| { | |
| "best_metric": 0.7515735626220703, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-400", | |
| "epoch": 0.2536461636017755, | |
| "eval_steps": 50, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0006341154090044388, | |
| "grad_norm": 4.2821364402771, | |
| "learning_rate": 1e-05, | |
| "loss": 5.3646, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0006341154090044388, | |
| "eval_loss": 1.5361766815185547, | |
| "eval_runtime": 296.5915, | |
| "eval_samples_per_second": 8.955, | |
| "eval_steps_per_second": 2.239, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0012682308180088776, | |
| "grad_norm": 4.619045734405518, | |
| "learning_rate": 2e-05, | |
| "loss": 5.7104, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0019023462270133164, | |
| "grad_norm": 6.930375099182129, | |
| "learning_rate": 3e-05, | |
| "loss": 5.8588, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0025364616360177552, | |
| "grad_norm": 3.5030791759490967, | |
| "learning_rate": 4e-05, | |
| "loss": 5.5698, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0031705770450221942, | |
| "grad_norm": 2.9358534812927246, | |
| "learning_rate": 5e-05, | |
| "loss": 5.4895, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.003804692454026633, | |
| "grad_norm": 3.55356502532959, | |
| "learning_rate": 6e-05, | |
| "loss": 4.971, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.004438807863031071, | |
| "grad_norm": 3.4778835773468018, | |
| "learning_rate": 7e-05, | |
| "loss": 4.8542, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0050729232720355105, | |
| "grad_norm": 2.8969311714172363, | |
| "learning_rate": 8e-05, | |
| "loss": 4.3794, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0057070386810399495, | |
| "grad_norm": 2.9659183025360107, | |
| "learning_rate": 9e-05, | |
| "loss": 4.1775, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0063411540900443885, | |
| "grad_norm": 2.9083251953125, | |
| "learning_rate": 0.0001, | |
| "loss": 4.0787, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.006975269499048827, | |
| "grad_norm": 2.144608736038208, | |
| "learning_rate": 9.99983777858264e-05, | |
| "loss": 3.96, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.007609384908053266, | |
| "grad_norm": 2.0248992443084717, | |
| "learning_rate": 9.999351124856874e-05, | |
| "loss": 3.7824, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.008243500317057704, | |
| "grad_norm": 3.3520021438598633, | |
| "learning_rate": 9.998540070400966e-05, | |
| "loss": 3.6636, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.008877615726062143, | |
| "grad_norm": 1.616699457168579, | |
| "learning_rate": 9.997404667843075e-05, | |
| "loss": 3.8141, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.009511731135066582, | |
| "grad_norm": 1.5765314102172852, | |
| "learning_rate": 9.995944990857849e-05, | |
| "loss": 3.8561, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.010145846544071021, | |
| "grad_norm": 1.6568735837936401, | |
| "learning_rate": 9.994161134161634e-05, | |
| "loss": 3.796, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.01077996195307546, | |
| "grad_norm": 1.4620311260223389, | |
| "learning_rate": 9.992053213506334e-05, | |
| "loss": 3.6469, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.011414077362079899, | |
| "grad_norm": 1.6220909357070923, | |
| "learning_rate": 9.989621365671902e-05, | |
| "loss": 3.712, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.012048192771084338, | |
| "grad_norm": 1.3328747749328613, | |
| "learning_rate": 9.986865748457457e-05, | |
| "loss": 3.6251, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.012682308180088777, | |
| "grad_norm": 1.767844319343567, | |
| "learning_rate": 9.983786540671051e-05, | |
| "loss": 3.6623, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.013316423589093214, | |
| "grad_norm": 1.543691873550415, | |
| "learning_rate": 9.980383942118066e-05, | |
| "loss": 3.7249, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.013950538998097653, | |
| "grad_norm": 1.1989096403121948, | |
| "learning_rate": 9.976658173588244e-05, | |
| "loss": 3.5418, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.014584654407102092, | |
| "grad_norm": 1.437749981880188, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 3.5405, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.015218769816106531, | |
| "grad_norm": 1.3658512830734253, | |
| "learning_rate": 9.968238114591566e-05, | |
| "loss": 3.4703, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.01585288522511097, | |
| "grad_norm": 1.3563766479492188, | |
| "learning_rate": 9.96354437049027e-05, | |
| "loss": 3.6742, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.016487000634115408, | |
| "grad_norm": 1.4246914386749268, | |
| "learning_rate": 9.95852854910781e-05, | |
| "loss": 3.5858, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.017121116043119847, | |
| "grad_norm": 1.4282399415969849, | |
| "learning_rate": 9.953190975913647e-05, | |
| "loss": 3.4487, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.017755231452124286, | |
| "grad_norm": 1.4253621101379395, | |
| "learning_rate": 9.947531997255256e-05, | |
| "loss": 3.415, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.018389346861128725, | |
| "grad_norm": 1.4151033163070679, | |
| "learning_rate": 9.941551980335652e-05, | |
| "loss": 3.5219, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.019023462270133164, | |
| "grad_norm": 1.5081020593643188, | |
| "learning_rate": 9.935251313189564e-05, | |
| "loss": 3.5514, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.019657577679137603, | |
| "grad_norm": 1.3580255508422852, | |
| "learning_rate": 9.928630404658255e-05, | |
| "loss": 3.4421, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.020291693088142042, | |
| "grad_norm": 1.3717368841171265, | |
| "learning_rate": 9.921689684362989e-05, | |
| "loss": 3.3642, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.02092580849714648, | |
| "grad_norm": 1.3913062810897827, | |
| "learning_rate": 9.914429602677162e-05, | |
| "loss": 3.4983, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.02155992390615092, | |
| "grad_norm": 1.3363810777664185, | |
| "learning_rate": 9.906850630697068e-05, | |
| "loss": 3.3011, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.02219403931515536, | |
| "grad_norm": 1.3653063774108887, | |
| "learning_rate": 9.898953260211338e-05, | |
| "loss": 3.2758, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.022828154724159798, | |
| "grad_norm": 1.2513288259506226, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 3.3022, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.023462270133164237, | |
| "grad_norm": 1.301414966583252, | |
| "learning_rate": 9.882205394146361e-05, | |
| "loss": 3.3013, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.024096385542168676, | |
| "grad_norm": 1.2664551734924316, | |
| "learning_rate": 9.87335598531214e-05, | |
| "loss": 3.5038, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.024730500951173115, | |
| "grad_norm": 1.233097791671753, | |
| "learning_rate": 9.864190351391822e-05, | |
| "loss": 3.2793, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.025364616360177554, | |
| "grad_norm": 1.2509174346923828, | |
| "learning_rate": 9.85470908713026e-05, | |
| "loss": 3.3181, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.02599873176918199, | |
| "grad_norm": 1.3066641092300415, | |
| "learning_rate": 9.844912807753104e-05, | |
| "loss": 3.2021, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.02663284717818643, | |
| "grad_norm": 1.2607433795928955, | |
| "learning_rate": 9.834802148926882e-05, | |
| "loss": 3.0709, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.027266962587190868, | |
| "grad_norm": 1.3766813278198242, | |
| "learning_rate": 9.824377766717759e-05, | |
| "loss": 3.175, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.027901077996195307, | |
| "grad_norm": 1.3962171077728271, | |
| "learning_rate": 9.813640337548954e-05, | |
| "loss": 3.2248, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.028535193405199746, | |
| "grad_norm": 1.5123058557510376, | |
| "learning_rate": 9.802590558156862e-05, | |
| "loss": 3.0562, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.029169308814204185, | |
| "grad_norm": 1.3668205738067627, | |
| "learning_rate": 9.791229145545831e-05, | |
| "loss": 2.8726, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.029803424223208624, | |
| "grad_norm": 1.6242258548736572, | |
| "learning_rate": 9.779556836941645e-05, | |
| "loss": 2.9806, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.030437539632213063, | |
| "grad_norm": 1.840408205986023, | |
| "learning_rate": 9.767574389743682e-05, | |
| "loss": 2.766, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.031071655041217502, | |
| "grad_norm": 1.5403425693511963, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 2.825, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.03170577045022194, | |
| "grad_norm": 1.665044903755188, | |
| "learning_rate": 9.742682209735727e-05, | |
| "loss": 2.9077, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03170577045022194, | |
| "eval_loss": 0.8847057223320007, | |
| "eval_runtime": 297.6178, | |
| "eval_samples_per_second": 8.924, | |
| "eval_steps_per_second": 2.231, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.032339885859226376, | |
| "grad_norm": 3.129869222640991, | |
| "learning_rate": 9.729774092143627e-05, | |
| "loss": 3.7664, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.032974001268230815, | |
| "grad_norm": 2.0350043773651123, | |
| "learning_rate": 9.716559066288715e-05, | |
| "loss": 3.6251, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.033608116677235254, | |
| "grad_norm": 1.3122864961624146, | |
| "learning_rate": 9.703037989675087e-05, | |
| "loss": 3.5988, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.03424223208623969, | |
| "grad_norm": 1.3547205924987793, | |
| "learning_rate": 9.689211739666023e-05, | |
| "loss": 3.4852, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.03487634749524413, | |
| "grad_norm": 1.131834864616394, | |
| "learning_rate": 9.675081213427076e-05, | |
| "loss": 3.4682, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.03551046290424857, | |
| "grad_norm": 1.118656039237976, | |
| "learning_rate": 9.66064732786784e-05, | |
| "loss": 3.4695, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.03614457831325301, | |
| "grad_norm": 1.095979928970337, | |
| "learning_rate": 9.645911019582467e-05, | |
| "loss": 3.3906, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.03677869372225745, | |
| "grad_norm": 1.330224633216858, | |
| "learning_rate": 9.630873244788883e-05, | |
| "loss": 3.5858, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.03741280913126189, | |
| "grad_norm": 1.1612094640731812, | |
| "learning_rate": 9.615534979266745e-05, | |
| "loss": 3.5082, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.03804692454026633, | |
| "grad_norm": 1.1439632177352905, | |
| "learning_rate": 9.599897218294122e-05, | |
| "loss": 3.4478, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.03868103994927077, | |
| "grad_norm": 1.480470895767212, | |
| "learning_rate": 9.583960976582913e-05, | |
| "loss": 3.3747, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.039315155358275206, | |
| "grad_norm": 1.1436293125152588, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 3.3507, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.039949270767279645, | |
| "grad_norm": 1.1501903533935547, | |
| "learning_rate": 9.551197206565173e-05, | |
| "loss": 3.3903, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.040583386176284084, | |
| "grad_norm": 1.1135419607162476, | |
| "learning_rate": 9.534371804252728e-05, | |
| "loss": 3.2974, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.04121750158528852, | |
| "grad_norm": 1.1008535623550415, | |
| "learning_rate": 9.517252173051911e-05, | |
| "loss": 3.3343, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.04185161699429296, | |
| "grad_norm": 1.1229270696640015, | |
| "learning_rate": 9.49983942383106e-05, | |
| "loss": 3.3202, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.0424857324032974, | |
| "grad_norm": 1.129658579826355, | |
| "learning_rate": 9.482134686478519e-05, | |
| "loss": 3.4031, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.04311984781230184, | |
| "grad_norm": 1.1309914588928223, | |
| "learning_rate": 9.464139109829321e-05, | |
| "loss": 3.3727, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.04375396322130628, | |
| "grad_norm": 1.1195210218429565, | |
| "learning_rate": 9.445853861590647e-05, | |
| "loss": 3.3391, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.04438807863031072, | |
| "grad_norm": 1.14125657081604, | |
| "learning_rate": 9.42728012826605e-05, | |
| "loss": 3.2907, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.04502219403931516, | |
| "grad_norm": 1.1160163879394531, | |
| "learning_rate": 9.408419115078471e-05, | |
| "loss": 3.2616, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.045656309448319596, | |
| "grad_norm": 1.1449788808822632, | |
| "learning_rate": 9.389272045892024e-05, | |
| "loss": 3.3349, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.046290424857324035, | |
| "grad_norm": 1.096792221069336, | |
| "learning_rate": 9.36984016313259e-05, | |
| "loss": 3.2215, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.046924540266328474, | |
| "grad_norm": 1.1410120725631714, | |
| "learning_rate": 9.350124727707197e-05, | |
| "loss": 3.3399, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.04755865567533291, | |
| "grad_norm": 1.1690126657485962, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 3.354, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.04819277108433735, | |
| "grad_norm": 1.2746645212173462, | |
| "learning_rate": 9.309848334400246e-05, | |
| "loss": 3.3572, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.04882688649334179, | |
| "grad_norm": 1.1271394491195679, | |
| "learning_rate": 9.289289989996133e-05, | |
| "loss": 3.3658, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.04946100190234623, | |
| "grad_norm": 1.2954553365707397, | |
| "learning_rate": 9.268453319711363e-05, | |
| "loss": 3.2274, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.05009511731135067, | |
| "grad_norm": 1.1216834783554077, | |
| "learning_rate": 9.247339675607605e-05, | |
| "loss": 3.267, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.05072923272035511, | |
| "grad_norm": 1.1674860715866089, | |
| "learning_rate": 9.225950427718975e-05, | |
| "loss": 3.3223, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.05136334812935954, | |
| "grad_norm": 1.0981123447418213, | |
| "learning_rate": 9.204286963963111e-05, | |
| "loss": 3.1553, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.05199746353836398, | |
| "grad_norm": 1.2098571062088013, | |
| "learning_rate": 9.182350690051133e-05, | |
| "loss": 3.287, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.05263157894736842, | |
| "grad_norm": 1.128381371498108, | |
| "learning_rate": 9.160143029396422e-05, | |
| "loss": 3.26, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.05326569435637286, | |
| "grad_norm": 1.2328784465789795, | |
| "learning_rate": 9.13766542302225e-05, | |
| "loss": 3.3351, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.053899809765377296, | |
| "grad_norm": 1.1560747623443604, | |
| "learning_rate": 9.114919329468282e-05, | |
| "loss": 3.268, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.054533925174381735, | |
| "grad_norm": 1.191244125366211, | |
| "learning_rate": 9.091906224695935e-05, | |
| "loss": 3.1744, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.055168040583386174, | |
| "grad_norm": 1.1172856092453003, | |
| "learning_rate": 9.068627601992598e-05, | |
| "loss": 3.1647, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.05580215599239061, | |
| "grad_norm": 1.1811506748199463, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 3.1507, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.05643627140139505, | |
| "grad_norm": 1.2562776803970337, | |
| "learning_rate": 9.021279861989885e-05, | |
| "loss": 3.1243, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.05707038681039949, | |
| "grad_norm": 1.2073371410369873, | |
| "learning_rate": 8.997213817017507e-05, | |
| "loss": 3.2071, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.05770450221940393, | |
| "grad_norm": 1.179689884185791, | |
| "learning_rate": 8.972888398568772e-05, | |
| "loss": 3.1118, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.05833861762840837, | |
| "grad_norm": 1.1589559316635132, | |
| "learning_rate": 8.948305185085225e-05, | |
| "loss": 3.0971, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.05897273303741281, | |
| "grad_norm": 1.5334442853927612, | |
| "learning_rate": 8.92346577173636e-05, | |
| "loss": 2.9583, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.05960684844641725, | |
| "grad_norm": 1.1919718980789185, | |
| "learning_rate": 8.898371770316111e-05, | |
| "loss": 2.9499, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.060240963855421686, | |
| "grad_norm": 1.2710996866226196, | |
| "learning_rate": 8.873024809138272e-05, | |
| "loss": 3.0539, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.060875079264426125, | |
| "grad_norm": 1.2094289064407349, | |
| "learning_rate": 8.847426532930831e-05, | |
| "loss": 2.9798, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.061509194673430564, | |
| "grad_norm": 1.2845017910003662, | |
| "learning_rate": 8.821578602729242e-05, | |
| "loss": 2.7427, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.062143310082435003, | |
| "grad_norm": 1.26568603515625, | |
| "learning_rate": 8.795482695768658e-05, | |
| "loss": 2.7497, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.06277742549143944, | |
| "grad_norm": 1.4030022621154785, | |
| "learning_rate": 8.769140505375085e-05, | |
| "loss": 2.6945, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.06341154090044387, | |
| "grad_norm": 1.5023269653320312, | |
| "learning_rate": 8.742553740855506e-05, | |
| "loss": 2.8778, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06341154090044387, | |
| "eval_loss": 0.8155827522277832, | |
| "eval_runtime": 297.7632, | |
| "eval_samples_per_second": 8.92, | |
| "eval_steps_per_second": 2.23, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06404565630944832, | |
| "grad_norm": 1.5373103618621826, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 3.477, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.06467977171845275, | |
| "grad_norm": 1.278006911277771, | |
| "learning_rate": 8.688653405904652e-05, | |
| "loss": 3.476, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.0653138871274572, | |
| "grad_norm": 1.0467379093170166, | |
| "learning_rate": 8.661343332988869e-05, | |
| "loss": 3.3155, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.06594800253646163, | |
| "grad_norm": 1.3283839225769043, | |
| "learning_rate": 8.633795680751116e-05, | |
| "loss": 3.4751, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.06658211794546608, | |
| "grad_norm": 1.1208345890045166, | |
| "learning_rate": 8.606012236719073e-05, | |
| "loss": 3.228, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.06721623335447051, | |
| "grad_norm": 1.0446490049362183, | |
| "learning_rate": 8.577994803720606e-05, | |
| "loss": 3.3327, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.06785034876347495, | |
| "grad_norm": 1.0248538255691528, | |
| "learning_rate": 8.549745199766792e-05, | |
| "loss": 3.348, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.06848446417247939, | |
| "grad_norm": 1.1333471536636353, | |
| "learning_rate": 8.521265257933948e-05, | |
| "loss": 3.3869, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.06911857958148383, | |
| "grad_norm": 1.1693687438964844, | |
| "learning_rate": 8.492556826244687e-05, | |
| "loss": 3.326, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.06975269499048826, | |
| "grad_norm": 1.0264679193496704, | |
| "learning_rate": 8.463621767547998e-05, | |
| "loss": 3.2675, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.07038681039949271, | |
| "grad_norm": 1.103921890258789, | |
| "learning_rate": 8.434461959398376e-05, | |
| "loss": 3.354, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.07102092580849714, | |
| "grad_norm": 1.0422513484954834, | |
| "learning_rate": 8.405079293933986e-05, | |
| "loss": 3.2956, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.07165504121750159, | |
| "grad_norm": 1.0689914226531982, | |
| "learning_rate": 8.375475677753881e-05, | |
| "loss": 3.3304, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.07228915662650602, | |
| "grad_norm": 1.0220685005187988, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 3.3593, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.07292327203551047, | |
| "grad_norm": 1.1142345666885376, | |
| "learning_rate": 8.315613291203976e-05, | |
| "loss": 3.295, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.0735573874445149, | |
| "grad_norm": 1.048160433769226, | |
| "learning_rate": 8.285358405218655e-05, | |
| "loss": 3.2746, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.07419150285351934, | |
| "grad_norm": 1.0649545192718506, | |
| "learning_rate": 8.25489033703452e-05, | |
| "loss": 3.3655, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.07482561826252378, | |
| "grad_norm": 1.0890159606933594, | |
| "learning_rate": 8.224211063680853e-05, | |
| "loss": 3.3308, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.07545973367152822, | |
| "grad_norm": 1.0687116384506226, | |
| "learning_rate": 8.19332257589174e-05, | |
| "loss": 3.1868, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.07609384908053266, | |
| "grad_norm": 1.1016745567321777, | |
| "learning_rate": 8.162226877976887e-05, | |
| "loss": 3.2142, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.0767279644895371, | |
| "grad_norm": 1.051974892616272, | |
| "learning_rate": 8.130925987691569e-05, | |
| "loss": 3.3221, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.07736207989854153, | |
| "grad_norm": 1.0410847663879395, | |
| "learning_rate": 8.099421936105702e-05, | |
| "loss": 3.1397, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.07799619530754598, | |
| "grad_norm": 1.072290301322937, | |
| "learning_rate": 8.067716767472045e-05, | |
| "loss": 3.3434, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.07863031071655041, | |
| "grad_norm": 1.0379129648208618, | |
| "learning_rate": 8.035812539093557e-05, | |
| "loss": 3.25, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.07926442612555486, | |
| "grad_norm": 1.0454907417297363, | |
| "learning_rate": 8.003711321189895e-05, | |
| "loss": 3.2054, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.07989854153455929, | |
| "grad_norm": 1.0810647010803223, | |
| "learning_rate": 7.971415196763088e-05, | |
| "loss": 3.1932, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.08053265694356374, | |
| "grad_norm": 1.0763037204742432, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 3.2082, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.08116677235256817, | |
| "grad_norm": 1.1298437118530273, | |
| "learning_rate": 7.906246623448183e-05, | |
| "loss": 3.1894, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.0818008877615726, | |
| "grad_norm": 1.0926628112792969, | |
| "learning_rate": 7.873378403255419e-05, | |
| "loss": 3.1941, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.08243500317057705, | |
| "grad_norm": 1.0685158967971802, | |
| "learning_rate": 7.840323733655778e-05, | |
| "loss": 3.169, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.08306911857958148, | |
| "grad_norm": 1.0912131071090698, | |
| "learning_rate": 7.807084759519405e-05, | |
| "loss": 3.1239, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.08370323398858592, | |
| "grad_norm": 1.1438196897506714, | |
| "learning_rate": 7.773663637675694e-05, | |
| "loss": 3.273, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.08433734939759036, | |
| "grad_norm": 1.1781691312789917, | |
| "learning_rate": 7.740062536773352e-05, | |
| "loss": 3.3073, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.0849714648065948, | |
| "grad_norm": 1.149361252784729, | |
| "learning_rate": 7.706283637139658e-05, | |
| "loss": 3.2185, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.08560558021559923, | |
| "grad_norm": 1.1654759645462036, | |
| "learning_rate": 7.672329130639005e-05, | |
| "loss": 3.1729, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.08623969562460368, | |
| "grad_norm": 1.3615312576293945, | |
| "learning_rate": 7.638201220530665e-05, | |
| "loss": 3.1193, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.08687381103360811, | |
| "grad_norm": 1.0837875604629517, | |
| "learning_rate": 7.603902121325813e-05, | |
| "loss": 3.1017, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.08750792644261256, | |
| "grad_norm": 1.105128526687622, | |
| "learning_rate": 7.569434058643844e-05, | |
| "loss": 3.1515, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.08814204185161699, | |
| "grad_norm": 1.0796869993209839, | |
| "learning_rate": 7.534799269067953e-05, | |
| "loss": 3.1271, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.08877615726062144, | |
| "grad_norm": 1.2005115747451782, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 3.131, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.08941027266962587, | |
| "grad_norm": 1.122515082359314, | |
| "learning_rate": 7.465038509514688e-05, | |
| "loss": 3.1134, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.09004438807863031, | |
| "grad_norm": 1.1972311735153198, | |
| "learning_rate": 7.42991706621303e-05, | |
| "loss": 3.0486, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.09067850348763475, | |
| "grad_norm": 1.140086054801941, | |
| "learning_rate": 7.394637949075154e-05, | |
| "loss": 2.9324, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.09131261889663919, | |
| "grad_norm": 1.3415659666061401, | |
| "learning_rate": 7.35920344731241e-05, | |
| "loss": 3.0243, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.09194673430564362, | |
| "grad_norm": 1.196213722229004, | |
| "learning_rate": 7.323615860218843e-05, | |
| "loss": 3.0405, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.09258084971464807, | |
| "grad_norm": 1.3614180088043213, | |
| "learning_rate": 7.287877497021978e-05, | |
| "loss": 2.8349, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.0932149651236525, | |
| "grad_norm": 1.288540244102478, | |
| "learning_rate": 7.251990676732984e-05, | |
| "loss": 2.7361, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.09384908053265695, | |
| "grad_norm": 1.2313977479934692, | |
| "learning_rate": 7.215957727996207e-05, | |
| "loss": 2.8024, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.09448319594166138, | |
| "grad_norm": 1.2483394145965576, | |
| "learning_rate": 7.179780988938051e-05, | |
| "loss": 2.5696, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.09511731135066583, | |
| "grad_norm": 1.4693834781646729, | |
| "learning_rate": 7.143462807015271e-05, | |
| "loss": 2.6114, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.09511731135066583, | |
| "eval_loss": 0.8000059723854065, | |
| "eval_runtime": 299.1522, | |
| "eval_samples_per_second": 8.878, | |
| "eval_steps_per_second": 2.22, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.09575142675967026, | |
| "grad_norm": 1.7044169902801514, | |
| "learning_rate": 7.107005538862646e-05, | |
| "loss": 4.1348, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.0963855421686747, | |
| "grad_norm": 1.5047065019607544, | |
| "learning_rate": 7.07041155014006e-05, | |
| "loss": 3.3954, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.09701965757767914, | |
| "grad_norm": 1.2422665357589722, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 3.3153, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.09765377298668358, | |
| "grad_norm": 5.329709053039551, | |
| "learning_rate": 6.996822917828477e-05, | |
| "loss": 3.3014, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.09828788839568801, | |
| "grad_norm": 0.9941044449806213, | |
| "learning_rate": 6.959833049300377e-05, | |
| "loss": 3.1952, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.09892200380469246, | |
| "grad_norm": 1.0377179384231567, | |
| "learning_rate": 6.922716010014255e-05, | |
| "loss": 3.3715, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.09955611921369689, | |
| "grad_norm": 1.0825616121292114, | |
| "learning_rate": 6.885474208441603e-05, | |
| "loss": 3.238, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.10019023462270134, | |
| "grad_norm": 1.0486564636230469, | |
| "learning_rate": 6.848110061149556e-05, | |
| "loss": 3.2671, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.10082435003170577, | |
| "grad_norm": 1.0950785875320435, | |
| "learning_rate": 6.810625992644085e-05, | |
| "loss": 3.387, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.10145846544071022, | |
| "grad_norm": 1.0026979446411133, | |
| "learning_rate": 6.773024435212678e-05, | |
| "loss": 3.2341, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.10209258084971465, | |
| "grad_norm": 1.0109604597091675, | |
| "learning_rate": 6.735307828766515e-05, | |
| "loss": 3.1852, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.10272669625871908, | |
| "grad_norm": 1.0047856569290161, | |
| "learning_rate": 6.697478620682137e-05, | |
| "loss": 3.316, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.10336081166772353, | |
| "grad_norm": 1.0846145153045654, | |
| "learning_rate": 6.659539265642643e-05, | |
| "loss": 3.2932, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.10399492707672796, | |
| "grad_norm": 1.1783442497253418, | |
| "learning_rate": 6.621492225478414e-05, | |
| "loss": 3.2566, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.1046290424857324, | |
| "grad_norm": 1.1924242973327637, | |
| "learning_rate": 6.583339969007363e-05, | |
| "loss": 3.2032, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.10526315789473684, | |
| "grad_norm": 1.0744683742523193, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 3.2983, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.10589727330374128, | |
| "grad_norm": 1.0177804231643677, | |
| "learning_rate": 6.506729716392481e-05, | |
| "loss": 3.1444, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.10653138871274571, | |
| "grad_norm": 1.000748872756958, | |
| "learning_rate": 6.468276691378155e-05, | |
| "loss": 3.1848, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.10716550412175016, | |
| "grad_norm": 0.9723776578903198, | |
| "learning_rate": 6.429728391993446e-05, | |
| "loss": 3.0663, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.10779961953075459, | |
| "grad_norm": 1.0768924951553345, | |
| "learning_rate": 6.391087319582264e-05, | |
| "loss": 3.4086, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.10843373493975904, | |
| "grad_norm": 1.0122095346450806, | |
| "learning_rate": 6.35235598150842e-05, | |
| "loss": 3.1676, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.10906785034876347, | |
| "grad_norm": 0.9996663331985474, | |
| "learning_rate": 6.313536890992935e-05, | |
| "loss": 3.226, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.10970196575776792, | |
| "grad_norm": 1.062509298324585, | |
| "learning_rate": 6.274632566950967e-05, | |
| "loss": 3.2257, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.11033608116677235, | |
| "grad_norm": 1.0289890766143799, | |
| "learning_rate": 6.235645533828349e-05, | |
| "loss": 3.2668, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.1109701965757768, | |
| "grad_norm": 1.0083286762237549, | |
| "learning_rate": 6.19657832143779e-05, | |
| "loss": 3.1725, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.11160431198478123, | |
| "grad_norm": 1.052514672279358, | |
| "learning_rate": 6.157433464794716e-05, | |
| "loss": 3.2222, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.11223842739378567, | |
| "grad_norm": 1.099542260169983, | |
| "learning_rate": 6.118213503952779e-05, | |
| "loss": 3.2265, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.1128725428027901, | |
| "grad_norm": 1.0204391479492188, | |
| "learning_rate": 6.078920983839031e-05, | |
| "loss": 3.221, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.11350665821179455, | |
| "grad_norm": 1.131372094154358, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 3.1018, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.11414077362079898, | |
| "grad_norm": 1.0413641929626465, | |
| "learning_rate": 6.0001284688802226e-05, | |
| "loss": 3.1977, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.11477488902980343, | |
| "grad_norm": 1.0400148630142212, | |
| "learning_rate": 5.960633586768543e-05, | |
| "loss": 3.2413, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.11540900443880786, | |
| "grad_norm": 1.1186673641204834, | |
| "learning_rate": 5.921076370520058e-05, | |
| "loss": 3.1271, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.1160431198478123, | |
| "grad_norm": 5.378909587860107, | |
| "learning_rate": 5.8814593869458455e-05, | |
| "loss": 3.017, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.11667723525681674, | |
| "grad_norm": 1.0467098951339722, | |
| "learning_rate": 5.841785206735192e-05, | |
| "loss": 3.1646, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.11731135066582118, | |
| "grad_norm": 1.2054715156555176, | |
| "learning_rate": 5.8020564042888015e-05, | |
| "loss": 3.1296, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.11794546607482562, | |
| "grad_norm": 1.1292742490768433, | |
| "learning_rate": 5.762275557551727e-05, | |
| "loss": 3.0876, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.11857958148383006, | |
| "grad_norm": 1.0788121223449707, | |
| "learning_rate": 5.7224452478461064e-05, | |
| "loss": 3.0394, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.1192136968928345, | |
| "grad_norm": 1.117724895477295, | |
| "learning_rate": 5.682568059703659e-05, | |
| "loss": 3.2083, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.11984781230183894, | |
| "grad_norm": 1.1697856187820435, | |
| "learning_rate": 5.642646580697973e-05, | |
| "loss": 3.1939, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.12048192771084337, | |
| "grad_norm": 1.1887933015823364, | |
| "learning_rate": 5.602683401276615e-05, | |
| "loss": 3.022, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.12111604311984782, | |
| "grad_norm": 1.121009111404419, | |
| "learning_rate": 5.562681114593028e-05, | |
| "loss": 2.9906, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.12175015852885225, | |
| "grad_norm": 1.1305354833602905, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 3.0668, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.1223842739378567, | |
| "grad_norm": 1.2224658727645874, | |
| "learning_rate": 5.482569604572576e-05, | |
| "loss": 3.1373, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.12301838934686113, | |
| "grad_norm": 1.1487756967544556, | |
| "learning_rate": 5.442465579556793e-05, | |
| "loss": 2.9432, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.12365250475586556, | |
| "grad_norm": 1.124640703201294, | |
| "learning_rate": 5.402332843583631e-05, | |
| "loss": 2.8273, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.12428662016487001, | |
| "grad_norm": 1.1626677513122559, | |
| "learning_rate": 5.3621740008088126e-05, | |
| "loss": 2.7136, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.12492073557387444, | |
| "grad_norm": 1.2303059101104736, | |
| "learning_rate": 5.321991657082097e-05, | |
| "loss": 2.5001, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.12555485098287889, | |
| "grad_norm": 1.3254761695861816, | |
| "learning_rate": 5.281788419778187e-05, | |
| "loss": 2.637, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.12618896639188332, | |
| "grad_norm": 1.3234314918518066, | |
| "learning_rate": 5.2415668976275355e-05, | |
| "loss": 2.5926, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.12682308180088775, | |
| "grad_norm": 1.525715708732605, | |
| "learning_rate": 5.201329700547076e-05, | |
| "loss": 2.5566, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.12682308180088775, | |
| "eval_loss": 0.7819186449050903, | |
| "eval_runtime": 298.2576, | |
| "eval_samples_per_second": 8.905, | |
| "eval_steps_per_second": 2.226, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.1274571972098922, | |
| "grad_norm": 1.2543182373046875, | |
| "learning_rate": 5.161079439470866e-05, | |
| "loss": 3.6287, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.12809131261889664, | |
| "grad_norm": 1.2177551984786987, | |
| "learning_rate": 5.1208187261806615e-05, | |
| "loss": 3.3176, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.12872542802790107, | |
| "grad_norm": 1.089298129081726, | |
| "learning_rate": 5.080550173136457e-05, | |
| "loss": 3.3302, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.1293595434369055, | |
| "grad_norm": 0.9786179065704346, | |
| "learning_rate": 5.0402763933069496e-05, | |
| "loss": 3.2146, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.12999365884590997, | |
| "grad_norm": 0.9849365949630737, | |
| "learning_rate": 5e-05, | |
| "loss": 3.2046, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.1306277742549144, | |
| "grad_norm": 0.9649850130081177, | |
| "learning_rate": 4.9597236066930516e-05, | |
| "loss": 3.3047, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.13126188966391883, | |
| "grad_norm": 0.9902065992355347, | |
| "learning_rate": 4.919449826863544e-05, | |
| "loss": 3.2021, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.13189600507292326, | |
| "grad_norm": 0.9828966856002808, | |
| "learning_rate": 4.87918127381934e-05, | |
| "loss": 3.2409, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.13253012048192772, | |
| "grad_norm": 1.0174707174301147, | |
| "learning_rate": 4.8389205605291365e-05, | |
| "loss": 3.3236, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.13316423589093215, | |
| "grad_norm": 0.9697160124778748, | |
| "learning_rate": 4.798670299452926e-05, | |
| "loss": 3.2682, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.13379835129993659, | |
| "grad_norm": 1.0232049226760864, | |
| "learning_rate": 4.758433102372466e-05, | |
| "loss": 3.2173, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.13443246670894102, | |
| "grad_norm": 0.9926899671554565, | |
| "learning_rate": 4.7182115802218126e-05, | |
| "loss": 3.2237, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.13506658211794548, | |
| "grad_norm": 1.0102065801620483, | |
| "learning_rate": 4.678008342917903e-05, | |
| "loss": 3.1939, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.1357006975269499, | |
| "grad_norm": 1.0068438053131104, | |
| "learning_rate": 4.6378259991911886e-05, | |
| "loss": 3.2868, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.13633481293595434, | |
| "grad_norm": 0.9734184741973877, | |
| "learning_rate": 4.597667156416371e-05, | |
| "loss": 3.1584, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.13696892834495877, | |
| "grad_norm": 15.009803771972656, | |
| "learning_rate": 4.5575344204432084e-05, | |
| "loss": 3.2934, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.13760304375396323, | |
| "grad_norm": 0.9599727988243103, | |
| "learning_rate": 4.5174303954274244e-05, | |
| "loss": 3.1804, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.13823715916296767, | |
| "grad_norm": 1.007860779762268, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 3.2044, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.1388712745719721, | |
| "grad_norm": 0.9852535128593445, | |
| "learning_rate": 4.437318885406973e-05, | |
| "loss": 3.0663, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.13950538998097653, | |
| "grad_norm": 0.9651182293891907, | |
| "learning_rate": 4.397316598723385e-05, | |
| "loss": 3.2841, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.140139505389981, | |
| "grad_norm": 0.9738379120826721, | |
| "learning_rate": 4.3573534193020274e-05, | |
| "loss": 3.1556, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.14077362079898542, | |
| "grad_norm": 1.0059845447540283, | |
| "learning_rate": 4.317431940296343e-05, | |
| "loss": 3.1792, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.14140773620798985, | |
| "grad_norm": 1.044323444366455, | |
| "learning_rate": 4.277554752153895e-05, | |
| "loss": 3.029, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.14204185161699429, | |
| "grad_norm": 1.0486642122268677, | |
| "learning_rate": 4.237724442448273e-05, | |
| "loss": 3.1771, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.14267596702599875, | |
| "grad_norm": 1.0943423509597778, | |
| "learning_rate": 4.197943595711198e-05, | |
| "loss": 3.0797, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.14331008243500318, | |
| "grad_norm": 1.032805323600769, | |
| "learning_rate": 4.1582147932648074e-05, | |
| "loss": 3.2204, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.1439441978440076, | |
| "grad_norm": 1.0199164152145386, | |
| "learning_rate": 4.118540613054156e-05, | |
| "loss": 3.1531, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.14457831325301204, | |
| "grad_norm": 1.031343698501587, | |
| "learning_rate": 4.078923629479943e-05, | |
| "loss": 3.1324, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.14521242866201647, | |
| "grad_norm": 1.0214942693710327, | |
| "learning_rate": 4.039366413231458e-05, | |
| "loss": 3.0392, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.14584654407102093, | |
| "grad_norm": 1.0384721755981445, | |
| "learning_rate": 3.9998715311197785e-05, | |
| "loss": 3.1602, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.14648065948002537, | |
| "grad_norm": 1.062154769897461, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 3.1477, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.1471147748890298, | |
| "grad_norm": 1.1326991319656372, | |
| "learning_rate": 3.92107901616097e-05, | |
| "loss": 3.0429, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.14774889029803423, | |
| "grad_norm": 1.0269993543624878, | |
| "learning_rate": 3.8817864960472236e-05, | |
| "loss": 3.038, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.1483830057070387, | |
| "grad_norm": 1.0221335887908936, | |
| "learning_rate": 3.842566535205286e-05, | |
| "loss": 3.0781, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.14901712111604312, | |
| "grad_norm": 1.4944452047348022, | |
| "learning_rate": 3.803421678562213e-05, | |
| "loss": 3.1389, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.14965123652504755, | |
| "grad_norm": 1.0359796285629272, | |
| "learning_rate": 3.764354466171652e-05, | |
| "loss": 3.0269, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.15028535193405199, | |
| "grad_norm": 1.1374757289886475, | |
| "learning_rate": 3.725367433049033e-05, | |
| "loss": 3.0241, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.15091946734305645, | |
| "grad_norm": 1.0923086404800415, | |
| "learning_rate": 3.6864631090070655e-05, | |
| "loss": 2.9826, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.15155358275206088, | |
| "grad_norm": 1.085671305656433, | |
| "learning_rate": 3.6476440184915815e-05, | |
| "loss": 3.0977, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.1521876981610653, | |
| "grad_norm": 1.0791702270507812, | |
| "learning_rate": 3.608912680417737e-05, | |
| "loss": 2.9639, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.15282181357006974, | |
| "grad_norm": 1.1274312734603882, | |
| "learning_rate": 3.570271608006555e-05, | |
| "loss": 2.8279, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.1534559289790742, | |
| "grad_norm": 1.0965379476547241, | |
| "learning_rate": 3.531723308621847e-05, | |
| "loss": 2.9191, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.15409004438807863, | |
| "grad_norm": 1.2408086061477661, | |
| "learning_rate": 3.493270283607522e-05, | |
| "loss": 2.8455, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.15472415979708307, | |
| "grad_norm": 1.1654704809188843, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 2.8552, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.1553582752060875, | |
| "grad_norm": 1.2121531963348389, | |
| "learning_rate": 3.4166600309926387e-05, | |
| "loss": 2.8707, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.15599239061509196, | |
| "grad_norm": 1.23041832447052, | |
| "learning_rate": 3.3785077745215873e-05, | |
| "loss": 2.8064, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.1566265060240964, | |
| "grad_norm": 1.237444281578064, | |
| "learning_rate": 3.340460734357359e-05, | |
| "loss": 2.5598, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.15726062143310082, | |
| "grad_norm": 1.2903790473937988, | |
| "learning_rate": 3.3025213793178646e-05, | |
| "loss": 2.6563, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.15789473684210525, | |
| "grad_norm": 1.252323865890503, | |
| "learning_rate": 3.264692171233485e-05, | |
| "loss": 2.4976, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.15852885225110971, | |
| "grad_norm": 1.4426056146621704, | |
| "learning_rate": 3.226975564787322e-05, | |
| "loss": 2.5239, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.15852885225110971, | |
| "eval_loss": 0.7680420875549316, | |
| "eval_runtime": 299.1281, | |
| "eval_samples_per_second": 8.879, | |
| "eval_steps_per_second": 2.22, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.15916296766011415, | |
| "grad_norm": 1.0694372653961182, | |
| "learning_rate": 3.189374007355917e-05, | |
| "loss": 3.8561, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.15979708306911858, | |
| "grad_norm": 1.0747451782226562, | |
| "learning_rate": 3.151889938850445e-05, | |
| "loss": 3.1765, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.160431198478123, | |
| "grad_norm": 1.0591131448745728, | |
| "learning_rate": 3.114525791558398e-05, | |
| "loss": 3.2664, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.16106531388712747, | |
| "grad_norm": 0.950308620929718, | |
| "learning_rate": 3.0772839899857464e-05, | |
| "loss": 3.0991, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.1616994292961319, | |
| "grad_norm": 0.9497650265693665, | |
| "learning_rate": 3.0401669506996256e-05, | |
| "loss": 3.3147, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.16233354470513633, | |
| "grad_norm": 0.9711596369743347, | |
| "learning_rate": 3.003177082171523e-05, | |
| "loss": 3.3452, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.16296766011414077, | |
| "grad_norm": 0.9367339015007019, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 3.1852, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.1636017755231452, | |
| "grad_norm": 0.9764601588249207, | |
| "learning_rate": 2.9295884498599414e-05, | |
| "loss": 3.2691, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.16423589093214966, | |
| "grad_norm": 0.958132266998291, | |
| "learning_rate": 2.8929944611373554e-05, | |
| "loss": 3.2167, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.1648700063411541, | |
| "grad_norm": 1.0170295238494873, | |
| "learning_rate": 2.8565371929847284e-05, | |
| "loss": 3.2041, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.16550412175015852, | |
| "grad_norm": 0.997759222984314, | |
| "learning_rate": 2.8202190110619493e-05, | |
| "loss": 3.2763, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.16613823715916295, | |
| "grad_norm": 1.0056345462799072, | |
| "learning_rate": 2.784042272003794e-05, | |
| "loss": 3.263, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.16677235256816741, | |
| "grad_norm": 0.9892311692237854, | |
| "learning_rate": 2.7480093232670158e-05, | |
| "loss": 3.0584, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.16740646797717185, | |
| "grad_norm": 0.9892306923866272, | |
| "learning_rate": 2.712122502978024e-05, | |
| "loss": 3.2246, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.16804058338617628, | |
| "grad_norm": 1.1950269937515259, | |
| "learning_rate": 2.6763841397811573e-05, | |
| "loss": 3.2072, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.1686746987951807, | |
| "grad_norm": 1.0062364339828491, | |
| "learning_rate": 2.64079655268759e-05, | |
| "loss": 3.2419, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.16930881420418517, | |
| "grad_norm": 0.9735838770866394, | |
| "learning_rate": 2.605362050924848e-05, | |
| "loss": 3.1027, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.1699429296131896, | |
| "grad_norm": 1.0225987434387207, | |
| "learning_rate": 2.57008293378697e-05, | |
| "loss": 3.1776, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.17057704502219403, | |
| "grad_norm": 1.0058813095092773, | |
| "learning_rate": 2.534961490485313e-05, | |
| "loss": 3.255, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.17121116043119847, | |
| "grad_norm": 0.9906468391418457, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 3.1952, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.17184527584020293, | |
| "grad_norm": 0.985464870929718, | |
| "learning_rate": 2.4652007309320498e-05, | |
| "loss": 3.1027, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.17247939124920736, | |
| "grad_norm": 0.9994264841079712, | |
| "learning_rate": 2.430565941356157e-05, | |
| "loss": 3.1101, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.1731135066582118, | |
| "grad_norm": 0.9816799759864807, | |
| "learning_rate": 2.3960978786741877e-05, | |
| "loss": 3.1194, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.17374762206721622, | |
| "grad_norm": 1.069852590560913, | |
| "learning_rate": 2.361798779469336e-05, | |
| "loss": 3.1427, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.17438173747622068, | |
| "grad_norm": 1.0157891511917114, | |
| "learning_rate": 2.3276708693609943e-05, | |
| "loss": 3.0912, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.17501585288522511, | |
| "grad_norm": 1.4053800106048584, | |
| "learning_rate": 2.2937163628603435e-05, | |
| "loss": 3.0539, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.17564996829422955, | |
| "grad_norm": 0.9908464550971985, | |
| "learning_rate": 2.259937463226651e-05, | |
| "loss": 3.0635, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.17628408370323398, | |
| "grad_norm": 1.0277864933013916, | |
| "learning_rate": 2.2263363623243054e-05, | |
| "loss": 3.0571, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.17691819911223844, | |
| "grad_norm": 1.0323623418807983, | |
| "learning_rate": 2.192915240480596e-05, | |
| "loss": 3.1435, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.17755231452124287, | |
| "grad_norm": 1.00318443775177, | |
| "learning_rate": 2.1596762663442218e-05, | |
| "loss": 2.9892, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.1781864299302473, | |
| "grad_norm": 1.02731454372406, | |
| "learning_rate": 2.1266215967445824e-05, | |
| "loss": 3.0905, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.17882054533925174, | |
| "grad_norm": 1.0222443342208862, | |
| "learning_rate": 2.0937533765518187e-05, | |
| "loss": 3.1027, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.1794546607482562, | |
| "grad_norm": 1.115246057510376, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 3.0372, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.18008877615726063, | |
| "grad_norm": 1.0481406450271606, | |
| "learning_rate": 2.0285848032369137e-05, | |
| "loss": 3.0176, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.18072289156626506, | |
| "grad_norm": 1.0405553579330444, | |
| "learning_rate": 1.996288678810105e-05, | |
| "loss": 2.92, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.1813570069752695, | |
| "grad_norm": 1.0391370058059692, | |
| "learning_rate": 1.9641874609064443e-05, | |
| "loss": 2.9469, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.18199112238427395, | |
| "grad_norm": 1.0537941455841064, | |
| "learning_rate": 1.932283232527956e-05, | |
| "loss": 3.0066, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.18262523779327838, | |
| "grad_norm": 1.1033146381378174, | |
| "learning_rate": 1.9005780638942982e-05, | |
| "loss": 3.1014, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.18325935320228282, | |
| "grad_norm": 1.0815560817718506, | |
| "learning_rate": 1.8690740123084316e-05, | |
| "loss": 2.9473, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.18389346861128725, | |
| "grad_norm": 1.1662791967391968, | |
| "learning_rate": 1.837773122023114e-05, | |
| "loss": 3.0494, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.18452758402029168, | |
| "grad_norm": 1.1125469207763672, | |
| "learning_rate": 1.8066774241082612e-05, | |
| "loss": 3.1307, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.18516169942929614, | |
| "grad_norm": 1.1446367502212524, | |
| "learning_rate": 1.7757889363191483e-05, | |
| "loss": 2.9742, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.18579581483830057, | |
| "grad_norm": 1.1044974327087402, | |
| "learning_rate": 1.745109662965481e-05, | |
| "loss": 2.8596, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.186429930247305, | |
| "grad_norm": 1.292734980583191, | |
| "learning_rate": 1.714641594781347e-05, | |
| "loss": 2.9186, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.18706404565630944, | |
| "grad_norm": 1.1280649900436401, | |
| "learning_rate": 1.684386708796025e-05, | |
| "loss": 2.7486, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.1876981610653139, | |
| "grad_norm": 1.157607913017273, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 2.6953, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.18833227647431833, | |
| "grad_norm": 1.1997601985931396, | |
| "learning_rate": 1.62452432224612e-05, | |
| "loss": 2.5338, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.18896639188332276, | |
| "grad_norm": 1.5699822902679443, | |
| "learning_rate": 1.5949207060660138e-05, | |
| "loss": 2.728, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.1896005072923272, | |
| "grad_norm": 1.2640992403030396, | |
| "learning_rate": 1.5655380406016235e-05, | |
| "loss": 2.4079, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.19023462270133165, | |
| "grad_norm": 1.4184346199035645, | |
| "learning_rate": 1.536378232452003e-05, | |
| "loss": 2.5856, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.19023462270133165, | |
| "eval_loss": 0.7586191296577454, | |
| "eval_runtime": 299.2736, | |
| "eval_samples_per_second": 8.875, | |
| "eval_steps_per_second": 2.219, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.19086873811033608, | |
| "grad_norm": 1.0150848627090454, | |
| "learning_rate": 1.5074431737553157e-05, | |
| "loss": 3.2563, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.19150285351934052, | |
| "grad_norm": 1.2016209363937378, | |
| "learning_rate": 1.4787347420660541e-05, | |
| "loss": 3.2713, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.19213696892834495, | |
| "grad_norm": 1.01749849319458, | |
| "learning_rate": 1.4502548002332088e-05, | |
| "loss": 3.182, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.1927710843373494, | |
| "grad_norm": 0.988182783126831, | |
| "learning_rate": 1.422005196279395e-05, | |
| "loss": 3.1752, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.19340519974635384, | |
| "grad_norm": 0.9728803038597107, | |
| "learning_rate": 1.3939877632809278e-05, | |
| "loss": 3.2968, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.19403931515535827, | |
| "grad_norm": 0.935308039188385, | |
| "learning_rate": 1.3662043192488849e-05, | |
| "loss": 3.0506, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.1946734305643627, | |
| "grad_norm": 0.9322471618652344, | |
| "learning_rate": 1.338656667011134e-05, | |
| "loss": 3.2073, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.19530754597336716, | |
| "grad_norm": 0.9291597008705139, | |
| "learning_rate": 1.3113465940953495e-05, | |
| "loss": 3.1834, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.1959416613823716, | |
| "grad_norm": 0.927598774433136, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 3.1462, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.19657577679137603, | |
| "grad_norm": 0.9430285096168518, | |
| "learning_rate": 1.257446259144494e-05, | |
| "loss": 3.1203, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.19720989220038046, | |
| "grad_norm": 1.4846646785736084, | |
| "learning_rate": 1.2308594946249163e-05, | |
| "loss": 3.1325, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.19784400760938492, | |
| "grad_norm": 0.9747286438941956, | |
| "learning_rate": 1.204517304231343e-05, | |
| "loss": 3.0351, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.19847812301838935, | |
| "grad_norm": 0.9820675849914551, | |
| "learning_rate": 1.178421397270758e-05, | |
| "loss": 3.1296, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.19911223842739378, | |
| "grad_norm": 0.9933830499649048, | |
| "learning_rate": 1.1525734670691701e-05, | |
| "loss": 3.1196, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.19974635383639822, | |
| "grad_norm": 0.9653772711753845, | |
| "learning_rate": 1.1269751908617277e-05, | |
| "loss": 3.0227, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.20038046924540268, | |
| "grad_norm": 1.0138531923294067, | |
| "learning_rate": 1.1016282296838887e-05, | |
| "loss": 3.0659, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.2010145846544071, | |
| "grad_norm": 0.9920913577079773, | |
| "learning_rate": 1.0765342282636416e-05, | |
| "loss": 3.1804, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.20164870006341154, | |
| "grad_norm": 0.9812908172607422, | |
| "learning_rate": 1.0516948149147754e-05, | |
| "loss": 3.1213, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.20228281547241597, | |
| "grad_norm": 1.0075311660766602, | |
| "learning_rate": 1.0271116014312293e-05, | |
| "loss": 3.1012, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.20291693088142043, | |
| "grad_norm": 1.024266242980957, | |
| "learning_rate": 1.0027861829824952e-05, | |
| "loss": 3.1247, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.20355104629042486, | |
| "grad_norm": 0.9915530681610107, | |
| "learning_rate": 9.787201380101157e-06, | |
| "loss": 3.1249, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.2041851616994293, | |
| "grad_norm": 1.0081713199615479, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 3.044, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.20481927710843373, | |
| "grad_norm": 1.0550603866577148, | |
| "learning_rate": 9.313723980074018e-06, | |
| "loss": 3.2099, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.20545339251743816, | |
| "grad_norm": 1.0586062669754028, | |
| "learning_rate": 9.080937753040646e-06, | |
| "loss": 3.2203, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.20608750792644262, | |
| "grad_norm": 0.9967835545539856, | |
| "learning_rate": 8.850806705317183e-06, | |
| "loss": 3.0307, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.20672162333544705, | |
| "grad_norm": 1.0282031297683716, | |
| "learning_rate": 8.623345769777514e-06, | |
| "loss": 3.0631, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.20735573874445148, | |
| "grad_norm": 0.9924097061157227, | |
| "learning_rate": 8.398569706035792e-06, | |
| "loss": 2.9257, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.20798985415345592, | |
| "grad_norm": 1.0178431272506714, | |
| "learning_rate": 8.176493099488663e-06, | |
| "loss": 3.0914, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.20862396956246038, | |
| "grad_norm": 1.0140167474746704, | |
| "learning_rate": 7.957130360368898e-06, | |
| "loss": 3.1761, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.2092580849714648, | |
| "grad_norm": 1.1238445043563843, | |
| "learning_rate": 7.740495722810271e-06, | |
| "loss": 2.9859, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.20989220038046924, | |
| "grad_norm": 1.0198321342468262, | |
| "learning_rate": 7.526603243923957e-06, | |
| "loss": 3.1234, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 1.0335346460342407, | |
| "learning_rate": 7.315466802886401e-06, | |
| "loss": 3.0115, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.21116043119847813, | |
| "grad_norm": 1.0279099941253662, | |
| "learning_rate": 7.107100100038671e-06, | |
| "loss": 3.0316, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.21179454660748256, | |
| "grad_norm": 1.0411081314086914, | |
| "learning_rate": 6.901516655997536e-06, | |
| "loss": 3.0541, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.212428662016487, | |
| "grad_norm": 1.0638669729232788, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 3.1495, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.21306277742549143, | |
| "grad_norm": 1.0645053386688232, | |
| "learning_rate": 6.498752722928042e-06, | |
| "loss": 3.13, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.2136968928344959, | |
| "grad_norm": 1.1958521604537964, | |
| "learning_rate": 6.301598368674105e-06, | |
| "loss": 2.9266, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.21433100824350032, | |
| "grad_norm": 1.0360051393508911, | |
| "learning_rate": 6.107279541079769e-06, | |
| "loss": 3.0758, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.21496512365250475, | |
| "grad_norm": 1.070273518562317, | |
| "learning_rate": 5.915808849215304e-06, | |
| "loss": 3.0503, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.21559923906150918, | |
| "grad_norm": 1.0638577938079834, | |
| "learning_rate": 5.727198717339511e-06, | |
| "loss": 2.9783, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.21623335447051364, | |
| "grad_norm": 1.0641688108444214, | |
| "learning_rate": 5.54146138409355e-06, | |
| "loss": 3.0227, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.21686746987951808, | |
| "grad_norm": 1.0784544944763184, | |
| "learning_rate": 5.358608901706802e-06, | |
| "loss": 2.9654, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.2175015852885225, | |
| "grad_norm": 1.1119335889816284, | |
| "learning_rate": 5.178653135214812e-06, | |
| "loss": 2.8726, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.21813570069752694, | |
| "grad_norm": 1.095287799835205, | |
| "learning_rate": 5.001605761689398e-06, | |
| "loss": 2.8824, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.2187698161065314, | |
| "grad_norm": 1.1402324438095093, | |
| "learning_rate": 4.827478269480895e-06, | |
| "loss": 2.8718, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.21940393151553583, | |
| "grad_norm": 1.1678264141082764, | |
| "learning_rate": 4.65628195747273e-06, | |
| "loss": 2.6562, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.22003804692454026, | |
| "grad_norm": 1.156401515007019, | |
| "learning_rate": 4.488027934348271e-06, | |
| "loss": 2.624, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.2206721623335447, | |
| "grad_norm": 1.2031807899475098, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 2.465, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.22130627774254916, | |
| "grad_norm": 1.2508282661437988, | |
| "learning_rate": 4.16039023417088e-06, | |
| "loss": 2.4946, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.2219403931515536, | |
| "grad_norm": 1.4422476291656494, | |
| "learning_rate": 4.001027817058789e-06, | |
| "loss": 2.5989, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.2219403931515536, | |
| "eval_loss": 0.7529736161231995, | |
| "eval_runtime": 298.6761, | |
| "eval_samples_per_second": 8.893, | |
| "eval_steps_per_second": 2.223, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.22257450856055802, | |
| "grad_norm": 0.920236349105835, | |
| "learning_rate": 3.844650207332562e-06, | |
| "loss": 3.2697, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.22320862396956245, | |
| "grad_norm": 0.9723142385482788, | |
| "learning_rate": 3.691267552111183e-06, | |
| "loss": 3.1764, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.2238427393785669, | |
| "grad_norm": 0.9397497177124023, | |
| "learning_rate": 3.54088980417534e-06, | |
| "loss": 3.067, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.22447685478757134, | |
| "grad_norm": 0.9813097715377808, | |
| "learning_rate": 3.393526721321616e-06, | |
| "loss": 3.319, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.22511097019657578, | |
| "grad_norm": 0.9807033538818359, | |
| "learning_rate": 3.249187865729264e-06, | |
| "loss": 3.1667, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.2257450856055802, | |
| "grad_norm": 0.9774824976921082, | |
| "learning_rate": 3.1078826033397843e-06, | |
| "loss": 3.1495, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.22637920101458464, | |
| "grad_norm": 0.9869019389152527, | |
| "learning_rate": 2.9696201032491434e-06, | |
| "loss": 3.1348, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.2270133164235891, | |
| "grad_norm": 0.9642476439476013, | |
| "learning_rate": 2.8344093371128424e-06, | |
| "loss": 3.0795, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.22764743183259353, | |
| "grad_norm": 0.9823551177978516, | |
| "learning_rate": 2.70225907856374e-06, | |
| "loss": 3.21, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.22828154724159797, | |
| "grad_norm": 0.9760318398475647, | |
| "learning_rate": 2.573177902642726e-06, | |
| "loss": 3.1709, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.2289156626506024, | |
| "grad_norm": 0.947401762008667, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 3.1004, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.22954977805960686, | |
| "grad_norm": 0.9643566012382507, | |
| "learning_rate": 2.324256102563188e-06, | |
| "loss": 3.1597, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.2301838934686113, | |
| "grad_norm": 0.970553994178772, | |
| "learning_rate": 2.204431630583548e-06, | |
| "loss": 3.0536, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.23081800887761572, | |
| "grad_norm": 0.9732277393341064, | |
| "learning_rate": 2.087708544541689e-06, | |
| "loss": 3.1504, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.23145212428662015, | |
| "grad_norm": 0.9547330141067505, | |
| "learning_rate": 1.974094418431388e-06, | |
| "loss": 3.0721, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.2320862396956246, | |
| "grad_norm": 0.9943172931671143, | |
| "learning_rate": 1.8635966245104664e-06, | |
| "loss": 3.1663, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.23272035510462905, | |
| "grad_norm": 0.9682246446609497, | |
| "learning_rate": 1.7562223328224325e-06, | |
| "loss": 3.0412, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.23335447051363348, | |
| "grad_norm": 0.9810298085212708, | |
| "learning_rate": 1.6519785107311891e-06, | |
| "loss": 3.219, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.2339885859226379, | |
| "grad_norm": 0.972686767578125, | |
| "learning_rate": 1.5508719224689717e-06, | |
| "loss": 3.0453, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.23462270133164237, | |
| "grad_norm": 0.9679805040359497, | |
| "learning_rate": 1.4529091286973995e-06, | |
| "loss": 3.0846, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.2352568167406468, | |
| "grad_norm": 1.3851791620254517, | |
| "learning_rate": 1.358096486081778e-06, | |
| "loss": 3.0721, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.23589093214965123, | |
| "grad_norm": 0.9714539647102356, | |
| "learning_rate": 1.2664401468786114e-06, | |
| "loss": 3.057, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.23652504755865567, | |
| "grad_norm": 1.0152660608291626, | |
| "learning_rate": 1.1779460585363944e-06, | |
| "loss": 3.1163, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.23715916296766013, | |
| "grad_norm": 0.9702284932136536, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 2.9209, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.23779327837666456, | |
| "grad_norm": 1.0239858627319336, | |
| "learning_rate": 1.0104673978866164e-06, | |
| "loss": 3.0793, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.238427393785669, | |
| "grad_norm": 1.0014803409576416, | |
| "learning_rate": 9.314936930293283e-07, | |
| "loss": 3.1433, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.23906150919467342, | |
| "grad_norm": 1.0843709707260132, | |
| "learning_rate": 8.557039732283944e-07, | |
| "loss": 3.0487, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.23969562460367788, | |
| "grad_norm": 1.011506199836731, | |
| "learning_rate": 7.83103156370113e-07, | |
| "loss": 2.96, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.2403297400126823, | |
| "grad_norm": 1.371131420135498, | |
| "learning_rate": 7.136959534174592e-07, | |
| "loss": 3.0234, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.24096385542168675, | |
| "grad_norm": 1.011451005935669, | |
| "learning_rate": 6.474868681043578e-07, | |
| "loss": 3.1524, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.24159797083069118, | |
| "grad_norm": 1.0045251846313477, | |
| "learning_rate": 5.844801966434832e-07, | |
| "loss": 3.0862, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.24223208623969564, | |
| "grad_norm": 1.0423835515975952, | |
| "learning_rate": 5.246800274474439e-07, | |
| "loss": 3.1457, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.24286620164870007, | |
| "grad_norm": 1.0296357870101929, | |
| "learning_rate": 4.680902408635335e-07, | |
| "loss": 3.0892, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.2435003170577045, | |
| "grad_norm": 1.0572096109390259, | |
| "learning_rate": 4.1471450892189846e-07, | |
| "loss": 3.1505, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.24413443246670893, | |
| "grad_norm": 1.027784824371338, | |
| "learning_rate": 3.6455629509730136e-07, | |
| "loss": 3.0082, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.2447685478757134, | |
| "grad_norm": 1.03461754322052, | |
| "learning_rate": 3.1761885408435054e-07, | |
| "loss": 3.0415, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.24540266328471783, | |
| "grad_norm": 1.0629018545150757, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 3.1103, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.24603677869372226, | |
| "grad_norm": 1.0713993310928345, | |
| "learning_rate": 2.334182641175686e-07, | |
| "loss": 3.1734, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.2466708941027267, | |
| "grad_norm": 1.0982544422149658, | |
| "learning_rate": 1.9616057881935436e-07, | |
| "loss": 2.9949, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.24730500951173112, | |
| "grad_norm": 1.064578890800476, | |
| "learning_rate": 1.6213459328950352e-07, | |
| "loss": 2.8567, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.24793912492073558, | |
| "grad_norm": 1.0758435726165771, | |
| "learning_rate": 1.3134251542544774e-07, | |
| "loss": 2.9331, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.24857324032974001, | |
| "grad_norm": 1.1049171686172485, | |
| "learning_rate": 1.0378634328099269e-07, | |
| "loss": 2.8457, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.24920735573874445, | |
| "grad_norm": 1.0962225198745728, | |
| "learning_rate": 7.946786493666647e-08, | |
| "loss": 2.8366, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.24984147114774888, | |
| "grad_norm": 1.1087645292282104, | |
| "learning_rate": 5.838865838366792e-08, | |
| "loss": 2.6736, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.25047558655675334, | |
| "grad_norm": 1.1320421695709229, | |
| "learning_rate": 4.055009142152067e-08, | |
| "loss": 2.5256, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.25110970196575777, | |
| "grad_norm": 1.1543631553649902, | |
| "learning_rate": 2.595332156925534e-08, | |
| "loss": 2.6751, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.2517438173747622, | |
| "grad_norm": 1.2271840572357178, | |
| "learning_rate": 1.4599295990352924e-08, | |
| "loss": 2.4729, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.25237793278376663, | |
| "grad_norm": 1.232959270477295, | |
| "learning_rate": 6.488751431266149e-09, | |
| "loss": 2.5886, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.25301204819277107, | |
| "grad_norm": 1.39299476146698, | |
| "learning_rate": 1.622214173602199e-09, | |
| "loss": 2.5761, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.2536461636017755, | |
| "grad_norm": 1.4202094078063965, | |
| "learning_rate": 0.0, | |
| "loss": 2.5205, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.2536461636017755, | |
| "eval_loss": 0.7515735626220703, | |
| "eval_runtime": 299.3765, | |
| "eval_samples_per_second": 8.872, | |
| "eval_steps_per_second": 2.218, | |
| "step": 400 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.35326756330537e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |