| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9902912621359223, | |
| "eval_steps": 500, | |
| "global_step": 462, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.006472491909385114, | |
| "grad_norm": 0.8681851528027766, | |
| "learning_rate": 4.255319148936171e-06, | |
| "loss": 1.2864, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.012944983818770227, | |
| "grad_norm": 0.8603766114064247, | |
| "learning_rate": 8.510638297872341e-06, | |
| "loss": 1.3029, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.019417475728155338, | |
| "grad_norm": 0.8460211658943029, | |
| "learning_rate": 1.2765957446808511e-05, | |
| "loss": 1.2676, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.025889967637540454, | |
| "grad_norm": 0.823879412927583, | |
| "learning_rate": 1.7021276595744682e-05, | |
| "loss": 1.2794, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.032362459546925564, | |
| "grad_norm": 0.7999755689798974, | |
| "learning_rate": 2.1276595744680852e-05, | |
| "loss": 1.2632, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.038834951456310676, | |
| "grad_norm": 0.779503875119536, | |
| "learning_rate": 2.5531914893617022e-05, | |
| "loss": 1.2356, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.045307443365695796, | |
| "grad_norm": 0.7113356719326077, | |
| "learning_rate": 2.9787234042553192e-05, | |
| "loss": 1.2075, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.05177993527508091, | |
| "grad_norm": 0.5733112055032882, | |
| "learning_rate": 3.4042553191489365e-05, | |
| "loss": 1.1078, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.05825242718446602, | |
| "grad_norm": 0.4686147590427965, | |
| "learning_rate": 3.829787234042553e-05, | |
| "loss": 1.0108, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.06472491909385113, | |
| "grad_norm": 0.4960364471309283, | |
| "learning_rate": 4.2553191489361704e-05, | |
| "loss": 0.9991, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07119741100323625, | |
| "grad_norm": 0.5462124995641008, | |
| "learning_rate": 4.680851063829788e-05, | |
| "loss": 0.9427, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.07766990291262135, | |
| "grad_norm": 0.546681134551426, | |
| "learning_rate": 5.1063829787234044e-05, | |
| "loss": 0.8893, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.08414239482200647, | |
| "grad_norm": 0.556093704754962, | |
| "learning_rate": 5.531914893617022e-05, | |
| "loss": 0.8444, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.09061488673139159, | |
| "grad_norm": 0.4982225051601507, | |
| "learning_rate": 5.9574468085106384e-05, | |
| "loss": 0.7642, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0970873786407767, | |
| "grad_norm": 0.48253513519932356, | |
| "learning_rate": 6.382978723404256e-05, | |
| "loss": 0.7228, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.10355987055016182, | |
| "grad_norm": 0.44162352908966596, | |
| "learning_rate": 6.808510638297873e-05, | |
| "loss": 0.6525, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.11003236245954692, | |
| "grad_norm": 0.3918381633741909, | |
| "learning_rate": 7.23404255319149e-05, | |
| "loss": 0.604, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.11650485436893204, | |
| "grad_norm": 0.2648043458799554, | |
| "learning_rate": 7.659574468085106e-05, | |
| "loss": 0.5693, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.12297734627831715, | |
| "grad_norm": 0.243819913231238, | |
| "learning_rate": 8.085106382978723e-05, | |
| "loss": 0.5205, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.12944983818770225, | |
| "grad_norm": 0.23168627433032857, | |
| "learning_rate": 8.510638297872341e-05, | |
| "loss": 0.5296, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.13592233009708737, | |
| "grad_norm": 0.1994751857050266, | |
| "learning_rate": 8.936170212765958e-05, | |
| "loss": 0.5269, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.1423948220064725, | |
| "grad_norm": 0.2530061947493179, | |
| "learning_rate": 9.361702127659576e-05, | |
| "loss": 0.5091, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.1488673139158576, | |
| "grad_norm": 0.24049496280327698, | |
| "learning_rate": 9.787234042553192e-05, | |
| "loss": 0.5169, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.1553398058252427, | |
| "grad_norm": 0.22948907942121893, | |
| "learning_rate": 0.00010212765957446809, | |
| "loss": 0.4954, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.16181229773462782, | |
| "grad_norm": 0.17253285827605366, | |
| "learning_rate": 0.00010638297872340425, | |
| "loss": 0.4678, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.16828478964401294, | |
| "grad_norm": 0.1938226429644451, | |
| "learning_rate": 0.00011063829787234043, | |
| "loss": 0.4857, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.17475728155339806, | |
| "grad_norm": 0.13721502402963923, | |
| "learning_rate": 0.00011489361702127661, | |
| "loss": 0.466, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.18122977346278318, | |
| "grad_norm": 0.12808136935864992, | |
| "learning_rate": 0.00011914893617021277, | |
| "loss": 0.4652, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.18770226537216828, | |
| "grad_norm": 0.12041781390556437, | |
| "learning_rate": 0.00012340425531914893, | |
| "loss": 0.4527, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.1941747572815534, | |
| "grad_norm": 0.11745727258563804, | |
| "learning_rate": 0.00012765957446808513, | |
| "loss": 0.4392, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.20064724919093851, | |
| "grad_norm": 0.11168371739950012, | |
| "learning_rate": 0.00013191489361702127, | |
| "loss": 0.4578, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.20711974110032363, | |
| "grad_norm": 0.11295507271229074, | |
| "learning_rate": 0.00013617021276595746, | |
| "loss": 0.4384, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.21359223300970873, | |
| "grad_norm": 0.11084156792638247, | |
| "learning_rate": 0.00014042553191489363, | |
| "loss": 0.4292, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.22006472491909385, | |
| "grad_norm": 0.10746985136176236, | |
| "learning_rate": 0.0001446808510638298, | |
| "loss": 0.4202, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.22653721682847897, | |
| "grad_norm": 0.10723808557482935, | |
| "learning_rate": 0.00014893617021276596, | |
| "loss": 0.4345, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.23300970873786409, | |
| "grad_norm": 0.10308761677232968, | |
| "learning_rate": 0.00015319148936170213, | |
| "loss": 0.4318, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.23948220064724918, | |
| "grad_norm": 0.1044215437881733, | |
| "learning_rate": 0.00015744680851063832, | |
| "loss": 0.4289, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2459546925566343, | |
| "grad_norm": 0.10783517564137016, | |
| "learning_rate": 0.00016170212765957446, | |
| "loss": 0.4226, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.2524271844660194, | |
| "grad_norm": 0.11058134845234333, | |
| "learning_rate": 0.00016595744680851065, | |
| "loss": 0.4392, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.2588996763754045, | |
| "grad_norm": 0.10721298987840806, | |
| "learning_rate": 0.00017021276595744682, | |
| "loss": 0.4097, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.26537216828478966, | |
| "grad_norm": 0.10203447619356035, | |
| "learning_rate": 0.00017446808510638298, | |
| "loss": 0.4371, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.27184466019417475, | |
| "grad_norm": 0.09611299924461707, | |
| "learning_rate": 0.00017872340425531915, | |
| "loss": 0.4264, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.2783171521035599, | |
| "grad_norm": 0.09911922773823158, | |
| "learning_rate": 0.00018297872340425532, | |
| "loss": 0.4177, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.284789644012945, | |
| "grad_norm": 0.10195593280531762, | |
| "learning_rate": 0.0001872340425531915, | |
| "loss": 0.4322, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.2912621359223301, | |
| "grad_norm": 0.0990363778451009, | |
| "learning_rate": 0.00019148936170212768, | |
| "loss": 0.4272, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2977346278317152, | |
| "grad_norm": 0.10312296274780797, | |
| "learning_rate": 0.00019574468085106384, | |
| "loss": 0.3997, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3042071197411003, | |
| "grad_norm": 0.09688182042517872, | |
| "learning_rate": 0.0002, | |
| "loss": 0.3955, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3106796116504854, | |
| "grad_norm": 0.09540799755823594, | |
| "learning_rate": 0.00019999713469087867, | |
| "loss": 0.4078, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.31715210355987056, | |
| "grad_norm": 0.10596966214446712, | |
| "learning_rate": 0.00019998853892771453, | |
| "loss": 0.4102, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.32362459546925565, | |
| "grad_norm": 0.10086883515116485, | |
| "learning_rate": 0.00019997421320309795, | |
| "loss": 0.4048, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3300970873786408, | |
| "grad_norm": 0.10400887250356652, | |
| "learning_rate": 0.00019995415833798158, | |
| "loss": 0.403, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.3365695792880259, | |
| "grad_norm": 0.09067444994385492, | |
| "learning_rate": 0.00019992837548163316, | |
| "loss": 0.3867, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.343042071197411, | |
| "grad_norm": 0.09109660691369531, | |
| "learning_rate": 0.00019989686611156972, | |
| "loss": 0.3956, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.34951456310679613, | |
| "grad_norm": 0.09162094937144005, | |
| "learning_rate": 0.000199859632033473, | |
| "loss": 0.407, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.3559870550161812, | |
| "grad_norm": 0.09041901984453071, | |
| "learning_rate": 0.00019981667538108587, | |
| "loss": 0.3913, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.36245954692556637, | |
| "grad_norm": 0.0948520973493128, | |
| "learning_rate": 0.00019976799861609008, | |
| "loss": 0.3959, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.36893203883495146, | |
| "grad_norm": 0.09239535228342881, | |
| "learning_rate": 0.00019971360452796522, | |
| "loss": 0.3976, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.37540453074433655, | |
| "grad_norm": 0.09430664951378093, | |
| "learning_rate": 0.0001996534962338288, | |
| "loss": 0.3984, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.3818770226537217, | |
| "grad_norm": 0.08478493523992439, | |
| "learning_rate": 0.0001995876771782577, | |
| "loss": 0.3862, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.3883495145631068, | |
| "grad_norm": 0.08736157516174174, | |
| "learning_rate": 0.00019951615113309075, | |
| "loss": 0.3768, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3948220064724919, | |
| "grad_norm": 0.08833359616935596, | |
| "learning_rate": 0.00019943892219721253, | |
| "loss": 0.3832, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.40129449838187703, | |
| "grad_norm": 0.09634861288945641, | |
| "learning_rate": 0.0001993559947963185, | |
| "loss": 0.3975, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.4077669902912621, | |
| "grad_norm": 0.08474514858686731, | |
| "learning_rate": 0.00019926737368266144, | |
| "loss": 0.3803, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.41423948220064727, | |
| "grad_norm": 0.0878556667338793, | |
| "learning_rate": 0.00019917306393477907, | |
| "loss": 0.3874, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.42071197411003236, | |
| "grad_norm": 0.08906183743795007, | |
| "learning_rate": 0.00019907307095720303, | |
| "loss": 0.384, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.42718446601941745, | |
| "grad_norm": 0.08738087160835142, | |
| "learning_rate": 0.00019896740048014908, | |
| "loss": 0.3883, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.4336569579288026, | |
| "grad_norm": 0.09070551920354154, | |
| "learning_rate": 0.00019885605855918885, | |
| "loss": 0.38, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.4401294498381877, | |
| "grad_norm": 0.08922144963205411, | |
| "learning_rate": 0.00019873905157490285, | |
| "loss": 0.3854, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.44660194174757284, | |
| "grad_norm": 0.09304718374139792, | |
| "learning_rate": 0.0001986163862325146, | |
| "loss": 0.3804, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.45307443365695793, | |
| "grad_norm": 0.09290101551147209, | |
| "learning_rate": 0.0001984880695615066, | |
| "loss": 0.3876, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.459546925566343, | |
| "grad_norm": 0.08862476310706492, | |
| "learning_rate": 0.0001983541089152174, | |
| "loss": 0.3701, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.46601941747572817, | |
| "grad_norm": 0.0864643410147614, | |
| "learning_rate": 0.00019821451197042026, | |
| "loss": 0.3625, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.47249190938511326, | |
| "grad_norm": 0.0889683558231721, | |
| "learning_rate": 0.0001980692867268832, | |
| "loss": 0.3635, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.47896440129449835, | |
| "grad_norm": 0.090198299336284, | |
| "learning_rate": 0.0001979184415069104, | |
| "loss": 0.3804, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.4854368932038835, | |
| "grad_norm": 0.09214744360572043, | |
| "learning_rate": 0.00019776198495486565, | |
| "loss": 0.367, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.4919093851132686, | |
| "grad_norm": 0.08545016444696638, | |
| "learning_rate": 0.00019759992603667667, | |
| "loss": 0.3671, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.49838187702265374, | |
| "grad_norm": 0.08836432043627582, | |
| "learning_rate": 0.00019743227403932134, | |
| "loss": 0.3759, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.5048543689320388, | |
| "grad_norm": 0.09220525179921107, | |
| "learning_rate": 0.00019725903857029564, | |
| "loss": 0.3759, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.511326860841424, | |
| "grad_norm": 0.09021919927906268, | |
| "learning_rate": 0.00019708022955706292, | |
| "loss": 0.3748, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.517799352750809, | |
| "grad_norm": 0.0857417469759507, | |
| "learning_rate": 0.00019689585724648516, | |
| "loss": 0.3778, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5242718446601942, | |
| "grad_norm": 0.08899327468436542, | |
| "learning_rate": 0.00019670593220423558, | |
| "loss": 0.3711, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.5307443365695793, | |
| "grad_norm": 0.09420631205003693, | |
| "learning_rate": 0.00019651046531419332, | |
| "loss": 0.3678, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.5372168284789643, | |
| "grad_norm": 0.09175386946113716, | |
| "learning_rate": 0.00019630946777781966, | |
| "loss": 0.3669, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5436893203883495, | |
| "grad_norm": 0.08372986124983309, | |
| "learning_rate": 0.0001961029511135161, | |
| "loss": 0.3593, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.5501618122977346, | |
| "grad_norm": 0.08923405383972152, | |
| "learning_rate": 0.00019589092715596417, | |
| "loss": 0.3729, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5566343042071198, | |
| "grad_norm": 0.08853783615052746, | |
| "learning_rate": 0.00019567340805544758, | |
| "loss": 0.3512, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.5631067961165048, | |
| "grad_norm": 0.09158487904428567, | |
| "learning_rate": 0.0001954504062771555, | |
| "loss": 0.3673, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.56957928802589, | |
| "grad_norm": 0.09709095338695828, | |
| "learning_rate": 0.00019522193460046864, | |
| "loss": 0.3758, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.5760517799352751, | |
| "grad_norm": 0.09252319787935548, | |
| "learning_rate": 0.00019498800611822645, | |
| "loss": 0.3588, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.5825242718446602, | |
| "grad_norm": 0.10229380875651542, | |
| "learning_rate": 0.00019474863423597728, | |
| "loss": 0.3859, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5889967637540453, | |
| "grad_norm": 0.09361866653135192, | |
| "learning_rate": 0.00019450383267120982, | |
| "loss": 0.3739, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.5954692556634305, | |
| "grad_norm": 0.0919566522897771, | |
| "learning_rate": 0.00019425361545256727, | |
| "loss": 0.3648, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.6019417475728155, | |
| "grad_norm": 0.10007063382515365, | |
| "learning_rate": 0.0001939979969190432, | |
| "loss": 0.3628, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.6084142394822006, | |
| "grad_norm": 0.08642815308708984, | |
| "learning_rate": 0.00019373699171915988, | |
| "loss": 0.3607, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.6148867313915858, | |
| "grad_norm": 0.08898100138256641, | |
| "learning_rate": 0.00019347061481012894, | |
| "loss": 0.3649, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6213592233009708, | |
| "grad_norm": 0.09929344608921754, | |
| "learning_rate": 0.00019319888145699415, | |
| "loss": 0.3707, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.627831715210356, | |
| "grad_norm": 0.08479565009317201, | |
| "learning_rate": 0.00019292180723175654, | |
| "loss": 0.3596, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6343042071197411, | |
| "grad_norm": 0.08536934965094412, | |
| "learning_rate": 0.00019263940801248226, | |
| "loss": 0.3693, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6407766990291263, | |
| "grad_norm": 0.09264353563725661, | |
| "learning_rate": 0.0001923516999823925, | |
| "loss": 0.3679, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.6472491909385113, | |
| "grad_norm": 0.09075116775864867, | |
| "learning_rate": 0.00019205869962893605, | |
| "loss": 0.3594, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6537216828478964, | |
| "grad_norm": 0.08696915654679535, | |
| "learning_rate": 0.0001917604237428447, | |
| "loss": 0.3699, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.6601941747572816, | |
| "grad_norm": 0.08387717136900205, | |
| "learning_rate": 0.00019145688941717075, | |
| "loss": 0.3569, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.09727169698788522, | |
| "learning_rate": 0.00019114811404630762, | |
| "loss": 0.3725, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.6731391585760518, | |
| "grad_norm": 0.08785848039003973, | |
| "learning_rate": 0.0001908341153249931, | |
| "loss": 0.3625, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.6796116504854369, | |
| "grad_norm": 0.09111196159253629, | |
| "learning_rate": 0.00019051491124729512, | |
| "loss": 0.3658, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.686084142394822, | |
| "grad_norm": 0.09173527522438048, | |
| "learning_rate": 0.00019019052010558088, | |
| "loss": 0.3605, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.6925566343042071, | |
| "grad_norm": 0.08349190333555147, | |
| "learning_rate": 0.00018986096048946824, | |
| "loss": 0.3714, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.6990291262135923, | |
| "grad_norm": 0.08795475348576975, | |
| "learning_rate": 0.0001895262512847607, | |
| "loss": 0.3646, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.7055016181229773, | |
| "grad_norm": 0.0834930063587362, | |
| "learning_rate": 0.00018918641167236505, | |
| "loss": 0.3518, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.7119741100323624, | |
| "grad_norm": 0.08299225873111288, | |
| "learning_rate": 0.00018884146112719207, | |
| "loss": 0.3498, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7184466019417476, | |
| "grad_norm": 0.08676290151306544, | |
| "learning_rate": 0.00018849141941704067, | |
| "loss": 0.3612, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7249190938511327, | |
| "grad_norm": 0.09107846226354271, | |
| "learning_rate": 0.00018813630660146488, | |
| "loss": 0.3508, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7313915857605178, | |
| "grad_norm": 0.09040771142681107, | |
| "learning_rate": 0.00018777614303062457, | |
| "loss": 0.358, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7378640776699029, | |
| "grad_norm": 0.08473984809675678, | |
| "learning_rate": 0.000187410949344119, | |
| "loss": 0.3567, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.7443365695792881, | |
| "grad_norm": 0.09166724031633931, | |
| "learning_rate": 0.00018704074646980415, | |
| "loss": 0.3474, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7508090614886731, | |
| "grad_norm": 0.09150633413591353, | |
| "learning_rate": 0.00018666555562259356, | |
| "loss": 0.3718, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.7572815533980582, | |
| "grad_norm": 0.09139304143928584, | |
| "learning_rate": 0.00018628539830324229, | |
| "loss": 0.3583, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.7637540453074434, | |
| "grad_norm": 0.08557332691225407, | |
| "learning_rate": 0.00018590029629711506, | |
| "loss": 0.3481, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.7702265372168284, | |
| "grad_norm": 0.08670915390405, | |
| "learning_rate": 0.00018551027167293768, | |
| "loss": 0.3486, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.7766990291262136, | |
| "grad_norm": 0.085015484261482, | |
| "learning_rate": 0.00018511534678153244, | |
| "loss": 0.367, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.7831715210355987, | |
| "grad_norm": 0.09010500519447748, | |
| "learning_rate": 0.0001847155442545372, | |
| "loss": 0.3673, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.7896440129449838, | |
| "grad_norm": 0.08749250339043911, | |
| "learning_rate": 0.00018431088700310844, | |
| "loss": 0.3581, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.7961165048543689, | |
| "grad_norm": 0.09294869349273092, | |
| "learning_rate": 0.00018390139821660855, | |
| "loss": 0.3502, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.8025889967637541, | |
| "grad_norm": 0.08999457767358941, | |
| "learning_rate": 0.00018348710136127655, | |
| "loss": 0.3604, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.8090614886731392, | |
| "grad_norm": 0.08450140095669934, | |
| "learning_rate": 0.0001830680201788836, | |
| "loss": 0.3585, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.8155339805825242, | |
| "grad_norm": 0.0885173201152873, | |
| "learning_rate": 0.00018264417868537244, | |
| "loss": 0.36, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.8220064724919094, | |
| "grad_norm": 0.08249252827967349, | |
| "learning_rate": 0.00018221560116948103, | |
| "loss": 0.3456, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8284789644012945, | |
| "grad_norm": 0.09259010876584428, | |
| "learning_rate": 0.0001817823121913506, | |
| "loss": 0.3523, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.8349514563106796, | |
| "grad_norm": 0.0829493165024249, | |
| "learning_rate": 0.00018134433658111845, | |
| "loss": 0.3502, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.8414239482200647, | |
| "grad_norm": 0.0870233230162085, | |
| "learning_rate": 0.00018090169943749476, | |
| "loss": 0.3503, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8478964401294499, | |
| "grad_norm": 0.09013336960665569, | |
| "learning_rate": 0.00018045442612632444, | |
| "loss": 0.355, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.8543689320388349, | |
| "grad_norm": 0.0887494126249745, | |
| "learning_rate": 0.00018000254227913348, | |
| "loss": 0.3492, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.86084142394822, | |
| "grad_norm": 0.08245510607718151, | |
| "learning_rate": 0.00017954607379166, | |
| "loss": 0.3509, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.8673139158576052, | |
| "grad_norm": 0.08652334474586301, | |
| "learning_rate": 0.00017908504682237047, | |
| "loss": 0.3464, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.8737864077669902, | |
| "grad_norm": 0.0832923145834622, | |
| "learning_rate": 0.00017861948779096046, | |
| "loss": 0.3392, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.8802588996763754, | |
| "grad_norm": 0.0812465660556935, | |
| "learning_rate": 0.0001781494233768408, | |
| "loss": 0.333, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.8867313915857605, | |
| "grad_norm": 0.088984541060325, | |
| "learning_rate": 0.00017767488051760857, | |
| "loss": 0.3615, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.8932038834951457, | |
| "grad_norm": 0.0867087441619348, | |
| "learning_rate": 0.00017719588640750336, | |
| "loss": 0.3553, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.8996763754045307, | |
| "grad_norm": 0.08382301700407316, | |
| "learning_rate": 0.00017671246849584903, | |
| "loss": 0.3592, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.9061488673139159, | |
| "grad_norm": 0.0865944397225724, | |
| "learning_rate": 0.0001762246544854807, | |
| "loss": 0.3605, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.912621359223301, | |
| "grad_norm": 0.08628029323884819, | |
| "learning_rate": 0.00017573247233115694, | |
| "loss": 0.371, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.919093851132686, | |
| "grad_norm": 0.08863123995414322, | |
| "learning_rate": 0.00017523595023795813, | |
| "loss": 0.3504, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.9255663430420712, | |
| "grad_norm": 0.08627162896846262, | |
| "learning_rate": 0.00017473511665966993, | |
| "loss": 0.3545, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.9320388349514563, | |
| "grad_norm": 0.08839402497347855, | |
| "learning_rate": 0.00017423000029715267, | |
| "loss": 0.3641, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.9385113268608414, | |
| "grad_norm": 0.08651737654873981, | |
| "learning_rate": 0.00017372063009669686, | |
| "loss": 0.3431, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9449838187702265, | |
| "grad_norm": 0.07938521602972776, | |
| "learning_rate": 0.00017320703524836405, | |
| "loss": 0.3541, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.9514563106796117, | |
| "grad_norm": 0.08679861689760528, | |
| "learning_rate": 0.00017268924518431438, | |
| "loss": 0.3492, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.9579288025889967, | |
| "grad_norm": 0.08562044946225537, | |
| "learning_rate": 0.00017216728957711967, | |
| "loss": 0.3497, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.9644012944983819, | |
| "grad_norm": 0.08302055814489633, | |
| "learning_rate": 0.0001716411983380632, | |
| "loss": 0.3472, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.970873786407767, | |
| "grad_norm": 0.08275430206926206, | |
| "learning_rate": 0.00017111100161542545, | |
| "loss": 0.3383, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.9773462783171522, | |
| "grad_norm": 0.08114036438803068, | |
| "learning_rate": 0.00017057672979275656, | |
| "loss": 0.36, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.9838187702265372, | |
| "grad_norm": 0.0875761383748598, | |
| "learning_rate": 0.0001700384134871351, | |
| "loss": 0.3727, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.9902912621359223, | |
| "grad_norm": 0.0826665996105745, | |
| "learning_rate": 0.0001694960835474134, | |
| "loss": 0.3443, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.9967637540453075, | |
| "grad_norm": 0.08226814015705675, | |
| "learning_rate": 0.00016894977105244997, | |
| "loss": 0.3653, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.9967637540453075, | |
| "eval_loss": 0.3545331358909607, | |
| "eval_runtime": 38.077, | |
| "eval_samples_per_second": 27.287, | |
| "eval_steps_per_second": 0.867, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.0032362459546926, | |
| "grad_norm": 0.08335200234700724, | |
| "learning_rate": 0.0001683995073093283, | |
| "loss": 0.3492, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.0097087378640777, | |
| "grad_norm": 0.0865080297629732, | |
| "learning_rate": 0.00016784532385156285, | |
| "loss": 0.3377, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.0161812297734627, | |
| "grad_norm": 0.08441339355225498, | |
| "learning_rate": 0.0001672872524372919, | |
| "loss": 0.34, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.022653721682848, | |
| "grad_norm": 0.0801171782758195, | |
| "learning_rate": 0.00016672532504745778, | |
| "loss": 0.3354, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.029126213592233, | |
| "grad_norm": 0.08282538959029859, | |
| "learning_rate": 0.00016615957388397399, | |
| "loss": 0.3497, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.035598705501618, | |
| "grad_norm": 0.0883355169428613, | |
| "learning_rate": 0.00016559003136787988, | |
| "loss": 0.338, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.0420711974110033, | |
| "grad_norm": 0.08616766891600182, | |
| "learning_rate": 0.00016501673013748284, | |
| "loss": 0.3491, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.0485436893203883, | |
| "grad_norm": 0.08188626612719885, | |
| "learning_rate": 0.0001644397030464877, | |
| "loss": 0.3359, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.0550161812297734, | |
| "grad_norm": 0.08630887969447615, | |
| "learning_rate": 0.00016385898316211426, | |
| "loss": 0.3268, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.0614886731391586, | |
| "grad_norm": 0.09109351353175275, | |
| "learning_rate": 0.0001632746037632021, | |
| "loss": 0.3356, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.0679611650485437, | |
| "grad_norm": 0.08843718177405194, | |
| "learning_rate": 0.00016268659833830367, | |
| "loss": 0.3205, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.074433656957929, | |
| "grad_norm": 0.09214475868602304, | |
| "learning_rate": 0.00016209500058376515, | |
| "loss": 0.3433, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.080906148867314, | |
| "grad_norm": 0.08469945389831732, | |
| "learning_rate": 0.00016149984440179537, | |
| "loss": 0.3252, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.087378640776699, | |
| "grad_norm": 0.09167891300803974, | |
| "learning_rate": 0.00016090116389852306, | |
| "loss": 0.3372, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.0938511326860842, | |
| "grad_norm": 0.08439372223738191, | |
| "learning_rate": 0.00016029899338204233, | |
| "loss": 0.3339, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.1003236245954693, | |
| "grad_norm": 0.09671189992628862, | |
| "learning_rate": 0.0001596933673604467, | |
| "loss": 0.331, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.1067961165048543, | |
| "grad_norm": 0.08926808998386844, | |
| "learning_rate": 0.00015908432053985143, | |
| "loss": 0.3438, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.1132686084142396, | |
| "grad_norm": 0.08651511861187883, | |
| "learning_rate": 0.0001584718878224047, | |
| "loss": 0.3199, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.1197411003236246, | |
| "grad_norm": 0.09087672552517187, | |
| "learning_rate": 0.00015785610430428762, | |
| "loss": 0.3312, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.1262135922330097, | |
| "grad_norm": 0.08473819565220535, | |
| "learning_rate": 0.00015723700527370268, | |
| "loss": 0.3265, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.132686084142395, | |
| "grad_norm": 0.08959082008669542, | |
| "learning_rate": 0.00015661462620885199, | |
| "loss": 0.3361, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.13915857605178, | |
| "grad_norm": 0.08820985748174735, | |
| "learning_rate": 0.0001559890027759037, | |
| "loss": 0.3352, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.145631067961165, | |
| "grad_norm": 0.08959538258978733, | |
| "learning_rate": 0.00015536017082694846, | |
| "loss": 0.353, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.1521035598705502, | |
| "grad_norm": 0.08556029404315754, | |
| "learning_rate": 0.0001547281663979446, | |
| "loss": 0.338, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.1585760517799353, | |
| "grad_norm": 0.08651722009131164, | |
| "learning_rate": 0.00015409302570665325, | |
| "loss": 0.3428, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.1650485436893203, | |
| "grad_norm": 0.09127001503693886, | |
| "learning_rate": 0.00015345478515056267, | |
| "loss": 0.3196, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.1715210355987056, | |
| "grad_norm": 0.08856796253981307, | |
| "learning_rate": 0.00015281348130480272, | |
| "loss": 0.3309, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.1779935275080906, | |
| "grad_norm": 0.08742855227811815, | |
| "learning_rate": 0.00015216915092004847, | |
| "loss": 0.3273, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.1844660194174756, | |
| "grad_norm": 0.09076657129408205, | |
| "learning_rate": 0.0001515218309204145, | |
| "loss": 0.3324, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.190938511326861, | |
| "grad_norm": 0.09296734178135176, | |
| "learning_rate": 0.00015087155840133888, | |
| "loss": 0.3443, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.197411003236246, | |
| "grad_norm": 0.08431288630485151, | |
| "learning_rate": 0.00015021837062745714, | |
| "loss": 0.3383, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.203883495145631, | |
| "grad_norm": 0.08374691953765048, | |
| "learning_rate": 0.00014956230503046703, | |
| "loss": 0.3368, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.2103559870550162, | |
| "grad_norm": 0.08899017726880568, | |
| "learning_rate": 0.00014890339920698334, | |
| "loss": 0.3368, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.2168284789644013, | |
| "grad_norm": 0.08562973948918146, | |
| "learning_rate": 0.00014824169091638337, | |
| "loss": 0.3397, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.2233009708737863, | |
| "grad_norm": 0.08827756751106554, | |
| "learning_rate": 0.00014757721807864317, | |
| "loss": 0.3405, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.2297734627831716, | |
| "grad_norm": 0.08736187981084419, | |
| "learning_rate": 0.0001469100187721644, | |
| "loss": 0.3262, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.2362459546925566, | |
| "grad_norm": 0.09050203121679472, | |
| "learning_rate": 0.0001462401312315922, | |
| "loss": 0.3345, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.2427184466019416, | |
| "grad_norm": 0.08534856926861663, | |
| "learning_rate": 0.00014556759384562416, | |
| "loss": 0.3277, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.249190938511327, | |
| "grad_norm": 0.08611809442550349, | |
| "learning_rate": 0.00014489244515481046, | |
| "loss": 0.3253, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.255663430420712, | |
| "grad_norm": 0.08688819258876039, | |
| "learning_rate": 0.0001442147238493451, | |
| "loss": 0.3349, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.262135922330097, | |
| "grad_norm": 0.08416654255653569, | |
| "learning_rate": 0.00014353446876684892, | |
| "loss": 0.3308, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.2686084142394822, | |
| "grad_norm": 0.09183514403063862, | |
| "learning_rate": 0.0001428517188901437, | |
| "loss": 0.3383, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.2750809061488673, | |
| "grad_norm": 0.08749863979244098, | |
| "learning_rate": 0.0001421665133450184, | |
| "loss": 0.3294, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.2815533980582523, | |
| "grad_norm": 0.08786155414487208, | |
| "learning_rate": 0.00014147889139798708, | |
| "loss": 0.332, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.2880258899676376, | |
| "grad_norm": 0.08754391459720788, | |
| "learning_rate": 0.00014078889245403844, | |
| "loss": 0.3268, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.2944983818770226, | |
| "grad_norm": 0.09026975504709552, | |
| "learning_rate": 0.0001400965560543778, | |
| "loss": 0.3344, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.3009708737864076, | |
| "grad_norm": 0.08953266507881985, | |
| "learning_rate": 0.0001394019218741612, | |
| "loss": 0.3355, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.307443365695793, | |
| "grad_norm": 0.08868590943641125, | |
| "learning_rate": 0.00013870502972022173, | |
| "loss": 0.3282, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.313915857605178, | |
| "grad_norm": 0.09182148747745249, | |
| "learning_rate": 0.00013800591952878825, | |
| "loss": 0.3394, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.3203883495145632, | |
| "grad_norm": 0.08835010573043636, | |
| "learning_rate": 0.00013730463136319692, | |
| "loss": 0.3316, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.3268608414239482, | |
| "grad_norm": 0.08573944970518392, | |
| "learning_rate": 0.00013660120541159537, | |
| "loss": 0.337, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.08411384694221385, | |
| "learning_rate": 0.00013589568198463944, | |
| "loss": 0.3356, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.3398058252427185, | |
| "grad_norm": 0.08762468166968049, | |
| "learning_rate": 0.0001351881015131833, | |
| "loss": 0.3388, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.3462783171521036, | |
| "grad_norm": 0.08837885567018512, | |
| "learning_rate": 0.00013447850454596265, | |
| "loss": 0.3354, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.3527508090614886, | |
| "grad_norm": 0.08823334999060412, | |
| "learning_rate": 0.00013376693174727065, | |
| "loss": 0.3341, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.3592233009708738, | |
| "grad_norm": 0.08665353483827595, | |
| "learning_rate": 0.00013305342389462792, | |
| "loss": 0.3282, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.3656957928802589, | |
| "grad_norm": 0.08780215390647159, | |
| "learning_rate": 0.00013233802187644566, | |
| "loss": 0.3245, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.3721682847896441, | |
| "grad_norm": 0.0880005148626301, | |
| "learning_rate": 0.0001316207666896824, | |
| "loss": 0.3264, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.3786407766990292, | |
| "grad_norm": 0.09350061568656749, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 0.33, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.3851132686084142, | |
| "grad_norm": 0.0869291546207145, | |
| "learning_rate": 0.00013018086132688184, | |
| "loss": 0.3219, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.3915857605177995, | |
| "grad_norm": 0.0926767409167509, | |
| "learning_rate": 0.0001294582936663239, | |
| "loss": 0.3294, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.3980582524271845, | |
| "grad_norm": 0.08850918948495869, | |
| "learning_rate": 0.00012873403786341513, | |
| "loss": 0.3362, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.4045307443365695, | |
| "grad_norm": 0.08797553458327365, | |
| "learning_rate": 0.00012800813542249072, | |
| "loss": 0.3293, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.4110032362459548, | |
| "grad_norm": 0.08816787591032364, | |
| "learning_rate": 0.00012728062794224832, | |
| "loss": 0.3297, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.4174757281553398, | |
| "grad_norm": 0.08844464048062073, | |
| "learning_rate": 0.0001265515571133643, | |
| "loss": 0.3405, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.4239482200647249, | |
| "grad_norm": 0.08777743550690727, | |
| "learning_rate": 0.00012582096471610467, | |
| "loss": 0.33, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.4304207119741101, | |
| "grad_norm": 0.0841701096943823, | |
| "learning_rate": 0.00012508889261793059, | |
| "loss": 0.3243, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.4368932038834952, | |
| "grad_norm": 0.08211677315785172, | |
| "learning_rate": 0.0001243553827710992, | |
| "loss": 0.3162, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.4433656957928802, | |
| "grad_norm": 0.09229614871738043, | |
| "learning_rate": 0.00012362047721025968, | |
| "loss": 0.3214, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.4498381877022655, | |
| "grad_norm": 0.08827293639891444, | |
| "learning_rate": 0.00012288421805004414, | |
| "loss": 0.3388, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.4563106796116505, | |
| "grad_norm": 0.08650939834799051, | |
| "learning_rate": 0.0001221466474826543, | |
| "loss": 0.3336, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.4627831715210355, | |
| "grad_norm": 0.08745414194563839, | |
| "learning_rate": 0.00012140780777544367, | |
| "loss": 0.3249, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.4692556634304208, | |
| "grad_norm": 0.08966472430098649, | |
| "learning_rate": 0.00012066774126849529, | |
| "loss": 0.3358, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.4757281553398058, | |
| "grad_norm": 0.08547479634425736, | |
| "learning_rate": 0.00011992649037219545, | |
| "loss": 0.3415, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.4822006472491909, | |
| "grad_norm": 0.08630753448523301, | |
| "learning_rate": 0.0001191840975648032, | |
| "loss": 0.3369, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.4886731391585761, | |
| "grad_norm": 0.08171192229188709, | |
| "learning_rate": 0.00011844060539001618, | |
| "loss": 0.3207, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.4951456310679612, | |
| "grad_norm": 0.08413278558608812, | |
| "learning_rate": 0.00011769605645453265, | |
| "loss": 0.3219, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.5016181229773462, | |
| "grad_norm": 0.08498283806676993, | |
| "learning_rate": 0.00011695049342560968, | |
| "loss": 0.3339, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.5080906148867315, | |
| "grad_norm": 0.08365977359578973, | |
| "learning_rate": 0.00011620395902861822, | |
| "loss": 0.3297, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.5145631067961165, | |
| "grad_norm": 0.0858081054599164, | |
| "learning_rate": 0.00011545649604459466, | |
| "loss": 0.3305, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.5210355987055015, | |
| "grad_norm": 0.08916196136213667, | |
| "learning_rate": 0.00011470814730778905, | |
| "loss": 0.3352, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.5275080906148868, | |
| "grad_norm": 0.08650692110960809, | |
| "learning_rate": 0.00011395895570321064, | |
| "loss": 0.3347, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.5339805825242718, | |
| "grad_norm": 0.08839636357345756, | |
| "learning_rate": 0.00011320896416417026, | |
| "loss": 0.3284, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.5404530744336569, | |
| "grad_norm": 0.08477612347310413, | |
| "learning_rate": 0.00011245821566981976, | |
| "loss": 0.3331, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.5469255663430421, | |
| "grad_norm": 0.08158083954977735, | |
| "learning_rate": 0.00011170675324268942, | |
| "loss": 0.3284, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.5533980582524272, | |
| "grad_norm": 0.08210674030802527, | |
| "learning_rate": 0.00011095461994622209, | |
| "loss": 0.3152, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.5598705501618122, | |
| "grad_norm": 0.08164700760175722, | |
| "learning_rate": 0.00011020185888230571, | |
| "loss": 0.3212, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.5663430420711975, | |
| "grad_norm": 0.08252073365071766, | |
| "learning_rate": 0.00010944851318880314, | |
| "loss": 0.3349, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.5728155339805825, | |
| "grad_norm": 0.0847504452125806, | |
| "learning_rate": 0.00010869462603708011, | |
| "loss": 0.3355, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.5792880258899675, | |
| "grad_norm": 0.08651133299846145, | |
| "learning_rate": 0.00010794024062953123, | |
| "loss": 0.3295, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.5857605177993528, | |
| "grad_norm": 0.08848888104315868, | |
| "learning_rate": 0.00010718540019710432, | |
| "loss": 0.334, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.5922330097087378, | |
| "grad_norm": 0.0838746079221116, | |
| "learning_rate": 0.00010643014799682296, | |
| "loss": 0.3236, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.5987055016181229, | |
| "grad_norm": 0.0834887968620315, | |
| "learning_rate": 0.00010567452730930743, | |
| "loss": 0.3231, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.6051779935275081, | |
| "grad_norm": 0.08903656754039865, | |
| "learning_rate": 0.00010491858143629469, | |
| "loss": 0.3303, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.6116504854368932, | |
| "grad_norm": 0.08833888604572854, | |
| "learning_rate": 0.00010416235369815693, | |
| "loss": 0.3417, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.6181229773462782, | |
| "grad_norm": 0.08946556956131413, | |
| "learning_rate": 0.00010340588743141879, | |
| "loss": 0.3341, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.6245954692556634, | |
| "grad_norm": 0.09195837741425467, | |
| "learning_rate": 0.00010264922598627418, | |
| "loss": 0.3292, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.6310679611650487, | |
| "grad_norm": 0.08669211913864286, | |
| "learning_rate": 0.0001018924127241019, | |
| "loss": 0.329, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.6375404530744335, | |
| "grad_norm": 0.08493981781947878, | |
| "learning_rate": 0.00010113549101498086, | |
| "loss": 0.317, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.6440129449838188, | |
| "grad_norm": 0.08859940503358235, | |
| "learning_rate": 0.00010037850423520454, | |
| "loss": 0.3239, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.650485436893204, | |
| "grad_norm": 0.08691997093628756, | |
| "learning_rate": 9.962149576479545e-05, | |
| "loss": 0.3276, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.6569579288025889, | |
| "grad_norm": 0.08182220008549855, | |
| "learning_rate": 9.886450898501917e-05, | |
| "loss": 0.322, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.6634304207119741, | |
| "grad_norm": 0.08983867569693547, | |
| "learning_rate": 9.810758727589813e-05, | |
| "loss": 0.3289, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.6699029126213594, | |
| "grad_norm": 0.08440426328661797, | |
| "learning_rate": 9.735077401372583e-05, | |
| "loss": 0.335, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.6763754045307442, | |
| "grad_norm": 0.09081741368893546, | |
| "learning_rate": 9.659411256858122e-05, | |
| "loss": 0.3285, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.6828478964401294, | |
| "grad_norm": 0.08638692257401422, | |
| "learning_rate": 9.583764630184311e-05, | |
| "loss": 0.3283, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.6893203883495147, | |
| "grad_norm": 0.08661785904833329, | |
| "learning_rate": 9.508141856370532e-05, | |
| "loss": 0.3315, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.6957928802588995, | |
| "grad_norm": 0.08867942067172753, | |
| "learning_rate": 9.432547269069261e-05, | |
| "loss": 0.3298, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.7022653721682848, | |
| "grad_norm": 0.08121634636493791, | |
| "learning_rate": 9.356985200317709e-05, | |
| "loss": 0.3264, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.70873786407767, | |
| "grad_norm": 0.0857656466474692, | |
| "learning_rate": 9.281459980289567e-05, | |
| "loss": 0.3265, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.715210355987055, | |
| "grad_norm": 0.08325063625580009, | |
| "learning_rate": 9.205975937046879e-05, | |
| "loss": 0.3248, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.72168284789644, | |
| "grad_norm": 0.09125541887572731, | |
| "learning_rate": 9.130537396291994e-05, | |
| "loss": 0.3336, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.7281553398058254, | |
| "grad_norm": 0.0858478153556312, | |
| "learning_rate": 9.055148681119688e-05, | |
| "loss": 0.3335, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.7346278317152104, | |
| "grad_norm": 0.08412291629199992, | |
| "learning_rate": 8.979814111769431e-05, | |
| "loss": 0.3247, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.7411003236245954, | |
| "grad_norm": 0.08342610941802514, | |
| "learning_rate": 8.904538005377794e-05, | |
| "loss": 0.3271, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.7475728155339807, | |
| "grad_norm": 0.08566344893624273, | |
| "learning_rate": 8.829324675731059e-05, | |
| "loss": 0.3342, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.7540453074433657, | |
| "grad_norm": 0.08349328064040033, | |
| "learning_rate": 8.754178433018025e-05, | |
| "loss": 0.3169, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.7605177993527508, | |
| "grad_norm": 0.08290650287411376, | |
| "learning_rate": 8.679103583582979e-05, | |
| "loss": 0.32, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.766990291262136, | |
| "grad_norm": 0.08355664601674428, | |
| "learning_rate": 8.604104429678935e-05, | |
| "loss": 0.3284, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.773462783171521, | |
| "grad_norm": 0.08088172666023426, | |
| "learning_rate": 8.529185269221097e-05, | |
| "loss": 0.3189, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.779935275080906, | |
| "grad_norm": 0.08494057369308265, | |
| "learning_rate": 8.45435039554054e-05, | |
| "loss": 0.3175, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.7864077669902914, | |
| "grad_norm": 0.08435898747241219, | |
| "learning_rate": 8.379604097138179e-05, | |
| "loss": 0.3359, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.7928802588996764, | |
| "grad_norm": 0.08369793345019513, | |
| "learning_rate": 8.304950657439033e-05, | |
| "loss": 0.3277, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.7993527508090614, | |
| "grad_norm": 0.08234674894169963, | |
| "learning_rate": 8.230394354546737e-05, | |
| "loss": 0.3106, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.8058252427184467, | |
| "grad_norm": 0.08480659421208028, | |
| "learning_rate": 8.15593946099838e-05, | |
| "loss": 0.3219, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.8122977346278317, | |
| "grad_norm": 0.087986920120639, | |
| "learning_rate": 8.08159024351968e-05, | |
| "loss": 0.3442, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.8187702265372168, | |
| "grad_norm": 0.0839196545152189, | |
| "learning_rate": 8.007350962780456e-05, | |
| "loss": 0.3216, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.825242718446602, | |
| "grad_norm": 0.08396185408224062, | |
| "learning_rate": 7.93322587315047e-05, | |
| "loss": 0.3249, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.831715210355987, | |
| "grad_norm": 0.08739956374343912, | |
| "learning_rate": 7.859219222455634e-05, | |
| "loss": 0.3366, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.838187702265372, | |
| "grad_norm": 0.08539428449034953, | |
| "learning_rate": 7.785335251734573e-05, | |
| "loss": 0.3399, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.8446601941747574, | |
| "grad_norm": 0.08443259129123379, | |
| "learning_rate": 7.711578194995589e-05, | |
| "loss": 0.326, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.8511326860841424, | |
| "grad_norm": 0.08226727888900415, | |
| "learning_rate": 7.637952278974034e-05, | |
| "loss": 0.3122, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.8576051779935274, | |
| "grad_norm": 0.08664703990759018, | |
| "learning_rate": 7.564461722890081e-05, | |
| "loss": 0.3276, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.8640776699029127, | |
| "grad_norm": 0.08284483357408984, | |
| "learning_rate": 7.491110738206942e-05, | |
| "loss": 0.3331, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.8705501618122977, | |
| "grad_norm": 0.08547813142706377, | |
| "learning_rate": 7.417903528389534e-05, | |
| "loss": 0.3108, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.8770226537216828, | |
| "grad_norm": 0.08375066019242061, | |
| "learning_rate": 7.344844288663571e-05, | |
| "loss": 0.3356, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.883495145631068, | |
| "grad_norm": 0.08564167180521602, | |
| "learning_rate": 7.27193720577517e-05, | |
| "loss": 0.3264, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.889967637540453, | |
| "grad_norm": 0.08560457179852256, | |
| "learning_rate": 7.19918645775093e-05, | |
| "loss": 0.3274, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.896440129449838, | |
| "grad_norm": 0.08594335919666654, | |
| "learning_rate": 7.126596213658488e-05, | |
| "loss": 0.3268, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.9029126213592233, | |
| "grad_norm": 0.086743773673858, | |
| "learning_rate": 7.05417063336761e-05, | |
| "loss": 0.3291, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.9093851132686084, | |
| "grad_norm": 0.08512277813785797, | |
| "learning_rate": 6.981913867311819e-05, | |
| "loss": 0.3199, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.9158576051779934, | |
| "grad_norm": 0.08482302988433711, | |
| "learning_rate": 6.909830056250527e-05, | |
| "loss": 0.3308, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.9223300970873787, | |
| "grad_norm": 0.08336748744395403, | |
| "learning_rate": 6.83792333103176e-05, | |
| "loss": 0.3303, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.9288025889967637, | |
| "grad_norm": 0.08685899308663286, | |
| "learning_rate": 6.766197812355438e-05, | |
| "loss": 0.3156, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.9352750809061487, | |
| "grad_norm": 0.08670025171649215, | |
| "learning_rate": 6.69465761053721e-05, | |
| "loss": 0.3263, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.941747572815534, | |
| "grad_norm": 0.08367058440741144, | |
| "learning_rate": 6.623306825272937e-05, | |
| "loss": 0.3184, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.948220064724919, | |
| "grad_norm": 0.08265849338783704, | |
| "learning_rate": 6.552149545403739e-05, | |
| "loss": 0.3141, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.954692556634304, | |
| "grad_norm": 0.08667064648503052, | |
| "learning_rate": 6.48118984868167e-05, | |
| "loss": 0.3176, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.9611650485436893, | |
| "grad_norm": 0.08624477625370294, | |
| "learning_rate": 6.410431801536058e-05, | |
| "loss": 0.3245, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.9676375404530746, | |
| "grad_norm": 0.0863914502801744, | |
| "learning_rate": 6.339879458840465e-05, | |
| "loss": 0.3345, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.9741100323624594, | |
| "grad_norm": 0.0841627898701783, | |
| "learning_rate": 6.269536863680307e-05, | |
| "loss": 0.3194, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.9805825242718447, | |
| "grad_norm": 0.0854832505200848, | |
| "learning_rate": 6.199408047121174e-05, | |
| "loss": 0.3396, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.98705501618123, | |
| "grad_norm": 0.07970724191486213, | |
| "learning_rate": 6.129497027977829e-05, | |
| "loss": 0.3228, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.9935275080906147, | |
| "grad_norm": 0.08216914347596244, | |
| "learning_rate": 6.059807812583883e-05, | |
| "loss": 0.3119, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.08150138639289872, | |
| "learning_rate": 5.990344394562226e-05, | |
| "loss": 0.3166, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.3374881148338318, | |
| "eval_runtime": 35.4418, | |
| "eval_samples_per_second": 29.316, | |
| "eval_steps_per_second": 0.931, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.0064724919093853, | |
| "grad_norm": 0.08131373994961688, | |
| "learning_rate": 5.92111075459616e-05, | |
| "loss": 0.3175, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.01294498381877, | |
| "grad_norm": 0.08591014053929732, | |
| "learning_rate": 5.852110860201294e-05, | |
| "loss": 0.3109, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.0194174757281553, | |
| "grad_norm": 0.08486358862414822, | |
| "learning_rate": 5.7833486654981606e-05, | |
| "loss": 0.3056, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.0258899676375406, | |
| "grad_norm": 0.08216589197196723, | |
| "learning_rate": 5.714828110985635e-05, | |
| "loss": 0.308, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 2.0323624595469254, | |
| "grad_norm": 0.08472956112082734, | |
| "learning_rate": 5.6465531233151126e-05, | |
| "loss": 0.3063, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.0388349514563107, | |
| "grad_norm": 0.08715839296976118, | |
| "learning_rate": 5.578527615065492e-05, | |
| "loss": 0.3146, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.045307443365696, | |
| "grad_norm": 0.08426241524579575, | |
| "learning_rate": 5.510755484518955e-05, | |
| "loss": 0.305, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 2.0517799352750807, | |
| "grad_norm": 0.08998957963984325, | |
| "learning_rate": 5.443240615437586e-05, | |
| "loss": 0.3082, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 2.058252427184466, | |
| "grad_norm": 0.09157974577803293, | |
| "learning_rate": 5.375986876840784e-05, | |
| "loss": 0.3187, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 2.0647249190938513, | |
| "grad_norm": 0.09110362940023115, | |
| "learning_rate": 5.30899812278356e-05, | |
| "loss": 0.3089, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 2.071197411003236, | |
| "grad_norm": 0.08924632615443999, | |
| "learning_rate": 5.2422781921356826e-05, | |
| "loss": 0.3108, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.0776699029126213, | |
| "grad_norm": 0.08653114959635026, | |
| "learning_rate": 5.1758309083616673e-05, | |
| "loss": 0.3033, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.0841423948220066, | |
| "grad_norm": 0.0897639664776455, | |
| "learning_rate": 5.109660079301668e-05, | |
| "loss": 0.3044, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 2.0906148867313914, | |
| "grad_norm": 0.08654171610929062, | |
| "learning_rate": 5.043769496953299e-05, | |
| "loss": 0.3033, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 2.0970873786407767, | |
| "grad_norm": 0.0883439805695338, | |
| "learning_rate": 4.9781629372542895e-05, | |
| "loss": 0.3081, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 2.103559870550162, | |
| "grad_norm": 0.08874070518690325, | |
| "learning_rate": 4.912844159866112e-05, | |
| "loss": 0.3029, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.1100323624595467, | |
| "grad_norm": 0.08970836183938685, | |
| "learning_rate": 4.847816907958549e-05, | |
| "loss": 0.3051, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 2.116504854368932, | |
| "grad_norm": 0.08702359256692323, | |
| "learning_rate": 4.783084907995156e-05, | |
| "loss": 0.3083, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 2.1229773462783172, | |
| "grad_norm": 0.08620252985570646, | |
| "learning_rate": 4.718651869519731e-05, | |
| "loss": 0.3009, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 2.129449838187702, | |
| "grad_norm": 0.09026535273366211, | |
| "learning_rate": 4.654521484943735e-05, | |
| "loss": 0.3095, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 2.1359223300970873, | |
| "grad_norm": 0.08797218673185163, | |
| "learning_rate": 4.59069742933468e-05, | |
| "loss": 0.3037, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.1423948220064726, | |
| "grad_norm": 0.08992206831884317, | |
| "learning_rate": 4.527183360205541e-05, | |
| "loss": 0.3058, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 2.148867313915858, | |
| "grad_norm": 0.09019684013585973, | |
| "learning_rate": 4.4639829173051554e-05, | |
| "loss": 0.3091, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 2.1553398058252426, | |
| "grad_norm": 0.08846989758042821, | |
| "learning_rate": 4.401099722409631e-05, | |
| "loss": 0.2968, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 2.161812297734628, | |
| "grad_norm": 0.09000600047149172, | |
| "learning_rate": 4.338537379114801e-05, | |
| "loss": 0.322, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 2.168284789644013, | |
| "grad_norm": 0.09084909315613375, | |
| "learning_rate": 4.2762994726297346e-05, | |
| "loss": 0.3104, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.174757281553398, | |
| "grad_norm": 0.08897581163092526, | |
| "learning_rate": 4.2143895695712444e-05, | |
| "loss": 0.3029, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 2.1812297734627832, | |
| "grad_norm": 0.08806995729841001, | |
| "learning_rate": 4.152811217759529e-05, | |
| "loss": 0.3127, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 2.1877022653721685, | |
| "grad_norm": 0.0909392143685305, | |
| "learning_rate": 4.091567946014858e-05, | |
| "loss": 0.3077, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 2.1941747572815533, | |
| "grad_norm": 0.08526830124486821, | |
| "learning_rate": 4.0306632639553323e-05, | |
| "loss": 0.3092, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 2.2006472491909386, | |
| "grad_norm": 0.0889661858323667, | |
| "learning_rate": 3.970100661795766e-05, | |
| "loss": 0.306, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.207119741100324, | |
| "grad_norm": 0.08742864887103546, | |
| "learning_rate": 3.909883610147696e-05, | |
| "loss": 0.3021, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 2.2135922330097086, | |
| "grad_norm": 0.08736016824506748, | |
| "learning_rate": 3.8500155598204644e-05, | |
| "loss": 0.2974, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 2.220064724919094, | |
| "grad_norm": 0.0895276470969271, | |
| "learning_rate": 3.7904999416234864e-05, | |
| "loss": 0.3121, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 2.226537216828479, | |
| "grad_norm": 0.08707410830498415, | |
| "learning_rate": 3.731340166169635e-05, | |
| "loss": 0.304, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 2.233009708737864, | |
| "grad_norm": 0.09084514125957084, | |
| "learning_rate": 3.6725396236797935e-05, | |
| "loss": 0.3139, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.2394822006472492, | |
| "grad_norm": 0.08595616841814938, | |
| "learning_rate": 3.614101683788575e-05, | |
| "loss": 0.3034, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 2.2459546925566345, | |
| "grad_norm": 0.08918897946715092, | |
| "learning_rate": 3.5560296953512295e-05, | |
| "loss": 0.2871, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 2.2524271844660193, | |
| "grad_norm": 0.08951525733240658, | |
| "learning_rate": 3.498326986251717e-05, | |
| "loss": 0.3064, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 2.2588996763754046, | |
| "grad_norm": 0.08959019506927517, | |
| "learning_rate": 3.4409968632120126e-05, | |
| "loss": 0.3062, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 2.26537216828479, | |
| "grad_norm": 0.09072525944669337, | |
| "learning_rate": 3.3840426116026044e-05, | |
| "loss": 0.3113, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.2718446601941746, | |
| "grad_norm": 0.09370535186596111, | |
| "learning_rate": 3.327467495254225e-05, | |
| "loss": 0.3195, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 2.27831715210356, | |
| "grad_norm": 0.08681410960013289, | |
| "learning_rate": 3.2712747562708115e-05, | |
| "loss": 0.3039, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 2.284789644012945, | |
| "grad_norm": 0.0884544764695178, | |
| "learning_rate": 3.215467614843719e-05, | |
| "loss": 0.3028, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 2.29126213592233, | |
| "grad_norm": 0.08974826283064652, | |
| "learning_rate": 3.160049269067174e-05, | |
| "loss": 0.3073, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 2.2977346278317152, | |
| "grad_norm": 0.08791055069767624, | |
| "learning_rate": 3.105022894755003e-05, | |
| "loss": 0.3011, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 2.3042071197411005, | |
| "grad_norm": 0.08951734113600135, | |
| "learning_rate": 3.0503916452586612e-05, | |
| "loss": 0.3056, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 2.3106796116504853, | |
| "grad_norm": 0.08762327614071365, | |
| "learning_rate": 2.9961586512864947e-05, | |
| "loss": 0.3138, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 2.3171521035598706, | |
| "grad_norm": 0.09032122428603973, | |
| "learning_rate": 2.9423270207243437e-05, | |
| "loss": 0.2978, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 2.323624595469256, | |
| "grad_norm": 0.0917821868399838, | |
| "learning_rate": 2.888899838457455e-05, | |
| "loss": 0.3084, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 2.3300970873786406, | |
| "grad_norm": 0.09193242261010433, | |
| "learning_rate": 2.835880166193683e-05, | |
| "loss": 0.3064, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.336569579288026, | |
| "grad_norm": 0.08883763145021951, | |
| "learning_rate": 2.7832710422880328e-05, | |
| "loss": 0.2993, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 2.343042071197411, | |
| "grad_norm": 0.08991962234888566, | |
| "learning_rate": 2.7310754815685624e-05, | |
| "loss": 0.3034, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 2.349514563106796, | |
| "grad_norm": 0.08994982579652418, | |
| "learning_rate": 2.679296475163595e-05, | |
| "loss": 0.3072, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 2.355987055016181, | |
| "grad_norm": 0.08970808859923911, | |
| "learning_rate": 2.6279369903303175e-05, | |
| "loss": 0.3023, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 2.3624595469255665, | |
| "grad_norm": 0.08588546055096836, | |
| "learning_rate": 2.5769999702847346e-05, | |
| "loss": 0.296, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 2.3689320388349513, | |
| "grad_norm": 0.08990846870006015, | |
| "learning_rate": 2.5264883340330113e-05, | |
| "loss": 0.3022, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 2.3754045307443366, | |
| "grad_norm": 0.09151580193749502, | |
| "learning_rate": 2.4764049762041874e-05, | |
| "loss": 0.2978, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 2.381877022653722, | |
| "grad_norm": 0.08959471455140089, | |
| "learning_rate": 2.426752766884306e-05, | |
| "loss": 0.3099, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 2.3883495145631066, | |
| "grad_norm": 0.0887364245299765, | |
| "learning_rate": 2.377534551451932e-05, | |
| "loss": 0.3016, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 2.394822006472492, | |
| "grad_norm": 0.08882940905923181, | |
| "learning_rate": 2.328753150415094e-05, | |
| "loss": 0.3004, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.401294498381877, | |
| "grad_norm": 0.08843646634108779, | |
| "learning_rate": 2.280411359249668e-05, | |
| "loss": 0.2933, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 2.407766990291262, | |
| "grad_norm": 0.09301769304474337, | |
| "learning_rate": 2.2325119482391467e-05, | |
| "loss": 0.3134, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 2.414239482200647, | |
| "grad_norm": 0.08707266390112682, | |
| "learning_rate": 2.185057662315918e-05, | |
| "loss": 0.3003, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 2.4207119741100325, | |
| "grad_norm": 0.0868807473282313, | |
| "learning_rate": 2.1380512209039528e-05, | |
| "loss": 0.302, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 2.4271844660194173, | |
| "grad_norm": 0.09363452018708772, | |
| "learning_rate": 2.0914953177629548e-05, | |
| "loss": 0.3148, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.4336569579288025, | |
| "grad_norm": 0.08938831583372829, | |
| "learning_rate": 2.0453926208340003e-05, | |
| "loss": 0.3055, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 2.440129449838188, | |
| "grad_norm": 0.09230125425366141, | |
| "learning_rate": 1.999745772086655e-05, | |
| "loss": 0.3041, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 2.4466019417475726, | |
| "grad_norm": 0.08932842130151852, | |
| "learning_rate": 1.954557387367557e-05, | |
| "loss": 0.3037, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 2.453074433656958, | |
| "grad_norm": 0.08923122998718774, | |
| "learning_rate": 1.9098300562505266e-05, | |
| "loss": 0.3055, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 2.459546925566343, | |
| "grad_norm": 0.0917592218140953, | |
| "learning_rate": 1.8655663418881584e-05, | |
| "loss": 0.3006, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.466019417475728, | |
| "grad_norm": 0.08856315870273591, | |
| "learning_rate": 1.821768780864943e-05, | |
| "loss": 0.3072, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 2.472491909385113, | |
| "grad_norm": 0.08867770851651259, | |
| "learning_rate": 1.7784398830519e-05, | |
| "loss": 0.298, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 2.4789644012944985, | |
| "grad_norm": 0.08895980547301872, | |
| "learning_rate": 1.7355821314627564e-05, | |
| "loss": 0.3087, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 2.4854368932038833, | |
| "grad_norm": 0.09119431484259279, | |
| "learning_rate": 1.6931979821116418e-05, | |
| "loss": 0.305, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 2.4919093851132685, | |
| "grad_norm": 0.09318125516492613, | |
| "learning_rate": 1.6512898638723497e-05, | |
| "loss": 0.3095, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 2.498381877022654, | |
| "grad_norm": 0.08770991666106868, | |
| "learning_rate": 1.6098601783391487e-05, | |
| "loss": 0.292, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 2.5048543689320386, | |
| "grad_norm": 0.08730384586077941, | |
| "learning_rate": 1.5689112996891576e-05, | |
| "loss": 0.2982, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 2.511326860841424, | |
| "grad_norm": 0.09025336858704619, | |
| "learning_rate": 1.5284455745462834e-05, | |
| "loss": 0.3175, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 2.517799352750809, | |
| "grad_norm": 0.09170439944127616, | |
| "learning_rate": 1.4884653218467571e-05, | |
| "loss": 0.3131, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 2.524271844660194, | |
| "grad_norm": 0.08975472589683904, | |
| "learning_rate": 1.4489728327062324e-05, | |
| "loss": 0.3077, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.530744336569579, | |
| "grad_norm": 0.09151247517360593, | |
| "learning_rate": 1.4099703702884936e-05, | |
| "loss": 0.2987, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 2.5372168284789645, | |
| "grad_norm": 0.08707773026948659, | |
| "learning_rate": 1.3714601696757712e-05, | |
| "loss": 0.301, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 2.5436893203883493, | |
| "grad_norm": 0.09236291075818845, | |
| "learning_rate": 1.3334444377406452e-05, | |
| "loss": 0.3095, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 2.5501618122977345, | |
| "grad_norm": 0.09003080825117314, | |
| "learning_rate": 1.2959253530195836e-05, | |
| "loss": 0.301, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 2.55663430420712, | |
| "grad_norm": 0.08899294818724815, | |
| "learning_rate": 1.258905065588103e-05, | |
| "loss": 0.2996, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 2.5631067961165046, | |
| "grad_norm": 0.08642480774869472, | |
| "learning_rate": 1.2223856969375447e-05, | |
| "loss": 0.2972, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 2.56957928802589, | |
| "grad_norm": 0.09069465990264014, | |
| "learning_rate": 1.1863693398535114e-05, | |
| "loss": 0.3028, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 2.576051779935275, | |
| "grad_norm": 0.09007456341294559, | |
| "learning_rate": 1.1508580582959349e-05, | |
| "loss": 0.3026, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 2.58252427184466, | |
| "grad_norm": 0.0874176261621062, | |
| "learning_rate": 1.1158538872807933e-05, | |
| "loss": 0.302, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 2.588996763754045, | |
| "grad_norm": 0.08762089070699175, | |
| "learning_rate": 1.0813588327634961e-05, | |
| "loss": 0.2845, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.5954692556634305, | |
| "grad_norm": 0.08875402737182131, | |
| "learning_rate": 1.0473748715239307e-05, | |
| "loss": 0.3016, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 2.6019417475728153, | |
| "grad_norm": 0.09129343356719948, | |
| "learning_rate": 1.01390395105318e-05, | |
| "loss": 0.3154, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 2.6084142394822005, | |
| "grad_norm": 0.08810297072389008, | |
| "learning_rate": 9.809479894419149e-06, | |
| "loss": 0.3033, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 2.614886731391586, | |
| "grad_norm": 0.09018340343849121, | |
| "learning_rate": 9.485088752704885e-06, | |
| "loss": 0.3088, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 2.6213592233009706, | |
| "grad_norm": 0.08968883260752107, | |
| "learning_rate": 9.16588467500693e-06, | |
| "loss": 0.2982, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 2.627831715210356, | |
| "grad_norm": 0.08957666144874589, | |
| "learning_rate": 8.851885953692374e-06, | |
| "loss": 0.3082, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 2.634304207119741, | |
| "grad_norm": 0.09198641057091544, | |
| "learning_rate": 8.543110582829272e-06, | |
| "loss": 0.3015, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 2.6407766990291264, | |
| "grad_norm": 0.08915392372194866, | |
| "learning_rate": 8.239576257155334e-06, | |
| "loss": 0.3068, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 2.647249190938511, | |
| "grad_norm": 0.09135989217201736, | |
| "learning_rate": 7.941300371063953e-06, | |
| "loss": 0.3118, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 2.6537216828478964, | |
| "grad_norm": 0.09150707922550998, | |
| "learning_rate": 7.648300017607534e-06, | |
| "loss": 0.3049, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.6601941747572817, | |
| "grad_norm": 0.08679197622255307, | |
| "learning_rate": 7.360591987517762e-06, | |
| "loss": 0.2973, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.09043247718192007, | |
| "learning_rate": 7.078192768243486e-06, | |
| "loss": 0.3016, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 2.6731391585760518, | |
| "grad_norm": 0.09227814811133576, | |
| "learning_rate": 6.80111854300588e-06, | |
| "loss": 0.3109, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 2.679611650485437, | |
| "grad_norm": 0.08752395500418549, | |
| "learning_rate": 6.5293851898710625e-06, | |
| "loss": 0.301, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 2.686084142394822, | |
| "grad_norm": 0.08959180387680037, | |
| "learning_rate": 6.2630082808401326e-06, | |
| "loss": 0.2992, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 2.692556634304207, | |
| "grad_norm": 0.08743171313620991, | |
| "learning_rate": 6.00200308095682e-06, | |
| "loss": 0.301, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 2.6990291262135924, | |
| "grad_norm": 0.08980865803761802, | |
| "learning_rate": 5.746384547432737e-06, | |
| "loss": 0.3174, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 2.705501618122977, | |
| "grad_norm": 0.08661713786701486, | |
| "learning_rate": 5.496167328790191e-06, | |
| "loss": 0.2939, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 2.7119741100323624, | |
| "grad_norm": 0.08786309161255973, | |
| "learning_rate": 5.251365764022753e-06, | |
| "loss": 0.2978, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 2.7184466019417477, | |
| "grad_norm": 0.08733419882477451, | |
| "learning_rate": 5.011993881773569e-06, | |
| "loss": 0.3002, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.724919093851133, | |
| "grad_norm": 0.08926182705644836, | |
| "learning_rate": 4.778065399531395e-06, | |
| "loss": 0.2975, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 2.7313915857605178, | |
| "grad_norm": 0.0894176370961602, | |
| "learning_rate": 4.549593722844492e-06, | |
| "loss": 0.3067, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 2.737864077669903, | |
| "grad_norm": 0.0876923455017712, | |
| "learning_rate": 4.326591944552438e-06, | |
| "loss": 0.3097, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 2.7443365695792883, | |
| "grad_norm": 0.08572253083705766, | |
| "learning_rate": 4.109072844035844e-06, | |
| "loss": 0.2911, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 2.750809061488673, | |
| "grad_norm": 0.08883901045408762, | |
| "learning_rate": 3.8970488864839334e-06, | |
| "loss": 0.2963, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.7572815533980584, | |
| "grad_norm": 0.09240186844770727, | |
| "learning_rate": 3.690532222180343e-06, | |
| "loss": 0.3024, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 2.7637540453074436, | |
| "grad_norm": 0.09051220479105677, | |
| "learning_rate": 3.4895346858066724e-06, | |
| "loss": 0.3089, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 2.7702265372168284, | |
| "grad_norm": 0.0891289977534799, | |
| "learning_rate": 3.2940677957644215e-06, | |
| "loss": 0.3047, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 2.7766990291262137, | |
| "grad_norm": 0.08973918683676917, | |
| "learning_rate": 3.104142753514849e-06, | |
| "loss": 0.3151, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 2.783171521035599, | |
| "grad_norm": 0.08839195406678332, | |
| "learning_rate": 2.9197704429370977e-06, | |
| "loss": 0.2995, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.7896440129449838, | |
| "grad_norm": 0.08711643702101148, | |
| "learning_rate": 2.7409614297043806e-06, | |
| "loss": 0.2983, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 2.796116504854369, | |
| "grad_norm": 0.08827751306740735, | |
| "learning_rate": 2.5677259606786684e-06, | |
| "loss": 0.3071, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 2.8025889967637543, | |
| "grad_norm": 0.08989812034868257, | |
| "learning_rate": 2.4000739633233347e-06, | |
| "loss": 0.3185, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 2.809061488673139, | |
| "grad_norm": 0.08814993018203572, | |
| "learning_rate": 2.238015045134334e-06, | |
| "loss": 0.303, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 2.8155339805825244, | |
| "grad_norm": 0.08731684328369535, | |
| "learning_rate": 2.0815584930895972e-06, | |
| "loss": 0.3053, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 2.8220064724919096, | |
| "grad_norm": 0.09248953078205228, | |
| "learning_rate": 1.9307132731168352e-06, | |
| "loss": 0.3048, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 2.8284789644012944, | |
| "grad_norm": 0.08979246033823442, | |
| "learning_rate": 1.7854880295797405e-06, | |
| "loss": 0.3046, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 2.8349514563106797, | |
| "grad_norm": 0.08615293987508164, | |
| "learning_rate": 1.6458910847826026e-06, | |
| "loss": 0.2991, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 2.841423948220065, | |
| "grad_norm": 0.08795133012526277, | |
| "learning_rate": 1.5119304384934252e-06, | |
| "loss": 0.2922, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 2.8478964401294498, | |
| "grad_norm": 0.0862166031371393, | |
| "learning_rate": 1.3836137674854255e-06, | |
| "loss": 0.2921, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.854368932038835, | |
| "grad_norm": 0.08955236492173343, | |
| "learning_rate": 1.2609484250971749e-06, | |
| "loss": 0.2997, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 2.8608414239482203, | |
| "grad_norm": 0.08843363251406085, | |
| "learning_rate": 1.143941440811147e-06, | |
| "loss": 0.3088, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 2.867313915857605, | |
| "grad_norm": 0.09009839385408548, | |
| "learning_rate": 1.0325995198509409e-06, | |
| "loss": 0.3109, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 2.8737864077669903, | |
| "grad_norm": 0.08785361574469933, | |
| "learning_rate": 9.269290427969868e-07, | |
| "loss": 0.2971, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 2.8802588996763756, | |
| "grad_norm": 0.08841485850159143, | |
| "learning_rate": 8.26936065220929e-07, | |
| "loss": 0.3035, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 2.8867313915857604, | |
| "grad_norm": 0.08779257157327024, | |
| "learning_rate": 7.326263173385584e-07, | |
| "loss": 0.3013, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 2.8932038834951457, | |
| "grad_norm": 0.0883658852563661, | |
| "learning_rate": 6.440052036815081e-07, | |
| "loss": 0.305, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 2.899676375404531, | |
| "grad_norm": 0.08862179969635353, | |
| "learning_rate": 5.610778027874908e-07, | |
| "loss": 0.2952, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 2.9061488673139158, | |
| "grad_norm": 0.08863424887187872, | |
| "learning_rate": 4.838488669092534e-07, | |
| "loss": 0.2956, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 2.912621359223301, | |
| "grad_norm": 0.08998809785580231, | |
| "learning_rate": 4.123228217422948e-07, | |
| "loss": 0.3015, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.9190938511326863, | |
| "grad_norm": 0.08707387881008244, | |
| "learning_rate": 3.465037661712134e-07, | |
| "loss": 0.3095, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 2.925566343042071, | |
| "grad_norm": 0.08740571796909039, | |
| "learning_rate": 2.86395472034795e-07, | |
| "loss": 0.3033, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 2.9320388349514563, | |
| "grad_norm": 0.08599743041374118, | |
| "learning_rate": 2.3200138390993e-07, | |
| "loss": 0.2904, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 2.9385113268608416, | |
| "grad_norm": 0.08726631646259621, | |
| "learning_rate": 1.83324618914138e-07, | |
| "loss": 0.3007, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 2.9449838187702264, | |
| "grad_norm": 0.08898358262047706, | |
| "learning_rate": 1.4036796652701078e-07, | |
| "loss": 0.3017, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 2.9514563106796117, | |
| "grad_norm": 0.08765396985812648, | |
| "learning_rate": 1.031338884302846e-07, | |
| "loss": 0.2887, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 2.957928802588997, | |
| "grad_norm": 0.09001756499046232, | |
| "learning_rate": 7.162451836685291e-08, | |
| "loss": 0.309, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 2.9644012944983817, | |
| "grad_norm": 0.09045952221990206, | |
| "learning_rate": 4.584166201841988e-08, | |
| "loss": 0.2996, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 2.970873786407767, | |
| "grad_norm": 0.08709990251871128, | |
| "learning_rate": 2.578679690204977e-08, | |
| "loss": 0.2982, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 2.9773462783171523, | |
| "grad_norm": 0.08844965304987025, | |
| "learning_rate": 1.1461072285490204e-08, | |
| "loss": 0.3136, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.983818770226537, | |
| "grad_norm": 0.0892220185111147, | |
| "learning_rate": 2.865309121358184e-09, | |
| "loss": 0.3011, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 2.9902912621359223, | |
| "grad_norm": 0.08562579597167659, | |
| "learning_rate": 0.0, | |
| "loss": 0.2951, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 2.9902912621359223, | |
| "eval_loss": 0.3344117999076843, | |
| "eval_runtime": 35.3894, | |
| "eval_samples_per_second": 29.359, | |
| "eval_steps_per_second": 0.932, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 2.9902912621359223, | |
| "step": 462, | |
| "total_flos": 1.8679451641472614e+17, | |
| "train_loss": 0.36376328850205325, | |
| "train_runtime": 5739.3163, | |
| "train_samples_per_second": 10.311, | |
| "train_steps_per_second": 0.08 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 462, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.8679451641472614e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |