loss": 1.1306, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.43103448275862066, | |
| "grad_norm": 0.7841258467886394, | |
| "learning_rate": 3.2667272798874038e-06, | |
| "loss": 1.1147, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.437192118226601, | |
| "grad_norm": 0.8698506517311302, | |
| "learning_rate": 3.2180923189681367e-06, | |
| "loss": 1.1261, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.4433497536945813, | |
| "grad_norm": 0.8917545084157941, | |
| "learning_rate": 3.169159303155017e-06, | |
| "loss": 1.1601, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.44950738916256155, | |
| "grad_norm": 0.8038943568517608, | |
| "learning_rate": 3.119948542823812e-06, | |
| "loss": 1.1297, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.45566502463054187, | |
| "grad_norm": 0.7809840947233494, | |
| "learning_rate": 3.0704804636322818e-06, | |
| "loss": 1.1319, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.4618226600985222, | |
| "grad_norm": 0.882437416337753, | |
| "learning_rate": 3.0207755980422036e-06, | |
| "loss": 1.1091, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.46798029556650245, | |
| "grad_norm": 0.7790616445808474, | |
| "learning_rate": 2.970854576797073e-06, | |
| "loss": 1.1239, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.47413793103448276, | |
| "grad_norm": 0.8320522532641044, | |
| "learning_rate": 2.920738120359e-06, | |
| "loss": 1.1291, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.4802955665024631, | |
| "grad_norm": 0.7761936531276284, | |
| "learning_rate": 2.8704470303083808e-06, | |
| "loss": 1.1214, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.48645320197044334, | |
| "grad_norm": 1.0348640934242699, | |
| "learning_rate": 2.820002180709888e-06, | |
| "loss": 1.139, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.49261083743842365, | |
| "grad_norm": 0.7529470220643377, | |
| "learning_rate": 2.769424509448379e-06, | |
| "loss": 1.1184, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.49261083743842365, | |
| "eval_loss": 1.1359272003173828, | |
| "eval_runtime": 145.5192, | |
| "eval_samples_per_second": 158.811, | |
| "eval_steps_per_second": 2.488, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.4987684729064039, | |
| "grad_norm": 0.7319049955009511, | |
| "learning_rate": 2.7187350095383197e-06, | |
| "loss": 1.1307, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.5049261083743842, | |
| "grad_norm": 0.7944174165988256, | |
| "learning_rate": 2.6679547204103174e-06, | |
| "loss": 1.1281, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.5110837438423645, | |
| "grad_norm": 0.7354074462564278, | |
| "learning_rate": 2.617104719178395e-06, | |
| "loss": 1.1268, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.5172413793103449, | |
| "grad_norm": 9.765135572157094, | |
| "learning_rate": 2.566206111891621e-06, | |
| "loss": 1.1517, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.5233990147783252, | |
| "grad_norm": 0.7361882082841237, | |
| "learning_rate": 2.5152800247737293e-06, | |
| "loss": 1.144, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.5295566502463054, | |
| "grad_norm": 0.7189583602976664, | |
| "learning_rate": 2.4643475954543657e-06, | |
| "loss": 1.1309, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.5357142857142857, | |
| "grad_norm": 0.6885666200051201, | |
| "learning_rate": 2.413429964195603e-06, | |
| "loss": 1.1287, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.541871921182266, | |
| "grad_norm": 0.7267908131449855, | |
| "learning_rate": 2.362548265117355e-06, | |
| "loss": 1.1205, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.5480295566502463, | |
| "grad_norm": 0.7416217316864918, | |
| "learning_rate": 2.3117236174253556e-06, | |
| "loss": 1.1205, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.5541871921182266, | |
| "grad_norm": 0.7715703971394161, | |
| "learning_rate": 2.260977116645306e-06, | |
| "loss": 1.1207, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.5603448275862069, | |
| "grad_norm": 0.7397928670597758, | |
| "learning_rate": 2.2103298258668755e-06, | |
| "loss": 1.1362, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.5665024630541872, | |
| "grad_norm": 0.6628570241953711, | |
| "learning_rate": 2.1598027670011427e-06, | |
| "loss": 1.1233, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.5726600985221675, | |
| "grad_norm": 0.6835018842239253, | |
| "learning_rate": 2.109416912055145e-06, | |
| "loss": 1.1104, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.5788177339901478, | |
| "grad_norm": 0.7202412027245217, | |
| "learning_rate": 2.0591931744271343e-06, | |
| "loss": 1.1126, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.5849753694581281, | |
| "grad_norm": 0.6827660438057419, | |
| "learning_rate": 2.009152400226159e-06, | |
| "loss": 1.145, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.5911330049261084, | |
| "grad_norm": 0.6937537769666066, | |
| "learning_rate": 1.959315359619575e-06, | |
| "loss": 1.1353, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.5972906403940886, | |
| "grad_norm": 0.6907502777860367, | |
| "learning_rate": 1.909702738212082e-06, | |
| "loss": 1.123, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.603448275862069, | |
| "grad_norm": 0.693175546587098, | |
| "learning_rate": 1.860335128459853e-06, | |
| "loss": 1.1011, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.6096059113300493, | |
| "grad_norm": 0.6890094920324707, | |
| "learning_rate": 1.8112330211233345e-06, | |
| "loss": 1.1225, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.6157635467980296, | |
| "grad_norm": 0.6283864984460129, | |
| "learning_rate": 1.7624167967622435e-06, | |
| "loss": 1.1304, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6219211822660099, | |
| "grad_norm": 0.651421242999227, | |
| "learning_rate": 1.7139067172763182e-06, | |
| "loss": 1.1178, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.6280788177339901, | |
| "grad_norm": 0.6411203916290963, | |
| "learning_rate": 1.6657229174953165e-06, | |
| "loss": 1.1178, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.6342364532019704, | |
| "grad_norm": 0.7064149412583947, | |
| "learning_rate": 1.6178853968217507e-06, | |
| "loss": 1.1096, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.6403940886699507, | |
| "grad_norm": 0.6785493165534039, | |
| "learning_rate": 1.5704140109298446e-06, | |
| "loss": 1.1017, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.646551724137931, | |
| "grad_norm": 0.6263668598279826, | |
| "learning_rate": 1.5233284635241335e-06, | |
| "loss": 1.1213, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.6527093596059114, | |
| "grad_norm": 0.6024047933221067, | |
| "learning_rate": 1.4766482981611538e-06, | |
| "loss": 1.135, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.6588669950738916, | |
| "grad_norm": 0.6074862524743537, | |
| "learning_rate": 1.4303928901375975e-06, | |
| "loss": 1.1152, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.6650246305418719, | |
| "grad_norm": 0.6687985677124222, | |
| "learning_rate": 1.3845814384483069e-06, | |
| "loss": 1.1238, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.6711822660098522, | |
| "grad_norm": 0.6938913634473785, | |
| "learning_rate": 1.3392329578174495e-06, | |
| "loss": 1.1163, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.6773399014778325, | |
| "grad_norm": 0.6485860550936319, | |
| "learning_rate": 1.2943662708061677e-06, | |
| "loss": 1.0949, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.6834975369458128, | |
| "grad_norm": 0.6651175170212851, | |
| "learning_rate": 1.2500000000000007e-06, | |
| "loss": 1.1052, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.6896551724137931, | |
| "grad_norm": 0.6699641632812873, | |
| "learning_rate": 1.2061525602792994e-06, | |
| "loss": 1.1276, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.6958128078817734, | |
| "grad_norm": 0.6569148767495105, | |
| "learning_rate": 1.1628421511758623e-06, | |
| "loss": 1.1168, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.7019704433497537, | |
| "grad_norm": 0.6414910470715921, | |
| "learning_rate": 1.1200867493189418e-06, | |
| "loss": 1.11, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.708128078817734, | |
| "grad_norm": 0.678067098458089, | |
| "learning_rate": 1.0779041009737814e-06, | |
| "loss": 1.0847, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 0.632585459457676, | |
| "learning_rate": 1.036311714675766e-06, | |
| "loss": 1.1182, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.7204433497536946, | |
| "grad_norm": 0.6691822937441232, | |
| "learning_rate": 9.953268539632373e-07, | |
| "loss": 1.0964, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.7266009852216748, | |
| "grad_norm": 0.6343579710582603, | |
| "learning_rate": 9.549665302120146e-07, | |
| "loss": 1.132, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.7327586206896551, | |
| "grad_norm": 0.6180396452041287, | |
| "learning_rate": 9.152474955745622e-07, | |
| "loss": 1.116, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.7389162561576355, | |
| "grad_norm": 0.6616169682957934, | |
| "learning_rate": 8.761862360267637e-07, | |
| "loss": 1.114, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.7389162561576355, | |
| "eval_loss": 1.116376280784607, | |
| "eval_runtime": 145.6375, | |
| "eval_samples_per_second": 158.682, | |
| "eval_steps_per_second": 2.486, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.7450738916256158, | |
| "grad_norm": 0.6289490113051065, | |
| "learning_rate": 8.377989645251719e-07, | |
| "loss": 1.1138, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.7512315270935961, | |
| "grad_norm": 0.660622290747466, | |
| "learning_rate": 8.001016142775788e-07, | |
| "loss": 1.112, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.7573891625615764, | |
| "grad_norm": 0.6134951955946074, | |
| "learning_rate": 7.631098321297043e-07, | |
| "loss": 1.1011, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.7635467980295566, | |
| "grad_norm": 0.5730948804433411, | |
| "learning_rate": 7.26838972070737e-07, | |
| "loss": 1.0975, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.7697044334975369, | |
| "grad_norm": 0.6367402038336315, | |
| "learning_rate": 6.913040888604319e-07, | |
| "loss": 1.1128, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.7758620689655172, | |
| "grad_norm": 0.6241201816149435, | |
| "learning_rate": 6.565199317804119e-07, | |
| "loss": 1.1261, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.7820197044334976, | |
| "grad_norm": 0.5653870478435855, | |
| "learning_rate": 6.22500938512256e-07, | |
| "loss": 1.0881, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.7881773399014779, | |
| "grad_norm": 0.6434562851528702, | |
| "learning_rate": 5.892612291449285e-07, | |
| "loss": 1.0967, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.7943349753694581, | |
| "grad_norm": 0.6554180876220743, | |
| "learning_rate": 5.568146003140224e-07, | |
| "loss": 1.1182, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.8004926108374384, | |
| "grad_norm": 0.5661893064494111, | |
| "learning_rate": 5.251745194752622e-07, | |
| "loss": 1.0855, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.8066502463054187, | |
| "grad_norm": 0.6316395392717625, | |
| "learning_rate": 4.943541193146359e-07, | |
| "loss": 1.0839, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.812807881773399, | |
| "grad_norm": 0.6880997274985746, | |
| "learning_rate": 4.6436619229748043e-07, | |
| "loss": 1.0974, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.8189655172413793, | |
| "grad_norm": 0.5859005991287833, | |
| "learning_rate": 4.352231853587763e-07, | |
| "loss": 1.0785, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.8251231527093597, | |
| "grad_norm": 0.6092650651655916, | |
| "learning_rate": 4.069371947368619e-07, | |
| "loss": 1.1242, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.8312807881773399, | |
| "grad_norm": 0.6169798037117927, | |
| "learning_rate": 3.795199609527117e-07, | |
| "loss": 1.0934, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.8374384236453202, | |
| "grad_norm": 0.6427617998792151, | |
| "learning_rate": 3.529828639368568e-07, | |
| "loss": 1.1091, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.8435960591133005, | |
| "grad_norm": 0.6027641515622904, | |
| "learning_rate": 3.273369183059782e-07, | |
| "loss": 1.1217, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.8497536945812808, | |
| "grad_norm": 0.6421164647547414, | |
| "learning_rate": 3.025927687911229e-07, | |
| "loss": 1.124, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.8559113300492611, | |
| "grad_norm": 0.6400304022478517, | |
| "learning_rate": 2.7876068581945195e-07, | |
| "loss": 1.1062, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.8620689655172413, | |
| "grad_norm": 0.560059894959134, | |
| "learning_rate": 2.558505612513479e-07, | |
| "loss": 1.1075, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.8682266009852216, | |
| "grad_norm": 0.5812782574180876, | |
| "learning_rate": 2.3387190427464817e-07, | |
| "loss": 1.1047, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.874384236453202, | |
| "grad_norm": 0.6572393162811876, | |
| "learning_rate": 2.1283383745771853e-07, | |
| "loss": 1.1214, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.8805418719211823, | |
| "grad_norm": 0.5786561583568361, | |
| "learning_rate": 1.9274509296299315e-07, | |
| "loss": 1.1158, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.8866995073891626, | |
| "grad_norm": 0.6313748204740087, | |
| "learning_rate": 1.736140089225613e-07, | |
| "loss": 1.0931, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.8928571428571429, | |
| "grad_norm": 0.6201628701455402, | |
| "learning_rate": 1.554485259773017e-07, | |
| "loss": 1.1147, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.8990147783251231, | |
| "grad_norm": 0.6261299289179968, | |
| "learning_rate": 1.3825618398100232e-07, | |
| "loss": 1.1198, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.9051724137931034, | |
| "grad_norm": 0.6418749663067701, | |
| "learning_rate": 1.2204411887083072e-07, | |
| "loss": 1.1194, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.9113300492610837, | |
| "grad_norm": 0.6473208096028387, | |
| "learning_rate": 1.068190597054583e-07, | |
| "loss": 1.1239, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.9174876847290641, | |
| "grad_norm": 0.6016869160420317, | |
| "learning_rate": 9.25873258720636e-08, | |
| "loss": 1.0837, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.9236453201970444, | |
| "grad_norm": 0.6065116668251945, | |
| "learning_rate": 7.935482446337628e-08, | |
| "loss": 1.0971, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.9298029556650246, | |
| "grad_norm": 0.6250083320216442, | |
| "learning_rate": 6.712704782585206e-08, | |
| "loss": 1.1142, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.9359605911330049, | |
| "grad_norm": 0.5942550368266071, | |
| "learning_rate": 5.590907127999173e-08, | |
| "loss": 1.1005, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.9421182266009852, | |
| "grad_norm": 0.6169133768850116, | |
| "learning_rate": 4.570555101375357e-08, | |
| "loss": 1.1132, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.9482758620689655, | |
| "grad_norm": 0.6063506272173839, | |
| "learning_rate": 3.652072214993335e-08, | |
| "loss": 1.1131, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.9544334975369458, | |
| "grad_norm": 0.6342757392800364, | |
| "learning_rate": 2.8358396988315272e-08, | |
| "loss": 1.108, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.9605911330049262, | |
| "grad_norm": 0.5872762183067076, | |
| "learning_rate": 2.122196342331767e-08, | |
| "loss": 1.0929, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.9667487684729064, | |
| "grad_norm": 0.6555117987749753, | |
| "learning_rate": 1.511438353779898e-08, | |
| "loss": 1.1045, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.9729064039408867, | |
| "grad_norm": 0.6586688843353297, | |
| "learning_rate": 1.0038192373600652e-08, | |
| "loss": 1.1009, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.979064039408867, | |
| "grad_norm": 0.605823715542427, | |
| "learning_rate": 5.995496879339924e-09, | |
| "loss": 1.1045, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.9852216748768473, | |
| "grad_norm": 0.6242883286900868, | |
| "learning_rate": 2.9879750358896565e-09, | |
| "loss": 1.1017, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.9852216748768473, | |
| "eval_loss": 1.1130995750427246, | |
| "eval_runtime": 145.5288, | |
| "eval_samples_per_second": 158.8, | |
| "eval_steps_per_second": 2.487, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.9913793103448276, | |
| "grad_norm": 0.5597207689632175, | |
| "learning_rate": 1.0168751599085035e-09, | |
| "loss": 1.1155, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.9975369458128078, | |
| "grad_norm": 0.6086338500014004, | |
| "learning_rate": 8.301538570676393e-11, | |
| "loss": 1.0802, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 812, | |
| "total_flos": 277556883980288.0, | |
| "train_loss": 1.2735300551494355, | |
| "train_runtime": 5425.0163, | |
| "train_samples_per_second": 38.316, | |
| "train_steps_per_second": 0.15 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 812, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 277556883980288.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |