| { |
| "architectures": [ |
| "HYV4ForCausalLM" |
| ], |
| "attention_bias": false, |
| "attention_dropout": 0.0, |
| "bos_token_id": 1, |
| "dtype": "float32", |
| "enable_ihc": true, |
| "enable_lm_head_fp32": true, |
| "eos_token_id": 2, |
| "gated_mla": true, |
| "gating_type": "elementwise", |
| "hc_eps": 1e-06, |
| "hc_magnitude": 2.0, |
| "hc_mult": 4, |
| "head_dim": 4, |
| "hidden_act": "silu", |
| "hidden_size": 32, |
| "index_head_dim": 8, |
| "index_n_heads": 4, |
| "index_topk": 8, |
| "indexer_types": [ |
| "full", |
| "full", |
| "shared", |
| "full" |
| ], |
| "initializer_range": 0.006, |
| "intermediate_size": 64, |
| "kv_lora_rank": 8, |
| "layer_types": [ |
| "deepseek_sparse_attention", |
| "deepseek_sparse_attention", |
| "deepseek_sparse_attention", |
| "deepseek_sparse_attention" |
| ], |
| "learnable_sink": true, |
| "learnable_sink_init": 0.0, |
| "max_position_embeddings": 262144, |
| "mlp_layer_types": [ |
| "dense", |
| "sparse", |
| "sparse", |
| "sparse" |
| ], |
| "model_type": "hy_v4", |
| "moe_intermediate_size": 16, |
| "n_group": 1, |
| "n_routed_experts": 4, |
| "n_shared_experts": 1, |
| "norm_topk_prob": true, |
| "num_attention_heads": 4, |
| "num_experts_per_tok": 2, |
| "num_hidden_layers": 4, |
| "num_key_value_heads": 4, |
| "pad_token_id": 0, |
| "q_lora_rank": 16, |
| "qk_head_dim": 8, |
| "qk_nope_head_dim": 4, |
| "qk_rope_head_dim": 4, |
| "rms_norm_eps": 1e-05, |
| "rope_parameters": { |
| "rope_theta": 10000.0, |
| "rope_type": "default" |
| }, |
| "routed_scaling_factor": 2.827, |
| "swiglu_limit": 10.0, |
| "tie_word_embeddings": false, |
| "topk_group": 1, |
| "transformers_version": "5.15.0.dev0", |
| "use_cache": true, |
| "v_head_dim": 8, |
| "vocab_size": 99 |
| } |
|
|
