| { | |
| "best_metric": 1.1534959077835083, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-350", | |
| "epoch": 0.47538200339558573, | |
| "eval_steps": 50, | |
| "global_step": 350, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0013582342954159593, | |
| "grad_norm": 0.7695918679237366, | |
| "learning_rate": 1e-05, | |
| "loss": 1.3733, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0013582342954159593, | |
| "eval_loss": 1.4100896120071411, | |
| "eval_runtime": 74.7293, | |
| "eval_samples_per_second": 16.593, | |
| "eval_steps_per_second": 4.148, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0027164685908319186, | |
| "grad_norm": 0.7374516129493713, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4754, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0040747028862478775, | |
| "grad_norm": 0.7155938148498535, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3174, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.005432937181663837, | |
| "grad_norm": 0.6224644184112549, | |
| "learning_rate": 4e-05, | |
| "loss": 1.3707, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.006791171477079796, | |
| "grad_norm": 0.4430236220359802, | |
| "learning_rate": 5e-05, | |
| "loss": 1.3729, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.008149405772495755, | |
| "grad_norm": 0.40971505641937256, | |
| "learning_rate": 6e-05, | |
| "loss": 1.3221, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.009507640067911714, | |
| "grad_norm": 0.3615075945854187, | |
| "learning_rate": 7e-05, | |
| "loss": 1.3454, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.010865874363327675, | |
| "grad_norm": 0.35050615668296814, | |
| "learning_rate": 8e-05, | |
| "loss": 1.309, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.012224108658743633, | |
| "grad_norm": 0.30065447092056274, | |
| "learning_rate": 9e-05, | |
| "loss": 1.2385, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.013582342954159592, | |
| "grad_norm": 0.3258337378501892, | |
| "learning_rate": 0.0001, | |
| "loss": 1.2086, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.014940577249575551, | |
| "grad_norm": 0.3569473326206207, | |
| "learning_rate": 9.99983777858264e-05, | |
| "loss": 1.3262, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.01629881154499151, | |
| "grad_norm": 0.37945055961608887, | |
| "learning_rate": 9.999351124856874e-05, | |
| "loss": 1.3609, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.01765704584040747, | |
| "grad_norm": 0.3099724352359772, | |
| "learning_rate": 9.998540070400966e-05, | |
| "loss": 1.2673, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.019015280135823428, | |
| "grad_norm": 0.31235983967781067, | |
| "learning_rate": 9.997404667843075e-05, | |
| "loss": 1.2393, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.02037351443123939, | |
| "grad_norm": 0.27202108502388, | |
| "learning_rate": 9.995944990857849e-05, | |
| "loss": 1.2544, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.02173174872665535, | |
| "grad_norm": 0.27280139923095703, | |
| "learning_rate": 9.994161134161634e-05, | |
| "loss": 1.2061, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.023089983022071308, | |
| "grad_norm": 0.293769896030426, | |
| "learning_rate": 9.992053213506334e-05, | |
| "loss": 1.1964, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.024448217317487267, | |
| "grad_norm": 0.28996336460113525, | |
| "learning_rate": 9.989621365671902e-05, | |
| "loss": 1.2483, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.025806451612903226, | |
| "grad_norm": 0.271930456161499, | |
| "learning_rate": 9.986865748457457e-05, | |
| "loss": 1.2141, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.027164685908319185, | |
| "grad_norm": 0.2686532735824585, | |
| "learning_rate": 9.983786540671051e-05, | |
| "loss": 1.2347, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.028522920203735144, | |
| "grad_norm": 0.2654222249984741, | |
| "learning_rate": 9.980383942118066e-05, | |
| "loss": 1.2133, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.029881154499151102, | |
| "grad_norm": 0.2612136900424957, | |
| "learning_rate": 9.976658173588244e-05, | |
| "loss": 1.2274, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.03123938879456706, | |
| "grad_norm": 0.24660049378871918, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 1.2137, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.03259762308998302, | |
| "grad_norm": 0.2399974763393402, | |
| "learning_rate": 9.968238114591566e-05, | |
| "loss": 1.1527, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.03395585738539898, | |
| "grad_norm": 0.24953867495059967, | |
| "learning_rate": 9.96354437049027e-05, | |
| "loss": 1.2337, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.03531409168081494, | |
| "grad_norm": 0.25382018089294434, | |
| "learning_rate": 9.95852854910781e-05, | |
| "loss": 1.2282, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.0366723259762309, | |
| "grad_norm": 0.26546919345855713, | |
| "learning_rate": 9.953190975913647e-05, | |
| "loss": 1.2031, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.038030560271646856, | |
| "grad_norm": 0.2536904513835907, | |
| "learning_rate": 9.947531997255256e-05, | |
| "loss": 1.218, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.03938879456706282, | |
| "grad_norm": 0.2628403902053833, | |
| "learning_rate": 9.941551980335652e-05, | |
| "loss": 1.1947, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.04074702886247878, | |
| "grad_norm": 0.26902860403060913, | |
| "learning_rate": 9.935251313189564e-05, | |
| "loss": 1.2258, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.042105263157894736, | |
| "grad_norm": 0.264118492603302, | |
| "learning_rate": 9.928630404658255e-05, | |
| "loss": 1.2215, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.0434634974533107, | |
| "grad_norm": 0.2503666579723358, | |
| "learning_rate": 9.921689684362989e-05, | |
| "loss": 1.2607, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.044821731748726654, | |
| "grad_norm": 0.2590126693248749, | |
| "learning_rate": 9.914429602677162e-05, | |
| "loss": 1.236, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.046179966044142616, | |
| "grad_norm": 0.2683681845664978, | |
| "learning_rate": 9.906850630697068e-05, | |
| "loss": 1.195, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.04753820033955857, | |
| "grad_norm": 0.2629988491535187, | |
| "learning_rate": 9.898953260211338e-05, | |
| "loss": 1.1939, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.048896434634974534, | |
| "grad_norm": 0.28124624490737915, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 1.277, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.05025466893039049, | |
| "grad_norm": 0.27577537298202515, | |
| "learning_rate": 9.882205394146361e-05, | |
| "loss": 1.1631, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.05161290322580645, | |
| "grad_norm": 0.26586708426475525, | |
| "learning_rate": 9.87335598531214e-05, | |
| "loss": 1.1817, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.052971137521222414, | |
| "grad_norm": 0.28580763936042786, | |
| "learning_rate": 9.864190351391822e-05, | |
| "loss": 1.2785, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.05432937181663837, | |
| "grad_norm": 0.28752267360687256, | |
| "learning_rate": 9.85470908713026e-05, | |
| "loss": 1.1675, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.05568760611205433, | |
| "grad_norm": 0.2884860634803772, | |
| "learning_rate": 9.844912807753104e-05, | |
| "loss": 1.1855, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.05704584040747029, | |
| "grad_norm": 0.2822176218032837, | |
| "learning_rate": 9.834802148926882e-05, | |
| "loss": 1.2276, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.05840407470288625, | |
| "grad_norm": 0.28766173124313354, | |
| "learning_rate": 9.824377766717759e-05, | |
| "loss": 1.1655, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.059762308998302205, | |
| "grad_norm": 0.30563387274742126, | |
| "learning_rate": 9.813640337548954e-05, | |
| "loss": 1.2651, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.06112054329371817, | |
| "grad_norm": 0.3028828799724579, | |
| "learning_rate": 9.802590558156862e-05, | |
| "loss": 1.1419, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.06247877758913412, | |
| "grad_norm": 0.3264331817626953, | |
| "learning_rate": 9.791229145545831e-05, | |
| "loss": 1.1629, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.06383701188455009, | |
| "grad_norm": 0.3301156163215637, | |
| "learning_rate": 9.779556836941645e-05, | |
| "loss": 1.2186, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.06519524617996604, | |
| "grad_norm": 0.3410678505897522, | |
| "learning_rate": 9.767574389743682e-05, | |
| "loss": 1.2162, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.06655348047538201, | |
| "grad_norm": 0.36292213201522827, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 1.1575, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.06791171477079797, | |
| "grad_norm": 0.5214879512786865, | |
| "learning_rate": 9.742682209735727e-05, | |
| "loss": 1.1233, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.06791171477079797, | |
| "eval_loss": 1.2765874862670898, | |
| "eval_runtime": 75.9116, | |
| "eval_samples_per_second": 16.335, | |
| "eval_steps_per_second": 4.084, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.06926994906621392, | |
| "grad_norm": 0.7125508189201355, | |
| "learning_rate": 9.729774092143627e-05, | |
| "loss": 1.3574, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.07062818336162988, | |
| "grad_norm": 0.5578199028968811, | |
| "learning_rate": 9.716559066288715e-05, | |
| "loss": 1.2801, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.07198641765704585, | |
| "grad_norm": 0.4072265923023224, | |
| "learning_rate": 9.703037989675087e-05, | |
| "loss": 1.1512, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.0733446519524618, | |
| "grad_norm": 0.2695824205875397, | |
| "learning_rate": 9.689211739666023e-05, | |
| "loss": 1.2332, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.07470288624787776, | |
| "grad_norm": 0.2340139001607895, | |
| "learning_rate": 9.675081213427076e-05, | |
| "loss": 1.1724, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.07606112054329371, | |
| "grad_norm": 0.23189423978328705, | |
| "learning_rate": 9.66064732786784e-05, | |
| "loss": 1.1846, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.07741935483870968, | |
| "grad_norm": 0.2402050495147705, | |
| "learning_rate": 9.645911019582467e-05, | |
| "loss": 1.207, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.07877758913412564, | |
| "grad_norm": 0.25576063990592957, | |
| "learning_rate": 9.630873244788883e-05, | |
| "loss": 1.2574, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.08013582342954159, | |
| "grad_norm": 0.25986823439598083, | |
| "learning_rate": 9.615534979266745e-05, | |
| "loss": 1.2114, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.08149405772495756, | |
| "grad_norm": 0.250826358795166, | |
| "learning_rate": 9.599897218294122e-05, | |
| "loss": 1.1968, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.08285229202037352, | |
| "grad_norm": 0.25176748633384705, | |
| "learning_rate": 9.583960976582913e-05, | |
| "loss": 1.1929, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.08421052631578947, | |
| "grad_norm": 0.2325923889875412, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 1.2047, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.08556876061120543, | |
| "grad_norm": 0.22908318042755127, | |
| "learning_rate": 9.551197206565173e-05, | |
| "loss": 1.1473, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.0869269949066214, | |
| "grad_norm": 0.21206900477409363, | |
| "learning_rate": 9.534371804252728e-05, | |
| "loss": 1.1855, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.08828522920203735, | |
| "grad_norm": 0.22555455565452576, | |
| "learning_rate": 9.517252173051911e-05, | |
| "loss": 1.1698, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.08964346349745331, | |
| "grad_norm": 0.23330341279506683, | |
| "learning_rate": 9.49983942383106e-05, | |
| "loss": 1.1946, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.09100169779286928, | |
| "grad_norm": 0.24638806283473969, | |
| "learning_rate": 9.482134686478519e-05, | |
| "loss": 1.209, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.09235993208828523, | |
| "grad_norm": 0.2484908103942871, | |
| "learning_rate": 9.464139109829321e-05, | |
| "loss": 1.1865, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.09371816638370119, | |
| "grad_norm": 0.23545753955841064, | |
| "learning_rate": 9.445853861590647e-05, | |
| "loss": 1.2447, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.09507640067911714, | |
| "grad_norm": 0.22677823901176453, | |
| "learning_rate": 9.42728012826605e-05, | |
| "loss": 1.1572, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.09643463497453311, | |
| "grad_norm": 0.22840863466262817, | |
| "learning_rate": 9.408419115078471e-05, | |
| "loss": 1.2223, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.09779286926994907, | |
| "grad_norm": 0.23291656374931335, | |
| "learning_rate": 9.389272045892024e-05, | |
| "loss": 1.2053, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.09915110356536502, | |
| "grad_norm": 0.24530543386936188, | |
| "learning_rate": 9.36984016313259e-05, | |
| "loss": 1.2103, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.10050933786078098, | |
| "grad_norm": 0.24790343642234802, | |
| "learning_rate": 9.350124727707197e-05, | |
| "loss": 1.1125, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.10186757215619695, | |
| "grad_norm": 0.23026898503303528, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 1.2442, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.1032258064516129, | |
| "grad_norm": 0.24416302144527435, | |
| "learning_rate": 9.309848334400246e-05, | |
| "loss": 1.1953, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.10458404074702886, | |
| "grad_norm": 0.22965535521507263, | |
| "learning_rate": 9.289289989996133e-05, | |
| "loss": 1.2232, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.10594227504244483, | |
| "grad_norm": 0.2312672883272171, | |
| "learning_rate": 9.268453319711363e-05, | |
| "loss": 1.1542, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.10730050933786078, | |
| "grad_norm": 0.2299879491329193, | |
| "learning_rate": 9.247339675607605e-05, | |
| "loss": 1.209, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.10865874363327674, | |
| "grad_norm": 0.23572081327438354, | |
| "learning_rate": 9.225950427718975e-05, | |
| "loss": 1.2257, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1100169779286927, | |
| "grad_norm": 0.23509450256824493, | |
| "learning_rate": 9.204286963963111e-05, | |
| "loss": 1.1656, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.11137521222410866, | |
| "grad_norm": 0.2482880800962448, | |
| "learning_rate": 9.182350690051133e-05, | |
| "loss": 1.1448, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.11273344651952462, | |
| "grad_norm": 0.24817270040512085, | |
| "learning_rate": 9.160143029396422e-05, | |
| "loss": 1.2322, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.11409168081494057, | |
| "grad_norm": 0.24748308956623077, | |
| "learning_rate": 9.13766542302225e-05, | |
| "loss": 1.1227, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.11544991511035653, | |
| "grad_norm": 0.25305241346359253, | |
| "learning_rate": 9.114919329468282e-05, | |
| "loss": 1.1337, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1168081494057725, | |
| "grad_norm": 0.24612681567668915, | |
| "learning_rate": 9.091906224695935e-05, | |
| "loss": 1.114, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.11816638370118845, | |
| "grad_norm": 0.28515782952308655, | |
| "learning_rate": 9.068627601992598e-05, | |
| "loss": 1.1734, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.11952461799660441, | |
| "grad_norm": 0.2633991539478302, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 1.1993, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.12088285229202038, | |
| "grad_norm": 0.26582807302474976, | |
| "learning_rate": 9.021279861989885e-05, | |
| "loss": 1.1467, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.12224108658743633, | |
| "grad_norm": 0.26425063610076904, | |
| "learning_rate": 8.997213817017507e-05, | |
| "loss": 1.2081, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.12359932088285229, | |
| "grad_norm": 0.26553940773010254, | |
| "learning_rate": 8.972888398568772e-05, | |
| "loss": 1.1629, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.12495755517826825, | |
| "grad_norm": 0.2655147910118103, | |
| "learning_rate": 8.948305185085225e-05, | |
| "loss": 1.1801, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.12631578947368421, | |
| "grad_norm": 0.2920195162296295, | |
| "learning_rate": 8.92346577173636e-05, | |
| "loss": 1.1947, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.12767402376910017, | |
| "grad_norm": 0.2780720591545105, | |
| "learning_rate": 8.898371770316111e-05, | |
| "loss": 1.209, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.12903225806451613, | |
| "grad_norm": 0.27526527643203735, | |
| "learning_rate": 8.873024809138272e-05, | |
| "loss": 1.173, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.13039049235993208, | |
| "grad_norm": 0.29257383942604065, | |
| "learning_rate": 8.847426532930831e-05, | |
| "loss": 1.2088, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.13174872665534804, | |
| "grad_norm": 0.29788464307785034, | |
| "learning_rate": 8.821578602729242e-05, | |
| "loss": 1.135, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.13310696095076402, | |
| "grad_norm": 0.31753095984458923, | |
| "learning_rate": 8.795482695768658e-05, | |
| "loss": 1.1596, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.13446519524617997, | |
| "grad_norm": 0.37438032031059265, | |
| "learning_rate": 8.769140505375085e-05, | |
| "loss": 1.1788, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.13582342954159593, | |
| "grad_norm": 0.4673005938529968, | |
| "learning_rate": 8.742553740855506e-05, | |
| "loss": 1.2225, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.13582342954159593, | |
| "eval_loss": 1.2262212038040161, | |
| "eval_runtime": 75.8474, | |
| "eval_samples_per_second": 16.349, | |
| "eval_steps_per_second": 4.087, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.13718166383701189, | |
| "grad_norm": 0.5559870600700378, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 1.2997, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.13853989813242784, | |
| "grad_norm": 0.4555906057357788, | |
| "learning_rate": 8.688653405904652e-05, | |
| "loss": 1.301, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.1398981324278438, | |
| "grad_norm": 0.31413426995277405, | |
| "learning_rate": 8.661343332988869e-05, | |
| "loss": 1.2188, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.14125636672325975, | |
| "grad_norm": 0.2615854740142822, | |
| "learning_rate": 8.633795680751116e-05, | |
| "loss": 1.1284, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.14261460101867574, | |
| "grad_norm": 0.2208709567785263, | |
| "learning_rate": 8.606012236719073e-05, | |
| "loss": 1.2063, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.1439728353140917, | |
| "grad_norm": 0.20494621992111206, | |
| "learning_rate": 8.577994803720606e-05, | |
| "loss": 1.2157, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.14533106960950765, | |
| "grad_norm": 0.2158403843641281, | |
| "learning_rate": 8.549745199766792e-05, | |
| "loss": 1.1396, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.1466893039049236, | |
| "grad_norm": 0.22656506299972534, | |
| "learning_rate": 8.521265257933948e-05, | |
| "loss": 1.1591, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.14804753820033956, | |
| "grad_norm": 0.23627708852291107, | |
| "learning_rate": 8.492556826244687e-05, | |
| "loss": 1.1998, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.1494057724957555, | |
| "grad_norm": 0.23139552772045135, | |
| "learning_rate": 8.463621767547998e-05, | |
| "loss": 1.2165, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15076400679117147, | |
| "grad_norm": 0.237015500664711, | |
| "learning_rate": 8.434461959398376e-05, | |
| "loss": 1.1609, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.15212224108658742, | |
| "grad_norm": 0.2324032187461853, | |
| "learning_rate": 8.405079293933986e-05, | |
| "loss": 1.2107, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.1534804753820034, | |
| "grad_norm": 0.22056780755519867, | |
| "learning_rate": 8.375475677753881e-05, | |
| "loss": 1.1398, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.15483870967741936, | |
| "grad_norm": 0.21998582780361176, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 1.1788, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.15619694397283532, | |
| "grad_norm": 0.220359206199646, | |
| "learning_rate": 8.315613291203976e-05, | |
| "loss": 1.2086, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.15755517826825127, | |
| "grad_norm": 0.2111772894859314, | |
| "learning_rate": 8.285358405218655e-05, | |
| "loss": 1.1961, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.15891341256366723, | |
| "grad_norm": 0.21595050394535065, | |
| "learning_rate": 8.25489033703452e-05, | |
| "loss": 1.2173, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.16027164685908318, | |
| "grad_norm": 0.22538748383522034, | |
| "learning_rate": 8.224211063680853e-05, | |
| "loss": 1.1223, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.16162988115449914, | |
| "grad_norm": 0.22196736931800842, | |
| "learning_rate": 8.19332257589174e-05, | |
| "loss": 1.1796, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.16298811544991512, | |
| "grad_norm": 0.22292838990688324, | |
| "learning_rate": 8.162226877976887e-05, | |
| "loss": 1.1198, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.16434634974533108, | |
| "grad_norm": 0.22580811381340027, | |
| "learning_rate": 8.130925987691569e-05, | |
| "loss": 1.1997, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.16570458404074703, | |
| "grad_norm": 0.22446706891059875, | |
| "learning_rate": 8.099421936105702e-05, | |
| "loss": 1.0938, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.167062818336163, | |
| "grad_norm": 0.24063196778297424, | |
| "learning_rate": 8.067716767472045e-05, | |
| "loss": 1.272, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.16842105263157894, | |
| "grad_norm": 0.2342248260974884, | |
| "learning_rate": 8.035812539093557e-05, | |
| "loss": 1.1282, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.1697792869269949, | |
| "grad_norm": 0.23322924971580505, | |
| "learning_rate": 8.003711321189895e-05, | |
| "loss": 1.1774, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.17113752122241085, | |
| "grad_norm": 0.227143794298172, | |
| "learning_rate": 7.971415196763088e-05, | |
| "loss": 1.1614, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.17249575551782684, | |
| "grad_norm": 0.2375260442495346, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 1.0878, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.1738539898132428, | |
| "grad_norm": 0.23172912001609802, | |
| "learning_rate": 7.906246623448183e-05, | |
| "loss": 1.114, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.17521222410865875, | |
| "grad_norm": 0.2283349484205246, | |
| "learning_rate": 7.873378403255419e-05, | |
| "loss": 1.1243, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.1765704584040747, | |
| "grad_norm": 0.22897659242153168, | |
| "learning_rate": 7.840323733655778e-05, | |
| "loss": 1.1494, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.17792869269949066, | |
| "grad_norm": 0.2336939573287964, | |
| "learning_rate": 7.807084759519405e-05, | |
| "loss": 1.2519, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.17928692699490661, | |
| "grad_norm": 0.23826266825199127, | |
| "learning_rate": 7.773663637675694e-05, | |
| "loss": 1.1086, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.18064516129032257, | |
| "grad_norm": 0.2425377517938614, | |
| "learning_rate": 7.740062536773352e-05, | |
| "loss": 1.1654, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.18200339558573855, | |
| "grad_norm": 0.2372315227985382, | |
| "learning_rate": 7.706283637139658e-05, | |
| "loss": 1.1962, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.1833616298811545, | |
| "grad_norm": 0.24342741072177887, | |
| "learning_rate": 7.672329130639005e-05, | |
| "loss": 1.1181, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.18471986417657046, | |
| "grad_norm": 0.2434910535812378, | |
| "learning_rate": 7.638201220530665e-05, | |
| "loss": 1.2285, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.18607809847198642, | |
| "grad_norm": 0.24890004098415375, | |
| "learning_rate": 7.603902121325813e-05, | |
| "loss": 1.1714, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.18743633276740237, | |
| "grad_norm": 0.25163108110427856, | |
| "learning_rate": 7.569434058643844e-05, | |
| "loss": 1.1956, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.18879456706281833, | |
| "grad_norm": 0.24931834638118744, | |
| "learning_rate": 7.534799269067953e-05, | |
| "loss": 1.1483, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.19015280135823429, | |
| "grad_norm": 0.25025516748428345, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 1.2085, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.19151103565365024, | |
| "grad_norm": 0.2583712339401245, | |
| "learning_rate": 7.465038509514688e-05, | |
| "loss": 1.2139, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.19286926994906622, | |
| "grad_norm": 0.26586151123046875, | |
| "learning_rate": 7.42991706621303e-05, | |
| "loss": 1.2006, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.19422750424448218, | |
| "grad_norm": 0.2643159031867981, | |
| "learning_rate": 7.394637949075154e-05, | |
| "loss": 1.1739, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.19558573853989814, | |
| "grad_norm": 0.2751873731613159, | |
| "learning_rate": 7.35920344731241e-05, | |
| "loss": 1.2109, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.1969439728353141, | |
| "grad_norm": 0.2727619707584381, | |
| "learning_rate": 7.323615860218843e-05, | |
| "loss": 1.1851, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.19830220713073005, | |
| "grad_norm": 0.2831723690032959, | |
| "learning_rate": 7.287877497021978e-05, | |
| "loss": 1.1864, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.199660441426146, | |
| "grad_norm": 0.28427502512931824, | |
| "learning_rate": 7.251990676732984e-05, | |
| "loss": 1.1347, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.20101867572156196, | |
| "grad_norm": 0.2935657203197479, | |
| "learning_rate": 7.215957727996207e-05, | |
| "loss": 1.1435, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.20237691001697794, | |
| "grad_norm": 0.3436415195465088, | |
| "learning_rate": 7.179780988938051e-05, | |
| "loss": 1.182, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.2037351443123939, | |
| "grad_norm": 0.41627824306488037, | |
| "learning_rate": 7.143462807015271e-05, | |
| "loss": 1.2069, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2037351443123939, | |
| "eval_loss": 1.1833840608596802, | |
| "eval_runtime": 75.9037, | |
| "eval_samples_per_second": 16.336, | |
| "eval_steps_per_second": 4.084, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.20509337860780985, | |
| "grad_norm": 0.2699250876903534, | |
| "learning_rate": 7.107005538862646e-05, | |
| "loss": 1.2426, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.2064516129032258, | |
| "grad_norm": 0.29538774490356445, | |
| "learning_rate": 7.07041155014006e-05, | |
| "loss": 1.2144, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.20780984719864176, | |
| "grad_norm": 0.2714727222919464, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 1.2891, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.20916808149405772, | |
| "grad_norm": 0.26055291295051575, | |
| "learning_rate": 6.996822917828477e-05, | |
| "loss": 1.1651, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 0.2369328737258911, | |
| "learning_rate": 6.959833049300377e-05, | |
| "loss": 1.1572, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.21188455008488966, | |
| "grad_norm": 0.215769425034523, | |
| "learning_rate": 6.922716010014255e-05, | |
| "loss": 1.1521, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.2132427843803056, | |
| "grad_norm": 0.20645937323570251, | |
| "learning_rate": 6.885474208441603e-05, | |
| "loss": 1.1744, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.21460101867572157, | |
| "grad_norm": 0.21098513901233673, | |
| "learning_rate": 6.848110061149556e-05, | |
| "loss": 1.1768, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.21595925297113752, | |
| "grad_norm": 0.20801562070846558, | |
| "learning_rate": 6.810625992644085e-05, | |
| "loss": 1.1577, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.21731748726655348, | |
| "grad_norm": 0.22393718361854553, | |
| "learning_rate": 6.773024435212678e-05, | |
| "loss": 1.2199, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.21867572156196943, | |
| "grad_norm": 0.20644541084766388, | |
| "learning_rate": 6.735307828766515e-05, | |
| "loss": 1.1711, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.2200339558573854, | |
| "grad_norm": 0.21023766696453094, | |
| "learning_rate": 6.697478620682137e-05, | |
| "loss": 1.1736, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.22139219015280137, | |
| "grad_norm": 0.21561093628406525, | |
| "learning_rate": 6.659539265642643e-05, | |
| "loss": 1.1502, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.22275042444821733, | |
| "grad_norm": 0.2196631133556366, | |
| "learning_rate": 6.621492225478414e-05, | |
| "loss": 1.1813, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.22410865874363328, | |
| "grad_norm": 0.2159351408481598, | |
| "learning_rate": 6.583339969007363e-05, | |
| "loss": 1.1997, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.22546689303904924, | |
| "grad_norm": 0.215266153216362, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 1.0927, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.2268251273344652, | |
| "grad_norm": 0.22314059734344482, | |
| "learning_rate": 6.506729716392481e-05, | |
| "loss": 1.2044, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.22818336162988115, | |
| "grad_norm": 0.22275614738464355, | |
| "learning_rate": 6.468276691378155e-05, | |
| "loss": 1.2058, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.2295415959252971, | |
| "grad_norm": 0.22310510277748108, | |
| "learning_rate": 6.429728391993446e-05, | |
| "loss": 1.1506, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.23089983022071306, | |
| "grad_norm": 0.21441923081874847, | |
| "learning_rate": 6.391087319582264e-05, | |
| "loss": 1.1346, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.23225806451612904, | |
| "grad_norm": 0.2179337441921234, | |
| "learning_rate": 6.35235598150842e-05, | |
| "loss": 1.1355, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.233616298811545, | |
| "grad_norm": 0.2206660360097885, | |
| "learning_rate": 6.313536890992935e-05, | |
| "loss": 1.1346, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.23497453310696095, | |
| "grad_norm": 0.21749481558799744, | |
| "learning_rate": 6.274632566950967e-05, | |
| "loss": 1.1424, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.2363327674023769, | |
| "grad_norm": 0.21882691979408264, | |
| "learning_rate": 6.235645533828349e-05, | |
| "loss": 1.2118, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.23769100169779286, | |
| "grad_norm": 0.22239045798778534, | |
| "learning_rate": 6.19657832143779e-05, | |
| "loss": 1.1532, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.23904923599320882, | |
| "grad_norm": 0.22943776845932007, | |
| "learning_rate": 6.157433464794716e-05, | |
| "loss": 1.1658, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.24040747028862477, | |
| "grad_norm": 0.22224801778793335, | |
| "learning_rate": 6.118213503952779e-05, | |
| "loss": 1.2036, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.24176570458404076, | |
| "grad_norm": 0.23055648803710938, | |
| "learning_rate": 6.078920983839031e-05, | |
| "loss": 1.1496, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.2431239388794567, | |
| "grad_norm": 0.22937384247779846, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 1.1748, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.24448217317487267, | |
| "grad_norm": 0.22974710166454315, | |
| "learning_rate": 6.0001284688802226e-05, | |
| "loss": 1.1668, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.24584040747028862, | |
| "grad_norm": 0.2364954799413681, | |
| "learning_rate": 5.960633586768543e-05, | |
| "loss": 1.1893, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.24719864176570458, | |
| "grad_norm": 0.2343706637620926, | |
| "learning_rate": 5.921076370520058e-05, | |
| "loss": 1.1655, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.24855687606112054, | |
| "grad_norm": 0.23308633267879486, | |
| "learning_rate": 5.8814593869458455e-05, | |
| "loss": 1.1332, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.2499151103565365, | |
| "grad_norm": 0.2331320345401764, | |
| "learning_rate": 5.841785206735192e-05, | |
| "loss": 1.088, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.25127334465195245, | |
| "grad_norm": 0.23977631330490112, | |
| "learning_rate": 5.8020564042888015e-05, | |
| "loss": 1.2078, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.25263157894736843, | |
| "grad_norm": 0.23800607025623322, | |
| "learning_rate": 5.762275557551727e-05, | |
| "loss": 1.1134, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.25398981324278436, | |
| "grad_norm": 0.24236267805099487, | |
| "learning_rate": 5.7224452478461064e-05, | |
| "loss": 1.2159, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.25534804753820034, | |
| "grad_norm": 0.24531729519367218, | |
| "learning_rate": 5.682568059703659e-05, | |
| "loss": 1.1084, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.2567062818336163, | |
| "grad_norm": 0.2564159333705902, | |
| "learning_rate": 5.642646580697973e-05, | |
| "loss": 1.0714, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.25806451612903225, | |
| "grad_norm": 0.2536662817001343, | |
| "learning_rate": 5.602683401276615e-05, | |
| "loss": 1.1332, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.25942275042444823, | |
| "grad_norm": 0.2573738098144531, | |
| "learning_rate": 5.562681114593028e-05, | |
| "loss": 1.1732, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.26078098471986416, | |
| "grad_norm": 0.28553032875061035, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 1.1813, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.26213921901528014, | |
| "grad_norm": 0.2642425000667572, | |
| "learning_rate": 5.482569604572576e-05, | |
| "loss": 1.1169, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.2634974533106961, | |
| "grad_norm": 0.27143600583076477, | |
| "learning_rate": 5.442465579556793e-05, | |
| "loss": 1.17, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.26485568760611206, | |
| "grad_norm": 0.27368876338005066, | |
| "learning_rate": 5.402332843583631e-05, | |
| "loss": 1.1743, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.26621392190152804, | |
| "grad_norm": 0.28669026494026184, | |
| "learning_rate": 5.3621740008088126e-05, | |
| "loss": 1.1476, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.26757215619694397, | |
| "grad_norm": 0.3066455125808716, | |
| "learning_rate": 5.321991657082097e-05, | |
| "loss": 1.2093, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.26893039049235995, | |
| "grad_norm": 0.3169887661933899, | |
| "learning_rate": 5.281788419778187e-05, | |
| "loss": 1.1844, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.2702886247877759, | |
| "grad_norm": 0.34708085656166077, | |
| "learning_rate": 5.2415668976275355e-05, | |
| "loss": 1.154, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.27164685908319186, | |
| "grad_norm": 0.39115723967552185, | |
| "learning_rate": 5.201329700547076e-05, | |
| "loss": 1.1696, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.27164685908319186, | |
| "eval_loss": 1.169433355331421, | |
| "eval_runtime": 75.7878, | |
| "eval_samples_per_second": 16.361, | |
| "eval_steps_per_second": 4.09, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2730050933786078, | |
| "grad_norm": 0.22767877578735352, | |
| "learning_rate": 5.161079439470866e-05, | |
| "loss": 1.2705, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.27436332767402377, | |
| "grad_norm": 0.23202009499073029, | |
| "learning_rate": 5.1208187261806615e-05, | |
| "loss": 1.1301, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.27572156196943975, | |
| "grad_norm": 0.2392009049654007, | |
| "learning_rate": 5.080550173136457e-05, | |
| "loss": 1.2333, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.2770797962648557, | |
| "grad_norm": 0.2345076948404312, | |
| "learning_rate": 5.0402763933069496e-05, | |
| "loss": 1.1608, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.27843803056027167, | |
| "grad_norm": 0.2247873842716217, | |
| "learning_rate": 5e-05, | |
| "loss": 1.1777, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.2797962648556876, | |
| "grad_norm": 0.22494539618492126, | |
| "learning_rate": 4.9597236066930516e-05, | |
| "loss": 1.2278, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.2811544991511036, | |
| "grad_norm": 0.20997026562690735, | |
| "learning_rate": 4.919449826863544e-05, | |
| "loss": 1.1539, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.2825127334465195, | |
| "grad_norm": 0.20957647264003754, | |
| "learning_rate": 4.87918127381934e-05, | |
| "loss": 1.1532, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.2838709677419355, | |
| "grad_norm": 0.21663248538970947, | |
| "learning_rate": 4.8389205605291365e-05, | |
| "loss": 1.2097, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.28522920203735147, | |
| "grad_norm": 0.2205830067396164, | |
| "learning_rate": 4.798670299452926e-05, | |
| "loss": 1.1235, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.2865874363327674, | |
| "grad_norm": 0.21743406355381012, | |
| "learning_rate": 4.758433102372466e-05, | |
| "loss": 1.0935, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.2879456706281834, | |
| "grad_norm": 0.21099945902824402, | |
| "learning_rate": 4.7182115802218126e-05, | |
| "loss": 1.2179, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.2893039049235993, | |
| "grad_norm": 0.2079070806503296, | |
| "learning_rate": 4.678008342917903e-05, | |
| "loss": 1.1754, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.2906621392190153, | |
| "grad_norm": 0.2137642800807953, | |
| "learning_rate": 4.6378259991911886e-05, | |
| "loss": 1.2339, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.2920203735144312, | |
| "grad_norm": 0.22139067947864532, | |
| "learning_rate": 4.597667156416371e-05, | |
| "loss": 1.0929, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.2933786078098472, | |
| "grad_norm": 0.21825869381427765, | |
| "learning_rate": 4.5575344204432084e-05, | |
| "loss": 1.1509, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.29473684210526313, | |
| "grad_norm": 0.217268168926239, | |
| "learning_rate": 4.5174303954274244e-05, | |
| "loss": 1.1935, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.2960950764006791, | |
| "grad_norm": 0.21918080747127533, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 1.1205, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.2974533106960951, | |
| "grad_norm": 0.21873408555984497, | |
| "learning_rate": 4.437318885406973e-05, | |
| "loss": 1.0912, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.298811544991511, | |
| "grad_norm": 0.22883340716362, | |
| "learning_rate": 4.397316598723385e-05, | |
| "loss": 1.1726, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.300169779286927, | |
| "grad_norm": 0.22588761150836945, | |
| "learning_rate": 4.3573534193020274e-05, | |
| "loss": 1.1836, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.30152801358234294, | |
| "grad_norm": 0.22106902301311493, | |
| "learning_rate": 4.317431940296343e-05, | |
| "loss": 1.1351, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.3028862478777589, | |
| "grad_norm": 0.22283874452114105, | |
| "learning_rate": 4.277554752153895e-05, | |
| "loss": 1.1228, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.30424448217317485, | |
| "grad_norm": 0.21989330649375916, | |
| "learning_rate": 4.237724442448273e-05, | |
| "loss": 1.1015, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.30560271646859083, | |
| "grad_norm": 0.2293674349784851, | |
| "learning_rate": 4.197943595711198e-05, | |
| "loss": 1.1147, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.3069609507640068, | |
| "grad_norm": 0.2231401950120926, | |
| "learning_rate": 4.1582147932648074e-05, | |
| "loss": 1.1674, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.30831918505942274, | |
| "grad_norm": 0.2178943008184433, | |
| "learning_rate": 4.118540613054156e-05, | |
| "loss": 1.1364, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.3096774193548387, | |
| "grad_norm": 0.22939598560333252, | |
| "learning_rate": 4.078923629479943e-05, | |
| "loss": 1.0993, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.31103565365025465, | |
| "grad_norm": 0.22620512545108795, | |
| "learning_rate": 4.039366413231458e-05, | |
| "loss": 1.1105, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.31239388794567063, | |
| "grad_norm": 0.23713254928588867, | |
| "learning_rate": 3.9998715311197785e-05, | |
| "loss": 1.1572, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.31375212224108656, | |
| "grad_norm": 0.23715730011463165, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 1.1551, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.31511035653650254, | |
| "grad_norm": 0.2342255711555481, | |
| "learning_rate": 3.92107901616097e-05, | |
| "loss": 1.1296, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.31646859083191853, | |
| "grad_norm": 0.2432532161474228, | |
| "learning_rate": 3.8817864960472236e-05, | |
| "loss": 1.1381, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.31782682512733446, | |
| "grad_norm": 0.23796603083610535, | |
| "learning_rate": 3.842566535205286e-05, | |
| "loss": 1.1609, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.31918505942275044, | |
| "grad_norm": 0.23435711860656738, | |
| "learning_rate": 3.803421678562213e-05, | |
| "loss": 1.2078, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.32054329371816637, | |
| "grad_norm": 0.23656964302062988, | |
| "learning_rate": 3.764354466171652e-05, | |
| "loss": 1.1346, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.32190152801358235, | |
| "grad_norm": 0.24362805485725403, | |
| "learning_rate": 3.725367433049033e-05, | |
| "loss": 1.1695, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.3232597623089983, | |
| "grad_norm": 0.2416103184223175, | |
| "learning_rate": 3.6864631090070655e-05, | |
| "loss": 1.1603, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.32461799660441426, | |
| "grad_norm": 0.2516671121120453, | |
| "learning_rate": 3.6476440184915815e-05, | |
| "loss": 1.1832, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.32597623089983024, | |
| "grad_norm": 0.25407183170318604, | |
| "learning_rate": 3.608912680417737e-05, | |
| "loss": 1.1619, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.32733446519524617, | |
| "grad_norm": 0.24797658622264862, | |
| "learning_rate": 3.570271608006555e-05, | |
| "loss": 1.1321, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.32869269949066215, | |
| "grad_norm": 0.26799654960632324, | |
| "learning_rate": 3.531723308621847e-05, | |
| "loss": 1.2035, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.3300509337860781, | |
| "grad_norm": 0.27492526173591614, | |
| "learning_rate": 3.493270283607522e-05, | |
| "loss": 1.171, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.33140916808149407, | |
| "grad_norm": 0.2613578736782074, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 1.1238, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.33276740237691, | |
| "grad_norm": 0.28807345032691956, | |
| "learning_rate": 3.4166600309926387e-05, | |
| "loss": 1.1153, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.334125636672326, | |
| "grad_norm": 0.28468626737594604, | |
| "learning_rate": 3.3785077745215873e-05, | |
| "loss": 1.1984, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.33548387096774196, | |
| "grad_norm": 0.28014445304870605, | |
| "learning_rate": 3.340460734357359e-05, | |
| "loss": 1.1793, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.3368421052631579, | |
| "grad_norm": 0.2991635203361511, | |
| "learning_rate": 3.3025213793178646e-05, | |
| "loss": 1.1491, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.33820033955857387, | |
| "grad_norm": 0.34155353903770447, | |
| "learning_rate": 3.264692171233485e-05, | |
| "loss": 1.1472, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.3395585738539898, | |
| "grad_norm": 0.39328497648239136, | |
| "learning_rate": 3.226975564787322e-05, | |
| "loss": 1.1768, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.3395585738539898, | |
| "eval_loss": 1.1618001461029053, | |
| "eval_runtime": 75.9188, | |
| "eval_samples_per_second": 16.333, | |
| "eval_steps_per_second": 4.083, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.3409168081494058, | |
| "grad_norm": 0.21301254630088806, | |
| "learning_rate": 3.189374007355917e-05, | |
| "loss": 1.1747, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.3422750424448217, | |
| "grad_norm": 0.2225201576948166, | |
| "learning_rate": 3.151889938850445e-05, | |
| "loss": 1.1631, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.3436332767402377, | |
| "grad_norm": 0.2226182520389557, | |
| "learning_rate": 3.114525791558398e-05, | |
| "loss": 1.1984, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.3449915110356537, | |
| "grad_norm": 0.23307554423809052, | |
| "learning_rate": 3.0772839899857464e-05, | |
| "loss": 1.1848, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.3463497453310696, | |
| "grad_norm": 0.22128012776374817, | |
| "learning_rate": 3.0401669506996256e-05, | |
| "loss": 1.1897, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.3477079796264856, | |
| "grad_norm": 0.22312457859516144, | |
| "learning_rate": 3.003177082171523e-05, | |
| "loss": 1.1978, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.3490662139219015, | |
| "grad_norm": 0.21898281574249268, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 1.1822, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.3504244482173175, | |
| "grad_norm": 0.2056150883436203, | |
| "learning_rate": 2.9295884498599414e-05, | |
| "loss": 1.2227, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.3517826825127334, | |
| "grad_norm": 0.221930593252182, | |
| "learning_rate": 2.8929944611373554e-05, | |
| "loss": 1.1336, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.3531409168081494, | |
| "grad_norm": 0.21197658777236938, | |
| "learning_rate": 2.8565371929847284e-05, | |
| "loss": 1.205, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.3544991511035654, | |
| "grad_norm": 0.20999324321746826, | |
| "learning_rate": 2.8202190110619493e-05, | |
| "loss": 1.1551, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.3558573853989813, | |
| "grad_norm": 0.20450879633426666, | |
| "learning_rate": 2.784042272003794e-05, | |
| "loss": 1.1924, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.3572156196943973, | |
| "grad_norm": 0.20566685497760773, | |
| "learning_rate": 2.7480093232670158e-05, | |
| "loss": 1.1541, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.35857385398981323, | |
| "grad_norm": 0.2097368687391281, | |
| "learning_rate": 2.712122502978024e-05, | |
| "loss": 1.1414, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.3599320882852292, | |
| "grad_norm": 0.21455705165863037, | |
| "learning_rate": 2.6763841397811573e-05, | |
| "loss": 1.1273, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.36129032258064514, | |
| "grad_norm": 0.20965005457401276, | |
| "learning_rate": 2.64079655268759e-05, | |
| "loss": 1.2223, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.3626485568760611, | |
| "grad_norm": 0.21943199634552002, | |
| "learning_rate": 2.605362050924848e-05, | |
| "loss": 1.2121, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.3640067911714771, | |
| "grad_norm": 0.21387051045894623, | |
| "learning_rate": 2.57008293378697e-05, | |
| "loss": 1.1135, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.36536502546689303, | |
| "grad_norm": 0.22179877758026123, | |
| "learning_rate": 2.534961490485313e-05, | |
| "loss": 1.1187, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.366723259762309, | |
| "grad_norm": 0.21317237615585327, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 1.1579, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.36808149405772495, | |
| "grad_norm": 0.2269050031900406, | |
| "learning_rate": 2.4652007309320498e-05, | |
| "loss": 1.1958, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.36943972835314093, | |
| "grad_norm": 0.22643712162971497, | |
| "learning_rate": 2.430565941356157e-05, | |
| "loss": 1.1358, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.37079796264855686, | |
| "grad_norm": 0.22138290107250214, | |
| "learning_rate": 2.3960978786741877e-05, | |
| "loss": 1.199, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.37215619694397284, | |
| "grad_norm": 0.22572052478790283, | |
| "learning_rate": 2.361798779469336e-05, | |
| "loss": 1.1746, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.3735144312393888, | |
| "grad_norm": 0.2312777042388916, | |
| "learning_rate": 2.3276708693609943e-05, | |
| "loss": 1.1201, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.37487266553480475, | |
| "grad_norm": 0.21996109187602997, | |
| "learning_rate": 2.2937163628603435e-05, | |
| "loss": 1.0967, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.37623089983022073, | |
| "grad_norm": 0.2212551087141037, | |
| "learning_rate": 2.259937463226651e-05, | |
| "loss": 1.118, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.37758913412563666, | |
| "grad_norm": 0.22271205484867096, | |
| "learning_rate": 2.2263363623243054e-05, | |
| "loss": 1.1563, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.37894736842105264, | |
| "grad_norm": 0.23066063225269318, | |
| "learning_rate": 2.192915240480596e-05, | |
| "loss": 1.1351, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.38030560271646857, | |
| "grad_norm": 0.24335499107837677, | |
| "learning_rate": 2.1596762663442218e-05, | |
| "loss": 1.2207, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.38166383701188455, | |
| "grad_norm": 0.23115313053131104, | |
| "learning_rate": 2.1266215967445824e-05, | |
| "loss": 1.1353, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.3830220713073005, | |
| "grad_norm": 0.23002904653549194, | |
| "learning_rate": 2.0937533765518187e-05, | |
| "loss": 1.1279, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.38438030560271647, | |
| "grad_norm": 0.2410050332546234, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 1.1076, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.38573853989813245, | |
| "grad_norm": 0.23403891921043396, | |
| "learning_rate": 2.0285848032369137e-05, | |
| "loss": 1.1247, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.3870967741935484, | |
| "grad_norm": 0.23966725170612335, | |
| "learning_rate": 1.996288678810105e-05, | |
| "loss": 1.1294, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.38845500848896436, | |
| "grad_norm": 0.2466222494840622, | |
| "learning_rate": 1.9641874609064443e-05, | |
| "loss": 1.158, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.3898132427843803, | |
| "grad_norm": 0.23464591801166534, | |
| "learning_rate": 1.932283232527956e-05, | |
| "loss": 1.1227, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.39117147707979627, | |
| "grad_norm": 0.2456170916557312, | |
| "learning_rate": 1.9005780638942982e-05, | |
| "loss": 1.1333, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.3925297113752122, | |
| "grad_norm": 0.25727927684783936, | |
| "learning_rate": 1.8690740123084316e-05, | |
| "loss": 1.148, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.3938879456706282, | |
| "grad_norm": 0.2539084851741791, | |
| "learning_rate": 1.837773122023114e-05, | |
| "loss": 1.2192, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.39524617996604416, | |
| "grad_norm": 0.2602689862251282, | |
| "learning_rate": 1.8066774241082612e-05, | |
| "loss": 1.1582, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.3966044142614601, | |
| "grad_norm": 0.2653775215148926, | |
| "learning_rate": 1.7757889363191483e-05, | |
| "loss": 1.122, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.3979626485568761, | |
| "grad_norm": 0.26318901777267456, | |
| "learning_rate": 1.745109662965481e-05, | |
| "loss": 1.0884, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.399320882852292, | |
| "grad_norm": 0.28393861651420593, | |
| "learning_rate": 1.714641594781347e-05, | |
| "loss": 1.1316, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.400679117147708, | |
| "grad_norm": 0.2693468928337097, | |
| "learning_rate": 1.684386708796025e-05, | |
| "loss": 1.1191, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.4020373514431239, | |
| "grad_norm": 0.28361696004867554, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 1.1576, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.4033955857385399, | |
| "grad_norm": 0.30078738927841187, | |
| "learning_rate": 1.62452432224612e-05, | |
| "loss": 1.1782, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.4047538200339559, | |
| "grad_norm": 0.31263673305511475, | |
| "learning_rate": 1.5949207060660138e-05, | |
| "loss": 1.1739, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.4061120543293718, | |
| "grad_norm": 0.33561959862709045, | |
| "learning_rate": 1.5655380406016235e-05, | |
| "loss": 1.1671, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.4074702886247878, | |
| "grad_norm": 0.4124099314212799, | |
| "learning_rate": 1.536378232452003e-05, | |
| "loss": 1.1998, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4074702886247878, | |
| "eval_loss": 1.1560478210449219, | |
| "eval_runtime": 75.727, | |
| "eval_samples_per_second": 16.375, | |
| "eval_steps_per_second": 4.094, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4088285229202037, | |
| "grad_norm": 0.20203496515750885, | |
| "learning_rate": 1.5074431737553157e-05, | |
| "loss": 1.1235, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.4101867572156197, | |
| "grad_norm": 0.2083783745765686, | |
| "learning_rate": 1.4787347420660541e-05, | |
| "loss": 1.1932, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.41154499151103563, | |
| "grad_norm": 0.2198968380689621, | |
| "learning_rate": 1.4502548002332088e-05, | |
| "loss": 1.1904, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.4129032258064516, | |
| "grad_norm": 0.21494299173355103, | |
| "learning_rate": 1.422005196279395e-05, | |
| "loss": 1.1966, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.4142614601018676, | |
| "grad_norm": 0.20090734958648682, | |
| "learning_rate": 1.3939877632809278e-05, | |
| "loss": 1.1352, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.4156196943972835, | |
| "grad_norm": 0.21259278059005737, | |
| "learning_rate": 1.3662043192488849e-05, | |
| "loss": 1.1839, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.4169779286926995, | |
| "grad_norm": 0.2208050638437271, | |
| "learning_rate": 1.338656667011134e-05, | |
| "loss": 1.147, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.41833616298811543, | |
| "grad_norm": 0.2121581882238388, | |
| "learning_rate": 1.3113465940953495e-05, | |
| "loss": 1.0805, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.4196943972835314, | |
| "grad_norm": 0.21660660207271576, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 1.1498, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 0.2145054191350937, | |
| "learning_rate": 1.257446259144494e-05, | |
| "loss": 1.1337, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.42241086587436333, | |
| "grad_norm": 0.20337048172950745, | |
| "learning_rate": 1.2308594946249163e-05, | |
| "loss": 1.1421, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.4237691001697793, | |
| "grad_norm": 0.21610523760318756, | |
| "learning_rate": 1.204517304231343e-05, | |
| "loss": 1.1509, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.42512733446519524, | |
| "grad_norm": 0.21116265654563904, | |
| "learning_rate": 1.178421397270758e-05, | |
| "loss": 1.1868, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.4264855687606112, | |
| "grad_norm": 0.21202139556407928, | |
| "learning_rate": 1.1525734670691701e-05, | |
| "loss": 1.1499, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.42784380305602715, | |
| "grad_norm": 0.21659952402114868, | |
| "learning_rate": 1.1269751908617277e-05, | |
| "loss": 1.149, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.42920203735144313, | |
| "grad_norm": 0.20819424092769623, | |
| "learning_rate": 1.1016282296838887e-05, | |
| "loss": 1.131, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.43056027164685906, | |
| "grad_norm": 0.21869328618049622, | |
| "learning_rate": 1.0765342282636416e-05, | |
| "loss": 1.0854, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.43191850594227504, | |
| "grad_norm": 0.2176653891801834, | |
| "learning_rate": 1.0516948149147754e-05, | |
| "loss": 1.1614, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.433276740237691, | |
| "grad_norm": 0.21377655863761902, | |
| "learning_rate": 1.0271116014312293e-05, | |
| "loss": 1.0566, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.43463497453310695, | |
| "grad_norm": 0.21560655534267426, | |
| "learning_rate": 1.0027861829824952e-05, | |
| "loss": 1.0961, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.43599320882852294, | |
| "grad_norm": 0.21113787591457367, | |
| "learning_rate": 9.787201380101157e-06, | |
| "loss": 1.1845, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.43735144312393887, | |
| "grad_norm": 0.21498249471187592, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 1.2006, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.43870967741935485, | |
| "grad_norm": 0.22041656076908112, | |
| "learning_rate": 9.313723980074018e-06, | |
| "loss": 1.1301, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.4400679117147708, | |
| "grad_norm": 0.22150786221027374, | |
| "learning_rate": 9.080937753040646e-06, | |
| "loss": 1.0695, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.44142614601018676, | |
| "grad_norm": 0.21770867705345154, | |
| "learning_rate": 8.850806705317183e-06, | |
| "loss": 1.1373, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.44278438030560274, | |
| "grad_norm": 0.21694478392601013, | |
| "learning_rate": 8.623345769777514e-06, | |
| "loss": 1.1801, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.44414261460101867, | |
| "grad_norm": 0.2310832142829895, | |
| "learning_rate": 8.398569706035792e-06, | |
| "loss": 1.1082, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.44550084889643465, | |
| "grad_norm": 0.23200993239879608, | |
| "learning_rate": 8.176493099488663e-06, | |
| "loss": 1.1414, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.4468590831918506, | |
| "grad_norm": 0.2326376736164093, | |
| "learning_rate": 7.957130360368898e-06, | |
| "loss": 1.1599, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.44821731748726656, | |
| "grad_norm": 0.2363159954547882, | |
| "learning_rate": 7.740495722810271e-06, | |
| "loss": 1.1786, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.4495755517826825, | |
| "grad_norm": 0.2400091290473938, | |
| "learning_rate": 7.526603243923957e-06, | |
| "loss": 1.1777, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.4509337860780985, | |
| "grad_norm": 0.23490336537361145, | |
| "learning_rate": 7.315466802886401e-06, | |
| "loss": 1.1933, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.45229202037351446, | |
| "grad_norm": 0.24406562745571136, | |
| "learning_rate": 7.107100100038671e-06, | |
| "loss": 1.2079, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.4536502546689304, | |
| "grad_norm": 0.2354699969291687, | |
| "learning_rate": 6.901516655997536e-06, | |
| "loss": 1.1893, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.45500848896434637, | |
| "grad_norm": 0.2448936104774475, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 1.1109, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.4563667232597623, | |
| "grad_norm": 0.24912843108177185, | |
| "learning_rate": 6.498752722928042e-06, | |
| "loss": 1.1893, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.4577249575551783, | |
| "grad_norm": 0.24521379172801971, | |
| "learning_rate": 6.301598368674105e-06, | |
| "loss": 1.1255, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.4590831918505942, | |
| "grad_norm": 0.255632609128952, | |
| "learning_rate": 6.107279541079769e-06, | |
| "loss": 1.1292, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.4604414261460102, | |
| "grad_norm": 0.25223416090011597, | |
| "learning_rate": 5.915808849215304e-06, | |
| "loss": 1.1416, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.4617996604414261, | |
| "grad_norm": 0.2556709051132202, | |
| "learning_rate": 5.727198717339511e-06, | |
| "loss": 1.1466, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.4631578947368421, | |
| "grad_norm": 0.25671127438545227, | |
| "learning_rate": 5.54146138409355e-06, | |
| "loss": 1.1387, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.4645161290322581, | |
| "grad_norm": 0.2614463269710541, | |
| "learning_rate": 5.358608901706802e-06, | |
| "loss": 1.1486, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.465874363327674, | |
| "grad_norm": 0.2740238308906555, | |
| "learning_rate": 5.178653135214812e-06, | |
| "loss": 1.133, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.46723259762309, | |
| "grad_norm": 0.27351686358451843, | |
| "learning_rate": 5.001605761689398e-06, | |
| "loss": 1.1442, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.4685908319185059, | |
| "grad_norm": 0.2915915250778198, | |
| "learning_rate": 4.827478269480895e-06, | |
| "loss": 1.1289, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.4699490662139219, | |
| "grad_norm": 0.28379565477371216, | |
| "learning_rate": 4.65628195747273e-06, | |
| "loss": 1.1466, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.47130730050933783, | |
| "grad_norm": 0.29930734634399414, | |
| "learning_rate": 4.488027934348271e-06, | |
| "loss": 1.1425, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.4726655348047538, | |
| "grad_norm": 0.32029014825820923, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 1.1495, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.4740237691001698, | |
| "grad_norm": 0.34378722310066223, | |
| "learning_rate": 4.16039023417088e-06, | |
| "loss": 1.1866, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.47538200339558573, | |
| "grad_norm": 0.43387606739997864, | |
| "learning_rate": 4.001027817058789e-06, | |
| "loss": 1.1076, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.47538200339558573, | |
| "eval_loss": 1.1534959077835083, | |
| "eval_runtime": 75.7386, | |
| "eval_samples_per_second": 16.372, | |
| "eval_steps_per_second": 4.093, | |
| "step": 350 | |
| } | |
| ], | |
| "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": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.813516281188188e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |