| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 25.778732545649838, | |
| "eval_steps": 500, | |
| "global_step": 1500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.6707318425178528, | |
| "learning_rate": 1.97816091954023e-05, | |
| "loss": 2.2888, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": NaN, | |
| "learning_rate": 1.9574712643678162e-05, | |
| "loss": 2.0848, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 0.7721680402755737, | |
| "learning_rate": 1.9344827586206897e-05, | |
| "loss": 2.0044, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 1.1140433549880981, | |
| "learning_rate": 1.9126436781609195e-05, | |
| "loss": 1.8016, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 0.7205075621604919, | |
| "learning_rate": 1.8896551724137934e-05, | |
| "loss": 1.7217, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 0.8933233618736267, | |
| "learning_rate": 1.866666666666667e-05, | |
| "loss": 1.5705, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "grad_norm": 0.7114273905754089, | |
| "learning_rate": 1.8436781609195404e-05, | |
| "loss": 1.4006, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 0.7229479551315308, | |
| "learning_rate": 1.820689655172414e-05, | |
| "loss": 1.3137, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "grad_norm": 0.9370490908622742, | |
| "learning_rate": 1.7977011494252874e-05, | |
| "loss": 1.1898, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "grad_norm": 0.6051978468894958, | |
| "learning_rate": 1.774712643678161e-05, | |
| "loss": 1.1229, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.78, | |
| "grad_norm": 0.6857028007507324, | |
| "learning_rate": 1.7517241379310347e-05, | |
| "loss": 1.051, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 4.12, | |
| "grad_norm": 0.6715748310089111, | |
| "learning_rate": 1.7287356321839082e-05, | |
| "loss": 0.9894, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.47, | |
| "grad_norm": 0.5918118953704834, | |
| "learning_rate": 1.7057471264367817e-05, | |
| "loss": 0.9687, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.81, | |
| "grad_norm": 0.6621690392494202, | |
| "learning_rate": 1.6827586206896552e-05, | |
| "loss": 0.9199, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 5.16, | |
| "grad_norm": 0.6697206497192383, | |
| "learning_rate": 1.659770114942529e-05, | |
| "loss": 0.9303, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 0.8184316158294678, | |
| "learning_rate": 1.6367816091954025e-05, | |
| "loss": 0.8898, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 5.84, | |
| "grad_norm": 0.6429987549781799, | |
| "learning_rate": 1.613793103448276e-05, | |
| "loss": 0.8623, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 6.19, | |
| "grad_norm": 0.7518043518066406, | |
| "learning_rate": 1.5908045977011495e-05, | |
| "loss": 0.8239, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 6.53, | |
| "grad_norm": 0.6667824983596802, | |
| "learning_rate": 1.567816091954023e-05, | |
| "loss": 0.8119, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 6.87, | |
| "grad_norm": 0.8569457530975342, | |
| "learning_rate": 1.5448275862068965e-05, | |
| "loss": 0.8139, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 7.22, | |
| "grad_norm": 0.7754850387573242, | |
| "learning_rate": 1.5218390804597702e-05, | |
| "loss": 0.7835, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 7.56, | |
| "grad_norm": 1.159196138381958, | |
| "learning_rate": 1.4988505747126439e-05, | |
| "loss": 0.7546, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 7.91, | |
| "grad_norm": 1.119764804840088, | |
| "learning_rate": 1.4758620689655174e-05, | |
| "loss": 0.7571, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 8.25, | |
| "grad_norm": 1.3600786924362183, | |
| "learning_rate": 1.452873563218391e-05, | |
| "loss": 0.7451, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 8.59, | |
| "grad_norm": 0.7608994245529175, | |
| "learning_rate": 1.4298850574712644e-05, | |
| "loss": 0.7109, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 8.94, | |
| "grad_norm": 1.0172290802001953, | |
| "learning_rate": 1.406896551724138e-05, | |
| "loss": 0.7228, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 9.28, | |
| "grad_norm": 1.042607069015503, | |
| "learning_rate": 1.3839080459770115e-05, | |
| "loss": 0.6939, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 9.62, | |
| "grad_norm": 0.8913071751594543, | |
| "learning_rate": 1.3609195402298852e-05, | |
| "loss": 0.6721, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 9.97, | |
| "grad_norm": 1.4283536672592163, | |
| "learning_rate": 1.3379310344827587e-05, | |
| "loss": 0.681, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 10.31, | |
| "grad_norm": 1.1445728540420532, | |
| "learning_rate": 1.3149425287356324e-05, | |
| "loss": 0.6484, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 10.66, | |
| "grad_norm": 1.425697684288025, | |
| "learning_rate": 1.2919540229885059e-05, | |
| "loss": 0.6558, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.8931305408477783, | |
| "learning_rate": 1.2689655172413795e-05, | |
| "loss": 0.6642, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 11.34, | |
| "grad_norm": 1.0374151468276978, | |
| "learning_rate": 1.2459770114942529e-05, | |
| "loss": 0.6202, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 11.69, | |
| "grad_norm": 1.628758430480957, | |
| "learning_rate": 1.2229885057471265e-05, | |
| "loss": 0.6163, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 12.03, | |
| "grad_norm": 1.3881452083587646, | |
| "learning_rate": 1.2e-05, | |
| "loss": 0.6364, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 12.37, | |
| "grad_norm": 1.0961302518844604, | |
| "learning_rate": 1.1770114942528737e-05, | |
| "loss": 0.5963, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 12.72, | |
| "grad_norm": 1.1812736988067627, | |
| "learning_rate": 1.1540229885057472e-05, | |
| "loss": 0.6102, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 13.06, | |
| "grad_norm": 1.103151559829712, | |
| "learning_rate": 1.1310344827586209e-05, | |
| "loss": 0.5965, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 13.4, | |
| "grad_norm": 1.108560562133789, | |
| "learning_rate": 1.1080459770114944e-05, | |
| "loss": 0.58, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 13.75, | |
| "grad_norm": 1.322364091873169, | |
| "learning_rate": 1.085057471264368e-05, | |
| "loss": 0.5707, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 14.09, | |
| "grad_norm": 1.2036404609680176, | |
| "learning_rate": 1.0620689655172414e-05, | |
| "loss": 0.5781, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 14.44, | |
| "grad_norm": 1.46902596950531, | |
| "learning_rate": 1.039080459770115e-05, | |
| "loss": 0.5413, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 14.78, | |
| "grad_norm": 0.9223589301109314, | |
| "learning_rate": 1.0160919540229885e-05, | |
| "loss": 0.5686, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 15.12, | |
| "grad_norm": 1.7452529668807983, | |
| "learning_rate": 9.931034482758622e-06, | |
| "loss": 0.5538, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 15.47, | |
| "grad_norm": 1.0680702924728394, | |
| "learning_rate": 9.701149425287357e-06, | |
| "loss": 0.5402, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 15.81, | |
| "grad_norm": 1.4106242656707764, | |
| "learning_rate": 9.471264367816094e-06, | |
| "loss": 0.5629, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 16.15, | |
| "grad_norm": 1.7341551780700684, | |
| "learning_rate": 9.241379310344829e-06, | |
| "loss": 0.5538, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 16.5, | |
| "grad_norm": 2.115643262863159, | |
| "learning_rate": 9.011494252873564e-06, | |
| "loss": 0.5481, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 16.84, | |
| "grad_norm": 1.1589787006378174, | |
| "learning_rate": 8.7816091954023e-06, | |
| "loss": 0.4981, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 17.19, | |
| "grad_norm": 1.0696042776107788, | |
| "learning_rate": 8.551724137931035e-06, | |
| "loss": 0.5041, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 17.53, | |
| "grad_norm": 1.892269253730774, | |
| "learning_rate": 8.32183908045977e-06, | |
| "loss": 0.4956, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 17.87, | |
| "grad_norm": 2.214688301086426, | |
| "learning_rate": 8.091954022988507e-06, | |
| "loss": 0.5363, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 18.22, | |
| "grad_norm": 1.271893858909607, | |
| "learning_rate": 7.862068965517242e-06, | |
| "loss": 0.5195, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 18.56, | |
| "grad_norm": 0.9383485913276672, | |
| "learning_rate": 7.632183908045979e-06, | |
| "loss": 0.5002, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 18.9, | |
| "grad_norm": 1.749745488166809, | |
| "learning_rate": 7.402298850574713e-06, | |
| "loss": 0.4841, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 19.25, | |
| "grad_norm": 1.2551345825195312, | |
| "learning_rate": 7.172413793103449e-06, | |
| "loss": 0.5033, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 19.59, | |
| "grad_norm": 1.0362412929534912, | |
| "learning_rate": 6.9425287356321845e-06, | |
| "loss": 0.5138, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 19.94, | |
| "grad_norm": 1.483361840248108, | |
| "learning_rate": 6.71264367816092e-06, | |
| "loss": 0.4826, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 20.28, | |
| "grad_norm": 1.792438268661499, | |
| "learning_rate": 6.482758620689655e-06, | |
| "loss": 0.4544, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 20.62, | |
| "grad_norm": 1.1103723049163818, | |
| "learning_rate": 6.252873563218391e-06, | |
| "loss": 0.5097, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 20.97, | |
| "grad_norm": 1.3724958896636963, | |
| "learning_rate": 6.022988505747127e-06, | |
| "loss": 0.4827, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 21.31, | |
| "grad_norm": 1.9351897239685059, | |
| "learning_rate": 5.793103448275863e-06, | |
| "loss": 0.4772, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 21.65, | |
| "grad_norm": 1.3644013404846191, | |
| "learning_rate": 5.563218390804598e-06, | |
| "loss": 0.485, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "grad_norm": 1.4544987678527832, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 0.468, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 22.34, | |
| "grad_norm": 1.7704741954803467, | |
| "learning_rate": 5.1034482758620695e-06, | |
| "loss": 0.4653, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 22.69, | |
| "grad_norm": 1.2633339166641235, | |
| "learning_rate": 4.873563218390805e-06, | |
| "loss": 0.477, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 23.03, | |
| "grad_norm": 1.43886399269104, | |
| "learning_rate": 4.643678160919541e-06, | |
| "loss": 0.4724, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 23.37, | |
| "grad_norm": 1.1346815824508667, | |
| "learning_rate": 4.413793103448276e-06, | |
| "loss": 0.4712, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 23.72, | |
| "grad_norm": 1.289389967918396, | |
| "learning_rate": 4.183908045977012e-06, | |
| "loss": 0.4661, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 24.06, | |
| "grad_norm": 1.477211833000183, | |
| "learning_rate": 3.954022988505747e-06, | |
| "loss": 0.4443, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 24.4, | |
| "grad_norm": 2.224083185195923, | |
| "learning_rate": 3.7241379310344832e-06, | |
| "loss": 0.4445, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 24.75, | |
| "grad_norm": 1.6573207378387451, | |
| "learning_rate": 3.4942528735632187e-06, | |
| "loss": 0.4756, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 25.09, | |
| "grad_norm": 2.269866466522217, | |
| "learning_rate": 3.2643678160919545e-06, | |
| "loss": 0.4535, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 25.44, | |
| "grad_norm": 1.2528423070907593, | |
| "learning_rate": 3.03448275862069e-06, | |
| "loss": 0.4881, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 25.78, | |
| "grad_norm": 1.2892448902130127, | |
| "learning_rate": 2.8045977011494257e-06, | |
| "loss": 0.4226, | |
| "step": 1500 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 1740, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 500, | |
| "total_flos": 3.8986916806656e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |