audio测试

准备环境

cd ./DataFlow-MM
conda create -n df_audio_test python=3.12
pip install -e .
pip install -e ".[audio]"
pip install -e ".[vllm]"

测试命令

python /data0/gty/DataFlow-MM/test/test_whisper_promptedvqa.py
python /data0/gty/DataFlow-MM/test/test_audio_promptedvqa.py

nano-banana (gemini-v2.5-image)测试

测试命令

python test/test_image_editing.py --api_key < your api key >

we utilize the api from yucha