| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 1294, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.015455950540958269, | |
| "grad_norm": 0.357046514749527, | |
| "learning_rate": 0.0003846153846153846, | |
| "loss": 1.5383, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.030911901081916538, | |
| "grad_norm": 0.11198900640010834, | |
| "learning_rate": 0.0007692307692307692, | |
| "loss": 1.1091, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04636785162287481, | |
| "grad_norm": 0.056582603603601456, | |
| "learning_rate": 0.001153846153846154, | |
| "loss": 0.7939, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.061823802163833076, | |
| "grad_norm": 0.0210476852953434, | |
| "learning_rate": 0.0015384615384615385, | |
| "loss": 0.6165, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07727975270479134, | |
| "grad_norm": 0.012083015404641628, | |
| "learning_rate": 0.0019230769230769232, | |
| "loss": 0.5624, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.09273570324574962, | |
| "grad_norm": 0.008505144156515598, | |
| "learning_rate": 0.002307692307692308, | |
| "loss": 0.5269, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.10819165378670788, | |
| "grad_norm": 0.0063809980638325214, | |
| "learning_rate": 0.0026923076923076926, | |
| "loss": 0.5059, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.12364760432766615, | |
| "grad_norm": 0.005832094699144363, | |
| "learning_rate": 0.0029950657894736842, | |
| "loss": 0.5202, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1391035548686244, | |
| "grad_norm": 0.004662094172090292, | |
| "learning_rate": 0.0029703947368421055, | |
| "loss": 0.5087, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.1545595054095827, | |
| "grad_norm": 0.004813206382095814, | |
| "learning_rate": 0.0029457236842105267, | |
| "loss": 0.4979, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.17001545595054096, | |
| "grad_norm": 0.003981301095336676, | |
| "learning_rate": 0.0029210526315789475, | |
| "loss": 0.4833, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.18547140649149924, | |
| "grad_norm": 0.0037942214403301477, | |
| "learning_rate": 0.0028963815789473687, | |
| "loss": 0.4842, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2009273570324575, | |
| "grad_norm": 0.0041742450557649136, | |
| "learning_rate": 0.0028717105263157895, | |
| "loss": 0.4818, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.21638330757341576, | |
| "grad_norm": 0.005099099595099688, | |
| "learning_rate": 0.0028470394736842108, | |
| "loss": 0.4809, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.23183925811437403, | |
| "grad_norm": 0.0031047100201249123, | |
| "learning_rate": 0.0028223684210526316, | |
| "loss": 0.5016, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2472952086553323, | |
| "grad_norm": 0.0036040199920535088, | |
| "learning_rate": 0.002797697368421053, | |
| "loss": 0.4775, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.26275115919629055, | |
| "grad_norm": 0.0033861789852380753, | |
| "learning_rate": 0.0027730263157894736, | |
| "loss": 0.4784, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2782071097372488, | |
| "grad_norm": 0.003118926426395774, | |
| "learning_rate": 0.002748355263157895, | |
| "loss": 0.4962, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2936630602782071, | |
| "grad_norm": 0.0035265563055872917, | |
| "learning_rate": 0.002723684210526316, | |
| "loss": 0.4829, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.3091190108191654, | |
| "grad_norm": 0.0035475995391607285, | |
| "learning_rate": 0.002699013157894737, | |
| "loss": 0.485, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.32457496136012365, | |
| "grad_norm": 0.0030264686793088913, | |
| "learning_rate": 0.002674342105263158, | |
| "loss": 0.4681, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3400309119010819, | |
| "grad_norm": 0.0033854299690574408, | |
| "learning_rate": 0.002649671052631579, | |
| "loss": 0.4805, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3554868624420402, | |
| "grad_norm": 0.0029569112230092287, | |
| "learning_rate": 0.002625, | |
| "loss": 0.4688, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.37094281298299847, | |
| "grad_norm": 0.0032272525131702423, | |
| "learning_rate": 0.002600328947368421, | |
| "loss": 0.4752, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.38639876352395675, | |
| "grad_norm": 0.003502602456137538, | |
| "learning_rate": 0.002575657894736842, | |
| "loss": 0.4699, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.401854714064915, | |
| "grad_norm": 0.0031522298231720924, | |
| "learning_rate": 0.002550986842105263, | |
| "loss": 0.4756, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.41731066460587324, | |
| "grad_norm": 0.003098264569416642, | |
| "learning_rate": 0.0025263157894736842, | |
| "loss": 0.4574, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4327666151468315, | |
| "grad_norm": 0.0025676521472632885, | |
| "learning_rate": 0.0025016447368421055, | |
| "loss": 0.4779, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.4482225656877898, | |
| "grad_norm": 0.0034302272833883762, | |
| "learning_rate": 0.0024769736842105263, | |
| "loss": 0.4729, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.46367851622874806, | |
| "grad_norm": 0.003159865504130721, | |
| "learning_rate": 0.0024523026315789475, | |
| "loss": 0.4715, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.47913446676970634, | |
| "grad_norm": 0.003168923780322075, | |
| "learning_rate": 0.0024276315789473683, | |
| "loss": 0.4764, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.4945904173106646, | |
| "grad_norm": 0.0034859515726566315, | |
| "learning_rate": 0.0024029605263157896, | |
| "loss": 0.4652, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5100463678516228, | |
| "grad_norm": 0.003067239187657833, | |
| "learning_rate": 0.0023782894736842104, | |
| "loss": 0.4648, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5255023183925811, | |
| "grad_norm": 0.0032223982270807028, | |
| "learning_rate": 0.0023536184210526316, | |
| "loss": 0.4725, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.5409582689335394, | |
| "grad_norm": 0.0027090355288237333, | |
| "learning_rate": 0.0023289473684210524, | |
| "loss": 0.4704, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5564142194744977, | |
| "grad_norm": 0.003484300570562482, | |
| "learning_rate": 0.0023042763157894736, | |
| "loss": 0.4616, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5718701700154559, | |
| "grad_norm": 0.003339330432936549, | |
| "learning_rate": 0.0022796052631578944, | |
| "loss": 0.4665, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5873261205564142, | |
| "grad_norm": 0.0029797593597322702, | |
| "learning_rate": 0.002254934210526316, | |
| "loss": 0.4573, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.6027820710973725, | |
| "grad_norm": 0.0030033981893211603, | |
| "learning_rate": 0.002230263157894737, | |
| "loss": 0.4618, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.6182380216383307, | |
| "grad_norm": 0.005113155115395784, | |
| "learning_rate": 0.002205592105263158, | |
| "loss": 0.4589, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.633693972179289, | |
| "grad_norm": 0.002975397277623415, | |
| "learning_rate": 0.002180921052631579, | |
| "loss": 0.4765, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6491499227202473, | |
| "grad_norm": 0.004753004759550095, | |
| "learning_rate": 0.00215625, | |
| "loss": 0.4631, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6646058732612056, | |
| "grad_norm": 0.003564928425475955, | |
| "learning_rate": 0.002131578947368421, | |
| "loss": 0.4488, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6800618238021638, | |
| "grad_norm": 0.0032665496692061424, | |
| "learning_rate": 0.0021069078947368422, | |
| "loss": 0.457, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6955177743431221, | |
| "grad_norm": 0.0030079709831625223, | |
| "learning_rate": 0.002082236842105263, | |
| "loss": 0.4667, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.7109737248840804, | |
| "grad_norm": 0.0025733078364282846, | |
| "learning_rate": 0.0020575657894736843, | |
| "loss": 0.4667, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.7264296754250387, | |
| "grad_norm": 0.00270587345585227, | |
| "learning_rate": 0.0020328947368421055, | |
| "loss": 0.4679, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7418856259659969, | |
| "grad_norm": 0.00273908581584692, | |
| "learning_rate": 0.0020082236842105263, | |
| "loss": 0.4694, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7573415765069552, | |
| "grad_norm": 0.002720112446695566, | |
| "learning_rate": 0.0019835526315789475, | |
| "loss": 0.4513, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7727975270479135, | |
| "grad_norm": 0.0028910296969115734, | |
| "learning_rate": 0.0019588815789473683, | |
| "loss": 0.4592, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7882534775888718, | |
| "grad_norm": 0.003034258494153619, | |
| "learning_rate": 0.0019342105263157896, | |
| "loss": 0.4611, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.80370942812983, | |
| "grad_norm": 0.0030727661214768887, | |
| "learning_rate": 0.0019095394736842106, | |
| "loss": 0.463, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.8191653786707882, | |
| "grad_norm": 0.0026562565471976995, | |
| "learning_rate": 0.0018848684210526316, | |
| "loss": 0.469, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.8346213292117465, | |
| "grad_norm": 0.002964869374409318, | |
| "learning_rate": 0.0018601973684210526, | |
| "loss": 0.462, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.8500772797527048, | |
| "grad_norm": 0.0029133500065654516, | |
| "learning_rate": 0.0018355263157894736, | |
| "loss": 0.4678, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.865533230293663, | |
| "grad_norm": 0.002890991512686014, | |
| "learning_rate": 0.0018108552631578947, | |
| "loss": 0.4646, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.8809891808346213, | |
| "grad_norm": 0.0025943962391465902, | |
| "learning_rate": 0.001786184210526316, | |
| "loss": 0.4645, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.8964451313755796, | |
| "grad_norm": 0.0027450553607195616, | |
| "learning_rate": 0.001761513157894737, | |
| "loss": 0.4691, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.9119010819165378, | |
| "grad_norm": 0.0027814917266368866, | |
| "learning_rate": 0.001736842105263158, | |
| "loss": 0.4549, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.9273570324574961, | |
| "grad_norm": 0.0030699821654707193, | |
| "learning_rate": 0.001712171052631579, | |
| "loss": 0.4694, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9428129829984544, | |
| "grad_norm": 0.002675461582839489, | |
| "learning_rate": 0.0016875, | |
| "loss": 0.4593, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.9582689335394127, | |
| "grad_norm": 0.0026055267080664635, | |
| "learning_rate": 0.001662828947368421, | |
| "loss": 0.4681, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.973724884080371, | |
| "grad_norm": 0.002623435575515032, | |
| "learning_rate": 0.001638157894736842, | |
| "loss": 0.46, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.9891808346213292, | |
| "grad_norm": 0.0023981425911188126, | |
| "learning_rate": 0.001613486842105263, | |
| "loss": 0.4587, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.9165626764297485, | |
| "eval_runtime": 1353.6255, | |
| "eval_samples_per_second": 7.648, | |
| "eval_steps_per_second": 0.478, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 1.0046367851622875, | |
| "grad_norm": 0.002703867619857192, | |
| "learning_rate": 0.001588815789473684, | |
| "loss": 0.4565, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.0200927357032457, | |
| "grad_norm": 0.002686214866116643, | |
| "learning_rate": 0.0015641447368421055, | |
| "loss": 0.4656, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.035548686244204, | |
| "grad_norm": 0.002944634296000004, | |
| "learning_rate": 0.0015394736842105265, | |
| "loss": 0.4609, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.0510046367851622, | |
| "grad_norm": 0.002919598249718547, | |
| "learning_rate": 0.0015148026315789475, | |
| "loss": 0.462, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.0664605873261206, | |
| "grad_norm": 0.0027975935954600573, | |
| "learning_rate": 0.0014901315789473686, | |
| "loss": 0.455, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.0819165378670788, | |
| "grad_norm": 0.002747561549767852, | |
| "learning_rate": 0.0014654605263157896, | |
| "loss": 0.4556, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0973724884080371, | |
| "grad_norm": 0.0029210918582975864, | |
| "learning_rate": 0.0014407894736842106, | |
| "loss": 0.4486, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.1128284389489953, | |
| "grad_norm": 0.0026143298018723726, | |
| "learning_rate": 0.0014161184210526316, | |
| "loss": 0.4593, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.1282843894899537, | |
| "grad_norm": 0.002351797418668866, | |
| "learning_rate": 0.0013914473684210526, | |
| "loss": 0.4419, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.1437403400309119, | |
| "grad_norm": 0.0029544688295572996, | |
| "learning_rate": 0.0013667763157894737, | |
| "loss": 0.4535, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.1591962905718702, | |
| "grad_norm": 0.0025962782092392445, | |
| "learning_rate": 0.0013421052631578947, | |
| "loss": 0.4454, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.1746522411128284, | |
| "grad_norm": 0.0026020314544439316, | |
| "learning_rate": 0.0013174342105263157, | |
| "loss": 0.4569, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.1901081916537868, | |
| "grad_norm": 0.00269780564121902, | |
| "learning_rate": 0.0012927631578947367, | |
| "loss": 0.4657, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.205564142194745, | |
| "grad_norm": 0.0027555320411920547, | |
| "learning_rate": 0.001268092105263158, | |
| "loss": 0.4546, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.2210200927357033, | |
| "grad_norm": 0.002603672444820404, | |
| "learning_rate": 0.001243421052631579, | |
| "loss": 0.4547, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.2364760432766615, | |
| "grad_norm": 0.002504123141989112, | |
| "learning_rate": 0.00121875, | |
| "loss": 0.4627, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2519319938176197, | |
| "grad_norm": 0.003078003181144595, | |
| "learning_rate": 0.001194078947368421, | |
| "loss": 0.4504, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.267387944358578, | |
| "grad_norm": 0.0032601715065538883, | |
| "learning_rate": 0.001169407894736842, | |
| "loss": 0.4508, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.2828438948995364, | |
| "grad_norm": 0.003114907769486308, | |
| "learning_rate": 0.0011447368421052633, | |
| "loss": 0.4448, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.2982998454404946, | |
| "grad_norm": 0.002611513016745448, | |
| "learning_rate": 0.0011200657894736843, | |
| "loss": 0.4611, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.3137557959814528, | |
| "grad_norm": 0.0028356926050037146, | |
| "learning_rate": 0.0010953947368421053, | |
| "loss": 0.4543, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.3292117465224111, | |
| "grad_norm": 0.0028607589192688465, | |
| "learning_rate": 0.0010707236842105263, | |
| "loss": 0.4729, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.3446676970633695, | |
| "grad_norm": 0.00289659365080297, | |
| "learning_rate": 0.0010460526315789474, | |
| "loss": 0.4567, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.3601236476043277, | |
| "grad_norm": 0.002599680330604315, | |
| "learning_rate": 0.0010213815789473686, | |
| "loss": 0.456, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.3755795981452859, | |
| "grad_norm": 0.0026973742060363293, | |
| "learning_rate": 0.0009967105263157896, | |
| "loss": 0.4552, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.3910355486862442, | |
| "grad_norm": 0.0025935762096196413, | |
| "learning_rate": 0.0009720394736842105, | |
| "loss": 0.4589, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.4064914992272024, | |
| "grad_norm": 0.0026942496187984943, | |
| "learning_rate": 0.0009473684210526315, | |
| "loss": 0.455, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.4219474497681608, | |
| "grad_norm": 0.0027507098857313395, | |
| "learning_rate": 0.0009226973684210528, | |
| "loss": 0.4627, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.437403400309119, | |
| "grad_norm": 0.0029431770090013742, | |
| "learning_rate": 0.0008980263157894738, | |
| "loss": 0.4557, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.4528593508500773, | |
| "grad_norm": 0.002638396341353655, | |
| "learning_rate": 0.0008733552631578948, | |
| "loss": 0.4546, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.4683153013910355, | |
| "grad_norm": 0.0026191219221800566, | |
| "learning_rate": 0.0008486842105263158, | |
| "loss": 0.4557, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.4837712519319939, | |
| "grad_norm": 0.002677230630069971, | |
| "learning_rate": 0.0008240131578947368, | |
| "loss": 0.4632, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.499227202472952, | |
| "grad_norm": 0.00285891885869205, | |
| "learning_rate": 0.000799342105263158, | |
| "loss": 0.4581, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.5146831530139102, | |
| "grad_norm": 0.0026907038409262896, | |
| "learning_rate": 0.000774671052631579, | |
| "loss": 0.4509, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.5301391035548686, | |
| "grad_norm": 0.0024703822564333677, | |
| "learning_rate": 0.00075, | |
| "loss": 0.4481, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.545595054095827, | |
| "grad_norm": 0.002853216603398323, | |
| "learning_rate": 0.000725328947368421, | |
| "loss": 0.4438, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5610510046367851, | |
| "grad_norm": 0.002754925051704049, | |
| "learning_rate": 0.0007006578947368422, | |
| "loss": 0.4656, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.5765069551777433, | |
| "grad_norm": 0.002468267921358347, | |
| "learning_rate": 0.0006759868421052632, | |
| "loss": 0.4515, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.5919629057187017, | |
| "grad_norm": 0.002553540049120784, | |
| "learning_rate": 0.0006513157894736842, | |
| "loss": 0.4506, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.60741885625966, | |
| "grad_norm": 0.0030136853456497192, | |
| "learning_rate": 0.0006266447368421052, | |
| "loss": 0.4552, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.6228748068006182, | |
| "grad_norm": 0.002497268607839942, | |
| "learning_rate": 0.0006019736842105263, | |
| "loss": 0.4506, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.6383307573415764, | |
| "grad_norm": 0.002612602198496461, | |
| "learning_rate": 0.0005773026315789474, | |
| "loss": 0.4519, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.6537867078825348, | |
| "grad_norm": 0.0023622452281415462, | |
| "learning_rate": 0.0005526315789473684, | |
| "loss": 0.4446, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.6692426584234932, | |
| "grad_norm": 0.0023850842844694853, | |
| "learning_rate": 0.0005279605263157895, | |
| "loss": 0.4533, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.6846986089644513, | |
| "grad_norm": 0.002770961495116353, | |
| "learning_rate": 0.0005032894736842105, | |
| "loss": 0.4523, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.7001545595054095, | |
| "grad_norm": 0.002901398576796055, | |
| "learning_rate": 0.0004786184210526316, | |
| "loss": 0.4472, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.7156105100463679, | |
| "grad_norm": 0.002840624190866947, | |
| "learning_rate": 0.0004539473684210526, | |
| "loss": 0.4496, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.7310664605873263, | |
| "grad_norm": 0.002452421234920621, | |
| "learning_rate": 0.00042927631578947365, | |
| "loss": 0.4553, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.7465224111282844, | |
| "grad_norm": 0.0025857773143798113, | |
| "learning_rate": 0.0004046052631578948, | |
| "loss": 0.4504, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.7619783616692426, | |
| "grad_norm": 0.002501491457223892, | |
| "learning_rate": 0.0003799342105263158, | |
| "loss": 0.4479, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.7774343122102008, | |
| "grad_norm": 0.002639561193063855, | |
| "learning_rate": 0.0003552631578947368, | |
| "loss": 0.4627, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.7928902627511591, | |
| "grad_norm": 0.002627423033118248, | |
| "learning_rate": 0.0003305921052631579, | |
| "loss": 0.4496, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.8083462132921175, | |
| "grad_norm": 0.0027368138544261456, | |
| "learning_rate": 0.00030592105263157896, | |
| "loss": 0.4567, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.8238021638330757, | |
| "grad_norm": 0.0022819822188466787, | |
| "learning_rate": 0.00028125000000000003, | |
| "loss": 0.4546, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.8392581143740339, | |
| "grad_norm": 0.0024920173455029726, | |
| "learning_rate": 0.00025657894736842105, | |
| "loss": 0.4571, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.8547140649149922, | |
| "grad_norm": 0.002788729965686798, | |
| "learning_rate": 0.00023190789473684213, | |
| "loss": 0.4597, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.8701700154559506, | |
| "grad_norm": 0.00281331455335021, | |
| "learning_rate": 0.00020723684210526317, | |
| "loss": 0.4561, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.8856259659969088, | |
| "grad_norm": 0.0028766875620931387, | |
| "learning_rate": 0.00018256578947368422, | |
| "loss": 0.4626, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.901081916537867, | |
| "grad_norm": 0.0027803743723779917, | |
| "learning_rate": 0.00015789473684210527, | |
| "loss": 0.4441, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.9165378670788253, | |
| "grad_norm": 0.0023476784117519855, | |
| "learning_rate": 0.0001332236842105263, | |
| "loss": 0.4529, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.9319938176197837, | |
| "grad_norm": 0.003161880187690258, | |
| "learning_rate": 0.00010855263157894737, | |
| "loss": 0.4489, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.947449768160742, | |
| "grad_norm": 0.0026641101576387882, | |
| "learning_rate": 8.388157894736842e-05, | |
| "loss": 0.4608, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.9629057187017, | |
| "grad_norm": 0.002320705447345972, | |
| "learning_rate": 5.921052631578947e-05, | |
| "loss": 0.4608, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.9783616692426584, | |
| "grad_norm": 0.0025752289220690727, | |
| "learning_rate": 3.4539473684210524e-05, | |
| "loss": 0.4466, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.9938176197836168, | |
| "grad_norm": 0.0026244991458952427, | |
| "learning_rate": 9.868421052631579e-06, | |
| "loss": 0.4555, | |
| "step": 1290 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1294, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.3257051724162499e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |