GitHub / microsoft/DirectML / commits
DirectML is a high-performance, hardware-accelerated DirectX 12 library for machine learning. DirectML provides GPU acceleration for common machine learning tasks across a broad range of supported hardware and drivers, including all DirectX 12-capable GPUs from vendors such as AMD, Intel, NVIDIA, and Qualcomm.
| SHA | Message | Author | Date | Stats |
|---|---|---|---|---|
| 8700779f | Harden the buffer size and strides calculation against numerical overflow (#734) |
Chai Chaoweeraprasit <w****5@h****m>
Committed by: GitHub <n****y@g****m> |
4 months ago | |
| a9614e82 |
Update Releases.md removing TBD's (#711)
Co-authored-by: Justin Stoecker <j****k@m****m>, Justin Stoecker <j****k@m****m> |
Dwayne Robinson <d****r@m****m>
Committed by: GitHub <n****y@g****m> |
about 1 year ago | |
| d2c4f114 | Update README to indicate DML is in maintenance mode (#710) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
about 1 year ago | |
| 674540ff | Print hex as hex (#707) |
Dwayne Robinson <d****r@m****m>
Committed by: GitHub <n****y@g****m> |
about 1 year ago | |
| ae698971 | DxDispatch add print verbose use fmt instead of snprintf (#706) |
Dwayne Robinson <d****r@m****m>
Committed by: GitHub <n****y@g****m> |
about 1 year ago | |
| 17c28291 | DxDispatch add print verbose option including exact bit patterns via hex (#705) |
Dwayne Robinson <d****r@m****m>
Committed by: GitHub <n****y@g****m> |
about 1 year ago | |
| 86647bd4 | Basic image I/O support for dxdispatch (#695) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
over 1 year ago | |
| a2b2dd2d | DirectML 1.15.4 is released (#661) |
Indy Zhu <i****t@u****m>
Committed by: GitHub <n****y@g****m> |
almost 2 years ago | |
| 8fd4bc3e | Update Releases.md for 1.15.3 (#658) |
Indy Zhu <i****t@u****m>
Committed by: GitHub <n****y@g****m> |
almost 2 years ago | |
| 94fafe58 |
For NPU Sample, add flexibility in device creation options and fix Generic ML...
Co-authored-by: Zu Shi <z****i@m****m> |
Zu-shi <z****i@g****m>
Committed by: GitHub <n****y@g****m> |
almost 2 years ago | |
| e6e55974 | Use DML_TENSOR_DATA_TYPE_UINT8 for ONNX_TENSOR_ELEMENT_DATA_TYPE_BOOL (#651) |
Indy Zhu <i****t@u****m>
Committed by: GitHub <n****y@g****m> |
almost 2 years ago | |
| 7138c1c0 | Update Readme (#649) |
Sumit Agarwal <s****0@g****m>
Committed by: GitHub <n****y@g****m> |
almost 2 years ago | |
| 2b952c3f | Update ort (#641) |
Sumit Agarwal <s****0@g****m>
Committed by: GitHub <n****y@g****m> |
about 2 years ago | |
| 53e1a0e9 | Update SHA256 for DML1.15.2 in ESRGAN sample (#639) |
Sumit Agarwal <s****0@g****m>
Committed by: GitHub <n****y@g****m> |
about 2 years ago | |
| b8730648 | Update Samples and Release.md (#636) |
Sumit Agarwal <s****0@g****m>
Committed by: GitHub <n****y@g****m> |
about 2 years ago | |
| f9d47d00 | Per Ashrit Shetty: Update readme for error message (#637) |
Indy Zhu <i****t@u****m>
Committed by: GitHub <n****y@g****m> |
about 2 years ago | |
| 1c6427aa | compute queue for core/ml adapters only (#635) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
about 2 years ago | |
| 2ac1dd0b | C++ sample for ESRGAN (#634) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
about 2 years ago | |
| ff5c0908 | Update Release Notes and Samples for DML1.15.1 (#633) |
Sumit Agarwal <s****0@g****m>
Committed by: GitHub <n****y@g****m> |
about 2 years ago | |
| 32e3d129 |
Add DmlSerialized Dispatchable to DxDispatch (#601)
Co-authored-by: Bereket Nigussie <t****e@m****m>, Sumit Agarwal <s****l@m****m> |
Bereket Nigussie <a****9@g****m>
Committed by: GitHub <n****y@g****m> |
about 2 years ago | |
| 6645ec13 | DxDispatch: remove half dependency from sourceforge (#631) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
about 2 years ago | |
| 07377a60 | update whisper sample (#629) |
Xiang Zhang <x****z@m****m>
Committed by: GitHub <n****y@g****m> |
about 2 years ago | |
| a87891eb | Add DxDispatch PIX timing capture command line support on windows (#617) |
Cole Brooking <1****T@u****m>
Committed by: GitHub <n****y@g****m> |
about 2 years ago | |
| 1d738f7a |
Update Torch-DirectML samples and docs for Torch-DirectML 2.3.0 (#610)
Co-authored-by: Dwayne Robinson <d****r@m****m>, Dwayne Robinson <d****r@m****m>, Dwayne Robinson <d****r@m****m>, Dwayne Robinson <d****r@m****m>, Dwayne Robinson <d****r@m****m>, Dwayne Robinson <d****r@m****m>, Sheil Kumar <s****k@m****m>, Dwayne Robinson <d****r@m****m> |
Sheil Kumar <s****7@g****m>
Committed by: GitHub <n****y@g****m> |
about 2 years ago | |
| 61a1a508 |
User/chrila/enable graph compile (#606)
Co-authored-by: Christian Larson <2****T@u****m> |
Christian Larson <c****T@u****m>
Committed by: GitHub <n****y@g****m> |
about 2 years ago | |
| 72ad224f |
Update samples, docs and release notes for DirectML 1.15.0 (#605)
Co-authored-by: Sheil Kumar <s****k@m****m> |
Sheil Kumar <s****7@g****m>
Committed by: GitHub <n****y@g****m> |
about 2 years ago | |
| 46f3e141 | Add code for creating Generic ML devices to NPU starter sample (#599) |
Zu-shi <a****n@g****m>
Committed by: GitHub <n****y@g****m> |
about 2 years ago | |
| 372a6228 |
Update docs for torch-directml 0.2.2 (#593)
Co-authored-by: Sheil Kumar <s****k@m****m>, Dwayne Robinson <d****r@m****m> |
Sheil Kumar <s****7@g****m>
Committed by: GitHub <n****y@g****m> |
about 2 years ago | |
| 4d65cad0 | Add MVN2 and Dequantize support (#591) |
Patrice Vignola <v****e@g****m>
Committed by: GitHub <n****y@g****m> |
about 2 years ago | |
| e3a75c1f | Upgrade dxdispatch to ORT 1.18 (#584) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| 3e69d6a4 |
Samples and docs updates for Torch-DirectML and add new LLM samples. (#585)
Co-authored-by: Sheil Kumar <s****k@m****m> |
Sheil Kumar <s****7@g****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| 2aaba74b | Minor fixes for formatting and present seperator (#583) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| cc9606b1 | Support for --present_separator, enabling some external tools perceive infere... |
NvVlad <9****d@u****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| 1c3ade67 | Fix debug callback not showing output (#581) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| 6a5919ce | Use custom heaps by default for cache-coherent UMA devices (#580) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| 5897d34a | Change feature level of Resample3 to 6.4 (#577) |
Patrice Vignola <v****e@g****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| d3918ea6 | Add SoftMax and GELU activation functions to DMLX (#573) |
Patrice Vignola <v****e@g****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| 0bd9f4f0 |
Enable generic feature level devices in DXDispatch (#571)
Co-authored-by: Justin Stoecker <j****k@m****m>, Justin Stoecker <j****k@m****m> |
Jeff Bloomfield <3****o@u****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| c6fe6945 | DxDispatch update to FL6.2 and a few bug fixes (#576) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| 7be19e21 |
Update Microsoft.ML.OnnxRuntime.DirectML version to 1.15.0->1.17.1 (#574)
Co-authored-by: Christian Larson <2****T@u****m> |
Christian Larson <c****T@u****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| 68d3ab6c | Add int4 support to DMLX (#572) |
Patrice Vignola <v****e@g****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| 224ceb74 | Fix if spacing (#562) |
Dwayne Robinson <d****r@m****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| 12abc68e | Typo "Pytorch"→"PyTorch" (#514) |
Hiroshi Yoshioka <4****8@u****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| 4f9df568 | fix bug (#565) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| d1299e61 | Support constant data nodes in DirectMLX (#564) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| da3abe62 |
Add sample showing how to enumerate NPU devices (#552)
Co-authored-by: Sheil Kumar <s****k@m****m> |
Sheil Kumar <s****7@g****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| 287cd032 |
Update compile and link commands to address binskim (#561)
Co-authored-by: Christian Larson (from Dev Box) <2****T@u****m> |
