-
Notifications
You must be signed in to change notification settings - Fork 1.2k
Pull requests: modelscope/DiffSynth-Studio
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
wan: add dual-GPU model-parallel path for Wan 2.x LoRA training (depends on #1435)
#1436
opened May 11, 2026 by
genno-whittlery
Loading…
4 of 5 tasks
fix(wan): apply patchify flatten in forward/consumers (fixes #1063)
#1435
opened May 11, 2026 by
genno-whittlery
Loading…
2 of 3 tasks
flux2: dual-GPU model-parallel + transformers 5.8 compat + Mistral-on-CPU data_process
#1434
opened May 11, 2026 by
genno-whittlery
Loading…
4 of 6 tasks
feat: add resume_from_checkpoint option to training configuration
#1406
opened Apr 23, 2026 by
hoangvictor
Loading…
feat(wan): Add prior-based diffusion step skip for ~70% fewer inference steps
#1325
opened Mar 4, 2026 by
Efrat-Taig
Loading…
[New Feasture]: Support SP + DP parallal on Wan training
#1223
opened Jan 21, 2026 by
mahaocong90
Loading…
feat: Add multi-GPU support with data/model/tensor parallelism
#1153
opened Dec 21, 2025 by
longkeyy
Loading…
3 of 5 tasks
[New Feasture]: Support for Flow-GRPO/Dance-GRPO
#1141
opened Dec 17, 2025 by
Jayce-Ping
•
Draft
2 of 10 tasks
支持int8-quanto量化微调Qwen-Image-Edit-2509,48G显存即可微调🎉,训练脚本添加result_image_field_name参数.
#1101
opened Dec 7, 2025 by
Deng-Xian-Sheng
Loading…
feat: Add LoRA-based knowledge distillation for Wan-Video
#922
opened Sep 16, 2025 by
MaulanaYusufIkhsanRobbani
Loading…
I have created the demonstration script `examples/wanvideo/model_infe…
#867
opened Aug 28, 2025 by
MaulanaYusufIkhsanRobbani
Loading…
ProTip!
Follow long discussions with comments:>50.