JLNTV_IPD/CogVideo
4
0
Fork 0
mirror of https://github.com/THUDM/CogVideo.git synced 2025-04-05 19:41:59 +08:00
Code Issues Packages Projects Releases Wiki Activity

Add resolution warning

Browse Source
This commit is contained in:
OleehyO 2024-12-16 11:34:51 +00:00
parent 7b4c9db6d9
commit d3a7d2dc91
1 changed files with 6 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
inference/cli_demo.py
View File

@ -17,6 +17,7 @@ $ python cli_demo.py --prompt "A girl riding a bike." --model_path THUDM/CogVide
Additional options are available to specify the model path, guidance scale, number of inference steps, video generation type, and output paths.
"""
import warnings
import argparse
from typing import Literal
@ -77,6 +78,11 @@ def generate_video(
image = None
video = None
if (width != 1360 or height != 768) and "cogvideox1.5-5b-i2v" in model_path.lower():
warnings.warn(f"The width({width}) and height({height}) are not recommended for CogVideoX1.5-5B-I2V. The best resolution for CogVideoX1.5-5B-I2V is 1360x768.")
elif (width != 720 or height != 480) and "cogvideox-5b-i2v" in model_path.lower():
warnings.warn(f"The width({width}) and height({height}) are not recommended for CogVideo-5B-I2V. The best resolution for CogVideo-5B-I2V is 720x480.")
if generate_type == "i2v":
pipe = CogVideoXImageToVideoPipeline.from_pretrained(model_path, torch_dtype=dtype)
image = load_image(image=image_or_video_path)