Skip to content

Pull requests: microsoft/onnxruntime-genai

New pull request New
45 Open 1,416 Closed
45 Open 1,416 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

Add Config::shared_assets_path for shared tokenizer/processor assets
#2126 opened May 6, 2026 by xiaoyu-work Contributor Loading…
1
Add JSON DOM and RFC 7386 merge_patch to src/json
#2125 opened May 6, 2026 by xiaoyu-work Contributor Loading…
3
Add granitemode support
#2124 opened May 6, 2026 by amdrajeevp1 Contributor Loading…
4 tasks
2
Fix multimodal CUDA pipeline: embedding output persistence causes shape mismatch
#2123 opened May 5, 2026 by justinchuby Contributor Loading…
1
10
Add Seed-OSS architecture support (SeedOssForCausalLM)
#2116 opened May 3, 2026 by PMeeske • Draft
2
Pipeline-as-Config: Declarative model dispatch replacing model_type string registry
#2115 opened May 2, 2026 by justinchuby Contributor • Draft
1
8
Cohere Transcribe Support
#2112 opened May 2, 2026 by nenad1002 Contributor • Draft
10
Fix AppendNextTokensToSequences heap overflow
#2111 opened Apr 30, 2026 by apsonawane Contributor Loading…
3
Fix heap overflow issue
#2110 opened Apr 30, 2026 by apsonawane Contributor Loading…
5
Add linux-aarch64 support
#2107 opened Apr 29, 2026 by baijumeswani Collaborator • Draft
Release captured graph resources when generator is destroyed
#2106 opened Apr 29, 2026 by qjia7 Contributor • Draft
WIP: Enable zero-copy for QNN GPU
#2105 opened Apr 27, 2026 by qti-mattsinc • Draft
3
Update generic example scripts for plugin QNN EP
#2104 opened Apr 27, 2026 by qti-kromero Loading…
6
[Qwen3.5] dedup position ids
#2102 opened Apr 27, 2026 by daijh Contributor Loading…
11
Qwen3.5: drop fp32 cast around RMSNorm in builder
#2101 opened Apr 27, 2026 by xiaofeihan1 Contributor Loading…
7
Enable graph capture for WebGPU models and DML continuous decoding tests
#2099 opened Apr 24, 2026 by qjia7 Contributor Loading…
10
Add AMDGPU execution provider support
#2093 opened Apr 20, 2026 by aditya-dl Loading…
11
Gemma 4 builder scaffolding (Part 1 of N, #2062)
#2088 opened Apr 16, 2026 by jaburges • Draft
3
WIP: TurboQuant for ORT WebGPU
#2084 opened Apr 14, 2026 by sushraja-msft Contributor • Draft
6
[Don't merge it] Fix quark quantize weight loading for Qwen3-VL-4B text model
#2082 opened Apr 13, 2026 by Tianping-amd Loading…
1
[Draft] First chunk handling and drop count for Nemotron
#2081 opened Apr 11, 2026 by jiafatom Contributor • Draft
12
extend modelbuilder to build Olmo3, SmolLM3 and other models
#2078 opened Apr 10, 2026 by xadupre Member Loading…
170
Enable CUDA graph capture for CUDA EP to improve decode throughput
#2070 opened Apr 7, 2026 by apsonawane Contributor Loading…
17
Fix CUDA build with MSVC by enabling /Zc:preprocessor for nvcc host compilation on VS 16.5 or greater
#2054 opened Apr 1, 2026 by nsubaru Loading…
2
Fix: Win32 build failure when paths contain spaces
#2053 opened Apr 1, 2026 by nsubaru Loading…
3
ProTip! Adding no:label will show everything without a label.