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

fix the example scripts in parallel_inference_xdit.py

Browse Source
This commit is contained in:
Xibo Sun 2024-09-26 14:12:26 +08:00
parent dcb6795b20
commit 3304ea7275
1 changed files with 10 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
tools/parallel_inference/parallel_inference_xdit.py
View File

@ -7,16 +7,20 @@ https://github.com/xdit-project/xDiT/blob/main/examples/cogvideox_example.py
By using this code, the inference process is parallelized on multiple GPUs,
and thus speeded up.
You can also use the run.sh file in the same folder to automate running this
code for batch generation of videos.
Usage:
1. pip install xfuser
2. run the following command to generate video
torchrun --nproc_per_node=4 cogvideox_xdit.py --model <cogvideox-model-path> \
--ring_degree 2 --use_cfg_parallel --height 480 --width 720 --num_frames 9 \
2. mkdir results
3. run the following command to generate video
torchrun --nproc_per_node=4 parallel_inference_xdit.py \
--model <cogvideox-model-path> --ulysses_degree 1 --ring_degree 2 \
--use_cfg_parallel --height 480 --width 720 --num_frames 9 \
--prompt 'A small dog.'
You can also use the run.sh file in the same folder to automate running this
code for batch generation of videos, by running:
sh ./run.sh
"""
import time