dependabot[bot] opened a new pull request, #1202: URL: https://github.com/apache/opennlp/pull/1202
Bumps `onnxruntime.version` from 1.27.0 to 1.28.0. Updates `com.microsoft.onnxruntime:onnxruntime` from 1.27.0 to 1.28.0 <details> <summary>Release notes</summary> <p><em>Sourced from <a href="https://github.com/microsoft/onnxruntime/releases";>com.microsoft.onnxruntime:onnxruntime's releases</a>.</em></p> <blockquote> <h2>ONNX Runtime v1.28.0</h2> <h2>Announcements & Breaking Changes</h2> <ul> <li>Upgraded to <strong>ONNX 1.22.0</strong> and protobuf 6.33.5 (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28754";>#28754</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29606";>#29606</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/28967";>#28967</a>). Graph optimizer opset version checks were updated accordingly (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28966";>#28966</a>).</li> <li><strong>cuDNN and cuFFT are now optional at runtime</strong> for the CUDA EP, and <code>nvrtc</code> is no longer linked, which significantly reduces the required CUDA redistributable footprint (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29252";>#29252</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29808";>#29808</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29705";>#29705</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29620";>#29620</a>).</li> <li>An <strong>experimental C/C++ API surface</strong> was introduced. <code>OrtModelPackageApi</code> now lives in the experimental C API and may change in future releases (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28746";>#28746</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29142";>#29142</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/28990";>#28990</a>).</li> <li><strong>Deprecated / removed:</strong> <ul> <li>SkipLayerNorm strict mode is deprecated (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29388";>#29388</a>).</li> <li>The TensorRT fused causal attention kernels were removed from the CUDA EP (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29143";>#29143</a>).</li> <li>The dynamic WGSL generator (duktape/Node) path was removed in favor of the Python <code>wgsl-gen</code> implementation (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29141";>#29141</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/28355";>#28355</a>).</li> <li><code>CUDA_QUANT_PREPROCESS</code> is off by default (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29687";>#29687</a>).</li> </ul> </li> <li>NPM packages are now published from the CUDA 13 pipeline (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28773";>#28773</a>).</li> <li>The CUDA 12.8 package architecture list was refreshed for this release (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29711";>#29711</a>).</li> </ul> <h2>Security Fixes</h2> <h3>Memory safety & input validation</h3> <ul> <li>Hardened the ORT FlatBuffer model loader against malformed buffers, and removed now-redundant table offset validation (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28186";>#28186</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29068";>#29068</a>)</li> <li>Fixed type confusion in raw-pointer <code>bind_input</code> causing an out-of-bounds write (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28839";>#28839</a>)</li> <li>Fixed out-of-bounds pointer in <code>TensorAt</code> for sub-byte packed types (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28973";>#28973</a>)</li> <li>Fixed arbitrary memory read, out-of-bounds dereference, and other OOB accesses in kernels (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28991";>#28991</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29011";>#29011</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29012";>#29012</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29014";>#29014</a>)</li> <li>Validated <code>Col2Im</code> inputs to prevent heap over-read (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28706";>#28706</a>)</li> <li>Hardened <code>CropAndResize</code> against malformed <code>crop_size</code> tensors (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28766";>#28766</a>)</li> <li>Validated <code>BeamSearch</code> <code>vocab_size</code> against logits width (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28774";>#28774</a>)</li> <li>Fixed bounds in <code>WhisperDecoderSubgraph::CreateInitialFeeds</code> (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29239";>#29239</a>)</li> <li>Validated <code>SparseAttention</code> CSR indices/key lengths and rejected zero-dimension <code>block_row_indices</code> (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29015";>#29015</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29242";>#29242</a>)</li> <li>Clamped derived sequence lengths and KV-cache index in CUDA GroupQueryAttention, and fixed a CPU GQA out-of-bounds read in the past-KV buffer (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29240";>#29240</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29447";>#29447</a>)</li> <li>Clamped 1D attention <code>mask_index</code> to valid bounds (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29449";>#29449</a>)</li> <li>Validated <code>MaxpoolWithMask</code> kernel rank against input spatial rank (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29253";>#29253</a>)</li> <li>Rejected CUDA BERT <code>EmbedLayerNorm</code>/<code>SkipLayerNorm</code> shapes exceeding 32-bit output indexing (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29264";>#29264</a>)</li> <li>Fixed the optional-output guard in <code>DecoderAttention</code>/<code>MultiHeadAttention</code> shape inference and negative-axis handling in <code>ExpandDims</code> shape inference (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29268";>#29268</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29448";>#29448</a>)</li> <li>Fixed <code>TreeEnsemble</code> target id validation and added input validation to <code>LinearClassifier</code> (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29293";>#29293</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29060";>#29060</a>)</li> <li>Fixed <code>DynamicQuantizeLSTM</code> zero-point/scale validation typos (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29462";>#29462</a>)</li> <li>Handled non-trivially-copyable types in <code>Loop</code>/<code>Scan</code> output concatenation (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29397";>#29397</a>)</li> <li>Normalized bool tensor <code>raw_data</code> to <code>{0, 1}</code> on unpack (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29238";>#29238</a>)</li> <li>Addressed hardening gaps in <code>Resize</code>, <code>PadFusion</code>, and LoRA handling (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28779";>#28779</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/28780";>#28780</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/28801";>#28801</a>)</li> <li>Fixed unbounded lifetime on <code>WithOutputTensor</code> in the Rust bindings (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29251";>#29251</a>)</li> </ul> <h3>Integer overflow & allocation size</h3> <ul> <li>Guarded <code>MlasConvPrepare</code> working-buffer products and <code>ConvTranspose</code> pad computation with SafeInt (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29444";>#29444</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29446";>#29446</a>)</li> <li>Fixed signed-int overflow in <code>SamplingState::Init</code> that could cause a heap buffer overflow (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29443";>#29443</a>)</li> <li>Hardened QMoE against integer overflow and partial K tiles (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29067";>#29067</a>)</li> <li>Validated <code>B</code>/scales/zero-points shape in <code>MatMulNBits::PrePack</code> (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29445";>#29445</a>)</li> <li>Pre-checked <code>ConstantOfShape</code> output size against the input initializer before constant folding (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28751";>#28751</a>)</li> <li>Fixed integer overflow in RKNPU implicit bias allocation (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29249";>#29249</a>)</li> <li>Fixed WebGPU out-of-bounds reads in <code>Pad</code> (int64/int32 truncation), <code>Slice</code>, and <code>GatherBlockQuantized</code> (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28721";>#28721</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/28704";>#28704</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/28718";>#28718</a>)</li> </ul> <h3>Supply chain & tooling</h3> <!-- raw HTML omitted --> </blockquote> <p>... (truncated)</p> </details> <details> <summary>Commits</summary> <ul> <li><a href="https://github.com/microsoft/onnxruntime/commit/da9b5e364c465de65c49d91e696cd6485270757f";><code>da9b5e3</code></a> Fix Windows zip artifact to include .inc header files (<a href="https://redirect.github.com/microsoft/onnxruntime/issues/29874";>#29874</a>)</li> <li><a href="https://github.com/microsoft/onnxruntime/commit/45de2a8b06d62989b3ab55ba7dc58a27ca83f9fc";><code>45de2a8</code></a> Fix NuGet packaging to include .inc files alongside .h headers (<a href="https://redirect.github.com/microsoft/onnxruntime/issues/29868";>#29868</a>)</li> <li><a href="https://github.com/microsoft/onnxruntime/commit/0368187f8403b9050f8dbc55b16966883bc93fb9";><code>0368187</code></a> ORT 1.28.0 release cherry-pick round 2 (<a href="https://redirect.github.com/microsoft/onnxruntime/issues/29821";>#29821</a>)</li> <li><a href="https://github.com/microsoft/onnxruntime/commit/e1bbb649d53189fd995f3975c8048b7067104eea";><code>e1bbb64</code></a> ORT 1.28.0 release cherry-pick round 1 (<a href="https://redirect.github.com/microsoft/onnxruntime/issues/29771";>#29771</a>)</li> <li><a href="https://github.com/microsoft/onnxruntime/commit/a06675e9a412692dae5805d1acdb4bd72bbd4f84";><code>a06675e</code></a> Fix web e2e (npm/vite) and Python DML CI pipelines (<a href="https://redirect.github.com/microsoft/onnxruntime/issues/29609";>#29609</a>)</li> <li><a href="https://github.com/microsoft/onnxruntime/commit/c4f19617bca2567ebd8a35a750c14158f3fa4a40";><code>c4f1961</code></a> Bump onnx to 1.22.0 and protobuf to 6.33.5 to fix security CVEs (<a href="https://redirect.github.com/microsoft/onnxruntime/issues/29606";>#29606</a>)</li> <li><a href="https://github.com/microsoft/onnxruntime/commit/e0ad071825027e478c701296afc64a04aa42e0ad";><code>e0ad071</code></a> [CUDA] Enable native SM90, block_size=32, and fused bias for fpA_intB MatMulN...</li> <li><a href="https://github.com/microsoft/onnxruntime/commit/a1fc71e707706346b97682dbe1acee00219a7e57";><code>a1fc71e</code></a> [CUDA] Fix QMoE profiler cross-stream race and CUDA-graph-capture safety (<a href="https://redirect.github.com/microsoft/onnxruntime/issues/29";>#29</a>...</li> <li><a href="https://github.com/microsoft/onnxruntime/commit/5eb4aeebf886c7f0b101281c0866765a5d56a700";><code>5eb4aee</code></a> Fix CustomOp forward compatibility: cap version instead of rejecting (<a href="https://redirect.github.com/microsoft/onnxruntime/issues/29574";>#29574</a>)</li> <li><a href="https://github.com/microsoft/onnxruntime/commit/7a1237141a30e6795bd5269a803c23ce8734d5c0";><code>7a12371</code></a> adjusts conv kernel to use get/set by offset helpers (<a href="https://redirect.github.com/microsoft/onnxruntime/issues/29463";>#29463</a>)</li> <li>Additional commits viewable in <a href="https://github.com/microsoft/onnxruntime/compare/v1.27.0...v1.28.0";>compare view</a></li> </ul> </details> <br /> Updates `com.microsoft.onnxruntime:onnxruntime_gpu` from 1.27.0 to 1.28.0 <details> <summary>Release notes</summary> <p><em>Sourced from <a href="https://github.com/microsoft/onnxruntime/releases";>com.microsoft.onnxruntime:onnxruntime_gpu's releases</a>.</em></p> <blockquote> <h2>ONNX Runtime v1.28.0</h2> <h2>Announcements & Breaking Changes</h2> <ul> <li>Upgraded to <strong>ONNX 1.22.0</strong> and protobuf 6.33.5 (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28754";>#28754</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29606";>#29606</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/28967";>#28967</a>). Graph optimizer opset version checks were updated accordingly (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28966";>#28966</a>).</li> <li><strong>cuDNN and cuFFT are now optional at runtime</strong> for the CUDA EP, and <code>nvrtc</code> is no longer linked, which significantly reduces the required CUDA redistributable footprint (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29252";>#29252</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29808";>#29808</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29705";>#29705</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29620";>#29620</a>).</li> <li>An <strong>experimental C/C++ API surface</strong> was introduced. <code>OrtModelPackageApi</code> now lives in the experimental C API and may change in future releases (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28746";>#28746</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29142";>#29142</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/28990";>#28990</a>).</li> <li><strong>Deprecated / removed:</strong> <ul> <li>SkipLayerNorm strict mode is deprecated (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29388";>#29388</a>).</li> <li>The TensorRT fused causal attention kernels were removed from the CUDA EP (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29143";>#29143</a>).</li> <li>The dynamic WGSL generator (duktape/Node) path was removed in favor of the Python <code>wgsl-gen</code> implementation (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29141";>#29141</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/28355";>#28355</a>).</li> <li><code>CUDA_QUANT_PREPROCESS</code> is off by default (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29687";>#29687</a>).</li> </ul> </li> <li>NPM packages are now published from the CUDA 13 pipeline (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28773";>#28773</a>).</li> <li>The CUDA 12.8 package architecture list was refreshed for this release (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29711";>#29711</a>).</li> </ul> <h2>Security Fixes</h2> <h3>Memory safety & input validation</h3> <ul> <li>Hardened the ORT FlatBuffer model loader against malformed buffers, and removed now-redundant table offset validation (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28186";>#28186</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29068";>#29068</a>)</li> <li>Fixed type confusion in raw-pointer <code>bind_input</code> causing an out-of-bounds write (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28839";>#28839</a>)</li> <li>Fixed out-of-bounds pointer in <code>TensorAt</code> for sub-byte packed types (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28973";>#28973</a>)</li> <li>Fixed arbitrary memory read, out-of-bounds dereference, and other OOB accesses in kernels (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28991";>#28991</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29011";>#29011</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29012";>#29012</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29014";>#29014</a>)</li> <li>Validated <code>Col2Im</code> inputs to prevent heap over-read (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28706";>#28706</a>)</li> <li>Hardened <code>CropAndResize</code> against malformed <code>crop_size</code> tensors (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28766";>#28766</a>)</li> <li>Validated <code>BeamSearch</code> <code>vocab_size</code> against logits width (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28774";>#28774</a>)</li> <li>Fixed bounds in <code>WhisperDecoderSubgraph::CreateInitialFeeds</code> (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29239";>#29239</a>)</li> <li>Validated <code>SparseAttention</code> CSR indices/key lengths and rejected zero-dimension <code>block_row_indices</code> (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29015";>#29015</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29242";>#29242</a>)</li> <li>Clamped derived sequence lengths and KV-cache index in CUDA GroupQueryAttention, and fixed a CPU GQA out-of-bounds read in the past-KV buffer (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29240";>#29240</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29447";>#29447</a>)</li> <li>Clamped 1D attention <code>mask_index</code> to valid bounds (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29449";>#29449</a>)</li> <li>Validated <code>MaxpoolWithMask</code> kernel rank against input spatial rank (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29253";>#29253</a>)</li> <li>Rejected CUDA BERT <code>EmbedLayerNorm</code>/<code>SkipLayerNorm</code> shapes exceeding 32-bit output indexing (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29264";>#29264</a>)</li> <li>Fixed the optional-output guard in <code>DecoderAttention</code>/<code>MultiHeadAttention</code> shape inference and negative-axis handling in <code>ExpandDims</code> shape inference (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29268";>#29268</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29448";>#29448</a>)</li> <li>Fixed <code>TreeEnsemble</code> target id validation and added input validation to <code>LinearClassifier</code> (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29293";>#29293</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29060";>#29060</a>)</li> <li>Fixed <code>DynamicQuantizeLSTM</code> zero-point/scale validation typos (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29462";>#29462</a>)</li> <li>Handled non-trivially-copyable types in <code>Loop</code>/<code>Scan</code> output concatenation (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29397";>#29397</a>)</li> <li>Normalized bool tensor <code>raw_data</code> to <code>{0, 1}</code> on unpack (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29238";>#29238</a>)</li> <li>Addressed hardening gaps in <code>Resize</code>, <code>PadFusion</code>, and LoRA handling (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28779";>#28779</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/28780";>#28780</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/28801";>#28801</a>)</li> <li>Fixed unbounded lifetime on <code>WithOutputTensor</code> in the Rust bindings (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29251";>#29251</a>)</li> </ul> <h3>Integer overflow & allocation size</h3> <ul> <li>Guarded <code>MlasConvPrepare</code> working-buffer products and <code>ConvTranspose</code> pad computation with SafeInt (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29444";>#29444</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/29446";>#29446</a>)</li> <li>Fixed signed-int overflow in <code>SamplingState::Init</code> that could cause a heap buffer overflow (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29443";>#29443</a>)</li> <li>Hardened QMoE against integer overflow and partial K tiles (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29067";>#29067</a>)</li> <li>Validated <code>B</code>/scales/zero-points shape in <code>MatMulNBits::PrePack</code> (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29445";>#29445</a>)</li> <li>Pre-checked <code>ConstantOfShape</code> output size against the input initializer before constant folding (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28751";>#28751</a>)</li> <li>Fixed integer overflow in RKNPU implicit bias allocation (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/29249";>#29249</a>)</li> <li>Fixed WebGPU out-of-bounds reads in <code>Pad</code> (int64/int32 truncation), <code>Slice</code>, and <code>GatherBlockQuantized</code> (<a href="https://redirect.github.com/microsoft/onnxruntime/pull/28721";>#28721</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/28704";>#28704</a>, <a href="https://redirect.github.com/microsoft/onnxruntime/pull/28718";>#28718</a>)</li> </ul> <h3>Supply chain & tooling</h3> <!-- raw HTML omitted --> </blockquote> <p>... (truncated)</p> </details> <details> <summary>Commits</summary> <ul> <li><a href="https://github.com/microsoft/onnxruntime/commit/da9b5e364c465de65c49d91e696cd6485270757f";><code>da9b5e3</code></a> Fix Windows zip artifact to include .inc header files (<a href="https://redirect.github.com/microsoft/onnxruntime/issues/29874";>#29874</a>)</li> <li><a href="https://github.com/microsoft/onnxruntime/commit/45de2a8b06d62989b3ab55ba7dc58a27ca83f9fc";><code>45de2a8</code></a> Fix NuGet packaging to include .inc files alongside .h headers (<a href="https://redirect.github.com/microsoft/onnxruntime/issues/29868";>#29868</a>)</li> <li><a href="https://github.com/microsoft/onnxruntime/commit/0368187f8403b9050f8dbc55b16966883bc93fb9";><code>0368187</code></a> ORT 1.28.0 release cherry-pick round 2 (<a href="https://redirect.github.com/microsoft/onnxruntime/issues/29821";>#29821</a>)</li> <li><a href="https://github.com/microsoft/onnxruntime/commit/e1bbb649d53189fd995f3975c8048b7067104eea";><code>e1bbb64</code></a> ORT 1.28.0 release cherry-pick round 1 (<a href="https://redirect.github.com/microsoft/onnxruntime/issues/29771";>#29771</a>)</li> <li><a href="https://github.com/microsoft/onnxruntime/commit/a06675e9a412692dae5805d1acdb4bd72bbd4f84";><code>a06675e</code></a> Fix web e2e (npm/vite) and Python DML CI pipelines (<a href="https://redirect.github.com/microsoft/onnxruntime/issues/29609";>#29609</a>)</li> <li><a href="https://github.com/microsoft/onnxruntime/commit/c4f19617bca2567ebd8a35a750c14158f3fa4a40";><code>c4f1961</code></a> Bump onnx to 1.22.0 and protobuf to 6.33.5 to fix security CVEs (<a href="https://redirect.github.com/microsoft/onnxruntime/issues/29606";>#29606</a>)</li> <li><a href="https://github.com/microsoft/onnxruntime/commit/e0ad071825027e478c701296afc64a04aa42e0ad";><code>e0ad071</code></a> [CUDA] Enable native SM90, block_size=32, and fused bias for fpA_intB MatMulN...</li> <li><a href="https://github.com/microsoft/onnxruntime/commit/a1fc71e707706346b97682dbe1acee00219a7e57";><code>a1fc71e</code></a> [CUDA] Fix QMoE profiler cross-stream race and CUDA-graph-capture safety (<a href="https://redirect.github.com/microsoft/onnxruntime/issues/29";>#29</a>...</li> <li><a href="https://github.com/microsoft/onnxruntime/commit/5eb4aeebf886c7f0b101281c0866765a5d56a700";><code>5eb4aee</code></a> Fix CustomOp forward compatibility: cap version instead of rejecting (<a href="https://redirect.github.com/microsoft/onnxruntime/issues/29574";>#29574</a>)</li> <li><a href="https://github.com/microsoft/onnxruntime/commit/7a1237141a30e6795bd5269a803c23ce8734d5c0";><code>7a12371</code></a> adjusts conv kernel to use get/set by offset helpers (<a href="https://redirect.github.com/microsoft/onnxruntime/issues/29463";>#29463</a>)</li> <li>Additional commits viewable in <a href="https://github.com/microsoft/onnxruntime/compare/v1.27.0...v1.28.0";>compare view</a></li> </ul> </details> <br /> Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting `@dependabot rebase`. [//]: # (dependabot-automerge-start) [//]: # (dependabot-automerge-end) --- <details> <summary>Dependabot commands and options</summary> <br /> You can trigger Dependabot actions by commenting on this PR: - `@dependabot rebase` will rebase this PR - `@dependabot recreate` will recreate this PR, overwriting any edits that have been made to it - `@dependabot show <dependency name> ignore conditions` will show all of the ignore conditions of the specified dependency - `@dependabot ignore this major version` will close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself) - `@dependabot ignore this minor version` will close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself) - `@dependabot ignore this dependency` will close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself) </details> -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
