| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 29.12621359223301, | |
| "eval_steps": 500, | |
| "global_step": 6000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_bleu": 46.9286, | |
| "eval_gen_len": 17.6388, | |
| "eval_loss": 0.2930362820625305, | |
| "eval_meteor": 0.6401, | |
| "eval_runtime": 21.9119, | |
| "eval_samples_per_second": 24.005, | |
| "eval_steps_per_second": 1.232, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_bleu": 55.0709, | |
| "eval_gen_len": 17.481, | |
| "eval_loss": 0.17119793593883514, | |
| "eval_meteor": 0.7436, | |
| "eval_runtime": 17.1858, | |
| "eval_samples_per_second": 30.607, | |
| "eval_steps_per_second": 1.571, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "learning_rate": 0.0002757281553398058, | |
| "loss": 1.4069, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_bleu": 56.5722, | |
| "eval_gen_len": 17.635, | |
| "eval_loss": 0.12680290639400482, | |
| "eval_meteor": 0.7412, | |
| "eval_runtime": 17.5013, | |
| "eval_samples_per_second": 30.055, | |
| "eval_steps_per_second": 1.543, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_bleu": 59.7915, | |
| "eval_gen_len": 17.5513, | |
| "eval_loss": 0.10222692042589188, | |
| "eval_meteor": 0.7776, | |
| "eval_runtime": 17.1596, | |
| "eval_samples_per_second": 30.653, | |
| "eval_steps_per_second": 1.573, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 4.85, | |
| "learning_rate": 0.00025145631067961165, | |
| "loss": 0.2339, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_bleu": 60.1489, | |
| "eval_gen_len": 17.6255, | |
| "eval_loss": 0.09315221011638641, | |
| "eval_meteor": 0.7764, | |
| "eval_runtime": 17.1827, | |
| "eval_samples_per_second": 30.612, | |
| "eval_steps_per_second": 1.571, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_bleu": 60.441, | |
| "eval_gen_len": 17.6749, | |
| "eval_loss": 0.08468244969844818, | |
| "eval_meteor": 0.7696, | |
| "eval_runtime": 17.5475, | |
| "eval_samples_per_second": 29.976, | |
| "eval_steps_per_second": 1.539, | |
| "step": 1236 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_bleu": 61.5099, | |
| "eval_gen_len": 17.6027, | |
| "eval_loss": 0.07199732959270477, | |
| "eval_meteor": 0.7875, | |
| "eval_runtime": 17.5698, | |
| "eval_samples_per_second": 29.938, | |
| "eval_steps_per_second": 1.537, | |
| "step": 1442 | |
| }, | |
| { | |
| "epoch": 7.28, | |
| "learning_rate": 0.00022718446601941746, | |
| "loss": 0.1661, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_bleu": 61.471, | |
| "eval_gen_len": 17.6901, | |
| "eval_loss": 0.07024037092924118, | |
| "eval_meteor": 0.7778, | |
| "eval_runtime": 17.3131, | |
| "eval_samples_per_second": 30.382, | |
| "eval_steps_per_second": 1.56, | |
| "step": 1648 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_bleu": 62.8426, | |
| "eval_gen_len": 17.6274, | |
| "eval_loss": 0.05566093325614929, | |
| "eval_meteor": 0.7954, | |
| "eval_runtime": 18.4096, | |
| "eval_samples_per_second": 28.572, | |
| "eval_steps_per_second": 1.467, | |
| "step": 1854 | |
| }, | |
| { | |
| "epoch": 9.71, | |
| "learning_rate": 0.00020291262135922327, | |
| "loss": 0.1289, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_bleu": 62.5415, | |
| "eval_gen_len": 17.6825, | |
| "eval_loss": 0.05298692733049393, | |
| "eval_meteor": 0.7883, | |
| "eval_runtime": 17.3063, | |
| "eval_samples_per_second": 30.394, | |
| "eval_steps_per_second": 1.56, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_bleu": 62.8065, | |
| "eval_gen_len": 17.711, | |
| "eval_loss": 0.049979548901319504, | |
| "eval_meteor": 0.7881, | |
| "eval_runtime": 17.359, | |
| "eval_samples_per_second": 30.301, | |
| "eval_steps_per_second": 1.555, | |
| "step": 2266 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_bleu": 62.6083, | |
| "eval_gen_len": 17.7167, | |
| "eval_loss": 0.04917723312973976, | |
| "eval_meteor": 0.7826, | |
| "eval_runtime": 17.3203, | |
| "eval_samples_per_second": 30.369, | |
| "eval_steps_per_second": 1.559, | |
| "step": 2472 | |
| }, | |
| { | |
| "epoch": 12.14, | |
| "learning_rate": 0.00017864077669902913, | |
| "loss": 0.1034, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_bleu": 63.5012, | |
| "eval_gen_len": 17.6844, | |
| "eval_loss": 0.042322106659412384, | |
| "eval_meteor": 0.7943, | |
| "eval_runtime": 17.3469, | |
| "eval_samples_per_second": 30.322, | |
| "eval_steps_per_second": 1.556, | |
| "step": 2678 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_bleu": 62.9858, | |
| "eval_gen_len": 17.7091, | |
| "eval_loss": 0.04112717881798744, | |
| "eval_meteor": 0.7876, | |
| "eval_runtime": 17.2586, | |
| "eval_samples_per_second": 30.478, | |
| "eval_steps_per_second": 1.564, | |
| "step": 2884 | |
| }, | |
| { | |
| "epoch": 14.56, | |
| "learning_rate": 0.00015436893203883494, | |
| "loss": 0.0862, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_bleu": 64.4301, | |
| "eval_gen_len": 17.6806, | |
| "eval_loss": 0.03431503847241402, | |
| "eval_meteor": 0.8036, | |
| "eval_runtime": 17.1867, | |
| "eval_samples_per_second": 30.605, | |
| "eval_steps_per_second": 1.571, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_bleu": 65.3347, | |
| "eval_gen_len": 17.654, | |
| "eval_loss": 0.031859882175922394, | |
| "eval_meteor": 0.8153, | |
| "eval_runtime": 17.2343, | |
| "eval_samples_per_second": 30.52, | |
| "eval_steps_per_second": 1.567, | |
| "step": 3296 | |
| }, | |
| { | |
| "epoch": 16.99, | |
| "learning_rate": 0.00013009708737864078, | |
| "loss": 0.073, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_bleu": 62.9474, | |
| "eval_gen_len": 17.7433, | |
| "eval_loss": 0.03299812600016594, | |
| "eval_meteor": 0.7817, | |
| "eval_runtime": 17.342, | |
| "eval_samples_per_second": 30.331, | |
| "eval_steps_per_second": 1.557, | |
| "step": 3502 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_bleu": 66.3794, | |
| "eval_gen_len": 17.6217, | |
| "eval_loss": 0.02293582074344158, | |
| "eval_meteor": 0.8285, | |
| "eval_runtime": 17.2657, | |
| "eval_samples_per_second": 30.465, | |
| "eval_steps_per_second": 1.564, | |
| "step": 3708 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_bleu": 64.1881, | |
| "eval_gen_len": 17.711, | |
| "eval_loss": 0.02707734704017639, | |
| "eval_meteor": 0.7974, | |
| "eval_runtime": 17.8576, | |
| "eval_samples_per_second": 29.455, | |
| "eval_steps_per_second": 1.512, | |
| "step": 3914 | |
| }, | |
| { | |
| "epoch": 19.42, | |
| "learning_rate": 0.00010582524271844659, | |
| "loss": 0.0609, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_bleu": 65.4997, | |
| "eval_gen_len": 17.6616, | |
| "eval_loss": 0.022358747199177742, | |
| "eval_meteor": 0.8156, | |
| "eval_runtime": 17.5005, | |
| "eval_samples_per_second": 30.056, | |
| "eval_steps_per_second": 1.543, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_bleu": 64.8991, | |
| "eval_gen_len": 17.6863, | |
| "eval_loss": 0.022536493837833405, | |
| "eval_meteor": 0.8061, | |
| "eval_runtime": 17.556, | |
| "eval_samples_per_second": 29.961, | |
| "eval_steps_per_second": 1.538, | |
| "step": 4326 | |
| }, | |
| { | |
| "epoch": 21.84, | |
| "learning_rate": 8.155339805825241e-05, | |
| "loss": 0.0538, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_bleu": 66.3962, | |
| "eval_gen_len": 17.6407, | |
| "eval_loss": 0.018802011385560036, | |
| "eval_meteor": 0.8251, | |
| "eval_runtime": 17.5884, | |
| "eval_samples_per_second": 29.906, | |
| "eval_steps_per_second": 1.535, | |
| "step": 4532 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_bleu": 65.0479, | |
| "eval_gen_len": 17.6863, | |
| "eval_loss": 0.02161051519215107, | |
| "eval_meteor": 0.8069, | |
| "eval_runtime": 17.501, | |
| "eval_samples_per_second": 30.055, | |
| "eval_steps_per_second": 1.543, | |
| "step": 4738 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_bleu": 66.0078, | |
| "eval_gen_len": 17.6559, | |
| "eval_loss": 0.01826297491788864, | |
| "eval_meteor": 0.8205, | |
| "eval_runtime": 17.773, | |
| "eval_samples_per_second": 29.595, | |
| "eval_steps_per_second": 1.519, | |
| "step": 4944 | |
| }, | |
| { | |
| "epoch": 24.27, | |
| "learning_rate": 5.728155339805825e-05, | |
| "loss": 0.0476, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_bleu": 65.1378, | |
| "eval_gen_len": 17.6863, | |
| "eval_loss": 0.021101167425513268, | |
| "eval_meteor": 0.8087, | |
| "eval_runtime": 17.5024, | |
| "eval_samples_per_second": 30.053, | |
| "eval_steps_per_second": 1.543, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_bleu": 65.5523, | |
| "eval_gen_len": 17.6749, | |
| "eval_loss": 0.018090983852744102, | |
| "eval_meteor": 0.8134, | |
| "eval_runtime": 17.7903, | |
| "eval_samples_per_second": 29.567, | |
| "eval_steps_per_second": 1.518, | |
| "step": 5356 | |
| }, | |
| { | |
| "epoch": 26.7, | |
| "learning_rate": 3.3009708737864073e-05, | |
| "loss": 0.044, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "eval_bleu": 66.1951, | |
| "eval_gen_len": 17.6597, | |
| "eval_loss": 0.016722695901989937, | |
| "eval_meteor": 0.8212, | |
| "eval_runtime": 17.6724, | |
| "eval_samples_per_second": 29.764, | |
| "eval_steps_per_second": 1.528, | |
| "step": 5562 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_bleu": 66.2812, | |
| "eval_gen_len": 17.6559, | |
| "eval_loss": 0.0156955998390913, | |
| "eval_meteor": 0.8226, | |
| "eval_runtime": 17.5211, | |
| "eval_samples_per_second": 30.021, | |
| "eval_steps_per_second": 1.541, | |
| "step": 5768 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "eval_bleu": 65.4079, | |
| "eval_gen_len": 17.6863, | |
| "eval_loss": 0.017170317471027374, | |
| "eval_meteor": 0.8103, | |
| "eval_runtime": 17.494, | |
| "eval_samples_per_second": 30.067, | |
| "eval_steps_per_second": 1.543, | |
| "step": 5974 | |
| }, | |
| { | |
| "epoch": 29.13, | |
| "learning_rate": 8.737864077669902e-06, | |
| "loss": 0.04, | |
| "step": 6000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 6180, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 500, | |
| "total_flos": 1.55075881506816e+16, | |
| "train_batch_size": 20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |