| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9998676022772408, | |
| "eval_steps": 500, | |
| "global_step": 1888, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0005295908910366742, | |
| "grad_norm": 0.41111565141978, | |
| "learning_rate": 1.0582010582010582e-06, | |
| "loss": 1.3689, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0026479544551833707, | |
| "grad_norm": 0.4260948383357496, | |
| "learning_rate": 5.291005291005291e-06, | |
| "loss": 1.3797, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.005295908910366741, | |
| "grad_norm": 0.42426082631444323, | |
| "learning_rate": 1.0582010582010582e-05, | |
| "loss": 1.4097, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.007943863365550112, | |
| "grad_norm": 0.4917431791111332, | |
| "learning_rate": 1.5873015873015872e-05, | |
| "loss": 1.4192, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.010591817820733483, | |
| "grad_norm": 0.21890435137541137, | |
| "learning_rate": 2.1164021164021164e-05, | |
| "loss": 1.3564, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.013239772275916854, | |
| "grad_norm": 0.1563118955046167, | |
| "learning_rate": 2.6455026455026456e-05, | |
| "loss": 1.3689, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.015887726731100223, | |
| "grad_norm": 0.14441954891998668, | |
| "learning_rate": 3.1746031746031745e-05, | |
| "loss": 1.3235, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.018535681186283594, | |
| "grad_norm": 0.13273151212388007, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 1.2872, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.021183635641466966, | |
| "grad_norm": 0.13964432905217536, | |
| "learning_rate": 4.232804232804233e-05, | |
| "loss": 1.2703, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.023831590096650337, | |
| "grad_norm": 0.14877385709600144, | |
| "learning_rate": 4.761904761904762e-05, | |
| "loss": 1.286, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.026479544551833708, | |
| "grad_norm": 0.11433768860303765, | |
| "learning_rate": 5.291005291005291e-05, | |
| "loss": 1.2649, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.02912749900701708, | |
| "grad_norm": 0.09609426563043175, | |
| "learning_rate": 5.82010582010582e-05, | |
| "loss": 1.2586, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.03177545346220045, | |
| "grad_norm": 0.08894905109962493, | |
| "learning_rate": 6.349206349206349e-05, | |
| "loss": 1.2265, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.03442340791738382, | |
| "grad_norm": 0.07490077864915894, | |
| "learning_rate": 6.878306878306878e-05, | |
| "loss": 1.2306, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.03707136237256719, | |
| "grad_norm": 0.07227158023314931, | |
| "learning_rate": 7.407407407407407e-05, | |
| "loss": 1.2407, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.03971931682775056, | |
| "grad_norm": 0.06869740415301212, | |
| "learning_rate": 7.936507936507937e-05, | |
| "loss": 1.2122, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.04236727128293393, | |
| "grad_norm": 0.06853429613019985, | |
| "learning_rate": 8.465608465608466e-05, | |
| "loss": 1.2229, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0450152257381173, | |
| "grad_norm": 0.0712797771766177, | |
| "learning_rate": 8.994708994708995e-05, | |
| "loss": 1.1861, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.047663180193300674, | |
| "grad_norm": 0.06827620857683864, | |
| "learning_rate": 9.523809523809524e-05, | |
| "loss": 1.1737, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.050311134648484045, | |
| "grad_norm": 0.07511648347979284, | |
| "learning_rate": 0.00010052910052910055, | |
| "loss": 1.1917, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.052959089103667416, | |
| "grad_norm": 0.07161483769838682, | |
| "learning_rate": 0.00010582010582010582, | |
| "loss": 1.1657, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.05560704355885079, | |
| "grad_norm": 0.0828998822809963, | |
| "learning_rate": 0.00011111111111111112, | |
| "loss": 1.2058, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.05825499801403416, | |
| "grad_norm": 0.08432989244956403, | |
| "learning_rate": 0.0001164021164021164, | |
| "loss": 1.1959, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.06090295246921753, | |
| "grad_norm": 0.08546417778154296, | |
| "learning_rate": 0.0001216931216931217, | |
| "loss": 1.1939, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.0635509069244009, | |
| "grad_norm": 0.08953022960897895, | |
| "learning_rate": 0.00012698412698412698, | |
| "loss": 1.152, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.06619886137958426, | |
| "grad_norm": 0.09090497875499474, | |
| "learning_rate": 0.00013227513227513228, | |
| "loss": 1.162, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.06884681583476764, | |
| "grad_norm": 0.09067803395184647, | |
| "learning_rate": 0.00013756613756613756, | |
| "loss": 1.1564, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.071494770289951, | |
| "grad_norm": 0.09746896349076282, | |
| "learning_rate": 0.00014285714285714287, | |
| "loss": 1.153, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.07414272474513438, | |
| "grad_norm": 0.09151621194870359, | |
| "learning_rate": 0.00014814814814814815, | |
| "loss": 1.1733, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.07679067920031775, | |
| "grad_norm": 0.09449259114304437, | |
| "learning_rate": 0.00015343915343915345, | |
| "loss": 1.1674, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.07943863365550112, | |
| "grad_norm": 0.09545666936190501, | |
| "learning_rate": 0.00015873015873015873, | |
| "loss": 1.1701, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.08208658811068449, | |
| "grad_norm": 0.0933818650565856, | |
| "learning_rate": 0.00016402116402116404, | |
| "loss": 1.1243, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.08473454256586786, | |
| "grad_norm": 0.09593001870262655, | |
| "learning_rate": 0.00016931216931216931, | |
| "loss": 1.1347, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.08738249702105123, | |
| "grad_norm": 0.09994210185656571, | |
| "learning_rate": 0.00017460317460317462, | |
| "loss": 1.1522, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.0900304514762346, | |
| "grad_norm": 0.09791765751683923, | |
| "learning_rate": 0.0001798941798941799, | |
| "loss": 1.1597, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.09267840593141798, | |
| "grad_norm": 0.09168024517745389, | |
| "learning_rate": 0.0001851851851851852, | |
| "loss": 1.1296, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.09532636038660135, | |
| "grad_norm": 0.09617452615612945, | |
| "learning_rate": 0.00019047619047619048, | |
| "loss": 1.1261, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.09797431484178472, | |
| "grad_norm": 0.10949818534072465, | |
| "learning_rate": 0.0001957671957671958, | |
| "loss": 1.1437, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.10062226929696809, | |
| "grad_norm": 0.09788854738358516, | |
| "learning_rate": 0.00019999982904458238, | |
| "loss": 1.1351, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.10327022375215146, | |
| "grad_norm": 0.09651211835063897, | |
| "learning_rate": 0.00019999384566633964, | |
| "loss": 1.1285, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.10591817820733483, | |
| "grad_norm": 0.09895389323355416, | |
| "learning_rate": 0.00019997931510172217, | |
| "loss": 1.1592, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.1085661326625182, | |
| "grad_norm": 0.09504405875958917, | |
| "learning_rate": 0.0001999562385927608, | |
| "loss": 1.127, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.11121408711770157, | |
| "grad_norm": 0.08454569193883055, | |
| "learning_rate": 0.00019992461811196922, | |
| "loss": 1.1163, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.11386204157288494, | |
| "grad_norm": 0.10371923765022571, | |
| "learning_rate": 0.00019988445636217513, | |
| "loss": 1.1286, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.11650999602806832, | |
| "grad_norm": 0.09773134958204137, | |
| "learning_rate": 0.00019983575677628946, | |
| "loss": 1.1547, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.11915795048325169, | |
| "grad_norm": 0.08853073845684854, | |
| "learning_rate": 0.0001997785235170127, | |
| "loss": 1.1391, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.12180590493843506, | |
| "grad_norm": 0.11043850318148267, | |
| "learning_rate": 0.00019971276147647937, | |
| "loss": 1.1265, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.12445385939361843, | |
| "grad_norm": 0.08955488246598574, | |
| "learning_rate": 0.00019963847627583952, | |
| "loss": 1.1293, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.1271018138488018, | |
| "grad_norm": 0.08677646900814183, | |
| "learning_rate": 0.00019955567426477847, | |
| "loss": 1.1665, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.12974976830398516, | |
| "grad_norm": 0.0827435794481087, | |
| "learning_rate": 0.00019946436252097401, | |
| "loss": 1.1517, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.13239772275916853, | |
| "grad_norm": 0.08325869934671043, | |
| "learning_rate": 0.00019936454884949146, | |
| "loss": 1.1491, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.1350456772143519, | |
| "grad_norm": 0.08582838361774685, | |
| "learning_rate": 0.0001992562417821164, | |
| "loss": 1.1447, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.13769363166953527, | |
| "grad_norm": 0.0784880184114551, | |
| "learning_rate": 0.00019913945057662547, | |
| "loss": 1.1321, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.14034158612471864, | |
| "grad_norm": 0.09123702006719589, | |
| "learning_rate": 0.00019901418521599506, | |
| "loss": 1.0955, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.142989540579902, | |
| "grad_norm": 0.08265522736674327, | |
| "learning_rate": 0.000198880456407548, | |
| "loss": 1.1493, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.14563749503508538, | |
| "grad_norm": 0.09073002430436583, | |
| "learning_rate": 0.00019873827558203826, | |
| "loss": 1.1101, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.14828544949026876, | |
| "grad_norm": 0.08138379527510481, | |
| "learning_rate": 0.0001985876548926739, | |
| "loss": 1.0982, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.15093340394545213, | |
| "grad_norm": 0.09137910087915257, | |
| "learning_rate": 0.00019842860721407837, | |
| "loss": 1.1312, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.1535813584006355, | |
| "grad_norm": 0.08477969588194346, | |
| "learning_rate": 0.0001982611461411898, | |
| "loss": 1.1413, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.15622931285581887, | |
| "grad_norm": 0.08817787188441822, | |
| "learning_rate": 0.0001980852859880992, | |
| "loss": 1.1394, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.15887726731100224, | |
| "grad_norm": 0.08829334171222576, | |
| "learning_rate": 0.00019790104178682666, | |
| "loss": 1.1236, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.1615252217661856, | |
| "grad_norm": 0.09023785166170276, | |
| "learning_rate": 0.00019770842928603673, | |
| "loss": 1.1205, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.16417317622136898, | |
| "grad_norm": 0.09500691769663268, | |
| "learning_rate": 0.000197507464949692, | |
| "loss": 1.1167, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.16682113067655235, | |
| "grad_norm": 0.08426021874529936, | |
| "learning_rate": 0.00019729816595564614, | |
| "loss": 1.1106, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.16946908513173572, | |
| "grad_norm": 0.08147138058421277, | |
| "learning_rate": 0.00019708055019417522, | |
| "loss": 1.11, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.1721170395869191, | |
| "grad_norm": 0.08138544612829161, | |
| "learning_rate": 0.00019685463626644873, | |
| "loss": 1.1186, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.17476499404210247, | |
| "grad_norm": 0.08666272634258464, | |
| "learning_rate": 0.00019662044348293955, | |
| "loss": 1.1372, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.17741294849728584, | |
| "grad_norm": 0.0843383524929789, | |
| "learning_rate": 0.00019637799186177332, | |
| "loss": 1.1079, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.1800609029524692, | |
| "grad_norm": 0.0772317749907282, | |
| "learning_rate": 0.00019612730212701746, | |
| "loss": 1.1475, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.18270885740765258, | |
| "grad_norm": 0.0774927041377668, | |
| "learning_rate": 0.00019586839570690957, | |
| "loss": 1.1152, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.18535681186283595, | |
| "grad_norm": 0.07478193615591532, | |
| "learning_rate": 0.00019560129473202587, | |
| "loss": 1.1267, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.18800476631801932, | |
| "grad_norm": 0.08480217654240685, | |
| "learning_rate": 0.00019532602203338958, | |
| "loss": 1.1239, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.1906527207732027, | |
| "grad_norm": 0.07940327835709149, | |
| "learning_rate": 0.00019504260114051945, | |
| "loss": 1.1309, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.19330067522838607, | |
| "grad_norm": 0.07586072518369895, | |
| "learning_rate": 0.00019475105627941828, | |
| "loss": 1.134, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.19594862968356944, | |
| "grad_norm": 0.07748779749674514, | |
| "learning_rate": 0.00019445141237050246, | |
| "loss": 1.1257, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.1985965841387528, | |
| "grad_norm": 0.0749023019530035, | |
| "learning_rate": 0.00019414369502647163, | |
| "loss": 1.1194, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.20124453859393618, | |
| "grad_norm": 0.07770003396390662, | |
| "learning_rate": 0.00019382793055011942, | |
| "loss": 1.1199, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.20389249304911955, | |
| "grad_norm": 0.07762689051643917, | |
| "learning_rate": 0.00019350414593208526, | |
| "loss": 1.1427, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.20654044750430292, | |
| "grad_norm": 0.07896343783224463, | |
| "learning_rate": 0.00019317236884854714, | |
| "loss": 1.1399, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.2091884019594863, | |
| "grad_norm": 0.08341099352486485, | |
| "learning_rate": 0.00019283262765885614, | |
| "loss": 1.0938, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.21183635641466966, | |
| "grad_norm": 0.08207858561332201, | |
| "learning_rate": 0.0001924849514031121, | |
| "loss": 1.1024, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.21448431086985303, | |
| "grad_norm": 0.07902580455460373, | |
| "learning_rate": 0.00019212936979968158, | |
| "loss": 1.1302, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.2171322653250364, | |
| "grad_norm": 0.08126318805436292, | |
| "learning_rate": 0.00019176591324265746, | |
| "loss": 1.1256, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.21978021978021978, | |
| "grad_norm": 0.08262896761244223, | |
| "learning_rate": 0.0001913946127992611, | |
| "loss": 1.1168, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.22242817423540315, | |
| "grad_norm": 0.08486099971998615, | |
| "learning_rate": 0.0001910155002071866, | |
| "loss": 1.1093, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.22507612869058652, | |
| "grad_norm": 0.08249982374177663, | |
| "learning_rate": 0.0001906286078718881, | |
| "loss": 1.1281, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.2277240831457699, | |
| "grad_norm": 0.08241790745995113, | |
| "learning_rate": 0.00019023396886380987, | |
| "loss": 1.1306, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.23037203760095326, | |
| "grad_norm": 0.08583672564162287, | |
| "learning_rate": 0.00018983161691555947, | |
| "loss": 1.1391, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.23301999205613663, | |
| "grad_norm": 0.07678631371522364, | |
| "learning_rate": 0.00018942158641902434, | |
| "loss": 1.1249, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.23566794651132, | |
| "grad_norm": 0.07687501040269233, | |
| "learning_rate": 0.00018900391242243224, | |
| "loss": 1.1275, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.23831590096650337, | |
| "grad_norm": 0.07734964302998311, | |
| "learning_rate": 0.00018857863062735525, | |
| "loss": 1.1419, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.24096385542168675, | |
| "grad_norm": 0.07789633231442543, | |
| "learning_rate": 0.00018814577738565826, | |
| "loss": 1.1313, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.24361180987687012, | |
| "grad_norm": 0.0786175058826523, | |
| "learning_rate": 0.00018770538969639162, | |
| "loss": 1.0941, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.2462597643320535, | |
| "grad_norm": 0.07662475568587293, | |
| "learning_rate": 0.0001872575052026286, | |
| "loss": 1.0895, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.24890771878723686, | |
| "grad_norm": 0.07891566103155145, | |
| "learning_rate": 0.00018680216218824763, | |
| "loss": 1.1195, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.25155567324242023, | |
| "grad_norm": 0.08801022195618644, | |
| "learning_rate": 0.0001863393995746603, | |
| "loss": 1.1123, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.2542036276976036, | |
| "grad_norm": 0.07439626465552164, | |
| "learning_rate": 0.00018586925691748394, | |
| "loss": 1.105, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.256851582152787, | |
| "grad_norm": 0.08921415332985139, | |
| "learning_rate": 0.00018539177440316096, | |
| "loss": 1.1183, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.2594995366079703, | |
| "grad_norm": 0.0722928891718278, | |
| "learning_rate": 0.00018490699284552352, | |
| "loss": 1.1225, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.2621474910631537, | |
| "grad_norm": 0.08123610651623102, | |
| "learning_rate": 0.00018441495368230518, | |
| "loss": 1.12, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.26479544551833706, | |
| "grad_norm": 0.08591833713679868, | |
| "learning_rate": 0.00018391569897159853, | |
| "loss": 1.1164, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.26744339997352046, | |
| "grad_norm": 0.0766094125907269, | |
| "learning_rate": 0.0001834092713882606, | |
| "loss": 1.1197, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.2700913544287038, | |
| "grad_norm": 0.07548163044316222, | |
| "learning_rate": 0.00018289571422026488, | |
| "loss": 1.111, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.2727393088838872, | |
| "grad_norm": 0.07972346954097494, | |
| "learning_rate": 0.00018237507136500126, | |
| "loss": 1.101, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.27538726333907054, | |
| "grad_norm": 0.08328304497316567, | |
| "learning_rate": 0.00018184738732552374, | |
| "loss": 1.1116, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.27803521779425394, | |
| "grad_norm": 0.0972393144531956, | |
| "learning_rate": 0.00018131270720674658, | |
| "loss": 1.1116, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.2806831722494373, | |
| "grad_norm": 0.07637369852545137, | |
| "learning_rate": 0.00018077107671158878, | |
| "loss": 1.1091, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.2833311267046207, | |
| "grad_norm": 0.07677607153425371, | |
| "learning_rate": 0.00018022254213706745, | |
| "loss": 1.0991, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.285979081159804, | |
| "grad_norm": 0.07646177666293459, | |
| "learning_rate": 0.0001796671503703406, | |
| "loss": 1.1015, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.2886270356149874, | |
| "grad_norm": 0.07666293082207067, | |
| "learning_rate": 0.00017910494888469938, | |
| "loss": 1.1007, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.29127499007017077, | |
| "grad_norm": 0.075595023851119, | |
| "learning_rate": 0.00017853598573550997, | |
| "loss": 1.1323, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.29392294452535417, | |
| "grad_norm": 0.07622954120001359, | |
| "learning_rate": 0.0001779603095561062, | |
| "loss": 1.1397, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.2965708989805375, | |
| "grad_norm": 0.07515451588120746, | |
| "learning_rate": 0.00017737796955363244, | |
| "loss": 1.0916, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.2992188534357209, | |
| "grad_norm": 0.07859697697366838, | |
| "learning_rate": 0.0001767890155048375, | |
| "loss": 1.0997, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.30186680789090425, | |
| "grad_norm": 0.07702712175636178, | |
| "learning_rate": 0.0001761934977518197, | |
| "loss": 1.0988, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.30451476234608765, | |
| "grad_norm": 0.0786046652278411, | |
| "learning_rate": 0.00017559146719772417, | |
| "loss": 1.1111, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.307162716801271, | |
| "grad_norm": 0.07520793055259228, | |
| "learning_rate": 0.00017498297530239132, | |
| "loss": 1.0989, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.3098106712564544, | |
| "grad_norm": 0.07578073058602562, | |
| "learning_rate": 0.00017436807407795852, | |
| "loss": 1.0948, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.31245862571163774, | |
| "grad_norm": 0.07604308595840065, | |
| "learning_rate": 0.00017374681608441422, | |
| "loss": 1.1036, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.31510658016682114, | |
| "grad_norm": 0.08158146756884099, | |
| "learning_rate": 0.0001731192544251051, | |
| "loss": 1.1225, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.3177545346220045, | |
| "grad_norm": 0.07498544543280497, | |
| "learning_rate": 0.00017248544274219715, | |
| "loss": 1.1201, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.3204024890771879, | |
| "grad_norm": 0.08493466237400855, | |
| "learning_rate": 0.00017184543521209036, | |
| "loss": 1.103, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.3230504435323712, | |
| "grad_norm": 0.08527784775049631, | |
| "learning_rate": 0.00017119928654078792, | |
| "loss": 1.1194, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.3256983979875546, | |
| "grad_norm": 0.08169595986359099, | |
| "learning_rate": 0.00017054705195922008, | |
| "loss": 1.0933, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.32834635244273797, | |
| "grad_norm": 0.08197785273219102, | |
| "learning_rate": 0.00016988878721852334, | |
| "loss": 1.1117, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.33099430689792136, | |
| "grad_norm": 0.07335268515860134, | |
| "learning_rate": 0.0001692245485852747, | |
| "loss": 1.0865, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.3336422613531047, | |
| "grad_norm": 0.08352503426226525, | |
| "learning_rate": 0.00016855439283668233, | |
| "loss": 1.1084, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.3362902158082881, | |
| "grad_norm": 0.07593742235391308, | |
| "learning_rate": 0.00016787837725573253, | |
| "loss": 1.107, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.33893817026347145, | |
| "grad_norm": 0.07457902202614533, | |
| "learning_rate": 0.0001671965596262931, | |
| "loss": 1.0935, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.34158612471865485, | |
| "grad_norm": 0.08114151429238778, | |
| "learning_rate": 0.00016650899822817433, | |
| "loss": 1.0969, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.3442340791738382, | |
| "grad_norm": 0.0717382651710646, | |
| "learning_rate": 0.00016581575183214722, | |
| "loss": 1.1352, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.3468820336290216, | |
| "grad_norm": 0.09884940877067575, | |
| "learning_rate": 0.00016511687969492024, | |
| "loss": 1.0922, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.34952998808420493, | |
| "grad_norm": 0.08048808120404499, | |
| "learning_rate": 0.00016441244155407386, | |
| "loss": 1.1225, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.35217794253938833, | |
| "grad_norm": 0.07768291438253641, | |
| "learning_rate": 0.00016370249762295464, | |
| "loss": 1.1163, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.3548258969945717, | |
| "grad_norm": 0.07669697580314543, | |
| "learning_rate": 0.00016298710858552817, | |
| "loss": 1.1454, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.3574738514497551, | |
| "grad_norm": 0.07654243142581256, | |
| "learning_rate": 0.0001622663355911922, | |
| "loss": 1.0954, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.3601218059049384, | |
| "grad_norm": 0.07308865070979989, | |
| "learning_rate": 0.0001615402402495495, | |
| "loss": 1.0849, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.3627697603601218, | |
| "grad_norm": 0.07098787881731487, | |
| "learning_rate": 0.00016080888462514185, | |
| "loss": 1.115, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.36541771481530516, | |
| "grad_norm": 0.07985291028148536, | |
| "learning_rate": 0.00016007233123214486, | |
| "loss": 1.0802, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.36806566927048856, | |
| "grad_norm": 0.07631811342549887, | |
| "learning_rate": 0.00015933064302902446, | |
| "loss": 1.12, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.3707136237256719, | |
| "grad_norm": 0.0716109218501906, | |
| "learning_rate": 0.00015858388341315535, | |
| "loss": 1.1223, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.3733615781808553, | |
| "grad_norm": 0.07770179033811983, | |
| "learning_rate": 0.00015783211621540208, | |
| "loss": 1.1206, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.37600953263603865, | |
| "grad_norm": 0.08008613710569794, | |
| "learning_rate": 0.00015707540569466282, | |
| "loss": 1.0927, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.37865748709122204, | |
| "grad_norm": 0.07909376018837275, | |
| "learning_rate": 0.00015631381653237683, | |
| "loss": 1.1208, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.3813054415464054, | |
| "grad_norm": 0.08795687346992156, | |
| "learning_rate": 0.00015554741382699556, | |
| "loss": 1.1158, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.3839533960015888, | |
| "grad_norm": 0.07837037230945529, | |
| "learning_rate": 0.00015477626308841832, | |
| "loss": 1.0996, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.38660135045677213, | |
| "grad_norm": 0.07641657124459816, | |
| "learning_rate": 0.0001540004302323927, | |
| "loss": 1.1014, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.38924930491195553, | |
| "grad_norm": 0.08137671044593779, | |
| "learning_rate": 0.00015321998157488014, | |
| "loss": 1.1078, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.3918972593671389, | |
| "grad_norm": 0.08041109732972, | |
| "learning_rate": 0.0001524349838263875, | |
| "loss": 1.1118, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.39454521382232227, | |
| "grad_norm": 0.07625661557827795, | |
| "learning_rate": 0.00015164550408626495, | |
| "loss": 1.102, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.3971931682775056, | |
| "grad_norm": 0.07622512218698127, | |
| "learning_rate": 0.00015085160983697033, | |
| "loss": 1.0988, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.399841122732689, | |
| "grad_norm": 0.07118368595229167, | |
| "learning_rate": 0.00015005336893830106, | |
| "loss": 1.1268, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.40248907718787236, | |
| "grad_norm": 0.07416345532311784, | |
| "learning_rate": 0.0001492508496215937, | |
| "loss": 1.0972, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.40513703164305576, | |
| "grad_norm": 0.07934756335393346, | |
| "learning_rate": 0.0001484441204838916, | |
| "loss": 1.1218, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.4077849860982391, | |
| "grad_norm": 0.07517515602393443, | |
| "learning_rate": 0.00014763325048208154, | |
| "loss": 1.1191, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.4104329405534225, | |
| "grad_norm": 0.07599570184456439, | |
| "learning_rate": 0.00014681830892699946, | |
| "loss": 1.1095, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.41308089500860584, | |
| "grad_norm": 0.07910162374766451, | |
| "learning_rate": 0.00014599936547750593, | |
| "loss": 1.0652, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.41572884946378924, | |
| "grad_norm": 0.07982743824380482, | |
| "learning_rate": 0.000145176490134532, | |
| "loss": 1.1065, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.4183768039189726, | |
| "grad_norm": 0.07714607042652752, | |
| "learning_rate": 0.00014434975323509566, | |
| "loss": 1.1059, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.421024758374156, | |
| "grad_norm": 0.07606594821793712, | |
| "learning_rate": 0.00014351922544628959, | |
| "loss": 1.0735, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.4236727128293393, | |
| "grad_norm": 0.07739034795460513, | |
| "learning_rate": 0.00014268497775924078, | |
| "loss": 1.1292, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.4263206672845227, | |
| "grad_norm": 0.08112666266940216, | |
| "learning_rate": 0.00014184708148304255, | |
| "loss": 1.0935, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.42896862173970607, | |
| "grad_norm": 0.07417441389696992, | |
| "learning_rate": 0.000141005608238659, | |
| "loss": 1.0917, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.43161657619488947, | |
| "grad_norm": 0.07327160038369271, | |
| "learning_rate": 0.0001401606299528033, | |
| "loss": 1.1082, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.4342645306500728, | |
| "grad_norm": 0.07329251639640971, | |
| "learning_rate": 0.0001393122188517893, | |
| "loss": 1.0983, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.4369124851052562, | |
| "grad_norm": 0.07652688725710881, | |
| "learning_rate": 0.00013846044745535822, | |
| "loss": 1.1055, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.43956043956043955, | |
| "grad_norm": 0.07594879575700789, | |
| "learning_rate": 0.00013760538857047957, | |
| "loss": 1.0881, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.44220839401562295, | |
| "grad_norm": 0.07795025881048111, | |
| "learning_rate": 0.00013674711528512793, | |
| "loss": 1.135, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.4448563484708063, | |
| "grad_norm": 0.07525588786296598, | |
| "learning_rate": 0.00013588570096203554, | |
| "loss": 1.0939, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.4475043029259897, | |
| "grad_norm": 0.07793636775238774, | |
| "learning_rate": 0.00013502121923242164, | |
| "loss": 1.0912, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.45015225738117304, | |
| "grad_norm": 0.07426019016116724, | |
| "learning_rate": 0.00013415374398969844, | |
| "loss": 1.1182, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.45280021183635644, | |
| "grad_norm": 0.07684359220337984, | |
| "learning_rate": 0.00013328334938315513, | |
| "loss": 1.0972, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.4554481662915398, | |
| "grad_norm": 0.08240740529852435, | |
| "learning_rate": 0.0001324101098116196, | |
| "loss": 1.1124, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.4580961207467232, | |
| "grad_norm": 0.0763891854972936, | |
| "learning_rate": 0.00013153409991709933, | |
| "loss": 1.0883, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.4607440752019065, | |
| "grad_norm": 0.08060164757564352, | |
| "learning_rate": 0.0001306553945784009, | |
| "loss": 1.1097, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.4633920296570899, | |
| "grad_norm": 0.0769697871980839, | |
| "learning_rate": 0.00012977406890472975, | |
| "loss": 1.0881, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.46603998411227326, | |
| "grad_norm": 0.07383145111875959, | |
| "learning_rate": 0.0001288901982292701, | |
| "loss": 1.0872, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.46868793856745666, | |
| "grad_norm": 0.07293602517307703, | |
| "learning_rate": 0.0001280038581027455, | |
| "loss": 1.0712, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.47133589302264, | |
| "grad_norm": 0.07479709674097254, | |
| "learning_rate": 0.00012711512428696114, | |
| "loss": 1.0922, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.4739838474778234, | |
| "grad_norm": 0.0755945597589927, | |
| "learning_rate": 0.0001262240727483278, | |
| "loss": 1.0925, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.47663180193300675, | |
| "grad_norm": 0.07530841667815956, | |
| "learning_rate": 0.0001253307796513686, | |
| "loss": 1.1226, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.47927975638819015, | |
| "grad_norm": 0.08059154491334077, | |
| "learning_rate": 0.0001244353213522087, | |
| "loss": 1.1334, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.4819277108433735, | |
| "grad_norm": 0.07234568635036283, | |
| "learning_rate": 0.00012353777439204838, | |
| "loss": 1.1304, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.4845756652985569, | |
| "grad_norm": 0.07303744281420084, | |
| "learning_rate": 0.00012263821549062074, | |
| "loss": 1.0934, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.48722361975374023, | |
| "grad_norm": 0.07710065386256926, | |
| "learning_rate": 0.00012173672153963377, | |
| "loss": 1.1153, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.48987157420892363, | |
| "grad_norm": 0.07830816633015714, | |
| "learning_rate": 0.00012083336959619802, | |
| "loss": 1.112, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.492519528664107, | |
| "grad_norm": 0.0756022313506664, | |
| "learning_rate": 0.00011992823687623977, | |
| "loss": 1.122, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.4951674831192904, | |
| "grad_norm": 0.07163741166179009, | |
| "learning_rate": 0.00011902140074790112, | |
| "loss": 1.0669, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.4978154375744737, | |
| "grad_norm": 0.08576324173799038, | |
| "learning_rate": 0.00011811293872492644, | |
| "loss": 1.121, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.5004633920296571, | |
| "grad_norm": 0.0762857316535538, | |
| "learning_rate": 0.00011720292846003706, | |
| "loss": 1.1249, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.5031113464848405, | |
| "grad_norm": 0.07839664604260674, | |
| "learning_rate": 0.00011629144773829344, | |
| "loss": 1.1062, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.5057593009400239, | |
| "grad_norm": 0.08031109603032427, | |
| "learning_rate": 0.00011537857447044656, | |
| "loss": 1.1013, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.5084072553952071, | |
| "grad_norm": 0.07756013246128472, | |
| "learning_rate": 0.00011446438668627819, | |
| "loss": 1.0908, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.5110552098503905, | |
| "grad_norm": 0.07362707242031638, | |
| "learning_rate": 0.00011354896252793117, | |
| "loss": 1.0951, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.513703164305574, | |
| "grad_norm": 0.07249015180621271, | |
| "learning_rate": 0.0001126323802432301, | |
| "loss": 1.0966, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.5163511187607573, | |
| "grad_norm": 0.07541567661268533, | |
| "learning_rate": 0.00011171471817899282, | |
| "loss": 1.1181, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.5189990732159406, | |
| "grad_norm": 0.08071128877259519, | |
| "learning_rate": 0.0001107960547743337, | |
| "loss": 1.0811, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.521647027671124, | |
| "grad_norm": 0.08258197653732914, | |
| "learning_rate": 0.00010987646855395872, | |
| "loss": 1.1417, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.5242949821263074, | |
| "grad_norm": 0.07840010957906275, | |
| "learning_rate": 0.00010895603812145359, | |
| "loss": 1.1253, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.5269429365814908, | |
| "grad_norm": 0.07562413230243624, | |
| "learning_rate": 0.00010803484215256471, | |
| "loss": 1.1022, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.5295908910366741, | |
| "grad_norm": 0.0748436684540439, | |
| "learning_rate": 0.00010711295938847452, | |
| "loss": 1.0963, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5322388454918575, | |
| "grad_norm": 0.07432957206360974, | |
| "learning_rate": 0.00010619046862907057, | |
| "loss": 1.0804, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.5348867999470409, | |
| "grad_norm": 0.07745940840059426, | |
| "learning_rate": 0.00010526744872621013, | |
| "loss": 1.1189, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.5375347544022243, | |
| "grad_norm": 0.07755117796467678, | |
| "learning_rate": 0.00010434397857698025, | |
| "loss": 1.0881, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.5401827088574076, | |
| "grad_norm": 0.07851509890246185, | |
| "learning_rate": 0.00010342013711695356, | |
| "loss": 1.1108, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.542830663312591, | |
| "grad_norm": 0.07480756214232848, | |
| "learning_rate": 0.00010249600331344144, | |
| "loss": 1.1214, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.5454786177677744, | |
| "grad_norm": 0.07388322231422537, | |
| "learning_rate": 0.00010157165615874384, | |
| "loss": 1.1129, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.5481265722229578, | |
| "grad_norm": 0.07729477706022746, | |
| "learning_rate": 0.00010064717466339746, | |
| "loss": 1.1024, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.5507745266781411, | |
| "grad_norm": 0.07418682486286104, | |
| "learning_rate": 9.972263784942191e-05, | |
| "loss": 1.0897, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.5534224811333245, | |
| "grad_norm": 0.07818304753661111, | |
| "learning_rate": 9.879812474356538e-05, | |
| "loss": 1.1209, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.5560704355885079, | |
| "grad_norm": 0.07539756214766863, | |
| "learning_rate": 9.78737143705495e-05, | |
| "loss": 1.0842, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.5587183900436913, | |
| "grad_norm": 0.08000396424013696, | |
| "learning_rate": 9.69494857463147e-05, | |
| "loss": 1.1022, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.5613663444988746, | |
| "grad_norm": 0.0780502131683453, | |
| "learning_rate": 9.602551787126593e-05, | |
| "loss": 1.0828, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.564014298954058, | |
| "grad_norm": 0.0756338343630312, | |
| "learning_rate": 9.510188972352003e-05, | |
| "loss": 1.1017, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.5666622534092414, | |
| "grad_norm": 0.0744886822174254, | |
| "learning_rate": 9.417868025215508e-05, | |
| "loss": 1.0982, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.5693102078644248, | |
| "grad_norm": 0.0743961983677683, | |
| "learning_rate": 9.325596837046177e-05, | |
| "loss": 1.1129, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.571958162319608, | |
| "grad_norm": 0.07437524331159841, | |
| "learning_rate": 9.233383294919832e-05, | |
| "loss": 1.0775, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.5746061167747915, | |
| "grad_norm": 0.0717481378509695, | |
| "learning_rate": 9.141235280984873e-05, | |
| "loss": 1.1017, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.5772540712299749, | |
| "grad_norm": 0.07661308153340447, | |
| "learning_rate": 9.04916067178854e-05, | |
| "loss": 1.0641, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.5799020256851583, | |
| "grad_norm": 0.0770623797115787, | |
| "learning_rate": 8.957167337603652e-05, | |
| "loss": 1.1182, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.5825499801403415, | |
| "grad_norm": 0.07405657746416461, | |
| "learning_rate": 8.865263141755865e-05, | |
| "loss": 1.1057, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.5851979345955249, | |
| "grad_norm": 0.07499634726366083, | |
| "learning_rate": 8.773455939951548e-05, | |
| "loss": 1.091, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.5878458890507083, | |
| "grad_norm": 0.08117162442341291, | |
| "learning_rate": 8.681753579606312e-05, | |
| "loss": 1.125, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.5904938435058917, | |
| "grad_norm": 0.07499243002348535, | |
| "learning_rate": 8.590163899174202e-05, | |
| "loss": 1.0797, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.593141797961075, | |
| "grad_norm": 0.0757499666860512, | |
| "learning_rate": 8.498694727477712e-05, | |
| "loss": 1.101, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.5957897524162584, | |
| "grad_norm": 0.0790409067216223, | |
| "learning_rate": 8.40735388303861e-05, | |
| "loss": 1.1147, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.5984377068714418, | |
| "grad_norm": 0.07791231584194006, | |
| "learning_rate": 8.316149173409604e-05, | |
| "loss": 1.0985, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.6010856613266252, | |
| "grad_norm": 0.07490124978826228, | |
| "learning_rate": 8.225088394506998e-05, | |
| "loss": 1.1174, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 0.6037336157818085, | |
| "grad_norm": 0.07693659200902671, | |
| "learning_rate": 8.134179329944308e-05, | |
| "loss": 1.1346, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.6063815702369919, | |
| "grad_norm": 0.07605699796790699, | |
| "learning_rate": 8.043429750366941e-05, | |
| "loss": 1.1001, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 0.6090295246921753, | |
| "grad_norm": 0.07462322261103259, | |
| "learning_rate": 7.952847412787997e-05, | |
| "loss": 1.0907, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.6116774791473587, | |
| "grad_norm": 0.07830257859960496, | |
| "learning_rate": 7.8624400599252e-05, | |
| "loss": 1.079, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.614325433602542, | |
| "grad_norm": 0.07768781511626291, | |
| "learning_rate": 7.772215419539087e-05, | |
| "loss": 1.092, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.6169733880577254, | |
| "grad_norm": 0.07237763916654574, | |
| "learning_rate": 7.682181203772462e-05, | |
| "loss": 1.1278, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 0.6196213425129088, | |
| "grad_norm": 0.07421572650644807, | |
| "learning_rate": 7.592345108491177e-05, | |
| "loss": 1.0914, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.6222692969680922, | |
| "grad_norm": 0.07647986250910133, | |
| "learning_rate": 7.502714812626306e-05, | |
| "loss": 1.0973, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.6249172514232755, | |
| "grad_norm": 0.07673429125787028, | |
| "learning_rate": 7.413297977517804e-05, | |
| "loss": 1.113, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.6275652058784589, | |
| "grad_norm": 0.07847988781414174, | |
| "learning_rate": 7.324102246259588e-05, | |
| "loss": 1.1106, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.6302131603336423, | |
| "grad_norm": 0.08602822092227703, | |
| "learning_rate": 7.235135243046266e-05, | |
| "loss": 1.0951, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.6328611147888257, | |
| "grad_norm": 0.07679615832799559, | |
| "learning_rate": 7.146404572521436e-05, | |
| "loss": 1.0822, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 0.635509069244009, | |
| "grad_norm": 0.07328313009689218, | |
| "learning_rate": 7.057917819127646e-05, | |
| "loss": 1.1144, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.6381570236991924, | |
| "grad_norm": 0.07561356041759124, | |
| "learning_rate": 6.969682546458119e-05, | |
| "loss": 1.0949, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 0.6408049781543758, | |
| "grad_norm": 0.07601697316878776, | |
| "learning_rate": 6.881706296610222e-05, | |
| "loss": 1.0881, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.6434529326095592, | |
| "grad_norm": 0.07424508033316624, | |
| "learning_rate": 6.793996589540794e-05, | |
| "loss": 1.0894, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 0.6461008870647424, | |
| "grad_norm": 0.0774103399193041, | |
| "learning_rate": 6.706560922423375e-05, | |
| "loss": 1.0889, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.6487488415199258, | |
| "grad_norm": 0.07704030882428428, | |
| "learning_rate": 6.619406769007347e-05, | |
| "loss": 1.1173, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.6513967959751092, | |
| "grad_norm": 0.07986295404405998, | |
| "learning_rate": 6.532541578979114e-05, | |
| "loss": 1.0828, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.6540447504302926, | |
| "grad_norm": 0.08205457004998916, | |
| "learning_rate": 6.445972777325322e-05, | |
| "loss": 1.1107, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 0.6566927048854759, | |
| "grad_norm": 0.07502810371034836, | |
| "learning_rate": 6.359707763698192e-05, | |
| "loss": 1.0947, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.6593406593406593, | |
| "grad_norm": 0.07480668675284598, | |
| "learning_rate": 6.27375391178301e-05, | |
| "loss": 1.129, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 0.6619886137958427, | |
| "grad_norm": 0.075324303274979, | |
| "learning_rate": 6.188118568667869e-05, | |
| "loss": 1.1049, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.6646365682510261, | |
| "grad_norm": 0.07716903325547174, | |
| "learning_rate": 6.102809054215626e-05, | |
| "loss": 1.0853, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 0.6672845227062094, | |
| "grad_norm": 0.08902130376105476, | |
| "learning_rate": 6.017832660438261e-05, | |
| "loss": 1.1014, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.6699324771613928, | |
| "grad_norm": 0.07427630543595487, | |
| "learning_rate": 5.9331966508735416e-05, | |
| "loss": 1.0715, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 0.6725804316165762, | |
| "grad_norm": 0.07483794256035971, | |
| "learning_rate": 5.848908259964176e-05, | |
| "loss": 1.1228, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.6752283860717596, | |
| "grad_norm": 0.07249249950069708, | |
| "learning_rate": 5.7649746924394374e-05, | |
| "loss": 1.1036, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.6778763405269429, | |
| "grad_norm": 0.07406719285536154, | |
| "learning_rate": 5.681403122699306e-05, | |
| "loss": 1.0975, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.6805242949821263, | |
| "grad_norm": 0.07322431567652883, | |
| "learning_rate": 5.598200694201228e-05, | |
| "loss": 1.1062, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 0.6831722494373097, | |
| "grad_norm": 0.07815522826559648, | |
| "learning_rate": 5.5153745188495395e-05, | |
| "loss": 1.0813, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.6858202038924931, | |
| "grad_norm": 0.07426469053250764, | |
| "learning_rate": 5.432931676387516e-05, | |
| "loss": 1.104, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 0.6884681583476764, | |
| "grad_norm": 0.08432084820000514, | |
| "learning_rate": 5.350879213792255e-05, | |
| "loss": 1.1244, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.6911161128028598, | |
| "grad_norm": 0.07573346123023687, | |
| "learning_rate": 5.2692241446723003e-05, | |
| "loss": 1.1084, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 0.6937640672580432, | |
| "grad_norm": 0.07843026074586353, | |
| "learning_rate": 5.18797344866814e-05, | |
| "loss": 1.102, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.6964120217132266, | |
| "grad_norm": 0.07870397238308142, | |
| "learning_rate": 5.107134070855617e-05, | |
| "loss": 1.0843, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 0.6990599761684099, | |
| "grad_norm": 0.07508751345932353, | |
| "learning_rate": 5.0267129211522834e-05, | |
| "loss": 1.0998, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.7017079306235933, | |
| "grad_norm": 0.07699932244077094, | |
| "learning_rate": 4.9467168737267376e-05, | |
| "loss": 1.1199, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.7043558850787767, | |
| "grad_norm": 0.07341861390300569, | |
| "learning_rate": 4.867152766411075e-05, | |
| "loss": 1.1099, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.7070038395339601, | |
| "grad_norm": 0.07886199590243474, | |
| "learning_rate": 4.788027400116387e-05, | |
| "loss": 1.1213, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 0.7096517939891434, | |
| "grad_norm": 0.07400224087388893, | |
| "learning_rate": 4.7093475382514343e-05, | |
| "loss": 1.1089, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.7122997484443268, | |
| "grad_norm": 0.07628400003826659, | |
| "learning_rate": 4.631119906144562e-05, | |
| "loss": 1.1101, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 0.7149477028995102, | |
| "grad_norm": 0.07655029779945569, | |
| "learning_rate": 4.5533511904687977e-05, | |
| "loss": 1.1464, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.7175956573546936, | |
| "grad_norm": 0.07408511396805334, | |
| "learning_rate": 4.4760480386703085e-05, | |
| "loss": 1.1036, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 0.7202436118098768, | |
| "grad_norm": 0.07722020217908175, | |
| "learning_rate": 4.399217058400225e-05, | |
| "loss": 1.1229, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.7228915662650602, | |
| "grad_norm": 0.08311731202657702, | |
| "learning_rate": 4.322864816949785e-05, | |
| "loss": 1.111, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 0.7255395207202436, | |
| "grad_norm": 0.07589827212004248, | |
| "learning_rate": 4.2469978406890265e-05, | |
| "loss": 1.1205, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.728187475175427, | |
| "grad_norm": 0.07793858759229856, | |
| "learning_rate": 4.171622614508911e-05, | |
| "loss": 1.1104, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.7308354296306103, | |
| "grad_norm": 0.07696199883343936, | |
| "learning_rate": 4.0967455812670095e-05, | |
| "loss": 1.1055, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.7334833840857937, | |
| "grad_norm": 0.07532988662602433, | |
| "learning_rate": 4.022373141236801e-05, | |
| "loss": 1.1052, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 0.7361313385409771, | |
| "grad_norm": 0.07570188583228322, | |
| "learning_rate": 3.948511651560593e-05, | |
| "loss": 1.0996, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.7387792929961605, | |
| "grad_norm": 0.07840832338236073, | |
| "learning_rate": 3.87516742570611e-05, | |
| "loss": 1.0989, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 0.7414272474513438, | |
| "grad_norm": 0.07688842225193993, | |
| "learning_rate": 3.802346732926866e-05, | |
| "loss": 1.094, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.7440752019065272, | |
| "grad_norm": 0.07667453813475356, | |
| "learning_rate": 3.730055797726271e-05, | |
| "loss": 1.1138, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 0.7467231563617106, | |
| "grad_norm": 0.07256620195156938, | |
| "learning_rate": 3.658300799325568e-05, | |
| "loss": 1.096, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.749371110816894, | |
| "grad_norm": 0.07556178320717383, | |
| "learning_rate": 3.5870878711356906e-05, | |
| "loss": 1.0752, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 0.7520190652720773, | |
| "grad_norm": 0.0755712488652408, | |
| "learning_rate": 3.516423100232943e-05, | |
| "loss": 1.0738, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.7546670197272607, | |
| "grad_norm": 0.07976556756122673, | |
| "learning_rate": 3.446312526838726e-05, | |
| "loss": 1.0695, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.7573149741824441, | |
| "grad_norm": 0.07904402049081768, | |
| "learning_rate": 3.376762143803249e-05, | |
| "loss": 1.0979, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.7599629286376274, | |
| "grad_norm": 0.07776446642649297, | |
| "learning_rate": 3.307777896093238e-05, | |
| "loss": 1.1018, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 0.7626108830928108, | |
| "grad_norm": 0.07527759382185836, | |
| "learning_rate": 3.239365680283812e-05, | |
| "loss": 1.0968, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.7652588375479942, | |
| "grad_norm": 0.0776648894196293, | |
| "learning_rate": 3.171531344054448e-05, | |
| "loss": 1.0907, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 0.7679067920031776, | |
| "grad_norm": 0.07845291586662016, | |
| "learning_rate": 3.104280685689128e-05, | |
| "loss": 1.0988, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.7705547464583609, | |
| "grad_norm": 0.07774267939899475, | |
| "learning_rate": 3.037619453580741e-05, | |
| "loss": 1.1025, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 0.7732027009135443, | |
| "grad_norm": 0.07875923469354247, | |
| "learning_rate": 2.9715533457397115e-05, | |
| "loss": 1.1106, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.7758506553687277, | |
| "grad_norm": 0.07692024929981921, | |
| "learning_rate": 2.906088009306942e-05, | |
| "loss": 1.0852, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 0.7784986098239111, | |
| "grad_norm": 0.07785685577501977, | |
| "learning_rate": 2.841229040071145e-05, | |
| "loss": 1.0682, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.7811465642790943, | |
| "grad_norm": 0.0760319263173443, | |
| "learning_rate": 2.7769819819904876e-05, | |
| "loss": 1.1018, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.7837945187342777, | |
| "grad_norm": 0.076129060223867, | |
| "learning_rate": 2.7133523267187344e-05, | |
| "loss": 1.1027, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.7864424731894611, | |
| "grad_norm": 0.07951511310110099, | |
| "learning_rate": 2.650345513135849e-05, | |
| "loss": 1.1047, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 0.7890904276446445, | |
| "grad_norm": 0.0766458764221786, | |
| "learning_rate": 2.587966926883063e-05, | |
| "loss": 1.1019, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.7917383820998278, | |
| "grad_norm": 0.07749498443053499, | |
| "learning_rate": 2.526221899902539e-05, | |
| "loss": 1.1006, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 0.7943863365550112, | |
| "grad_norm": 0.08039092314611351, | |
| "learning_rate": 2.465115709981636e-05, | |
| "loss": 1.0787, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.7970342910101946, | |
| "grad_norm": 0.07771691438169755, | |
| "learning_rate": 2.4046535803017344e-05, | |
| "loss": 1.1049, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 0.799682245465378, | |
| "grad_norm": 0.07950511536169985, | |
| "learning_rate": 2.3448406789918133e-05, | |
| "loss": 1.0991, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.8023301999205613, | |
| "grad_norm": 0.08027602014187285, | |
| "learning_rate": 2.285682118686674e-05, | |
| "loss": 1.1072, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 0.8049781543757447, | |
| "grad_norm": 0.07696693283704584, | |
| "learning_rate": 2.227182956089924e-05, | |
| "loss": 1.0927, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.8076261088309281, | |
| "grad_norm": 0.07963101925667078, | |
| "learning_rate": 2.169348191541757e-05, | |
| "loss": 1.0934, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.8102740632861115, | |
| "grad_norm": 0.08049222177031838, | |
| "learning_rate": 2.112182768591534e-05, | |
| "loss": 1.1383, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.8129220177412948, | |
| "grad_norm": 0.07530307843428516, | |
| "learning_rate": 2.0556915735752114e-05, | |
| "loss": 1.0875, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 0.8155699721964782, | |
| "grad_norm": 0.0773611919685891, | |
| "learning_rate": 1.9998794351976958e-05, | |
| "loss": 1.107, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.8182179266516616, | |
| "grad_norm": 0.07662227830919144, | |
| "learning_rate": 1.9447511241200722e-05, | |
| "loss": 1.1042, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 0.820865881106845, | |
| "grad_norm": 0.07670312386682876, | |
| "learning_rate": 1.8903113525518334e-05, | |
| "loss": 1.0996, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.8235138355620283, | |
| "grad_norm": 0.07637261789090262, | |
| "learning_rate": 1.836564773848112e-05, | |
| "loss": 1.1056, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 0.8261617900172117, | |
| "grad_norm": 0.0778692527235825, | |
| "learning_rate": 1.783515982111885e-05, | |
| "loss": 1.1013, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.8288097444723951, | |
| "grad_norm": 0.07646896656564574, | |
| "learning_rate": 1.7311695118013237e-05, | |
| "loss": 1.123, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 0.8314576989275785, | |
| "grad_norm": 0.07814676271315506, | |
| "learning_rate": 1.6795298373421776e-05, | |
| "loss": 1.1167, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.8341056533827618, | |
| "grad_norm": 0.07623300427172139, | |
| "learning_rate": 1.6286013727453154e-05, | |
| "loss": 1.1107, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.8367536078379452, | |
| "grad_norm": 0.0770697500075822, | |
| "learning_rate": 1.5783884712294362e-05, | |
| "loss": 1.083, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.8394015622931286, | |
| "grad_norm": 0.07692133101024615, | |
| "learning_rate": 1.5288954248489685e-05, | |
| "loss": 1.1099, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 0.842049516748312, | |
| "grad_norm": 0.0755198202012927, | |
| "learning_rate": 1.4801264641271827e-05, | |
| "loss": 1.0753, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.8446974712034953, | |
| "grad_norm": 0.07515799406054297, | |
| "learning_rate": 1.4320857576946e-05, | |
| "loss": 1.1014, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 0.8473454256586787, | |
| "grad_norm": 0.07479012037004316, | |
| "learning_rate": 1.3847774119326573e-05, | |
| "loss": 1.0942, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.849993380113862, | |
| "grad_norm": 0.07721255980030046, | |
| "learning_rate": 1.3382054706226987e-05, | |
| "loss": 1.1133, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 0.8526413345690455, | |
| "grad_norm": 0.07642790610807988, | |
| "learning_rate": 1.2923739146003489e-05, | |
| "loss": 1.0951, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.8552892890242287, | |
| "grad_norm": 0.07855286960878372, | |
| "learning_rate": 1.2472866614152124e-05, | |
| "loss": 1.0965, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 0.8579372434794121, | |
| "grad_norm": 0.07681456927547702, | |
| "learning_rate": 1.2029475649960353e-05, | |
| "loss": 1.0866, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.8605851979345955, | |
| "grad_norm": 0.07738703516789706, | |
| "learning_rate": 1.1593604153212767e-05, | |
| "loss": 1.1066, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 0.8632331523897789, | |
| "grad_norm": 0.07595140973388262, | |
| "learning_rate": 1.1165289380951416e-05, | |
| "loss": 1.1071, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.8658811068449622, | |
| "grad_norm": 0.07583738961687224, | |
| "learning_rate": 1.074456794429135e-05, | |
| "loss": 1.1228, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 0.8685290613001456, | |
| "grad_norm": 0.0759554158788922, | |
| "learning_rate": 1.0331475805291136e-05, | |
| "loss": 1.1145, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.871177015755329, | |
| "grad_norm": 0.07558370365509604, | |
| "learning_rate": 9.926048273878819e-06, | |
| "loss": 1.0867, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 0.8738249702105124, | |
| "grad_norm": 0.07447741708164467, | |
| "learning_rate": 9.528320004833935e-06, | |
| "loss": 1.0982, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.8764729246656957, | |
| "grad_norm": 0.07741539189144307, | |
| "learning_rate": 9.138324994825186e-06, | |
| "loss": 1.0949, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 0.8791208791208791, | |
| "grad_norm": 0.07536616901375505, | |
| "learning_rate": 8.75609657950448e-06, | |
| "loss": 1.0941, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.8817688335760625, | |
| "grad_norm": 0.07569809049947313, | |
| "learning_rate": 8.381667430657603e-06, | |
| "loss": 1.0907, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 0.8844167880312459, | |
| "grad_norm": 0.07776927737637536, | |
| "learning_rate": 8.015069553411469e-06, | |
| "loss": 1.1134, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.8870647424864292, | |
| "grad_norm": 0.07389019327060092, | |
| "learning_rate": 7.656334283498334e-06, | |
| "loss": 1.0757, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 0.8897126969416126, | |
| "grad_norm": 0.07792778854648327, | |
| "learning_rate": 7.305492284577531e-06, | |
| "loss": 1.0795, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.892360651396796, | |
| "grad_norm": 0.07609958493651112, | |
| "learning_rate": 6.962573545614159e-06, | |
| "loss": 1.0975, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 0.8950086058519794, | |
| "grad_norm": 0.07556244044567743, | |
| "learning_rate": 6.627607378315903e-06, | |
| "loss": 1.1235, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.8976565603071627, | |
| "grad_norm": 0.07801379209812083, | |
| "learning_rate": 6.3006224146274994e-06, | |
| "loss": 1.1113, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 0.9003045147623461, | |
| "grad_norm": 0.07594137231580138, | |
| "learning_rate": 5.9816466042833085e-06, | |
| "loss": 1.1, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.9029524692175295, | |
| "grad_norm": 0.0765503105906239, | |
| "learning_rate": 5.670707212418313e-06, | |
| "loss": 1.0999, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 0.9056004236727129, | |
| "grad_norm": 0.0784364193602198, | |
| "learning_rate": 5.367830817237596e-06, | |
| "loss": 1.0789, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.9082483781278962, | |
| "grad_norm": 0.07687073522520732, | |
| "learning_rate": 5.073043307744396e-06, | |
| "loss": 1.0941, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 0.9108963325830796, | |
| "grad_norm": 0.07581921026660658, | |
| "learning_rate": 4.7863698815273195e-06, | |
| "loss": 1.1097, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.913544287038263, | |
| "grad_norm": 0.07789236711738227, | |
| "learning_rate": 4.507835042606456e-06, | |
| "loss": 1.0713, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 0.9161922414934464, | |
| "grad_norm": 0.07967848637291239, | |
| "learning_rate": 4.237462599338837e-06, | |
| "loss": 1.114, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.9188401959486296, | |
| "grad_norm": 0.07753641186277775, | |
| "learning_rate": 3.975275662383404e-06, | |
| "loss": 1.086, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 0.921488150403813, | |
| "grad_norm": 0.07788205466367802, | |
| "learning_rate": 3.721296642725569e-06, | |
| "loss": 1.1076, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.9241361048589964, | |
| "grad_norm": 0.07489418836802078, | |
| "learning_rate": 3.475547249761557e-06, | |
| "loss": 1.0876, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 0.9267840593141798, | |
| "grad_norm": 0.07633053121238625, | |
| "learning_rate": 3.23804848944278e-06, | |
| "loss": 1.1186, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.9294320137693631, | |
| "grad_norm": 0.0749545234975198, | |
| "learning_rate": 3.0088206624802716e-06, | |
| "loss": 1.1176, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 0.9320799682245465, | |
| "grad_norm": 0.07806751123666335, | |
| "learning_rate": 2.7878833626095224e-06, | |
| "loss": 1.1104, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.9347279226797299, | |
| "grad_norm": 0.07350877409736523, | |
| "learning_rate": 2.575255474915561e-06, | |
| "loss": 1.1033, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 0.9373758771349133, | |
| "grad_norm": 0.07754465442688381, | |
| "learning_rate": 2.3709551742187586e-06, | |
| "loss": 1.0918, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.9400238315900966, | |
| "grad_norm": 0.07943198880463362, | |
| "learning_rate": 2.1749999235213147e-06, | |
| "loss": 1.0957, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 0.94267178604528, | |
| "grad_norm": 0.0738980478341624, | |
| "learning_rate": 1.9874064725145746e-06, | |
| "loss": 1.0965, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.9453197405004634, | |
| "grad_norm": 0.07568301411412179, | |
| "learning_rate": 1.8081908561472406e-06, | |
| "loss": 1.0928, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 0.9479676949556468, | |
| "grad_norm": 0.07808219875357872, | |
| "learning_rate": 1.6373683932548455e-06, | |
| "loss": 1.0951, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.9506156494108301, | |
| "grad_norm": 0.0768674233407858, | |
| "learning_rate": 1.4749536852502687e-06, | |
| "loss": 1.0992, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 0.9532636038660135, | |
| "grad_norm": 0.0768177210422567, | |
| "learning_rate": 1.3209606148756993e-06, | |
| "loss": 1.1247, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.9559115583211969, | |
| "grad_norm": 0.076755922231129, | |
| "learning_rate": 1.1754023450159523e-06, | |
| "loss": 1.0955, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 0.9585595127763803, | |
| "grad_norm": 0.07587714541337397, | |
| "learning_rate": 1.0382913175733587e-06, | |
| "loss": 1.1138, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.9612074672315636, | |
| "grad_norm": 0.07977716829550086, | |
| "learning_rate": 9.096392524042374e-07, | |
| "loss": 1.1018, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 0.963855421686747, | |
| "grad_norm": 0.07805623190906948, | |
| "learning_rate": 7.894571463171851e-07, | |
| "loss": 1.0995, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.9665033761419304, | |
| "grad_norm": 0.07827918723132034, | |
| "learning_rate": 6.77755272132985e-07, | |
| "loss": 1.1072, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 0.9691513305971138, | |
| "grad_norm": 0.07504095329870522, | |
| "learning_rate": 5.745431778066412e-07, | |
| "loss": 1.0945, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.9717992850522971, | |
| "grad_norm": 0.07543686944798814, | |
| "learning_rate": 4.798296856111772e-07, | |
| "loss": 1.1205, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 0.9744472395074805, | |
| "grad_norm": 0.0775313859719524, | |
| "learning_rate": 3.9362289138351604e-07, | |
| "loss": 1.0969, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.9770951939626639, | |
| "grad_norm": 0.08399677573379252, | |
| "learning_rate": 3.15930163832534e-07, | |
| "loss": 1.1326, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 0.9797431484178473, | |
| "grad_norm": 0.08167978103101492, | |
| "learning_rate": 2.4675814390916483e-07, | |
| "loss": 1.1006, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.9823911028730306, | |
| "grad_norm": 0.07627980254031101, | |
| "learning_rate": 1.8611274423874226e-07, | |
| "loss": 1.1026, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 0.985039057328214, | |
| "grad_norm": 0.07798059783576732, | |
| "learning_rate": 1.339991486156489e-07, | |
| "loss": 1.112, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.9876870117833974, | |
| "grad_norm": 0.07847559734854997, | |
| "learning_rate": 9.04218115601485e-08, | |
| "loss": 1.0969, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 0.9903349662385807, | |
| "grad_norm": 0.07602247898283418, | |
| "learning_rate": 5.5384457937723843e-08, | |
| "loss": 1.0857, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.992982920693764, | |
| "grad_norm": 0.07830104923923938, | |
| "learning_rate": 2.8890082640598004e-08, | |
| "loss": 1.0924, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 0.9956308751489474, | |
| "grad_norm": 0.0737777189572401, | |
| "learning_rate": 1.0940950331772559e-08, | |
| "loss": 1.0997, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.9982788296041308, | |
| "grad_norm": 0.07376783137923695, | |
| "learning_rate": 1.5385952514712644e-09, | |
| "loss": 1.1069, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 0.9998676022772408, | |
| "eval_loss": 1.0810853242874146, | |
| "eval_runtime": 4476.4649, | |
| "eval_samples_per_second": 3.006, | |
| "eval_steps_per_second": 0.752, | |
| "step": 1888 | |
| }, | |
| { | |
| "epoch": 0.9998676022772408, | |
| "step": 1888, | |
| "total_flos": 2.3745805538033664e+16, | |
| "train_loss": 1.1182949897848953, | |
| "train_runtime": 43452.0384, | |
| "train_samples_per_second": 2.781, | |
| "train_steps_per_second": 0.043 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1888, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.3745805538033664e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |