| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.81132075471698, | |
| "eval_steps": 500, | |
| "global_step": 260, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03773584905660377, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 2.9895, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.18867924528301888, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 3.0316, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.37735849056603776, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 2.9227, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.5660377358490566, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 0.00011538461538461538, | |
| "loss": 2.5841, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.7547169811320755, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 0.00015384615384615385, | |
| "loss": 2.3138, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.9433962264150944, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 0.00019230769230769233, | |
| "loss": 2.0246, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.9811320754716981, | |
| "eval_loss": 2.6613192558288574, | |
| "eval_runtime": 0.7072, | |
| "eval_samples_per_second": 14.139, | |
| "eval_steps_per_second": 1.414, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.1320754716981132, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 0.00019985583705641418, | |
| "loss": 1.7376, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.320754716981132, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 0.0001992708874098054, | |
| "loss": 1.5704, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.509433962264151, | |
| "grad_norm": 0.498046875, | |
| "learning_rate": 0.00019823877374156647, | |
| "loss": 1.4624, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.6981132075471699, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00019676414519013781, | |
| "loss": 1.3792, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.8867924528301887, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00019485364419471454, | |
| "loss": 1.3202, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 2.540452480316162, | |
| "eval_runtime": 0.5421, | |
| "eval_samples_per_second": 18.446, | |
| "eval_steps_per_second": 1.845, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.0754716981132075, | |
| "grad_norm": 0.2314453125, | |
| "learning_rate": 0.00019251587657449236, | |
| "loss": 1.2675, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.2641509433962264, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 0.0001897613727639014, | |
| "loss": 1.2218, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.452830188679245, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00018660254037844388, | |
| "loss": 1.2149, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.641509433962264, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00018305360832480117, | |
| "loss": 1.179, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.830188679245283, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.0001791305627069662, | |
| "loss": 1.1694, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.981132075471698, | |
| "eval_loss": 2.5124902725219727, | |
| "eval_runtime": 0.6642, | |
| "eval_samples_per_second": 15.055, | |
| "eval_steps_per_second": 1.506, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 3.018867924528302, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00017485107481711012, | |
| "loss": 1.1675, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.207547169811321, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.00017023442153554777, | |
| "loss": 1.1345, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 3.3962264150943398, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 0.0001653013984983585, | |
| "loss": 1.1304, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.5849056603773586, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.0001600742264237979, | |
| "loss": 1.1194, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 3.7735849056603774, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00015457645101945046, | |
| "loss": 1.0993, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.9622641509433962, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.00014883283692099112, | |
| "loss": 1.1076, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 2.5138261318206787, | |
| "eval_runtime": 0.5374, | |
| "eval_samples_per_second": 18.609, | |
| "eval_steps_per_second": 1.861, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 4.150943396226415, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 0.00014286925614030542, | |
| "loss": 1.0832, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.339622641509434, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.00013671257152545277, | |
| "loss": 1.0718, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 4.528301886792453, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 0.0001303905157574247, | |
| "loss": 1.0737, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.716981132075472, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.0001239315664287558, | |
| "loss": 1.0708, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 4.90566037735849, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 1.0651, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 4.981132075471698, | |
| "eval_loss": 2.5085928440093994, | |
| "eval_runtime": 0.6726, | |
| "eval_samples_per_second": 14.868, | |
| "eval_steps_per_second": 1.487, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 5.09433962264151, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 0.00011071984957874479, | |
| "loss": 1.055, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 5.283018867924528, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 0.00010402659401094152, | |
| "loss": 1.0371, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 5.471698113207547, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 9.73152007189939e-05, | |
| "loss": 1.0452, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 5.660377358490566, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 9.061590105968208e-05, | |
| "loss": 1.0461, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 5.849056603773585, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 8.395887191422397e-05, | |
| "loss": 1.0394, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 2.5248091220855713, | |
| "eval_runtime": 0.5366, | |
| "eval_samples_per_second": 18.636, | |
| "eval_steps_per_second": 1.864, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 6.037735849056604, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 7.73740997570278e-05, | |
| "loss": 1.0302, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 6.226415094339623, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 7.089124558212871e-05, | |
| "loss": 1.0164, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 6.415094339622642, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 6.453951129574644e-05, | |
| "loss": 1.0265, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 6.60377358490566, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 5.834750817679606e-05, | |
| "loss": 1.019, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 6.7924528301886795, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 5.234312799786921e-05, | |
| "loss": 1.0159, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 6.981132075471698, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 4.6553417387219886e-05, | |
| "loss": 1.0232, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 6.981132075471698, | |
| "eval_loss": 2.526435136795044, | |
| "eval_runtime": 0.6455, | |
| "eval_samples_per_second": 15.492, | |
| "eval_steps_per_second": 1.549, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 7.169811320754717, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 4.100445599768774e-05, | |
| "loss": 1.0125, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 7.3584905660377355, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 3.5721239031346066e-05, | |
| "loss": 1.0212, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 7.547169811320755, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 3.072756464904006e-05, | |
| "loss": 1.0052, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 7.735849056603773, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 2.6045926771976303e-05, | |
| "loss": 1.0067, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 7.9245283018867925, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 2.1697413758237784e-05, | |
| "loss": 1.0042, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 2.5296316146850586, | |
| "eval_runtime": 0.534, | |
| "eval_samples_per_second": 18.728, | |
| "eval_steps_per_second": 1.873, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 8.11320754716981, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 1.7701613410634365e-05, | |
| "loss": 1.0077, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 8.30188679245283, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 1.4076524743778319e-05, | |
| "loss": 0.999, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 8.49056603773585, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 1.083847690782972e-05, | |
| "loss": 1.0027, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 8.679245283018869, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 8.002055634117578e-06, | |
| "loss": 0.9984, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 8.867924528301886, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 5.580037533961546e-06, | |
| "loss": 1.0109, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 8.981132075471699, | |
| "eval_loss": 2.531932830810547, | |
| "eval_runtime": 0.6509, | |
| "eval_samples_per_second": 15.363, | |
| "eval_steps_per_second": 1.536, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 9.056603773584905, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 3.5833325466437694e-06, | |
| "loss": 1.0008, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 9.245283018867925, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 2.0209347957732328e-06, | |
| "loss": 1.0094, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 9.433962264150944, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 8.998820754091531e-07, | |
| "loss": 0.9987, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 9.622641509433961, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 2.2522414843748618e-07, | |
| "loss": 1.0006, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 9.81132075471698, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.0, | |
| "loss": 1.0064, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 9.81132075471698, | |
| "eval_loss": 2.531856060028076, | |
| "eval_runtime": 0.5289, | |
| "eval_samples_per_second": 18.906, | |
| "eval_steps_per_second": 1.891, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 9.81132075471698, | |
| "step": 260, | |
| "total_flos": 2.036227620470784e+17, | |
| "train_loss": 1.2490827266986553, | |
| "train_runtime": 901.3115, | |
| "train_samples_per_second": 18.473, | |
| "train_steps_per_second": 0.288 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 260, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 100, | |
| "total_flos": 2.036227620470784e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |