| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 1370, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.072992700729927, | |
| "grad_norm": 5.749124526977539, | |
| "learning_rate": 1.4492753623188407e-05, | |
| "loss": 0.9214, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.145985401459854, | |
| "grad_norm": 2.569692611694336, | |
| "learning_rate": 2.8985507246376814e-05, | |
| "loss": 0.506, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.21897810218978103, | |
| "grad_norm": 1.676221489906311, | |
| "learning_rate": 4.347826086956522e-05, | |
| "loss": 0.3337, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.291970802919708, | |
| "grad_norm": 1.1421884298324585, | |
| "learning_rate": 5.797101449275363e-05, | |
| "loss": 0.2416, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.36496350364963503, | |
| "grad_norm": 0.7032585144042969, | |
| "learning_rate": 7.246376811594203e-05, | |
| "loss": 0.1709, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.43795620437956206, | |
| "grad_norm": 0.8790648579597473, | |
| "learning_rate": 8.695652173913044e-05, | |
| "loss": 0.1541, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5109489051094891, | |
| "grad_norm": 0.9456900358200073, | |
| "learning_rate": 9.999985422436231e-05, | |
| "loss": 0.1368, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.583941605839416, | |
| "grad_norm": 0.8083729147911072, | |
| "learning_rate": 9.998236217634196e-05, | |
| "loss": 0.1253, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.656934306569343, | |
| "grad_norm": 0.5535995960235596, | |
| "learning_rate": 9.993572668745786e-05, | |
| "loss": 0.1117, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.7299270072992701, | |
| "grad_norm": 0.7210536599159241, | |
| "learning_rate": 9.985997494967441e-05, | |
| "loss": 0.0905, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8029197080291971, | |
| "grad_norm": 0.5236138105392456, | |
| "learning_rate": 9.975515113189827e-05, | |
| "loss": 0.0917, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.8759124087591241, | |
| "grad_norm": 0.4034636318683624, | |
| "learning_rate": 9.962131635422462e-05, | |
| "loss": 0.0802, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.948905109489051, | |
| "grad_norm": 0.6598319411277771, | |
| "learning_rate": 9.945854865229965e-05, | |
| "loss": 0.0744, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.0218978102189782, | |
| "grad_norm": 0.6162580847740173, | |
| "learning_rate": 9.926694293181986e-05, | |
| "loss": 0.0808, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.094890510948905, | |
| "grad_norm": 0.5971336960792542, | |
| "learning_rate": 9.904661091319503e-05, | |
| "loss": 0.0765, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.167883211678832, | |
| "grad_norm": 0.7087730765342712, | |
| "learning_rate": 9.879768106640687e-05, | |
| "loss": 0.0713, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.2408759124087592, | |
| "grad_norm": 0.6969349980354309, | |
| "learning_rate": 9.852029853610148e-05, | |
| "loss": 0.0665, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.313868613138686, | |
| "grad_norm": 0.6811894178390503, | |
| "learning_rate": 9.821462505695917e-05, | |
| "loss": 0.0631, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.3868613138686132, | |
| "grad_norm": 0.34767240285873413, | |
| "learning_rate": 9.788083885939116e-05, | |
| "loss": 0.064, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.4598540145985401, | |
| "grad_norm": 0.3470080494880676, | |
| "learning_rate": 9.75191345656179e-05, | |
| "loss": 0.0557, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.5328467153284673, | |
| "grad_norm": 0.39414817094802856, | |
| "learning_rate": 9.712972307618981e-05, | |
| "loss": 0.0627, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.6058394160583942, | |
| "grad_norm": 0.5173900723457336, | |
| "learning_rate": 9.671283144701663e-05, | |
| "loss": 0.0573, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.6788321167883211, | |
| "grad_norm": 0.34683912992477417, | |
| "learning_rate": 9.626870275697682e-05, | |
| "loss": 0.0603, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.7518248175182483, | |
| "grad_norm": 0.5500655174255371, | |
| "learning_rate": 9.579759596618454e-05, | |
| "loss": 0.062, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.8248175182481752, | |
| "grad_norm": 0.3647141456604004, | |
| "learning_rate": 9.529978576499652e-05, | |
| "loss": 0.0504, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.897810218978102, | |
| "grad_norm": 0.3856804072856903, | |
| "learning_rate": 9.477556241384724e-05, | |
| "loss": 0.0543, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.9708029197080292, | |
| "grad_norm": 0.35589075088500977, | |
| "learning_rate": 9.422523157400533e-05, | |
| "loss": 0.0496, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.0437956204379564, | |
| "grad_norm": 0.37966451048851013, | |
| "learning_rate": 9.36491141293504e-05, | |
| "loss": 0.0475, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.116788321167883, | |
| "grad_norm": 0.4046926200389862, | |
| "learning_rate": 9.304754599927377e-05, | |
| "loss": 0.0466, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.18978102189781, | |
| "grad_norm": 0.5988451242446899, | |
| "learning_rate": 9.242087794281243e-05, | |
| "loss": 0.0529, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.2627737226277373, | |
| "grad_norm": 0.3989510238170624, | |
| "learning_rate": 9.176947535413046e-05, | |
| "loss": 0.0509, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.335766423357664, | |
| "grad_norm": 0.4888112246990204, | |
| "learning_rate": 9.10937180494669e-05, | |
| "loss": 0.0405, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.408759124087591, | |
| "grad_norm": 0.5207938551902771, | |
| "learning_rate": 9.039400004567469e-05, | |
| "loss": 0.0472, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.4817518248175183, | |
| "grad_norm": 0.41739702224731445, | |
| "learning_rate": 8.967072933047945e-05, | |
| "loss": 0.0523, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.554744525547445, | |
| "grad_norm": 0.5469692349433899, | |
| "learning_rate": 8.892432762459221e-05, | |
| "loss": 0.0466, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.627737226277372, | |
| "grad_norm": 0.44405660033226013, | |
| "learning_rate": 8.815523013581488e-05, | |
| "loss": 0.0475, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.7007299270072993, | |
| "grad_norm": 0.42182251811027527, | |
| "learning_rate": 8.736388530528162e-05, | |
| "loss": 0.0423, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.7737226277372264, | |
| "grad_norm": 0.323812335729599, | |
| "learning_rate": 8.655075454598426e-05, | |
| "loss": 0.0421, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.846715328467153, | |
| "grad_norm": 0.44694793224334717, | |
| "learning_rate": 8.571631197373422e-05, | |
| "loss": 0.0447, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.9197080291970803, | |
| "grad_norm": 0.2432471662759781, | |
| "learning_rate": 8.486104413071755e-05, | |
| "loss": 0.0438, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.9927007299270074, | |
| "grad_norm": 0.42758315801620483, | |
| "learning_rate": 8.398544970180469e-05, | |
| "loss": 0.0466, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.065693430656934, | |
| "grad_norm": 0.3887377679347992, | |
| "learning_rate": 8.309003922377996e-05, | |
| "loss": 0.04, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.1386861313868613, | |
| "grad_norm": 0.33644479513168335, | |
| "learning_rate": 8.217533478766068e-05, | |
| "loss": 0.0455, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.2116788321167884, | |
| "grad_norm": 0.29482313990592957, | |
| "learning_rate": 8.124186973427911e-05, | |
| "loss": 0.0398, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.2846715328467155, | |
| "grad_norm": 0.46486881375312805, | |
| "learning_rate": 8.029018834330506e-05, | |
| "loss": 0.0432, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.3576642335766422, | |
| "grad_norm": 0.4706019163131714, | |
| "learning_rate": 7.932084551589027e-05, | |
| "loss": 0.036, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 3.4306569343065694, | |
| "grad_norm": 0.502895176410675, | |
| "learning_rate": 7.833440645111975e-05, | |
| "loss": 0.0372, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 3.5036496350364965, | |
| "grad_norm": 0.37358635663986206, | |
| "learning_rate": 7.733144631645852e-05, | |
| "loss": 0.0415, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 3.576642335766423, | |
| "grad_norm": 0.38203373551368713, | |
| "learning_rate": 7.631254991238621e-05, | |
| "loss": 0.0362, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 3.6496350364963503, | |
| "grad_norm": 0.38717857003211975, | |
| "learning_rate": 7.527831133141476e-05, | |
| "loss": 0.0416, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.7226277372262775, | |
| "grad_norm": 0.32451218366622925, | |
| "learning_rate": 7.422933361168825e-05, | |
| "loss": 0.0378, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 3.795620437956204, | |
| "grad_norm": 0.3965657651424408, | |
| "learning_rate": 7.316622838536673e-05, | |
| "loss": 0.0374, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 3.8686131386861313, | |
| "grad_norm": 0.5710194110870361, | |
| "learning_rate": 7.208961552199913e-05, | |
| "loss": 0.0372, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 3.9416058394160585, | |
| "grad_norm": 0.3017335832118988, | |
| "learning_rate": 7.100012276709302e-05, | |
| "loss": 0.0411, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 4.014598540145985, | |
| "grad_norm": 0.42834600806236267, | |
| "learning_rate": 6.98983853760924e-05, | |
| "loss": 0.0324, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.087591240875913, | |
| "grad_norm": 0.2624529302120209, | |
| "learning_rate": 6.878504574397626e-05, | |
| "loss": 0.0362, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 4.160583941605839, | |
| "grad_norm": 0.3475762605667114, | |
| "learning_rate": 6.766075303069459e-05, | |
| "loss": 0.0398, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 4.233576642335766, | |
| "grad_norm": 0.5206601023674011, | |
| "learning_rate": 6.65261627826597e-05, | |
| "loss": 0.0367, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 4.306569343065694, | |
| "grad_norm": 0.4349361062049866, | |
| "learning_rate": 6.538193655051381e-05, | |
| "loss": 0.0365, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 4.37956204379562, | |
| "grad_norm": 0.5204529166221619, | |
| "learning_rate": 6.422874150339579e-05, | |
| "loss": 0.0358, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.452554744525547, | |
| "grad_norm": 0.27471956610679626, | |
| "learning_rate": 6.30672500399318e-05, | |
| "loss": 0.0336, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 4.525547445255475, | |
| "grad_norm": 0.3909718990325928, | |
| "learning_rate": 6.189813939617682e-05, | |
| "loss": 0.0363, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 4.598540145985401, | |
| "grad_norm": 0.24090121686458588, | |
| "learning_rate": 6.072209125073561e-05, | |
| "loss": 0.0363, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 4.671532846715328, | |
| "grad_norm": 0.3554544448852539, | |
| "learning_rate": 5.95397913272932e-05, | |
| "loss": 0.0359, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 4.744525547445256, | |
| "grad_norm": 0.31269019842147827, | |
| "learning_rate": 5.8351928994787006e-05, | |
| "loss": 0.0405, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 4.817518248175182, | |
| "grad_norm": 0.40918299555778503, | |
| "learning_rate": 5.7159196865453294e-05, | |
| "loss": 0.0361, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 4.89051094890511, | |
| "grad_norm": 0.522287130355835, | |
| "learning_rate": 5.596229039098271e-05, | |
| "loss": 0.0336, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 4.963503649635037, | |
| "grad_norm": 0.48848339915275574, | |
| "learning_rate": 5.4761907457020077e-05, | |
| "loss": 0.0317, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 5.036496350364963, | |
| "grad_norm": 0.4906352162361145, | |
| "learning_rate": 5.355874797624515e-05, | |
| "loss": 0.0295, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 5.109489051094891, | |
| "grad_norm": 0.24832546710968018, | |
| "learning_rate": 5.235351348027129e-05, | |
| "loss": 0.0299, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 5.182481751824818, | |
| "grad_norm": 0.3912707567214966, | |
| "learning_rate": 5.1146906710600306e-05, | |
| "loss": 0.0306, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 5.255474452554744, | |
| "grad_norm": 0.23428483307361603, | |
| "learning_rate": 4.993963120887183e-05, | |
| "loss": 0.0312, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 5.328467153284672, | |
| "grad_norm": 0.39608365297317505, | |
| "learning_rate": 4.8732390906646097e-05, | |
| "loss": 0.0323, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 5.401459854014599, | |
| "grad_norm": 0.24625132977962494, | |
| "learning_rate": 4.75258897149594e-05, | |
| "loss": 0.0346, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 5.474452554744525, | |
| "grad_norm": 0.4109422564506531, | |
| "learning_rate": 4.632083111389153e-05, | |
| "loss": 0.0305, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 5.547445255474453, | |
| "grad_norm": 0.4575289487838745, | |
| "learning_rate": 4.5117917742384456e-05, | |
| "loss": 0.0335, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 5.62043795620438, | |
| "grad_norm": 0.3231428265571594, | |
| "learning_rate": 4.391785098855156e-05, | |
| "loss": 0.032, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 5.693430656934306, | |
| "grad_norm": 0.22695614397525787, | |
| "learning_rate": 4.272133058071595e-05, | |
| "loss": 0.0341, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 5.766423357664234, | |
| "grad_norm": 0.3635699450969696, | |
| "learning_rate": 4.1529054179416875e-05, | |
| "loss": 0.0303, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 5.839416058394161, | |
| "grad_norm": 0.44931599497795105, | |
| "learning_rate": 4.034171697062157e-05, | |
| "loss": 0.0312, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 5.912408759124087, | |
| "grad_norm": 0.38062986731529236, | |
| "learning_rate": 3.916001126038008e-05, | |
| "loss": 0.0298, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 5.985401459854015, | |
| "grad_norm": 0.44949817657470703, | |
| "learning_rate": 3.7984626071159224e-05, | |
| "loss": 0.0301, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 6.0583941605839415, | |
| "grad_norm": 0.2569969892501831, | |
| "learning_rate": 3.681624674009121e-05, | |
| "loss": 0.03, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 6.131386861313868, | |
| "grad_norm": 0.3157390356063843, | |
| "learning_rate": 3.5655554519370956e-05, | |
| "loss": 0.0303, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 6.204379562043796, | |
| "grad_norm": 0.27934950590133667, | |
| "learning_rate": 3.450322617903543e-05, | |
| "loss": 0.0281, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 6.2773722627737225, | |
| "grad_norm": 0.2728220224380493, | |
| "learning_rate": 3.3359933612356156e-05, | |
| "loss": 0.0287, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 6.350364963503649, | |
| "grad_norm": 0.27907413244247437, | |
| "learning_rate": 3.2226343444075465e-05, | |
| "loss": 0.0268, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 6.423357664233577, | |
| "grad_norm": 0.2804538309574127, | |
| "learning_rate": 3.110311664171458e-05, | |
| "loss": 0.03, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 6.4963503649635035, | |
| "grad_norm": 0.5569368600845337, | |
| "learning_rate": 2.999090813018035e-05, | |
| "loss": 0.028, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 6.569343065693431, | |
| "grad_norm": 0.7178159952163696, | |
| "learning_rate": 2.8890366409895148e-05, | |
| "loss": 0.0325, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 6.642335766423358, | |
| "grad_norm": 0.19943037629127502, | |
| "learning_rate": 2.780213317867292e-05, | |
| "loss": 0.025, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 6.7153284671532845, | |
| "grad_norm": 0.1796846240758896, | |
| "learning_rate": 2.672684295756147e-05, | |
| "loss": 0.0257, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 6.788321167883212, | |
| "grad_norm": 0.23872807621955872, | |
| "learning_rate": 2.566512272086945e-05, | |
| "loss": 0.0304, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 6.861313868613139, | |
| "grad_norm": 0.2793905735015869, | |
| "learning_rate": 2.4617591530593613e-05, | |
| "loss": 0.0245, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 6.934306569343065, | |
| "grad_norm": 0.2569703161716461, | |
| "learning_rate": 2.3584860175459584e-05, | |
| "loss": 0.0269, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 7.007299270072993, | |
| "grad_norm": 0.27794763445854187, | |
| "learning_rate": 2.2567530814786463e-05, | |
| "loss": 0.0287, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 7.08029197080292, | |
| "grad_norm": 0.237870454788208, | |
| "learning_rate": 2.156619662738319e-05, | |
| "loss": 0.0261, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 7.153284671532846, | |
| "grad_norm": 0.3608488142490387, | |
| "learning_rate": 2.0581441465680986e-05, | |
| "loss": 0.0259, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 7.226277372262774, | |
| "grad_norm": 0.3045063316822052, | |
| "learning_rate": 1.961383951530394e-05, | |
| "loss": 0.0286, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 7.299270072992701, | |
| "grad_norm": 0.2131444364786148, | |
| "learning_rate": 1.866395496027602e-05, | |
| "loss": 0.0254, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 7.372262773722627, | |
| "grad_norm": 0.25216737389564514, | |
| "learning_rate": 1.7732341654059785e-05, | |
| "loss": 0.0222, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 7.445255474452555, | |
| "grad_norm": 0.2963661849498749, | |
| "learning_rate": 1.6819542796618487e-05, | |
| "loss": 0.0292, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 7.518248175182482, | |
| "grad_norm": 0.4280974864959717, | |
| "learning_rate": 1.592609061769004e-05, | |
| "loss": 0.0263, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 7.591240875912408, | |
| "grad_norm": 0.23306907713413239, | |
| "learning_rate": 1.5052506066457461e-05, | |
| "loss": 0.0211, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 7.664233576642336, | |
| "grad_norm": 0.2898852825164795, | |
| "learning_rate": 1.4199298507796698e-05, | |
| "loss": 0.0209, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 7.737226277372263, | |
| "grad_norm": 0.22745266556739807, | |
| "learning_rate": 1.3366965425278899e-05, | |
| "loss": 0.0218, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 7.81021897810219, | |
| "grad_norm": 0.244761660695076, | |
| "learning_rate": 1.2555992131100457e-05, | |
| "loss": 0.0213, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 7.883211678832117, | |
| "grad_norm": 0.3487051725387573, | |
| "learning_rate": 1.1766851483109858e-05, | |
| "loss": 0.0251, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 7.956204379562044, | |
| "grad_norm": 0.38351985812187195, | |
| "learning_rate": 1.1000003609096337e-05, | |
| "loss": 0.0233, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 8.02919708029197, | |
| "grad_norm": 0.2925453782081604, | |
| "learning_rate": 1.0255895638501045e-05, | |
| "loss": 0.0258, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 8.102189781021897, | |
| "grad_norm": 0.2964305281639099, | |
| "learning_rate": 9.534961441707307e-06, | |
| "loss": 0.0193, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 8.175182481751825, | |
| "grad_norm": 0.4270758628845215, | |
| "learning_rate": 8.837621377061877e-06, | |
| "loss": 0.0223, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 8.248175182481752, | |
| "grad_norm": 0.22104066610336304, | |
| "learning_rate": 8.16428204577468e-06, | |
| "loss": 0.0217, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 8.321167883211679, | |
| "grad_norm": 0.24666981399059296, | |
| "learning_rate": 7.515336054840022e-06, | |
| "loss": 0.0254, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 8.394160583941606, | |
| "grad_norm": 0.16614589095115662, | |
| "learning_rate": 6.8911617881174725e-06, | |
| "loss": 0.02, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 8.467153284671532, | |
| "grad_norm": 0.2146356999874115, | |
| "learning_rate": 6.292123185705867e-06, | |
| "loss": 0.024, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 8.540145985401459, | |
| "grad_norm": 0.1602909117937088, | |
| "learning_rate": 5.718569531739154e-06, | |
| "loss": 0.0267, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 8.613138686131387, | |
| "grad_norm": 0.4400380849838257, | |
| "learning_rate": 5.170835250727663e-06, | |
| "loss": 0.0231, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 8.686131386861314, | |
| "grad_norm": 0.18121108412742615, | |
| "learning_rate": 4.6492397125637525e-06, | |
| "loss": 0.0192, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 8.75912408759124, | |
| "grad_norm": 0.26767614483833313, | |
| "learning_rate": 4.154087046305322e-06, | |
| "loss": 0.0249, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 8.832116788321168, | |
| "grad_norm": 0.2408941090106964, | |
| "learning_rate": 3.6856659628459912e-06, | |
| "loss": 0.0214, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 8.905109489051094, | |
| "grad_norm": 0.2535434663295746, | |
| "learning_rate": 3.244249586575038e-06, | |
| "loss": 0.0203, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 8.978102189781023, | |
| "grad_norm": 0.6523481011390686, | |
| "learning_rate": 2.830095296125612e-06, | |
| "loss": 0.0222, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 9.05109489051095, | |
| "grad_norm": 0.16764762997627258, | |
| "learning_rate": 2.4434445743037713e-06, | |
| "loss": 0.0225, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 9.124087591240876, | |
| "grad_norm": 0.35821276903152466, | |
| "learning_rate": 2.0845228672860538e-06, | |
| "loss": 0.0206, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 9.197080291970803, | |
| "grad_norm": 0.38857728242874146, | |
| "learning_rate": 1.7535394531675187e-06, | |
| "loss": 0.0222, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 9.27007299270073, | |
| "grad_norm": 0.2351900190114975, | |
| "learning_rate": 1.4506873199370497e-06, | |
| "loss": 0.021, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 9.343065693430656, | |
| "grad_norm": 0.33225876092910767, | |
| "learning_rate": 1.1761430529509899e-06, | |
| "loss": 0.0258, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 9.416058394160585, | |
| "grad_norm": 0.2640361785888672, | |
| "learning_rate": 9.300667319706857e-07, | |
| "loss": 0.0202, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 9.489051094890511, | |
| "grad_norm": 0.15353083610534668, | |
| "learning_rate": 7.126018378241062e-07, | |
| "loss": 0.0215, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 9.562043795620438, | |
| "grad_norm": 0.33280253410339355, | |
| "learning_rate": 5.238751687458021e-07, | |
| "loss": 0.0217, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 9.635036496350365, | |
| "grad_norm": 0.18690501153469086, | |
| "learning_rate": 3.639967664440802e-07, | |
| "loss": 0.0194, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 9.708029197080291, | |
| "grad_norm": 0.6559237837791443, | |
| "learning_rate": 2.3305985193852742e-07, | |
| "loss": 0.0242, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 9.78102189781022, | |
| "grad_norm": 0.3035587966442108, | |
| "learning_rate": 1.3114077120517376e-07, | |
| "loss": 0.0271, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 9.854014598540147, | |
| "grad_norm": 0.30467328429222107, | |
| "learning_rate": 5.8298950661112017e-08, | |
| "loss": 0.0187, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 9.927007299270073, | |
| "grad_norm": 0.22451713681221008, | |
| "learning_rate": 1.4576862514487089e-08, | |
| "loss": 0.022, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.33209139108657837, | |
| "learning_rate": 0.0, | |
| "loss": 0.0227, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 1370, | |
| "total_flos": 1.9634851190645568e+17, | |
| "train_loss": 0.05370959477485531, | |
| "train_runtime": 1800.9203, | |
| "train_samples_per_second": 48.631, | |
| "train_steps_per_second": 0.761 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1370, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 10000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.9634851190645568e+17, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |