| { | |
| "best_metric": 28.9507, | |
| "best_model_checkpoint": "./ko-en_mbartLarge_exp20p/checkpoint-48000", | |
| "epoch": 5.569091541942221, | |
| "eval_steps": 4000, | |
| "global_step": 48000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 1.25e-05, | |
| "loss": 1.9162, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.8085, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 1.6542, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 5e-05, | |
| "loss": 1.543, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.999973747616574e-05, | |
| "loss": 1.4815, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.999894991017647e-05, | |
| "loss": 1.4549, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 4.999763731857256e-05, | |
| "loss": 1.4233, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 4.9995799728920956e-05, | |
| "loss": 1.4008, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "eval_bleu": 22.7174, | |
| "eval_gen_len": 18.7094, | |
| "eval_loss": 1.3739314079284668, | |
| "eval_runtime": 1153.4033, | |
| "eval_samples_per_second": 14.944, | |
| "eval_steps_per_second": 0.935, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 4.9993437179814526e-05, | |
| "loss": 1.3868, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 4.999054972087132e-05, | |
| "loss": 1.3612, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 4.9987137412733476e-05, | |
| "loss": 1.3519, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 4.998320032706597e-05, | |
| "loss": 1.351, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 4.997873854655512e-05, | |
| "loss": 1.3285, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 4.9973752164906806e-05, | |
| "loss": 1.2959, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 4.9968241286844556e-05, | |
| "loss": 1.3041, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 4.996220602810732e-05, | |
| "loss": 1.2847, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "eval_bleu": 24.8557, | |
| "eval_gen_len": 18.7254, | |
| "eval_loss": 1.2651795148849487, | |
| "eval_runtime": 1152.6426, | |
| "eval_samples_per_second": 14.954, | |
| "eval_steps_per_second": 0.935, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 4.995564651544704e-05, | |
| "loss": 1.2796, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 4.994856288662599e-05, | |
| "loss": 1.2328, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 4.9940955290413884e-05, | |
| "loss": 1.2225, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 4.993282388658474e-05, | |
| "loss": 1.2061, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 4.992416884591354e-05, | |
| "loss": 1.2152, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 4.991499035017266e-05, | |
| "loss": 1.1955, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 4.9905288592127996e-05, | |
| "loss": 1.1855, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 4.9895063775534965e-05, | |
| "loss": 1.2009, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "eval_bleu": 26.2074, | |
| "eval_gen_len": 18.7513, | |
| "eval_loss": 1.20820152759552, | |
| "eval_runtime": 1153.2349, | |
| "eval_samples_per_second": 14.947, | |
| "eval_steps_per_second": 0.935, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 4.9884316115134224e-05, | |
| "loss": 1.1941, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 4.987304583664712e-05, | |
| "loss": 1.1982, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 4.986125317677099e-05, | |
| "loss": 1.188, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 4.98489383831742e-05, | |
| "loss": 1.1825, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "learning_rate": 4.983610171449087e-05, | |
| "loss": 1.1948, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 4.982274344031553e-05, | |
| "loss": 1.1476, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 4.980886384119741e-05, | |
| "loss": 1.1627, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 4.979446320863455e-05, | |
| "loss": 1.1686, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "eval_bleu": 26.304, | |
| "eval_gen_len": 19.161, | |
| "eval_loss": 1.1840966939926147, | |
| "eval_runtime": 1166.1525, | |
| "eval_samples_per_second": 14.781, | |
| "eval_steps_per_second": 0.924, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "learning_rate": 4.97795418450677e-05, | |
| "loss": 1.1675, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "learning_rate": 4.976410006387394e-05, | |
| "loss": 1.1585, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "learning_rate": 4.974813818936012e-05, | |
| "loss": 1.0833, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 4.973165655675605e-05, | |
| "loss": 1.0134, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "learning_rate": 4.9714655512207414e-05, | |
| "loss": 1.0257, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 4.969713541276859e-05, | |
| "loss": 1.0206, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "learning_rate": 4.967909662639506e-05, | |
| "loss": 1.0503, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 4.9660539531935746e-05, | |
| "loss": 1.0205, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "eval_bleu": 27.8937, | |
| "eval_gen_len": 18.6638, | |
| "eval_loss": 1.144053339958191, | |
| "eval_runtime": 1147.812, | |
| "eval_samples_per_second": 15.017, | |
| "eval_steps_per_second": 0.939, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "learning_rate": 4.9641464519125e-05, | |
| "loss": 1.032, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 4.962187198857447e-05, | |
| "loss": 1.0121, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "learning_rate": 4.960176235176465e-05, | |
| "loss": 1.013, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "learning_rate": 4.958113603103627e-05, | |
| "loss": 1.0169, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 4.9559993459581375e-05, | |
| "loss": 1.0315, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "learning_rate": 4.953833508143429e-05, | |
| "loss": 1.0431, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "learning_rate": 4.951616135146226e-05, | |
| "loss": 1.0398, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "learning_rate": 4.949347273535588e-05, | |
| "loss": 1.0217, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "eval_bleu": 28.4149, | |
| "eval_gen_len": 18.6666, | |
| "eval_loss": 1.1301122903823853, | |
| "eval_runtime": 1149.3346, | |
| "eval_samples_per_second": 14.997, | |
| "eval_steps_per_second": 0.938, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "learning_rate": 4.9470269709619356e-05, | |
| "loss": 1.0313, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "learning_rate": 4.944655276156047e-05, | |
| "loss": 1.0127, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "learning_rate": 4.9422322389280354e-05, | |
| "loss": 1.0183, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "learning_rate": 4.939757910166303e-05, | |
| "loss": 0.9723, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "learning_rate": 4.937232341836471e-05, | |
| "loss": 0.8652, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "learning_rate": 4.9346555869802904e-05, | |
| "loss": 0.8811, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "learning_rate": 4.9320276997145255e-05, | |
| "loss": 0.8824, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "learning_rate": 4.929348735229821e-05, | |
| "loss": 0.8876, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "eval_bleu": 28.5803, | |
| "eval_gen_len": 18.6229, | |
| "eval_loss": 1.1270363330841064, | |
| "eval_runtime": 1153.2981, | |
| "eval_samples_per_second": 14.946, | |
| "eval_steps_per_second": 0.935, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "learning_rate": 4.926618749789539e-05, | |
| "loss": 0.8734, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "learning_rate": 4.923837800728578e-05, | |
| "loss": 0.8772, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "learning_rate": 4.921005946452171e-05, | |
| "loss": 0.8821, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "learning_rate": 4.918123246434658e-05, | |
| "loss": 0.8774, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "learning_rate": 4.915189761218237e-05, | |
| "loss": 0.8932, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "learning_rate": 4.912205552411688e-05, | |
| "loss": 0.8745, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "learning_rate": 4.9091706826890896e-05, | |
| "loss": 0.8965, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "learning_rate": 4.9060852157884893e-05, | |
| "loss": 0.9024, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "eval_bleu": 28.852, | |
| "eval_gen_len": 18.7813, | |
| "eval_loss": 1.118120551109314, | |
| "eval_runtime": 1155.1982, | |
| "eval_samples_per_second": 14.921, | |
| "eval_steps_per_second": 0.933, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "learning_rate": 4.9029492165105776e-05, | |
| "loss": 0.8909, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "learning_rate": 4.899762750717318e-05, | |
| "loss": 0.8989, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "learning_rate": 4.8965258853305676e-05, | |
| "loss": 0.9054, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "learning_rate": 4.893238688330673e-05, | |
| "loss": 0.8852, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 4.889901228755036e-05, | |
| "loss": 0.8879, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 4.06, | |
| "learning_rate": 4.886513576696674e-05, | |
| "loss": 0.7955, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 4.12, | |
| "learning_rate": 4.8830758033027385e-05, | |
| "loss": 0.795, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "learning_rate": 4.879587980773026e-05, | |
| "loss": 0.7927, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "eval_bleu": 28.3975, | |
| "eval_gen_len": 18.4863, | |
| "eval_loss": 1.1393358707427979, | |
| "eval_runtime": 1137.7417, | |
| "eval_samples_per_second": 15.15, | |
| "eval_steps_per_second": 0.947, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 4.23, | |
| "learning_rate": 4.8760501823584607e-05, | |
| "loss": 0.8031, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "learning_rate": 4.8724624823595535e-05, | |
| "loss": 0.7891, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "learning_rate": 4.8688249561248456e-05, | |
| "loss": 0.8028, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 4.41, | |
| "learning_rate": 4.8651376800493245e-05, | |
| "loss": 0.8059, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 4.47, | |
| "learning_rate": 4.861400731572818e-05, | |
| "loss": 0.8093, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 4.52, | |
| "learning_rate": 4.857614189178369e-05, | |
| "loss": 0.8044, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 4.58, | |
| "learning_rate": 4.8537781323905896e-05, | |
| "loss": 0.8098, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "learning_rate": 4.849892641773984e-05, | |
| "loss": 0.8174, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "eval_bleu": 28.6313, | |
| "eval_gen_len": 18.3916, | |
| "eval_loss": 1.124934434890747, | |
| "eval_runtime": 1132.4078, | |
| "eval_samples_per_second": 15.222, | |
| "eval_steps_per_second": 0.952, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "learning_rate": 4.845957798931265e-05, | |
| "loss": 0.821, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "learning_rate": 4.841973686501636e-05, | |
| "loss": 0.8104, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 4.81, | |
| "learning_rate": 4.8379403881590544e-05, | |
| "loss": 0.8155, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 4.87, | |
| "learning_rate": 4.833857988610474e-05, | |
| "loss": 0.8213, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 4.93, | |
| "learning_rate": 4.8297265735940714e-05, | |
| "loss": 0.819, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 4.99, | |
| "learning_rate": 4.825546229877439e-05, | |
| "loss": 0.8255, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 5.05, | |
| "learning_rate": 4.8213170452557655e-05, | |
| "loss": 0.7598, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 5.11, | |
| "learning_rate": 4.817039108549991e-05, | |
| "loss": 0.7434, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 5.11, | |
| "eval_bleu": 28.2898, | |
| "eval_gen_len": 18.7739, | |
| "eval_loss": 1.1695784330368042, | |
| "eval_runtime": 1160.4522, | |
| "eval_samples_per_second": 14.854, | |
| "eval_steps_per_second": 0.929, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 5.16, | |
| "learning_rate": 4.812712509604945e-05, | |
| "loss": 0.7564, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 5.22, | |
| "learning_rate": 4.808337339287453e-05, | |
| "loss": 0.7496, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 5.28, | |
| "learning_rate": 4.803913689484436e-05, | |
| "loss": 0.7381, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 5.34, | |
| "learning_rate": 4.7994416531009745e-05, | |
| "loss": 0.7394, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "learning_rate": 4.794921324058358e-05, | |
| "loss": 0.7468, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 5.45, | |
| "learning_rate": 4.790352797292116e-05, | |
| "loss": 0.7448, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 5.51, | |
| "learning_rate": 4.785736168750022e-05, | |
| "loss": 0.7339, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "learning_rate": 4.781071535390078e-05, | |
| "loss": 0.7416, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "eval_bleu": 28.9507, | |
| "eval_gen_len": 18.6744, | |
| "eval_loss": 1.1450848579406738, | |
| "eval_runtime": 1148.6011, | |
| "eval_samples_per_second": 15.007, | |
| "eval_steps_per_second": 0.939, | |
| "step": 48000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 344760, | |
| "num_train_epochs": 40, | |
| "save_steps": 4000, | |
| "total_flos": 1.6644596837352735e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |