ViBES-T2M / config.json
JuzeZhang's picture
ViBES-T2M: 9B text-to-motion expert (expert1-only, mot_vocab 514, HumanML3D)
5abef14 verified
Raw
History Blame Contribute Delete
1.68 kB
{
"_name_or_path": "JuzeZhang/ViBES-T2M",
"add_bias_linear": false,
"add_qkv_bias": true,
"apply_query_key_layer_scaling": true,
"apply_residual_connection_post_layernorm": false,
"architectures": [
"ChatGLMForConditionalGenerationMotExpertNum2"
],
"attention_dropout": 0.0,
"attention_softmax_in_fp32": true,
"auto_map": {
"AutoConfig": "configuration_chatglm.ChatGLMConfig",
"AutoModel": "modeling_chatglm.ChatGLMForConditionalGenerationMotExpertNum2",
"AutoModelForCausalLM": "modeling_chatglm.ChatGLMForConditionalGenerationMotExpertNum2",
"AutoModelForSeq2SeqLM": "modeling_chatglm.ChatGLMForConditionalGenerationMotExpertNum2",
"AutoModelForSequenceClassification": "modeling_chatglm.ChatGLMForSequenceClassification"
},
"bias_dropout_fusion": true,
"classifier_dropout": null,
"eos_token_id": [
151329,
151336,
151338
],
"ffn_hidden_size": 13696,
"fp32_residual_connection": false,
"hidden_dropout": 0.0,
"hidden_size": 4096,
"kv_channels": 128,
"layernorm_epsilon": 3.90625e-08,
"model_type": "chatglm",
"mot_settings": {
"mot_ffn_hidden_size": 4096,
"mot_hidden_size": 512,
"mot_vocab_size": 514
},
"multi_query_attention": true,
"multi_query_group_num": 2,
"num_attention_heads": 32,
"num_hidden_layers": 40,
"num_layers": 40,
"original_rope": true,
"pad_token_id": 151329,
"padded_vocab_size": 168960,
"post_layer_norm": true,
"rmsnorm": true,
"rope_ratio": 1,
"seq_length": 8192,
"t1_layer_limit": 40,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.48.3",
"use_cache": true,
"vocab_size": 168960
}