| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.980952380952381, | |
| "eval_steps": 500, | |
| "global_step": 104, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 5e-06, | |
| "loss": 4.4886, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 1e-05, | |
| "loss": 4.3555, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 3.6464, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 2e-05, | |
| "loss": 2.3598, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 1.9995065603657317e-05, | |
| "loss": 1.4315, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 1.9980267284282718e-05, | |
| "loss": 0.6266, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 1.99556196460308e-05, | |
| "loss": 0.4461, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 1.9921147013144782e-05, | |
| "loss": 0.3908, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 1.9876883405951378e-05, | |
| "loss": 0.4065, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 1.982287250728689e-05, | |
| "loss": 0.3487, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 1.9759167619387474e-05, | |
| "loss": 0.3474, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 1.9685831611286312e-05, | |
| "loss": 0.3515, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 1.9602936856769432e-05, | |
| "loss": 0.3129, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 1.9510565162951538e-05, | |
| "loss": 0.3053, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 1.9408807689542257e-05, | |
| "loss": 0.2763, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 1.9297764858882516e-05, | |
| "loss": 0.3204, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 1.9177546256839814e-05, | |
| "loss": 0.2495, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 1.9048270524660197e-05, | |
| "loss": 0.2724, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 1.891006524188368e-05, | |
| "loss": 0.246, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 1.8763066800438638e-05, | |
| "loss": 0.2728, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 1.860742027003944e-05, | |
| "loss": 0.2043, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 1.8443279255020153e-05, | |
| "loss": 0.2391, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 1.827080574274562e-05, | |
| "loss": 0.2308, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 1.8090169943749477e-05, | |
| "loss": 0.2563, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 1.7901550123756906e-05, | |
| "loss": 0.2158, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 1.7705132427757895e-05, | |
| "loss": 0.1841, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 1.7501110696304598e-05, | |
| "loss": 0.2227, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 1.7289686274214116e-05, | |
| "loss": 0.2535, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 1.7071067811865477e-05, | |
| "loss": 0.2374, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 1.684547105928689e-05, | |
| "loss": 0.2521, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 1.661311865323652e-05, | |
| "loss": 0.2235, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.63742398974869e-05, | |
| "loss": 0.2182, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.6129070536529767e-05, | |
| "loss": 0.2209, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.5877852522924733e-05, | |
| "loss": 0.1769, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.5620833778521306e-05, | |
| "loss": 0.1795, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.5358267949789968e-05, | |
| "loss": 0.1864, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.5090414157503715e-05, | |
| "loss": 0.1636, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 1.4817536741017153e-05, | |
| "loss": 0.1761, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 1.4539904997395468e-05, | |
| "loss": 0.1851, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 1.4257792915650728e-05, | |
| "loss": 0.2, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 1.3971478906347806e-05, | |
| "loss": 0.1913, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 1.3681245526846782e-05, | |
| "loss": 0.1821, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 1.3387379202452917e-05, | |
| "loss": 0.1956, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 1.3090169943749475e-05, | |
| "loss": 0.1732, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 1.2789911060392295e-05, | |
| "loss": 0.1666, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 1.2486898871648552e-05, | |
| "loss": 0.1479, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.2181432413965428e-05, | |
| "loss": 0.1501, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.187381314585725e-05, | |
| "loss": 0.1632, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 1.156434465040231e-05, | |
| "loss": 0.1672, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 1.1253332335643043e-05, | |
| "loss": 0.1497, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.0941083133185146e-05, | |
| "loss": 0.1863, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 1.0627905195293135e-05, | |
| "loss": 0.1461, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 1.0314107590781284e-05, | |
| "loss": 0.1628, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1345, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 9.685892409218718e-06, | |
| "loss": 0.1451, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 9.372094804706867e-06, | |
| "loss": 0.1364, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 9.058916866814857e-06, | |
| "loss": 0.123, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 8.746667664356957e-06, | |
| "loss": 0.121, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 8.43565534959769e-06, | |
| "loss": 0.1503, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 8.126186854142752e-06, | |
| "loss": 0.1112, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 7.818567586034578e-06, | |
| "loss": 0.1138, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 7.513101128351454e-06, | |
| "loss": 0.1525, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 7.210088939607709e-06, | |
| "loss": 0.1119, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 6.909830056250527e-06, | |
| "loss": 0.1044, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 6.612620797547087e-06, | |
| "loss": 0.1273, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 6.318754473153221e-06, | |
| "loss": 0.1192, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 6.028521093652195e-06, | |
| "loss": 0.1106, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 5.742207084349274e-06, | |
| "loss": 0.164, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 5.460095002604533e-06, | |
| "loss": 0.1123, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 5.1824632589828465e-06, | |
| "loss": 0.1235, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 4.909585842496287e-06, | |
| "loss": 0.108, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 4.641732050210032e-06, | |
| "loss": 0.1306, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 4.379166221478697e-06, | |
| "loss": 0.1087, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 4.12214747707527e-06, | |
| "loss": 0.1247, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 3.8709294634702374e-06, | |
| "loss": 0.1127, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 3.625760102513103e-06, | |
| "loss": 0.113, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 3.3868813467634833e-06, | |
| "loss": 0.102, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "learning_rate": 3.1545289407131128e-06, | |
| "loss": 0.1071, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 2.9289321881345257e-06, | |
| "loss": 0.1285, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 2.7103137257858867e-06, | |
| "loss": 0.1074, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 2.4988893036954045e-06, | |
| "loss": 0.1414, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "learning_rate": 2.2948675722421086e-06, | |
| "loss": 0.1405, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "learning_rate": 2.098449876243096e-06, | |
| "loss": 0.1313, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 1.9098300562505266e-06, | |
| "loss": 0.1096, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 1.7291942572543806e-06, | |
| "loss": 0.109, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 1.5567207449798517e-06, | |
| "loss": 0.1042, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "learning_rate": 1.3925797299605649e-06, | |
| "loss": 0.1155, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "learning_rate": 1.2369331995613664e-06, | |
| "loss": 0.095, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 1.0899347581163222e-06, | |
| "loss": 0.1146, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 9.517294753398066e-07, | |
| "loss": 0.1166, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "learning_rate": 8.224537431601886e-07, | |
| "loss": 0.0988, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "learning_rate": 7.022351411174866e-07, | |
| "loss": 0.1031, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "learning_rate": 5.911923104577455e-07, | |
| "loss": 0.089, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 4.894348370484648e-07, | |
| "loss": 0.1128, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "learning_rate": 3.9706314323056936e-07, | |
| "loss": 0.1201, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 3.1416838871368925e-07, | |
| "loss": 0.1249, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "learning_rate": 2.4083238061252565e-07, | |
| "loss": 0.1032, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 1.7712749271311392e-07, | |
| "loss": 0.1126, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 1.231165940486234e-07, | |
| "loss": 0.0813, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "learning_rate": 7.885298685522235e-08, | |
| "loss": 0.1213, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 4.438035396920004e-08, | |
| "loss": 0.0951, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "learning_rate": 1.973271571728441e-08, | |
| "loss": 0.0943, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 4.9343963426840006e-09, | |
| "loss": 0.1023, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "learning_rate": 0.0, | |
| "loss": 0.1173, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "step": 104, | |
| "total_flos": 6593597865984.0, | |
| "train_loss": 0.32519721426069736, | |
| "train_runtime": 1026.2127, | |
| "train_samples_per_second": 6.488, | |
| "train_steps_per_second": 0.101 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 104, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 50000, | |
| "total_flos": 6593597865984.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |