星海智算知识库

Appearance

【智谱「清影」同源-CogVideoX】部署教程 ​

1. 前置准备 ​

注册算力云平台：星海智算 https://gpu.spacehpc.com/

由于对于硬件的需求，推荐使用显存≥18GB的卡进行创建

1.1 创建并启动实例 ​

2. 相关依赖以及库文件安装 ​

2.1 设置学术加速 ​

由于需要访问外网，因此需要设置学术加速，学术加速的相关规则请见：学术加速

bash
source /etc/network_turbo

2.2 安装库以及相关依赖 ​

bash
pip install modelscope
pip install torch
pip install accelerate
pip install sentencepiece

bash
pip install --upgrade opencv-python transformers 
pip install git+https://github.com/huggingface/diffusers.git@878f609aa5ce4a78fea0f048726889debde1d7e8#egg=diffusers # Still in PR

3. 配置相关文件 ​

打开Notebook

复制代码进ipynb文件

py
import torch
from diffusers import CogVideoXPipeline
from diffusers.utils import export_to_video
from modelscope import snapshot_download

prompt = "A panda, dressed in a small, red jacket and a tiny hat, sits on a wooden stool in a serene bamboo forest. The panda's fluffy paws strum a miniature acoustic guitar, producing soft, melodic tunes. Nearby, a few other pandas gather, watching curiously and some clapping in rhythm. Sunlight filters through the tall bamboo, casting a gentle glow on the scene. The panda's face is expressive, showing concentration and joy as it plays. The background includes a small, flowing stream and vibrant green foliage, enhancing the peaceful and magical atmosphere of this unique musical performance."

model_dir = snapshot_download("ZhipuAI/CogVideoX-2b")

pipe = CogVideoXPipeline.from_pretrained(
    model_dir,
    torch_dtype=torch.float16
).to("cuda")

prompt_embeds, _ = pipe.encode_prompt(
    prompt=prompt,
    do_classifier_free_guidance=True,
    num_videos_per_prompt=1,
    max_sequence_length=226,
    device="cuda",
    dtype=torch.float16,
)

video = pipe(
    num_inference_steps=50,
    guidance_scale=6,
    prompt_embeds=prompt_embeds,
).frames[0]

export_to_video(video, "output.mp4", fps=8)

4. 结果输出 ​

点击运行

output.mp4为输出的视频文件，注意在后续的输出时需要更改输出文件的名称，防止对之前的文件进行覆盖

bash
export_to_video(video, "output.mp4", fps=8) //修改output.mp4为你喜欢的名称 例如test.mp4

Copyright@ 2024 spacehpc.com