Audio-Text-to-Text
Transformers
Safetensors
English
qwen2_audio
text2text-generation
audio
speech
audio-llm
paralinguistic
pclm
dpo
voxparadox
Instructions to use IHP-Lab/Qwen2-Audio_PCLM_DPO with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use IHP-Lab/Qwen2-Audio_PCLM_DPO with Transformers:
# Load model directly from transformers import AutoProcessor, AutoModelForSeq2SeqLM processor = AutoProcessor.from_pretrained("IHP-Lab/Qwen2-Audio_PCLM_DPO") model = AutoModelForSeq2SeqLM.from_pretrained("IHP-Lab/Qwen2-Audio_PCLM_DPO") - Notebooks
- Google Colab
- Kaggle
| { | |
| "_name_or_path": "IHP-Lab/Qwen2-Audio_PCLM_DPO", | |
| "architectures": [ | |
| "Qwen2AudioForConditionalGeneration" | |
| ], | |
| "audio_config": { | |
| "model_type": "qwen2_audio_encoder" | |
| }, | |
| "audio_token_index": 151646, | |
| "expose_layers": [ | |
| 5, | |
| 15, | |
| 25, | |
| 30 | |
| ], | |
| "ignore_index": -100, | |
| "model_type": "qwen2_audio", | |
| "pclm_exclude_final": false, | |
| "pclm_text_model": "prajjwal1/bert-small", | |
| "text_config": { | |
| "bos_token_id": 151643, | |
| "eos_token_id": 151645, | |
| "intermediate_size": 11008, | |
| "max_position_embeddings": 8192, | |
| "model_type": "qwen2", | |
| "rms_norm_eps": 1e-05, | |
| "torch_dtype": "bfloat16", | |
| "use_mrope": false, | |
| "vocab_size": 156032 | |
| }, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.46.0", | |
| "use_pclm": true, | |
| "vocab_size": 156032 | |
| } |