microsoft/onnxruntime-genai v0.11.0

v0.11.0

Repository: microsoft/onnxruntime-genai

Tag: v0.11.0

Published: 2025-11-14T02:51:44Z

Prerelease: no

Release notes:

What's Changed

• ADO - Update WinML build pipeline by @chrisdMSFT in https://github.com/microsoft/onnxruntime-genai/pull/1768
• Fix CMakeLists.txt auto-detection of library directory by @anujj in https://github.com/microsoft/onnxruntime-genai/pull/1774
• Fix new/delete override and Enable cuda kernel test in Windows by @tianleiwu in https://github.com/microsoft/onnxruntime-genai/pull/1772
• Use abbreviation for TensorRT RTX EP by @kunal-vaishnavi in https://github.com/microsoft/onnxruntime-genai/pull/1763
• Add trust remote code option to model builder by @kunal-vaishnavi in https://github.com/microsoft/onnxruntime-genai/pull/1766
• Support block-wise quant in qmoe op by @apsonawane in https://github.com/microsoft/onnxruntime-genai/pull/1746
• Change the status for TRT-RTX EP by @gaugarg-nv in https://github.com/microsoft/onnxruntime-genai/pull/1780
• Cherry-Pick changes from rel 0.10.0 back to main. by @chrisdMSFT in https://github.com/microsoft/onnxruntime-genai/pull/1782
• Fix /CETCOMPAT Usage for Cross-Compiling by @sayanshaw24 in https://github.com/microsoft/onnxruntime-genai/pull/1779
• Provide distributed version of improved TopK kernel by @hariharans29 in https://github.com/microsoft/onnxruntime-genai/pull/1710
• [TRT-RTX] Disable KV cache re-computation for Phi models by @gaugarg-nv in https://github.com/microsoft/onnxruntime-genai/pull/1787
• [CUDA] Add high-performance Top-K kernels and online benchmarking by @tianleiwu in https://github.com/microsoft/onnxruntime-genai/pull/1748
• Change shared indices array type from float to int by @hariharans29 in https://github.com/microsoft/onnxruntime-genai/pull/1789
• Enable bfloat16 multi-modal models by @kunal-vaishnavi in https://github.com/microsoft/onnxruntime-genai/pull/1786
• Disable lmhead while prompt processing by @qti-ashimaj in https://github.com/microsoft/onnxruntime-genai/pull/1762
• Introduce support for dynamic batching by @baijumeswani in https://github.com/microsoft/onnxruntime-genai/pull/1662
• Generate pyd type info by @chemwolf6922 in https://github.com/microsoft/onnxruntime-genai/pull/1742
• Add trt-rtx c packages in c example by @anujj in https://github.com/microsoft/onnxruntime-genai/pull/1794
• [CUDA] Fix build with CUDA >= 12.9 by @tianleiwu in https://github.com/microsoft/onnxruntime-genai/pull/1802
• [CUDA] topk kernels v2 by @tianleiwu in https://github.com/microsoft/onnxruntime-genai/pull/1798
• Add prefill Chunking Support for NvTensorRtRtx and Cuda Providers by @anujj in https://github.com/microsoft/onnxruntime-genai/pull/1765
• Add TRT-RTX EP support, keep NvTensorRtRtx as user facing name, and force QDQ by @anujj in https://github.com/microsoft/onnxruntime-genai/pull/1791
• [CUDA] Add static assert to suppress windows build warnings by @tianleiwu in https://github.com/microsoft/onnxruntime-genai/pull/1804
• Revert "Generate pyd type info" by @baijumeswani in https://github.com/microsoft/onnxruntime-genai/pull/1805
• [QNN] Support continuous decoding by @baijumeswani in https://github.com/microsoft/onnxruntime-genai/pull/1808
• ADO Pipeline - nuget_winml_package_reference_version is configured at build time. by @chrisdMSFT in https://github.com/microsoft/onnxruntime-genai/pull/1811
• Update version to 0.11.0-dev by @baijumeswani in https://github.com/microsoft/onnxruntime-genai/pull/1815
• Add Support For Tokenizer Options by @sayanshaw24 in https://github.com/microsoft/onnxruntime-genai/pull/1785
• Fix exit call in README example by @justinchuby in https://github.com/microsoft/onnxruntime-genai/pull/1823
• Add tokenizer APIs for accessing important ids by @kunal-vaishnavi in https://github.com/microsoft/onnxruntime-genai/pull/1822
• Use correct classes for config-only usage in model builder by @kunal-vaishnavi in https://github.com/microsoft/onnxruntime-genai/pull/1828
• Fix packaging pipeline by @baijumeswani in https://github.com/microsoft/onnxruntime-genai/pull/1829
• Add missing tokenizer methods in java by @baijumeswani in https://github.com/microsoft/onnxruntime-genai/pull/1833
• Add run options to ONNX Runtime GenAI by @kunal-vaishnavi in https://github.com/microsoft/onnxruntime-genai/pull/1795
• Avoid Processing EOS Token During Continuous Decoding by @baijumeswani in https://github.com/microsoft/onnxruntime-genai/pull/1814
• Fix nuget packaging pipeline for dev builds by @baijumeswani in https://github.com/microsoft/onnxruntime-genai/pull/1837
• Add tool normalization for tool calling by @kunal-vaishnavi in https://github.com/microsoft/onnxruntime-genai/pull/1838
• Refactor past_present_share_buffer logic into reusable function by @anujj in https://github.com/microsoft/onnxruntime-genai/pull/1839
• Fix nuget packaging pipeline by @baijumeswani in https://github.com/microsoft/onnxruntime-genai/pull/1841
• Add enable_webgpu_graph in extra_options by @qjia7 in https://github.com/microsoft/onnxruntime-genai/pull/1788
• Update tool normalization in ORT GenAI by @kunal-vaishnavi in https://github.com/microsoft/onnxruntime-genai/pull/1842
• Support RotaryEmbedding in GQA for webgpu ep by @xiaofeihan1 in https://github.com/microsoft/onnxruntime-genai/pull/1847
• Enable guidance ff tokens for faster inference by @JC1DA in https://github.com/microsoft/onnxruntime-genai/pull/1803
• Support pre-registered plug-in cuda execution provider library by @baijumeswani in https://github.com/microsoft/onnxruntime-genai/pull/1850
• ADO: Update pipeline to publish onnxruntime-genai. for relwithdebinfo builds. by @chrisdMSFT in https://github.com/microsoft/onnxruntime-genai/pull/1855
• Layer-wise KV Cache Allocation for Models with Alternating Attention Patterns by @anujj in https://github.com/microsoft/onnxruntime-genai/pull/1832
• Mpasumarthi/nvtrt test suite by @mpasumarthi-git in https://github.com/microsoft/onnxruntime-genai/pull/1756
• bugfix: fix a memory issue in Whisper by @fs-eire in https://github.com/microsoft/onnxruntime-genai/pull/1859
• Add disable cuda graph when num_beams > 1 and fix set_provider_option bug by @anujj in https://github.com/microsoft/onnxruntime-genai/pull/1846
• Mixed precision export support for gptq quantized model by @rM-planet in https://github.com/microsoft/onnxruntime-genai/pull/1853
• Enable If Node Support for TRT-RTX in Phi-3.5/Phi-4 LongRoPE Models by @anujj in https://github.com/microsoft/onnxruntime-genai/pull/1851
• Fix handling...