Christian Larson <c****T@u****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| 586759ee |
Update DxDispatch to fix Bin Skim errors (#557)
Co-authored-by: Christian Larson (from Dev Box) <2****T@u****m> |
Christian Larson <c****T@u****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| aa9982c1 |
Update Code to work with agility SDK, added Dred, update agilitySdk (#548)
Co-authored-by: Christian Larson <2****T@u****m> |
Christian Larson <c****T@u****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| 81ec7d7b | Fix links for TF and PT with DirectML (#549) |
Hayden Barnes <h****s@h****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| 88ecc6c1 | Update release notes for DML 1.13.1 (#546) |
tbqh <1****h@u****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| 293dc808 | Update samples and release notes for DML 1.13 (#534) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| 63ef3a6b |
[DirectMLX] Fixed the local variable initialized but not referenced issue. (#...
Signed-off-by: Austin Hu <a****u@i****m> |
Austin Hu <a****u@i****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| 2534446a | DxDispatch profiling improvements (#530) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| 99daf0f6 | [directmlx] Move resample3 to feature level 6.3 (#532) |
Linnea May <l****e@g****m>
Committed by: GitHub <n****y@g****m> |
over 2 years ago | |
| 275d1d1f |
Add support for 0 size tensor (#527)
Co-authored-by: Christian Larson <2****T@u****m> |
Christian Larson <c****T@u****m>
Committed by: GitHub <n****y@g****m> |
almost 3 years ago | |
| 294bfc8c | Fix cpack for dxdispatch (#523) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
almost 3 years ago | |
| f2d46f94 |
Add export library to install command (#522)
Co-authored-by: Christian Larson <2****T@u****m> |
Christian Larson <c****T@u****m>
Committed by: GitHub <n****y@g****m> |
almost 3 years ago | |
| 270a4005 |
Add dll interface to DxDispatch for test scripting (#519)
Co-authored-by: Christian Larson <2****T@u****m> |
Christian Larson <c****T@u****m>
Committed by: GitHub <n****y@g****m> |
almost 3 years ago | |
| 100465ec |
[DirectMLX] fix build issue in Resample when TARGET_VERSION < 0x6200 (#520)
Co-authored-by: Linnea May <l****y@m****m>, Dwayne Robinson <f****r@h****m> |
Linnea May <l****e@g****m>
Committed by: GitHub <n****y@g****m> |
almost 3 years ago | |
| ac9f01c2 |
Add deferred binding and core adapter (#515)
Co-authored-by: Christian Larson <2****T@u****m> |
Christian Larson <c****T@u****m>
Committed by: GitHub <n****y@g****m> |
almost 3 years ago | |
| ab04d68c |
update dml::Resample to include RESAMPLE3 api (#517)
Co-authored-by: Linnea May <l****y@m****m> |
Linnea May <l****e@g****m>
Committed by: GitHub <n****y@g****m> |
almost 3 years ago | |
| 4069e9f0 | Add support for 3d convolution (#507) |
Damien Menigaux <d****x@p****u>
Committed by: GitHub <n****y@g****m> |
almost 3 years ago | |
| 02fe4b55 | Update dxdispatch pipeline to produce nugets for internal testing (#505) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
almost 3 years ago | |
| ce57292b | Add profiling option to DxDispatch (#499) |
Patrice Vignola <v****e@g****m>
Committed by: GitHub <n****y@g****m> |
about 3 years ago | |
| 3b843ae5 |
Update readme, samples, and tools to DML 1.12.1 (#498)
Co-authored-by: Sheil Kumar <s****k@m****m> |
Sheil Kumar <s****7@g****m>
Committed by: GitHub <n****y@g****m> |
about 3 years ago | |
| 151e57c3 | guard additional parameter of resample2 (#496) |
Xiang Zhang <x****z@m****m>
Committed by: GitHub <n****y@g****m> |
about 3 years ago | |
| ae44b189 | update resample in DirectMLX (#495) |
Xiang Zhang <x****z@m****m>
Committed by: GitHub <n****y@g****m> |
about 3 years ago | |
| 46a71aa9 | Adding AvgPooling1 (#491) |
raoanag <1****g@u****m>
Committed by: GitHub <n****y@g****m> |
about 3 years ago | |
| 24e0ccb6 | Dequantize by scale data type (#478) |
raoanag <1****g@u****m>
Committed by: GitHub <n****y@g****m> |
about 3 years ago | |
| 5aaec405 | Support ORT extensions in DxDispatch (#472) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
about 3 years ago | |
| 7cdf6582 | Update DxDispatch (#460) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
about 3 years ago | |
| 52e24420 | Update readme, samples, and tools to DML 1.12.0 (#455) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| 52423556 |
Update DML samples and release notes for DML1.11.0 (#453)
Co-authored-by: Dwayne Robinson <d****r@m****m> |
Sumit Agarwal <s****0@g****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| a30875bc | DxDispatch microbenchmarking enhancements (#452) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| 70b0af99 | Dxdispatch: option to clear D3D shader caches (#441) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| 737b12cc | small fix to install different versions of spacy based on python version (#440) |
Xiang Zhang <x****z@m****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| 4a6920e0 | Update requirements.txt (#433) |
Linnea May <l****e@g****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| 4d2e7f41 | DxDispatch ONNX enhancements (#427) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| 088b99d6 | Print GPU/CPU timing of each iteration under -v flag (#428) |
mingmingtasd <m****u@i****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| 5b99d10d | Update to newer version of pybind11 to fix compile error (#406) |
Xiang Li <p****e@o****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| 4ad660c0 | Add accelerate to the package requirements (#419) |
Patrice Vignola <v****e@g****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| 7abf633e | Update stable diffusion instructions (#418) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| fd3fe8cc | Use OnnxStableDiffusionPipeline instead of StableDiffusionOnnxPipeline which ... |
tinyboxvk <t****k@u****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| 492471d9 | Increase CG task timeout (#401) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| 29c71923 | Instruction to set up Stable Diffusion to run on ORT-DML (#399) |
Chai Chaoweeraprasit <w****5@h****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| 0243fef9 | Support Xbox in HelloDirectML sample (#394) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| 96faab60 | Update DxDispatch dependencies and event duration method (#393) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| b95f3cf7 | Support for node names in DMLX graphs (#389) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| e491f5bc | Repoint Readme.md issue link to current TensorFlow plugin repo (#381) |
Dwayne Robinson <d****r@m****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| ecbc2046 | Comparison type fix in dxdispatch helper script (#336) |
Justin Stoecker <j****k@m****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| 70e625c0 | Update samples and dxdispatch to DML1.10.1 (#377) |
Sumit Agarwal <s****0@g****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| 2e7b8929 | Update attention_is_all_you_need sample readme |
Xiang Zhang <x****z@m****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| f7210197 |
add training/hyperparam yaml files- not the actual data (#365)
Co-authored-by: linnealovespie <l****y@m****m> |
Linnea May <l****e@g****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| 156cdb95 | Update train.py (#350) |
Linnea May <l****e@g****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| 5208c964 | Update README.md (#349) |
Linnea May <l****e@g****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| 983b5a91 | Update README.md (#342) |
Linnea May <l****e@g****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| a8e9d20b |
edit documentation and samples (#341)
Co-authored-by: Linnea May <l****y@m****m> |
Linnea May <l****e@g****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| 5610f64b | Update DML1.10 Release Note (#340) |
Sumit Agarwal <s****0@g****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago | |
| e24ac449 | DML 1.10 Release.md (#339) |
Sumit Agarwal <s****0@g****m>
Committed by: GitHub <n****y@g****m> |
over 3 years ago |