Skip to content

Pull requests: quic/efficient-transformers

New pull request New
52 Open 1,014 Closed
52 Open 1,014 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

Bug fixes for missing kv_cache_buffer
#1082 opened Jun 15, 2026 by quic-rishinr Contributor Loading…
Minor fix to export 2L onnx for Qwen3_5
#1081 opened Jun 15, 2026 by tv-karthikeya Contributor Loading…
Fix DeepSeekV3 transformers compatibility
#1078 opened Jun 14, 2026 by sudheepm-wq Contributor Loading…
Bug fixes for missing kv_cache_buffer
#1077 opened Jun 13, 2026 by quic-sanising Contributor Loading…
[v1.22_tmp] Add layerwise FP16 sensitivity recovery skill + NPI helper
#1076 opened Jun 12, 2026 by anujgupt-github Contributor Loading…
ci(0612): fast per-PR pipeline, xdist 4-card sharding + tiny model lane
#1075 opened Jun 12, 2026 by vbaddi Contributor Loading…
@vbaddi
examples: add qwen3.5-moe layerwise NPI YAML + wired decode example
#1074 opened Jun 12, 2026 by anujgupt-github Contributor Loading…
nit(0612): Refine production cleanup for PR 1029 1.22 Release 1.22 candidate enhancement New feature or request
#1073 opened Jun 12, 2026 by vbaddi Contributor Loading…
@vbaddi @ochougul @quic-rishinr
Add YAML-aware from_pretrained scaling + runtime transform wiring
#1072 opened Jun 11, 2026 by anujgupt-github Contributor Loading…
Adding unit test and ci tests for gemma4
#1071 opened Jun 11, 2026 by tchawada Contributor • Draft
Adds repo-level AI agent setup guidance and set of skills
#1070 opened Jun 11, 2026 by ochougul Contributor • Draft
Replicate kv transform code with changes and comment addressal from # 1037
#1069 opened Jun 11, 2026 by quic-dhirajku Contributor Loading…
Adding vision and text npi files for E2B, E4B and 31B model
#1068 opened Jun 11, 2026 by tchawada Contributor Loading…
Feature/add deepseek v4
#1058 opened Jun 9, 2026 by shagsood • Draft
Feature/add glm moe dsa
#1057 opened Jun 9, 2026 by shagsood • Draft
Add onnx-ir dependency to pyproject 1.22 Release 1.22 candidate
#1054 opened Jun 9, 2026 by quic-amitraj Contributor Loading…
Meta state model loading for weightless subsequent runs
#1053 opened Jun 8, 2026 by Shrubabati7 Loading…
Reduce whole-model ONNX export memory 1.22 Release 1.22 candidate enhancement New feature or request
#1052 opened Jun 8, 2026 by anujgupt-github Contributor Loading…
18
Reduce ONNX export memory with external initializers enhancement New feature or request
#1050 opened Jun 8, 2026 by anujgupt-github Contributor Loading…
Rewrite layer-wise ONNX export as an API -> adds CustomLoader and Loop inside export
#1048 opened Jun 5, 2026 by ochougul Contributor Loading…
Reranker & Embedding: single-QPC support with KV cache eliminated 1.22 Release 1.22 candidate
#1045 opened Jun 5, 2026 by quic-amitraj Contributor Loading…
2
Added final support of dream model
#1042 opened Jun 5, 2026 by jinkasaisagar • Draft
KV handoff with DMA slicing APIs to avoid KV input/output copies.
#1039 opened Jun 4, 2026 by quic-akuruvil Contributor Loading…
@quic-akuruvil
[EB] Qwen_3_5_Moe 1.22 Release 1.22 candidate
#1038 opened Jun 4, 2026 by mohiso22 Contributor Loading…
4
Repeatkv transform 1.22 Release 1.22 candidate
#1037 opened Jun 4, 2026 by quic-dhirajku Contributor Loading…
4
ProTip! Mix and match filters to narrow down what you’re looking for.