-
Notifications
You must be signed in to change notification settings - Fork 20
Pull requests: alibaba/ChatLearn
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
Support force free memory for policy model with no colocate.
#218
opened Jan 26, 2025 by
adoda
Loading…
parallel async call model replica to improve setup time
#217
opened Jan 26, 2025 by
Yancey1989
Loading…
feat[param_sync][WIP]: allow k/v replicate in policy generation
#215
opened Jan 23, 2025 by
haolin-nju
Loading…
fix[vllm benchmark]: avoid num input tokens overflow
#213
opened Jan 22, 2025 by
haolin-nju
Loading…
fix(parameter_sync): default regrouping parameters to all-gather instead of all-to-all
#202
opened Jan 6, 2025 by
haolin-nju
Loading…
feature(parameter_sync): offer 2 GPU memory optimization levels in broadcasting parameters
#195
opened Dec 27, 2024 by
haolin-nju
Loading…
[WIP]feature(mixtral): support Mixtral-8x7B SFT, Reward, and Alignment
#95
opened Sep 24, 2024 by
haolin-nju
Loading…
ProTip!
What’s not been updated in a month: updated:<2024-12-26.