Modelpublic

tranquangchung/qwen2-audio-dialogue

sourceHugging Faceapache-2.0updated 8mo agoView on Hugging Face

0likes12downloads

Model Card

license: apache-2.0 library_name: transformers tags:

—audio
—speech-to-speech
—multimodal
—qwen2-audio ---

Dialogue

This repository contains a fine-tuned version of Qwen2-Audio optimized for speech-to-speech dialogue interactions.

Model Description

—Architecture: Qwen2-Audio (0.5B Backbone)

Usage

Since this model uses a custom implementation, ensure the models directory is present in your working directory.

python

from models.modeling_qwen2_audio import Qwen2AudioForConditionalGeneration
from models.processing_qwen2_audio import Qwen2AudioProcessor

# Load model and processor
model = Qwen2AudioForConditionalGeneration.from_pretrained("tranquangchung/qwen2-audio-dialogue")
processor = Qwen2AudioProcessor.from_pretrained("tranquangchung/qwen2-audio-dialogue")