Text Generation
Transformers
Safetensors
minimax_m2
conversational
custom_code
fp8
MiniMax-M2 / config.json
AntonV's picture
AntonV HF Staff
v5 support
8bdd6aa
raw
history blame
1.05 kB
{
"attention_dropout": 0.0,
"bos_token_id": 200034,
"eos_token_id": 200020,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 3072,
"initializer_range": 0.02,
"intermediate_size": 1536,
"max_position_embeddings": 196608,
"model_type": "minimax_m2",
"num_attention_heads": 48,
"num_experts_per_tok": 8,
"num_hidden_layers": 62,
"num_key_value_heads": 8,
"num_local_experts": 256,
"output_router_logits": false,
"quantization_config": {
"activation_scheme": "dynamic",
"dequantize": false,
"modules_to_not_convert": [
"gate",
"e_score_correction_bias",
"lm_head"
],
"quant_method": "fp8",
"weight_block_size": [
128,
128
]
},
"rms_norm_eps": 1e-06,
"rope_parameters": {
"partial_rotary_factor": 0.5,
"rope_theta": 5000000.0,
"rope_type": "default"
},
"router_aux_loss_coef": 0.001,
"router_jitter_noise": 0.0,
"tie_word_embeddings": false,
"transformers_version": "5.0.0.dev0",
"use_cache": true,
"vocab_size": 200064
}