| { | |
| "architectures": [ | |
| "BailingMoeLinearV2ForCausalLM" | |
| ], | |
| "attention_dropout": 0.0, | |
| "auto_map": { | |
| "AutoConfig": "configuration_bailing_moe_linear_v2.BailingMoeLinearV2Config", | |
| "AutoModel": "modeling_bailing_moe_linear_v2.BailingMoeLinearV2Model", | |
| "AutoModelForCausalLM": "modeling_bailing_moe_linear_v2.BailingMoeLinearV2ForCausalLM" | |
| }, | |
| "num_hidden_layers": 20, | |
| "hidden_size": 2048, | |
| "intermediate_size": 5120, | |
| "pad_token_id": 156892, | |
| "eos_token_id": 156892, | |
| "first_k_dense_replace": 1, | |
| "hidden_act": "silu", | |
| "max_position_embeddings": 131072, | |
| "model_type": "bailing_moe_linear", | |
| "moe_intermediate_size": 512, | |
| "norm_topk_prob": true, | |
| "num_experts_per_tok": 8, | |
| "num_attention_heads": 16, | |
| "num_experts": 256, | |
| "num_key_value_heads": 4, | |
| "rope_theta": 1000000, | |
| "rope_scaling": null, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.56.1", | |
| "use_bias": false, | |
| "use_rmsnorm": true, | |
| "rms_norm_eps": 1e-06, | |
| "head_dim": 128, | |
| "num_shared_experts": 1, | |
| "use_cache": true, | |
| "use_qkv_bias": false, | |
| "embedding_dropout": 0.0, | |
| "output_dropout": 0.0, | |
| "vocab_size": 157184, | |
| "partial_rotary_factor": 0.5, | |
| "router_dtype": "fp32", | |
| "moe_router_enable_expert_bias": true, | |
| "routed_scaling_factor": 2.5, | |
| "n_group": 8, | |
| "topk_group": 4, | |
| "use_qk_norm": true, | |
| "score_function": "sigmoid", | |
| "moe_shared_expert_intermediate_size": 512, | |
| "num_nextn_predict_layers": 0, | |
| "layer_group_size": 5, | |
| "group_norm_size": 4, | |
| "linear_silu": false | |
| } |