| { | |
| "best_global_step": 2500, | |
| "best_metric": 8.384208161162686, | |
| "best_model_checkpoint": "./SALAMA_NEWMED2/checkpoint-2500", | |
| "epoch": 0.9987016878058524, | |
| "eval_steps": 2500, | |
| "global_step": 2500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00399480675122341, | |
| "grad_norm": 4.626885890960693, | |
| "learning_rate": 1.8e-07, | |
| "loss": 0.157, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.00798961350244682, | |
| "grad_norm": 4.6383256912231445, | |
| "learning_rate": 3.8e-07, | |
| "loss": 0.1612, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.011984420253670228, | |
| "grad_norm": 4.137997627258301, | |
| "learning_rate": 5.800000000000001e-07, | |
| "loss": 0.1601, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.01597922700489364, | |
| "grad_norm": 4.052377223968506, | |
| "learning_rate": 7.8e-07, | |
| "loss": 0.166, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.019974033756117046, | |
| "grad_norm": 4.292566299438477, | |
| "learning_rate": 9.800000000000001e-07, | |
| "loss": 0.1519, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.023968840507340456, | |
| "grad_norm": 5.116729736328125, | |
| "learning_rate": 1.1800000000000001e-06, | |
| "loss": 0.1697, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.027963647258563866, | |
| "grad_norm": 4.941128253936768, | |
| "learning_rate": 1.3800000000000001e-06, | |
| "loss": 0.1338, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.03195845400978728, | |
| "grad_norm": 4.003450870513916, | |
| "learning_rate": 1.5800000000000001e-06, | |
| "loss": 0.1506, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.035953260761010686, | |
| "grad_norm": 4.099050521850586, | |
| "learning_rate": 1.7800000000000001e-06, | |
| "loss": 0.128, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.03994806751223409, | |
| "grad_norm": 3.813715696334839, | |
| "learning_rate": 1.98e-06, | |
| "loss": 0.1519, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.043942874263457506, | |
| "grad_norm": 3.1228017807006836, | |
| "learning_rate": 2.1800000000000003e-06, | |
| "loss": 0.1602, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.04793768101468091, | |
| "grad_norm": 3.2234630584716797, | |
| "learning_rate": 2.38e-06, | |
| "loss": 0.1573, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.051932487765904325, | |
| "grad_norm": 3.196345806121826, | |
| "learning_rate": 2.5800000000000003e-06, | |
| "loss": 0.1594, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.05592729451712773, | |
| "grad_norm": 3.428943634033203, | |
| "learning_rate": 2.7800000000000005e-06, | |
| "loss": 0.1528, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.059922101268351145, | |
| "grad_norm": 4.3613505363464355, | |
| "learning_rate": 2.9800000000000003e-06, | |
| "loss": 0.1764, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.06391690801957456, | |
| "grad_norm": 3.6310315132141113, | |
| "learning_rate": 3.1800000000000005e-06, | |
| "loss": 0.1472, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.06791171477079797, | |
| "grad_norm": 4.561097621917725, | |
| "learning_rate": 3.3800000000000007e-06, | |
| "loss": 0.158, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.07190652152202137, | |
| "grad_norm": 2.8412837982177734, | |
| "learning_rate": 3.58e-06, | |
| "loss": 0.1561, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.07590132827324478, | |
| "grad_norm": 3.9358673095703125, | |
| "learning_rate": 3.7800000000000002e-06, | |
| "loss": 0.1682, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.07989613502446818, | |
| "grad_norm": 3.9661779403686523, | |
| "learning_rate": 3.980000000000001e-06, | |
| "loss": 0.1375, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.0838909417756916, | |
| "grad_norm": 4.195302963256836, | |
| "learning_rate": 4.18e-06, | |
| "loss": 0.1557, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.08788574852691501, | |
| "grad_norm": 4.319852352142334, | |
| "learning_rate": 4.38e-06, | |
| "loss": 0.1828, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.09188055527813842, | |
| "grad_norm": 3.6552112102508545, | |
| "learning_rate": 4.58e-06, | |
| "loss": 0.1814, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.09587536202936182, | |
| "grad_norm": 4.2657365798950195, | |
| "learning_rate": 4.78e-06, | |
| "loss": 0.1829, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.09987016878058524, | |
| "grad_norm": 4.424646377563477, | |
| "learning_rate": 4.980000000000001e-06, | |
| "loss": 0.1734, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.10386497553180865, | |
| "grad_norm": 3.4387409687042236, | |
| "learning_rate": 5.18e-06, | |
| "loss": 0.1548, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.10785978228303206, | |
| "grad_norm": 3.510521173477173, | |
| "learning_rate": 5.380000000000001e-06, | |
| "loss": 0.1401, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.11185458903425546, | |
| "grad_norm": 3.2406973838806152, | |
| "learning_rate": 5.580000000000001e-06, | |
| "loss": 0.1573, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.11584939578547888, | |
| "grad_norm": 3.9534926414489746, | |
| "learning_rate": 5.78e-06, | |
| "loss": 0.1768, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.11984420253670229, | |
| "grad_norm": 4.111250400543213, | |
| "learning_rate": 5.98e-06, | |
| "loss": 0.1754, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.1238390092879257, | |
| "grad_norm": 3.7990970611572266, | |
| "learning_rate": 6.18e-06, | |
| "loss": 0.1594, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.12783381603914912, | |
| "grad_norm": 3.062520980834961, | |
| "learning_rate": 6.380000000000001e-06, | |
| "loss": 0.1506, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.13182862279037252, | |
| "grad_norm": 4.121057987213135, | |
| "learning_rate": 6.5800000000000005e-06, | |
| "loss": 0.1713, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.13582342954159593, | |
| "grad_norm": 3.127715826034546, | |
| "learning_rate": 6.780000000000001e-06, | |
| "loss": 0.1598, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.13981823629281934, | |
| "grad_norm": 3.2151925563812256, | |
| "learning_rate": 6.98e-06, | |
| "loss": 0.1491, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.14381304304404274, | |
| "grad_norm": 2.5872230529785156, | |
| "learning_rate": 7.180000000000001e-06, | |
| "loss": 0.1691, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.14780784979526615, | |
| "grad_norm": 4.3994059562683105, | |
| "learning_rate": 7.3800000000000005e-06, | |
| "loss": 0.1867, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.15180265654648956, | |
| "grad_norm": 4.190740585327148, | |
| "learning_rate": 7.58e-06, | |
| "loss": 0.1752, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.15579746329771296, | |
| "grad_norm": 4.055376052856445, | |
| "learning_rate": 7.78e-06, | |
| "loss": 0.1728, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.15979227004893637, | |
| "grad_norm": 4.507293224334717, | |
| "learning_rate": 7.980000000000002e-06, | |
| "loss": 0.1755, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.1637870768001598, | |
| "grad_norm": 3.873270034790039, | |
| "learning_rate": 8.18e-06, | |
| "loss": 0.1678, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.1677818835513832, | |
| "grad_norm": 4.0611491203308105, | |
| "learning_rate": 8.380000000000001e-06, | |
| "loss": 0.1682, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.17177669030260662, | |
| "grad_norm": 3.7498176097869873, | |
| "learning_rate": 8.580000000000001e-06, | |
| "loss": 0.1621, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.17577149705383002, | |
| "grad_norm": 2.854010581970215, | |
| "learning_rate": 8.78e-06, | |
| "loss": 0.1514, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.17976630380505343, | |
| "grad_norm": 3.098860025405884, | |
| "learning_rate": 8.98e-06, | |
| "loss": 0.1386, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.18376111055627684, | |
| "grad_norm": 5.511478424072266, | |
| "learning_rate": 9.180000000000002e-06, | |
| "loss": 0.176, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.18775591730750024, | |
| "grad_norm": 3.8882782459259033, | |
| "learning_rate": 9.38e-06, | |
| "loss": 0.1742, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.19175072405872365, | |
| "grad_norm": 5.161660194396973, | |
| "learning_rate": 9.58e-06, | |
| "loss": 0.1495, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.19574553080994708, | |
| "grad_norm": 3.9864652156829834, | |
| "learning_rate": 9.780000000000001e-06, | |
| "loss": 0.1869, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.1997403375611705, | |
| "grad_norm": 3.3368163108825684, | |
| "learning_rate": 9.980000000000001e-06, | |
| "loss": 0.1589, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.2037351443123939, | |
| "grad_norm": 4.298588752746582, | |
| "learning_rate": 9.98716486023959e-06, | |
| "loss": 0.1675, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.2077299510636173, | |
| "grad_norm": 3.3363137245178223, | |
| "learning_rate": 9.972903593839133e-06, | |
| "loss": 0.186, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.2117247578148407, | |
| "grad_norm": 3.345853090286255, | |
| "learning_rate": 9.958642327438678e-06, | |
| "loss": 0.1736, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.21571956456606411, | |
| "grad_norm": 3.282945156097412, | |
| "learning_rate": 9.944381061038221e-06, | |
| "loss": 0.1711, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.21971437131728752, | |
| "grad_norm": 5.176159858703613, | |
| "learning_rate": 9.930119794637765e-06, | |
| "loss": 0.2024, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.22370917806851093, | |
| "grad_norm": 3.2021000385284424, | |
| "learning_rate": 9.91585852823731e-06, | |
| "loss": 0.1515, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.22770398481973433, | |
| "grad_norm": 3.4485816955566406, | |
| "learning_rate": 9.901597261836851e-06, | |
| "loss": 0.1673, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.23169879157095777, | |
| "grad_norm": 3.681199789047241, | |
| "learning_rate": 9.887335995436396e-06, | |
| "loss": 0.1713, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.23569359832218117, | |
| "grad_norm": 4.410940647125244, | |
| "learning_rate": 9.87307472903594e-06, | |
| "loss": 0.1653, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.23968840507340458, | |
| "grad_norm": 3.284226655960083, | |
| "learning_rate": 9.858813462635483e-06, | |
| "loss": 0.1442, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.243683211824628, | |
| "grad_norm": 3.73500394821167, | |
| "learning_rate": 9.844552196235026e-06, | |
| "loss": 0.1851, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.2476780185758514, | |
| "grad_norm": 4.125741958618164, | |
| "learning_rate": 9.83029092983457e-06, | |
| "loss": 0.1965, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.2516728253270748, | |
| "grad_norm": 4.371180057525635, | |
| "learning_rate": 9.816029663434114e-06, | |
| "loss": 0.1623, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.25566763207829823, | |
| "grad_norm": 5.782146453857422, | |
| "learning_rate": 9.801768397033657e-06, | |
| "loss": 0.1821, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.2596624388295216, | |
| "grad_norm": 3.964480400085449, | |
| "learning_rate": 9.787507130633202e-06, | |
| "loss": 0.1726, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.26365724558074505, | |
| "grad_norm": 3.8023924827575684, | |
| "learning_rate": 9.773245864232744e-06, | |
| "loss": 0.1771, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.2676520523319684, | |
| "grad_norm": 3.616413116455078, | |
| "learning_rate": 9.758984597832289e-06, | |
| "loss": 0.1913, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.27164685908319186, | |
| "grad_norm": 3.1415390968322754, | |
| "learning_rate": 9.744723331431832e-06, | |
| "loss": 0.1783, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.27564166583441524, | |
| "grad_norm": 3.9891934394836426, | |
| "learning_rate": 9.730462065031375e-06, | |
| "loss": 0.1686, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.2796364725856387, | |
| "grad_norm": 3.3664565086364746, | |
| "learning_rate": 9.71620079863092e-06, | |
| "loss": 0.1928, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.28363127933686205, | |
| "grad_norm": 3.149634599685669, | |
| "learning_rate": 9.701939532230463e-06, | |
| "loss": 0.1839, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.2876260860880855, | |
| "grad_norm": 3.469912052154541, | |
| "learning_rate": 9.687678265830007e-06, | |
| "loss": 0.1699, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.2916208928393089, | |
| "grad_norm": 4.075517654418945, | |
| "learning_rate": 9.67341699942955e-06, | |
| "loss": 0.1807, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.2956156995905323, | |
| "grad_norm": 3.139695405960083, | |
| "learning_rate": 9.659155733029095e-06, | |
| "loss": 0.1618, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.29961050634175573, | |
| "grad_norm": 4.296646595001221, | |
| "learning_rate": 9.644894466628636e-06, | |
| "loss": 0.1912, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.3036053130929791, | |
| "grad_norm": 4.1294379234313965, | |
| "learning_rate": 9.630633200228181e-06, | |
| "loss": 0.1559, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.30760011984420255, | |
| "grad_norm": 3.1996731758117676, | |
| "learning_rate": 9.616371933827725e-06, | |
| "loss": 0.1868, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.3115949265954259, | |
| "grad_norm": 3.0223231315612793, | |
| "learning_rate": 9.602110667427268e-06, | |
| "loss": 0.176, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.31558973334664936, | |
| "grad_norm": 3.7960736751556396, | |
| "learning_rate": 9.587849401026813e-06, | |
| "loss": 0.1617, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.31958454009787274, | |
| "grad_norm": 3.7317583560943604, | |
| "learning_rate": 9.573588134626356e-06, | |
| "loss": 0.1461, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.32357934684909617, | |
| "grad_norm": 2.9020590782165527, | |
| "learning_rate": 9.5593268682259e-06, | |
| "loss": 0.1498, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.3275741536003196, | |
| "grad_norm": 3.0017457008361816, | |
| "learning_rate": 9.545065601825442e-06, | |
| "loss": 0.143, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.331568960351543, | |
| "grad_norm": 2.892198324203491, | |
| "learning_rate": 9.530804335424987e-06, | |
| "loss": 0.1781, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.3355637671027664, | |
| "grad_norm": 3.424417734146118, | |
| "learning_rate": 9.516543069024529e-06, | |
| "loss": 0.1565, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.3395585738539898, | |
| "grad_norm": 3.7211880683898926, | |
| "learning_rate": 9.502281802624074e-06, | |
| "loss": 0.1654, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.34355338060521323, | |
| "grad_norm": 3.9410080909729004, | |
| "learning_rate": 9.488020536223617e-06, | |
| "loss": 0.17, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.3475481873564366, | |
| "grad_norm": 3.70001482963562, | |
| "learning_rate": 9.47375926982316e-06, | |
| "loss": 0.166, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.35154299410766005, | |
| "grad_norm": 4.225161075592041, | |
| "learning_rate": 9.459498003422705e-06, | |
| "loss": 0.1844, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.3555378008588834, | |
| "grad_norm": 3.4117069244384766, | |
| "learning_rate": 9.445236737022249e-06, | |
| "loss": 0.1746, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.35953260761010686, | |
| "grad_norm": 3.593611717224121, | |
| "learning_rate": 9.430975470621792e-06, | |
| "loss": 0.1696, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.3635274143613303, | |
| "grad_norm": 4.341125965118408, | |
| "learning_rate": 9.416714204221335e-06, | |
| "loss": 0.1627, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.36752222111255367, | |
| "grad_norm": 4.328516960144043, | |
| "learning_rate": 9.40245293782088e-06, | |
| "loss": 0.1672, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.3715170278637771, | |
| "grad_norm": 3.9491498470306396, | |
| "learning_rate": 9.388191671420423e-06, | |
| "loss": 0.1741, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.3755118346150005, | |
| "grad_norm": 3.0523922443389893, | |
| "learning_rate": 9.373930405019966e-06, | |
| "loss": 0.1609, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.3795066413662239, | |
| "grad_norm": 3.96618914604187, | |
| "learning_rate": 9.35966913861951e-06, | |
| "loss": 0.1508, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.3835014481174473, | |
| "grad_norm": 3.1924965381622314, | |
| "learning_rate": 9.345407872219053e-06, | |
| "loss": 0.155, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.38749625486867073, | |
| "grad_norm": 4.26977014541626, | |
| "learning_rate": 9.331146605818598e-06, | |
| "loss": 0.1971, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.39149106161989417, | |
| "grad_norm": 3.3444809913635254, | |
| "learning_rate": 9.316885339418141e-06, | |
| "loss": 0.1643, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.39548586837111754, | |
| "grad_norm": 3.90372371673584, | |
| "learning_rate": 9.302624073017684e-06, | |
| "loss": 0.1655, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.399480675122341, | |
| "grad_norm": 3.3378031253814697, | |
| "learning_rate": 9.288362806617228e-06, | |
| "loss": 0.1752, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.40347548187356436, | |
| "grad_norm": 4.0089826583862305, | |
| "learning_rate": 9.274101540216773e-06, | |
| "loss": 0.1616, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.4074702886247878, | |
| "grad_norm": 3.947985887527466, | |
| "learning_rate": 9.259840273816316e-06, | |
| "loss": 0.1576, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.41146509537601117, | |
| "grad_norm": 3.615910768508911, | |
| "learning_rate": 9.245579007415859e-06, | |
| "loss": 0.1516, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.4154599021272346, | |
| "grad_norm": 4.045357704162598, | |
| "learning_rate": 9.231317741015402e-06, | |
| "loss": 0.1795, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.419454708878458, | |
| "grad_norm": 3.5269615650177, | |
| "learning_rate": 9.217056474614946e-06, | |
| "loss": 0.152, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.4234495156296814, | |
| "grad_norm": 3.298372983932495, | |
| "learning_rate": 9.20279520821449e-06, | |
| "loss": 0.1544, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.42744432238090485, | |
| "grad_norm": 3.652780055999756, | |
| "learning_rate": 9.188533941814034e-06, | |
| "loss": 0.1497, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.43143912913212823, | |
| "grad_norm": 3.1677920818328857, | |
| "learning_rate": 9.174272675413579e-06, | |
| "loss": 0.1817, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.43543393588335166, | |
| "grad_norm": 3.229947328567505, | |
| "learning_rate": 9.16001140901312e-06, | |
| "loss": 0.1438, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.43942874263457504, | |
| "grad_norm": 4.125128269195557, | |
| "learning_rate": 9.145750142612665e-06, | |
| "loss": 0.1547, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.4434235493857985, | |
| "grad_norm": 3.456726312637329, | |
| "learning_rate": 9.131488876212208e-06, | |
| "loss": 0.1731, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.44741835613702186, | |
| "grad_norm": 3.4217751026153564, | |
| "learning_rate": 9.117227609811752e-06, | |
| "loss": 0.1505, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.4514131628882453, | |
| "grad_norm": 2.805898666381836, | |
| "learning_rate": 9.102966343411297e-06, | |
| "loss": 0.1712, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.45540796963946867, | |
| "grad_norm": 3.565091133117676, | |
| "learning_rate": 9.08870507701084e-06, | |
| "loss": 0.1914, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.4594027763906921, | |
| "grad_norm": 3.3190317153930664, | |
| "learning_rate": 9.074443810610383e-06, | |
| "loss": 0.1567, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.46339758314191554, | |
| "grad_norm": 3.5254857540130615, | |
| "learning_rate": 9.060182544209926e-06, | |
| "loss": 0.1694, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.4673923898931389, | |
| "grad_norm": 4.437207221984863, | |
| "learning_rate": 9.045921277809471e-06, | |
| "loss": 0.1671, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.47138719664436235, | |
| "grad_norm": 4.538163185119629, | |
| "learning_rate": 9.031660011409013e-06, | |
| "loss": 0.1825, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.47538200339558573, | |
| "grad_norm": 3.300701141357422, | |
| "learning_rate": 9.017398745008558e-06, | |
| "loss": 0.1576, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.47937681014680916, | |
| "grad_norm": 3.468872308731079, | |
| "learning_rate": 9.003137478608101e-06, | |
| "loss": 0.1403, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.48337161689803254, | |
| "grad_norm": 3.9454572200775146, | |
| "learning_rate": 8.988876212207644e-06, | |
| "loss": 0.1604, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.487366423649256, | |
| "grad_norm": 4.445584297180176, | |
| "learning_rate": 8.97461494580719e-06, | |
| "loss": 0.1432, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.49136123040047935, | |
| "grad_norm": 4.145312309265137, | |
| "learning_rate": 8.960353679406733e-06, | |
| "loss": 0.1668, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.4953560371517028, | |
| "grad_norm": 4.525082588195801, | |
| "learning_rate": 8.946092413006276e-06, | |
| "loss": 0.1788, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.4993508439029262, | |
| "grad_norm": 3.0187716484069824, | |
| "learning_rate": 8.931831146605819e-06, | |
| "loss": 0.1603, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.5033456506541496, | |
| "grad_norm": 3.865370750427246, | |
| "learning_rate": 8.917569880205364e-06, | |
| "loss": 0.1598, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.507340457405373, | |
| "grad_norm": 3.384380340576172, | |
| "learning_rate": 8.903308613804906e-06, | |
| "loss": 0.1705, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.5113352641565965, | |
| "grad_norm": 2.700723648071289, | |
| "learning_rate": 8.88904734740445e-06, | |
| "loss": 0.1496, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.5153300709078198, | |
| "grad_norm": 3.309508800506592, | |
| "learning_rate": 8.874786081003994e-06, | |
| "loss": 0.186, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.5193248776590432, | |
| "grad_norm": 4.151066303253174, | |
| "learning_rate": 8.860524814603537e-06, | |
| "loss": 0.1506, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.5233196844102667, | |
| "grad_norm": 2.7802693843841553, | |
| "learning_rate": 8.846263548203082e-06, | |
| "loss": 0.1537, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.5273144911614901, | |
| "grad_norm": 3.3392629623413086, | |
| "learning_rate": 8.832002281802625e-06, | |
| "loss": 0.1521, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.5313092979127134, | |
| "grad_norm": 3.8402390480041504, | |
| "learning_rate": 8.817741015402168e-06, | |
| "loss": 0.1374, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.5353041046639369, | |
| "grad_norm": 4.220055103302002, | |
| "learning_rate": 8.803479749001712e-06, | |
| "loss": 0.1801, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.5392989114151603, | |
| "grad_norm": 4.223282814025879, | |
| "learning_rate": 8.789218482601257e-06, | |
| "loss": 0.152, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.5432937181663837, | |
| "grad_norm": 3.0471835136413574, | |
| "learning_rate": 8.7749572162008e-06, | |
| "loss": 0.1593, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.5472885249176072, | |
| "grad_norm": 4.133998870849609, | |
| "learning_rate": 8.760695949800343e-06, | |
| "loss": 0.1941, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.5512833316688305, | |
| "grad_norm": 4.070934295654297, | |
| "learning_rate": 8.746434683399886e-06, | |
| "loss": 0.1786, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.5552781384200539, | |
| "grad_norm": 3.6335694789886475, | |
| "learning_rate": 8.73217341699943e-06, | |
| "loss": 0.1446, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.5592729451712773, | |
| "grad_norm": 2.9351556301116943, | |
| "learning_rate": 8.717912150598975e-06, | |
| "loss": 0.1605, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.5632677519225008, | |
| "grad_norm": 4.384500503540039, | |
| "learning_rate": 8.703650884198518e-06, | |
| "loss": 0.1785, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.5672625586737241, | |
| "grad_norm": 4.204753398895264, | |
| "learning_rate": 8.689389617798061e-06, | |
| "loss": 0.1555, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.5712573654249475, | |
| "grad_norm": 2.983480930328369, | |
| "learning_rate": 8.675128351397604e-06, | |
| "loss": 0.1642, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.575252172176171, | |
| "grad_norm": 3.3433070182800293, | |
| "learning_rate": 8.66086708499715e-06, | |
| "loss": 0.1542, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.5792469789273944, | |
| "grad_norm": 3.154055118560791, | |
| "learning_rate": 8.646605818596692e-06, | |
| "loss": 0.1654, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.5832417856786178, | |
| "grad_norm": 2.8929858207702637, | |
| "learning_rate": 8.632344552196236e-06, | |
| "loss": 0.1423, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.5872365924298412, | |
| "grad_norm": 3.4675965309143066, | |
| "learning_rate": 8.618083285795779e-06, | |
| "loss": 0.1503, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.5912313991810646, | |
| "grad_norm": 3.996819019317627, | |
| "learning_rate": 8.603822019395322e-06, | |
| "loss": 0.1678, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.595226205932288, | |
| "grad_norm": 4.100089073181152, | |
| "learning_rate": 8.589560752994867e-06, | |
| "loss": 0.1576, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.5992210126835115, | |
| "grad_norm": 3.0981807708740234, | |
| "learning_rate": 8.57529948659441e-06, | |
| "loss": 0.1528, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.6032158194347348, | |
| "grad_norm": 3.0388245582580566, | |
| "learning_rate": 8.561038220193954e-06, | |
| "loss": 0.1529, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.6072106261859582, | |
| "grad_norm": 3.5439846515655518, | |
| "learning_rate": 8.546776953793497e-06, | |
| "loss": 0.1543, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.6112054329371817, | |
| "grad_norm": 4.001914978027344, | |
| "learning_rate": 8.532515687393042e-06, | |
| "loss": 0.1623, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.6152002396884051, | |
| "grad_norm": 3.801597833633423, | |
| "learning_rate": 8.518254420992585e-06, | |
| "loss": 0.1343, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.6191950464396285, | |
| "grad_norm": 3.7877979278564453, | |
| "learning_rate": 8.503993154592128e-06, | |
| "loss": 0.1556, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.6231898531908519, | |
| "grad_norm": 3.0071709156036377, | |
| "learning_rate": 8.489731888191672e-06, | |
| "loss": 0.1354, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.6271846599420753, | |
| "grad_norm": 2.960198163986206, | |
| "learning_rate": 8.475470621791215e-06, | |
| "loss": 0.1567, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.6311794666932987, | |
| "grad_norm": 3.769070863723755, | |
| "learning_rate": 8.46120935539076e-06, | |
| "loss": 0.1392, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.6351742734445222, | |
| "grad_norm": 4.615024089813232, | |
| "learning_rate": 8.446948088990303e-06, | |
| "loss": 0.17, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.6391690801957455, | |
| "grad_norm": 2.945762872695923, | |
| "learning_rate": 8.432686822589846e-06, | |
| "loss": 0.1551, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.6431638869469689, | |
| "grad_norm": 2.70076584815979, | |
| "learning_rate": 8.41842555618939e-06, | |
| "loss": 0.1404, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.6471586936981923, | |
| "grad_norm": 3.891033172607422, | |
| "learning_rate": 8.404164289788934e-06, | |
| "loss": 0.142, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.6511535004494158, | |
| "grad_norm": 3.177311658859253, | |
| "learning_rate": 8.389903023388478e-06, | |
| "loss": 0.1396, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.6551483072006392, | |
| "grad_norm": 3.1945362091064453, | |
| "learning_rate": 8.375641756988021e-06, | |
| "loss": 0.1553, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.6591431139518625, | |
| "grad_norm": 3.6035103797912598, | |
| "learning_rate": 8.361380490587566e-06, | |
| "loss": 0.1485, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.663137920703086, | |
| "grad_norm": 2.8870794773101807, | |
| "learning_rate": 8.347119224187107e-06, | |
| "loss": 0.1467, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.6671327274543094, | |
| "grad_norm": 2.89115834236145, | |
| "learning_rate": 8.332857957786652e-06, | |
| "loss": 0.1552, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.6711275342055328, | |
| "grad_norm": 3.7482142448425293, | |
| "learning_rate": 8.318596691386196e-06, | |
| "loss": 0.1562, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.6751223409567562, | |
| "grad_norm": 3.9963674545288086, | |
| "learning_rate": 8.304335424985739e-06, | |
| "loss": 0.1501, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.6791171477079796, | |
| "grad_norm": 2.8433284759521484, | |
| "learning_rate": 8.290074158585282e-06, | |
| "loss": 0.147, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.683111954459203, | |
| "grad_norm": 4.271126747131348, | |
| "learning_rate": 8.275812892184827e-06, | |
| "loss": 0.1608, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.6871067612104265, | |
| "grad_norm": 3.0002920627593994, | |
| "learning_rate": 8.26155162578437e-06, | |
| "loss": 0.1688, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.6911015679616499, | |
| "grad_norm": 2.862595796585083, | |
| "learning_rate": 8.247290359383914e-06, | |
| "loss": 0.1594, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.6950963747128732, | |
| "grad_norm": 2.760641098022461, | |
| "learning_rate": 8.233029092983458e-06, | |
| "loss": 0.1498, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.6990911814640967, | |
| "grad_norm": 3.4918088912963867, | |
| "learning_rate": 8.218767826583002e-06, | |
| "loss": 0.1644, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.7030859882153201, | |
| "grad_norm": 3.4492998123168945, | |
| "learning_rate": 8.204506560182545e-06, | |
| "loss": 0.1415, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.7070807949665435, | |
| "grad_norm": 3.371929407119751, | |
| "learning_rate": 8.190245293782088e-06, | |
| "loss": 0.1628, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.7110756017177668, | |
| "grad_norm": 3.80068039894104, | |
| "learning_rate": 8.175984027381633e-06, | |
| "loss": 0.207, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.7150704084689903, | |
| "grad_norm": 3.124359607696533, | |
| "learning_rate": 8.161722760981175e-06, | |
| "loss": 0.1328, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.7190652152202137, | |
| "grad_norm": 3.6702919006347656, | |
| "learning_rate": 8.14746149458072e-06, | |
| "loss": 0.1752, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.7230600219714372, | |
| "grad_norm": 2.942159414291382, | |
| "learning_rate": 8.133200228180263e-06, | |
| "loss": 0.1541, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.7270548287226606, | |
| "grad_norm": 3.441380262374878, | |
| "learning_rate": 8.118938961779806e-06, | |
| "loss": 0.1513, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.7310496354738839, | |
| "grad_norm": 3.824725866317749, | |
| "learning_rate": 8.104677695379351e-06, | |
| "loss": 0.1595, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.7350444422251073, | |
| "grad_norm": 2.577880382537842, | |
| "learning_rate": 8.090416428978894e-06, | |
| "loss": 0.1526, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.7390392489763308, | |
| "grad_norm": 3.2072184085845947, | |
| "learning_rate": 8.076155162578438e-06, | |
| "loss": 0.1601, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.7430340557275542, | |
| "grad_norm": 3.3650033473968506, | |
| "learning_rate": 8.06189389617798e-06, | |
| "loss": 0.1527, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.7470288624787776, | |
| "grad_norm": 3.575591564178467, | |
| "learning_rate": 8.047632629777526e-06, | |
| "loss": 0.1539, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.751023669230001, | |
| "grad_norm": 3.2429842948913574, | |
| "learning_rate": 8.033371363377069e-06, | |
| "loss": 0.17, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.7550184759812244, | |
| "grad_norm": 2.9099652767181396, | |
| "learning_rate": 8.019110096976612e-06, | |
| "loss": 0.1487, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.7590132827324478, | |
| "grad_norm": 3.7374091148376465, | |
| "learning_rate": 8.004848830576156e-06, | |
| "loss": 0.1491, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.7630080894836713, | |
| "grad_norm": 2.990825891494751, | |
| "learning_rate": 7.990587564175699e-06, | |
| "loss": 0.1495, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.7670028962348946, | |
| "grad_norm": 3.381781816482544, | |
| "learning_rate": 7.976326297775244e-06, | |
| "loss": 0.1265, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.770997702986118, | |
| "grad_norm": 3.0114705562591553, | |
| "learning_rate": 7.962065031374787e-06, | |
| "loss": 0.1421, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.7749925097373415, | |
| "grad_norm": 3.9027457237243652, | |
| "learning_rate": 7.94780376497433e-06, | |
| "loss": 0.1453, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.7789873164885649, | |
| "grad_norm": 5.936771869659424, | |
| "learning_rate": 7.933542498573873e-06, | |
| "loss": 0.1426, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.7829821232397883, | |
| "grad_norm": 3.449986219406128, | |
| "learning_rate": 7.919281232173418e-06, | |
| "loss": 0.1722, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.7869769299910117, | |
| "grad_norm": 3.3907039165496826, | |
| "learning_rate": 7.905019965772962e-06, | |
| "loss": 0.1464, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.7909717367422351, | |
| "grad_norm": 3.0017893314361572, | |
| "learning_rate": 7.890758699372505e-06, | |
| "loss": 0.1562, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.7949665434934585, | |
| "grad_norm": 3.4000141620635986, | |
| "learning_rate": 7.876497432972048e-06, | |
| "loss": 0.173, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.798961350244682, | |
| "grad_norm": 3.022404670715332, | |
| "learning_rate": 7.862236166571591e-06, | |
| "loss": 0.15, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.8029561569959053, | |
| "grad_norm": 3.237109899520874, | |
| "learning_rate": 7.847974900171136e-06, | |
| "loss": 0.1571, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.8069509637471287, | |
| "grad_norm": 2.5590929985046387, | |
| "learning_rate": 7.83371363377068e-06, | |
| "loss": 0.1596, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.8109457704983521, | |
| "grad_norm": 4.01102352142334, | |
| "learning_rate": 7.819452367370223e-06, | |
| "loss": 0.1511, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.8149405772495756, | |
| "grad_norm": 3.1152961254119873, | |
| "learning_rate": 7.805191100969766e-06, | |
| "loss": 0.1564, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.818935384000799, | |
| "grad_norm": 3.8849892616271973, | |
| "learning_rate": 7.790929834569311e-06, | |
| "loss": 0.1686, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.8229301907520223, | |
| "grad_norm": 4.077883243560791, | |
| "learning_rate": 7.776668568168854e-06, | |
| "loss": 0.1519, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.8269249975032458, | |
| "grad_norm": 3.1383252143859863, | |
| "learning_rate": 7.762407301768397e-06, | |
| "loss": 0.1572, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.8309198042544692, | |
| "grad_norm": 3.380702257156372, | |
| "learning_rate": 7.748146035367942e-06, | |
| "loss": 0.147, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.8349146110056926, | |
| "grad_norm": 3.4674811363220215, | |
| "learning_rate": 7.733884768967484e-06, | |
| "loss": 0.1802, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.838909417756916, | |
| "grad_norm": 3.4076437950134277, | |
| "learning_rate": 7.719623502567029e-06, | |
| "loss": 0.1562, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.8429042245081394, | |
| "grad_norm": 4.0926361083984375, | |
| "learning_rate": 7.705362236166572e-06, | |
| "loss": 0.1563, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.8468990312593628, | |
| "grad_norm": 3.7539260387420654, | |
| "learning_rate": 7.691100969766115e-06, | |
| "loss": 0.1474, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.8508938380105863, | |
| "grad_norm": 3.1815896034240723, | |
| "learning_rate": 7.676839703365659e-06, | |
| "loss": 0.1644, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.8548886447618097, | |
| "grad_norm": 4.006225109100342, | |
| "learning_rate": 7.662578436965204e-06, | |
| "loss": 0.1672, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.858883451513033, | |
| "grad_norm": 2.3754403591156006, | |
| "learning_rate": 7.648317170564747e-06, | |
| "loss": 0.1344, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.8628782582642565, | |
| "grad_norm": 3.58271861076355, | |
| "learning_rate": 7.63405590416429e-06, | |
| "loss": 0.1707, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.8668730650154799, | |
| "grad_norm": 3.560786247253418, | |
| "learning_rate": 7.619794637763834e-06, | |
| "loss": 0.1716, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.8708678717667033, | |
| "grad_norm": 3.7507925033569336, | |
| "learning_rate": 7.6055333713633774e-06, | |
| "loss": 0.1385, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.8748626785179267, | |
| "grad_norm": 3.3080689907073975, | |
| "learning_rate": 7.5912721049629215e-06, | |
| "loss": 0.1409, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.8788574852691501, | |
| "grad_norm": 2.7758231163024902, | |
| "learning_rate": 7.577010838562466e-06, | |
| "loss": 0.1389, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.8828522920203735, | |
| "grad_norm": 3.5948941707611084, | |
| "learning_rate": 7.562749572162008e-06, | |
| "loss": 0.1322, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.886847098771597, | |
| "grad_norm": 3.502706527709961, | |
| "learning_rate": 7.548488305761552e-06, | |
| "loss": 0.149, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.8908419055228204, | |
| "grad_norm": 3.414030075073242, | |
| "learning_rate": 7.534227039361096e-06, | |
| "loss": 0.1545, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.8948367122740437, | |
| "grad_norm": 3.6713764667510986, | |
| "learning_rate": 7.519965772960639e-06, | |
| "loss": 0.1423, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.8988315190252671, | |
| "grad_norm": 3.5644688606262207, | |
| "learning_rate": 7.505704506560183e-06, | |
| "loss": 0.1738, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.9028263257764906, | |
| "grad_norm": 3.30999755859375, | |
| "learning_rate": 7.491443240159727e-06, | |
| "loss": 0.1751, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.906821132527714, | |
| "grad_norm": 3.0740654468536377, | |
| "learning_rate": 7.47718197375927e-06, | |
| "loss": 0.1293, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.9108159392789373, | |
| "grad_norm": 2.7236621379852295, | |
| "learning_rate": 7.462920707358814e-06, | |
| "loss": 0.1669, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.9148107460301608, | |
| "grad_norm": 3.11482310295105, | |
| "learning_rate": 7.448659440958358e-06, | |
| "loss": 0.1239, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.9188055527813842, | |
| "grad_norm": 3.0215625762939453, | |
| "learning_rate": 7.434398174557901e-06, | |
| "loss": 0.1527, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.9228003595326076, | |
| "grad_norm": 3.896244764328003, | |
| "learning_rate": 7.420136908157445e-06, | |
| "loss": 0.1498, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.9267951662838311, | |
| "grad_norm": 4.210872173309326, | |
| "learning_rate": 7.405875641756989e-06, | |
| "loss": 0.1521, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.9307899730350544, | |
| "grad_norm": 4.199804306030273, | |
| "learning_rate": 7.391614375356533e-06, | |
| "loss": 0.1363, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.9347847797862778, | |
| "grad_norm": 3.445571184158325, | |
| "learning_rate": 7.377353108956075e-06, | |
| "loss": 0.141, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.9387795865375013, | |
| "grad_norm": 3.0716307163238525, | |
| "learning_rate": 7.3630918425556194e-06, | |
| "loss": 0.1502, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.9427743932887247, | |
| "grad_norm": 3.4230117797851562, | |
| "learning_rate": 7.3488305761551635e-06, | |
| "loss": 0.1544, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.946769200039948, | |
| "grad_norm": 3.4998185634613037, | |
| "learning_rate": 7.334569309754707e-06, | |
| "loss": 0.1414, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.9507640067911715, | |
| "grad_norm": 3.726682186126709, | |
| "learning_rate": 7.320308043354251e-06, | |
| "loss": 0.1272, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.9547588135423949, | |
| "grad_norm": 3.086432456970215, | |
| "learning_rate": 7.306046776953795e-06, | |
| "loss": 0.1373, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.9587536202936183, | |
| "grad_norm": 3.308708667755127, | |
| "learning_rate": 7.291785510553337e-06, | |
| "loss": 0.1434, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.9627484270448418, | |
| "grad_norm": 2.7734992504119873, | |
| "learning_rate": 7.2775242441528815e-06, | |
| "loss": 0.1575, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.9667432337960651, | |
| "grad_norm": 2.2432265281677246, | |
| "learning_rate": 7.2632629777524256e-06, | |
| "loss": 0.1208, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.9707380405472885, | |
| "grad_norm": 3.2374532222747803, | |
| "learning_rate": 7.249001711351969e-06, | |
| "loss": 0.1366, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.974732847298512, | |
| "grad_norm": 2.740648031234741, | |
| "learning_rate": 7.234740444951512e-06, | |
| "loss": 0.1514, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.9787276540497354, | |
| "grad_norm": 3.543654441833496, | |
| "learning_rate": 7.220479178551056e-06, | |
| "loss": 0.1261, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.9827224608009587, | |
| "grad_norm": 3.6213510036468506, | |
| "learning_rate": 7.206217912150599e-06, | |
| "loss": 0.1482, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.9867172675521821, | |
| "grad_norm": 2.9423608779907227, | |
| "learning_rate": 7.1919566457501435e-06, | |
| "loss": 0.158, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.9907120743034056, | |
| "grad_norm": 5.1055145263671875, | |
| "learning_rate": 7.1776953793496876e-06, | |
| "loss": 0.1245, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.994706881054629, | |
| "grad_norm": 3.547549247741699, | |
| "learning_rate": 7.16343411294923e-06, | |
| "loss": 0.1582, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.9987016878058524, | |
| "grad_norm": 2.7595136165618896, | |
| "learning_rate": 7.149172846548774e-06, | |
| "loss": 0.1446, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.9987016878058524, | |
| "eval_loss": 0.08971710503101349, | |
| "eval_runtime": 11133.1611, | |
| "eval_samples_per_second": 1.799, | |
| "eval_steps_per_second": 0.225, | |
| "eval_wer": 8.384208161162686, | |
| "step": 2500 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 7512, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 2500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.16483926016e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |