config.json

{
  "_name_or_path": "/path/to/Uni-MoE-speech-base",
  "architectures": [
    "LlavaLlamaForCausalLM"
  ],
  "bos_token_id": 1,
  "eos_token_id": 2,
  "freeze_mm_mlp_adapter": false,
  "freeze_mm_vision_resampler": false,
  "hidden_act": "silu",
  "hidden_size": 4096,
  "image_aspect_ratio": "pad",
  "initializer_range": 0.02,
  "intermediate_size": 11008,
  "language": "English",
  "local_files_only": false,
  "max_length": 4096,
  "max_position_embeddings": 4096,
  "mm_audio_hidden_size": 768,
  "mm_audio_select_feature": "patch",
  "mm_audio_select_layer": 0,
  "mm_audio_tower": "/path/to/whisper-small",
  "mm_hidden_size": 1024,
  "mm_projector_type": "mlp2x_gelu",
  "mm_query_tokens": true,
  "mm_resampler_type": null,
  "mm_use_im_patch_token": false,
  "mm_use_im_start_end": false,
  "mm_vision_select_feature": "patch",
  "mm_vision_select_layer": -2,
  "mm_vision_tower": "/path/to/clip-vit-large-patch14-336",
  "model_type": "llava",
  "moe": true,
  "num_attention_heads": 32,
  "num_experts": 4,
  "num_experts_per_tok": 2,
  "num_hidden_layers": 32,
  "num_key_value_heads": 32,
  "pad_token_id": 0,
  "pretraining_tp": 1,
  "query_tokens_size": 50,
  "rms_norm_eps": 1e-05,
  "rope_scaling": null,
  "task": "transcribe",
  "tie_word_embeddings": false,
  "torch_dtype": "float32",
  "transformers_version": "4.31.0",
  "tune_mm_audio_aligner": true,
  "tune_mm_mlp_adapter": false,
  "tune_mm_vision_resampler": false,
  "unfreeze_mm_vision_tower": false,
  "use_cache": true,
  "use_mm_proj": true,
  "vocab_size": 32000
}