| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 300.0, | |
| "global_step": 11700, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 12.82, | |
| "learning_rate": 1.2375000000000001e-05, | |
| "loss": 3.4354, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 12.82, | |
| "eval_bleu": 56.6427, | |
| "eval_em": 0.0, | |
| "eval_gen_len": 70.5947, | |
| "eval_loss": 1.5065408945083618, | |
| "eval_rm": 0.0, | |
| "eval_runtime": 175.1978, | |
| "eval_samples_per_second": 2.38, | |
| "eval_steps_per_second": 0.303, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 25.64, | |
| "learning_rate": 2.4875e-05, | |
| "loss": 0.8473, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 25.64, | |
| "eval_bleu": 90.5419, | |
| "eval_em": 0.0192, | |
| "eval_gen_len": 76.9736, | |
| "eval_loss": 0.3859139084815979, | |
| "eval_rm": 0.0216, | |
| "eval_runtime": 165.5619, | |
| "eval_samples_per_second": 2.519, | |
| "eval_steps_per_second": 0.32, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 38.46, | |
| "learning_rate": 3.737500000000001e-05, | |
| "loss": 0.2049, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 38.46, | |
| "eval_bleu": 93.6495, | |
| "eval_em": 0.0504, | |
| "eval_gen_len": 75.1655, | |
| "eval_loss": 0.24716846644878387, | |
| "eval_rm": 0.0671, | |
| "eval_runtime": 167.8405, | |
| "eval_samples_per_second": 2.485, | |
| "eval_steps_per_second": 0.316, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 51.28, | |
| "learning_rate": 4.9875000000000006e-05, | |
| "loss": 0.1222, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 51.28, | |
| "eval_bleu": 93.8388, | |
| "eval_em": 0.0959, | |
| "eval_gen_len": 75.6403, | |
| "eval_loss": 0.23381924629211426, | |
| "eval_rm": 0.1487, | |
| "eval_runtime": 164.0184, | |
| "eval_samples_per_second": 2.542, | |
| "eval_steps_per_second": 0.323, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 64.1, | |
| "learning_rate": 4.7448453608247423e-05, | |
| "loss": 0.0923, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 64.1, | |
| "eval_bleu": 94.71, | |
| "eval_em": 0.2158, | |
| "eval_gen_len": 75.8177, | |
| "eval_loss": 0.19438204169273376, | |
| "eval_rm": 0.2662, | |
| "eval_runtime": 166.6507, | |
| "eval_samples_per_second": 2.502, | |
| "eval_steps_per_second": 0.318, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 76.92, | |
| "learning_rate": 4.487113402061856e-05, | |
| "loss": 0.0752, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 76.92, | |
| "eval_bleu": 95.0458, | |
| "eval_em": 0.2662, | |
| "eval_gen_len": 75.2638, | |
| "eval_loss": 0.19899217784404755, | |
| "eval_rm": 0.3022, | |
| "eval_runtime": 165.9288, | |
| "eval_samples_per_second": 2.513, | |
| "eval_steps_per_second": 0.319, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 89.74, | |
| "learning_rate": 4.229381443298969e-05, | |
| "loss": 0.0627, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 89.74, | |
| "eval_bleu": 95.3518, | |
| "eval_em": 0.3429, | |
| "eval_gen_len": 76.9928, | |
| "eval_loss": 0.195655956864357, | |
| "eval_rm": 0.3957, | |
| "eval_runtime": 164.1213, | |
| "eval_samples_per_second": 2.541, | |
| "eval_steps_per_second": 0.323, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 102.56, | |
| "learning_rate": 3.9716494845360825e-05, | |
| "loss": 0.052, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 102.56, | |
| "eval_bleu": 95.5392, | |
| "eval_em": 0.3837, | |
| "eval_gen_len": 76.1007, | |
| "eval_loss": 0.18605293333530426, | |
| "eval_rm": 0.4508, | |
| "eval_runtime": 163.8256, | |
| "eval_samples_per_second": 2.545, | |
| "eval_steps_per_second": 0.324, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 115.38, | |
| "learning_rate": 3.713917525773196e-05, | |
| "loss": 0.0457, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 115.38, | |
| "eval_bleu": 95.6692, | |
| "eval_em": 0.4173, | |
| "eval_gen_len": 76.1727, | |
| "eval_loss": 0.187970370054245, | |
| "eval_rm": 0.4892, | |
| "eval_runtime": 165.0086, | |
| "eval_samples_per_second": 2.527, | |
| "eval_steps_per_second": 0.321, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 128.21, | |
| "learning_rate": 3.4561855670103095e-05, | |
| "loss": 0.0386, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 128.21, | |
| "eval_bleu": 95.9215, | |
| "eval_em": 0.446, | |
| "eval_gen_len": 76.0168, | |
| "eval_loss": 0.18496404588222504, | |
| "eval_rm": 0.5276, | |
| "eval_runtime": 160.7718, | |
| "eval_samples_per_second": 2.594, | |
| "eval_steps_per_second": 0.33, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 141.03, | |
| "learning_rate": 3.1984536082474226e-05, | |
| "loss": 0.0321, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 141.03, | |
| "eval_bleu": 95.931, | |
| "eval_em": 0.4964, | |
| "eval_gen_len": 75.2566, | |
| "eval_loss": 0.17244744300842285, | |
| "eval_rm": 0.5875, | |
| "eval_runtime": 162.2245, | |
| "eval_samples_per_second": 2.571, | |
| "eval_steps_per_second": 0.327, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 153.85, | |
| "learning_rate": 2.9407216494845364e-05, | |
| "loss": 0.026, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 153.85, | |
| "eval_bleu": 96.4317, | |
| "eval_em": 0.5348, | |
| "eval_gen_len": 75.741, | |
| "eval_loss": 0.16870950162410736, | |
| "eval_rm": 0.6499, | |
| "eval_runtime": 165.0932, | |
| "eval_samples_per_second": 2.526, | |
| "eval_steps_per_second": 0.321, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 166.67, | |
| "learning_rate": 2.6829896907216496e-05, | |
| "loss": 0.0242, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 166.67, | |
| "eval_bleu": 96.197, | |
| "eval_em": 0.5372, | |
| "eval_gen_len": 76.1127, | |
| "eval_loss": 0.17071698606014252, | |
| "eval_rm": 0.6403, | |
| "eval_runtime": 162.7041, | |
| "eval_samples_per_second": 2.563, | |
| "eval_steps_per_second": 0.326, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 179.49, | |
| "learning_rate": 2.425257731958763e-05, | |
| "loss": 0.0193, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 179.49, | |
| "eval_bleu": 96.3422, | |
| "eval_em": 0.5564, | |
| "eval_gen_len": 75.3933, | |
| "eval_loss": 0.1643209457397461, | |
| "eval_rm": 0.6691, | |
| "eval_runtime": 163.0211, | |
| "eval_samples_per_second": 2.558, | |
| "eval_steps_per_second": 0.325, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 192.31, | |
| "learning_rate": 2.1675257731958766e-05, | |
| "loss": 0.0164, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 192.31, | |
| "eval_bleu": 96.5278, | |
| "eval_em": 0.5779, | |
| "eval_gen_len": 75.4508, | |
| "eval_loss": 0.16497784852981567, | |
| "eval_rm": 0.693, | |
| "eval_runtime": 161.7709, | |
| "eval_samples_per_second": 2.578, | |
| "eval_steps_per_second": 0.328, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 205.13, | |
| "learning_rate": 1.9097938144329897e-05, | |
| "loss": 0.0139, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 205.13, | |
| "eval_bleu": 96.6382, | |
| "eval_em": 0.6091, | |
| "eval_gen_len": 75.9592, | |
| "eval_loss": 0.16682015359401703, | |
| "eval_rm": 0.7314, | |
| "eval_runtime": 160.3701, | |
| "eval_samples_per_second": 2.6, | |
| "eval_steps_per_second": 0.33, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 217.95, | |
| "learning_rate": 1.6520618556701032e-05, | |
| "loss": 0.012, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 217.95, | |
| "eval_bleu": 96.5488, | |
| "eval_em": 0.6163, | |
| "eval_gen_len": 76.0024, | |
| "eval_loss": 0.16442929208278656, | |
| "eval_rm": 0.729, | |
| "eval_runtime": 161.7705, | |
| "eval_samples_per_second": 2.578, | |
| "eval_steps_per_second": 0.328, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 230.77, | |
| "learning_rate": 1.3943298969072165e-05, | |
| "loss": 0.0106, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 230.77, | |
| "eval_bleu": 96.6353, | |
| "eval_em": 0.6091, | |
| "eval_gen_len": 75.5468, | |
| "eval_loss": 0.16534733772277832, | |
| "eval_rm": 0.7266, | |
| "eval_runtime": 158.2739, | |
| "eval_samples_per_second": 2.635, | |
| "eval_steps_per_second": 0.335, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 243.59, | |
| "learning_rate": 1.1365979381443299e-05, | |
| "loss": 0.0093, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 243.59, | |
| "eval_bleu": 96.8984, | |
| "eval_em": 0.6331, | |
| "eval_gen_len": 75.7242, | |
| "eval_loss": 0.16627563536167145, | |
| "eval_rm": 0.7482, | |
| "eval_runtime": 159.472, | |
| "eval_samples_per_second": 2.615, | |
| "eval_steps_per_second": 0.332, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 256.41, | |
| "learning_rate": 8.788659793814432e-06, | |
| "loss": 0.0084, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 256.41, | |
| "eval_bleu": 96.6199, | |
| "eval_em": 0.6331, | |
| "eval_gen_len": 75.3885, | |
| "eval_loss": 0.1675705760717392, | |
| "eval_rm": 0.7482, | |
| "eval_runtime": 157.5771, | |
| "eval_samples_per_second": 2.646, | |
| "eval_steps_per_second": 0.336, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 269.23, | |
| "learning_rate": 6.211340206185568e-06, | |
| "loss": 0.0076, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 269.23, | |
| "eval_bleu": 96.5038, | |
| "eval_em": 0.6283, | |
| "eval_gen_len": 75.3453, | |
| "eval_loss": 0.16782505810260773, | |
| "eval_rm": 0.7482, | |
| "eval_runtime": 181.3202, | |
| "eval_samples_per_second": 2.3, | |
| "eval_steps_per_second": 0.292, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 282.05, | |
| "learning_rate": 3.6340206185567013e-06, | |
| "loss": 0.007, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 282.05, | |
| "eval_bleu": 96.7187, | |
| "eval_em": 0.6355, | |
| "eval_gen_len": 75.9281, | |
| "eval_loss": 0.16688644886016846, | |
| "eval_rm": 0.7458, | |
| "eval_runtime": 181.4265, | |
| "eval_samples_per_second": 2.298, | |
| "eval_steps_per_second": 0.292, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 294.87, | |
| "learning_rate": 1.0567010309278351e-06, | |
| "loss": 0.0065, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 294.87, | |
| "eval_bleu": 96.7679, | |
| "eval_em": 0.6307, | |
| "eval_gen_len": 75.6355, | |
| "eval_loss": 0.16715963184833527, | |
| "eval_rm": 0.7482, | |
| "eval_runtime": 181.7054, | |
| "eval_samples_per_second": 2.295, | |
| "eval_steps_per_second": 0.292, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 300.0, | |
| "step": 11700, | |
| "total_flos": 9455707744902144.0, | |
| "train_loss": 0.001008551752465403, | |
| "train_runtime": 1876.4514, | |
| "train_samples_per_second": 196.008, | |
| "train_steps_per_second": 6.235 | |
| } | |
| ], | |
| "max_steps": 11700, | |
| "num_train_epochs": 300, | |
| "total_flos": 9455707744902144.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |