| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 1730, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 4.66322660446167, | |
| "learning_rate": 2.8901734104046245e-05, | |
| "loss": 2.8724, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 4.364796161651611, | |
| "learning_rate": 5.780346820809249e-05, | |
| "loss": 1.8669, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 7.0544962882995605, | |
| "learning_rate": 8.670520231213874e-05, | |
| "loss": 1.0394, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 2.588165521621704, | |
| "learning_rate": 0.00011560693641618498, | |
| "loss": 0.8095, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 2.311723470687866, | |
| "learning_rate": 0.00014450867052023122, | |
| "loss": 0.7673, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 2.0632615089416504, | |
| "learning_rate": 0.00017341040462427748, | |
| "loss": 0.7738, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 1.7875710725784302, | |
| "learning_rate": 0.00019974309569685292, | |
| "loss": 0.7151, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 1.7314724922180176, | |
| "learning_rate": 0.00019653179190751445, | |
| "loss": 0.6782, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 2.1410841941833496, | |
| "learning_rate": 0.00019332048811817598, | |
| "loss": 0.6733, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 1.6796475648880005, | |
| "learning_rate": 0.0001901091843288375, | |
| "loss": 0.711, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 1.9651137590408325, | |
| "learning_rate": 0.00018689788053949903, | |
| "loss": 0.6285, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 1.9334725141525269, | |
| "learning_rate": 0.00018368657675016056, | |
| "loss": 0.6387, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 1.9697507619857788, | |
| "learning_rate": 0.00018047527296082209, | |
| "loss": 0.6203, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 1.5927447080612183, | |
| "learning_rate": 0.0001772639691714836, | |
| "loss": 0.6036, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 1.5775573253631592, | |
| "learning_rate": 0.00017405266538214514, | |
| "loss": 0.5443, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 2.265123128890991, | |
| "learning_rate": 0.0001708413615928067, | |
| "loss": 0.5864, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 1.9081721305847168, | |
| "learning_rate": 0.00016763005780346822, | |
| "loss": 0.497, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 1.7822768688201904, | |
| "learning_rate": 0.00016441875401412975, | |
| "loss": 0.5652, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 1.585552453994751, | |
| "learning_rate": 0.00016120745022479128, | |
| "loss": 0.5356, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 1.8936752080917358, | |
| "learning_rate": 0.0001579961464354528, | |
| "loss": 0.4897, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 1.7502835988998413, | |
| "learning_rate": 0.00015478484264611433, | |
| "loss": 0.5051, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 1.6106928586959839, | |
| "learning_rate": 0.00015157353885677586, | |
| "loss": 0.5034, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 1.6488401889801025, | |
| "learning_rate": 0.00014836223506743738, | |
| "loss": 0.4692, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 1.5491576194763184, | |
| "learning_rate": 0.0001451509312780989, | |
| "loss": 0.5016, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 1.7753480672836304, | |
| "learning_rate": 0.00014193962748876044, | |
| "loss": 0.4836, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 1.5605947971343994, | |
| "learning_rate": 0.00013872832369942197, | |
| "loss": 0.5059, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 1.4864176511764526, | |
| "learning_rate": 0.0001355170199100835, | |
| "loss": 0.4667, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 1.6282180547714233, | |
| "learning_rate": 0.00013230571612074502, | |
| "loss": 0.4208, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 1.9214680194854736, | |
| "learning_rate": 0.00012909441233140655, | |
| "loss": 0.3733, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 1.9027807712554932, | |
| "learning_rate": 0.00012588310854206808, | |
| "loss": 0.3831, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 1.756855845451355, | |
| "learning_rate": 0.0001226718047527296, | |
| "loss": 0.3876, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "grad_norm": 1.5269505977630615, | |
| "learning_rate": 0.00011946050096339114, | |
| "loss": 0.3847, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 1.7476940155029297, | |
| "learning_rate": 0.00011624919717405267, | |
| "loss": 0.3604, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 1.8043824434280396, | |
| "learning_rate": 0.0001130378933847142, | |
| "loss": 0.384, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 1.4151132106781006, | |
| "learning_rate": 0.00010982658959537572, | |
| "loss": 0.3745, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 1.907926082611084, | |
| "learning_rate": 0.00010661528580603725, | |
| "loss": 0.372, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 1.8263049125671387, | |
| "learning_rate": 0.00010340398201669879, | |
| "loss": 0.3972, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 1.8957775831222534, | |
| "learning_rate": 0.00010019267822736032, | |
| "loss": 0.3967, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 1.772609829902649, | |
| "learning_rate": 9.698137443802185e-05, | |
| "loss": 0.3775, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 1.976365566253662, | |
| "learning_rate": 9.377007064868337e-05, | |
| "loss": 0.4043, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 1.689432978630066, | |
| "learning_rate": 9.05587668593449e-05, | |
| "loss": 0.3226, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 1.9092849493026733, | |
| "learning_rate": 8.734746307000643e-05, | |
| "loss": 0.3758, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "grad_norm": 1.7606756687164307, | |
| "learning_rate": 8.413615928066796e-05, | |
| "loss": 0.3095, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "grad_norm": 1.8509767055511475, | |
| "learning_rate": 8.092485549132948e-05, | |
| "loss": 0.2712, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 1.5414364337921143, | |
| "learning_rate": 7.771355170199101e-05, | |
| "loss": 0.2695, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "grad_norm": 1.8055434226989746, | |
| "learning_rate": 7.450224791265255e-05, | |
| "loss": 0.2828, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 2.0254344940185547, | |
| "learning_rate": 7.129094412331408e-05, | |
| "loss": 0.2729, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 3.47, | |
| "grad_norm": 2.2866463661193848, | |
| "learning_rate": 6.80796403339756e-05, | |
| "loss": 0.3022, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "grad_norm": 1.5647475719451904, | |
| "learning_rate": 6.486833654463712e-05, | |
| "loss": 0.2893, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "grad_norm": 1.9078121185302734, | |
| "learning_rate": 6.165703275529865e-05, | |
| "loss": 0.2751, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "grad_norm": 1.9310845136642456, | |
| "learning_rate": 5.844572896596018e-05, | |
| "loss": 0.2614, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "grad_norm": 1.5173165798187256, | |
| "learning_rate": 5.523442517662171e-05, | |
| "loss": 0.2781, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "grad_norm": 2.2908642292022705, | |
| "learning_rate": 5.2023121387283234e-05, | |
| "loss": 0.2889, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 2.0346012115478516, | |
| "learning_rate": 4.881181759794477e-05, | |
| "loss": 0.2997, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "grad_norm": 1.6963484287261963, | |
| "learning_rate": 4.56005138086063e-05, | |
| "loss": 0.2415, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "grad_norm": 1.552463173866272, | |
| "learning_rate": 4.238921001926782e-05, | |
| "loss": 0.229, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.12, | |
| "grad_norm": 2.3235974311828613, | |
| "learning_rate": 3.917790622992935e-05, | |
| "loss": 0.1927, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 4.19, | |
| "grad_norm": 1.6171674728393555, | |
| "learning_rate": 3.596660244059088e-05, | |
| "loss": 0.2116, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 4.26, | |
| "grad_norm": 1.9670432806015015, | |
| "learning_rate": 3.275529865125241e-05, | |
| "loss": 0.207, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "grad_norm": 2.1716980934143066, | |
| "learning_rate": 2.9543994861913938e-05, | |
| "loss": 0.2023, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.41, | |
| "grad_norm": 2.0103349685668945, | |
| "learning_rate": 2.6332691072575465e-05, | |
| "loss": 0.2081, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 2.488182544708252, | |
| "learning_rate": 2.3121387283236996e-05, | |
| "loss": 0.2206, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "grad_norm": 2.150543451309204, | |
| "learning_rate": 1.9910083493898523e-05, | |
| "loss": 0.1989, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "grad_norm": 1.952592134475708, | |
| "learning_rate": 1.6698779704560053e-05, | |
| "loss": 0.2074, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "grad_norm": 1.8111902475357056, | |
| "learning_rate": 1.348747591522158e-05, | |
| "loss": 0.2122, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 4.77, | |
| "grad_norm": 2.595923662185669, | |
| "learning_rate": 1.027617212588311e-05, | |
| "loss": 0.191, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 4.84, | |
| "grad_norm": 2.17901611328125, | |
| "learning_rate": 7.064868336544637e-06, | |
| "loss": 0.2205, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 4.91, | |
| "grad_norm": 2.284546136856079, | |
| "learning_rate": 3.853564547206165e-06, | |
| "loss": 0.2239, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 4.99, | |
| "grad_norm": 2.4415647983551025, | |
| "learning_rate": 6.422607578676943e-07, | |
| "loss": 0.1967, | |
| "step": 1725 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 1730, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 7.5248089823232e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |