| { | |
| "best_metric": 46.4318, | |
| "best_model_checkpoint": "outputs/modernisa-v2-byt5-base-lr0.0001/checkpoint-34000", | |
| "epoch": 5.0, | |
| "global_step": 57430, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 9.912937489117186e-05, | |
| "loss": 0.3717, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 9.825874978234372e-05, | |
| "loss": 0.2696, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "eval_bleu": 27.8571, | |
| "eval_cer": 34.4149, | |
| "eval_gen_len": 18.5, | |
| "eval_loss": 0.3027326464653015, | |
| "eval_runtime": 51.4554, | |
| "eval_samples_per_second": 45.593, | |
| "eval_steps_per_second": 2.857, | |
| "eval_wer": 49.5134, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 9.73881246735156e-05, | |
| "loss": 0.2564, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 9.651749956468745e-05, | |
| "loss": 0.2518, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "eval_bleu": 29.2213, | |
| "eval_cer": 34.6336, | |
| "eval_gen_len": 18.5371, | |
| "eval_loss": 0.28565752506256104, | |
| "eval_runtime": 53.6563, | |
| "eval_samples_per_second": 43.723, | |
| "eval_steps_per_second": 2.74, | |
| "eval_wer": 49.1981, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 9.564687445585931e-05, | |
| "loss": 0.2387, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 9.477624934703117e-05, | |
| "loss": 0.2343, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "eval_bleu": 29.5067, | |
| "eval_cer": 34.9795, | |
| "eval_gen_len": 18.5537, | |
| "eval_loss": 0.2730022966861725, | |
| "eval_runtime": 51.0673, | |
| "eval_samples_per_second": 45.939, | |
| "eval_steps_per_second": 2.879, | |
| "eval_wer": 49.117, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 9.390562423820304e-05, | |
| "loss": 0.2331, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 9.30349991293749e-05, | |
| "loss": 0.2292, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "eval_bleu": 29.884, | |
| "eval_cer": 34.8015, | |
| "eval_gen_len": 18.5516, | |
| "eval_loss": 0.26898515224456787, | |
| "eval_runtime": 52.4703, | |
| "eval_samples_per_second": 44.711, | |
| "eval_steps_per_second": 2.802, | |
| "eval_wer": 48.7025, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 9.216437402054676e-05, | |
| "loss": 0.2243, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 9.129374891171862e-05, | |
| "loss": 0.2243, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "eval_bleu": 29.9577, | |
| "eval_cer": 34.7218, | |
| "eval_gen_len": 18.5477, | |
| "eval_loss": 0.26465946435928345, | |
| "eval_runtime": 51.6529, | |
| "eval_samples_per_second": 45.419, | |
| "eval_steps_per_second": 2.846, | |
| "eval_wer": 48.8466, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 9.042312380289048e-05, | |
| "loss": 0.2185, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 8.955249869406234e-05, | |
| "loss": 0.2112, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "eval_bleu": 30.3115, | |
| "eval_cer": 34.4895, | |
| "eval_gen_len": 18.5477, | |
| "eval_loss": 0.2636098265647888, | |
| "eval_runtime": 53.1222, | |
| "eval_samples_per_second": 44.162, | |
| "eval_steps_per_second": 2.767, | |
| "eval_wer": 48.3871, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 8.86818735852342e-05, | |
| "loss": 0.2165, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 8.781124847640607e-05, | |
| "loss": 0.2118, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "eval_bleu": 30.6364, | |
| "eval_cer": 34.7455, | |
| "eval_gen_len": 18.5413, | |
| "eval_loss": 0.25554388761520386, | |
| "eval_runtime": 50.9916, | |
| "eval_samples_per_second": 46.008, | |
| "eval_steps_per_second": 2.883, | |
| "eval_wer": 48.3961, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 8.694062336757793e-05, | |
| "loss": 0.2105, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 8.606999825874978e-05, | |
| "loss": 0.205, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "eval_bleu": 31.0881, | |
| "eval_cer": 34.0759, | |
| "eval_gen_len": 18.5269, | |
| "eval_loss": 0.2507636547088623, | |
| "eval_runtime": 53.5946, | |
| "eval_samples_per_second": 43.773, | |
| "eval_steps_per_second": 2.743, | |
| "eval_wer": 47.468, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 8.519937314992164e-05, | |
| "loss": 0.2035, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 8.432874804109351e-05, | |
| "loss": 0.2049, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "eval_bleu": 31.1481, | |
| "eval_cer": 34.4133, | |
| "eval_gen_len": 18.5503, | |
| "eval_loss": 0.24714592099189758, | |
| "eval_runtime": 60.0036, | |
| "eval_samples_per_second": 39.098, | |
| "eval_steps_per_second": 2.45, | |
| "eval_wer": 47.5942, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 8.345812293226537e-05, | |
| "loss": 0.2074, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 8.258749782343723e-05, | |
| "loss": 0.2005, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "eval_bleu": 30.9375, | |
| "eval_cer": 34.281, | |
| "eval_gen_len": 18.5405, | |
| "eval_loss": 0.24682185053825378, | |
| "eval_runtime": 51.6353, | |
| "eval_samples_per_second": 45.434, | |
| "eval_steps_per_second": 2.847, | |
| "eval_wer": 47.6392, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 8.171687271460909e-05, | |
| "loss": 0.2049, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 8.084624760578095e-05, | |
| "loss": 0.1999, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "eval_bleu": 30.9692, | |
| "eval_cer": 34.4183, | |
| "eval_gen_len": 18.5405, | |
| "eval_loss": 0.24305607378482819, | |
| "eval_runtime": 53.15, | |
| "eval_samples_per_second": 44.139, | |
| "eval_steps_per_second": 2.766, | |
| "eval_wer": 47.7023, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 7.997562249695282e-05, | |
| "loss": 0.2037, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 7.910499738812468e-05, | |
| "loss": 0.161, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "eval_bleu": 31.2337, | |
| "eval_cer": 34.1878, | |
| "eval_gen_len": 18.5298, | |
| "eval_loss": 0.24913541972637177, | |
| "eval_runtime": 51.1684, | |
| "eval_samples_per_second": 45.849, | |
| "eval_steps_per_second": 2.873, | |
| "eval_wer": 47.3238, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 7.823437227929654e-05, | |
| "loss": 0.1568, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 7.73637471704684e-05, | |
| "loss": 0.1601, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "eval_bleu": 31.4422, | |
| "eval_cer": 34.1657, | |
| "eval_gen_len": 18.5371, | |
| "eval_loss": 0.24957244098186493, | |
| "eval_runtime": 52.0238, | |
| "eval_samples_per_second": 45.095, | |
| "eval_steps_per_second": 2.826, | |
| "eval_wer": 47.3689, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 7.649312206164027e-05, | |
| "loss": 0.1523, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 7.562249695281213e-05, | |
| "loss": 0.1606, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "eval_bleu": 31.4582, | |
| "eval_cer": 34.2386, | |
| "eval_gen_len": 18.5405, | |
| "eval_loss": 0.24588151276111603, | |
| "eval_runtime": 51.9336, | |
| "eval_samples_per_second": 45.173, | |
| "eval_steps_per_second": 2.831, | |
| "eval_wer": 47.3329, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 7.475187184398399e-05, | |
| "loss": 0.158, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 7.388124673515585e-05, | |
| "loss": 0.1594, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "eval_bleu": 31.386, | |
| "eval_cer": 34.2912, | |
| "eval_gen_len": 18.5375, | |
| "eval_loss": 0.24664774537086487, | |
| "eval_runtime": 51.7169, | |
| "eval_samples_per_second": 45.362, | |
| "eval_steps_per_second": 2.842, | |
| "eval_wer": 47.1166, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 7.30106216263277e-05, | |
| "loss": 0.1609, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 7.213999651749956e-05, | |
| "loss": 0.1617, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "eval_bleu": 31.6546, | |
| "eval_cer": 34.0149, | |
| "eval_gen_len": 18.5294, | |
| "eval_loss": 0.2411554753780365, | |
| "eval_runtime": 52.5275, | |
| "eval_samples_per_second": 44.662, | |
| "eval_steps_per_second": 2.799, | |
| "eval_wer": 46.8373, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 7.126937140867142e-05, | |
| "loss": 0.1594, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 7.03987462998433e-05, | |
| "loss": 0.1582, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "eval_bleu": 31.2924, | |
| "eval_cer": 34.2573, | |
| "eval_gen_len": 18.5503, | |
| "eval_loss": 0.24606570601463318, | |
| "eval_runtime": 50.8697, | |
| "eval_samples_per_second": 46.118, | |
| "eval_steps_per_second": 2.89, | |
| "eval_wer": 47.4139, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 6.952812119101515e-05, | |
| "loss": 0.1599, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 6.865749608218701e-05, | |
| "loss": 0.1572, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "eval_bleu": 31.1484, | |
| "eval_cer": 34.3675, | |
| "eval_gen_len": 18.5499, | |
| "eval_loss": 0.24250736832618713, | |
| "eval_runtime": 52.5057, | |
| "eval_samples_per_second": 44.681, | |
| "eval_steps_per_second": 2.8, | |
| "eval_wer": 47.45, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 6.778687097335887e-05, | |
| "loss": 0.1574, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "learning_rate": 6.691624586453074e-05, | |
| "loss": 0.1565, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "eval_bleu": 31.6967, | |
| "eval_cer": 34.1047, | |
| "eval_gen_len": 18.5388, | |
| "eval_loss": 0.2424342930316925, | |
| "eval_runtime": 51.0272, | |
| "eval_samples_per_second": 45.975, | |
| "eval_steps_per_second": 2.881, | |
| "eval_wer": 46.9724, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 6.60456207557026e-05, | |
| "loss": 0.1582, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 6.517499564687446e-05, | |
| "loss": 0.1585, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "eval_bleu": 31.9026, | |
| "eval_cer": 34.281, | |
| "eval_gen_len": 18.558, | |
| "eval_loss": 0.2381763756275177, | |
| "eval_runtime": 52.4669, | |
| "eval_samples_per_second": 44.714, | |
| "eval_steps_per_second": 2.802, | |
| "eval_wer": 47.0175, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "learning_rate": 6.430437053804632e-05, | |
| "loss": 0.1559, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 6.343374542921819e-05, | |
| "loss": 0.1522, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "eval_bleu": 32.1619, | |
| "eval_cer": 33.9369, | |
| "eval_gen_len": 18.5311, | |
| "eval_loss": 0.23654896020889282, | |
| "eval_runtime": 52.4567, | |
| "eval_samples_per_second": 44.723, | |
| "eval_steps_per_second": 2.802, | |
| "eval_wer": 46.5219, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 6.256312032039005e-05, | |
| "loss": 0.1578, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 6.169249521156191e-05, | |
| "loss": 0.156, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "eval_bleu": 31.7762, | |
| "eval_cer": 33.9572, | |
| "eval_gen_len": 18.5401, | |
| "eval_loss": 0.2381468415260315, | |
| "eval_runtime": 56.1509, | |
| "eval_samples_per_second": 41.78, | |
| "eval_steps_per_second": 2.618, | |
| "eval_wer": 46.7922, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 6.0821870102733766e-05, | |
| "loss": 0.1567, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 5.995124499390563e-05, | |
| "loss": 0.1538, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_bleu": 31.8785, | |
| "eval_cer": 34.2319, | |
| "eval_gen_len": 18.5516, | |
| "eval_loss": 0.24016974866390228, | |
| "eval_runtime": 53.089, | |
| "eval_samples_per_second": 44.19, | |
| "eval_steps_per_second": 2.769, | |
| "eval_wer": 46.8012, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "learning_rate": 5.908061988507749e-05, | |
| "loss": 0.1075, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 5.8209994776249355e-05, | |
| "loss": 0.1083, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "eval_bleu": 31.9905, | |
| "eval_cer": 34.0098, | |
| "eval_gen_len": 18.5384, | |
| "eval_loss": 0.2653577923774719, | |
| "eval_runtime": 52.9572, | |
| "eval_samples_per_second": 44.3, | |
| "eval_steps_per_second": 2.776, | |
| "eval_wer": 46.603, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 5.7339369667421214e-05, | |
| "loss": 0.108, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "learning_rate": 5.6468744558593066e-05, | |
| "loss": 0.1086, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "eval_bleu": 31.6257, | |
| "eval_cer": 34.2607, | |
| "eval_gen_len": 18.5409, | |
| "eval_loss": 0.26182088255882263, | |
| "eval_runtime": 51.8801, | |
| "eval_samples_per_second": 45.22, | |
| "eval_steps_per_second": 2.833, | |
| "eval_wer": 46.9995, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "learning_rate": 5.559811944976493e-05, | |
| "loss": 0.1085, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "learning_rate": 5.472749434093679e-05, | |
| "loss": 0.1092, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "eval_bleu": 31.4886, | |
| "eval_cer": 34.337, | |
| "eval_gen_len": 18.5422, | |
| "eval_loss": 0.2658332288265228, | |
| "eval_runtime": 51.9175, | |
| "eval_samples_per_second": 45.187, | |
| "eval_steps_per_second": 2.831, | |
| "eval_wer": 47.1436, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "learning_rate": 5.3856869232108655e-05, | |
| "loss": 0.1106, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "learning_rate": 5.2986244123280514e-05, | |
| "loss": 0.1086, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "eval_bleu": 31.8448, | |
| "eval_cer": 34.1217, | |
| "eval_gen_len": 18.5375, | |
| "eval_loss": 0.26663142442703247, | |
| "eval_runtime": 51.6292, | |
| "eval_samples_per_second": 45.439, | |
| "eval_steps_per_second": 2.847, | |
| "eval_wer": 46.6751, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "learning_rate": 5.211561901445238e-05, | |
| "loss": 0.1087, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 5.124499390562424e-05, | |
| "loss": 0.1098, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "eval_bleu": 31.709, | |
| "eval_cer": 34.1946, | |
| "eval_gen_len": 18.5452, | |
| "eval_loss": 0.2659294009208679, | |
| "eval_runtime": 58.8373, | |
| "eval_samples_per_second": 39.873, | |
| "eval_steps_per_second": 2.498, | |
| "eval_wer": 46.8913, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "learning_rate": 5.03743687967961e-05, | |
| "loss": 0.1106, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "learning_rate": 4.950374368796796e-05, | |
| "loss": 0.1117, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "eval_bleu": 31.8114, | |
| "eval_cer": 34.1708, | |
| "eval_gen_len": 18.5431, | |
| "eval_loss": 0.2648890018463135, | |
| "eval_runtime": 57.7914, | |
| "eval_samples_per_second": 40.594, | |
| "eval_steps_per_second": 2.544, | |
| "eval_wer": 46.8913, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "learning_rate": 4.863311857913983e-05, | |
| "loss": 0.1087, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 4.7762493470311686e-05, | |
| "loss": 0.1094, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "eval_bleu": 31.6955, | |
| "eval_cer": 34.1606, | |
| "eval_gen_len": 18.5375, | |
| "eval_loss": 0.26563677191734314, | |
| "eval_runtime": 53.0374, | |
| "eval_samples_per_second": 44.233, | |
| "eval_steps_per_second": 2.772, | |
| "eval_wer": 46.8643, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "learning_rate": 4.689186836148355e-05, | |
| "loss": 0.1079, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "learning_rate": 4.602124325265541e-05, | |
| "loss": 0.1077, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "eval_bleu": 31.5495, | |
| "eval_cer": 34.0064, | |
| "eval_gen_len": 18.5448, | |
| "eval_loss": 0.2636907398700714, | |
| "eval_runtime": 53.9667, | |
| "eval_samples_per_second": 43.471, | |
| "eval_steps_per_second": 2.724, | |
| "eval_wer": 46.8823, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "learning_rate": 4.5150618143827276e-05, | |
| "loss": 0.1103, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "learning_rate": 4.427999303499913e-05, | |
| "loss": 0.1088, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "eval_bleu": 32.0837, | |
| "eval_cer": 33.9504, | |
| "eval_gen_len": 18.5413, | |
| "eval_loss": 0.2668997645378113, | |
| "eval_runtime": 50.6471, | |
| "eval_samples_per_second": 46.321, | |
| "eval_steps_per_second": 2.902, | |
| "eval_wer": 46.612, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "learning_rate": 4.340936792617099e-05, | |
| "loss": 0.1087, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "learning_rate": 4.253874281734285e-05, | |
| "loss": 0.1087, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "eval_bleu": 31.5549, | |
| "eval_cer": 34.2149, | |
| "eval_gen_len": 18.5286, | |
| "eval_loss": 0.264612078666687, | |
| "eval_runtime": 52.8215, | |
| "eval_samples_per_second": 44.414, | |
| "eval_steps_per_second": 2.783, | |
| "eval_wer": 47.0806, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "learning_rate": 4.166811770851472e-05, | |
| "loss": 0.1091, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "learning_rate": 4.0797492599686575e-05, | |
| "loss": 0.1077, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "eval_bleu": 32.1129, | |
| "eval_cer": 33.9403, | |
| "eval_gen_len": 18.5452, | |
| "eval_loss": 0.26299673318862915, | |
| "eval_runtime": 53.065, | |
| "eval_samples_per_second": 44.21, | |
| "eval_steps_per_second": 2.77, | |
| "eval_wer": 46.4318, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 3.992686749085844e-05, | |
| "loss": 0.1037, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "learning_rate": 3.90562423820303e-05, | |
| "loss": 0.0652, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "eval_bleu": 31.3861, | |
| "eval_cer": 34.1149, | |
| "eval_gen_len": 18.5396, | |
| "eval_loss": 0.33602526783943176, | |
| "eval_runtime": 53.5525, | |
| "eval_samples_per_second": 43.807, | |
| "eval_steps_per_second": 2.745, | |
| "eval_wer": 47.1977, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "learning_rate": 3.8185617273202165e-05, | |
| "loss": 0.0648, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "learning_rate": 3.7314992164374024e-05, | |
| "loss": 0.0662, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "eval_bleu": 31.2372, | |
| "eval_cer": 34.203, | |
| "eval_gen_len": 18.552, | |
| "eval_loss": 0.3401012718677521, | |
| "eval_runtime": 51.9918, | |
| "eval_samples_per_second": 45.122, | |
| "eval_steps_per_second": 2.827, | |
| "eval_wer": 47.3869, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "learning_rate": 3.644436705554589e-05, | |
| "loss": 0.0635, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "learning_rate": 3.557374194671774e-05, | |
| "loss": 0.0666, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "eval_bleu": 31.3462, | |
| "eval_cer": 34.1759, | |
| "eval_gen_len": 18.5469, | |
| "eval_loss": 0.33890092372894287, | |
| "eval_runtime": 51.8417, | |
| "eval_samples_per_second": 45.253, | |
| "eval_steps_per_second": 2.836, | |
| "eval_wer": 47.2968, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 3.26, | |
| "learning_rate": 3.4703116837889606e-05, | |
| "loss": 0.0646, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "learning_rate": 3.3832491729061465e-05, | |
| "loss": 0.0648, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "eval_bleu": 30.835, | |
| "eval_cer": 34.381, | |
| "eval_gen_len": 18.552, | |
| "eval_loss": 0.3339092433452606, | |
| "eval_runtime": 51.3621, | |
| "eval_samples_per_second": 45.676, | |
| "eval_steps_per_second": 2.862, | |
| "eval_wer": 47.6753, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "learning_rate": 3.296186662023333e-05, | |
| "loss": 0.0648, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "learning_rate": 3.209124151140519e-05, | |
| "loss": 0.0654, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "eval_bleu": 31.0958, | |
| "eval_cer": 34.4692, | |
| "eval_gen_len": 18.5524, | |
| "eval_loss": 0.33954936265945435, | |
| "eval_runtime": 51.3892, | |
| "eval_samples_per_second": 45.652, | |
| "eval_steps_per_second": 2.861, | |
| "eval_wer": 47.7203, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "learning_rate": 3.1220616402577054e-05, | |
| "loss": 0.0645, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "learning_rate": 3.0349991293748913e-05, | |
| "loss": 0.0663, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "eval_bleu": 31.126, | |
| "eval_cer": 34.4539, | |
| "eval_gen_len": 18.5499, | |
| "eval_loss": 0.3317714333534241, | |
| "eval_runtime": 53.3556, | |
| "eval_samples_per_second": 43.969, | |
| "eval_steps_per_second": 2.755, | |
| "eval_wer": 47.5942, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 3.53, | |
| "learning_rate": 2.9479366184920775e-05, | |
| "loss": 0.0657, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "learning_rate": 2.8608741076092637e-05, | |
| "loss": 0.0648, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "eval_bleu": 31.0295, | |
| "eval_cer": 34.3539, | |
| "eval_gen_len": 18.5477, | |
| "eval_loss": 0.33970215916633606, | |
| "eval_runtime": 52.8717, | |
| "eval_samples_per_second": 44.372, | |
| "eval_steps_per_second": 2.78, | |
| "eval_wer": 47.5852, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "learning_rate": 2.77381159672645e-05, | |
| "loss": 0.0647, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "learning_rate": 2.6867490858436354e-05, | |
| "loss": 0.0635, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "eval_bleu": 31.1287, | |
| "eval_cer": 34.4285, | |
| "eval_gen_len": 18.5494, | |
| "eval_loss": 0.3414219617843628, | |
| "eval_runtime": 52.9358, | |
| "eval_samples_per_second": 44.318, | |
| "eval_steps_per_second": 2.777, | |
| "eval_wer": 47.5491, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "learning_rate": 2.5996865749608216e-05, | |
| "loss": 0.0657, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "learning_rate": 2.512624064078008e-05, | |
| "loss": 0.0656, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "eval_bleu": 30.9225, | |
| "eval_cer": 34.4285, | |
| "eval_gen_len": 18.5563, | |
| "eval_loss": 0.3393559455871582, | |
| "eval_runtime": 51.5264, | |
| "eval_samples_per_second": 45.53, | |
| "eval_steps_per_second": 2.853, | |
| "eval_wer": 47.6392, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "learning_rate": 2.425561553195194e-05, | |
| "loss": 0.0659, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "learning_rate": 2.3384990423123803e-05, | |
| "loss": 0.0625, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "eval_bleu": 31.2435, | |
| "eval_cer": 34.1674, | |
| "eval_gen_len": 18.5439, | |
| "eval_loss": 0.341974139213562, | |
| "eval_runtime": 51.6192, | |
| "eval_samples_per_second": 45.448, | |
| "eval_steps_per_second": 2.848, | |
| "eval_wer": 47.2968, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "learning_rate": 2.2514365314295665e-05, | |
| "loss": 0.0642, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "learning_rate": 2.1643740205467527e-05, | |
| "loss": 0.0636, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "eval_bleu": 31.0688, | |
| "eval_cer": 34.3743, | |
| "eval_gen_len": 18.5439, | |
| "eval_loss": 0.3447582423686981, | |
| "eval_runtime": 51.0408, | |
| "eval_samples_per_second": 45.963, | |
| "eval_steps_per_second": 2.88, | |
| "eval_wer": 47.6843, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "learning_rate": 2.0773115096639385e-05, | |
| "loss": 0.0644, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 1.9902489987811247e-05, | |
| "loss": 0.0586, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_bleu": 31.2353, | |
| "eval_cer": 34.2963, | |
| "eval_gen_len": 18.549, | |
| "eval_loss": 0.36747270822525024, | |
| "eval_runtime": 54.2264, | |
| "eval_samples_per_second": 43.263, | |
| "eval_steps_per_second": 2.711, | |
| "eval_wer": 47.441, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "learning_rate": 1.903186487898311e-05, | |
| "loss": 0.0326, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 4.09, | |
| "learning_rate": 1.816123977015497e-05, | |
| "loss": 0.0298, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 4.09, | |
| "eval_bleu": 30.698, | |
| "eval_cer": 34.4319, | |
| "eval_gen_len": 18.5512, | |
| "eval_loss": 0.45656564831733704, | |
| "eval_runtime": 53.4442, | |
| "eval_samples_per_second": 43.896, | |
| "eval_steps_per_second": 2.751, | |
| "eval_wer": 47.8555, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 4.14, | |
| "learning_rate": 1.7290614661326833e-05, | |
| "loss": 0.0308, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "learning_rate": 1.6419989552498692e-05, | |
| "loss": 0.0301, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "eval_bleu": 30.7773, | |
| "eval_cer": 34.3861, | |
| "eval_gen_len": 18.5507, | |
| "eval_loss": 0.4724096655845642, | |
| "eval_runtime": 52.6433, | |
| "eval_samples_per_second": 44.564, | |
| "eval_steps_per_second": 2.792, | |
| "eval_wer": 47.8374, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 4.22, | |
| "learning_rate": 1.5549364443670554e-05, | |
| "loss": 0.0307, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "learning_rate": 1.4678739334842418e-05, | |
| "loss": 0.0311, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "eval_bleu": 31.0878, | |
| "eval_cer": 34.3861, | |
| "eval_gen_len": 18.5503, | |
| "eval_loss": 0.4639967978000641, | |
| "eval_runtime": 51.8688, | |
| "eval_samples_per_second": 45.229, | |
| "eval_steps_per_second": 2.834, | |
| "eval_wer": 47.6212, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "learning_rate": 1.380811422601428e-05, | |
| "loss": 0.0313, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "learning_rate": 1.2937489117186142e-05, | |
| "loss": 0.03, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "eval_bleu": 30.8319, | |
| "eval_cer": 34.459, | |
| "eval_gen_len": 18.5529, | |
| "eval_loss": 0.46544739603996277, | |
| "eval_runtime": 56.982, | |
| "eval_samples_per_second": 41.171, | |
| "eval_steps_per_second": 2.58, | |
| "eval_wer": 47.8915, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "learning_rate": 1.2066864008358002e-05, | |
| "loss": 0.0306, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "learning_rate": 1.1196238899529864e-05, | |
| "loss": 0.0302, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "eval_bleu": 30.9236, | |
| "eval_cer": 34.4997, | |
| "eval_gen_len": 18.552, | |
| "eval_loss": 0.46650850772857666, | |
| "eval_runtime": 50.604, | |
| "eval_samples_per_second": 46.36, | |
| "eval_steps_per_second": 2.905, | |
| "eval_wer": 47.9276, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "learning_rate": 1.0325613790701725e-05, | |
| "loss": 0.0303, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 4.53, | |
| "learning_rate": 9.454988681873587e-06, | |
| "loss": 0.029, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 4.53, | |
| "eval_bleu": 30.8307, | |
| "eval_cer": 34.4997, | |
| "eval_gen_len": 18.5482, | |
| "eval_loss": 0.4756769835948944, | |
| "eval_runtime": 53.9866, | |
| "eval_samples_per_second": 43.455, | |
| "eval_steps_per_second": 2.723, | |
| "eval_wer": 47.9456, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "learning_rate": 8.584363573045447e-06, | |
| "loss": 0.031, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 4.61, | |
| "learning_rate": 7.713738464217309e-06, | |
| "loss": 0.0301, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 4.61, | |
| "eval_bleu": 30.7983, | |
| "eval_cer": 34.5218, | |
| "eval_gen_len": 18.5473, | |
| "eval_loss": 0.4672394096851349, | |
| "eval_runtime": 52.9301, | |
| "eval_samples_per_second": 44.323, | |
| "eval_steps_per_second": 2.777, | |
| "eval_wer": 47.9456, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 4.66, | |
| "learning_rate": 6.843113355389169e-06, | |
| "loss": 0.0296, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "learning_rate": 5.972488246561031e-06, | |
| "loss": 0.0294, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "eval_bleu": 30.8924, | |
| "eval_cer": 34.4353, | |
| "eval_gen_len": 18.5529, | |
| "eval_loss": 0.4714747667312622, | |
| "eval_runtime": 53.7737, | |
| "eval_samples_per_second": 43.627, | |
| "eval_steps_per_second": 2.734, | |
| "eval_wer": 47.7564, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 4.74, | |
| "learning_rate": 5.1018631377328925e-06, | |
| "loss": 0.0292, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "learning_rate": 4.2312380289047546e-06, | |
| "loss": 0.0288, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "eval_bleu": 30.7372, | |
| "eval_cer": 34.4675, | |
| "eval_gen_len": 18.5524, | |
| "eval_loss": 0.4751755893230438, | |
| "eval_runtime": 51.756, | |
| "eval_samples_per_second": 45.328, | |
| "eval_steps_per_second": 2.84, | |
| "eval_wer": 47.7924, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 4.83, | |
| "learning_rate": 3.3606129200766153e-06, | |
| "loss": 0.0293, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 4.88, | |
| "learning_rate": 2.4899878112484765e-06, | |
| "loss": 0.0289, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 4.88, | |
| "eval_bleu": 30.8554, | |
| "eval_cer": 34.459, | |
| "eval_gen_len": 18.5516, | |
| "eval_loss": 0.4744004011154175, | |
| "eval_runtime": 51.9465, | |
| "eval_samples_per_second": 45.162, | |
| "eval_steps_per_second": 2.83, | |
| "eval_wer": 47.8555, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 4.92, | |
| "learning_rate": 1.619362702420338e-06, | |
| "loss": 0.0285, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "learning_rate": 7.487375935921992e-07, | |
| "loss": 0.0288, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "eval_bleu": 30.8745, | |
| "eval_cer": 34.4895, | |
| "eval_gen_len": 18.5499, | |
| "eval_loss": 0.4744308888912201, | |
| "eval_runtime": 51.4462, | |
| "eval_samples_per_second": 45.601, | |
| "eval_steps_per_second": 2.857, | |
| "eval_wer": 47.8194, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 57430, | |
| "total_flos": 1.2162866037733786e+17, | |
| "train_loss": 0.11780740710349318, | |
| "train_runtime": 33856.2876, | |
| "train_samples_per_second": 27.14, | |
| "train_steps_per_second": 1.696 | |
| } | |
| ], | |
| "max_steps": 57430, | |
| "num_train_epochs": 5, | |
| "total_flos": 1.2162866037733786e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |