| { | |
| "model_type": "causal_diffusion", | |
| "interpretable": true, | |
| "n_layers": 32, | |
| "n_head": 32, | |
| "n_embd": 4096, | |
| "block_size": 4096, | |
| "n_kv_heads": 4, | |
| "diff_block_size": 64, | |
| "use_rms_norm": true, | |
| "norm_eps": 1e-05, | |
| "norm_order": "post", | |
| "use_qk_norm": true, | |
| "use_rope": true, | |
| "rope_base": 500000.0, | |
| "rope_full_precision": true, | |
| "mlp_type": "swiglu", | |
| "activation": "gelu", | |
| "mlp_ratio": 4, | |
| "intermediate_size": null, | |
| "use_bias": false, | |
| "clip_qkv": 10.0, | |
| "weight_sharing": true, | |
| "pad_token_id": 100277, | |
| "bos_token_id": 100278, | |
| "eos_token_id": 100257, | |
| "endofchunk_token_id": 100279, | |
| "mask_token_id": 100280, | |
| "vocab_size": 100281, | |
| "concept": { | |
| "n_concepts": 33732, | |
| "n_unknown_concepts": 101196, | |
| "max_concepts": 16, | |
| "concept_dim": 4096, | |
| "use_attention_known": false, | |
| "use_attention_unknown": false, | |
| "topk_known": 16, | |
| "topk_known_features": 32, | |
| "unknown_topk": 128, | |
| "use_unknown": true, | |
| "apply_topk_to_unknown": true, | |
| "topk_on_logits": false, | |
| "factorize_unknown": true, | |
| "factorize_rank": 256, | |
| "use_epsilon_correction": true, | |
| "block_size": 4096, | |
| "pad_multiple": 16, | |
| "store_unknown_weights": false, | |
| "inject_layer": 16, | |
| "inject_alpha": 1.0 | |
| } | |
| } |