| { |
| "run": "pilot_D_s2", |
| "step": 6000, |
| "model_config": { |
| "vocab_size": 256, |
| "d_enc": 256, |
| "d_main": 512, |
| "n_enc": 2, |
| "n_dec": 2, |
| "n_main": 6, |
| "n_heads_enc": 4, |
| "n_heads_main": 8, |
| "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": "D", |
| "seed": 2, |
| "out_dir": null, |
| "run_name": "pilot_D_s2", |
| "corpus": "corpus", |
| "langs": null, |
| "seq_len": 1024, |
| "batch_size": 32, |
| "grad_accum": 1, |
| "steps": 6000, |
| "lr": 0.0003, |
| "min_lr_frac": 0.1, |
| "warmup": 300, |
| "weight_decay": 0.1, |
| "grad_clip": 1.0, |
| "betas": [ |
| 0.9, |
| 0.95 |
| ], |
| "router_lr_mult": 1.0, |
| "d_enc": 256, |
| "d_main": 512, |
| "n_enc": 2, |
| "n_dec": 2, |
| "n_main": 6, |
| "n_heads_enc": 4, |
| "n_heads_main": 8, |
| "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": 200, |
| "eval_every": 200, |
| "log_every": 50, |
| "probe_sents": 128, |
| "eval_batches": 8, |
| "keep_ckpt": false, |
| "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 |
| }, |
| "probe_agg": { |
| "gini_cpc": 0.1757837534840907, |
| "gini_cps": 0.01599505125486056, |
| "bpic": 10.709968186638388, |
| "worst_lang": "zh", |
| "cpc_parity_vs_en": { |
| "en": 1.0, |
| "de": 0.8944228849244555, |
| "vi": 0.9569868292813777, |
| "tr": 0.9851015829444819, |
| "fi": 0.9070411258376411, |
| "ru": 0.9773387348139986, |
| "el": 0.8520522471973283, |
| "hi": 1.0783989190601373, |
| "ta": 0.9644563797395437, |
| "th": 1.1005169566212463, |
| "zh": 3.1702951901214265, |
| "my": 0.9099451006166094 |
| } |
| }, |
| "source": "AdaptiveChunking/hnet-chunking-pilots/pilot_D_s2/model.safetensors" |
| } |
