{ "model": { "name": "mamba", "d_model": 256, "n_layers": 7, "d_state": 16, "d_conv": 4, "expand": 2, "dt_rank": null, "dt_min": 0.001, "dt_max": 0.1, "dt_floor": 0.0001, "scan_chunk": 16, "scan_tokens": 4096, "grad_ckpt": false }, "train": { "steps": 10000, "batch_size": 16, "seq_len": 1024, "lr": 0.001, "min_lr_ratio": 0.1, "warmup": 1000, "weight_decay": 0.1, "grad_clip": 1.0, "log_every": 50, "eval_every": 1000, "eval_batch_tokens": 16384, "seed": 0, "align_crops": true }, "params": 3095068, "data": { "pairs": 569670, "dropped": 0, "null_symbols": "!'+,-./:;<=>?[]_{}" }, "best": { "char_accuracy": 0.18181128764556412, "word_accuracy": 0.010008523897961603, "precision": 0.7922315994950653, "recall": 0.05078616476954914, "f1": 0.03797654601370306, "levenshtein": 739.1352992194276, "normalized_levenshtein": 0.8180898377015771, "samples": 1153, "loss": 2.487926709028701, "step": 10000 } }