| { | |
| "_name_or_path": "CrossDNA-pretrain", | |
| "model_type": "crossdna", | |
| "architectures": ["CrossDNAForMaskedLM"], | |
| "auto_map": { | |
| "AutoConfig": "configuration_crossdna.CrossDNAConfig", | |
| "AutoModelForMaskedLM": "modeling_crossdna.CrossDNAForMaskedLM", | |
| "AutoTokenizer": "tokenization_crossdna.CrossDNATokenizer" | |
| }, | |
| "torch_dtype": "float32", | |
| "alphabet_size": 5, | |
| "d_model": 1024, | |
| "block_size": 1024, | |
| "depth": 12, | |
| "drop_path_rates": [0.0, 0.05], | |
| "dropout": 0.15, | |
| "pretrain": true, | |
| "for_representation": false, | |
| "use_s_scan": true, | |
| "use_bridge": true, | |
| "use_mem": false, | |
| "use_rc_kl": false, | |
| "use_barlow": false, | |
| "use_tv": false, | |
| "sem_max_weight": 0.12, | |
| "sem_warmup_steps": 10000, | |
| "aux_ce_weight": 0.0, | |
| "gate_freeze_steps": 5000, | |
| "detach_gate": false, | |
| "gate_sup_weight": 0.02, | |
| "gate_sup_warmup_steps": 1000, | |
| "gate_temp": 2.0, | |
| "transformer_cfg": { | |
| "hidden_size": 1024, | |
| "norm_eps": 1e-5, | |
| "max_position_embeddings": 1024, | |
| "hidden_ratio": 4.0, | |
| "hidden_act": "swish", | |
| "fuse_swiglu": true, | |
| "attn": { | |
| "num_heads": 8, | |
| "num_kv_heads": 8, | |
| "qkv_bias": false, | |
| "window_size": 512, | |
| "rope_theta": 10000 | |
| } | |
| }, | |
| "comba_cfg": { | |
| "hidden_size": 1024, | |
| "expand_v": 1, | |
| "head_dim": 128, | |
| "num_heads": 8, | |
| "use_gate": true, | |
| "mode": "chunk", | |
| "use_short_conv": true, | |
| "correction_factor": 0.02, | |
| "conv_size": 4, | |
| "norm_eps": 1e-5 | |
| }, | |
| "pad_token_id": 4, | |
| "bos_token_id": 2, | |
| "sep_token_id": 1, | |
| "cls_token_id": 0, | |
| "mask_token_id": 3, | |
| "vocab_size": 5 | |
| } | |