| { |
| "best_global_step": 7028, |
| "best_metric": 0.767158567905426, |
| "best_model_checkpoint": "./mcqa_qwen3_letter/checkpoint-7028", |
| "epoch": 0.9998932953939178, |
| "eval_steps": 500, |
| "global_step": 7028, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.007113640405477503, |
| "grad_norm": 20.374778747558594, |
| "learning_rate": 1.337126600284495e-06, |
| "loss": 2.342, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.014227280810955006, |
| "grad_norm": 23.720001220703125, |
| "learning_rate": 2.7596017069701283e-06, |
| "loss": 1.6092, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.02134092121643251, |
| "grad_norm": 14.003034591674805, |
| "learning_rate": 4.182076813655762e-06, |
| "loss": 1.4564, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.028454561621910013, |
| "grad_norm": 11.145376205444336, |
| "learning_rate": 5.604551920341395e-06, |
| "loss": 1.481, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.035568202027387516, |
| "grad_norm": 14.064179420471191, |
| "learning_rate": 7.027027027027028e-06, |
| "loss": 1.4483, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.04268184243286502, |
| "grad_norm": 14.6947021484375, |
| "learning_rate": 8.44950213371266e-06, |
| "loss": 1.4355, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.04979548283834252, |
| "grad_norm": 11.520552635192871, |
| "learning_rate": 9.871977240398294e-06, |
| "loss": 1.4314, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.056909123243820026, |
| "grad_norm": 11.234253883361816, |
| "learning_rate": 1.1294452347083926e-05, |
| "loss": 1.3494, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.06402276364929753, |
| "grad_norm": 9.99629020690918, |
| "learning_rate": 1.271692745376956e-05, |
| "loss": 1.3342, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.07113640405477503, |
| "grad_norm": 8.978431701660156, |
| "learning_rate": 1.4139402560455193e-05, |
| "loss": 1.2951, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.07825004446025254, |
| "grad_norm": 9.42021369934082, |
| "learning_rate": 1.5561877667140826e-05, |
| "loss": 1.3058, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.08536368486573004, |
| "grad_norm": 9.690632820129395, |
| "learning_rate": 1.698435277382646e-05, |
| "loss": 1.2561, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.09247732527120754, |
| "grad_norm": 10.151556015014648, |
| "learning_rate": 1.8406827880512092e-05, |
| "loss": 1.2415, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.09959096567668504, |
| "grad_norm": 8.446418762207031, |
| "learning_rate": 1.9829302987197725e-05, |
| "loss": 1.3303, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.10670460608216255, |
| "grad_norm": 9.451847076416016, |
| "learning_rate": 1.9860869565217395e-05, |
| "loss": 1.2735, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.11381824648764005, |
| "grad_norm": 9.343132019042969, |
| "learning_rate": 1.9702766798418973e-05, |
| "loss": 1.3471, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.12093188689311755, |
| "grad_norm": 9.069653511047363, |
| "learning_rate": 1.9544664031620554e-05, |
| "loss": 1.2535, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.12804552729859506, |
| "grad_norm": 5.554582595825195, |
| "learning_rate": 1.9386561264822135e-05, |
| "loss": 1.2887, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.13515916770407255, |
| "grad_norm": 8.713581085205078, |
| "learning_rate": 1.9228458498023717e-05, |
| "loss": 1.1681, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.14227280810955006, |
| "grad_norm": 8.742165565490723, |
| "learning_rate": 1.9070355731225298e-05, |
| "loss": 1.2271, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.14938644851502755, |
| "grad_norm": 8.755561828613281, |
| "learning_rate": 1.891225296442688e-05, |
| "loss": 1.2557, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.15650008892050507, |
| "grad_norm": 7.881649017333984, |
| "learning_rate": 1.875415019762846e-05, |
| "loss": 1.1999, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.16361372932598256, |
| "grad_norm": 9.955330848693848, |
| "learning_rate": 1.8596047430830042e-05, |
| "loss": 1.1375, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.17072736973146008, |
| "grad_norm": 9.453360557556152, |
| "learning_rate": 1.843794466403162e-05, |
| "loss": 1.2173, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.17784101013693757, |
| "grad_norm": 6.806760787963867, |
| "learning_rate": 1.8279841897233205e-05, |
| "loss": 1.2116, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.18495465054241508, |
| "grad_norm": 13.604923248291016, |
| "learning_rate": 1.8121739130434783e-05, |
| "loss": 1.0788, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.19206829094789257, |
| "grad_norm": 10.025015830993652, |
| "learning_rate": 1.7963636363636364e-05, |
| "loss": 1.149, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.1991819313533701, |
| "grad_norm": 9.246174812316895, |
| "learning_rate": 1.7805533596837945e-05, |
| "loss": 1.1171, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.20629557175884758, |
| "grad_norm": 9.556463241577148, |
| "learning_rate": 1.7647430830039527e-05, |
| "loss": 1.1744, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.2134092121643251, |
| "grad_norm": 13.477895736694336, |
| "learning_rate": 1.7489328063241108e-05, |
| "loss": 1.0723, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.22052285256980259, |
| "grad_norm": 7.583223342895508, |
| "learning_rate": 1.733122529644269e-05, |
| "loss": 1.1004, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.2276364929752801, |
| "grad_norm": 9.056886672973633, |
| "learning_rate": 1.717312252964427e-05, |
| "loss": 1.1413, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.2347501333807576, |
| "grad_norm": 8.818636894226074, |
| "learning_rate": 1.701501976284585e-05, |
| "loss": 1.0248, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.2418637737862351, |
| "grad_norm": 11.717618942260742, |
| "learning_rate": 1.6856916996047434e-05, |
| "loss": 1.0386, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.2489774141917126, |
| "grad_norm": 7.414971828460693, |
| "learning_rate": 1.6698814229249015e-05, |
| "loss": 1.0885, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.2560910545971901, |
| "grad_norm": 6.960204601287842, |
| "learning_rate": 1.6540711462450593e-05, |
| "loss": 1.053, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.26320469500266763, |
| "grad_norm": 6.861100673675537, |
| "learning_rate": 1.6382608695652174e-05, |
| "loss": 1.0398, |
| "step": 1850 |
| }, |
| { |
| "epoch": 0.2703183354081451, |
| "grad_norm": 7.645370960235596, |
| "learning_rate": 1.6224505928853756e-05, |
| "loss": 0.9453, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.2774319758136226, |
| "grad_norm": 7.169046878814697, |
| "learning_rate": 1.6066403162055337e-05, |
| "loss": 0.9847, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.28454561621910013, |
| "grad_norm": 6.290506362915039, |
| "learning_rate": 1.5908300395256918e-05, |
| "loss": 0.9592, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.29165925662457765, |
| "grad_norm": 8.001459121704102, |
| "learning_rate": 1.57501976284585e-05, |
| "loss": 0.9674, |
| "step": 2050 |
| }, |
| { |
| "epoch": 0.2987728970300551, |
| "grad_norm": 8.543235778808594, |
| "learning_rate": 1.559525691699605e-05, |
| "loss": 1.003, |
| "step": 2100 |
| }, |
| { |
| "epoch": 0.3058865374355326, |
| "grad_norm": 10.702069282531738, |
| "learning_rate": 1.543715415019763e-05, |
| "loss": 0.9914, |
| "step": 2150 |
| }, |
| { |
| "epoch": 0.31300017784101014, |
| "grad_norm": 5.234061241149902, |
| "learning_rate": 1.527905138339921e-05, |
| "loss": 1.0809, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.32011381824648766, |
| "grad_norm": 8.00466251373291, |
| "learning_rate": 1.5120948616600793e-05, |
| "loss": 0.9713, |
| "step": 2250 |
| }, |
| { |
| "epoch": 0.3272274586519651, |
| "grad_norm": 7.484270095825195, |
| "learning_rate": 1.4962845849802372e-05, |
| "loss": 0.906, |
| "step": 2300 |
| }, |
| { |
| "epoch": 0.33434109905744264, |
| "grad_norm": 17.150543212890625, |
| "learning_rate": 1.4804743083003954e-05, |
| "loss": 1.0206, |
| "step": 2350 |
| }, |
| { |
| "epoch": 0.34145473946292015, |
| "grad_norm": 8.58785343170166, |
| "learning_rate": 1.4646640316205533e-05, |
| "loss": 0.9868, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.34856837986839767, |
| "grad_norm": 7.040744304656982, |
| "learning_rate": 1.4488537549407116e-05, |
| "loss": 0.9225, |
| "step": 2450 |
| }, |
| { |
| "epoch": 0.35568202027387513, |
| "grad_norm": 8.559785842895508, |
| "learning_rate": 1.4330434782608698e-05, |
| "loss": 0.9345, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.36279566067935265, |
| "grad_norm": 9.140971183776855, |
| "learning_rate": 1.4172332015810277e-05, |
| "loss": 1.004, |
| "step": 2550 |
| }, |
| { |
| "epoch": 0.36990930108483017, |
| "grad_norm": 8.379377365112305, |
| "learning_rate": 1.4014229249011859e-05, |
| "loss": 1.0199, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.3770229414903077, |
| "grad_norm": 6.753004550933838, |
| "learning_rate": 1.3856126482213438e-05, |
| "loss": 0.9463, |
| "step": 2650 |
| }, |
| { |
| "epoch": 0.38413658189578515, |
| "grad_norm": 7.884097576141357, |
| "learning_rate": 1.3698023715415021e-05, |
| "loss": 0.9367, |
| "step": 2700 |
| }, |
| { |
| "epoch": 0.39125022230126266, |
| "grad_norm": 8.405110359191895, |
| "learning_rate": 1.3539920948616603e-05, |
| "loss": 0.9036, |
| "step": 2750 |
| }, |
| { |
| "epoch": 0.3983638627067402, |
| "grad_norm": 7.3707661628723145, |
| "learning_rate": 1.3381818181818182e-05, |
| "loss": 0.9868, |
| "step": 2800 |
| }, |
| { |
| "epoch": 0.4054775031122177, |
| "grad_norm": 9.472613334655762, |
| "learning_rate": 1.3223715415019764e-05, |
| "loss": 0.9809, |
| "step": 2850 |
| }, |
| { |
| "epoch": 0.41259114351769516, |
| "grad_norm": 8.191240310668945, |
| "learning_rate": 1.3065612648221343e-05, |
| "loss": 0.9249, |
| "step": 2900 |
| }, |
| { |
| "epoch": 0.4197047839231727, |
| "grad_norm": 7.100245952606201, |
| "learning_rate": 1.2907509881422926e-05, |
| "loss": 0.9303, |
| "step": 2950 |
| }, |
| { |
| "epoch": 0.4268184243286502, |
| "grad_norm": 6.30331563949585, |
| "learning_rate": 1.2749407114624508e-05, |
| "loss": 0.9459, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.4339320647341277, |
| "grad_norm": 5.755526065826416, |
| "learning_rate": 1.2591304347826087e-05, |
| "loss": 0.8818, |
| "step": 3050 |
| }, |
| { |
| "epoch": 0.44104570513960517, |
| "grad_norm": 7.514272689819336, |
| "learning_rate": 1.243320158102767e-05, |
| "loss": 0.8817, |
| "step": 3100 |
| }, |
| { |
| "epoch": 0.4481593455450827, |
| "grad_norm": 6.345920562744141, |
| "learning_rate": 1.227509881422925e-05, |
| "loss": 0.8257, |
| "step": 3150 |
| }, |
| { |
| "epoch": 0.4552729859505602, |
| "grad_norm": 5.915317058563232, |
| "learning_rate": 1.2120158102766799e-05, |
| "loss": 0.9691, |
| "step": 3200 |
| }, |
| { |
| "epoch": 0.4623866263560377, |
| "grad_norm": 4.785401821136475, |
| "learning_rate": 1.1962055335968382e-05, |
| "loss": 0.97, |
| "step": 3250 |
| }, |
| { |
| "epoch": 0.4695002667615152, |
| "grad_norm": 8.403907775878906, |
| "learning_rate": 1.1803952569169962e-05, |
| "loss": 0.8699, |
| "step": 3300 |
| }, |
| { |
| "epoch": 0.4766139071669927, |
| "grad_norm": 8.583006858825684, |
| "learning_rate": 1.1645849802371543e-05, |
| "loss": 0.9233, |
| "step": 3350 |
| }, |
| { |
| "epoch": 0.4837275475724702, |
| "grad_norm": 7.836755752563477, |
| "learning_rate": 1.1487747035573123e-05, |
| "loss": 0.8716, |
| "step": 3400 |
| }, |
| { |
| "epoch": 0.49084118797794773, |
| "grad_norm": 4.714875221252441, |
| "learning_rate": 1.1329644268774704e-05, |
| "loss": 0.8499, |
| "step": 3450 |
| }, |
| { |
| "epoch": 0.4979548283834252, |
| "grad_norm": 7.097923755645752, |
| "learning_rate": 1.1171541501976287e-05, |
| "loss": 0.912, |
| "step": 3500 |
| }, |
| { |
| "epoch": 0.5050684687889028, |
| "grad_norm": 4.853347301483154, |
| "learning_rate": 1.1013438735177867e-05, |
| "loss": 0.8411, |
| "step": 3550 |
| }, |
| { |
| "epoch": 0.5121821091943802, |
| "grad_norm": 6.642279148101807, |
| "learning_rate": 1.0855335968379448e-05, |
| "loss": 0.9325, |
| "step": 3600 |
| }, |
| { |
| "epoch": 0.5192957495998577, |
| "grad_norm": 6.702445030212402, |
| "learning_rate": 1.0697233201581028e-05, |
| "loss": 0.8354, |
| "step": 3650 |
| }, |
| { |
| "epoch": 0.5264093900053353, |
| "grad_norm": 6.634207725524902, |
| "learning_rate": 1.0539130434782609e-05, |
| "loss": 0.8813, |
| "step": 3700 |
| }, |
| { |
| "epoch": 0.5335230304108127, |
| "grad_norm": 4.122372627258301, |
| "learning_rate": 1.0381027667984192e-05, |
| "loss": 0.7811, |
| "step": 3750 |
| }, |
| { |
| "epoch": 0.5406366708162902, |
| "grad_norm": 8.899633407592773, |
| "learning_rate": 1.0222924901185772e-05, |
| "loss": 0.8707, |
| "step": 3800 |
| }, |
| { |
| "epoch": 0.5477503112217678, |
| "grad_norm": 5.374997615814209, |
| "learning_rate": 1.0064822134387353e-05, |
| "loss": 0.8854, |
| "step": 3850 |
| }, |
| { |
| "epoch": 0.5548639516272452, |
| "grad_norm": 7.9275102615356445, |
| "learning_rate": 9.906719367588934e-06, |
| "loss": 0.9255, |
| "step": 3900 |
| }, |
| { |
| "epoch": 0.5619775920327228, |
| "grad_norm": 6.925314903259277, |
| "learning_rate": 9.748616600790514e-06, |
| "loss": 0.9027, |
| "step": 3950 |
| }, |
| { |
| "epoch": 0.5690912324382003, |
| "grad_norm": 9.117793083190918, |
| "learning_rate": 9.590513833992095e-06, |
| "loss": 0.8831, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.5762048728436777, |
| "grad_norm": 5.861302852630615, |
| "learning_rate": 9.432411067193677e-06, |
| "loss": 0.8329, |
| "step": 4050 |
| }, |
| { |
| "epoch": 0.5833185132491553, |
| "grad_norm": 6.011074066162109, |
| "learning_rate": 9.274308300395256e-06, |
| "loss": 0.8766, |
| "step": 4100 |
| }, |
| { |
| "epoch": 0.5904321536546328, |
| "grad_norm": 5.763038158416748, |
| "learning_rate": 9.11620553359684e-06, |
| "loss": 0.9374, |
| "step": 4150 |
| }, |
| { |
| "epoch": 0.5975457940601102, |
| "grad_norm": 5.520111083984375, |
| "learning_rate": 8.95810276679842e-06, |
| "loss": 0.8813, |
| "step": 4200 |
| }, |
| { |
| "epoch": 0.6046594344655878, |
| "grad_norm": 7.351500511169434, |
| "learning_rate": 8.8e-06, |
| "loss": 0.8677, |
| "step": 4250 |
| }, |
| { |
| "epoch": 0.6117730748710652, |
| "grad_norm": 5.451608657836914, |
| "learning_rate": 8.641897233201582e-06, |
| "loss": 0.8606, |
| "step": 4300 |
| }, |
| { |
| "epoch": 0.6188867152765428, |
| "grad_norm": 5.458895683288574, |
| "learning_rate": 8.483794466403163e-06, |
| "loss": 0.8522, |
| "step": 4350 |
| }, |
| { |
| "epoch": 0.6260003556820203, |
| "grad_norm": 5.554731369018555, |
| "learning_rate": 8.325691699604743e-06, |
| "loss": 0.9042, |
| "step": 4400 |
| }, |
| { |
| "epoch": 0.6331139960874977, |
| "grad_norm": 3.6643424034118652, |
| "learning_rate": 8.167588932806326e-06, |
| "loss": 0.8696, |
| "step": 4450 |
| }, |
| { |
| "epoch": 0.6402276364929753, |
| "grad_norm": 5.593344688415527, |
| "learning_rate": 8.009486166007906e-06, |
| "loss": 0.7663, |
| "step": 4500 |
| }, |
| { |
| "epoch": 0.6473412768984528, |
| "grad_norm": 7.597021102905273, |
| "learning_rate": 7.851383399209487e-06, |
| "loss": 0.7752, |
| "step": 4550 |
| }, |
| { |
| "epoch": 0.6544549173039302, |
| "grad_norm": 4.095972537994385, |
| "learning_rate": 7.693280632411068e-06, |
| "loss": 0.897, |
| "step": 4600 |
| }, |
| { |
| "epoch": 0.6615685577094078, |
| "grad_norm": 7.817249298095703, |
| "learning_rate": 7.535177865612649e-06, |
| "loss": 0.7448, |
| "step": 4650 |
| }, |
| { |
| "epoch": 0.6686821981148853, |
| "grad_norm": 3.296755313873291, |
| "learning_rate": 7.37707509881423e-06, |
| "loss": 0.8065, |
| "step": 4700 |
| }, |
| { |
| "epoch": 0.6757958385203628, |
| "grad_norm": 5.186858177185059, |
| "learning_rate": 7.218972332015811e-06, |
| "loss": 0.8618, |
| "step": 4750 |
| }, |
| { |
| "epoch": 0.6829094789258403, |
| "grad_norm": 6.484861373901367, |
| "learning_rate": 7.060869565217392e-06, |
| "loss": 0.8549, |
| "step": 4800 |
| }, |
| { |
| "epoch": 0.6900231193313178, |
| "grad_norm": 4.90764856338501, |
| "learning_rate": 6.902766798418972e-06, |
| "loss": 0.9441, |
| "step": 4850 |
| }, |
| { |
| "epoch": 0.6971367597367953, |
| "grad_norm": 3.6083099842071533, |
| "learning_rate": 6.744664031620554e-06, |
| "loss": 0.7467, |
| "step": 4900 |
| }, |
| { |
| "epoch": 0.7042504001422728, |
| "grad_norm": 4.79904317855835, |
| "learning_rate": 6.586561264822134e-06, |
| "loss": 0.7974, |
| "step": 4950 |
| }, |
| { |
| "epoch": 0.7113640405477503, |
| "grad_norm": 8.30154800415039, |
| "learning_rate": 6.4284584980237165e-06, |
| "loss": 0.8927, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.7184776809532278, |
| "grad_norm": 5.582029819488525, |
| "learning_rate": 6.270355731225297e-06, |
| "loss": 0.8348, |
| "step": 5050 |
| }, |
| { |
| "epoch": 0.7255913213587053, |
| "grad_norm": 7.575035572052002, |
| "learning_rate": 6.112252964426878e-06, |
| "loss": 0.8691, |
| "step": 5100 |
| }, |
| { |
| "epoch": 0.7327049617641829, |
| "grad_norm": 6.100114345550537, |
| "learning_rate": 5.954150197628459e-06, |
| "loss": 0.788, |
| "step": 5150 |
| }, |
| { |
| "epoch": 0.7398186021696603, |
| "grad_norm": 7.05925178527832, |
| "learning_rate": 5.796047430830039e-06, |
| "loss": 0.8325, |
| "step": 5200 |
| }, |
| { |
| "epoch": 0.7469322425751378, |
| "grad_norm": 6.5520920753479, |
| "learning_rate": 5.6379446640316215e-06, |
| "loss": 0.7792, |
| "step": 5250 |
| }, |
| { |
| "epoch": 0.7540458829806154, |
| "grad_norm": 6.9081807136535645, |
| "learning_rate": 5.479841897233202e-06, |
| "loss": 0.805, |
| "step": 5300 |
| }, |
| { |
| "epoch": 0.7611595233860928, |
| "grad_norm": 7.752345085144043, |
| "learning_rate": 5.321739130434783e-06, |
| "loss": 0.7998, |
| "step": 5350 |
| }, |
| { |
| "epoch": 0.7682731637915703, |
| "grad_norm": 3.8013064861297607, |
| "learning_rate": 5.163636363636364e-06, |
| "loss": 0.7748, |
| "step": 5400 |
| }, |
| { |
| "epoch": 0.7753868041970479, |
| "grad_norm": 5.316600322723389, |
| "learning_rate": 5.005533596837945e-06, |
| "loss": 0.7644, |
| "step": 5450 |
| }, |
| { |
| "epoch": 0.7825004446025253, |
| "grad_norm": 8.250207901000977, |
| "learning_rate": 4.8474308300395266e-06, |
| "loss": 0.7766, |
| "step": 5500 |
| }, |
| { |
| "epoch": 0.7896140850080029, |
| "grad_norm": 5.134970664978027, |
| "learning_rate": 4.689328063241107e-06, |
| "loss": 0.79, |
| "step": 5550 |
| }, |
| { |
| "epoch": 0.7967277254134804, |
| "grad_norm": 5.832414627075195, |
| "learning_rate": 4.531225296442688e-06, |
| "loss": 0.9154, |
| "step": 5600 |
| }, |
| { |
| "epoch": 0.8038413658189578, |
| "grad_norm": 6.164233684539795, |
| "learning_rate": 4.373122529644269e-06, |
| "loss": 0.8288, |
| "step": 5650 |
| }, |
| { |
| "epoch": 0.8109550062244354, |
| "grad_norm": 11.60009479522705, |
| "learning_rate": 4.21501976284585e-06, |
| "loss": 0.7047, |
| "step": 5700 |
| }, |
| { |
| "epoch": 0.8180686466299129, |
| "grad_norm": 3.422287702560425, |
| "learning_rate": 4.056916996047432e-06, |
| "loss": 0.7387, |
| "step": 5750 |
| }, |
| { |
| "epoch": 0.8251822870353903, |
| "grad_norm": 3.786428689956665, |
| "learning_rate": 3.898814229249012e-06, |
| "loss": 0.806, |
| "step": 5800 |
| }, |
| { |
| "epoch": 0.8322959274408679, |
| "grad_norm": 3.7693772315979004, |
| "learning_rate": 3.740711462450593e-06, |
| "loss": 0.8136, |
| "step": 5850 |
| }, |
| { |
| "epoch": 0.8394095678463454, |
| "grad_norm": 3.2537221908569336, |
| "learning_rate": 3.5826086956521744e-06, |
| "loss": 0.7834, |
| "step": 5900 |
| }, |
| { |
| "epoch": 0.8465232082518229, |
| "grad_norm": 2.938720226287842, |
| "learning_rate": 3.4245059288537553e-06, |
| "loss": 0.7607, |
| "step": 5950 |
| }, |
| { |
| "epoch": 0.8536368486573004, |
| "grad_norm": 6.700362205505371, |
| "learning_rate": 3.266403162055336e-06, |
| "loss": 0.8713, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.8607504890627778, |
| "grad_norm": 4.768782615661621, |
| "learning_rate": 3.1083003952569176e-06, |
| "loss": 0.8086, |
| "step": 6050 |
| }, |
| { |
| "epoch": 0.8678641294682554, |
| "grad_norm": 6.537806034088135, |
| "learning_rate": 2.950197628458498e-06, |
| "loss": 0.8722, |
| "step": 6100 |
| }, |
| { |
| "epoch": 0.8749777698737329, |
| "grad_norm": 6.817727088928223, |
| "learning_rate": 2.792094861660079e-06, |
| "loss": 0.7126, |
| "step": 6150 |
| }, |
| { |
| "epoch": 0.8820914102792103, |
| "grad_norm": 5.932295799255371, |
| "learning_rate": 2.6339920948616604e-06, |
| "loss": 0.778, |
| "step": 6200 |
| }, |
| { |
| "epoch": 0.8892050506846879, |
| "grad_norm": 6.475405693054199, |
| "learning_rate": 2.4758893280632413e-06, |
| "loss": 0.7812, |
| "step": 6250 |
| }, |
| { |
| "epoch": 0.8963186910901654, |
| "grad_norm": 8.57321834564209, |
| "learning_rate": 2.3177865612648222e-06, |
| "loss": 0.7654, |
| "step": 6300 |
| }, |
| { |
| "epoch": 0.903432331495643, |
| "grad_norm": 1.9723538160324097, |
| "learning_rate": 2.1596837944664036e-06, |
| "loss": 0.7701, |
| "step": 6350 |
| }, |
| { |
| "epoch": 0.9105459719011204, |
| "grad_norm": 6.585085391998291, |
| "learning_rate": 2.001581027667984e-06, |
| "loss": 0.8419, |
| "step": 6400 |
| }, |
| { |
| "epoch": 0.9176596123065979, |
| "grad_norm": 4.754767894744873, |
| "learning_rate": 1.8434782608695654e-06, |
| "loss": 0.9006, |
| "step": 6450 |
| }, |
| { |
| "epoch": 0.9247732527120754, |
| "grad_norm": 4.661940574645996, |
| "learning_rate": 1.6853754940711464e-06, |
| "loss": 0.7784, |
| "step": 6500 |
| }, |
| { |
| "epoch": 0.9318868931175529, |
| "grad_norm": 4.972583770751953, |
| "learning_rate": 1.5272727272727275e-06, |
| "loss": 0.8112, |
| "step": 6550 |
| }, |
| { |
| "epoch": 0.9390005335230304, |
| "grad_norm": 6.641800403594971, |
| "learning_rate": 1.3691699604743084e-06, |
| "loss": 0.7466, |
| "step": 6600 |
| }, |
| { |
| "epoch": 0.9461141739285079, |
| "grad_norm": 5.098995208740234, |
| "learning_rate": 1.2110671936758893e-06, |
| "loss": 0.824, |
| "step": 6650 |
| }, |
| { |
| "epoch": 0.9532278143339854, |
| "grad_norm": 7.060574054718018, |
| "learning_rate": 1.0529644268774705e-06, |
| "loss": 0.8502, |
| "step": 6700 |
| }, |
| { |
| "epoch": 0.960341454739463, |
| "grad_norm": 4.678645610809326, |
| "learning_rate": 8.948616600790515e-07, |
| "loss": 0.8004, |
| "step": 6750 |
| }, |
| { |
| "epoch": 0.9674550951449404, |
| "grad_norm": 5.895313739776611, |
| "learning_rate": 7.367588932806324e-07, |
| "loss": 0.6811, |
| "step": 6800 |
| }, |
| { |
| "epoch": 0.9745687355504179, |
| "grad_norm": 5.316034317016602, |
| "learning_rate": 5.786561264822135e-07, |
| "loss": 0.7955, |
| "step": 6850 |
| }, |
| { |
| "epoch": 0.9816823759558955, |
| "grad_norm": 5.032801628112793, |
| "learning_rate": 4.205533596837945e-07, |
| "loss": 0.8206, |
| "step": 6900 |
| }, |
| { |
| "epoch": 0.9887960163613729, |
| "grad_norm": 5.486656188964844, |
| "learning_rate": 2.6245059288537554e-07, |
| "loss": 0.795, |
| "step": 6950 |
| }, |
| { |
| "epoch": 0.9959096567668504, |
| "grad_norm": 6.862821102142334, |
| "learning_rate": 1.0434782608695654e-07, |
| "loss": 0.676, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.9998932953939178, |
| "eval_loss": 0.767158567905426, |
| "eval_runtime": 341.0802, |
| "eval_samples_per_second": 15.454, |
| "eval_steps_per_second": 1.932, |
| "step": 7028 |
| } |
| ], |
| "logging_steps": 50, |
| "max_steps": 7028, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 1, |
| "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": 3.803879769007718e+16, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
