Skip to content

Pull requests: meta-recsys/generative-recommenders

New pull request New
110 Open 350 Closed
110 Open 350 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

Swap -c CUDA version flags for the constraint modifier CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#569 opened Aug 7, 2026 by wenxin0319 Loading…
1
num_stages=0 -> 1 (meta-recsys/generative-recommenders) CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#568 opened Aug 4, 2026 by njriasan Contributor Loading…
1
Fix num_stages range CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#567 opened Aug 4, 2026 by njriasan Contributor Loading…
1
Fix Triton layer norm u activation forwarding CLA Signed This label is managed by the Meta Open Source bot.
#566 opened Aug 4, 2026 by wang1ang Loading…
Use one compiler stage for H100 TLX configs CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#564 opened Jul 31, 2026 by mark14wu Loading…
2
Add local attention kernels and coverage CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#562 opened Jul 30, 2026 by njriasan Contributor Loading…
1
Fix: use bitwise operators instead of logical operators in Triton kernels CLA Signed This label is managed by the Meta Open Source bot.
#561 opened Jul 28, 2026 by wang1ang Loading…
Make torchrec.modules.sorted_index_select import defensive CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#560 opened Jul 24, 2026 by LinjianMa Contributor Loading…
1
fbcode/generative_recommenders/ops/cpp/hstu_attention/flash_common.h CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#559 opened Jul 24, 2026 by meta-codesync Bot Loading…
HSTU CA-overarch ranker architecture, config providers, and kernels CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#558 opened Jul 22, 2026 by XialuLi Loading…
1
Optimize jagged-jagged BMM reduce-sum kernel CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#557 opened Jul 21, 2026 by arsatis Loading…
1
Fix int32 overflow in fused SwiGLU forward kernel output store for large M CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#556 opened Jul 17, 2026 by Zhihan-Lu Loading…
1
Enable multirow concat/split 2D jagged kernels on H100 (sm90+) CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#555 opened Jul 16, 2026 by mrshenli Loading…
1
Add tlx based FP8 addm kernel for AMD MI350 CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#554 opened Jul 16, 2026 by rguo-aws Loading…
1
sorted_index_select: APP-layer callers (hammer/dper_lib/generative_recommenders) CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#553 opened Jul 16, 2026 by hpnhxxwn Contributor Loading…
1
Change cuda hstu max_q_len type in CPU/meta dispatch CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#552 opened Jul 15, 2026 by Indicator Loading…
1
Use triton multi-row and helion for concat/split 2D jagged CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#551 opened Jul 14, 2026 by rafaykhurram Contributor Loading…
1
sorted_index_select: atomic-light segment-reduce backward for index_select CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#550 opened Jul 14, 2026 by hpnhxxwn Contributor Loading…
1
Add bf16 accumulation option for HSTU attention bwd CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#549 opened Jul 8, 2026 by rafaykhurram Contributor Loading…
1
Avoid int32 overflow in HSTU preprocessing CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#548 opened Jun 29, 2026 by tangyuhao Contributor Loading…
1
SymInt support in cuda attn kernel, bwd meta dispatch CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#547 opened Jun 24, 2026 by ruochen99 Contributor Loading…
1
Enable Pyrefly for generative_recommenders CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#546 opened Jun 18, 2026 by maggiemoss Contributor Loading…
1
Support 2-CTA addmm grid sizing CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#543 opened Jun 16, 2026 by njriasan Contributor Loading…
1
Support DP=2 with 2-CTA for addmm CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#542 opened Jun 15, 2026 by jananisriram Contributor Loading…
1
Optimize rms norm bwd by avoiding atomic_add CLA Signed This label is managed by the Meta Open Source bot. meta-exported
#541 opened Jun 12, 2026 by vinitX Contributor Loading…
1
ProTip! Follow long discussions with comments:>50.