| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.1, | |
| "global_step": 100000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 5.5406, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 3.7083, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 6e-06, | |
| "loss": 3.0627, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 2.7998, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 1e-05, | |
| "loss": 2.6425, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 1.2e-05, | |
| "loss": 2.5589, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 2.4647, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 2.432, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 1.8e-05, | |
| "loss": 2.3395, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 2e-05, | |
| "loss": 2.268, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 2.1699, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 2.4e-05, | |
| "loss": 2.2278, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 1.8529, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 1.7047, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6762, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 2.1632, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 2.0574, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 3.6e-05, | |
| "loss": 2.0309, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 3.8e-05, | |
| "loss": 2.0596, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 4e-05, | |
| "loss": 1.923, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 4.2e-05, | |
| "loss": 1.8665, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 1.8216, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 1.7746, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.738, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 5e-05, | |
| "loss": 1.7621, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 5.2000000000000004e-05, | |
| "loss": 1.4583, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 5.4000000000000005e-05, | |
| "loss": 1.3901, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 1.2028, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 5.8e-05, | |
| "loss": 1.1564, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 6e-05, | |
| "loss": 1.3146, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 6.2e-05, | |
| "loss": 1.253, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 1.4678, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 6.6e-05, | |
| "loss": 1.3828, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 6.800000000000001e-05, | |
| "loss": 1.3896, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 7e-05, | |
| "loss": 1.2786, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 7.2e-05, | |
| "loss": 1.2617, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 7.4e-05, | |
| "loss": 1.2258, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 7.6e-05, | |
| "loss": 1.1949, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 7.800000000000001e-05, | |
| "loss": 1.2313, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 8e-05, | |
| "loss": 1.166, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 8.2e-05, | |
| "loss": 1.135, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 8.4e-05, | |
| "loss": 1.1184, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 8.6e-05, | |
| "loss": 1.1188, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 1.0766, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 9e-05, | |
| "loss": 1.1467, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 9.200000000000001e-05, | |
| "loss": 1.2094, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 9.4e-05, | |
| "loss": 1.1808, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 9.6e-05, | |
| "loss": 1.1104, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 9.8e-05, | |
| "loss": 1.0752, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0001, | |
| "loss": 1.1378, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "eval_loss": 0.9520693421363831, | |
| "eval_runtime": 23.3729, | |
| "eval_samples_per_second": 213.923, | |
| "eval_steps_per_second": 1.711, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 9.999972660400536e-05, | |
| "loss": 1.062, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 9.999890641901125e-05, | |
| "loss": 1.213, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 9.999753945398704e-05, | |
| "loss": 1.2887, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 9.99956257238817e-05, | |
| "loss": 1.1635, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 1.0488, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 9.999015805811965e-05, | |
| "loss": 1.219, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 9.998660418225645e-05, | |
| "loss": 0.96, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 9.998250366089848e-05, | |
| "loss": 0.7237, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 9.997785653888835e-05, | |
| "loss": 0.7869, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 1.0635, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 9.996692270216947e-05, | |
| "loss": 1.1039, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 9.996063610703137e-05, | |
| "loss": 1.0344, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 9.995380315038119e-05, | |
| "loss": 1.0226, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 9.994642390694308e-05, | |
| "loss": 1.0823, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 1.0319, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 9.993002688846913e-05, | |
| "loss": 1.01, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 9.992100929274846e-05, | |
| "loss": 1.0433, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 9.991144576886823e-05, | |
| "loss": 1.0019, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 9.990133642141359e-05, | |
| "loss": 1.0647, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 0.9993, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 9.987948070396571e-05, | |
| "loss": 1.0423, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 9.986773457298311e-05, | |
| "loss": 1.106, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 9.985544309644475e-05, | |
| "loss": 0.906, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 9.984260640876821e-05, | |
| "loss": 1.0415, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 0.9777, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 9.981529796748134e-05, | |
| "loss": 1.1837, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 9.980082651251175e-05, | |
| "loss": 1.0263, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 9.97858104436822e-05, | |
| "loss": 1.0564, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 9.977024992520602e-05, | |
| "loss": 1.1334, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 0.9738, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 9.973749622593534e-05, | |
| "loss": 0.9563, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 9.972030340333001e-05, | |
| "loss": 0.9526, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 9.970256684745258e-05, | |
| "loss": 1.0087, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 9.968428675226714e-05, | |
| "loss": 0.9702, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 0.9558, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 9.964609674954696e-05, | |
| "loss": 0.9797, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 9.962618725965196e-05, | |
| "loss": 0.971, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 9.96057350657239e-05, | |
| "loss": 0.9795, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 9.95847403914247e-05, | |
| "loss": 1.0089, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 1.0765, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 9.954112452602045e-05, | |
| "loss": 1.1708, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 9.95185038118915e-05, | |
| "loss": 1.1234, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 9.949534157133844e-05, | |
| "loss": 1.059, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 9.94716380576598e-05, | |
| "loss": 1.0153, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 0.9817, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 9.942260825371358e-05, | |
| "loss": 1.0508, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 9.939728249962807e-05, | |
| "loss": 1.0584, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 9.937141654477528e-05, | |
| "loss": 1.0159, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 9.934501067202117e-05, | |
| "loss": 1.0099, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 0.9323, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "eval_loss": 0.8298289179801941, | |
| "eval_runtime": 18.3792, | |
| "eval_samples_per_second": 272.046, | |
| "eval_steps_per_second": 2.176, | |
| "step": 100000 | |
| } | |
| ], | |
| "max_steps": 1000000, | |
| "num_train_epochs": 9223372036854775807, | |
| "total_flos": 1.6845398605824e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |