FramePack实测

昨天晚上刷推看到敏神发布了FramePack，居然只要6G显存，于是来玩一下。
个人PC配置：

• 内存：16G*2 4800 MHz
• GPU：RTX 4060

运行过程中内存基本占满，GPU利用率也基本上100%

环境搭建

git clone https://github.com/lllyasviel/FramePack.git
cd FramePack
conda create -n framepack python=3.10
conda activate framepack
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu124
pip install -r requirements.txt

安装flash-attn，但作者安装的sage-attention。

pip install .\flash_attn-2.7.4%2Bcu124torch2.6.0cxx11abiFALSE-cp310-cp310-win_amd64.whl

运行过程会下载一堆模型，耗时很长

text_encoder = LlamaModel.from_pretrained("hunyuanvideo-community/HunyuanVideo", subfolder='text_encoder', torch_dtype=torch.float16).cpu()
text_encoder_2 = CLIPTextModel.from_pretrained("hunyuanvideo-community/HunyuanVideo", subfolder='text_encoder_2', torch_dtype=torch.float16).cpu()
tokenizer = LlamaTokenizerFast.from_pretrained("hunyuanvideo-community/HunyuanVideo", subfolder='tokenizer')
tokenizer_2 = CLIPTokenizer.from_pretrained("hunyuanvideo-community/HunyuanVideo", subfolder='tokenizer_2')
vae = AutoencoderKLHunyuanVideo.from_pretrained("hunyuanvideo-community/HunyuanVideo", subfolder='vae', torch_dtype=torch.float16).cpu()
feature_extractor = SiglipImageProcessor.from_pretrained("lllyasviel/flux_redux_bfl", subfolder='feature_extractor')
image_encoder = SiglipVisionModel.from_pretrained("lllyasviel/flux_redux_bfl", subfolder='image_encoder', torch_dtype=torch.float16).cpu()
transformer = HunyuanVideoTransformer3DModelPacked.from_pretrained('lllyasviel/FramePackI2V_HY', torch_dtype=torch.bfloat16).cpu()

运行，端口自己随意，没被占用就可以

python demo_gradio.py --port 7750

效果
输入图像

prompts：A man is waving a flag in the midst of the war