{ "best_metric": 62.7436, "best_model_checkpoint": "models/led-base-16384/finetune_led_4096_global_on_highlights_pretrained_CNNDM_full_and_duc/checkpoint-1300", "epoch": 5.439330543933054, "global_step": 1300, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.21, "learning_rate": 4.895397489539749e-05, "loss": 1.6206, "step": 50 }, { "epoch": 0.42, "learning_rate": 4.7907949790794984e-05, "loss": 1.5797, "step": 100 }, { "epoch": 0.42, "eval_gen_len": 126.07751937984496, "eval_gold_content__rouge1": 58.3427, "eval_gold_content__rouge1_precision": 59.85, "eval_gold_content__rouge1_recall": 57.438, "eval_gold_content__rouge2": 33.9535, "eval_gold_content__rouge2_precision": 34.901, "eval_gold_content__rouge2_recall": 33.41, "eval_gold_content__rougeL": 43.3591, "eval_gold_content__rougeL_precision": 44.5351, "eval_gold_content__rougeL_recall": 42.6937, "eval_gold_content__rougeLsum": 43.5008, "eval_gold_content__rougeLsum_precision": 44.6304, "eval_gold_content__rougeLsum_recall": 42.8334, "eval_gold_rouge1": 61.89, "eval_gold_rouge1_precision": 63.1594, "eval_gold_rouge1_recall": 61.1538, "eval_gold_rouge2": 36.4056, "eval_gold_rouge2_precision": 37.1918, "eval_gold_rouge2_recall": 35.9234, "eval_gold_rougeL": 43.5642, "eval_gold_rougeL_precision": 44.4054, "eval_gold_rougeL_recall": 43.0498, "eval_gold_rougeLsum": 58.5201, "eval_gold_rougeLsum_precision": 59.7136, "eval_gold_rougeLsum_recall": 57.7912, "eval_loss": 2.1532697677612305, "eval_runtime": 308.7055, "eval_samples_per_second": 0.418, "eval_steps_per_second": 0.418, "step": 100 }, { "epoch": 0.63, "learning_rate": 4.686192468619247e-05, "loss": 1.5908, "step": 150 }, { "epoch": 0.84, "learning_rate": 4.581589958158996e-05, "loss": 1.5953, "step": 200 }, { "epoch": 0.84, "eval_gen_len": 133.74418604651163, "eval_gold_content__rouge1": 58.4517, "eval_gold_content__rouge1_precision": 58.1392, "eval_gold_content__rouge1_recall": 59.3081, "eval_gold_content__rouge2": 34.4547, "eval_gold_content__rouge2_precision": 34.347, "eval_gold_content__rouge2_recall": 34.8401, "eval_gold_content__rougeL": 43.6875, "eval_gold_content__rougeL_precision": 43.5151, "eval_gold_content__rougeL_recall": 44.2304, "eval_gold_content__rougeLsum": 43.6718, "eval_gold_content__rougeLsum_precision": 43.5039, "eval_gold_content__rougeLsum_recall": 44.2415, "eval_gold_rouge1": 62.0262, "eval_gold_rouge1_precision": 61.4937, "eval_gold_rouge1_recall": 63.0726, "eval_gold_rouge2": 36.8819, "eval_gold_rouge2_precision": 36.6756, "eval_gold_rouge2_recall": 37.3674, "eval_gold_rougeL": 43.6689, "eval_gold_rougeL_precision": 43.3582, "eval_gold_rougeL_recall": 44.3137, "eval_gold_rougeLsum": 58.5109, "eval_gold_rougeLsum_precision": 58.0665, "eval_gold_rougeLsum_recall": 59.4408, "eval_loss": 2.1124985218048096, "eval_runtime": 327.1737, "eval_samples_per_second": 0.394, "eval_steps_per_second": 0.394, "step": 200 }, { "epoch": 1.05, "learning_rate": 4.476987447698745e-05, "loss": 1.5117, "step": 250 }, { "epoch": 1.26, "learning_rate": 4.372384937238494e-05, "loss": 1.2453, "step": 300 }, { "epoch": 1.26, "eval_gen_len": 127.17054263565892, "eval_gold_content__rouge1": 58.3354, "eval_gold_content__rouge1_precision": 58.9279, "eval_gold_content__rouge1_recall": 58.2257, "eval_gold_content__rouge2": 34.017, "eval_gold_content__rouge2_precision": 34.5263, "eval_gold_content__rouge2_recall": 33.778, "eval_gold_content__rougeL": 43.6636, "eval_gold_content__rougeL_precision": 44.2381, "eval_gold_content__rougeL_recall": 43.4675, "eval_gold_content__rougeLsum": 43.6821, "eval_gold_content__rougeLsum_precision": 44.3055, "eval_gold_content__rougeLsum_recall": 43.4534, "eval_gold_rouge1": 61.767, "eval_gold_rouge1_precision": 62.401, "eval_gold_rouge1_recall": 61.667, "eval_gold_rouge2": 36.1801, "eval_gold_rouge2_precision": 36.6528, "eval_gold_rouge2_recall": 35.9826, "eval_gold_rougeL": 43.7391, "eval_gold_rougeL_precision": 44.2889, "eval_gold_rougeL_recall": 43.5425, "eval_gold_rougeLsum": 58.1127, "eval_gold_rougeLsum_precision": 58.7827, "eval_gold_rougeLsum_recall": 57.9085, "eval_loss": 2.1830742359161377, "eval_runtime": 306.4961, "eval_samples_per_second": 0.421, "eval_steps_per_second": 0.421, "step": 300 }, { "epoch": 1.46, "learning_rate": 4.267782426778243e-05, "loss": 1.2287, "step": 350 }, { "epoch": 1.67, "learning_rate": 4.1631799163179915e-05, "loss": 1.2763, "step": 400 }, { "epoch": 1.67, "eval_gen_len": 129.26356589147287, "eval_gold_content__rouge1": 58.9265, "eval_gold_content__rouge1_precision": 59.6196, "eval_gold_content__rouge1_recall": 58.889, "eval_gold_content__rouge2": 34.0912, "eval_gold_content__rouge2_precision": 34.4828, "eval_gold_content__rouge2_recall": 34.0341, "eval_gold_content__rougeL": 43.4803, "eval_gold_content__rougeL_precision": 43.8369, "eval_gold_content__rougeL_recall": 43.5011, "eval_gold_content__rougeLsum": 43.4751, "eval_gold_content__rougeLsum_precision": 43.7794, "eval_gold_content__rougeLsum_recall": 43.4949, "eval_gold_rouge1": 62.3505, "eval_gold_rouge1_precision": 62.9951, "eval_gold_rouge1_recall": 62.2874, "eval_gold_rouge2": 36.4292, "eval_gold_rouge2_precision": 36.8521, "eval_gold_rouge2_recall": 36.3282, "eval_gold_rougeL": 43.8949, "eval_gold_rougeL_precision": 44.2859, "eval_gold_rougeL_recall": 43.8659, "eval_gold_rougeLsum": 58.9232, "eval_gold_rougeLsum_precision": 59.5425, "eval_gold_rougeLsum_recall": 58.8475, "eval_loss": 2.142529249191284, "eval_runtime": 305.8865, "eval_samples_per_second": 0.422, "eval_steps_per_second": 0.422, "step": 400 }, { "epoch": 1.88, "learning_rate": 4.0585774058577406e-05, "loss": 1.2922, "step": 450 }, { "epoch": 2.09, "learning_rate": 3.95397489539749e-05, "loss": 1.0938, "step": 500 }, { "epoch": 2.09, "eval_gen_len": 133.6046511627907, "eval_gold_content__rouge1": 58.6518, "eval_gold_content__rouge1_precision": 58.3803, "eval_gold_content__rouge1_recall": 59.3574, "eval_gold_content__rouge2": 33.2829, "eval_gold_content__rouge2_precision": 33.2277, "eval_gold_content__rouge2_recall": 33.588, "eval_gold_content__rougeL": 42.8525, "eval_gold_content__rougeL_precision": 42.7057, "eval_gold_content__rougeL_recall": 43.3466, "eval_gold_content__rougeLsum": 42.8745, "eval_gold_content__rougeLsum_precision": 42.7141, "eval_gold_content__rougeLsum_recall": 43.3652, "eval_gold_rouge1": 62.1112, "eval_gold_rouge1_precision": 61.6416, "eval_gold_rouge1_recall": 63.0257, "eval_gold_rouge2": 35.6622, "eval_gold_rouge2_precision": 35.5027, "eval_gold_rouge2_recall": 36.0691, "eval_gold_rougeL": 42.9354, "eval_gold_rougeL_precision": 42.6368, "eval_gold_rougeL_recall": 43.5109, "eval_gold_rougeLsum": 58.3501, "eval_gold_rougeLsum_precision": 57.9716, "eval_gold_rougeLsum_recall": 59.1286, "eval_loss": 2.240668535232544, "eval_runtime": 327.5299, "eval_samples_per_second": 0.394, "eval_steps_per_second": 0.394, "step": 500 }, { "epoch": 2.3, "learning_rate": 3.849372384937239e-05, "loss": 1.0109, "step": 550 }, { "epoch": 2.51, "learning_rate": 3.744769874476988e-05, "loss": 0.9934, "step": 600 }, { "epoch": 2.51, "eval_gen_len": 130.85271317829458, "eval_gold_content__rouge1": 58.5347, "eval_gold_content__rouge1_precision": 58.7796, "eval_gold_content__rouge1_recall": 58.8349, "eval_gold_content__rouge2": 33.8458, "eval_gold_content__rouge2_precision": 33.9726, "eval_gold_content__rouge2_recall": 34.0374, "eval_gold_content__rougeL": 43.289, "eval_gold_content__rougeL_precision": 43.4375, "eval_gold_content__rougeL_recall": 43.4558, "eval_gold_content__rougeLsum": 43.2637, "eval_gold_content__rougeLsum_precision": 43.4389, "eval_gold_content__rougeLsum_recall": 43.486, "eval_gold_rouge1": 62.1317, "eval_gold_rouge1_precision": 62.2568, "eval_gold_rouge1_recall": 62.4558, "eval_gold_rouge2": 36.5123, "eval_gold_rouge2_precision": 36.6028, "eval_gold_rouge2_recall": 36.7019, "eval_gold_rougeL": 43.3287, "eval_gold_rougeL_precision": 43.4139, "eval_gold_rougeL_recall": 43.5481, "eval_gold_rougeLsum": 58.7645, "eval_gold_rougeLsum_precision": 58.911, "eval_gold_rougeLsum_recall": 59.0483, "eval_loss": 2.2259631156921387, "eval_runtime": 327.0684, "eval_samples_per_second": 0.394, "eval_steps_per_second": 0.394, "step": 600 }, { "epoch": 2.72, "learning_rate": 3.640167364016737e-05, "loss": 1.012, "step": 650 }, { "epoch": 2.93, "learning_rate": 3.5355648535564854e-05, "loss": 1.04, "step": 700 }, { "epoch": 2.93, "eval_gen_len": 128.65891472868216, "eval_gold_content__rouge1": 58.99, "eval_gold_content__rouge1_precision": 59.5164, "eval_gold_content__rouge1_recall": 58.9141, "eval_gold_content__rouge2": 33.6019, "eval_gold_content__rouge2_precision": 33.946, "eval_gold_content__rouge2_recall": 33.5476, "eval_gold_content__rougeL": 43.6877, "eval_gold_content__rougeL_precision": 44.0417, "eval_gold_content__rougeL_recall": 43.7057, "eval_gold_content__rougeLsum": 43.621, "eval_gold_content__rougeLsum_precision": 44.0079, "eval_gold_content__rougeLsum_recall": 43.6015, "eval_gold_rouge1": 62.5394, "eval_gold_rouge1_precision": 63.0444, "eval_gold_rouge1_recall": 62.4707, "eval_gold_rouge2": 36.2644, "eval_gold_rouge2_precision": 36.6473, "eval_gold_rouge2_recall": 36.1616, "eval_gold_rougeL": 43.9778, "eval_gold_rougeL_precision": 44.2802, "eval_gold_rougeL_recall": 43.943, "eval_gold_rougeLsum": 58.8109, "eval_gold_rougeLsum_precision": 59.3138, "eval_gold_rougeLsum_recall": 58.7047, "eval_loss": 2.248823404312134, "eval_runtime": 305.3298, "eval_samples_per_second": 0.422, "eval_steps_per_second": 0.422, "step": 700 }, { "epoch": 3.14, "learning_rate": 3.4309623430962344e-05, "loss": 0.8781, "step": 750 }, { "epoch": 3.35, "learning_rate": 3.3263598326359835e-05, "loss": 0.8043, "step": 800 }, { "epoch": 3.35, "eval_gen_len": 132.66666666666666, "eval_gold_content__rouge1": 58.6355, "eval_gold_content__rouge1_precision": 58.401, "eval_gold_content__rouge1_recall": 59.3404, "eval_gold_content__rouge2": 32.7864, "eval_gold_content__rouge2_precision": 32.6957, "eval_gold_content__rouge2_recall": 33.0917, "eval_gold_content__rougeL": 42.5009, "eval_gold_content__rougeL_precision": 42.3598, "eval_gold_content__rougeL_recall": 42.9895, "eval_gold_content__rougeLsum": 42.5917, "eval_gold_content__rougeLsum_precision": 42.4082, "eval_gold_content__rougeLsum_recall": 43.0791, "eval_gold_rouge1": 62.2945, "eval_gold_rouge1_precision": 62.0757, "eval_gold_rouge1_recall": 62.9913, "eval_gold_rouge2": 35.3636, "eval_gold_rouge2_precision": 35.2951, "eval_gold_rouge2_recall": 35.6452, "eval_gold_rougeL": 42.9827, "eval_gold_rougeL_precision": 42.8331, "eval_gold_rougeL_recall": 43.4049, "eval_gold_rougeLsum": 58.3145, "eval_gold_rougeLsum_precision": 58.1325, "eval_gold_rougeLsum_recall": 58.895, "eval_loss": 2.3569345474243164, "eval_runtime": 317.4014, "eval_samples_per_second": 0.406, "eval_steps_per_second": 0.406, "step": 800 }, { "epoch": 3.56, "learning_rate": 3.2217573221757326e-05, "loss": 0.8277, "step": 850 }, { "epoch": 3.77, "learning_rate": 3.117154811715482e-05, "loss": 0.8269, "step": 900 }, { "epoch": 3.77, "eval_gen_len": 131.32558139534885, "eval_gold_content__rouge1": 58.3411, "eval_gold_content__rouge1_precision": 58.2642, "eval_gold_content__rouge1_recall": 58.9275, "eval_gold_content__rouge2": 33.3716, "eval_gold_content__rouge2_precision": 33.4297, "eval_gold_content__rouge2_recall": 33.6498, "eval_gold_content__rougeL": 42.2759, "eval_gold_content__rougeL_precision": 42.2765, "eval_gold_content__rougeL_recall": 42.687, "eval_gold_content__rougeLsum": 42.2536, "eval_gold_content__rougeLsum_precision": 42.2241, "eval_gold_content__rougeLsum_recall": 42.7188, "eval_gold_rouge1": 61.6417, "eval_gold_rouge1_precision": 61.6121, "eval_gold_rouge1_recall": 62.1201, "eval_gold_rouge2": 35.6839, "eval_gold_rouge2_precision": 35.7649, "eval_gold_rouge2_recall": 35.8464, "eval_gold_rougeL": 42.5074, "eval_gold_rougeL_precision": 42.5107, "eval_gold_rougeL_recall": 42.81, "eval_gold_rougeLsum": 58.0849, "eval_gold_rougeLsum_precision": 58.0593, "eval_gold_rougeLsum_recall": 58.511, "eval_loss": 2.332422971725464, "eval_runtime": 323.7426, "eval_samples_per_second": 0.398, "eval_steps_per_second": 0.398, "step": 900 }, { "epoch": 3.97, "learning_rate": 3.0125523012552304e-05, "loss": 0.8544, "step": 950 }, { "epoch": 4.18, "learning_rate": 2.9079497907949792e-05, "loss": 0.6838, "step": 1000 }, { "epoch": 4.18, "eval_gen_len": 131.14728682170542, "eval_gold_content__rouge1": 57.4335, "eval_gold_content__rouge1_precision": 57.3305, "eval_gold_content__rouge1_recall": 57.8738, "eval_gold_content__rouge2": 31.9398, "eval_gold_content__rouge2_precision": 31.9439, "eval_gold_content__rouge2_recall": 32.1494, "eval_gold_content__rougeL": 41.5105, "eval_gold_content__rougeL_precision": 41.3791, "eval_gold_content__rougeL_recall": 41.8825, "eval_gold_content__rougeLsum": 41.5048, "eval_gold_content__rougeLsum_precision": 41.3923, "eval_gold_content__rougeLsum_recall": 41.8949, "eval_gold_rouge1": 61.3986, "eval_gold_rouge1_precision": 61.2071, "eval_gold_rouge1_recall": 61.9882, "eval_gold_rouge2": 34.4586, "eval_gold_rouge2_precision": 34.403, "eval_gold_rouge2_recall": 34.7027, "eval_gold_rougeL": 41.8032, "eval_gold_rougeL_precision": 41.6313, "eval_gold_rougeL_recall": 42.2476, "eval_gold_rougeLsum": 57.302, "eval_gold_rougeLsum_precision": 57.1356, "eval_gold_rougeLsum_recall": 57.8116, "eval_loss": 2.4267024993896484, "eval_runtime": 321.0984, "eval_samples_per_second": 0.402, "eval_steps_per_second": 0.402, "step": 1000 }, { "epoch": 4.39, "learning_rate": 2.8033472803347283e-05, "loss": 0.6376, "step": 1050 }, { "epoch": 4.6, "learning_rate": 2.6987447698744773e-05, "loss": 0.6719, "step": 1100 }, { "epoch": 4.6, "eval_gen_len": 132.86046511627907, "eval_gold_content__rouge1": 58.6491, "eval_gold_content__rouge1_precision": 58.2647, "eval_gold_content__rouge1_recall": 59.4943, "eval_gold_content__rouge2": 33.6426, "eval_gold_content__rouge2_precision": 33.4593, "eval_gold_content__rouge2_recall": 34.0799, "eval_gold_content__rougeL": 43.163, "eval_gold_content__rougeL_precision": 42.8565, "eval_gold_content__rougeL_recall": 43.7869, "eval_gold_content__rougeLsum": 43.1082, "eval_gold_content__rougeLsum_precision": 42.7496, "eval_gold_content__rougeLsum_recall": 43.7595, "eval_gold_rouge1": 62.3384, "eval_gold_rouge1_precision": 61.9139, "eval_gold_rouge1_recall": 63.1585, "eval_gold_rouge2": 35.8046, "eval_gold_rouge2_precision": 35.5978, "eval_gold_rouge2_recall": 36.2046, "eval_gold_rougeL": 43.2641, "eval_gold_rougeL_precision": 42.9357, "eval_gold_rougeL_recall": 43.8159, "eval_gold_rougeLsum": 58.3628, "eval_gold_rougeLsum_precision": 57.9516, "eval_gold_rougeLsum_recall": 59.1191, "eval_loss": 2.427868366241455, "eval_runtime": 317.5232, "eval_samples_per_second": 0.406, "eval_steps_per_second": 0.406, "step": 1100 }, { "epoch": 4.81, "learning_rate": 2.594142259414226e-05, "loss": 0.6775, "step": 1150 }, { "epoch": 5.02, "learning_rate": 2.489539748953975e-05, "loss": 0.6835, "step": 1200 }, { "epoch": 5.02, "eval_gen_len": 129.37209302325581, "eval_gold_content__rouge1": 57.8101, "eval_gold_content__rouge1_precision": 58.1747, "eval_gold_content__rouge1_recall": 57.7813, "eval_gold_content__rouge2": 31.8646, "eval_gold_content__rouge2_precision": 32.0479, "eval_gold_content__rouge2_recall": 31.8628, "eval_gold_content__rougeL": 41.4763, "eval_gold_content__rougeL_precision": 41.7604, "eval_gold_content__rougeL_recall": 41.4871, "eval_gold_content__rougeLsum": 41.4962, "eval_gold_content__rougeLsum_precision": 41.7238, "eval_gold_content__rougeLsum_recall": 41.498, "eval_gold_rouge1": 61.4546, "eval_gold_rouge1_precision": 61.8021, "eval_gold_rouge1_recall": 61.3898, "eval_gold_rouge2": 34.5481, "eval_gold_rouge2_precision": 34.8132, "eval_gold_rouge2_recall": 34.5658, "eval_gold_rougeL": 41.9335, "eval_gold_rougeL_precision": 42.1417, "eval_gold_rougeL_recall": 41.9428, "eval_gold_rougeLsum": 57.518, "eval_gold_rougeLsum_precision": 57.9298, "eval_gold_rougeLsum_recall": 57.4184, "eval_loss": 2.4904870986938477, "eval_runtime": 306.5024, "eval_samples_per_second": 0.421, "eval_steps_per_second": 0.421, "step": 1200 }, { "epoch": 5.23, "learning_rate": 2.3849372384937242e-05, "loss": 0.5561, "step": 1250 }, { "epoch": 5.44, "learning_rate": 2.280334728033473e-05, "loss": 0.5495, "step": 1300 }, { "epoch": 5.44, "eval_gen_len": 132.88372093023256, "eval_gold_content__rouge1": 59.2827, "eval_gold_content__rouge1_precision": 58.931, "eval_gold_content__rouge1_recall": 60.0205, "eval_gold_content__rouge2": 34.1327, "eval_gold_content__rouge2_precision": 34.0164, "eval_gold_content__rouge2_recall": 34.4858, "eval_gold_content__rougeL": 43.5606, "eval_gold_content__rougeL_precision": 43.3704, "eval_gold_content__rougeL_recall": 44.1162, "eval_gold_content__rougeLsum": 43.5148, "eval_gold_content__rougeLsum_precision": 43.2937, "eval_gold_content__rougeLsum_recall": 44.1221, "eval_gold_rouge1": 62.7436, "eval_gold_rouge1_precision": 62.4107, "eval_gold_rouge1_recall": 63.4468, "eval_gold_rouge2": 36.4478, "eval_gold_rouge2_precision": 36.3739, "eval_gold_rouge2_recall": 36.8162, "eval_gold_rougeL": 43.7493, "eval_gold_rougeL_precision": 43.5387, "eval_gold_rougeL_recall": 44.2374, "eval_gold_rougeLsum": 58.9706, "eval_gold_rougeLsum_precision": 58.6953, "eval_gold_rougeLsum_recall": 59.5665, "eval_loss": 2.555025339126587, "eval_runtime": 323.5705, "eval_samples_per_second": 0.399, "eval_steps_per_second": 0.399, "step": 1300 } ], "max_steps": 2390, "num_train_epochs": 10, "total_flos": 1.0265751601127424e+16, "trial_name": null, "trial_params": null }