{ "args": { "bench": "benchmarks", "size": "base", "epochs": 12, "batch_size": 256, "lr": 0.0003, "weight_decay": 0.01, "warmup": 300, "augment": 0.25, "device": "auto", "out": "checkpoints/base", "seed": 0, "max_steps": 0, "dev_limit": 3000 }, "device": "mps", "git_commit": null, "parameters": 35054595, "best_dev_nucleus_exact": 0.6557333333333333, "wall_clock_seconds": 4925.1, "history": [ { "epoch": 0, "step": 350, "loss": 1.3525640233925411, "dev_nucleus_exact": 0.6162666666666666, "dev_boundary_f1": 0.9794019933554816, "seconds": 465.5 }, { "epoch": 1, "step": 700, "loss": 0.7800240731239318, "dev_nucleus_exact": 0.6346666666666667, "dev_boundary_f1": 0.999333777481679, "seconds": 908.4 }, { "epoch": 2, "step": 1050, "loss": 0.6399015586716789, "dev_nucleus_exact": 0.6461333333333333, "dev_boundary_f1": 0.9986666666666667, "seconds": 1333.5 }, { "epoch": 3, "step": 1400, "loss": 0.5727078880582537, "dev_nucleus_exact": 0.652, "dev_boundary_f1": 1.0, "seconds": 1727.4 }, { "epoch": 4, "step": 1750, "loss": 0.5232000116791043, "dev_nucleus_exact": 0.6490666666666667, "dev_boundary_f1": 1.0, "seconds": 2122.2 }, { "epoch": 5, "step": 2100, "loss": 0.482598266686712, "dev_nucleus_exact": 0.6514666666666666, "dev_boundary_f1": 1.0, "seconds": 2517.7 }, { "epoch": 6, "step": 2450, "loss": 0.4508444060598101, "dev_nucleus_exact": 0.6528, "dev_boundary_f1": 1.0, "seconds": 2921.6 }, { "epoch": 7, "step": 2800, "loss": 0.4187164889914649, "dev_nucleus_exact": 0.6525333333333333, "dev_boundary_f1": 1.0, "seconds": 3324.3 }, { "epoch": 8, "step": 3150, "loss": 0.39189900176865716, "dev_nucleus_exact": 0.6546666666666666, "dev_boundary_f1": 1.0, "seconds": 3718.0 }, { "epoch": 9, "step": 3500, "loss": 0.3702780089208058, "dev_nucleus_exact": 0.6538666666666667, "dev_boundary_f1": 1.0, "seconds": 4119.3 }, { "epoch": 10, "step": 3850, "loss": 0.3531410438673837, "dev_nucleus_exact": 0.6554666666666666, "dev_boundary_f1": 1.0, "seconds": 4519.3 }, { "epoch": 11, "step": 4200, "loss": 0.34747300050088337, "dev_nucleus_exact": 0.6557333333333333, "dev_boundary_f1": 1.0, "seconds": 4924.5 } ], "n_train_examples": 89430, "n_constructed": 17886 }