| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9981916817359855, | |
| "eval_steps": 500, | |
| "global_step": 138, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.007233273056057866, | |
| "grad_norm": 5.463847855317702, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.0902, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.014466546112115732, | |
| "grad_norm": 4.681007032225575, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.9968, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0216998191681736, | |
| "grad_norm": 4.218229671072367, | |
| "learning_rate": 1.2e-05, | |
| "loss": 0.9226, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.028933092224231464, | |
| "grad_norm": 9.650732793859843, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 0.896, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03616636528028933, | |
| "grad_norm": 5.578584837598941, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7791, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0433996383363472, | |
| "grad_norm": 4.789830044233945, | |
| "learning_rate": 1.9997210372120276e-05, | |
| "loss": 0.6718, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.05063291139240506, | |
| "grad_norm": 4.141804977350294, | |
| "learning_rate": 1.998884304488584e-05, | |
| "loss": 0.6076, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.05786618444846293, | |
| "grad_norm": 3.5515048983104056, | |
| "learning_rate": 1.997490268664256e-05, | |
| "loss": 0.4989, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0650994575045208, | |
| "grad_norm": 3.153711457430971, | |
| "learning_rate": 1.995539707507284e-05, | |
| "loss": 0.4152, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.07233273056057866, | |
| "grad_norm": 2.9236266977118777, | |
| "learning_rate": 1.9930337092856243e-05, | |
| "loss": 0.4002, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07956600361663653, | |
| "grad_norm": 2.876849628749717, | |
| "learning_rate": 1.9899736721597787e-05, | |
| "loss": 0.395, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0867992766726944, | |
| "grad_norm": 2.4107246576130668, | |
| "learning_rate": 1.9863613034027224e-05, | |
| "loss": 0.3166, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.09403254972875226, | |
| "grad_norm": 2.2339458170958726, | |
| "learning_rate": 1.9821986184473757e-05, | |
| "loss": 0.3097, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.10126582278481013, | |
| "grad_norm": 2.2197484192147043, | |
| "learning_rate": 1.9774879397621387e-05, | |
| "loss": 0.3013, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.10849909584086799, | |
| "grad_norm": 1.9826673969406083, | |
| "learning_rate": 1.9722318955551307e-05, | |
| "loss": 0.3094, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.11573236889692586, | |
| "grad_norm": 2.984906540777772, | |
| "learning_rate": 1.966433418307843e-05, | |
| "loss": 0.3331, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.12296564195298372, | |
| "grad_norm": 2.2412636771179346, | |
| "learning_rate": 1.960095743139033e-05, | |
| "loss": 0.3285, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1301989150090416, | |
| "grad_norm": 1.857679521391161, | |
| "learning_rate": 1.9532224059997693e-05, | |
| "loss": 0.2672, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.13743218806509946, | |
| "grad_norm": 2.0139486786788434, | |
| "learning_rate": 1.9458172417006347e-05, | |
| "loss": 0.2523, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.14466546112115733, | |
| "grad_norm": 1.7175026487258431, | |
| "learning_rate": 1.9378843817721856e-05, | |
| "loss": 0.2444, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1518987341772152, | |
| "grad_norm": 1.855948744613208, | |
| "learning_rate": 1.929428252159866e-05, | |
| "loss": 0.2489, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.15913200723327306, | |
| "grad_norm": 2.2399762935756637, | |
| "learning_rate": 1.9204535707546602e-05, | |
| "loss": 0.347, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.16636528028933092, | |
| "grad_norm": 1.8979316016809038, | |
| "learning_rate": 1.9109653447608607e-05, | |
| "loss": 0.2244, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.1735985533453888, | |
| "grad_norm": 1.8508155059344538, | |
| "learning_rate": 1.900968867902419e-05, | |
| "loss": 0.2151, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.18083182640144665, | |
| "grad_norm": 1.8054619356574957, | |
| "learning_rate": 1.8904697174694447e-05, | |
| "loss": 0.2414, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.18806509945750452, | |
| "grad_norm": 1.633010553246676, | |
| "learning_rate": 1.879473751206489e-05, | |
| "loss": 0.2211, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.19529837251356238, | |
| "grad_norm": 1.774632206684753, | |
| "learning_rate": 1.8679871040443632e-05, | |
| "loss": 0.2194, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.20253164556962025, | |
| "grad_norm": 1.48448462933083, | |
| "learning_rate": 1.8560161846773002e-05, | |
| "loss": 0.1607, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.20976491862567812, | |
| "grad_norm": 1.950497681891261, | |
| "learning_rate": 1.8435676719873828e-05, | |
| "loss": 0.1848, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.21699819168173598, | |
| "grad_norm": 1.646269811867694, | |
| "learning_rate": 1.830648511318223e-05, | |
| "loss": 0.1726, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.22423146473779385, | |
| "grad_norm": 1.3277544792281775, | |
| "learning_rate": 1.817265910599978e-05, | |
| "loss": 0.1492, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.2314647377938517, | |
| "grad_norm": 1.52978658678254, | |
| "learning_rate": 1.8034273363278615e-05, | |
| "loss": 0.1964, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.23869801084990958, | |
| "grad_norm": 1.8231843149019955, | |
| "learning_rate": 1.789140509396394e-05, | |
| "loss": 0.2086, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.24593128390596744, | |
| "grad_norm": 1.8149511918444887, | |
| "learning_rate": 1.7744134007917195e-05, | |
| "loss": 0.2206, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.25316455696202533, | |
| "grad_norm": 1.7081708732378678, | |
| "learning_rate": 1.7592542271443888e-05, | |
| "loss": 0.2302, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.2603978300180832, | |
| "grad_norm": 1.4494531226173395, | |
| "learning_rate": 1.74367144614509e-05, | |
| "loss": 0.1969, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.26763110307414106, | |
| "grad_norm": 1.4435352798148773, | |
| "learning_rate": 1.7276737518258865e-05, | |
| "loss": 0.1662, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.27486437613019893, | |
| "grad_norm": 1.5398089087755158, | |
| "learning_rate": 1.7112700697095955e-05, | |
| "loss": 0.1961, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.2820976491862568, | |
| "grad_norm": 1.5371251862624924, | |
| "learning_rate": 1.6944695518300087e-05, | |
| "loss": 0.1824, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.28933092224231466, | |
| "grad_norm": 1.2387761223387177, | |
| "learning_rate": 1.6772815716257414e-05, | |
| "loss": 0.1691, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2965641952983725, | |
| "grad_norm": 1.4819991639273267, | |
| "learning_rate": 1.6597157187105475e-05, | |
| "loss": 0.1817, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.3037974683544304, | |
| "grad_norm": 1.1328162832418003, | |
| "learning_rate": 1.6417817935230318e-05, | |
| "loss": 0.1418, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.31103074141048825, | |
| "grad_norm": 1.6609923134511753, | |
| "learning_rate": 1.6234898018587336e-05, | |
| "loss": 0.2262, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.3182640144665461, | |
| "grad_norm": 1.101183253773407, | |
| "learning_rate": 1.6048499492876378e-05, | |
| "loss": 0.1134, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.325497287522604, | |
| "grad_norm": 1.735418027034885, | |
| "learning_rate": 1.5858726354602248e-05, | |
| "loss": 0.175, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.33273056057866185, | |
| "grad_norm": 1.1925277558480014, | |
| "learning_rate": 1.5665684483052425e-05, | |
| "loss": 0.1415, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3399638336347197, | |
| "grad_norm": 1.7164775271513815, | |
| "learning_rate": 1.5469481581224274e-05, | |
| "loss": 0.2229, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3471971066907776, | |
| "grad_norm": 1.54944673239779, | |
| "learning_rate": 1.527022711573479e-05, | |
| "loss": 0.2009, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.35443037974683544, | |
| "grad_norm": 1.052952493457072, | |
| "learning_rate": 1.50680322557464e-05, | |
| "loss": 0.1244, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3616636528028933, | |
| "grad_norm": 1.2955834096511674, | |
| "learning_rate": 1.4863009810942814e-05, | |
| "loss": 0.1426, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3688969258589512, | |
| "grad_norm": 1.0647479266524689, | |
| "learning_rate": 1.4655274168589635e-05, | |
| "loss": 0.1171, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.37613019891500904, | |
| "grad_norm": 1.3531117302986653, | |
| "learning_rate": 1.444494122971476e-05, | |
| "loss": 0.144, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3833634719710669, | |
| "grad_norm": 1.5396377899251914, | |
| "learning_rate": 1.4232128344444251e-05, | |
| "loss": 0.1426, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.39059674502712477, | |
| "grad_norm": 1.6556906631060258, | |
| "learning_rate": 1.4016954246529697e-05, | |
| "loss": 0.1834, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.39783001808318263, | |
| "grad_norm": 1.4971003634614435, | |
| "learning_rate": 1.37995389871036e-05, | |
| "loss": 0.1696, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.4050632911392405, | |
| "grad_norm": 1.3952013105051067, | |
| "learning_rate": 1.3580003867699801e-05, | |
| "loss": 0.148, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.41229656419529837, | |
| "grad_norm": 1.146527628567494, | |
| "learning_rate": 1.3358471372576229e-05, | |
| "loss": 0.1225, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.41952983725135623, | |
| "grad_norm": 1.1979689508816695, | |
| "learning_rate": 1.3135065100377816e-05, | |
| "loss": 0.1629, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.4267631103074141, | |
| "grad_norm": 1.361570996632967, | |
| "learning_rate": 1.2909909695177647e-05, | |
| "loss": 0.1762, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.43399638336347196, | |
| "grad_norm": 1.459216322827295, | |
| "learning_rate": 1.268313077693485e-05, | |
| "loss": 0.1823, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.4412296564195298, | |
| "grad_norm": 0.9032128659262652, | |
| "learning_rate": 1.2454854871407993e-05, | |
| "loss": 0.1004, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.4484629294755877, | |
| "grad_norm": 1.2429585892566153, | |
| "learning_rate": 1.2225209339563144e-05, | |
| "loss": 0.1431, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.45569620253164556, | |
| "grad_norm": 1.2516084012000934, | |
| "learning_rate": 1.1994322306515926e-05, | |
| "loss": 0.1623, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.4629294755877034, | |
| "grad_norm": 1.196502381130855, | |
| "learning_rate": 1.176232259004722e-05, | |
| "loss": 0.1329, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.4701627486437613, | |
| "grad_norm": 1.3204205518835002, | |
| "learning_rate": 1.1529339628732462e-05, | |
| "loss": 0.1617, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.47739602169981915, | |
| "grad_norm": 1.033121940148869, | |
| "learning_rate": 1.1295503409724526e-05, | |
| "loss": 0.1311, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.484629294755877, | |
| "grad_norm": 1.169007325201895, | |
| "learning_rate": 1.1060944396230583e-05, | |
| "loss": 0.1597, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.4918625678119349, | |
| "grad_norm": 1.256316293799304, | |
| "learning_rate": 1.0825793454723325e-05, | |
| "loss": 0.1558, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.49909584086799275, | |
| "grad_norm": 0.9373040893556883, | |
| "learning_rate": 1.0590181781927229e-05, | |
| "loss": 0.1246, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.5063291139240507, | |
| "grad_norm": 1.193004996300962, | |
| "learning_rate": 1.0354240831620542e-05, | |
| "loss": 0.1626, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.5135623869801085, | |
| "grad_norm": 1.2648937590042812, | |
| "learning_rate": 1.0118102241293848e-05, | |
| "loss": 0.1252, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.5207956600361664, | |
| "grad_norm": 1.2964585274228753, | |
| "learning_rate": 9.881897758706155e-06, | |
| "loss": 0.1794, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.5280289330922242, | |
| "grad_norm": 1.0311616305251576, | |
| "learning_rate": 9.645759168379463e-06, | |
| "loss": 0.1029, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.5352622061482821, | |
| "grad_norm": 1.4796928099723654, | |
| "learning_rate": 9.409818218072774e-06, | |
| "loss": 0.1932, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.5424954792043399, | |
| "grad_norm": 1.422580485613523, | |
| "learning_rate": 9.174206545276678e-06, | |
| "loss": 0.138, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.5497287522603979, | |
| "grad_norm": 1.2903416870852855, | |
| "learning_rate": 8.93905560376942e-06, | |
| "loss": 0.1405, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.5569620253164557, | |
| "grad_norm": 0.8368005903323104, | |
| "learning_rate": 8.704496590275479e-06, | |
| "loss": 0.0839, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.5641952983725136, | |
| "grad_norm": 1.462710303908404, | |
| "learning_rate": 8.47066037126754e-06, | |
| "loss": 0.1234, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 1.5099630306756744, | |
| "learning_rate": 8.237677409952784e-06, | |
| "loss": 0.1632, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5786618444846293, | |
| "grad_norm": 1.1554759264435244, | |
| "learning_rate": 8.005677693484077e-06, | |
| "loss": 0.1189, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5858951175406871, | |
| "grad_norm": 0.9519115511119977, | |
| "learning_rate": 7.774790660436857e-06, | |
| "loss": 0.1184, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.593128390596745, | |
| "grad_norm": 1.1753813368542916, | |
| "learning_rate": 7.545145128592009e-06, | |
| "loss": 0.1469, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.6003616636528029, | |
| "grad_norm": 0.8852628329465493, | |
| "learning_rate": 7.316869223065156e-06, | |
| "loss": 0.1052, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.6075949367088608, | |
| "grad_norm": 1.2158386219830792, | |
| "learning_rate": 7.090090304822356e-06, | |
| "loss": 0.1615, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.6148282097649186, | |
| "grad_norm": 1.2467724590478952, | |
| "learning_rate": 6.864934899622191e-06, | |
| "loss": 0.1818, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.6220614828209765, | |
| "grad_norm": 1.3636775085007256, | |
| "learning_rate": 6.6415286274237744e-06, | |
| "loss": 0.1395, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.6292947558770343, | |
| "grad_norm": 1.282395006899603, | |
| "learning_rate": 6.419996132300203e-06, | |
| "loss": 0.1565, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.6365280289330922, | |
| "grad_norm": 1.0471222095326378, | |
| "learning_rate": 6.200461012896401e-06, | |
| "loss": 0.1234, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.64376130198915, | |
| "grad_norm": 1.0937728472618042, | |
| "learning_rate": 5.983045753470308e-06, | |
| "loss": 0.1249, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.650994575045208, | |
| "grad_norm": 0.9151031662804701, | |
| "learning_rate": 5.7678716555557515e-06, | |
| "loss": 0.1127, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.6582278481012658, | |
| "grad_norm": 1.0701635271377732, | |
| "learning_rate": 5.5550587702852465e-06, | |
| "loss": 0.1147, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.6654611211573237, | |
| "grad_norm": 1.4324884227966928, | |
| "learning_rate": 5.344725831410369e-06, | |
| "loss": 0.1539, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.6726943942133815, | |
| "grad_norm": 1.1833796524874158, | |
| "learning_rate": 5.136990189057187e-06, | |
| "loss": 0.1473, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.6799276672694394, | |
| "grad_norm": 1.3412949266346563, | |
| "learning_rate": 4.931967744253601e-06, | |
| "loss": 0.1825, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.6871609403254972, | |
| "grad_norm": 1.1919515749555813, | |
| "learning_rate": 4.729772884265212e-06, | |
| "loss": 0.1645, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6943942133815552, | |
| "grad_norm": 1.121963526401964, | |
| "learning_rate": 4.530518418775734e-06, | |
| "loss": 0.1669, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.701627486437613, | |
| "grad_norm": 1.106624227413996, | |
| "learning_rate": 4.33431551694758e-06, | |
| "loss": 0.1636, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.7088607594936709, | |
| "grad_norm": 1.1060273425542941, | |
| "learning_rate": 4.1412736453977545e-06, | |
| "loss": 0.1634, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.7160940325497287, | |
| "grad_norm": 1.0995104844138277, | |
| "learning_rate": 3.9515005071236274e-06, | |
| "loss": 0.1342, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.7233273056057866, | |
| "grad_norm": 1.3721525349668775, | |
| "learning_rate": 3.7651019814126656e-06, | |
| "loss": 0.1339, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7305605786618445, | |
| "grad_norm": 1.3147221154801705, | |
| "learning_rate": 3.582182064769687e-06, | |
| "loss": 0.1577, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.7377938517179023, | |
| "grad_norm": 1.1356847248978508, | |
| "learning_rate": 3.402842812894529e-06, | |
| "loss": 0.1131, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.7450271247739603, | |
| "grad_norm": 1.0016288001929678, | |
| "learning_rate": 3.2271842837425917e-06, | |
| "loss": 0.1225, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.7522603978300181, | |
| "grad_norm": 1.3052781470084382, | |
| "learning_rate": 3.0553044816999133e-06, | |
| "loss": 0.1927, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.759493670886076, | |
| "grad_norm": 1.3624560033988535, | |
| "learning_rate": 2.8872993029040506e-06, | |
| "loss": 0.1553, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.7667269439421338, | |
| "grad_norm": 1.2446145740957406, | |
| "learning_rate": 2.723262481741138e-06, | |
| "loss": 0.1378, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.7739602169981917, | |
| "grad_norm": 1.5537240215768289, | |
| "learning_rate": 2.563285538549104e-06, | |
| "loss": 0.1908, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.7811934900542495, | |
| "grad_norm": 1.2200310855744627, | |
| "learning_rate": 2.407457728556115e-06, | |
| "loss": 0.1441, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.7884267631103075, | |
| "grad_norm": 1.3767110614355733, | |
| "learning_rate": 2.2558659920828095e-06, | |
| "loss": 0.1794, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.7956600361663653, | |
| "grad_norm": 1.3646317595886468, | |
| "learning_rate": 2.1085949060360654e-06, | |
| "loss": 0.1761, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.8028933092224232, | |
| "grad_norm": 0.7397893286023046, | |
| "learning_rate": 1.96572663672139e-06, | |
| "loss": 0.1045, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.810126582278481, | |
| "grad_norm": 1.0092396426988175, | |
| "learning_rate": 1.8273408940002202e-06, | |
| "loss": 0.1162, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.8173598553345389, | |
| "grad_norm": 1.668265792932794, | |
| "learning_rate": 1.693514886817772e-06, | |
| "loss": 0.1777, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.8245931283905967, | |
| "grad_norm": 1.3589428752926316, | |
| "learning_rate": 1.5643232801261731e-06, | |
| "loss": 0.1687, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.8318264014466547, | |
| "grad_norm": 0.8841209658700555, | |
| "learning_rate": 1.4398381532270001e-06, | |
| "loss": 0.1079, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.8390596745027125, | |
| "grad_norm": 0.8642771280686085, | |
| "learning_rate": 1.3201289595563693e-06, | |
| "loss": 0.1247, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.8462929475587704, | |
| "grad_norm": 1.1725446043997025, | |
| "learning_rate": 1.2052624879351105e-06, | |
| "loss": 0.1118, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.8535262206148282, | |
| "grad_norm": 0.8430507728068868, | |
| "learning_rate": 1.0953028253055541e-06, | |
| "loss": 0.1154, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.8607594936708861, | |
| "grad_norm": 1.1514692393692905, | |
| "learning_rate": 9.903113209758098e-07, | |
| "loss": 0.1643, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.8679927667269439, | |
| "grad_norm": 1.0531754579482937, | |
| "learning_rate": 8.903465523913957e-07, | |
| "loss": 0.1195, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.8752260397830018, | |
| "grad_norm": 1.1752942657410133, | |
| "learning_rate": 7.954642924533995e-07, | |
| "loss": 0.1576, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.8824593128390597, | |
| "grad_norm": 1.5113189404876857, | |
| "learning_rate": 7.057174784013432e-07, | |
| "loss": 0.1954, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.8896925858951176, | |
| "grad_norm": 0.9763145279668668, | |
| "learning_rate": 6.211561822781476e-07, | |
| "loss": 0.1358, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.8969258589511754, | |
| "grad_norm": 1.0884981959132525, | |
| "learning_rate": 5.418275829936537e-07, | |
| "loss": 0.1455, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.9041591320072333, | |
| "grad_norm": 1.1897400189233847, | |
| "learning_rate": 4.6777594000230855e-07, | |
| "loss": 0.1669, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.9113924050632911, | |
| "grad_norm": 1.225645214005749, | |
| "learning_rate": 3.9904256860967436e-07, | |
| "loss": 0.149, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.918625678119349, | |
| "grad_norm": 1.465044305909909, | |
| "learning_rate": 3.356658169215743e-07, | |
| "loss": 0.1481, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.9258589511754068, | |
| "grad_norm": 1.2424255190165467, | |
| "learning_rate": 2.776810444486944e-07, | |
| "loss": 0.1579, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.9330922242314648, | |
| "grad_norm": 1.2063094148288678, | |
| "learning_rate": 2.2512060237861455e-07, | |
| "loss": 0.1326, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.9403254972875226, | |
| "grad_norm": 1.0826240342833202, | |
| "learning_rate": 1.7801381552624565e-07, | |
| "loss": 0.1157, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.9475587703435805, | |
| "grad_norm": 1.2012754222257493, | |
| "learning_rate": 1.3638696597277678e-07, | |
| "loss": 0.1554, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.9547920433996383, | |
| "grad_norm": 1.0701124216474254, | |
| "learning_rate": 1.0026327840221728e-07, | |
| "loss": 0.1543, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.9620253164556962, | |
| "grad_norm": 1.0800203595933195, | |
| "learning_rate": 6.966290714375934e-08, | |
| "loss": 0.1484, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.969258589511754, | |
| "grad_norm": 1.0589626716764733, | |
| "learning_rate": 4.460292492716512e-08, | |
| "loss": 0.1635, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.976491862567812, | |
| "grad_norm": 0.7207109463474409, | |
| "learning_rate": 2.509731335744281e-08, | |
| "loss": 0.0853, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.9837251356238698, | |
| "grad_norm": 1.3772819365603755, | |
| "learning_rate": 1.1156955114162149e-08, | |
| "loss": 0.1777, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.9909584086799277, | |
| "grad_norm": 1.4747350400123909, | |
| "learning_rate": 2.7896278797256983e-09, | |
| "loss": 0.1516, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.9981916817359855, | |
| "grad_norm": 1.3893837186428997, | |
| "learning_rate": 0.0, | |
| "loss": 0.1907, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.9981916817359855, | |
| "step": 138, | |
| "total_flos": 283290642284544.0, | |
| "train_loss": 0.2102620561701664, | |
| "train_runtime": 1187.4984, | |
| "train_samples_per_second": 14.899, | |
| "train_steps_per_second": 0.116 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 138, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 283290642284544.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |