Pull Requests microsoft/onnxruntime

Add transformer optimizer JSON report

#28445 opened 2026-05-10 03:10 by KOKOSde

fix: use relocatable path for safeint_interface when using vcpkg

#28444 opened 2026-05-09 18:46 by 2ssh

Fix f16 overflow in SkipLayerNormalization CUDA kernel

#28442 opened 2026-05-09 16:24 by justinchuby

Fix CPU LayerNormalization NaNs on large constant inputs

#28440 opened 2026-05-09 15:59 by Copilot

Bump @babel/plugin-transform-modules-systemjs from 7.25.0 to 7.29.4 in /js/react_native/e2e dependencies javascript

#28439 opened 2026-05-09 12:47 by dependabot[bot]

webgpu: support vec2 type for gemm-subgroup

#28436 opened 2026-05-09 07:49 by xhcao

Add CPU Cast op support for Float8E8M0 with round_mode

#28435 opened 2026-05-09 04:48 by tianleiwu

[WebGPU] Fix numerical stability issue in QMoE op

#28434 opened 2026-05-09 02:52 by xenova

Skip file path validation in CreateEpContextModel when the output is not a file

#28431 opened 2026-05-09 01:30 by shiyi9801

Add unit test for PR 28396

#28430 opened 2026-05-09 01:05 by adrianlizarraga

Validate NodeArg creation

#28429 opened 2026-05-09 00:15 by yuslepukhin

Address tokenizer shortcomings

#28428 opened 2026-05-08 23:23 by yuslepukhin

[WebGPU] Fix SkipSimplifiedLayerNormalization bias

#28427 opened 2026-05-08 22:44 by xenova

Fix CSE crash on string tensor attributes

#28422 opened 2026-05-08 21:54 by rajatmonga

Use weights_only=True for remaining torch.load() calls

#28421 opened 2026-05-08 21:47 by adrianlizarraga

Add CPU DynamicQuantMatMulFp8 contrib op with MLAS FP8 fallback

#28416 opened 2026-05-08 15:37 by melkap01-Arm

docs(quantization): clarify CPU parameter recommendations

#28415 opened 2026-05-08 11:30 by Rishi-Dave

fix: preserve Q/DQ around Resize when mode is linear or cubic

#28414 opened 2026-05-08 11:17 by Rishi-Dave

[webgpu] Optimize LinearAttention Op with subgroup ep:WebGPU

#28412 opened 2026-05-08 11:02 by daijh

Add RVV (RISC-V Vector Extension) optimized convolution and pooling kernels for the NCHWc blocked format in MLAS

#28411 opened 2026-05-08 10:06 by velonica0

Enable QuickGeluFusion on WebGPU EP and fix fp16 shader

#28410 opened 2026-05-08 08:52 by xiaofeihan1

[CUDA] PagedAttention: use exact max_query_len on FA path

#28409 opened 2026-05-08 04:58 by elwhyjay

Validate sparse tensor external file paths

#28408 opened 2026-05-08 01:21 by adrianlizarraga

Use abseil for readable POSIX stack traces in debug builds

#28405 opened 2026-05-07 23:07 by tianleiwu

Address Resize kernel shortcomings

#28402 opened 2026-05-07 22:29 by yuslepukhin

Bump diffusers from 0.28.0 to 0.38.0 in /onnxruntime/python/tools/transformers/models/stable_diffusion/requirements dependencies python

#28401 opened 2026-05-07 18:33 by dependabot[bot]

Fix Inverse kernel rank underflow before indexing trailing dims

#28400 opened 2026-05-07 17:56 by Copilot

Fix out-of-bounds read in CropBase: validate scale attribute length

#28399 opened 2026-05-07 17:54 by Copilot

Add fix for fastmath corner case with failing tests via sbgemm_neon_kernel

#28394 opened 2026-05-07 10:09 by JonathanC-ARM

[EP Context] Use ToPathString() for ep_context_file_path to correctly handle UTF-8 Unicode paths on Windows

#28390 opened 2026-05-07 04:33 by jnagi-intel