Feature Extraction
Transformers
Safetensors
English
Chinese
emova
Omni-modal-LLM
Multi-modal-LLM
Emotional-spoken-dialogue
custom_code
Eval Results
emova-qwen-2-5-72b-hf / config.json
KaiChen1998's picture
Update config.json
9591baa verified
{
"architectures": [
"EMOVAForConditionalGeneration"
],
"auto_map": {
"AutoConfig": "configuration_emova.EMOVAConfig",
"AutoModel": "modeling_emova.EMOVAForConditionalGeneration",
"AutoModelForCausalLM": "modeling_emova.EMOVAForConditionalGeneration"
},
"default_system_prompt": "You are a helpful assistant.",
"ignore_index": -100,
"image_token_index": 155761,
"mm_projector_config": {
"mlp_depth": 2,
"trainable": true,
"type": "MLPProjector"
},
"model_type": "emova",
"text_config": {
"architectures": [
"Qwen2ForCausalLM"
],
"bos_token_id": 151643,
"eos_token_id": 151645,
"hidden_size": 8192,
"intermediate_size": 29568,
"max_position_embeddings": 32768,
"max_window_layers": 70,
"model_type": "qwen2",
"num_attention_heads": 64,
"num_hidden_layers": 80,
"num_key_value_heads": 8,
"rope_theta": 1000000.0,
"sliding_window": null,
"tokenizer_model_max_length": 8192,
"tokenizer_padding_side": "left",
"torch_dtype": "float32",
"use_sliding_window": false,
"vocab_size": 155776
},
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.44.2",
"vision_config": {
"max_pixels": 3211264,
"model_type": "qwen2_vl",
"pretrained_model_name_or_path": "Emova-ollm/qwen2vit600m",
"trainable": true,
"type": "Qwen2VisionTower",
"unfreeze_mm_vision_tower": true
}
}