| { | |
| "architectures": [ | |
| "MiMoV2FlashModel" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "attention_value_scale": 0.707, | |
| "bos_token_id": 1, | |
| "dtype": "float32", | |
| "eos_token_id": 151645, | |
| "head_dim": 8, | |
| "hidden_act": "silu", | |
| "hidden_size": 32, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 64, | |
| "layer_types": [ | |
| "full_attention", | |
| "sliding_attention" | |
| ], | |
| "max_position_embeddings": 200, | |
| "mlp_layer_types": [ | |
| "dense", | |
| "sparse" | |
| ], | |
| "model_type": "mimo_v2_flash", | |
| "moe_intermediate_size": 16, | |
| "n_group": 1, | |
| "n_routed_experts": 4, | |
| "norm_topk_prob": true, | |
| "num_attention_heads": 4, | |
| "num_experts_per_tok": 2, | |
| "num_hidden_layers": 2, | |
| "num_key_value_heads": 2, | |
| "pad_token_id": 151643, | |
| "rms_norm_eps": 1e-05, | |
| "rope_parameters": { | |
| "full_attention": { | |
| "partial_rotary_factor": 0.5, | |
| "rope_theta": 5000000.0, | |
| "rope_type": "default" | |
| }, | |
| "sliding_attention": { | |
| "partial_rotary_factor": 0.5, | |
| "rope_theta": 10000.0, | |
| "rope_type": "default" | |
| } | |
| }, | |
| "routed_scaling_factor": 1.0, | |
| "sliding_window": 64, | |
| "tie_word_embeddings": false, | |
| "topk_group": 1, | |
| "transformers_version": "5.16.0.dev0", | |
| "use_cache": true, | |
| "v_head_dim": 16, | |
| "vocab_size": 151669 | |
| } | |