| { | |
| "best_metric": 3.7010114192962646, | |
| "best_model_checkpoint": "/scratch/ka2773/project/lm-mem/checkpoints/gpt2_full_12-768-1024_a_01/checkpoint-27500", | |
| "epoch": 2.9165340969349876, | |
| "global_step": 27500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 6e-05, | |
| "loss": 6.6816, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "eval_loss": 6.0154523849487305, | |
| "eval_runtime": 9.6363, | |
| "eval_samples_per_second": 24.491, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 6e-05, | |
| "loss": 6.014, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "eval_loss": 5.75693941116333, | |
| "eval_runtime": 9.6373, | |
| "eval_samples_per_second": 24.488, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 6e-05, | |
| "loss": 5.8347, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "eval_loss": 5.60043478012085, | |
| "eval_runtime": 9.6586, | |
| "eval_samples_per_second": 24.434, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 6e-05, | |
| "loss": 5.6946, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "eval_loss": 5.451482772827148, | |
| "eval_runtime": 9.6385, | |
| "eval_samples_per_second": 24.485, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 6e-05, | |
| "loss": 5.5628, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "eval_loss": 5.31380558013916, | |
| "eval_runtime": 9.6551, | |
| "eval_samples_per_second": 24.443, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 6e-05, | |
| "loss": 5.4328, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "eval_loss": 5.174571990966797, | |
| "eval_runtime": 9.6558, | |
| "eval_samples_per_second": 24.441, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 6e-05, | |
| "loss": 5.3018, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "eval_loss": 5.04353666305542, | |
| "eval_runtime": 9.6566, | |
| "eval_samples_per_second": 24.439, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 6e-05, | |
| "loss": 5.1768, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "eval_loss": 4.923619270324707, | |
| "eval_runtime": 9.6443, | |
| "eval_samples_per_second": 24.47, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 6e-05, | |
| "loss": 5.0737, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "eval_loss": 4.811069011688232, | |
| "eval_runtime": 9.6397, | |
| "eval_samples_per_second": 24.482, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 6e-05, | |
| "loss": 4.9689, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "eval_loss": 4.7268242835998535, | |
| "eval_runtime": 9.6562, | |
| "eval_samples_per_second": 24.44, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 6e-05, | |
| "loss": 4.8992, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "eval_loss": 4.646885871887207, | |
| "eval_runtime": 9.6343, | |
| "eval_samples_per_second": 24.496, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 6e-05, | |
| "loss": 4.8284, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "eval_loss": 4.5805535316467285, | |
| "eval_runtime": 9.6391, | |
| "eval_samples_per_second": 24.484, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 6e-05, | |
| "loss": 4.77, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "eval_loss": 4.520021915435791, | |
| "eval_runtime": 9.6435, | |
| "eval_samples_per_second": 24.472, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 6e-05, | |
| "loss": 4.7052, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "eval_loss": 4.462948322296143, | |
| "eval_runtime": 9.6368, | |
| "eval_samples_per_second": 24.489, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 6e-05, | |
| "loss": 4.6502, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_loss": 4.413123607635498, | |
| "eval_runtime": 9.6344, | |
| "eval_samples_per_second": 24.496, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 6e-05, | |
| "loss": 4.6089, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "eval_loss": 4.364911079406738, | |
| "eval_runtime": 9.6314, | |
| "eval_samples_per_second": 24.503, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 6e-05, | |
| "loss": 4.5489, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "eval_loss": 4.311524391174316, | |
| "eval_runtime": 9.6277, | |
| "eval_samples_per_second": 24.513, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 6e-05, | |
| "loss": 4.5042, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "eval_loss": 4.259948253631592, | |
| "eval_runtime": 9.6594, | |
| "eval_samples_per_second": 24.432, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 6e-05, | |
| "loss": 4.454, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "eval_loss": 4.214750289916992, | |
| "eval_runtime": 9.5772, | |
| "eval_samples_per_second": 24.642, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 6e-05, | |
| "loss": 4.3863, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "eval_loss": 4.168990612030029, | |
| "eval_runtime": 9.5953, | |
| "eval_samples_per_second": 24.595, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 6e-05, | |
| "loss": 4.3451, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "eval_loss": 4.131111145019531, | |
| "eval_runtime": 9.5871, | |
| "eval_samples_per_second": 24.616, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 6e-05, | |
| "loss": 4.3209, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "eval_loss": 4.101424217224121, | |
| "eval_runtime": 9.5799, | |
| "eval_samples_per_second": 24.635, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 6e-05, | |
| "loss": 4.2959, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "eval_loss": 4.075887203216553, | |
| "eval_runtime": 9.5788, | |
| "eval_samples_per_second": 24.638, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "learning_rate": 6e-05, | |
| "loss": 4.259, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "eval_loss": 4.046448230743408, | |
| "eval_runtime": 9.5828, | |
| "eval_samples_per_second": 24.628, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 6e-05, | |
| "loss": 4.2412, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "eval_loss": 4.025355339050293, | |
| "eval_runtime": 9.5605, | |
| "eval_samples_per_second": 24.685, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 6e-05, | |
| "loss": 4.2107, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "eval_loss": 4.008790493011475, | |
| "eval_runtime": 9.5742, | |
| "eval_samples_per_second": 24.65, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 6e-05, | |
| "loss": 4.1955, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "eval_loss": 3.980191707611084, | |
| "eval_runtime": 9.5885, | |
| "eval_samples_per_second": 24.613, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 6e-05, | |
| "loss": 4.1816, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "eval_loss": 3.955965995788574, | |
| "eval_runtime": 9.5843, | |
| "eval_samples_per_second": 24.624, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 6e-05, | |
| "loss": 4.1469, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "eval_loss": 3.9338672161102295, | |
| "eval_runtime": 9.5686, | |
| "eval_samples_per_second": 24.664, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "learning_rate": 6e-05, | |
| "loss": 4.1401, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "eval_loss": 3.91983962059021, | |
| "eval_runtime": 9.5696, | |
| "eval_samples_per_second": 24.661, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 6e-05, | |
| "loss": 4.1208, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "eval_loss": 3.906564712524414, | |
| "eval_runtime": 9.5628, | |
| "eval_samples_per_second": 24.679, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 6e-05, | |
| "loss": 4.1191, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "eval_loss": 3.89436674118042, | |
| "eval_runtime": 9.5623, | |
| "eval_samples_per_second": 24.68, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "learning_rate": 6e-05, | |
| "loss": 4.0999, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "eval_loss": 3.884828805923462, | |
| "eval_runtime": 9.5239, | |
| "eval_samples_per_second": 24.78, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 6e-05, | |
| "loss": 4.09, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "eval_loss": 3.877413034439087, | |
| "eval_runtime": 9.5641, | |
| "eval_samples_per_second": 24.675, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 6e-05, | |
| "loss": 4.084, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "eval_loss": 3.8598787784576416, | |
| "eval_runtime": 9.5727, | |
| "eval_samples_per_second": 24.653, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "learning_rate": 6e-05, | |
| "loss": 4.0619, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "eval_loss": 3.843931198120117, | |
| "eval_runtime": 9.5583, | |
| "eval_samples_per_second": 24.691, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 6e-05, | |
| "loss": 4.049, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "eval_loss": 3.8279693126678467, | |
| "eval_runtime": 9.5719, | |
| "eval_samples_per_second": 24.656, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "learning_rate": 6e-05, | |
| "loss": 4.0287, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "eval_loss": 3.817969799041748, | |
| "eval_runtime": 9.5289, | |
| "eval_samples_per_second": 24.767, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "learning_rate": 6e-05, | |
| "loss": 4.0008, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "eval_loss": 3.8063206672668457, | |
| "eval_runtime": 9.5741, | |
| "eval_samples_per_second": 24.65, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9993, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "eval_loss": 3.80794358253479, | |
| "eval_runtime": 9.5702, | |
| "eval_samples_per_second": 24.66, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9836, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "eval_loss": 3.7982654571533203, | |
| "eval_runtime": 9.5716, | |
| "eval_samples_per_second": 24.656, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9813, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "eval_loss": 3.7907371520996094, | |
| "eval_runtime": 9.5685, | |
| "eval_samples_per_second": 24.664, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9785, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "eval_loss": 3.7781858444213867, | |
| "eval_runtime": 9.5717, | |
| "eval_samples_per_second": 24.656, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9704, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "eval_loss": 3.772528648376465, | |
| "eval_runtime": 9.5802, | |
| "eval_samples_per_second": 24.634, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9741, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "eval_loss": 3.778170108795166, | |
| "eval_runtime": 9.5981, | |
| "eval_samples_per_second": 24.588, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9587, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "eval_loss": 3.7567873001098633, | |
| "eval_runtime": 9.5943, | |
| "eval_samples_per_second": 24.598, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9505, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "eval_loss": 3.7593204975128174, | |
| "eval_runtime": 9.589, | |
| "eval_samples_per_second": 24.611, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9358, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "eval_loss": 3.750127077102661, | |
| "eval_runtime": 9.5871, | |
| "eval_samples_per_second": 24.616, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9371, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "eval_loss": 3.7427163124084473, | |
| "eval_runtime": 9.5925, | |
| "eval_samples_per_second": 24.602, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9379, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "eval_loss": 3.7363016605377197, | |
| "eval_runtime": 9.5892, | |
| "eval_samples_per_second": 24.611, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9352, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "eval_loss": 3.7250401973724365, | |
| "eval_runtime": 9.5877, | |
| "eval_samples_per_second": 24.615, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9163, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "eval_loss": 3.7216522693634033, | |
| "eval_runtime": 9.5913, | |
| "eval_samples_per_second": 24.606, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9174, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "eval_loss": 3.7126359939575195, | |
| "eval_runtime": 9.5904, | |
| "eval_samples_per_second": 24.608, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9028, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "eval_loss": 3.7034430503845215, | |
| "eval_runtime": 9.599, | |
| "eval_samples_per_second": 24.586, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9076, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "eval_loss": 3.7010114192962646, | |
| "eval_runtime": 9.5974, | |
| "eval_samples_per_second": 24.59, | |
| "step": 27500 | |
| } | |
| ], | |
| "max_steps": 94290, | |
| "num_train_epochs": 10, | |
| "total_flos": 3969647640576000.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |