| { | |
| "best_metric": 0.6551458239555359, | |
| "best_model_checkpoint": "Action_model/checkpoint-600", | |
| "epoch": 10.0, | |
| "eval_steps": 100, | |
| "global_step": 3140, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.6790106296539307, | |
| "learning_rate": 9.936305732484077e-05, | |
| "loss": 2.1638, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.8781601190567017, | |
| "learning_rate": 9.872611464968153e-05, | |
| "loss": 1.8702, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 2.154752016067505, | |
| "learning_rate": 9.80891719745223e-05, | |
| "loss": 1.496, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 2.715329885482788, | |
| "learning_rate": 9.745222929936307e-05, | |
| "loss": 1.2633, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 2.8564321994781494, | |
| "learning_rate": 9.681528662420382e-05, | |
| "loss": 1.1382, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "eval_accuracy": 0.7676190476190476, | |
| "eval_loss": 1.0001901388168335, | |
| "eval_runtime": 19.3002, | |
| "eval_samples_per_second": 54.404, | |
| "eval_steps_per_second": 6.839, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 2.2369894981384277, | |
| "learning_rate": 9.617834394904459e-05, | |
| "loss": 1.0084, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 3.4310965538024902, | |
| "learning_rate": 9.554140127388536e-05, | |
| "loss": 0.9195, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 3.0611016750335693, | |
| "learning_rate": 9.490445859872612e-05, | |
| "loss": 0.9319, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 4.078617095947266, | |
| "learning_rate": 9.426751592356689e-05, | |
| "loss": 0.7825, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 2.7804274559020996, | |
| "learning_rate": 9.363057324840766e-05, | |
| "loss": 0.782, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "eval_accuracy": 0.7676190476190476, | |
| "eval_loss": 0.7673064470291138, | |
| "eval_runtime": 15.1028, | |
| "eval_samples_per_second": 69.523, | |
| "eval_steps_per_second": 8.74, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 1.4780933856964111, | |
| "learning_rate": 9.299363057324841e-05, | |
| "loss": 0.6899, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 2.280031681060791, | |
| "learning_rate": 9.238853503184714e-05, | |
| "loss": 0.6333, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 2.9191763401031494, | |
| "learning_rate": 9.17515923566879e-05, | |
| "loss": 0.7143, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 2.002995729446411, | |
| "learning_rate": 9.111464968152866e-05, | |
| "loss": 0.6578, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 3.061410427093506, | |
| "learning_rate": 9.047770700636943e-05, | |
| "loss": 0.6289, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "eval_accuracy": 0.7866666666666666, | |
| "eval_loss": 0.70728999376297, | |
| "eval_runtime": 15.1443, | |
| "eval_samples_per_second": 69.333, | |
| "eval_steps_per_second": 8.716, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 3.693359375, | |
| "learning_rate": 8.984076433121019e-05, | |
| "loss": 0.5495, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 3.992403507232666, | |
| "learning_rate": 8.920382165605096e-05, | |
| "loss": 0.5563, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 2.6919829845428467, | |
| "learning_rate": 8.856687898089173e-05, | |
| "loss": 0.5681, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 3.0533978939056396, | |
| "learning_rate": 8.796178343949045e-05, | |
| "loss": 0.5286, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 2.9173035621643066, | |
| "learning_rate": 8.732484076433122e-05, | |
| "loss": 0.5028, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "eval_accuracy": 0.7685714285714286, | |
| "eval_loss": 0.7260778546333313, | |
| "eval_runtime": 17.5337, | |
| "eval_samples_per_second": 59.885, | |
| "eval_steps_per_second": 7.528, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 1.7928365468978882, | |
| "learning_rate": 8.668789808917198e-05, | |
| "loss": 0.4379, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 4.260186195373535, | |
| "learning_rate": 8.605095541401275e-05, | |
| "loss": 0.48, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 2.8024277687072754, | |
| "learning_rate": 8.541401273885352e-05, | |
| "loss": 0.3689, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 4.045362949371338, | |
| "learning_rate": 8.477707006369427e-05, | |
| "loss": 0.3922, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 1.6472926139831543, | |
| "learning_rate": 8.414012738853504e-05, | |
| "loss": 0.4746, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "eval_accuracy": 0.7619047619047619, | |
| "eval_loss": 0.7463707327842712, | |
| "eval_runtime": 15.5218, | |
| "eval_samples_per_second": 67.647, | |
| "eval_steps_per_second": 8.504, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 4.001276969909668, | |
| "learning_rate": 8.350318471337581e-05, | |
| "loss": 0.4123, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 4.151864528656006, | |
| "learning_rate": 8.286624203821657e-05, | |
| "loss": 0.4626, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 3.43729567527771, | |
| "learning_rate": 8.222929936305733e-05, | |
| "loss": 0.4279, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 3.8993654251098633, | |
| "learning_rate": 8.159235668789809e-05, | |
| "loss": 0.4288, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 3.0417702198028564, | |
| "learning_rate": 8.095541401273886e-05, | |
| "loss": 0.4298, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "eval_accuracy": 0.799047619047619, | |
| "eval_loss": 0.6551458239555359, | |
| "eval_runtime": 15.0487, | |
| "eval_samples_per_second": 69.773, | |
| "eval_steps_per_second": 8.771, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 4.088045597076416, | |
| "learning_rate": 8.031847133757963e-05, | |
| "loss": 0.404, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 2.1314449310302734, | |
| "learning_rate": 7.968152866242038e-05, | |
| "loss": 0.4096, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 5.116842746734619, | |
| "learning_rate": 7.904458598726115e-05, | |
| "loss": 0.3332, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 4.045914649963379, | |
| "learning_rate": 7.840764331210192e-05, | |
| "loss": 0.2961, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 1.9349473714828491, | |
| "learning_rate": 7.777070063694268e-05, | |
| "loss": 0.3488, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "eval_accuracy": 0.7733333333333333, | |
| "eval_loss": 0.7358552813529968, | |
| "eval_runtime": 15.4659, | |
| "eval_samples_per_second": 67.891, | |
| "eval_steps_per_second": 8.535, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 4.3049845695495605, | |
| "learning_rate": 7.713375796178345e-05, | |
| "loss": 0.3078, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 2.4363415241241455, | |
| "learning_rate": 7.649681528662422e-05, | |
| "loss": 0.4005, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 3.518944501876831, | |
| "learning_rate": 7.585987261146497e-05, | |
| "loss": 0.3428, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 3.8007328510284424, | |
| "learning_rate": 7.522292993630574e-05, | |
| "loss": 0.3471, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 1.3215352296829224, | |
| "learning_rate": 7.45859872611465e-05, | |
| "loss": 0.266, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "eval_accuracy": 0.7514285714285714, | |
| "eval_loss": 0.829559862613678, | |
| "eval_runtime": 15.0559, | |
| "eval_samples_per_second": 69.74, | |
| "eval_steps_per_second": 8.767, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 1.9990004301071167, | |
| "learning_rate": 7.394904458598727e-05, | |
| "loss": 0.2918, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 5.201882362365723, | |
| "learning_rate": 7.331210191082802e-05, | |
| "loss": 0.3175, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 4.187939167022705, | |
| "learning_rate": 7.267515923566879e-05, | |
| "loss": 0.304, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 2.6038854122161865, | |
| "learning_rate": 7.203821656050955e-05, | |
| "loss": 0.3119, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 3.0886316299438477, | |
| "learning_rate": 7.140127388535032e-05, | |
| "loss": 0.3651, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "eval_accuracy": 0.7304761904761905, | |
| "eval_loss": 0.8660680651664734, | |
| "eval_runtime": 15.7752, | |
| "eval_samples_per_second": 66.56, | |
| "eval_steps_per_second": 8.368, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 3.3455846309661865, | |
| "learning_rate": 7.076433121019108e-05, | |
| "loss": 0.2739, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 3.0882346630096436, | |
| "learning_rate": 7.012738853503184e-05, | |
| "loss": 0.2327, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "grad_norm": 4.43536901473999, | |
| "learning_rate": 6.949044585987261e-05, | |
| "loss": 0.2699, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 3.8387985229492188, | |
| "learning_rate": 6.885350318471338e-05, | |
| "loss": 0.295, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "grad_norm": 2.8090808391571045, | |
| "learning_rate": 6.821656050955413e-05, | |
| "loss": 0.2796, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "eval_accuracy": 0.7866666666666666, | |
| "eval_loss": 0.7188078165054321, | |
| "eval_runtime": 15.2769, | |
| "eval_samples_per_second": 68.731, | |
| "eval_steps_per_second": 8.641, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 2.789705991744995, | |
| "learning_rate": 6.75796178343949e-05, | |
| "loss": 0.234, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "grad_norm": 4.02871561050415, | |
| "learning_rate": 6.694267515923567e-05, | |
| "loss": 0.2282, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 3.38, | |
| "grad_norm": 2.786869764328003, | |
| "learning_rate": 6.630573248407643e-05, | |
| "loss": 0.3052, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "grad_norm": 3.5847015380859375, | |
| "learning_rate": 6.56687898089172e-05, | |
| "loss": 0.2343, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 2.2771642208099365, | |
| "learning_rate": 6.503184713375797e-05, | |
| "loss": 0.2703, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "eval_accuracy": 0.7476190476190476, | |
| "eval_loss": 0.8421508073806763, | |
| "eval_runtime": 15.1919, | |
| "eval_samples_per_second": 69.116, | |
| "eval_steps_per_second": 8.689, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "grad_norm": 3.3567535877227783, | |
| "learning_rate": 6.439490445859872e-05, | |
| "loss": 0.2429, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "grad_norm": 4.021266937255859, | |
| "learning_rate": 6.375796178343949e-05, | |
| "loss": 0.2875, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "grad_norm": 1.5646867752075195, | |
| "learning_rate": 6.312101910828026e-05, | |
| "loss": 0.2355, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "grad_norm": 0.8209928274154663, | |
| "learning_rate": 6.248407643312102e-05, | |
| "loss": 0.2367, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 3.82, | |
| "grad_norm": 5.591761589050293, | |
| "learning_rate": 6.184713375796178e-05, | |
| "loss": 0.2608, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.82, | |
| "eval_accuracy": 0.7723809523809524, | |
| "eval_loss": 0.8207409381866455, | |
| "eval_runtime": 15.3778, | |
| "eval_samples_per_second": 68.28, | |
| "eval_steps_per_second": 8.584, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "grad_norm": 2.2955307960510254, | |
| "learning_rate": 6.121019108280255e-05, | |
| "loss": 0.2174, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "grad_norm": 4.40664005279541, | |
| "learning_rate": 6.057324840764331e-05, | |
| "loss": 0.2168, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "grad_norm": 4.821913719177246, | |
| "learning_rate": 5.993630573248408e-05, | |
| "loss": 0.251, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "grad_norm": 6.526182174682617, | |
| "learning_rate": 5.929936305732484e-05, | |
| "loss": 0.2424, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 4.14, | |
| "grad_norm": 1.996484398841858, | |
| "learning_rate": 5.86624203821656e-05, | |
| "loss": 0.251, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.14, | |
| "eval_accuracy": 0.7266666666666667, | |
| "eval_loss": 1.0251611471176147, | |
| "eval_runtime": 14.903, | |
| "eval_samples_per_second": 70.455, | |
| "eval_steps_per_second": 8.857, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 2.93237566947937, | |
| "learning_rate": 5.802547770700637e-05, | |
| "loss": 0.1727, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "grad_norm": 3.0160069465637207, | |
| "learning_rate": 5.7388535031847135e-05, | |
| "loss": 0.2049, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "grad_norm": 2.878361701965332, | |
| "learning_rate": 5.67515923566879e-05, | |
| "loss": 0.2221, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "grad_norm": 3.4867329597473145, | |
| "learning_rate": 5.6114649681528666e-05, | |
| "loss": 0.2459, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 4.46, | |
| "grad_norm": 4.972071170806885, | |
| "learning_rate": 5.547770700636943e-05, | |
| "loss": 0.2085, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.46, | |
| "eval_accuracy": 0.7171428571428572, | |
| "eval_loss": 1.0474802255630493, | |
| "eval_runtime": 15.353, | |
| "eval_samples_per_second": 68.391, | |
| "eval_steps_per_second": 8.598, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.52, | |
| "grad_norm": 3.7226266860961914, | |
| "learning_rate": 5.484076433121019e-05, | |
| "loss": 0.22, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "grad_norm": 3.0898613929748535, | |
| "learning_rate": 5.420382165605096e-05, | |
| "loss": 0.206, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "grad_norm": 5.401129722595215, | |
| "learning_rate": 5.356687898089172e-05, | |
| "loss": 0.2215, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "grad_norm": 3.430591344833374, | |
| "learning_rate": 5.2929936305732485e-05, | |
| "loss": 0.1883, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 4.78, | |
| "grad_norm": 0.33961915969848633, | |
| "learning_rate": 5.229299363057325e-05, | |
| "loss": 0.1715, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.78, | |
| "eval_accuracy": 0.7495238095238095, | |
| "eval_loss": 0.8852301836013794, | |
| "eval_runtime": 15.0006, | |
| "eval_samples_per_second": 69.997, | |
| "eval_steps_per_second": 8.8, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.84, | |
| "grad_norm": 3.065584182739258, | |
| "learning_rate": 5.1656050955414016e-05, | |
| "loss": 0.1687, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 4.120232582092285, | |
| "learning_rate": 5.101910828025478e-05, | |
| "loss": 0.214, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 4.97, | |
| "grad_norm": 0.5130987763404846, | |
| "learning_rate": 5.038216560509554e-05, | |
| "loss": 0.1631, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "grad_norm": 4.085451126098633, | |
| "learning_rate": 4.974522292993631e-05, | |
| "loss": 0.1907, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 3.028500556945801, | |
| "learning_rate": 4.910828025477707e-05, | |
| "loss": 0.2051, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "eval_accuracy": 0.7790476190476191, | |
| "eval_loss": 0.8164414763450623, | |
| "eval_runtime": 15.3889, | |
| "eval_samples_per_second": 68.231, | |
| "eval_steps_per_second": 8.578, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 5.16, | |
| "grad_norm": 2.5563900470733643, | |
| "learning_rate": 4.8471337579617835e-05, | |
| "loss": 0.187, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 5.22, | |
| "grad_norm": 3.853022813796997, | |
| "learning_rate": 4.7834394904458604e-05, | |
| "loss": 0.2186, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 5.29, | |
| "grad_norm": 6.835115909576416, | |
| "learning_rate": 4.7197452229299366e-05, | |
| "loss": 0.1717, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 5.35, | |
| "grad_norm": 3.7477526664733887, | |
| "learning_rate": 4.656050955414013e-05, | |
| "loss": 0.2352, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 5.41, | |
| "grad_norm": 3.5091373920440674, | |
| "learning_rate": 4.59235668789809e-05, | |
| "loss": 0.1481, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 5.41, | |
| "eval_accuracy": 0.7628571428571429, | |
| "eval_loss": 0.8825291991233826, | |
| "eval_runtime": 15.2694, | |
| "eval_samples_per_second": 68.765, | |
| "eval_steps_per_second": 8.645, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 5.48, | |
| "grad_norm": 4.236053943634033, | |
| "learning_rate": 4.528662420382166e-05, | |
| "loss": 0.1482, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 5.54, | |
| "grad_norm": 3.335090160369873, | |
| "learning_rate": 4.464968152866242e-05, | |
| "loss": 0.1399, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 5.61, | |
| "grad_norm": 2.238372802734375, | |
| "learning_rate": 4.4012738853503185e-05, | |
| "loss": 0.1664, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 5.67, | |
| "grad_norm": 1.2184184789657593, | |
| "learning_rate": 4.3375796178343954e-05, | |
| "loss": 0.179, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 5.73, | |
| "grad_norm": 7.519371509552002, | |
| "learning_rate": 4.2738853503184716e-05, | |
| "loss": 0.177, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 5.73, | |
| "eval_accuracy": 0.7866666666666666, | |
| "eval_loss": 0.8622841238975525, | |
| "eval_runtime": 16.3495, | |
| "eval_samples_per_second": 64.222, | |
| "eval_steps_per_second": 8.074, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "grad_norm": 3.128350019454956, | |
| "learning_rate": 4.210191082802548e-05, | |
| "loss": 0.1943, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 5.86, | |
| "grad_norm": 3.5789175033569336, | |
| "learning_rate": 4.146496815286625e-05, | |
| "loss": 0.1431, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 5.92, | |
| "grad_norm": 5.671403408050537, | |
| "learning_rate": 4.082802547770701e-05, | |
| "loss": 0.1761, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 5.99, | |
| "grad_norm": 2.8949170112609863, | |
| "learning_rate": 4.019108280254777e-05, | |
| "loss": 0.1639, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 6.05, | |
| "grad_norm": 6.811347961425781, | |
| "learning_rate": 3.955414012738854e-05, | |
| "loss": 0.1607, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 6.05, | |
| "eval_accuracy": 0.7609523809523809, | |
| "eval_loss": 0.948749303817749, | |
| "eval_runtime": 15.6299, | |
| "eval_samples_per_second": 67.179, | |
| "eval_steps_per_second": 8.445, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 6.11, | |
| "grad_norm": 3.2780776023864746, | |
| "learning_rate": 3.8917197452229304e-05, | |
| "loss": 0.2392, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 6.18, | |
| "grad_norm": 1.425671935081482, | |
| "learning_rate": 3.8280254777070066e-05, | |
| "loss": 0.1988, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 6.24, | |
| "grad_norm": 4.055123329162598, | |
| "learning_rate": 3.7643312101910836e-05, | |
| "loss": 0.1563, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 6.31, | |
| "grad_norm": 5.6207356452941895, | |
| "learning_rate": 3.700636942675159e-05, | |
| "loss": 0.1364, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 6.37, | |
| "grad_norm": 0.6465654373168945, | |
| "learning_rate": 3.6369426751592353e-05, | |
| "loss": 0.1273, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 6.37, | |
| "eval_accuracy": 0.7733333333333333, | |
| "eval_loss": 0.8984624743461609, | |
| "eval_runtime": 15.237, | |
| "eval_samples_per_second": 68.911, | |
| "eval_steps_per_second": 8.663, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 6.43, | |
| "grad_norm": 3.3592026233673096, | |
| "learning_rate": 3.573248407643312e-05, | |
| "loss": 0.1903, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "grad_norm": 2.187608480453491, | |
| "learning_rate": 3.5095541401273885e-05, | |
| "loss": 0.1684, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 6.56, | |
| "grad_norm": 2.657270908355713, | |
| "learning_rate": 3.445859872611465e-05, | |
| "loss": 0.1619, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 6.62, | |
| "grad_norm": 2.2679970264434814, | |
| "learning_rate": 3.3821656050955416e-05, | |
| "loss": 0.1556, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 6.69, | |
| "grad_norm": 1.9460710287094116, | |
| "learning_rate": 3.318471337579618e-05, | |
| "loss": 0.1609, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 6.69, | |
| "eval_accuracy": 0.7504761904761905, | |
| "eval_loss": 0.9624072313308716, | |
| "eval_runtime": 15.076, | |
| "eval_samples_per_second": 69.647, | |
| "eval_steps_per_second": 8.756, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 6.75, | |
| "grad_norm": 2.056673526763916, | |
| "learning_rate": 3.254777070063694e-05, | |
| "loss": 0.1779, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 6.82, | |
| "grad_norm": 2.4007034301757812, | |
| "learning_rate": 3.191082802547771e-05, | |
| "loss": 0.1359, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 6.88, | |
| "grad_norm": 6.746215343475342, | |
| "learning_rate": 3.127388535031847e-05, | |
| "loss": 0.1653, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 6.94, | |
| "grad_norm": 3.8807878494262695, | |
| "learning_rate": 3.0636942675159235e-05, | |
| "loss": 0.1434, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "grad_norm": 4.5821990966796875, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1583, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "eval_accuracy": 0.7780952380952381, | |
| "eval_loss": 0.9015449285507202, | |
| "eval_runtime": 15.3184, | |
| "eval_samples_per_second": 68.545, | |
| "eval_steps_per_second": 8.617, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 7.07, | |
| "grad_norm": 3.755427837371826, | |
| "learning_rate": 2.9363057324840763e-05, | |
| "loss": 0.1251, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 7.13, | |
| "grad_norm": 2.709980010986328, | |
| "learning_rate": 2.872611464968153e-05, | |
| "loss": 0.1233, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "grad_norm": 2.7730648517608643, | |
| "learning_rate": 2.8089171974522295e-05, | |
| "loss": 0.1019, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 7.26, | |
| "grad_norm": 4.75289249420166, | |
| "learning_rate": 2.7452229299363057e-05, | |
| "loss": 0.1862, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 7.32, | |
| "grad_norm": 2.336014747619629, | |
| "learning_rate": 2.6815286624203823e-05, | |
| "loss": 0.1178, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 7.32, | |
| "eval_accuracy": 0.7761904761904762, | |
| "eval_loss": 0.9142788648605347, | |
| "eval_runtime": 15.5959, | |
| "eval_samples_per_second": 67.325, | |
| "eval_steps_per_second": 8.464, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 7.39, | |
| "grad_norm": 5.943305015563965, | |
| "learning_rate": 2.617834394904459e-05, | |
| "loss": 0.146, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 7.45, | |
| "grad_norm": 1.1991711854934692, | |
| "learning_rate": 2.554140127388535e-05, | |
| "loss": 0.1185, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 7.52, | |
| "grad_norm": 1.5695414543151855, | |
| "learning_rate": 2.4904458598726117e-05, | |
| "loss": 0.1437, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 7.58, | |
| "grad_norm": 0.3648098111152649, | |
| "learning_rate": 2.426751592356688e-05, | |
| "loss": 0.1028, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 7.64, | |
| "grad_norm": 2.0215370655059814, | |
| "learning_rate": 2.3630573248407645e-05, | |
| "loss": 0.1175, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 7.64, | |
| "eval_accuracy": 0.7590476190476191, | |
| "eval_loss": 0.9670929908752441, | |
| "eval_runtime": 18.4333, | |
| "eval_samples_per_second": 56.962, | |
| "eval_steps_per_second": 7.161, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 7.71, | |
| "grad_norm": 5.889522075653076, | |
| "learning_rate": 2.299363057324841e-05, | |
| "loss": 0.152, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 7.77, | |
| "grad_norm": 3.53730845451355, | |
| "learning_rate": 2.2356687898089173e-05, | |
| "loss": 0.165, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 7.83, | |
| "grad_norm": 2.178981065750122, | |
| "learning_rate": 2.171974522292994e-05, | |
| "loss": 0.1128, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 7.9, | |
| "grad_norm": 0.19914887845516205, | |
| "learning_rate": 2.1114649681528666e-05, | |
| "loss": 0.1307, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 7.96, | |
| "grad_norm": 2.177159070968628, | |
| "learning_rate": 2.0477707006369428e-05, | |
| "loss": 0.1257, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 7.96, | |
| "eval_accuracy": 0.7838095238095238, | |
| "eval_loss": 0.8925411701202393, | |
| "eval_runtime": 17.5418, | |
| "eval_samples_per_second": 59.857, | |
| "eval_steps_per_second": 7.525, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 8.03, | |
| "grad_norm": 2.6388609409332275, | |
| "learning_rate": 1.9840764331210194e-05, | |
| "loss": 0.1199, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 8.09, | |
| "grad_norm": 4.0329155921936035, | |
| "learning_rate": 1.9203821656050956e-05, | |
| "loss": 0.1158, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 8.15, | |
| "grad_norm": 2.214768409729004, | |
| "learning_rate": 1.856687898089172e-05, | |
| "loss": 0.148, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 8.22, | |
| "grad_norm": 3.5198891162872314, | |
| "learning_rate": 1.7929936305732484e-05, | |
| "loss": 0.1107, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 8.28, | |
| "grad_norm": 1.0330649614334106, | |
| "learning_rate": 1.7292993630573247e-05, | |
| "loss": 0.0939, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 8.28, | |
| "eval_accuracy": 0.7704761904761904, | |
| "eval_loss": 0.9257068634033203, | |
| "eval_runtime": 15.0117, | |
| "eval_samples_per_second": 69.945, | |
| "eval_steps_per_second": 8.793, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 8.34, | |
| "grad_norm": 1.8858942985534668, | |
| "learning_rate": 1.6656050955414012e-05, | |
| "loss": 0.1063, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 8.41, | |
| "grad_norm": 2.4143009185791016, | |
| "learning_rate": 1.6019108280254778e-05, | |
| "loss": 0.1624, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 8.47, | |
| "grad_norm": 0.2723791003227234, | |
| "learning_rate": 1.538216560509554e-05, | |
| "loss": 0.1109, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 8.54, | |
| "grad_norm": 3.380007266998291, | |
| "learning_rate": 1.4745222929936306e-05, | |
| "loss": 0.1375, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "grad_norm": 1.9235315322875977, | |
| "learning_rate": 1.410828025477707e-05, | |
| "loss": 0.1238, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "eval_accuracy": 0.7647619047619048, | |
| "eval_loss": 0.9797086715698242, | |
| "eval_runtime": 15.6554, | |
| "eval_samples_per_second": 67.07, | |
| "eval_steps_per_second": 8.432, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 8.66, | |
| "grad_norm": 2.037036180496216, | |
| "learning_rate": 1.3471337579617834e-05, | |
| "loss": 0.1381, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 8.73, | |
| "grad_norm": 1.407027006149292, | |
| "learning_rate": 1.2834394904458598e-05, | |
| "loss": 0.0994, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 8.79, | |
| "grad_norm": 1.2764071226119995, | |
| "learning_rate": 1.2197452229299364e-05, | |
| "loss": 0.1033, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 8.85, | |
| "grad_norm": 2.9256701469421387, | |
| "learning_rate": 1.1560509554140128e-05, | |
| "loss": 0.0982, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 8.92, | |
| "grad_norm": 5.594937324523926, | |
| "learning_rate": 1.0923566878980892e-05, | |
| "loss": 0.1219, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 8.92, | |
| "eval_accuracy": 0.7723809523809524, | |
| "eval_loss": 0.939895510673523, | |
| "eval_runtime": 14.8747, | |
| "eval_samples_per_second": 70.589, | |
| "eval_steps_per_second": 8.874, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 8.98, | |
| "grad_norm": 3.3030786514282227, | |
| "learning_rate": 1.0286624203821656e-05, | |
| "loss": 0.0996, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 9.04, | |
| "grad_norm": 1.5488649606704712, | |
| "learning_rate": 9.649681528662422e-06, | |
| "loss": 0.1477, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 9.11, | |
| "grad_norm": 4.593501091003418, | |
| "learning_rate": 9.012738853503185e-06, | |
| "loss": 0.1142, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 9.17, | |
| "grad_norm": 6.411059379577637, | |
| "learning_rate": 8.375796178343949e-06, | |
| "loss": 0.1249, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 9.24, | |
| "grad_norm": 2.7087924480438232, | |
| "learning_rate": 7.738853503184713e-06, | |
| "loss": 0.0985, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 9.24, | |
| "eval_accuracy": 0.7647619047619048, | |
| "eval_loss": 0.9940046072006226, | |
| "eval_runtime": 15.2749, | |
| "eval_samples_per_second": 68.74, | |
| "eval_steps_per_second": 8.642, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 9.3, | |
| "grad_norm": 5.092051029205322, | |
| "learning_rate": 7.1019108280254775e-06, | |
| "loss": 0.1043, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 9.36, | |
| "grad_norm": 10.890628814697266, | |
| "learning_rate": 6.464968152866242e-06, | |
| "loss": 0.1174, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 9.43, | |
| "grad_norm": 3.2413973808288574, | |
| "learning_rate": 5.8280254777070065e-06, | |
| "loss": 0.105, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 9.49, | |
| "grad_norm": 3.838075876235962, | |
| "learning_rate": 5.191082802547771e-06, | |
| "loss": 0.0857, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 9.55, | |
| "grad_norm": 2.082455635070801, | |
| "learning_rate": 4.5541401273885346e-06, | |
| "loss": 0.1069, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 9.55, | |
| "eval_accuracy": 0.7742857142857142, | |
| "eval_loss": 0.9392004013061523, | |
| "eval_runtime": 15.3798, | |
| "eval_samples_per_second": 68.271, | |
| "eval_steps_per_second": 8.583, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 9.62, | |
| "grad_norm": 0.4810134768486023, | |
| "learning_rate": 3.9171974522292995e-06, | |
| "loss": 0.0748, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 9.68, | |
| "grad_norm": 1.6180094480514526, | |
| "learning_rate": 3.280254777070064e-06, | |
| "loss": 0.1049, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 9.75, | |
| "grad_norm": 0.6423608660697937, | |
| "learning_rate": 2.6433121019108284e-06, | |
| "loss": 0.126, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 9.81, | |
| "grad_norm": 1.139814853668213, | |
| "learning_rate": 2.0063694267515925e-06, | |
| "loss": 0.0813, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 9.87, | |
| "grad_norm": 0.03859005495905876, | |
| "learning_rate": 1.3694267515923567e-06, | |
| "loss": 0.0589, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 9.87, | |
| "eval_accuracy": 0.78, | |
| "eval_loss": 0.9408173561096191, | |
| "eval_runtime": 15.1635, | |
| "eval_samples_per_second": 69.245, | |
| "eval_steps_per_second": 8.705, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 9.94, | |
| "grad_norm": 2.202677011489868, | |
| "learning_rate": 7.324840764331211e-07, | |
| "loss": 0.0856, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 3.140180826187134, | |
| "learning_rate": 9.554140127388536e-08, | |
| "loss": 0.0997, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 3140, | |
| "total_flos": 7.776878731479245e+18, | |
| "train_loss": 0.2955047018209081, | |
| "train_runtime": 3022.5495, | |
| "train_samples_per_second": 33.2, | |
| "train_steps_per_second": 1.039 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 3140, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 100, | |
| "total_flos": 7.776878731479245e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |