diff --git "a/checkpoint-7500/trainer_state.json" "b/checkpoint-7500/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-7500/trainer_state.json" @@ -0,0 +1,5311 @@ +{ + "best_global_step": 7500, + "best_metric": 2.2808728914804304, + "best_model_checkpoint": "./SALAMA_NEWMED2/checkpoint-7500", + "epoch": 2.9955058424048735, + "eval_steps": 2500, + "global_step": 7500, + "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 + }, + { + "epoch": 1.002396884050734, + "grad_norm": 2.1444203853607178, + "learning_rate": 7.134911580148318e-06, + "loss": 0.0875, + "step": 2510 + }, + { + "epoch": 1.0063916908019575, + "grad_norm": 2.192164659500122, + "learning_rate": 7.120650313747861e-06, + "loss": 0.0748, + "step": 2520 + }, + { + "epoch": 1.0103864975531809, + "grad_norm": 2.3259665966033936, + "learning_rate": 7.1063890473474055e-06, + "loss": 0.0779, + "step": 2530 + }, + { + "epoch": 1.0143813043044043, + "grad_norm": 2.5470423698425293, + "learning_rate": 7.092127780946949e-06, + "loss": 0.0715, + "step": 2540 + }, + { + "epoch": 1.0183761110556278, + "grad_norm": 2.3078274726867676, + "learning_rate": 7.077866514546492e-06, + "loss": 0.0821, + "step": 2550 + }, + { + "epoch": 1.0223709178068512, + "grad_norm": 2.316812753677368, + "learning_rate": 7.063605248146036e-06, + "loss": 0.0694, + "step": 2560 + }, + { + "epoch": 1.0263657245580744, + "grad_norm": 2.5326666831970215, + "learning_rate": 7.04934398174558e-06, + "loss": 0.0624, + "step": 2570 + }, + { + "epoch": 1.0303605313092978, + "grad_norm": 2.4062130451202393, + "learning_rate": 7.035082715345123e-06, + "loss": 0.0807, + "step": 2580 + }, + { + "epoch": 1.0343553380605213, + "grad_norm": 2.9939708709716797, + "learning_rate": 7.020821448944667e-06, + "loss": 0.0898, + "step": 2590 + }, + { + "epoch": 1.0383501448117447, + "grad_norm": 2.683166980743408, + "learning_rate": 7.006560182544211e-06, + "loss": 0.0679, + "step": 2600 + }, + { + "epoch": 1.0423449515629681, + "grad_norm": 1.943904161453247, + "learning_rate": 6.992298916143754e-06, + "loss": 0.0794, + "step": 2610 + }, + { + "epoch": 1.0463397583141916, + "grad_norm": 2.627399444580078, + "learning_rate": 6.978037649743298e-06, + "loss": 0.0764, + "step": 2620 + }, + { + "epoch": 1.050334565065415, + "grad_norm": 2.662348747253418, + "learning_rate": 6.963776383342841e-06, + "loss": 0.0698, + "step": 2630 + }, + { + "epoch": 1.0543293718166384, + "grad_norm": 2.6556851863861084, + "learning_rate": 6.949515116942385e-06, + "loss": 0.0641, + "step": 2640 + }, + { + "epoch": 1.0583241785678619, + "grad_norm": 2.1765525341033936, + "learning_rate": 6.935253850541929e-06, + "loss": 0.0754, + "step": 2650 + }, + { + "epoch": 1.062318985319085, + "grad_norm": 2.0937461853027344, + "learning_rate": 6.920992584141473e-06, + "loss": 0.0713, + "step": 2660 + }, + { + "epoch": 1.0663137920703085, + "grad_norm": 2.6876823902130127, + "learning_rate": 6.906731317741015e-06, + "loss": 0.0776, + "step": 2670 + }, + { + "epoch": 1.070308598821532, + "grad_norm": 1.9541680812835693, + "learning_rate": 6.892470051340559e-06, + "loss": 0.0743, + "step": 2680 + }, + { + "epoch": 1.0743034055727554, + "grad_norm": 2.5399224758148193, + "learning_rate": 6.878208784940103e-06, + "loss": 0.0748, + "step": 2690 + }, + { + "epoch": 1.0782982123239788, + "grad_norm": 2.379944086074829, + "learning_rate": 6.863947518539647e-06, + "loss": 0.0721, + "step": 2700 + }, + { + "epoch": 1.0822930190752023, + "grad_norm": 2.211306571960449, + "learning_rate": 6.849686252139191e-06, + "loss": 0.0767, + "step": 2710 + }, + { + "epoch": 1.0862878258264257, + "grad_norm": 3.00471568107605, + "learning_rate": 6.835424985738735e-06, + "loss": 0.0821, + "step": 2720 + }, + { + "epoch": 1.0902826325776491, + "grad_norm": 2.261798143386841, + "learning_rate": 6.821163719338277e-06, + "loss": 0.0719, + "step": 2730 + }, + { + "epoch": 1.0942774393288726, + "grad_norm": 2.2622437477111816, + "learning_rate": 6.806902452937821e-06, + "loss": 0.0797, + "step": 2740 + }, + { + "epoch": 1.098272246080096, + "grad_norm": 2.405115842819214, + "learning_rate": 6.792641186537365e-06, + "loss": 0.072, + "step": 2750 + }, + { + "epoch": 1.1022670528313192, + "grad_norm": 2.5937540531158447, + "learning_rate": 6.778379920136909e-06, + "loss": 0.0849, + "step": 2760 + }, + { + "epoch": 1.1062618595825426, + "grad_norm": 2.0180816650390625, + "learning_rate": 6.764118653736452e-06, + "loss": 0.0754, + "step": 2770 + }, + { + "epoch": 1.110256666333766, + "grad_norm": 2.903346538543701, + "learning_rate": 6.749857387335996e-06, + "loss": 0.0679, + "step": 2780 + }, + { + "epoch": 1.1142514730849895, + "grad_norm": 2.69219970703125, + "learning_rate": 6.735596120935539e-06, + "loss": 0.0796, + "step": 2790 + }, + { + "epoch": 1.118246279836213, + "grad_norm": 2.6778345108032227, + "learning_rate": 6.721334854535083e-06, + "loss": 0.0757, + "step": 2800 + }, + { + "epoch": 1.1222410865874364, + "grad_norm": 1.992986798286438, + "learning_rate": 6.7070735881346274e-06, + "loss": 0.0706, + "step": 2810 + }, + { + "epoch": 1.1262358933386598, + "grad_norm": 2.307279348373413, + "learning_rate": 6.69281232173417e-06, + "loss": 0.0776, + "step": 2820 + }, + { + "epoch": 1.1302307000898832, + "grad_norm": 2.8715591430664062, + "learning_rate": 6.678551055333714e-06, + "loss": 0.0772, + "step": 2830 + }, + { + "epoch": 1.1342255068411067, + "grad_norm": 2.425602674484253, + "learning_rate": 6.664289788933258e-06, + "loss": 0.0823, + "step": 2840 + }, + { + "epoch": 1.13822031359233, + "grad_norm": 2.799405097961426, + "learning_rate": 6.650028522532801e-06, + "loss": 0.077, + "step": 2850 + }, + { + "epoch": 1.1422151203435533, + "grad_norm": 2.3349719047546387, + "learning_rate": 6.635767256132345e-06, + "loss": 0.0589, + "step": 2860 + }, + { + "epoch": 1.1462099270947768, + "grad_norm": 2.405153274536133, + "learning_rate": 6.621505989731889e-06, + "loss": 0.0798, + "step": 2870 + }, + { + "epoch": 1.1502047338460002, + "grad_norm": 2.737041473388672, + "learning_rate": 6.607244723331432e-06, + "loss": 0.0803, + "step": 2880 + }, + { + "epoch": 1.1541995405972236, + "grad_norm": 2.6586458683013916, + "learning_rate": 6.592983456930976e-06, + "loss": 0.0764, + "step": 2890 + }, + { + "epoch": 1.158194347348447, + "grad_norm": 2.114450693130493, + "learning_rate": 6.57872219053052e-06, + "loss": 0.0823, + "step": 2900 + }, + { + "epoch": 1.1621891540996705, + "grad_norm": 2.1893551349639893, + "learning_rate": 6.5644609241300625e-06, + "loss": 0.0632, + "step": 2910 + }, + { + "epoch": 1.166183960850894, + "grad_norm": 1.8855793476104736, + "learning_rate": 6.5501996577296066e-06, + "loss": 0.0741, + "step": 2920 + }, + { + "epoch": 1.1701787676021174, + "grad_norm": 2.348771572113037, + "learning_rate": 6.535938391329151e-06, + "loss": 0.0746, + "step": 2930 + }, + { + "epoch": 1.1741735743533406, + "grad_norm": 1.690022587776184, + "learning_rate": 6.521677124928694e-06, + "loss": 0.0652, + "step": 2940 + }, + { + "epoch": 1.178168381104564, + "grad_norm": 2.1050920486450195, + "learning_rate": 6.507415858528238e-06, + "loss": 0.0757, + "step": 2950 + }, + { + "epoch": 1.1821631878557874, + "grad_norm": 2.2572884559631348, + "learning_rate": 6.493154592127781e-06, + "loss": 0.0821, + "step": 2960 + }, + { + "epoch": 1.1861579946070109, + "grad_norm": 2.6294784545898438, + "learning_rate": 6.478893325727325e-06, + "loss": 0.0772, + "step": 2970 + }, + { + "epoch": 1.1901528013582343, + "grad_norm": 2.248670816421509, + "learning_rate": 6.464632059326869e-06, + "loss": 0.0725, + "step": 2980 + }, + { + "epoch": 1.1941476081094577, + "grad_norm": 2.916764736175537, + "learning_rate": 6.450370792926413e-06, + "loss": 0.0773, + "step": 2990 + }, + { + "epoch": 1.1981424148606812, + "grad_norm": 2.563652276992798, + "learning_rate": 6.436109526525957e-06, + "loss": 0.0711, + "step": 3000 + }, + { + "epoch": 1.2021372216119046, + "grad_norm": 2.4842357635498047, + "learning_rate": 6.421848260125499e-06, + "loss": 0.0808, + "step": 3010 + }, + { + "epoch": 1.206132028363128, + "grad_norm": 2.3920555114746094, + "learning_rate": 6.407586993725043e-06, + "loss": 0.064, + "step": 3020 + }, + { + "epoch": 1.2101268351143513, + "grad_norm": 2.3040075302124023, + "learning_rate": 6.393325727324587e-06, + "loss": 0.077, + "step": 3030 + }, + { + "epoch": 1.2141216418655747, + "grad_norm": 3.0777621269226074, + "learning_rate": 6.379064460924131e-06, + "loss": 0.0729, + "step": 3040 + }, + { + "epoch": 1.2181164486167981, + "grad_norm": 2.5559732913970947, + "learning_rate": 6.364803194523675e-06, + "loss": 0.0745, + "step": 3050 + }, + { + "epoch": 1.2221112553680216, + "grad_norm": 1.7757205963134766, + "learning_rate": 6.350541928123218e-06, + "loss": 0.0786, + "step": 3060 + }, + { + "epoch": 1.226106062119245, + "grad_norm": 2.1353683471679688, + "learning_rate": 6.336280661722761e-06, + "loss": 0.0658, + "step": 3070 + }, + { + "epoch": 1.2301008688704684, + "grad_norm": 2.2478456497192383, + "learning_rate": 6.322019395322305e-06, + "loss": 0.0716, + "step": 3080 + }, + { + "epoch": 1.2340956756216919, + "grad_norm": 1.8999541997909546, + "learning_rate": 6.307758128921849e-06, + "loss": 0.0691, + "step": 3090 + }, + { + "epoch": 1.2380904823729153, + "grad_norm": 1.6499491930007935, + "learning_rate": 6.293496862521392e-06, + "loss": 0.0691, + "step": 3100 + }, + { + "epoch": 1.2420852891241387, + "grad_norm": 2.789496898651123, + "learning_rate": 6.279235596120936e-06, + "loss": 0.0737, + "step": 3110 + }, + { + "epoch": 1.246080095875362, + "grad_norm": 2.9136650562286377, + "learning_rate": 6.26497432972048e-06, + "loss": 0.062, + "step": 3120 + }, + { + "epoch": 1.2500749026265854, + "grad_norm": 2.7884409427642822, + "learning_rate": 6.250713063320023e-06, + "loss": 0.0654, + "step": 3130 + }, + { + "epoch": 1.2540697093778088, + "grad_norm": 2.545773983001709, + "learning_rate": 6.236451796919567e-06, + "loss": 0.0746, + "step": 3140 + }, + { + "epoch": 1.2580645161290323, + "grad_norm": 1.842954158782959, + "learning_rate": 6.222190530519111e-06, + "loss": 0.0678, + "step": 3150 + }, + { + "epoch": 1.2620593228802557, + "grad_norm": 2.340939998626709, + "learning_rate": 6.207929264118654e-06, + "loss": 0.0724, + "step": 3160 + }, + { + "epoch": 1.2660541296314791, + "grad_norm": 2.3492343425750732, + "learning_rate": 6.193667997718198e-06, + "loss": 0.0773, + "step": 3170 + }, + { + "epoch": 1.2700489363827026, + "grad_norm": 2.6989333629608154, + "learning_rate": 6.179406731317742e-06, + "loss": 0.0754, + "step": 3180 + }, + { + "epoch": 1.2740437431339258, + "grad_norm": 1.8909000158309937, + "learning_rate": 6.165145464917285e-06, + "loss": 0.0638, + "step": 3190 + }, + { + "epoch": 1.2780385498851494, + "grad_norm": 3.751016616821289, + "learning_rate": 6.1508841985168285e-06, + "loss": 0.0918, + "step": 3200 + }, + { + "epoch": 1.2820333566363726, + "grad_norm": 2.805493116378784, + "learning_rate": 6.136622932116373e-06, + "loss": 0.0782, + "step": 3210 + }, + { + "epoch": 1.286028163387596, + "grad_norm": 2.417083978652954, + "learning_rate": 6.122361665715916e-06, + "loss": 0.0739, + "step": 3220 + }, + { + "epoch": 1.2900229701388195, + "grad_norm": 2.121392250061035, + "learning_rate": 6.10810039931546e-06, + "loss": 0.061, + "step": 3230 + }, + { + "epoch": 1.294017776890043, + "grad_norm": 1.9236271381378174, + "learning_rate": 6.093839132915004e-06, + "loss": 0.0809, + "step": 3240 + }, + { + "epoch": 1.2980125836412664, + "grad_norm": 2.1827476024627686, + "learning_rate": 6.0795778665145464e-06, + "loss": 0.06, + "step": 3250 + }, + { + "epoch": 1.3020073903924898, + "grad_norm": 2.33284854888916, + "learning_rate": 6.0653166001140905e-06, + "loss": 0.0653, + "step": 3260 + }, + { + "epoch": 1.3060021971437132, + "grad_norm": 1.6561416387557983, + "learning_rate": 6.051055333713635e-06, + "loss": 0.0778, + "step": 3270 + }, + { + "epoch": 1.3099970038949365, + "grad_norm": 2.6791889667510986, + "learning_rate": 6.036794067313178e-06, + "loss": 0.0755, + "step": 3280 + }, + { + "epoch": 1.31399181064616, + "grad_norm": 2.7164688110351562, + "learning_rate": 6.022532800912721e-06, + "loss": 0.0625, + "step": 3290 + }, + { + "epoch": 1.3179866173973833, + "grad_norm": 2.960965633392334, + "learning_rate": 6.008271534512265e-06, + "loss": 0.0701, + "step": 3300 + }, + { + "epoch": 1.3219814241486068, + "grad_norm": 2.447868585586548, + "learning_rate": 5.9940102681118085e-06, + "loss": 0.074, + "step": 3310 + }, + { + "epoch": 1.3259762308998302, + "grad_norm": 2.773967742919922, + "learning_rate": 5.9797490017113525e-06, + "loss": 0.0885, + "step": 3320 + }, + { + "epoch": 1.3299710376510536, + "grad_norm": 2.2736873626708984, + "learning_rate": 5.965487735310897e-06, + "loss": 0.0697, + "step": 3330 + }, + { + "epoch": 1.333965844402277, + "grad_norm": 1.7243731021881104, + "learning_rate": 5.951226468910439e-06, + "loss": 0.0637, + "step": 3340 + }, + { + "epoch": 1.3379606511535005, + "grad_norm": 2.63024640083313, + "learning_rate": 5.936965202509983e-06, + "loss": 0.0761, + "step": 3350 + }, + { + "epoch": 1.341955457904724, + "grad_norm": 1.550184965133667, + "learning_rate": 5.922703936109527e-06, + "loss": 0.0815, + "step": 3360 + }, + { + "epoch": 1.3459502646559471, + "grad_norm": 2.066818952560425, + "learning_rate": 5.9084426697090705e-06, + "loss": 0.0706, + "step": 3370 + }, + { + "epoch": 1.3499450714071708, + "grad_norm": 2.038065195083618, + "learning_rate": 5.8941814033086146e-06, + "loss": 0.0701, + "step": 3380 + }, + { + "epoch": 1.353939878158394, + "grad_norm": 2.8719944953918457, + "learning_rate": 5.879920136908158e-06, + "loss": 0.077, + "step": 3390 + }, + { + "epoch": 1.3579346849096174, + "grad_norm": 3.0827386379241943, + "learning_rate": 5.865658870507701e-06, + "loss": 0.08, + "step": 3400 + }, + { + "epoch": 1.3619294916608409, + "grad_norm": 2.600886344909668, + "learning_rate": 5.851397604107245e-06, + "loss": 0.0703, + "step": 3410 + }, + { + "epoch": 1.3659242984120643, + "grad_norm": 2.431089401245117, + "learning_rate": 5.837136337706789e-06, + "loss": 0.0694, + "step": 3420 + }, + { + "epoch": 1.3699191051632877, + "grad_norm": 1.9217971563339233, + "learning_rate": 5.822875071306332e-06, + "loss": 0.0649, + "step": 3430 + }, + { + "epoch": 1.3739139119145112, + "grad_norm": 1.802898645401001, + "learning_rate": 5.808613804905876e-06, + "loss": 0.0722, + "step": 3440 + }, + { + "epoch": 1.3779087186657346, + "grad_norm": 2.1869544982910156, + "learning_rate": 5.79435253850542e-06, + "loss": 0.0757, + "step": 3450 + }, + { + "epoch": 1.3819035254169578, + "grad_norm": 2.309509038925171, + "learning_rate": 5.780091272104963e-06, + "loss": 0.0658, + "step": 3460 + }, + { + "epoch": 1.3858983321681815, + "grad_norm": 2.2018911838531494, + "learning_rate": 5.765830005704507e-06, + "loss": 0.0597, + "step": 3470 + }, + { + "epoch": 1.3898931389194047, + "grad_norm": 2.274909019470215, + "learning_rate": 5.751568739304051e-06, + "loss": 0.0773, + "step": 3480 + }, + { + "epoch": 1.3938879456706281, + "grad_norm": 2.7324776649475098, + "learning_rate": 5.737307472903594e-06, + "loss": 0.0679, + "step": 3490 + }, + { + "epoch": 1.3978827524218516, + "grad_norm": 2.519226312637329, + "learning_rate": 5.723046206503138e-06, + "loss": 0.0614, + "step": 3500 + }, + { + "epoch": 1.401877559173075, + "grad_norm": 1.8280469179153442, + "learning_rate": 5.708784940102682e-06, + "loss": 0.0639, + "step": 3510 + }, + { + "epoch": 1.4058723659242984, + "grad_norm": 3.6149826049804688, + "learning_rate": 5.694523673702225e-06, + "loss": 0.0926, + "step": 3520 + }, + { + "epoch": 1.4098671726755219, + "grad_norm": 2.238356351852417, + "learning_rate": 5.680262407301768e-06, + "loss": 0.0692, + "step": 3530 + }, + { + "epoch": 1.4138619794267453, + "grad_norm": 2.1481969356536865, + "learning_rate": 5.6660011409013125e-06, + "loss": 0.068, + "step": 3540 + }, + { + "epoch": 1.4178567861779685, + "grad_norm": 2.112460136413574, + "learning_rate": 5.651739874500856e-06, + "loss": 0.071, + "step": 3550 + }, + { + "epoch": 1.4218515929291922, + "grad_norm": 2.6362290382385254, + "learning_rate": 5.6374786081004e-06, + "loss": 0.081, + "step": 3560 + }, + { + "epoch": 1.4258463996804154, + "grad_norm": 2.9250431060791016, + "learning_rate": 5.623217341699944e-06, + "loss": 0.0684, + "step": 3570 + }, + { + "epoch": 1.4298412064316388, + "grad_norm": 2.8909897804260254, + "learning_rate": 5.608956075299488e-06, + "loss": 0.0662, + "step": 3580 + }, + { + "epoch": 1.4338360131828622, + "grad_norm": 2.6137373447418213, + "learning_rate": 5.59469480889903e-06, + "loss": 0.0995, + "step": 3590 + }, + { + "epoch": 1.4378308199340857, + "grad_norm": 2.0842933654785156, + "learning_rate": 5.5804335424985745e-06, + "loss": 0.0732, + "step": 3600 + }, + { + "epoch": 1.4418256266853091, + "grad_norm": 1.304982304573059, + "learning_rate": 5.566172276098119e-06, + "loss": 0.0615, + "step": 3610 + }, + { + "epoch": 1.4458204334365325, + "grad_norm": 2.8772590160369873, + "learning_rate": 5.551911009697661e-06, + "loss": 0.0812, + "step": 3620 + }, + { + "epoch": 1.449815240187756, + "grad_norm": 1.5531134605407715, + "learning_rate": 5.537649743297205e-06, + "loss": 0.0627, + "step": 3630 + }, + { + "epoch": 1.4538100469389792, + "grad_norm": 2.48970627784729, + "learning_rate": 5.523388476896749e-06, + "loss": 0.0668, + "step": 3640 + }, + { + "epoch": 1.4578048536902029, + "grad_norm": 3.0162699222564697, + "learning_rate": 5.509127210496292e-06, + "loss": 0.0857, + "step": 3650 + }, + { + "epoch": 1.461799660441426, + "grad_norm": 3.163447618484497, + "learning_rate": 5.4948659440958365e-06, + "loss": 0.0705, + "step": 3660 + }, + { + "epoch": 1.4657944671926495, + "grad_norm": 7.272205352783203, + "learning_rate": 5.480604677695381e-06, + "loss": 0.0632, + "step": 3670 + }, + { + "epoch": 1.469789273943873, + "grad_norm": 1.673565149307251, + "learning_rate": 5.466343411294923e-06, + "loss": 0.0606, + "step": 3680 + }, + { + "epoch": 1.4737840806950964, + "grad_norm": 1.9384418725967407, + "learning_rate": 5.452082144894467e-06, + "loss": 0.0582, + "step": 3690 + }, + { + "epoch": 1.4777788874463198, + "grad_norm": 1.5022332668304443, + "learning_rate": 5.437820878494011e-06, + "loss": 0.0753, + "step": 3700 + }, + { + "epoch": 1.4817736941975432, + "grad_norm": 2.6111533641815186, + "learning_rate": 5.4235596120935544e-06, + "loss": 0.0729, + "step": 3710 + }, + { + "epoch": 1.4857685009487667, + "grad_norm": 2.3940229415893555, + "learning_rate": 5.409298345693098e-06, + "loss": 0.07, + "step": 3720 + }, + { + "epoch": 1.4897633076999899, + "grad_norm": 1.55259108543396, + "learning_rate": 5.395037079292642e-06, + "loss": 0.0645, + "step": 3730 + }, + { + "epoch": 1.4937581144512135, + "grad_norm": 1.9426193237304688, + "learning_rate": 5.380775812892185e-06, + "loss": 0.0706, + "step": 3740 + }, + { + "epoch": 1.4977529212024367, + "grad_norm": 2.144442319869995, + "learning_rate": 5.366514546491729e-06, + "loss": 0.0629, + "step": 3750 + }, + { + "epoch": 1.5017477279536604, + "grad_norm": 2.539332151412964, + "learning_rate": 5.352253280091273e-06, + "loss": 0.0699, + "step": 3760 + }, + { + "epoch": 1.5057425347048836, + "grad_norm": 2.933539867401123, + "learning_rate": 5.337992013690816e-06, + "loss": 0.079, + "step": 3770 + }, + { + "epoch": 1.509737341456107, + "grad_norm": 3.2602779865264893, + "learning_rate": 5.32373074729036e-06, + "loss": 0.0577, + "step": 3780 + }, + { + "epoch": 1.5137321482073305, + "grad_norm": 2.080461025238037, + "learning_rate": 5.309469480889904e-06, + "loss": 0.0548, + "step": 3790 + }, + { + "epoch": 1.517726954958554, + "grad_norm": 3.82892107963562, + "learning_rate": 5.295208214489447e-06, + "loss": 0.0681, + "step": 3800 + }, + { + "epoch": 1.5217217617097774, + "grad_norm": 2.3601534366607666, + "learning_rate": 5.280946948088991e-06, + "loss": 0.0692, + "step": 3810 + }, + { + "epoch": 1.5257165684610006, + "grad_norm": 2.4199206829071045, + "learning_rate": 5.266685681688534e-06, + "loss": 0.0623, + "step": 3820 + }, + { + "epoch": 1.5297113752122242, + "grad_norm": 1.8460701704025269, + "learning_rate": 5.252424415288078e-06, + "loss": 0.0628, + "step": 3830 + }, + { + "epoch": 1.5337061819634474, + "grad_norm": 2.4453554153442383, + "learning_rate": 5.238163148887622e-06, + "loss": 0.0681, + "step": 3840 + }, + { + "epoch": 1.537700988714671, + "grad_norm": 3.5866127014160156, + "learning_rate": 5.223901882487166e-06, + "loss": 0.0882, + "step": 3850 + }, + { + "epoch": 1.5416957954658943, + "grad_norm": 2.0650634765625, + "learning_rate": 5.209640616086708e-06, + "loss": 0.0801, + "step": 3860 + }, + { + "epoch": 1.5456906022171177, + "grad_norm": 2.9257614612579346, + "learning_rate": 5.195379349686252e-06, + "loss": 0.0618, + "step": 3870 + }, + { + "epoch": 1.5496854089683412, + "grad_norm": 2.5670325756073, + "learning_rate": 5.1811180832857964e-06, + "loss": 0.0689, + "step": 3880 + }, + { + "epoch": 1.5536802157195646, + "grad_norm": 2.218632221221924, + "learning_rate": 5.16685681688534e-06, + "loss": 0.0682, + "step": 3890 + }, + { + "epoch": 1.557675022470788, + "grad_norm": 2.3123788833618164, + "learning_rate": 5.152595550484884e-06, + "loss": 0.083, + "step": 3900 + }, + { + "epoch": 1.5616698292220113, + "grad_norm": 2.382164239883423, + "learning_rate": 5.138334284084428e-06, + "loss": 0.0577, + "step": 3910 + }, + { + "epoch": 1.565664635973235, + "grad_norm": 2.004746913909912, + "learning_rate": 5.12407301768397e-06, + "loss": 0.067, + "step": 3920 + }, + { + "epoch": 1.5696594427244581, + "grad_norm": 1.480615496635437, + "learning_rate": 5.109811751283514e-06, + "loss": 0.0594, + "step": 3930 + }, + { + "epoch": 1.5736542494756818, + "grad_norm": 2.4283230304718018, + "learning_rate": 5.0955504848830584e-06, + "loss": 0.0679, + "step": 3940 + }, + { + "epoch": 1.577649056226905, + "grad_norm": 2.572091817855835, + "learning_rate": 5.081289218482601e-06, + "loss": 0.0686, + "step": 3950 + }, + { + "epoch": 1.5816438629781284, + "grad_norm": 3.342036247253418, + "learning_rate": 5.067027952082145e-06, + "loss": 0.0756, + "step": 3960 + }, + { + "epoch": 1.5856386697293519, + "grad_norm": 2.1152732372283936, + "learning_rate": 5.052766685681689e-06, + "loss": 0.0761, + "step": 3970 + }, + { + "epoch": 1.5896334764805753, + "grad_norm": 1.9615633487701416, + "learning_rate": 5.038505419281232e-06, + "loss": 0.0593, + "step": 3980 + }, + { + "epoch": 1.5936282832317987, + "grad_norm": 2.9293763637542725, + "learning_rate": 5.024244152880776e-06, + "loss": 0.0679, + "step": 3990 + }, + { + "epoch": 1.597623089983022, + "grad_norm": 1.6003406047821045, + "learning_rate": 5.0099828864803205e-06, + "loss": 0.0601, + "step": 4000 + }, + { + "epoch": 1.6016178967342456, + "grad_norm": 2.523336410522461, + "learning_rate": 4.995721620079864e-06, + "loss": 0.0821, + "step": 4010 + }, + { + "epoch": 1.6056127034854688, + "grad_norm": 2.5983314514160156, + "learning_rate": 4.981460353679407e-06, + "loss": 0.0737, + "step": 4020 + }, + { + "epoch": 1.6096075102366925, + "grad_norm": 3.036498546600342, + "learning_rate": 4.96719908727895e-06, + "loss": 0.072, + "step": 4030 + }, + { + "epoch": 1.6136023169879157, + "grad_norm": 2.7793092727661133, + "learning_rate": 4.952937820878494e-06, + "loss": 0.065, + "step": 4040 + }, + { + "epoch": 1.617597123739139, + "grad_norm": 2.809861183166504, + "learning_rate": 4.9386765544780376e-06, + "loss": 0.0669, + "step": 4050 + }, + { + "epoch": 1.6215919304903625, + "grad_norm": 3.808748722076416, + "learning_rate": 4.924415288077582e-06, + "loss": 0.0695, + "step": 4060 + }, + { + "epoch": 1.625586737241586, + "grad_norm": 2.0795702934265137, + "learning_rate": 4.910154021677126e-06, + "loss": 0.0727, + "step": 4070 + }, + { + "epoch": 1.6295815439928094, + "grad_norm": 2.0030627250671387, + "learning_rate": 4.895892755276669e-06, + "loss": 0.0825, + "step": 4080 + }, + { + "epoch": 1.6335763507440326, + "grad_norm": 2.6505062580108643, + "learning_rate": 4.881631488876213e-06, + "loss": 0.0629, + "step": 4090 + }, + { + "epoch": 1.6375711574952563, + "grad_norm": 2.439034938812256, + "learning_rate": 4.867370222475756e-06, + "loss": 0.0705, + "step": 4100 + }, + { + "epoch": 1.6415659642464795, + "grad_norm": 2.9876742362976074, + "learning_rate": 4.8531089560753e-06, + "loss": 0.0666, + "step": 4110 + }, + { + "epoch": 1.6455607709977031, + "grad_norm": 1.9269096851348877, + "learning_rate": 4.838847689674844e-06, + "loss": 0.0498, + "step": 4120 + }, + { + "epoch": 1.6495555777489264, + "grad_norm": 2.0127387046813965, + "learning_rate": 4.824586423274387e-06, + "loss": 0.0761, + "step": 4130 + }, + { + "epoch": 1.6535503845001498, + "grad_norm": 2.3247992992401123, + "learning_rate": 4.810325156873931e-06, + "loss": 0.0614, + "step": 4140 + }, + { + "epoch": 1.6575451912513732, + "grad_norm": 4.898730278015137, + "learning_rate": 4.796063890473474e-06, + "loss": 0.0706, + "step": 4150 + }, + { + "epoch": 1.6615399980025967, + "grad_norm": 2.9095699787139893, + "learning_rate": 4.781802624073018e-06, + "loss": 0.0779, + "step": 4160 + }, + { + "epoch": 1.66553480475382, + "grad_norm": 1.6168105602264404, + "learning_rate": 4.767541357672562e-06, + "loss": 0.0672, + "step": 4170 + }, + { + "epoch": 1.6695296115050433, + "grad_norm": 2.98400616645813, + "learning_rate": 4.753280091272106e-06, + "loss": 0.062, + "step": 4180 + }, + { + "epoch": 1.673524418256267, + "grad_norm": 2.4829282760620117, + "learning_rate": 4.739018824871649e-06, + "loss": 0.0596, + "step": 4190 + }, + { + "epoch": 1.6775192250074902, + "grad_norm": 1.7955764532089233, + "learning_rate": 4.724757558471192e-06, + "loss": 0.0552, + "step": 4200 + }, + { + "epoch": 1.6815140317587138, + "grad_norm": 1.729946255683899, + "learning_rate": 4.710496292070736e-06, + "loss": 0.0716, + "step": 4210 + }, + { + "epoch": 1.685508838509937, + "grad_norm": 1.765784740447998, + "learning_rate": 4.6962350256702795e-06, + "loss": 0.0638, + "step": 4220 + }, + { + "epoch": 1.6895036452611605, + "grad_norm": 2.498591423034668, + "learning_rate": 4.681973759269824e-06, + "loss": 0.0677, + "step": 4230 + }, + { + "epoch": 1.693498452012384, + "grad_norm": 2.2129974365234375, + "learning_rate": 4.667712492869368e-06, + "loss": 0.0591, + "step": 4240 + }, + { + "epoch": 1.6974932587636073, + "grad_norm": 2.0700387954711914, + "learning_rate": 4.653451226468911e-06, + "loss": 0.0583, + "step": 4250 + }, + { + "epoch": 1.7014880655148308, + "grad_norm": 2.034822463989258, + "learning_rate": 4.639189960068454e-06, + "loss": 0.0645, + "step": 4260 + }, + { + "epoch": 1.705482872266054, + "grad_norm": 3.041527271270752, + "learning_rate": 4.624928693667998e-06, + "loss": 0.0613, + "step": 4270 + }, + { + "epoch": 1.7094776790172777, + "grad_norm": 2.3591227531433105, + "learning_rate": 4.6106674272675416e-06, + "loss": 0.0683, + "step": 4280 + }, + { + "epoch": 1.7134724857685009, + "grad_norm": 2.1549127101898193, + "learning_rate": 4.596406160867085e-06, + "loss": 0.0708, + "step": 4290 + }, + { + "epoch": 1.7174672925197245, + "grad_norm": 2.118690013885498, + "learning_rate": 4.582144894466629e-06, + "loss": 0.0717, + "step": 4300 + }, + { + "epoch": 1.7214620992709477, + "grad_norm": 3.4352660179138184, + "learning_rate": 4.567883628066173e-06, + "loss": 0.0764, + "step": 4310 + }, + { + "epoch": 1.7254569060221712, + "grad_norm": 2.167402744293213, + "learning_rate": 4.553622361665716e-06, + "loss": 0.0701, + "step": 4320 + }, + { + "epoch": 1.7294517127733946, + "grad_norm": 2.3733415603637695, + "learning_rate": 4.53936109526526e-06, + "loss": 0.0599, + "step": 4330 + }, + { + "epoch": 1.733446519524618, + "grad_norm": 2.605437994003296, + "learning_rate": 4.525099828864804e-06, + "loss": 0.0606, + "step": 4340 + }, + { + "epoch": 1.7374413262758415, + "grad_norm": 2.1807992458343506, + "learning_rate": 4.510838562464347e-06, + "loss": 0.0633, + "step": 4350 + }, + { + "epoch": 1.7414361330270647, + "grad_norm": 2.621598720550537, + "learning_rate": 4.496577296063891e-06, + "loss": 0.0609, + "step": 4360 + }, + { + "epoch": 1.7454309397782883, + "grad_norm": 1.5117213726043701, + "learning_rate": 4.482316029663434e-06, + "loss": 0.0588, + "step": 4370 + }, + { + "epoch": 1.7494257465295115, + "grad_norm": 2.744356393814087, + "learning_rate": 4.468054763262978e-06, + "loss": 0.0683, + "step": 4380 + }, + { + "epoch": 1.7534205532807352, + "grad_norm": 2.6566450595855713, + "learning_rate": 4.4537934968625215e-06, + "loss": 0.0748, + "step": 4390 + }, + { + "epoch": 1.7574153600319584, + "grad_norm": 2.3438005447387695, + "learning_rate": 4.439532230462066e-06, + "loss": 0.0739, + "step": 4400 + }, + { + "epoch": 1.7614101667831819, + "grad_norm": 2.7191786766052246, + "learning_rate": 4.425270964061609e-06, + "loss": 0.0621, + "step": 4410 + }, + { + "epoch": 1.7654049735344053, + "grad_norm": 1.86728036403656, + "learning_rate": 4.411009697661153e-06, + "loss": 0.0558, + "step": 4420 + }, + { + "epoch": 1.7693997802856287, + "grad_norm": 2.096303939819336, + "learning_rate": 4.396748431260696e-06, + "loss": 0.0624, + "step": 4430 + }, + { + "epoch": 1.7733945870368522, + "grad_norm": 2.400118350982666, + "learning_rate": 4.38248716486024e-06, + "loss": 0.0852, + "step": 4440 + }, + { + "epoch": 1.7773893937880754, + "grad_norm": 2.4184322357177734, + "learning_rate": 4.3682258984597835e-06, + "loss": 0.0623, + "step": 4450 + }, + { + "epoch": 1.781384200539299, + "grad_norm": 2.340954065322876, + "learning_rate": 4.353964632059327e-06, + "loss": 0.0609, + "step": 4460 + }, + { + "epoch": 1.7853790072905222, + "grad_norm": 2.7005741596221924, + "learning_rate": 4.339703365658871e-06, + "loss": 0.0597, + "step": 4470 + }, + { + "epoch": 1.789373814041746, + "grad_norm": 3.050964832305908, + "learning_rate": 4.325442099258414e-06, + "loss": 0.0727, + "step": 4480 + }, + { + "epoch": 1.793368620792969, + "grad_norm": 2.0838069915771484, + "learning_rate": 4.311180832857958e-06, + "loss": 0.0705, + "step": 4490 + }, + { + "epoch": 1.7973634275441925, + "grad_norm": 3.0605244636535645, + "learning_rate": 4.296919566457502e-06, + "loss": 0.0596, + "step": 4500 + }, + { + "epoch": 1.801358234295416, + "grad_norm": 2.4080119132995605, + "learning_rate": 4.2826583000570456e-06, + "loss": 0.06, + "step": 4510 + }, + { + "epoch": 1.8053530410466394, + "grad_norm": 2.491190195083618, + "learning_rate": 4.268397033656589e-06, + "loss": 0.0611, + "step": 4520 + }, + { + "epoch": 1.8093478477978628, + "grad_norm": 2.252627372741699, + "learning_rate": 4.254135767256133e-06, + "loss": 0.0595, + "step": 4530 + }, + { + "epoch": 1.813342654549086, + "grad_norm": 2.160486936569214, + "learning_rate": 4.239874500855676e-06, + "loss": 0.0667, + "step": 4540 + }, + { + "epoch": 1.8173374613003097, + "grad_norm": 2.1602847576141357, + "learning_rate": 4.225613234455219e-06, + "loss": 0.057, + "step": 4550 + }, + { + "epoch": 1.821332268051533, + "grad_norm": 5.302746295928955, + "learning_rate": 4.2113519680547635e-06, + "loss": 0.0759, + "step": 4560 + }, + { + "epoch": 1.8253270748027566, + "grad_norm": 1.646782398223877, + "learning_rate": 4.197090701654308e-06, + "loss": 0.0582, + "step": 4570 + }, + { + "epoch": 1.8293218815539798, + "grad_norm": 2.2816812992095947, + "learning_rate": 4.182829435253851e-06, + "loss": 0.069, + "step": 4580 + }, + { + "epoch": 1.8333166883052032, + "grad_norm": 1.8131096363067627, + "learning_rate": 4.168568168853395e-06, + "loss": 0.0622, + "step": 4590 + }, + { + "epoch": 1.8373114950564267, + "grad_norm": 2.061286687850952, + "learning_rate": 4.154306902452938e-06, + "loss": 0.0617, + "step": 4600 + }, + { + "epoch": 1.84130630180765, + "grad_norm": 2.60317063331604, + "learning_rate": 4.1400456360524814e-06, + "loss": 0.0451, + "step": 4610 + }, + { + "epoch": 1.8453011085588735, + "grad_norm": 2.2564353942871094, + "learning_rate": 4.1257843696520255e-06, + "loss": 0.0604, + "step": 4620 + }, + { + "epoch": 1.8492959153100967, + "grad_norm": 1.961222529411316, + "learning_rate": 4.111523103251569e-06, + "loss": 0.0665, + "step": 4630 + }, + { + "epoch": 1.8532907220613204, + "grad_norm": 2.076077938079834, + "learning_rate": 4.097261836851113e-06, + "loss": 0.0601, + "step": 4640 + }, + { + "epoch": 1.8572855288125436, + "grad_norm": 2.159482002258301, + "learning_rate": 4.083000570450656e-06, + "loss": 0.0597, + "step": 4650 + }, + { + "epoch": 1.8612803355637673, + "grad_norm": 1.7660284042358398, + "learning_rate": 4.0687393040502e-06, + "loss": 0.0618, + "step": 4660 + }, + { + "epoch": 1.8652751423149905, + "grad_norm": 2.1984660625457764, + "learning_rate": 4.054478037649744e-06, + "loss": 0.0613, + "step": 4670 + }, + { + "epoch": 1.869269949066214, + "grad_norm": 2.328014373779297, + "learning_rate": 4.0402167712492876e-06, + "loss": 0.0585, + "step": 4680 + }, + { + "epoch": 1.8732647558174373, + "grad_norm": 1.8420252799987793, + "learning_rate": 4.025955504848831e-06, + "loss": 0.0679, + "step": 4690 + }, + { + "epoch": 1.8772595625686608, + "grad_norm": 1.9796903133392334, + "learning_rate": 4.011694238448375e-06, + "loss": 0.0695, + "step": 4700 + }, + { + "epoch": 1.8812543693198842, + "grad_norm": 1.705096960067749, + "learning_rate": 3.997432972047918e-06, + "loss": 0.0598, + "step": 4710 + }, + { + "epoch": 1.8852491760711074, + "grad_norm": 2.8509531021118164, + "learning_rate": 3.983171705647461e-06, + "loss": 0.066, + "step": 4720 + }, + { + "epoch": 1.889243982822331, + "grad_norm": 1.7244213819503784, + "learning_rate": 3.9689104392470055e-06, + "loss": 0.0629, + "step": 4730 + }, + { + "epoch": 1.8932387895735543, + "grad_norm": 2.2036972045898438, + "learning_rate": 3.954649172846549e-06, + "loss": 0.0616, + "step": 4740 + }, + { + "epoch": 1.897233596324778, + "grad_norm": 2.4186980724334717, + "learning_rate": 3.940387906446093e-06, + "loss": 0.0639, + "step": 4750 + }, + { + "epoch": 1.9012284030760012, + "grad_norm": 2.057669162750244, + "learning_rate": 3.926126640045637e-06, + "loss": 0.06, + "step": 4760 + }, + { + "epoch": 1.9052232098272246, + "grad_norm": 1.5648422241210938, + "learning_rate": 3.91186537364518e-06, + "loss": 0.0517, + "step": 4770 + }, + { + "epoch": 1.909218016578448, + "grad_norm": 2.056607246398926, + "learning_rate": 3.897604107244723e-06, + "loss": 0.0596, + "step": 4780 + }, + { + "epoch": 1.9132128233296715, + "grad_norm": 2.0157339572906494, + "learning_rate": 3.8833428408442675e-06, + "loss": 0.0592, + "step": 4790 + }, + { + "epoch": 1.917207630080895, + "grad_norm": 2.3085403442382812, + "learning_rate": 3.869081574443811e-06, + "loss": 0.0656, + "step": 4800 + }, + { + "epoch": 1.921202436832118, + "grad_norm": 3.6553146839141846, + "learning_rate": 3.854820308043354e-06, + "loss": 0.0732, + "step": 4810 + }, + { + "epoch": 1.9251972435833418, + "grad_norm": 2.9719808101654053, + "learning_rate": 3.840559041642898e-06, + "loss": 0.0632, + "step": 4820 + }, + { + "epoch": 1.929192050334565, + "grad_norm": 2.629744052886963, + "learning_rate": 3.826297775242442e-06, + "loss": 0.0707, + "step": 4830 + }, + { + "epoch": 1.9331868570857886, + "grad_norm": 2.4601378440856934, + "learning_rate": 3.8120365088419854e-06, + "loss": 0.0578, + "step": 4840 + }, + { + "epoch": 1.9371816638370118, + "grad_norm": 1.9864972829818726, + "learning_rate": 3.797775242441529e-06, + "loss": 0.0605, + "step": 4850 + }, + { + "epoch": 1.9411764705882353, + "grad_norm": 2.0608339309692383, + "learning_rate": 3.7835139760410728e-06, + "loss": 0.0552, + "step": 4860 + }, + { + "epoch": 1.9451712773394587, + "grad_norm": 1.912820816040039, + "learning_rate": 3.769252709640616e-06, + "loss": 0.0663, + "step": 4870 + }, + { + "epoch": 1.9491660840906821, + "grad_norm": 2.3183212280273438, + "learning_rate": 3.75499144324016e-06, + "loss": 0.0812, + "step": 4880 + }, + { + "epoch": 1.9531608908419056, + "grad_norm": 2.262383222579956, + "learning_rate": 3.740730176839704e-06, + "loss": 0.0595, + "step": 4890 + }, + { + "epoch": 1.9571556975931288, + "grad_norm": 2.6391637325286865, + "learning_rate": 3.726468910439247e-06, + "loss": 0.066, + "step": 4900 + }, + { + "epoch": 1.9611505043443525, + "grad_norm": 2.2891080379486084, + "learning_rate": 3.712207644038791e-06, + "loss": 0.067, + "step": 4910 + }, + { + "epoch": 1.9651453110955757, + "grad_norm": 3.415213108062744, + "learning_rate": 3.6979463776383344e-06, + "loss": 0.0862, + "step": 4920 + }, + { + "epoch": 1.9691401178467993, + "grad_norm": 2.93517804145813, + "learning_rate": 3.683685111237878e-06, + "loss": 0.0794, + "step": 4930 + }, + { + "epoch": 1.9731349245980225, + "grad_norm": 2.152313470840454, + "learning_rate": 3.669423844837422e-06, + "loss": 0.0619, + "step": 4940 + }, + { + "epoch": 1.977129731349246, + "grad_norm": 2.799994945526123, + "learning_rate": 3.6551625784369654e-06, + "loss": 0.0715, + "step": 4950 + }, + { + "epoch": 1.9811245381004694, + "grad_norm": 1.8129743337631226, + "learning_rate": 3.640901312036509e-06, + "loss": 0.0635, + "step": 4960 + }, + { + "epoch": 1.9851193448516928, + "grad_norm": 3.1220531463623047, + "learning_rate": 3.6266400456360527e-06, + "loss": 0.0705, + "step": 4970 + }, + { + "epoch": 1.9891141516029163, + "grad_norm": 2.8412084579467773, + "learning_rate": 3.6123787792355964e-06, + "loss": 0.0602, + "step": 4980 + }, + { + "epoch": 1.9931089583541395, + "grad_norm": 1.7142022848129272, + "learning_rate": 3.5981175128351405e-06, + "loss": 0.07, + "step": 4990 + }, + { + "epoch": 1.9971037651053631, + "grad_norm": 1.686712622642517, + "learning_rate": 3.5838562464346838e-06, + "loss": 0.0605, + "step": 5000 + }, + { + "epoch": 1.9971037651053631, + "eval_loss": 0.04570349305868149, + "eval_runtime": 11145.7877, + "eval_samples_per_second": 1.797, + "eval_steps_per_second": 0.225, + "eval_wer": 3.9939936648508825, + "step": 5000 + }, + { + "epoch": 2.000798961350245, + "grad_norm": 1.432715654373169, + "learning_rate": 3.5695949800342274e-06, + "loss": 0.0471, + "step": 5010 + }, + { + "epoch": 2.004793768101468, + "grad_norm": 1.166917085647583, + "learning_rate": 3.555333713633771e-06, + "loss": 0.0224, + "step": 5020 + }, + { + "epoch": 2.0087885748526917, + "grad_norm": 1.1348071098327637, + "learning_rate": 3.5410724472333148e-06, + "loss": 0.0226, + "step": 5030 + }, + { + "epoch": 2.012783381603915, + "grad_norm": 0.8629413843154907, + "learning_rate": 3.526811180832858e-06, + "loss": 0.0245, + "step": 5040 + }, + { + "epoch": 2.016778188355138, + "grad_norm": 0.9867520332336426, + "learning_rate": 3.512549914432402e-06, + "loss": 0.0219, + "step": 5050 + }, + { + "epoch": 2.0207729951063618, + "grad_norm": 3.1294004917144775, + "learning_rate": 3.4982886480319454e-06, + "loss": 0.0248, + "step": 5060 + }, + { + "epoch": 2.024767801857585, + "grad_norm": 1.323643445968628, + "learning_rate": 3.484027381631489e-06, + "loss": 0.029, + "step": 5070 + }, + { + "epoch": 2.0287626086088086, + "grad_norm": 1.3909205198287964, + "learning_rate": 3.469766115231033e-06, + "loss": 0.0237, + "step": 5080 + }, + { + "epoch": 2.032757415360032, + "grad_norm": 1.3001748323440552, + "learning_rate": 3.4555048488305764e-06, + "loss": 0.0248, + "step": 5090 + }, + { + "epoch": 2.0367522221112555, + "grad_norm": 1.1780959367752075, + "learning_rate": 3.44124358243012e-06, + "loss": 0.0186, + "step": 5100 + }, + { + "epoch": 2.0407470288624787, + "grad_norm": 1.0736746788024902, + "learning_rate": 3.4269823160296637e-06, + "loss": 0.0226, + "step": 5110 + }, + { + "epoch": 2.0447418356137024, + "grad_norm": 1.5082242488861084, + "learning_rate": 3.4127210496292074e-06, + "loss": 0.0247, + "step": 5120 + }, + { + "epoch": 2.0487366423649256, + "grad_norm": 1.5528379678726196, + "learning_rate": 3.3984597832287506e-06, + "loss": 0.0228, + "step": 5130 + }, + { + "epoch": 2.052731449116149, + "grad_norm": 1.1033167839050293, + "learning_rate": 3.3841985168282947e-06, + "loss": 0.0245, + "step": 5140 + }, + { + "epoch": 2.0567262558673725, + "grad_norm": 1.5862687826156616, + "learning_rate": 3.3699372504278384e-06, + "loss": 0.0244, + "step": 5150 + }, + { + "epoch": 2.0607210626185957, + "grad_norm": 1.2361029386520386, + "learning_rate": 3.3556759840273816e-06, + "loss": 0.0247, + "step": 5160 + }, + { + "epoch": 2.0647158693698193, + "grad_norm": 2.002153158187866, + "learning_rate": 3.3414147176269257e-06, + "loss": 0.0265, + "step": 5170 + }, + { + "epoch": 2.0687106761210425, + "grad_norm": 0.7162835597991943, + "learning_rate": 3.327153451226469e-06, + "loss": 0.0201, + "step": 5180 + }, + { + "epoch": 2.072705482872266, + "grad_norm": 1.4344475269317627, + "learning_rate": 3.3128921848260127e-06, + "loss": 0.0328, + "step": 5190 + }, + { + "epoch": 2.0767002896234894, + "grad_norm": 1.3027048110961914, + "learning_rate": 3.2986309184255568e-06, + "loss": 0.0201, + "step": 5200 + }, + { + "epoch": 2.080695096374713, + "grad_norm": 1.60873281955719, + "learning_rate": 3.2843696520251e-06, + "loss": 0.0257, + "step": 5210 + }, + { + "epoch": 2.0846899031259363, + "grad_norm": 3.3106722831726074, + "learning_rate": 3.2701083856246437e-06, + "loss": 0.0338, + "step": 5220 + }, + { + "epoch": 2.0886847098771595, + "grad_norm": 1.4701682329177856, + "learning_rate": 3.2558471192241873e-06, + "loss": 0.0188, + "step": 5230 + }, + { + "epoch": 2.092679516628383, + "grad_norm": 1.71177077293396, + "learning_rate": 3.241585852823731e-06, + "loss": 0.0192, + "step": 5240 + }, + { + "epoch": 2.0966743233796064, + "grad_norm": 1.5331566333770752, + "learning_rate": 3.2273245864232743e-06, + "loss": 0.0256, + "step": 5250 + }, + { + "epoch": 2.10066913013083, + "grad_norm": 1.1146838665008545, + "learning_rate": 3.2130633200228184e-06, + "loss": 0.0231, + "step": 5260 + }, + { + "epoch": 2.104663936882053, + "grad_norm": 1.6950159072875977, + "learning_rate": 3.198802053622362e-06, + "loss": 0.0279, + "step": 5270 + }, + { + "epoch": 2.108658743633277, + "grad_norm": 2.177121162414551, + "learning_rate": 3.1845407872219053e-06, + "loss": 0.0225, + "step": 5280 + }, + { + "epoch": 2.1126535503845, + "grad_norm": 1.6834362745285034, + "learning_rate": 3.1702795208214494e-06, + "loss": 0.0292, + "step": 5290 + }, + { + "epoch": 2.1166483571357237, + "grad_norm": 1.275499939918518, + "learning_rate": 3.1560182544209926e-06, + "loss": 0.0214, + "step": 5300 + }, + { + "epoch": 2.120643163886947, + "grad_norm": 1.7945035696029663, + "learning_rate": 3.1417569880205367e-06, + "loss": 0.0208, + "step": 5310 + }, + { + "epoch": 2.12463797063817, + "grad_norm": 1.075283169746399, + "learning_rate": 3.1274957216200804e-06, + "loss": 0.0151, + "step": 5320 + }, + { + "epoch": 2.128632777389394, + "grad_norm": 1.7752685546875, + "learning_rate": 3.1132344552196236e-06, + "loss": 0.024, + "step": 5330 + }, + { + "epoch": 2.132627584140617, + "grad_norm": 1.9435361623764038, + "learning_rate": 3.0989731888191677e-06, + "loss": 0.0225, + "step": 5340 + }, + { + "epoch": 2.1366223908918407, + "grad_norm": 1.636218547821045, + "learning_rate": 3.084711922418711e-06, + "loss": 0.023, + "step": 5350 + }, + { + "epoch": 2.140617197643064, + "grad_norm": 1.1532355546951294, + "learning_rate": 3.0704506560182546e-06, + "loss": 0.0258, + "step": 5360 + }, + { + "epoch": 2.1446120043942876, + "grad_norm": 1.8517895936965942, + "learning_rate": 3.0561893896177987e-06, + "loss": 0.0263, + "step": 5370 + }, + { + "epoch": 2.1486068111455108, + "grad_norm": 1.0594685077667236, + "learning_rate": 3.041928123217342e-06, + "loss": 0.0316, + "step": 5380 + }, + { + "epoch": 2.1526016178967344, + "grad_norm": 1.580573320388794, + "learning_rate": 3.0276668568168852e-06, + "loss": 0.0212, + "step": 5390 + }, + { + "epoch": 2.1565964246479576, + "grad_norm": 1.9935818910598755, + "learning_rate": 3.0134055904164293e-06, + "loss": 0.0206, + "step": 5400 + }, + { + "epoch": 2.160591231399181, + "grad_norm": 1.528911828994751, + "learning_rate": 2.999144324015973e-06, + "loss": 0.0293, + "step": 5410 + }, + { + "epoch": 2.1645860381504045, + "grad_norm": 1.330037236213684, + "learning_rate": 2.9848830576155162e-06, + "loss": 0.0214, + "step": 5420 + }, + { + "epoch": 2.1685808449016277, + "grad_norm": 1.2586803436279297, + "learning_rate": 2.9706217912150603e-06, + "loss": 0.0343, + "step": 5430 + }, + { + "epoch": 2.1725756516528514, + "grad_norm": 1.6885484457015991, + "learning_rate": 2.9563605248146036e-06, + "loss": 0.0238, + "step": 5440 + }, + { + "epoch": 2.1765704584040746, + "grad_norm": 1.1901226043701172, + "learning_rate": 2.9420992584141473e-06, + "loss": 0.0217, + "step": 5450 + }, + { + "epoch": 2.1805652651552982, + "grad_norm": 1.4856065511703491, + "learning_rate": 2.9278379920136913e-06, + "loss": 0.0215, + "step": 5460 + }, + { + "epoch": 2.1845600719065215, + "grad_norm": 0.9664995670318604, + "learning_rate": 2.9135767256132346e-06, + "loss": 0.0195, + "step": 5470 + }, + { + "epoch": 2.188554878657745, + "grad_norm": 1.7648258209228516, + "learning_rate": 2.8993154592127783e-06, + "loss": 0.0223, + "step": 5480 + }, + { + "epoch": 2.1925496854089683, + "grad_norm": 1.3821454048156738, + "learning_rate": 2.885054192812322e-06, + "loss": 0.023, + "step": 5490 + }, + { + "epoch": 2.196544492160192, + "grad_norm": 0.885549783706665, + "learning_rate": 2.8707929264118656e-06, + "loss": 0.0252, + "step": 5500 + }, + { + "epoch": 2.200539298911415, + "grad_norm": 0.976472020149231, + "learning_rate": 2.856531660011409e-06, + "loss": 0.0235, + "step": 5510 + }, + { + "epoch": 2.2045341056626384, + "grad_norm": 1.4620656967163086, + "learning_rate": 2.842270393610953e-06, + "loss": 0.0205, + "step": 5520 + }, + { + "epoch": 2.208528912413862, + "grad_norm": 2.453080177307129, + "learning_rate": 2.8280091272104966e-06, + "loss": 0.0338, + "step": 5530 + }, + { + "epoch": 2.2125237191650853, + "grad_norm": 0.9037997126579285, + "learning_rate": 2.81374786081004e-06, + "loss": 0.023, + "step": 5540 + }, + { + "epoch": 2.216518525916309, + "grad_norm": 1.1964457035064697, + "learning_rate": 2.799486594409584e-06, + "loss": 0.0222, + "step": 5550 + }, + { + "epoch": 2.220513332667532, + "grad_norm": 1.7553346157073975, + "learning_rate": 2.785225328009127e-06, + "loss": 0.0246, + "step": 5560 + }, + { + "epoch": 2.224508139418756, + "grad_norm": 1.4776095151901245, + "learning_rate": 2.770964061608671e-06, + "loss": 0.0246, + "step": 5570 + }, + { + "epoch": 2.228502946169979, + "grad_norm": 1.3192557096481323, + "learning_rate": 2.756702795208215e-06, + "loss": 0.0233, + "step": 5580 + }, + { + "epoch": 2.2324977529212022, + "grad_norm": 1.866843819618225, + "learning_rate": 2.7424415288077582e-06, + "loss": 0.0266, + "step": 5590 + }, + { + "epoch": 2.236492559672426, + "grad_norm": 1.6727417707443237, + "learning_rate": 2.728180262407302e-06, + "loss": 0.0275, + "step": 5600 + }, + { + "epoch": 2.240487366423649, + "grad_norm": 1.2342602014541626, + "learning_rate": 2.7139189960068456e-06, + "loss": 0.03, + "step": 5610 + }, + { + "epoch": 2.2444821731748728, + "grad_norm": 1.5016778707504272, + "learning_rate": 2.6996577296063892e-06, + "loss": 0.0287, + "step": 5620 + }, + { + "epoch": 2.248476979926096, + "grad_norm": 1.8288167715072632, + "learning_rate": 2.6853964632059333e-06, + "loss": 0.0283, + "step": 5630 + }, + { + "epoch": 2.2524717866773196, + "grad_norm": 0.8718370795249939, + "learning_rate": 2.6711351968054766e-06, + "loss": 0.0189, + "step": 5640 + }, + { + "epoch": 2.256466593428543, + "grad_norm": 1.2181274890899658, + "learning_rate": 2.6568739304050202e-06, + "loss": 0.0217, + "step": 5650 + }, + { + "epoch": 2.2604614001797665, + "grad_norm": 1.3012434244155884, + "learning_rate": 2.642612664004564e-06, + "loss": 0.0228, + "step": 5660 + }, + { + "epoch": 2.2644562069309897, + "grad_norm": 1.0627079010009766, + "learning_rate": 2.6283513976041076e-06, + "loss": 0.0195, + "step": 5670 + }, + { + "epoch": 2.2684510136822134, + "grad_norm": 1.3637123107910156, + "learning_rate": 2.614090131203651e-06, + "loss": 0.0189, + "step": 5680 + }, + { + "epoch": 2.2724458204334366, + "grad_norm": 1.2548840045928955, + "learning_rate": 2.599828864803195e-06, + "loss": 0.0303, + "step": 5690 + }, + { + "epoch": 2.27644062718466, + "grad_norm": 1.2600808143615723, + "learning_rate": 2.5855675984027386e-06, + "loss": 0.0272, + "step": 5700 + }, + { + "epoch": 2.2804354339358834, + "grad_norm": 1.089040994644165, + "learning_rate": 2.571306332002282e-06, + "loss": 0.024, + "step": 5710 + }, + { + "epoch": 2.2844302406871067, + "grad_norm": 1.4958359003067017, + "learning_rate": 2.557045065601826e-06, + "loss": 0.0263, + "step": 5720 + }, + { + "epoch": 2.2884250474383303, + "grad_norm": 1.3157131671905518, + "learning_rate": 2.542783799201369e-06, + "loss": 0.0211, + "step": 5730 + }, + { + "epoch": 2.2924198541895535, + "grad_norm": 1.8704617023468018, + "learning_rate": 2.528522532800913e-06, + "loss": 0.027, + "step": 5740 + }, + { + "epoch": 2.296414660940777, + "grad_norm": 1.4256103038787842, + "learning_rate": 2.514261266400457e-06, + "loss": 0.0314, + "step": 5750 + }, + { + "epoch": 2.3004094676920004, + "grad_norm": 1.2393786907196045, + "learning_rate": 2.5e-06, + "loss": 0.0231, + "step": 5760 + }, + { + "epoch": 2.3044042744432236, + "grad_norm": 1.3867117166519165, + "learning_rate": 2.485738733599544e-06, + "loss": 0.0289, + "step": 5770 + }, + { + "epoch": 2.3083990811944473, + "grad_norm": 1.3235056400299072, + "learning_rate": 2.4714774671990875e-06, + "loss": 0.0262, + "step": 5780 + }, + { + "epoch": 2.3123938879456705, + "grad_norm": 0.9694569110870361, + "learning_rate": 2.4572162007986312e-06, + "loss": 0.0196, + "step": 5790 + }, + { + "epoch": 2.316388694696894, + "grad_norm": 0.7398777008056641, + "learning_rate": 2.442954934398175e-06, + "loss": 0.0284, + "step": 5800 + }, + { + "epoch": 2.3203835014481173, + "grad_norm": 1.5198540687561035, + "learning_rate": 2.428693667997718e-06, + "loss": 0.0184, + "step": 5810 + }, + { + "epoch": 2.324378308199341, + "grad_norm": 1.4485077857971191, + "learning_rate": 2.414432401597262e-06, + "loss": 0.017, + "step": 5820 + }, + { + "epoch": 2.328373114950564, + "grad_norm": 0.886734664440155, + "learning_rate": 2.400171135196806e-06, + "loss": 0.0224, + "step": 5830 + }, + { + "epoch": 2.332367921701788, + "grad_norm": 0.999384880065918, + "learning_rate": 2.3859098687963496e-06, + "loss": 0.0357, + "step": 5840 + }, + { + "epoch": 2.336362728453011, + "grad_norm": 2.2749531269073486, + "learning_rate": 2.371648602395893e-06, + "loss": 0.0178, + "step": 5850 + }, + { + "epoch": 2.3403575352042347, + "grad_norm": 1.3109686374664307, + "learning_rate": 2.3573873359954365e-06, + "loss": 0.0247, + "step": 5860 + }, + { + "epoch": 2.344352341955458, + "grad_norm": 1.4980666637420654, + "learning_rate": 2.34312606959498e-06, + "loss": 0.0221, + "step": 5870 + }, + { + "epoch": 2.348347148706681, + "grad_norm": 1.2611418962478638, + "learning_rate": 2.328864803194524e-06, + "loss": 0.0302, + "step": 5880 + }, + { + "epoch": 2.352341955457905, + "grad_norm": 1.4152772426605225, + "learning_rate": 2.3146035367940675e-06, + "loss": 0.0165, + "step": 5890 + }, + { + "epoch": 2.356336762209128, + "grad_norm": 2.642857551574707, + "learning_rate": 2.300342270393611e-06, + "loss": 0.0257, + "step": 5900 + }, + { + "epoch": 2.3603315689603517, + "grad_norm": 1.5568487644195557, + "learning_rate": 2.286081003993155e-06, + "loss": 0.0248, + "step": 5910 + }, + { + "epoch": 2.364326375711575, + "grad_norm": 0.7589583992958069, + "learning_rate": 2.2718197375926985e-06, + "loss": 0.0208, + "step": 5920 + }, + { + "epoch": 2.3683211824627985, + "grad_norm": 1.232725977897644, + "learning_rate": 2.257558471192242e-06, + "loss": 0.0296, + "step": 5930 + }, + { + "epoch": 2.3723159892140218, + "grad_norm": 1.2233147621154785, + "learning_rate": 2.2432972047917854e-06, + "loss": 0.0173, + "step": 5940 + }, + { + "epoch": 2.376310795965245, + "grad_norm": 1.1067813634872437, + "learning_rate": 2.229035938391329e-06, + "loss": 0.0185, + "step": 5950 + }, + { + "epoch": 2.3803056027164686, + "grad_norm": 1.5446206331253052, + "learning_rate": 2.214774671990873e-06, + "loss": 0.0207, + "step": 5960 + }, + { + "epoch": 2.384300409467692, + "grad_norm": 2.2748281955718994, + "learning_rate": 2.2005134055904164e-06, + "loss": 0.0259, + "step": 5970 + }, + { + "epoch": 2.3882952162189155, + "grad_norm": 1.4236910343170166, + "learning_rate": 2.18625213918996e-06, + "loss": 0.0167, + "step": 5980 + }, + { + "epoch": 2.3922900229701387, + "grad_norm": 1.2497378587722778, + "learning_rate": 2.171990872789504e-06, + "loss": 0.0255, + "step": 5990 + }, + { + "epoch": 2.3962848297213624, + "grad_norm": 1.3303847312927246, + "learning_rate": 2.1577296063890475e-06, + "loss": 0.0217, + "step": 6000 + }, + { + "epoch": 2.4002796364725856, + "grad_norm": 0.9109508395195007, + "learning_rate": 2.143468339988591e-06, + "loss": 0.0171, + "step": 6010 + }, + { + "epoch": 2.4042744432238092, + "grad_norm": 1.2678133249282837, + "learning_rate": 2.129207073588135e-06, + "loss": 0.0185, + "step": 6020 + }, + { + "epoch": 2.4082692499750324, + "grad_norm": 0.8435303568840027, + "learning_rate": 2.1149458071876785e-06, + "loss": 0.0161, + "step": 6030 + }, + { + "epoch": 2.412264056726256, + "grad_norm": 0.9425959587097168, + "learning_rate": 2.100684540787222e-06, + "loss": 0.0185, + "step": 6040 + }, + { + "epoch": 2.4162588634774793, + "grad_norm": 0.7956383228302002, + "learning_rate": 2.086423274386766e-06, + "loss": 0.0212, + "step": 6050 + }, + { + "epoch": 2.4202536702287025, + "grad_norm": 1.771776795387268, + "learning_rate": 2.0721620079863095e-06, + "loss": 0.0225, + "step": 6060 + }, + { + "epoch": 2.424248476979926, + "grad_norm": 1.2092336416244507, + "learning_rate": 2.0579007415858527e-06, + "loss": 0.0205, + "step": 6070 + }, + { + "epoch": 2.4282432837311494, + "grad_norm": 0.7379654049873352, + "learning_rate": 2.043639475185397e-06, + "loss": 0.0212, + "step": 6080 + }, + { + "epoch": 2.432238090482373, + "grad_norm": 1.7100214958190918, + "learning_rate": 2.0293782087849405e-06, + "loss": 0.0219, + "step": 6090 + }, + { + "epoch": 2.4362328972335963, + "grad_norm": 1.4732763767242432, + "learning_rate": 2.0151169423844837e-06, + "loss": 0.0218, + "step": 6100 + }, + { + "epoch": 2.44022770398482, + "grad_norm": 1.1913158893585205, + "learning_rate": 2.0008556759840274e-06, + "loss": 0.0217, + "step": 6110 + }, + { + "epoch": 2.444222510736043, + "grad_norm": 1.7059197425842285, + "learning_rate": 1.986594409583571e-06, + "loss": 0.0238, + "step": 6120 + }, + { + "epoch": 2.4482173174872663, + "grad_norm": 1.668705701828003, + "learning_rate": 1.9723331431831148e-06, + "loss": 0.0151, + "step": 6130 + }, + { + "epoch": 2.45221212423849, + "grad_norm": 2.046867609024048, + "learning_rate": 1.9580718767826584e-06, + "loss": 0.0252, + "step": 6140 + }, + { + "epoch": 2.456206930989713, + "grad_norm": 1.5397975444793701, + "learning_rate": 1.943810610382202e-06, + "loss": 0.0221, + "step": 6150 + }, + { + "epoch": 2.460201737740937, + "grad_norm": 1.1507179737091064, + "learning_rate": 1.9295493439817458e-06, + "loss": 0.024, + "step": 6160 + }, + { + "epoch": 2.46419654449216, + "grad_norm": 1.250187873840332, + "learning_rate": 1.9152880775812894e-06, + "loss": 0.0212, + "step": 6170 + }, + { + "epoch": 2.4681913512433837, + "grad_norm": 2.1282033920288086, + "learning_rate": 1.9010268111808331e-06, + "loss": 0.0191, + "step": 6180 + }, + { + "epoch": 2.472186157994607, + "grad_norm": 1.8202413320541382, + "learning_rate": 1.8867655447803768e-06, + "loss": 0.0229, + "step": 6190 + }, + { + "epoch": 2.4761809647458306, + "grad_norm": 1.2776498794555664, + "learning_rate": 1.8725042783799202e-06, + "loss": 0.0184, + "step": 6200 + }, + { + "epoch": 2.480175771497054, + "grad_norm": 1.3365709781646729, + "learning_rate": 1.858243011979464e-06, + "loss": 0.0164, + "step": 6210 + }, + { + "epoch": 2.4841705782482775, + "grad_norm": 1.5029335021972656, + "learning_rate": 1.8439817455790076e-06, + "loss": 0.0213, + "step": 6220 + }, + { + "epoch": 2.4881653849995007, + "grad_norm": 1.0199559926986694, + "learning_rate": 1.829720479178551e-06, + "loss": 0.0225, + "step": 6230 + }, + { + "epoch": 2.492160191750724, + "grad_norm": 1.1633626222610474, + "learning_rate": 1.815459212778095e-06, + "loss": 0.0198, + "step": 6240 + }, + { + "epoch": 2.4961549985019476, + "grad_norm": 0.9185020327568054, + "learning_rate": 1.8011979463776386e-06, + "loss": 0.0166, + "step": 6250 + }, + { + "epoch": 2.5001498052531708, + "grad_norm": 1.3016319274902344, + "learning_rate": 1.786936679977182e-06, + "loss": 0.0189, + "step": 6260 + }, + { + "epoch": 2.5041446120043944, + "grad_norm": 2.205761194229126, + "learning_rate": 1.7726754135767257e-06, + "loss": 0.0326, + "step": 6270 + }, + { + "epoch": 2.5081394187556176, + "grad_norm": 0.9863551259040833, + "learning_rate": 1.7584141471762694e-06, + "loss": 0.0243, + "step": 6280 + }, + { + "epoch": 2.5121342255068413, + "grad_norm": 1.3452067375183105, + "learning_rate": 1.7441528807758129e-06, + "loss": 0.0199, + "step": 6290 + }, + { + "epoch": 2.5161290322580645, + "grad_norm": 1.4086759090423584, + "learning_rate": 1.7298916143753565e-06, + "loss": 0.0215, + "step": 6300 + }, + { + "epoch": 2.5201238390092877, + "grad_norm": 1.327675700187683, + "learning_rate": 1.7156303479749004e-06, + "loss": 0.024, + "step": 6310 + }, + { + "epoch": 2.5241186457605114, + "grad_norm": 1.2988176345825195, + "learning_rate": 1.701369081574444e-06, + "loss": 0.0181, + "step": 6320 + }, + { + "epoch": 2.5281134525117346, + "grad_norm": 1.558910846710205, + "learning_rate": 1.6871078151739875e-06, + "loss": 0.0273, + "step": 6330 + }, + { + "epoch": 2.5321082592629582, + "grad_norm": 1.1715282201766968, + "learning_rate": 1.6728465487735312e-06, + "loss": 0.0214, + "step": 6340 + }, + { + "epoch": 2.5361030660141815, + "grad_norm": 1.2082953453063965, + "learning_rate": 1.6585852823730749e-06, + "loss": 0.0163, + "step": 6350 + }, + { + "epoch": 2.540097872765405, + "grad_norm": 1.5872927904129028, + "learning_rate": 1.6443240159726183e-06, + "loss": 0.0199, + "step": 6360 + }, + { + "epoch": 2.5440926795166283, + "grad_norm": 2.50032114982605, + "learning_rate": 1.6300627495721622e-06, + "loss": 0.0284, + "step": 6370 + }, + { + "epoch": 2.5480874862678515, + "grad_norm": 1.4885234832763672, + "learning_rate": 1.615801483171706e-06, + "loss": 0.019, + "step": 6380 + }, + { + "epoch": 2.552082293019075, + "grad_norm": 1.101858377456665, + "learning_rate": 1.6015402167712494e-06, + "loss": 0.0177, + "step": 6390 + }, + { + "epoch": 2.556077099770299, + "grad_norm": 2.104825735092163, + "learning_rate": 1.587278950370793e-06, + "loss": 0.0225, + "step": 6400 + }, + { + "epoch": 2.560071906521522, + "grad_norm": 1.8307040929794312, + "learning_rate": 1.5730176839703367e-06, + "loss": 0.022, + "step": 6410 + }, + { + "epoch": 2.5640667132727453, + "grad_norm": 1.7648673057556152, + "learning_rate": 1.5587564175698802e-06, + "loss": 0.0217, + "step": 6420 + }, + { + "epoch": 2.568061520023969, + "grad_norm": 0.8933414816856384, + "learning_rate": 1.544495151169424e-06, + "loss": 0.0197, + "step": 6430 + }, + { + "epoch": 2.572056326775192, + "grad_norm": 1.0381286144256592, + "learning_rate": 1.5302338847689677e-06, + "loss": 0.0204, + "step": 6440 + }, + { + "epoch": 2.576051133526416, + "grad_norm": 1.289479374885559, + "learning_rate": 1.5159726183685112e-06, + "loss": 0.024, + "step": 6450 + }, + { + "epoch": 2.580045940277639, + "grad_norm": 1.8248828649520874, + "learning_rate": 1.5017113519680548e-06, + "loss": 0.0213, + "step": 6460 + }, + { + "epoch": 2.5840407470288627, + "grad_norm": 1.7288215160369873, + "learning_rate": 1.4874500855675985e-06, + "loss": 0.0205, + "step": 6470 + }, + { + "epoch": 2.588035553780086, + "grad_norm": 1.9080641269683838, + "learning_rate": 1.4731888191671424e-06, + "loss": 0.0221, + "step": 6480 + }, + { + "epoch": 2.592030360531309, + "grad_norm": 1.5588740110397339, + "learning_rate": 1.4589275527666856e-06, + "loss": 0.0194, + "step": 6490 + }, + { + "epoch": 2.5960251672825327, + "grad_norm": 1.553518533706665, + "learning_rate": 1.4446662863662295e-06, + "loss": 0.0247, + "step": 6500 + }, + { + "epoch": 2.600019974033756, + "grad_norm": 1.5387547016143799, + "learning_rate": 1.4304050199657732e-06, + "loss": 0.0162, + "step": 6510 + }, + { + "epoch": 2.6040147807849796, + "grad_norm": 1.93451988697052, + "learning_rate": 1.4161437535653167e-06, + "loss": 0.018, + "step": 6520 + }, + { + "epoch": 2.608009587536203, + "grad_norm": 1.5397741794586182, + "learning_rate": 1.4018824871648603e-06, + "loss": 0.019, + "step": 6530 + }, + { + "epoch": 2.6120043942874265, + "grad_norm": 1.1058956384658813, + "learning_rate": 1.387621220764404e-06, + "loss": 0.0222, + "step": 6540 + }, + { + "epoch": 2.6159992010386497, + "grad_norm": 1.3490480184555054, + "learning_rate": 1.3733599543639475e-06, + "loss": 0.0296, + "step": 6550 + }, + { + "epoch": 2.619994007789873, + "grad_norm": 2.667679786682129, + "learning_rate": 1.3590986879634913e-06, + "loss": 0.022, + "step": 6560 + }, + { + "epoch": 2.6239888145410966, + "grad_norm": 1.6278321743011475, + "learning_rate": 1.344837421563035e-06, + "loss": 0.0209, + "step": 6570 + }, + { + "epoch": 2.62798362129232, + "grad_norm": 2.2078537940979004, + "learning_rate": 1.3305761551625785e-06, + "loss": 0.0194, + "step": 6580 + }, + { + "epoch": 2.6319784280435434, + "grad_norm": 1.520711898803711, + "learning_rate": 1.3163148887621221e-06, + "loss": 0.0228, + "step": 6590 + }, + { + "epoch": 2.6359732347947666, + "grad_norm": 1.4241026639938354, + "learning_rate": 1.3020536223616658e-06, + "loss": 0.0219, + "step": 6600 + }, + { + "epoch": 2.6399680415459903, + "grad_norm": 0.8276427388191223, + "learning_rate": 1.2877923559612093e-06, + "loss": 0.0186, + "step": 6610 + }, + { + "epoch": 2.6439628482972135, + "grad_norm": 1.381286382675171, + "learning_rate": 1.2735310895607532e-06, + "loss": 0.0181, + "step": 6620 + }, + { + "epoch": 2.647957655048437, + "grad_norm": 1.3304340839385986, + "learning_rate": 1.2592698231602968e-06, + "loss": 0.0164, + "step": 6630 + }, + { + "epoch": 2.6519524617996604, + "grad_norm": 0.8185485005378723, + "learning_rate": 1.2450085567598403e-06, + "loss": 0.014, + "step": 6640 + }, + { + "epoch": 2.655947268550884, + "grad_norm": 0.7885934710502625, + "learning_rate": 1.230747290359384e-06, + "loss": 0.0172, + "step": 6650 + }, + { + "epoch": 2.6599420753021072, + "grad_norm": 1.4677386283874512, + "learning_rate": 1.2164860239589276e-06, + "loss": 0.0147, + "step": 6660 + }, + { + "epoch": 2.6639368820533305, + "grad_norm": 0.8628529906272888, + "learning_rate": 1.2022247575584713e-06, + "loss": 0.0154, + "step": 6670 + }, + { + "epoch": 2.667931688804554, + "grad_norm": 1.708916187286377, + "learning_rate": 1.187963491158015e-06, + "loss": 0.0212, + "step": 6680 + }, + { + "epoch": 2.6719264955557773, + "grad_norm": 1.4769736528396606, + "learning_rate": 1.1737022247575586e-06, + "loss": 0.0204, + "step": 6690 + }, + { + "epoch": 2.675921302307001, + "grad_norm": 1.8435742855072021, + "learning_rate": 1.159440958357102e-06, + "loss": 0.0231, + "step": 6700 + }, + { + "epoch": 2.679916109058224, + "grad_norm": 1.2248367071151733, + "learning_rate": 1.145179691956646e-06, + "loss": 0.0207, + "step": 6710 + }, + { + "epoch": 2.683910915809448, + "grad_norm": 1.5767561197280884, + "learning_rate": 1.1309184255561894e-06, + "loss": 0.0246, + "step": 6720 + }, + { + "epoch": 2.687905722560671, + "grad_norm": 1.6061640977859497, + "learning_rate": 1.1166571591557331e-06, + "loss": 0.0221, + "step": 6730 + }, + { + "epoch": 2.6919005293118943, + "grad_norm": 1.0796889066696167, + "learning_rate": 1.1023958927552768e-06, + "loss": 0.02, + "step": 6740 + }, + { + "epoch": 2.695895336063118, + "grad_norm": 1.1353380680084229, + "learning_rate": 1.0881346263548204e-06, + "loss": 0.027, + "step": 6750 + }, + { + "epoch": 2.6998901428143416, + "grad_norm": 1.7669193744659424, + "learning_rate": 1.0738733599543641e-06, + "loss": 0.0218, + "step": 6760 + }, + { + "epoch": 2.703884949565565, + "grad_norm": 1.1086927652359009, + "learning_rate": 1.0596120935539076e-06, + "loss": 0.0201, + "step": 6770 + }, + { + "epoch": 2.707879756316788, + "grad_norm": 1.7246551513671875, + "learning_rate": 1.0453508271534513e-06, + "loss": 0.0207, + "step": 6780 + }, + { + "epoch": 2.7118745630680117, + "grad_norm": 2.2447052001953125, + "learning_rate": 1.031089560752995e-06, + "loss": 0.0227, + "step": 6790 + }, + { + "epoch": 2.715869369819235, + "grad_norm": 1.8295458555221558, + "learning_rate": 1.0168282943525386e-06, + "loss": 0.0173, + "step": 6800 + }, + { + "epoch": 2.7198641765704585, + "grad_norm": 0.8647915720939636, + "learning_rate": 1.0025670279520823e-06, + "loss": 0.0147, + "step": 6810 + }, + { + "epoch": 2.7238589833216817, + "grad_norm": 1.5187458992004395, + "learning_rate": 9.88305761551626e-07, + "loss": 0.021, + "step": 6820 + }, + { + "epoch": 2.7278537900729054, + "grad_norm": 1.4684723615646362, + "learning_rate": 9.740444951511694e-07, + "loss": 0.0165, + "step": 6830 + }, + { + "epoch": 2.7318485968241286, + "grad_norm": 0.7985019683837891, + "learning_rate": 9.597832287507133e-07, + "loss": 0.022, + "step": 6840 + }, + { + "epoch": 2.735843403575352, + "grad_norm": 1.4682860374450684, + "learning_rate": 9.455219623502567e-07, + "loss": 0.0191, + "step": 6850 + }, + { + "epoch": 2.7398382103265755, + "grad_norm": 1.9475513696670532, + "learning_rate": 9.312606959498004e-07, + "loss": 0.0206, + "step": 6860 + }, + { + "epoch": 2.7438330170777987, + "grad_norm": 0.5559619069099426, + "learning_rate": 9.169994295493441e-07, + "loss": 0.0155, + "step": 6870 + }, + { + "epoch": 2.7478278238290224, + "grad_norm": 1.3355700969696045, + "learning_rate": 9.027381631488876e-07, + "loss": 0.017, + "step": 6880 + }, + { + "epoch": 2.7518226305802456, + "grad_norm": 1.5415430068969727, + "learning_rate": 8.884768967484313e-07, + "loss": 0.0157, + "step": 6890 + }, + { + "epoch": 2.755817437331469, + "grad_norm": 1.518141508102417, + "learning_rate": 8.74215630347975e-07, + "loss": 0.0178, + "step": 6900 + }, + { + "epoch": 2.7598122440826924, + "grad_norm": 0.7671738266944885, + "learning_rate": 8.599543639475185e-07, + "loss": 0.0278, + "step": 6910 + }, + { + "epoch": 2.7638070508339156, + "grad_norm": 0.48593270778656006, + "learning_rate": 8.456930975470623e-07, + "loss": 0.0163, + "step": 6920 + }, + { + "epoch": 2.7678018575851393, + "grad_norm": 1.4314422607421875, + "learning_rate": 8.314318311466059e-07, + "loss": 0.0163, + "step": 6930 + }, + { + "epoch": 2.771796664336363, + "grad_norm": 1.5990930795669556, + "learning_rate": 8.171705647461495e-07, + "loss": 0.0169, + "step": 6940 + }, + { + "epoch": 2.775791471087586, + "grad_norm": 1.637716293334961, + "learning_rate": 8.029092983456932e-07, + "loss": 0.0178, + "step": 6950 + }, + { + "epoch": 2.7797862778388094, + "grad_norm": 0.7901973724365234, + "learning_rate": 7.886480319452368e-07, + "loss": 0.0198, + "step": 6960 + }, + { + "epoch": 2.783781084590033, + "grad_norm": 4.324735164642334, + "learning_rate": 7.743867655447804e-07, + "loss": 0.0244, + "step": 6970 + }, + { + "epoch": 2.7877758913412563, + "grad_norm": 1.5631135702133179, + "learning_rate": 7.601254991443241e-07, + "loss": 0.0209, + "step": 6980 + }, + { + "epoch": 2.79177069809248, + "grad_norm": 1.5763413906097412, + "learning_rate": 7.458642327438677e-07, + "loss": 0.0211, + "step": 6990 + }, + { + "epoch": 2.795765504843703, + "grad_norm": 1.613045334815979, + "learning_rate": 7.316029663434114e-07, + "loss": 0.0255, + "step": 7000 + }, + { + "epoch": 2.7997603115949268, + "grad_norm": 1.3384398221969604, + "learning_rate": 7.17341699942955e-07, + "loss": 0.0165, + "step": 7010 + }, + { + "epoch": 2.80375511834615, + "grad_norm": 1.0482898950576782, + "learning_rate": 7.030804335424986e-07, + "loss": 0.0183, + "step": 7020 + }, + { + "epoch": 2.807749925097373, + "grad_norm": 1.8749336004257202, + "learning_rate": 6.888191671420423e-07, + "loss": 0.0151, + "step": 7030 + }, + { + "epoch": 2.811744731848597, + "grad_norm": 0.6791527271270752, + "learning_rate": 6.745579007415858e-07, + "loss": 0.0187, + "step": 7040 + }, + { + "epoch": 2.81573953859982, + "grad_norm": 1.18306303024292, + "learning_rate": 6.602966343411295e-07, + "loss": 0.0278, + "step": 7050 + }, + { + "epoch": 2.8197343453510437, + "grad_norm": 0.9196826815605164, + "learning_rate": 6.460353679406732e-07, + "loss": 0.0178, + "step": 7060 + }, + { + "epoch": 2.823729152102267, + "grad_norm": 1.0838522911071777, + "learning_rate": 6.317741015402168e-07, + "loss": 0.023, + "step": 7070 + }, + { + "epoch": 2.8277239588534906, + "grad_norm": 1.0019241571426392, + "learning_rate": 6.175128351397604e-07, + "loss": 0.0194, + "step": 7080 + }, + { + "epoch": 2.831718765604714, + "grad_norm": 1.3708668947219849, + "learning_rate": 6.032515687393041e-07, + "loss": 0.0177, + "step": 7090 + }, + { + "epoch": 2.835713572355937, + "grad_norm": 1.0062038898468018, + "learning_rate": 5.889903023388478e-07, + "loss": 0.0271, + "step": 7100 + }, + { + "epoch": 2.8397083791071607, + "grad_norm": 1.5444132089614868, + "learning_rate": 5.747290359383914e-07, + "loss": 0.0228, + "step": 7110 + }, + { + "epoch": 2.8437031858583843, + "grad_norm": 1.549835443496704, + "learning_rate": 5.60467769537935e-07, + "loss": 0.0217, + "step": 7120 + }, + { + "epoch": 2.8476979926096075, + "grad_norm": 0.8351393342018127, + "learning_rate": 5.462065031374787e-07, + "loss": 0.0185, + "step": 7130 + }, + { + "epoch": 2.8516927993608308, + "grad_norm": 1.3424099683761597, + "learning_rate": 5.319452367370223e-07, + "loss": 0.0209, + "step": 7140 + }, + { + "epoch": 2.8556876061120544, + "grad_norm": 0.9143849611282349, + "learning_rate": 5.17683970336566e-07, + "loss": 0.0243, + "step": 7150 + }, + { + "epoch": 2.8596824128632776, + "grad_norm": 0.9082064032554626, + "learning_rate": 5.034227039361096e-07, + "loss": 0.022, + "step": 7160 + }, + { + "epoch": 2.8636772196145013, + "grad_norm": 1.0831217765808105, + "learning_rate": 4.891614375356533e-07, + "loss": 0.0152, + "step": 7170 + }, + { + "epoch": 2.8676720263657245, + "grad_norm": 0.8026896119117737, + "learning_rate": 4.7490017113519687e-07, + "loss": 0.0153, + "step": 7180 + }, + { + "epoch": 2.871666833116948, + "grad_norm": 1.4889538288116455, + "learning_rate": 4.6063890473474054e-07, + "loss": 0.016, + "step": 7190 + }, + { + "epoch": 2.8756616398681714, + "grad_norm": 0.5586309432983398, + "learning_rate": 4.463776383342841e-07, + "loss": 0.0119, + "step": 7200 + }, + { + "epoch": 2.8796564466193946, + "grad_norm": 1.3444619178771973, + "learning_rate": 4.321163719338278e-07, + "loss": 0.0167, + "step": 7210 + }, + { + "epoch": 2.8836512533706182, + "grad_norm": 1.4485714435577393, + "learning_rate": 4.178551055333714e-07, + "loss": 0.0204, + "step": 7220 + }, + { + "epoch": 2.8876460601218414, + "grad_norm": 1.0944844484329224, + "learning_rate": 4.0359383913291507e-07, + "loss": 0.0241, + "step": 7230 + }, + { + "epoch": 2.891640866873065, + "grad_norm": 1.5610586404800415, + "learning_rate": 3.8933257273245863e-07, + "loss": 0.0167, + "step": 7240 + }, + { + "epoch": 2.8956356736242883, + "grad_norm": 1.6713281869888306, + "learning_rate": 3.750713063320023e-07, + "loss": 0.0216, + "step": 7250 + }, + { + "epoch": 2.899630480375512, + "grad_norm": 1.646205186843872, + "learning_rate": 3.6081003993154597e-07, + "loss": 0.0195, + "step": 7260 + }, + { + "epoch": 2.903625287126735, + "grad_norm": 0.9843171834945679, + "learning_rate": 3.4654877353108964e-07, + "loss": 0.0254, + "step": 7270 + }, + { + "epoch": 2.9076200938779584, + "grad_norm": 1.0365681648254395, + "learning_rate": 3.322875071306332e-07, + "loss": 0.0186, + "step": 7280 + }, + { + "epoch": 2.911614900629182, + "grad_norm": 1.3323211669921875, + "learning_rate": 3.180262407301769e-07, + "loss": 0.0149, + "step": 7290 + }, + { + "epoch": 2.9156097073804057, + "grad_norm": 1.0428935289382935, + "learning_rate": 3.037649743297205e-07, + "loss": 0.015, + "step": 7300 + }, + { + "epoch": 2.919604514131629, + "grad_norm": 2.072652578353882, + "learning_rate": 2.895037079292641e-07, + "loss": 0.0231, + "step": 7310 + }, + { + "epoch": 2.923599320882852, + "grad_norm": 1.7197315692901611, + "learning_rate": 2.752424415288078e-07, + "loss": 0.015, + "step": 7320 + }, + { + "epoch": 2.927594127634076, + "grad_norm": 1.166914939880371, + "learning_rate": 2.609811751283514e-07, + "loss": 0.0324, + "step": 7330 + }, + { + "epoch": 2.931588934385299, + "grad_norm": 1.1400363445281982, + "learning_rate": 2.467199087278951e-07, + "loss": 0.0176, + "step": 7340 + }, + { + "epoch": 2.9355837411365227, + "grad_norm": 1.5482438802719116, + "learning_rate": 2.324586423274387e-07, + "loss": 0.017, + "step": 7350 + }, + { + "epoch": 2.939578547887746, + "grad_norm": 1.0702805519104004, + "learning_rate": 2.1819737592698234e-07, + "loss": 0.0202, + "step": 7360 + }, + { + "epoch": 2.9435733546389695, + "grad_norm": 1.5657294988632202, + "learning_rate": 2.0393610952652596e-07, + "loss": 0.0263, + "step": 7370 + }, + { + "epoch": 2.9475681613901927, + "grad_norm": 1.4721356630325317, + "learning_rate": 1.8967484312606963e-07, + "loss": 0.0217, + "step": 7380 + }, + { + "epoch": 2.951562968141416, + "grad_norm": 0.7691061496734619, + "learning_rate": 1.7541357672561324e-07, + "loss": 0.0182, + "step": 7390 + }, + { + "epoch": 2.9555577748926396, + "grad_norm": 1.4036718606948853, + "learning_rate": 1.611523103251569e-07, + "loss": 0.0161, + "step": 7400 + }, + { + "epoch": 2.959552581643863, + "grad_norm": 1.071810007095337, + "learning_rate": 1.4689104392470053e-07, + "loss": 0.0153, + "step": 7410 + }, + { + "epoch": 2.9635473883950865, + "grad_norm": 1.9158252477645874, + "learning_rate": 1.3262977752424415e-07, + "loss": 0.0244, + "step": 7420 + }, + { + "epoch": 2.9675421951463097, + "grad_norm": 2.0826191902160645, + "learning_rate": 1.183685111237878e-07, + "loss": 0.0206, + "step": 7430 + }, + { + "epoch": 2.9715370018975333, + "grad_norm": 1.3713316917419434, + "learning_rate": 1.0410724472333144e-07, + "loss": 0.0205, + "step": 7440 + }, + { + "epoch": 2.9755318086487565, + "grad_norm": 1.2220088243484497, + "learning_rate": 8.984597832287507e-08, + "loss": 0.017, + "step": 7450 + }, + { + "epoch": 2.9795266153999798, + "grad_norm": 1.025079369544983, + "learning_rate": 7.558471192241872e-08, + "loss": 0.0208, + "step": 7460 + }, + { + "epoch": 2.9835214221512034, + "grad_norm": 1.1116774082183838, + "learning_rate": 6.132344552196236e-08, + "loss": 0.0224, + "step": 7470 + }, + { + "epoch": 2.987516228902427, + "grad_norm": 1.1613423824310303, + "learning_rate": 4.7062179121506e-08, + "loss": 0.0249, + "step": 7480 + }, + { + "epoch": 2.9915110356536503, + "grad_norm": 1.4108824729919434, + "learning_rate": 3.280091272104963e-08, + "loss": 0.0162, + "step": 7490 + }, + { + "epoch": 2.9955058424048735, + "grad_norm": 1.3798424005508423, + "learning_rate": 1.853964632059327e-08, + "loss": 0.0212, + "step": 7500 + }, + { + "epoch": 2.9955058424048735, + "eval_loss": 0.03275377303361893, + "eval_runtime": 11023.9586, + "eval_samples_per_second": 1.816, + "eval_steps_per_second": 0.227, + "eval_wer": 2.2808728914804304, + "step": 7500 + } + ], + "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": 2.4488802392997888e+20, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +}