| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.987613542526837, | |
| "eval_steps": 500, | |
| "global_step": 1510, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 6.482018947601318, | |
| "learning_rate": 4.768211920529802e-06, | |
| "loss": 5.8299, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 4.826422691345215, | |
| "learning_rate": 9.735099337748345e-06, | |
| "loss": 5.1849, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 5.315689563751221, | |
| "learning_rate": 1.4701986754966887e-05, | |
| "loss": 4.4164, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 2.3576250076293945, | |
| "learning_rate": 1.9668874172185433e-05, | |
| "loss": 3.5949, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 1.602609634399414, | |
| "learning_rate": 2.4635761589403972e-05, | |
| "loss": 2.9948, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.4007989168167114, | |
| "learning_rate": 2.9602649006622518e-05, | |
| "loss": 2.6452, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 2.107353687286377, | |
| "learning_rate": 2.9492273730684327e-05, | |
| "loss": 2.4714, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 1.3134864568710327, | |
| "learning_rate": 2.894039735099338e-05, | |
| "loss": 2.3634, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 1.580680251121521, | |
| "learning_rate": 2.838852097130243e-05, | |
| "loss": 2.3032, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 1.3618550300598145, | |
| "learning_rate": 2.783664459161148e-05, | |
| "loss": 2.2501, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 1.531614899635315, | |
| "learning_rate": 2.728476821192053e-05, | |
| "loss": 2.2265, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 1.4849926233291626, | |
| "learning_rate": 2.673289183222958e-05, | |
| "loss": 2.1844, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 1.6642751693725586, | |
| "learning_rate": 2.6181015452538634e-05, | |
| "loss": 2.1166, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 1.7284085750579834, | |
| "learning_rate": 2.5629139072847685e-05, | |
| "loss": 2.1233, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 1.6671867370605469, | |
| "learning_rate": 2.5077262693156735e-05, | |
| "loss": 2.1212, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 1.756274938583374, | |
| "learning_rate": 2.4525386313465785e-05, | |
| "loss": 2.0726, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 1.8096344470977783, | |
| "learning_rate": 2.3973509933774835e-05, | |
| "loss": 2.0875, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 1.7899348735809326, | |
| "learning_rate": 2.3421633554083885e-05, | |
| "loss": 2.0833, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 1.917384386062622, | |
| "learning_rate": 2.286975717439294e-05, | |
| "loss": 2.0743, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 1.8895052671432495, | |
| "learning_rate": 2.231788079470199e-05, | |
| "loss": 2.0609, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 1.9963146448135376, | |
| "learning_rate": 2.176600441501104e-05, | |
| "loss": 2.0656, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 1.9002119302749634, | |
| "learning_rate": 2.121412803532009e-05, | |
| "loss": 2.0541, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 1.874874472618103, | |
| "learning_rate": 2.066225165562914e-05, | |
| "loss": 2.0416, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 1.7960671186447144, | |
| "learning_rate": 2.011037527593819e-05, | |
| "loss": 2.0311, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 2.1183598041534424, | |
| "learning_rate": 1.9558498896247242e-05, | |
| "loss": 2.0009, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 1.9861488342285156, | |
| "learning_rate": 1.9006622516556292e-05, | |
| "loss": 1.9688, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 2.235743284225464, | |
| "learning_rate": 1.8454746136865343e-05, | |
| "loss": 1.9989, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "grad_norm": 1.955734133720398, | |
| "learning_rate": 1.7902869757174393e-05, | |
| "loss": 1.9733, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "grad_norm": 2.044551372528076, | |
| "learning_rate": 1.7350993377483443e-05, | |
| "loss": 1.9574, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 2.117047071456909, | |
| "learning_rate": 1.6799116997792496e-05, | |
| "loss": 1.9654, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 2.02316951751709, | |
| "learning_rate": 1.6247240618101546e-05, | |
| "loss": 1.9498, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 2.2580747604370117, | |
| "learning_rate": 1.5695364238410596e-05, | |
| "loss": 1.9748, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 2.0818331241607666, | |
| "learning_rate": 1.5143487858719647e-05, | |
| "loss": 1.984, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 2.1940929889678955, | |
| "learning_rate": 1.4591611479028698e-05, | |
| "loss": 1.9646, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 2.1697773933410645, | |
| "learning_rate": 1.4039735099337748e-05, | |
| "loss": 1.958, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 2.0976200103759766, | |
| "learning_rate": 1.3487858719646799e-05, | |
| "loss": 1.946, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "grad_norm": 2.160543441772461, | |
| "learning_rate": 1.293598233995585e-05, | |
| "loss": 1.9398, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "grad_norm": 2.161280393600464, | |
| "learning_rate": 1.23841059602649e-05, | |
| "loss": 1.9008, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "grad_norm": 2.2662644386291504, | |
| "learning_rate": 1.183222958057395e-05, | |
| "loss": 1.9111, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 2.374934673309326, | |
| "learning_rate": 1.1280353200883002e-05, | |
| "loss": 1.9087, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "grad_norm": 2.3813648223876953, | |
| "learning_rate": 1.0728476821192052e-05, | |
| "loss": 1.9116, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 3.47, | |
| "grad_norm": 2.115699529647827, | |
| "learning_rate": 1.0176600441501102e-05, | |
| "loss": 1.92, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "grad_norm": 2.41668963432312, | |
| "learning_rate": 9.624724061810154e-06, | |
| "loss": 1.9452, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "grad_norm": 2.383557081222534, | |
| "learning_rate": 9.072847682119206e-06, | |
| "loss": 1.9209, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.72, | |
| "grad_norm": 2.2272610664367676, | |
| "learning_rate": 8.520971302428258e-06, | |
| "loss": 1.9057, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 2.5420844554901123, | |
| "learning_rate": 7.969094922737308e-06, | |
| "loss": 1.9025, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.88, | |
| "grad_norm": 2.2384705543518066, | |
| "learning_rate": 7.417218543046358e-06, | |
| "loss": 1.9072, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "grad_norm": 2.2767913341522217, | |
| "learning_rate": 6.865342163355408e-06, | |
| "loss": 1.8947, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "grad_norm": 2.2728919982910156, | |
| "learning_rate": 6.313465783664459e-06, | |
| "loss": 1.8819, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 4.13, | |
| "grad_norm": 2.4492197036743164, | |
| "learning_rate": 5.76158940397351e-06, | |
| "loss": 1.8914, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 4.21, | |
| "grad_norm": 2.238704204559326, | |
| "learning_rate": 5.20971302428256e-06, | |
| "loss": 1.8723, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "grad_norm": 2.341646194458008, | |
| "learning_rate": 4.657836644591612e-06, | |
| "loss": 1.8808, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.38, | |
| "grad_norm": 2.2996695041656494, | |
| "learning_rate": 4.105960264900663e-06, | |
| "loss": 1.88, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 4.46, | |
| "grad_norm": 2.213731050491333, | |
| "learning_rate": 3.554083885209713e-06, | |
| "loss": 1.8597, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 4.54, | |
| "grad_norm": 2.0925798416137695, | |
| "learning_rate": 3.002207505518764e-06, | |
| "loss": 1.853, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "grad_norm": 2.3262617588043213, | |
| "learning_rate": 2.450331125827815e-06, | |
| "loss": 1.8795, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "grad_norm": 2.3422157764434814, | |
| "learning_rate": 1.8984547461368651e-06, | |
| "loss": 1.865, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "grad_norm": 2.6429784297943115, | |
| "learning_rate": 1.3465783664459163e-06, | |
| "loss": 1.8871, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 4.87, | |
| "grad_norm": 2.334404945373535, | |
| "learning_rate": 7.947019867549669e-07, | |
| "loss": 1.8886, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 4.95, | |
| "grad_norm": 2.4886159896850586, | |
| "learning_rate": 2.428256070640177e-07, | |
| "loss": 1.8998, | |
| "step": 1500 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 1510, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 7.427503857598464e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |