| { |
| "run": "spec_A_s1", |
| "model_config": { |
| "vocab_size": 256, |
| "d_enc": 384, |
| "d_main": 768, |
| "n_enc": 3, |
| "n_dec": 3, |
| "n_main": 10, |
| "n_heads_enc": 6, |
| "n_heads_main": 12, |
| "seq_len": 1024, |
| "dropout": 0.0, |
| "target_ratio": 6.0, |
| "max_chunk_mult": 3.0, |
| "dechunk": "ema", |
| "router_noise": 0.0, |
| "rope_base": 10000.0 |
| }, |
| "tied_weights": { |
| "head.weight": "emb.weight" |
| }, |
| "train_config": { |
| "condition": "A", |
| "seed": 1, |
| "out_dir": null, |
| "run_name": "spec_A_s1", |
| "corpus": "corpus", |
| "langs": null, |
| "seq_len": 1024, |
| "batch_size": 64, |
| "grad_accum": 1, |
| "steps": 11000, |
| "lr": 0.0003, |
| "min_lr_frac": 0.1, |
| "warmup": 500, |
| "weight_decay": 0.1, |
| "grad_clip": 1.0, |
| "betas": [ |
| 0.9, |
| 0.95 |
| ], |
| "router_lr_mult": 1.0, |
| "d_enc": 384, |
| "d_main": 768, |
| "n_enc": 3, |
| "n_dec": 3, |
| "n_main": 10, |
| "n_heads_enc": 6, |
| "n_heads_main": 12, |
| "dechunk": "ema", |
| "target_ratio": 6.0, |
| "alpha": 0.03, |
| "alpha_parity_cap": 0.06, |
| "global_anchor": 0.0, |
| "premium_clip": 4.0, |
| "premium_kind": "flores_per_sent", |
| "mixture": null, |
| "doc_filters": null, |
| "tau1": 0.4, |
| "tau2": 0.7, |
| "freeze_scope": "router", |
| "ckpt_every": 1000, |
| "eval_every": 1000, |
| "log_every": 100, |
| "probe_sents": 128, |
| "eval_batches": 8, |
| "keep_ckpt": true, |
| "amp": "bf16", |
| "compile": false, |
| "cpc_decay": 0.9, |
| "device": "cuda", |
| "log_spaced_probe": true, |
| "probe_bpb_batches": 4, |
| "dump_reps": true, |
| "dump_reps_thresholds": [ |
| 0.3, |
| 0.4, |
| 0.5, |
| 0.6, |
| 0.7 |
| ], |
| "router_noise": 0.0, |
| "genreg_target_entropy": null, |
| "genreg_eta": 0.05 |
| }, |
| "checkpoints": [ |
| { |
| "tag": "000000", |
| "step": 0 |
| }, |
| { |
| "tag": "000001", |
| "step": 1 |
| }, |
| { |
| "tag": "000002", |
| "step": 2 |
| }, |
| { |
| "tag": "000004", |
| "step": 4 |
| }, |
| { |
| "tag": "000008", |
| "step": 8 |
| }, |
| { |
| "tag": "000016", |
| "step": 16 |
| }, |
| { |
| "tag": "000032", |
| "step": 32 |
| }, |
| { |
| "tag": "000064", |
| "step": 64 |
| }, |
| { |
| "tag": "000128", |
| "step": 128 |
| }, |
| { |
| "tag": "000256", |
| "step": 256 |
| }, |
| { |
| "tag": "001000", |
| "step": 1000 |
| }, |
| { |
| "tag": "002000", |
| "step": 2000 |
| }, |
| { |
| "tag": "003000", |
| "step": 3000 |
| }, |
| { |
| "tag": "004000", |
| "step": 4000 |
| }, |
| { |
| "tag": "005000", |
| "step": 5000 |
| }, |
| { |
| "tag": "006000", |
| "step": 6000 |
| }, |
| { |
| "tag": "007000", |
| "step": 7000 |
| }, |
| { |
| "tag": "008000", |
| "step": 8000 |
| }, |
| { |
| "tag": "009000", |
| "step": 9000 |
| }, |
| { |
| "tag": "010000", |
| "step": 10000 |
| }, |
| { |
| "tag": "011000", |
| "step": 11000 |
| } |
| ], |
| "step": 11000, |
| "source": "AdaptiveChunking/hnet-chunking-spec-82m/spec_A_s1/011000.safetensors" |
| } |
