Skip to content

Pull requests: deepseek-ai/DeepEP

New pull request New
13 Open 84 Closed
13 Open 84 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Support prefill with 2 GPUs
#331 opened Jul 27, 2025 by fzyzcjy Loading…
Add hidden_size 6144
#329 opened Jul 25, 2025 by jeejeelee Loading…
2
enhance warp copy efficiency in cached_notify()
#315 opened Jul 18, 2025 by ZhiyiHu1999 Loading…
2
Fix the address of dispatch_rdma_recv_count_buffer in LL Layout to avoid cleaning after each change in hidden_size/token_num.
#313 opened Jul 18, 2025 by yilin-void Loading…
2
Add diagnosis module for efficient and precise location of slow rank
#311 opened Jul 17, 2025 by wangfakang Loading…
1
support low latency dispatch tma
#293 opened Jul 10, 2025 by ayrnb Loading…
2
Tiny support custom nvcc flags
#280 opened Jul 5, 2025 by fzyzcjy Loading…
Allow using few SMs for low-latency mode
#277 opened Jul 3, 2025 by fzyzcjy Loading…
Tiny allow testing in MNNVL environments
#275 opened Jul 3, 2025 by fzyzcjy Loading…
4
Computation communication overlap
#249 opened Jun 24, 2025 by fzyzcjy • Draft
Optimize low latency combine recv kernel (about 3.0x speedup)
#248 opened Jun 23, 2025 by fzyzcjy Loading…
7
Support other NVLink scenarios
#218 opened Jun 17, 2025 by fzyzcjy Loading…
9
Support cuMem API in cross process shared memory management
#217 opened Jun 17, 2025 by fzyzcjy Loading…
4
ProTip! What’s not been updated in a month: updated:<2025-06-27.