{ "args": { "base": "jhu-clsp/mmBERT-base", "data": "see code/", "out": "model", "max_len": 512, "tokens_per_batch": 65536, "lr": 3e-05, "epochs": 1.0, "warmup": 0.05, "lower": 0.7, "max_train": 0, "max_val": 6000, "evals": 8, "seed": 0 }, "log": [ { "step": 538, "cased": { "punct_f1": 0.7856, ".": 0.8964, ",": 0.7074, "?": 0.6663, "-": 0.2025, ":": 0.462 }, "lower": { "punct_f1": 0.7217, ".": 0.8062, ",": 0.6682, "?": 0.5876, "-": 0.18, ":": 0.421 }, "score": 0.7536499999999999 }, { "step": 1076, "cased": { "punct_f1": 0.7975, ".": 0.9028, ",": 0.7225, "?": 0.6755, "-": 0.2805, ":": 0.5127 }, "lower": { "punct_f1": 0.7383, ".": 0.8199, ",": 0.683, "?": 0.605, "-": 0.2533, ":": 0.4819 }, "score": 0.7679 }, { "step": 1614, "cased": { "punct_f1": 0.7998, ".": 0.9097, ",": 0.7176, "?": 0.6738, "-": 0.2772, ":": 0.5288 }, "lower": { "punct_f1": 0.7414, ".": 0.8292, ",": 0.6807, "?": 0.593, "-": 0.2549, ":": 0.4996 }, "score": 0.7706 }, { "step": 2152, "cased": { "punct_f1": 0.8104, ".": 0.9118, ",": 0.7401, "?": 0.6981, "-": 0.309, ":": 0.548 }, "lower": { "punct_f1": 0.756, ".": 0.8359, ",": 0.7046, "?": 0.6259, "-": 0.2836, ":": 0.5138 }, "score": 0.7832 }, { "step": 2690, "cased": { "punct_f1": 0.8118, ".": 0.9133, ",": 0.7404, "?": 0.6912, "-": 0.3063, ":": 0.5581 }, "lower": { "punct_f1": 0.7586, ".": 0.8391, ",": 0.7056, "?": 0.6195, "-": 0.2996, ":": 0.5239 }, "score": 0.7852 }, { "step": 3228, "cased": { "punct_f1": 0.8151, ".": 0.9153, ",": 0.7452, "?": 0.6991, "-": 0.3425, ":": 0.5644 }, "lower": { "punct_f1": 0.7624, ".": 0.8419, ",": 0.7105, "?": 0.6263, "-": 0.3167, ":": 0.5375 }, "score": 0.7887500000000001 }, { "step": 3766, "cased": { "punct_f1": 0.8162, ".": 0.9161, ",": 0.747, "?": 0.7008, "-": 0.3396, ":": 0.5619 }, "lower": { "punct_f1": 0.7631, ".": 0.8429, ",": 0.7115, "?": 0.6289, "-": 0.3108, ":": 0.528 }, "score": 0.78965 }, { "step": 4304, "cased": { "punct_f1": 0.8164, ".": 0.9159, ",": 0.7478, "?": 0.6989, "-": 0.3422, ":": 0.5612 }, "lower": { "punct_f1": 0.7638, ".": 0.843, ",": 0.7131, "?": 0.6305, "-": 0.3146, ":": 0.5264 }, "score": 0.7901 }, { "step": 4307, "cased": { "punct_f1": 0.8165, ".": 0.916, ",": 0.7479, "?": 0.6993, "-": 0.342, ":": 0.5605 }, "lower": { "punct_f1": 0.7638, ".": 0.843, ",": 0.7132, "?": 0.6288, "-": 0.3143, ":": 0.5262 }, "score": 0.79015 } ], "best": 0.79015 }