{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.6451612903225806, "global_step": 80000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "learning_rate": 4.97983870967742e-05, "loss": 6.7059, "step": 500 }, { "epoch": 0.01, "learning_rate": 4.959677419354839e-05, "loss": 5.6289, "step": 1000 }, { "epoch": 0.01, "learning_rate": 4.939516129032259e-05, "loss": 5.1456, "step": 1500 }, { "epoch": 0.02, "learning_rate": 4.9193548387096775e-05, "loss": 4.7873, "step": 2000 }, { "epoch": 0.02, "learning_rate": 4.899193548387097e-05, "loss": 4.6802, "step": 2500 }, { "epoch": 0.02, "learning_rate": 4.8790322580645164e-05, "loss": 4.3627, "step": 3000 }, { "epoch": 0.03, "learning_rate": 4.858870967741936e-05, "loss": 4.196, "step": 3500 }, { "epoch": 0.03, "learning_rate": 4.8387096774193554e-05, "loss": 3.7748, "step": 4000 }, { "epoch": 0.04, "learning_rate": 4.818548387096775e-05, "loss": 3.8727, "step": 4500 }, { "epoch": 0.04, "learning_rate": 4.7983870967741937e-05, "loss": 3.646, "step": 5000 }, { "epoch": 0.04, "learning_rate": 4.778225806451613e-05, "loss": 3.6242, "step": 5500 }, { "epoch": 0.05, "learning_rate": 4.7580645161290326e-05, "loss": 3.5577, "step": 6000 }, { "epoch": 0.05, "learning_rate": 4.737903225806452e-05, "loss": 3.5461, "step": 6500 }, { "epoch": 0.06, "learning_rate": 4.7177419354838716e-05, "loss": 3.4455, "step": 7000 }, { "epoch": 0.06, "learning_rate": 4.697580645161291e-05, "loss": 3.4798, "step": 7500 }, { "epoch": 0.06, "learning_rate": 4.67741935483871e-05, "loss": 3.4644, "step": 8000 }, { "epoch": 0.07, "learning_rate": 4.657258064516129e-05, "loss": 3.3689, "step": 8500 }, { "epoch": 0.07, "learning_rate": 4.637096774193548e-05, "loss": 3.3166, "step": 9000 }, { "epoch": 0.08, "learning_rate": 4.6169354838709676e-05, "loss": 3.3226, "step": 9500 }, { "epoch": 0.08, "learning_rate": 4.596774193548387e-05, "loss": 3.283, "step": 10000 }, { "epoch": 0.08, "learning_rate": 4.5766129032258065e-05, "loss": 3.1365, "step": 10500 }, { "epoch": 0.09, "learning_rate": 4.556451612903226e-05, "loss": 3.23, "step": 11000 }, { "epoch": 0.09, "learning_rate": 4.5362903225806455e-05, "loss": 3.2074, "step": 11500 }, { "epoch": 0.1, "learning_rate": 4.516129032258064e-05, "loss": 3.221, "step": 12000 }, { "epoch": 0.1, "learning_rate": 4.495967741935484e-05, "loss": 3.1792, "step": 12500 }, { "epoch": 0.1, "learning_rate": 4.475806451612903e-05, "loss": 3.0746, "step": 13000 }, { "epoch": 0.11, "learning_rate": 4.455645161290323e-05, "loss": 3.1394, "step": 13500 }, { "epoch": 0.11, "learning_rate": 4.435483870967742e-05, "loss": 3.0609, "step": 14000 }, { "epoch": 0.12, "learning_rate": 4.415322580645162e-05, "loss": 3.1135, "step": 14500 }, { "epoch": 0.12, "learning_rate": 4.395161290322581e-05, "loss": 3.0392, "step": 15000 }, { "epoch": 0.12, "learning_rate": 4.375e-05, "loss": 3.0255, "step": 15500 }, { "epoch": 0.13, "learning_rate": 4.3548387096774194e-05, "loss": 3.0095, "step": 16000 }, { "epoch": 0.13, "learning_rate": 4.334677419354839e-05, "loss": 3.0491, "step": 16500 }, { "epoch": 0.14, "learning_rate": 4.3145161290322584e-05, "loss": 3.068, "step": 17000 }, { "epoch": 0.14, "learning_rate": 4.294354838709678e-05, "loss": 2.9763, "step": 17500 }, { "epoch": 0.15, "learning_rate": 4.2741935483870973e-05, "loss": 2.985, "step": 18000 }, { "epoch": 0.15, "learning_rate": 4.254032258064516e-05, "loss": 2.9917, "step": 18500 }, { "epoch": 0.15, "learning_rate": 4.2338709677419356e-05, "loss": 3.0161, "step": 19000 }, { "epoch": 0.16, "learning_rate": 4.213709677419355e-05, "loss": 2.9634, "step": 19500 }, { "epoch": 0.16, "learning_rate": 4.1935483870967746e-05, "loss": 2.9889, "step": 20000 }, { "epoch": 0.17, "learning_rate": 4.173387096774194e-05, "loss": 3.0213, "step": 20500 }, { "epoch": 0.17, "learning_rate": 4.1532258064516135e-05, "loss": 2.9616, "step": 21000 }, { "epoch": 0.17, "learning_rate": 4.133064516129033e-05, "loss": 2.8978, "step": 21500 }, { "epoch": 0.18, "learning_rate": 4.112903225806452e-05, "loss": 2.7943, "step": 22000 }, { "epoch": 0.18, "learning_rate": 4.092741935483871e-05, "loss": 2.9148, "step": 22500 }, { "epoch": 0.19, "learning_rate": 4.072580645161291e-05, "loss": 2.9236, "step": 23000 }, { "epoch": 0.19, "learning_rate": 4.0524193548387096e-05, "loss": 2.8025, "step": 23500 }, { "epoch": 0.19, "learning_rate": 4.032258064516129e-05, "loss": 2.8641, "step": 24000 }, { "epoch": 0.2, "learning_rate": 4.0120967741935485e-05, "loss": 2.8458, "step": 24500 }, { "epoch": 0.2, "learning_rate": 3.991935483870968e-05, "loss": 2.8769, "step": 25000 }, { "epoch": 0.21, "learning_rate": 3.971774193548387e-05, "loss": 2.9281, "step": 25500 }, { "epoch": 0.21, "learning_rate": 3.951612903225806e-05, "loss": 2.8934, "step": 26000 }, { "epoch": 0.21, "learning_rate": 3.931451612903226e-05, "loss": 2.8421, "step": 26500 }, { "epoch": 0.22, "learning_rate": 3.911290322580645e-05, "loss": 2.9583, "step": 27000 }, { "epoch": 0.22, "learning_rate": 3.891129032258065e-05, "loss": 2.8541, "step": 27500 }, { "epoch": 0.23, "learning_rate": 3.870967741935484e-05, "loss": 2.9187, "step": 28000 }, { "epoch": 0.23, "learning_rate": 3.8508064516129036e-05, "loss": 2.9765, "step": 28500 }, { "epoch": 0.23, "learning_rate": 3.8306451612903224e-05, "loss": 2.6983, "step": 29000 }, { "epoch": 0.24, "learning_rate": 3.810483870967742e-05, "loss": 2.8056, "step": 29500 }, { "epoch": 0.24, "learning_rate": 3.7903225806451614e-05, "loss": 2.7826, "step": 30000 }, { "epoch": 0.25, "learning_rate": 3.770161290322581e-05, "loss": 2.8993, "step": 30500 }, { "epoch": 0.25, "learning_rate": 3.7500000000000003e-05, "loss": 2.931, "step": 31000 }, { "epoch": 0.25, "learning_rate": 3.72983870967742e-05, "loss": 2.8276, "step": 31500 }, { "epoch": 0.26, "learning_rate": 3.7096774193548386e-05, "loss": 2.8348, "step": 32000 }, { "epoch": 0.26, "learning_rate": 3.689516129032258e-05, "loss": 2.8317, "step": 32500 }, { "epoch": 0.27, "learning_rate": 3.6693548387096776e-05, "loss": 2.8061, "step": 33000 }, { "epoch": 0.27, "learning_rate": 3.649193548387097e-05, "loss": 2.7982, "step": 33500 }, { "epoch": 0.27, "learning_rate": 3.6290322580645165e-05, "loss": 2.8439, "step": 34000 }, { "epoch": 0.28, "learning_rate": 3.608870967741936e-05, "loss": 2.7924, "step": 34500 }, { "epoch": 0.28, "learning_rate": 3.5887096774193555e-05, "loss": 2.7077, "step": 35000 }, { "epoch": 0.29, "learning_rate": 3.568548387096774e-05, "loss": 2.8267, "step": 35500 }, { "epoch": 0.29, "learning_rate": 3.548387096774194e-05, "loss": 2.8308, "step": 36000 }, { "epoch": 0.29, "learning_rate": 3.528225806451613e-05, "loss": 2.7554, "step": 36500 }, { "epoch": 0.3, "learning_rate": 3.508064516129033e-05, "loss": 2.7693, "step": 37000 }, { "epoch": 0.3, "learning_rate": 3.487903225806452e-05, "loss": 2.7417, "step": 37500 }, { "epoch": 0.31, "learning_rate": 3.467741935483872e-05, "loss": 2.6836, "step": 38000 }, { "epoch": 0.31, "learning_rate": 3.4475806451612905e-05, "loss": 2.7334, "step": 38500 }, { "epoch": 0.31, "learning_rate": 3.427419354838709e-05, "loss": 2.9839, "step": 39000 }, { "epoch": 0.32, "learning_rate": 3.407258064516129e-05, "loss": 2.7495, "step": 39500 }, { "epoch": 0.32, "learning_rate": 3.387096774193548e-05, "loss": 2.767, "step": 40000 }, { "epoch": 0.33, "learning_rate": 3.366935483870968e-05, "loss": 2.7316, "step": 40500 }, { "epoch": 0.33, "learning_rate": 3.346774193548387e-05, "loss": 2.7787, "step": 41000 }, { "epoch": 0.33, "learning_rate": 3.3266129032258067e-05, "loss": 2.7504, "step": 41500 }, { "epoch": 0.34, "learning_rate": 3.306451612903226e-05, "loss": 2.6913, "step": 42000 }, { "epoch": 0.34, "learning_rate": 3.286290322580645e-05, "loss": 2.7743, "step": 42500 }, { "epoch": 0.35, "learning_rate": 3.2661290322580644e-05, "loss": 2.7656, "step": 43000 }, { "epoch": 0.35, "learning_rate": 3.245967741935484e-05, "loss": 2.7828, "step": 43500 }, { "epoch": 0.35, "learning_rate": 3.2258064516129034e-05, "loss": 2.7927, "step": 44000 }, { "epoch": 0.36, "learning_rate": 3.205645161290323e-05, "loss": 2.7036, "step": 44500 }, { "epoch": 0.36, "learning_rate": 3.185483870967742e-05, "loss": 2.6734, "step": 45000 }, { "epoch": 0.37, "learning_rate": 3.165322580645161e-05, "loss": 2.7777, "step": 45500 }, { "epoch": 0.37, "learning_rate": 3.1451612903225806e-05, "loss": 2.7163, "step": 46000 }, { "epoch": 0.38, "learning_rate": 3.125e-05, "loss": 2.7279, "step": 46500 }, { "epoch": 0.38, "learning_rate": 3.1048387096774195e-05, "loss": 2.7209, "step": 47000 }, { "epoch": 0.38, "learning_rate": 3.084677419354839e-05, "loss": 2.6712, "step": 47500 }, { "epoch": 0.39, "learning_rate": 3.0645161290322585e-05, "loss": 2.7012, "step": 48000 }, { "epoch": 0.39, "learning_rate": 3.0443548387096776e-05, "loss": 2.7272, "step": 48500 }, { "epoch": 0.4, "learning_rate": 3.024193548387097e-05, "loss": 2.7072, "step": 49000 }, { "epoch": 0.4, "learning_rate": 3.0040322580645162e-05, "loss": 2.6696, "step": 49500 }, { "epoch": 0.4, "learning_rate": 2.9838709677419357e-05, "loss": 2.6467, "step": 50000 }, { "epoch": 0.41, "learning_rate": 2.9637096774193552e-05, "loss": 2.7601, "step": 50500 }, { "epoch": 0.41, "learning_rate": 2.9435483870967743e-05, "loss": 2.7173, "step": 51000 }, { "epoch": 0.42, "learning_rate": 2.9233870967741938e-05, "loss": 2.6568, "step": 51500 }, { "epoch": 0.42, "learning_rate": 2.9032258064516133e-05, "loss": 2.713, "step": 52000 }, { "epoch": 0.42, "learning_rate": 2.8830645161290328e-05, "loss": 2.7691, "step": 52500 }, { "epoch": 0.43, "learning_rate": 2.862903225806452e-05, "loss": 2.7268, "step": 53000 }, { "epoch": 0.43, "learning_rate": 2.8427419354838714e-05, "loss": 2.5954, "step": 53500 }, { "epoch": 0.44, "learning_rate": 2.822580645161291e-05, "loss": 2.6935, "step": 54000 }, { "epoch": 0.44, "learning_rate": 2.8024193548387097e-05, "loss": 2.6538, "step": 54500 }, { "epoch": 0.44, "learning_rate": 2.7822580645161288e-05, "loss": 2.7291, "step": 55000 }, { "epoch": 0.45, "learning_rate": 2.7620967741935483e-05, "loss": 2.6463, "step": 55500 }, { "epoch": 0.45, "learning_rate": 2.7419354838709678e-05, "loss": 2.7153, "step": 56000 }, { "epoch": 0.46, "learning_rate": 2.721774193548387e-05, "loss": 2.6529, "step": 56500 }, { "epoch": 0.46, "learning_rate": 2.7016129032258064e-05, "loss": 2.6653, "step": 57000 }, { "epoch": 0.46, "learning_rate": 2.681451612903226e-05, "loss": 2.6571, "step": 57500 }, { "epoch": 0.47, "learning_rate": 2.661290322580645e-05, "loss": 2.6362, "step": 58000 }, { "epoch": 0.47, "learning_rate": 2.6411290322580645e-05, "loss": 2.709, "step": 58500 }, { "epoch": 0.48, "learning_rate": 2.620967741935484e-05, "loss": 2.6777, "step": 59000 }, { "epoch": 0.48, "learning_rate": 2.6008064516129034e-05, "loss": 2.7051, "step": 59500 }, { "epoch": 0.48, "learning_rate": 2.5806451612903226e-05, "loss": 2.6077, "step": 60000 }, { "epoch": 0.49, "learning_rate": 2.560483870967742e-05, "loss": 2.6631, "step": 60500 }, { "epoch": 0.49, "learning_rate": 2.5403225806451615e-05, "loss": 2.6302, "step": 61000 }, { "epoch": 0.5, "learning_rate": 2.5201612903225806e-05, "loss": 2.605, "step": 61500 }, { "epoch": 0.5, "learning_rate": 2.5e-05, "loss": 2.5936, "step": 62000 }, { "epoch": 0.5, "learning_rate": 2.4798387096774196e-05, "loss": 2.6623, "step": 62500 }, { "epoch": 0.51, "learning_rate": 2.4596774193548387e-05, "loss": 2.7046, "step": 63000 }, { "epoch": 0.51, "learning_rate": 2.4395161290322582e-05, "loss": 2.5976, "step": 63500 }, { "epoch": 0.52, "learning_rate": 2.4193548387096777e-05, "loss": 2.6664, "step": 64000 }, { "epoch": 0.52, "learning_rate": 2.3991935483870968e-05, "loss": 2.6509, "step": 64500 }, { "epoch": 0.52, "learning_rate": 2.3790322580645163e-05, "loss": 2.6651, "step": 65000 }, { "epoch": 0.53, "learning_rate": 2.3588709677419358e-05, "loss": 2.6349, "step": 65500 }, { "epoch": 0.53, "learning_rate": 2.338709677419355e-05, "loss": 2.6451, "step": 66000 }, { "epoch": 0.54, "learning_rate": 2.318548387096774e-05, "loss": 2.6604, "step": 66500 }, { "epoch": 0.54, "learning_rate": 2.2983870967741935e-05, "loss": 2.6519, "step": 67000 }, { "epoch": 0.54, "learning_rate": 2.278225806451613e-05, "loss": 2.6072, "step": 67500 }, { "epoch": 0.55, "learning_rate": 2.258064516129032e-05, "loss": 2.6437, "step": 68000 }, { "epoch": 0.55, "learning_rate": 2.2379032258064516e-05, "loss": 2.6558, "step": 68500 }, { "epoch": 0.56, "learning_rate": 2.217741935483871e-05, "loss": 2.5988, "step": 69000 }, { "epoch": 0.56, "learning_rate": 2.1975806451612906e-05, "loss": 2.646, "step": 69500 }, { "epoch": 0.56, "learning_rate": 2.1774193548387097e-05, "loss": 2.6395, "step": 70000 }, { "epoch": 0.57, "learning_rate": 2.1572580645161292e-05, "loss": 2.6216, "step": 70500 }, { "epoch": 0.57, "learning_rate": 2.1370967741935487e-05, "loss": 2.6985, "step": 71000 }, { "epoch": 0.58, "learning_rate": 2.1169354838709678e-05, "loss": 2.6508, "step": 71500 }, { "epoch": 0.58, "learning_rate": 2.0967741935483873e-05, "loss": 2.6124, "step": 72000 }, { "epoch": 0.58, "learning_rate": 2.0766129032258068e-05, "loss": 2.5642, "step": 72500 }, { "epoch": 0.59, "learning_rate": 2.056451612903226e-05, "loss": 2.6297, "step": 73000 }, { "epoch": 0.59, "learning_rate": 2.0362903225806454e-05, "loss": 2.5099, "step": 73500 }, { "epoch": 0.6, "learning_rate": 2.0161290322580645e-05, "loss": 2.5825, "step": 74000 }, { "epoch": 0.6, "learning_rate": 1.995967741935484e-05, "loss": 2.6509, "step": 74500 }, { "epoch": 0.6, "learning_rate": 1.975806451612903e-05, "loss": 2.5801, "step": 75000 }, { "epoch": 0.61, "learning_rate": 1.9556451612903226e-05, "loss": 2.6328, "step": 75500 }, { "epoch": 0.61, "learning_rate": 1.935483870967742e-05, "loss": 2.5998, "step": 76000 }, { "epoch": 0.62, "learning_rate": 1.9153225806451612e-05, "loss": 2.6794, "step": 76500 }, { "epoch": 0.62, "learning_rate": 1.8951612903225807e-05, "loss": 2.6152, "step": 77000 }, { "epoch": 0.62, "learning_rate": 1.8750000000000002e-05, "loss": 2.5854, "step": 77500 }, { "epoch": 0.63, "learning_rate": 1.8548387096774193e-05, "loss": 2.587, "step": 78000 }, { "epoch": 0.63, "learning_rate": 1.8346774193548388e-05, "loss": 2.5791, "step": 78500 }, { "epoch": 0.64, "learning_rate": 1.8145161290322583e-05, "loss": 2.5613, "step": 79000 }, { "epoch": 0.64, "learning_rate": 1.7943548387096777e-05, "loss": 2.581, "step": 79500 }, { "epoch": 0.65, "learning_rate": 1.774193548387097e-05, "loss": 2.5466, "step": 80000 } ], "max_steps": 124000, "num_train_epochs": 9223372036854775807, "total_flos": 1.18874494205952e+18, "trial_name": null, "trial_params": null }