{ "architectures": [ "MiniMaxM2ForCausalLM" ], "attn_type_list": [ 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 ], "auto_map": { "AutoConfig": "configuration_minimax_m2.MiniMaxM2Config", "AutoModelForCausalLM": "modeling_minimax_m2.MiniMaxM2ForCausalLM" }, "dtype": "bfloat16", "head_dim": 128, "hidden_act": "silu", "hidden_size": 3072, "intermediate_size": 1536, "max_position_embeddings": 196608, "model_type": "minimax_m2", "mtp_transformer_layers": 0, "num_attention_heads": 48, "num_experts_per_tok": 8, "num_hidden_layers": 62, "num_key_value_heads": 8, "num_local_experts": 256, "num_mtp_modules": 0, "qk_norm_type": "per_layer", "rms_norm_eps": 1e-06, "rope_theta": 5000000, "rotary_dim": 64, "scoring_func": "sigmoid", "shared_intermediate_size": 0, "tie_word_embeddings": false, "transformers_version": "4.46.1", "use_cache": true, "use_mtp": false, "use_qk_norm": true, "use_routing_bias": true, "vocab_size": 200064, "_name_or_path": "M2.5-SLURPY", "quantization": { "bits": 8, "mode": "affine", "group_size": 64, "routed_expert_bits": { "gate_proj": 2, "down_proj": 4, "up_proj": 2 }, "mxtq_bits": { "routed_expert": { "gate_proj": 2, "down_proj": 4, "up_proj": 2 }, "attention": 8, "shared_expert": 8, "embed_tokens": 8, "lm_head": 8, "norms_router_biases": 16 } }, "mxtq_bits": { "routed_expert": { "gate_proj": 2, "down_proj": 4, "up_proj": 2 }, "attention": 8, "shared_expert": 8, "embed_tokens": 8, "lm_head": 8, "norms_router_biases": 16 }, "weight_format": "mxtq" }