microsoft/onnxruntime

Pull Requests Commits

Try loading onnxruntime_providers_cuda.dll from another directory

baijumeswani committed 121 days ago

63c588ac

Fix quantization in Whisper model export (#26353)

jiafatom committed 122 days ago

Verified ed7847c1

[webgpu] Don't use num_workgroups when use indirect dispatch (#26334)

qjia7 committed 122 days ago

Verified 8ab27d9f

[build] disable DAWN_FETCH_DEPENDENCIES when using custom dawn src path (#26346)

fs-eire committed 122 days ago

Verified b1fdf9d3

[LARCH64 CPU]Provide inference acceleration optimization for Loongson CPU with 4-bit quantized models (#26280)

movedancer committed 124 days ago

Verified 7cc28b0b

[QNN EP] Reshape Transpose Fusion for 6D tensors (#26338)

quic-ashwshan committed 124 days ago

Verified 7804b5c1

S390x simd implementation (#25757)

AlekseiNikiforovIBM committed 124 days ago

Verified 510dd14f

Refactor: Move fundamental data type headers to core/common (#26330)

Changming Sun committed 125 days ago

Verified e0569fd9

Quant tool: Use 2D scales and zp in default MatMulNBitsQuantizer (#26329)

jambayk committed 125 days ago

Verified 6dbd41b5

[QNN EP] Upgrade QNN to 2.39.0 (#26249)

qti-jkilpatrick committed 125 days ago

Verified 878863d0

Quant tool: Don't store original outputs in MinMax Calibrator (#26321)

jambayk committed 126 days ago

Verified bbd0a80d

Internal Dupe of #25255 - [MLAS] Optimize MlasConv using thread partition opt (#26103)

hariharans29 committed 126 days ago

Verified 992c5981

Add support to filter_ep_devices using ep or device metadata in perftest (#26255)

jnagi-intel committed 126 days ago

Verified 9e79b367

[webgpu] Support RotaryEmbedding in flash attention (#26297)

xiaofeihan1 committed 126 days ago

Verified 90d92f82

[Build] Lock torch, onnxscript and onnx-ir versions to latest (#26315)

tianleiwu committed 126 days ago

Verified 036fde03

[webgpu] And int64 to cast (#25610)

qjia7 committed 127 days ago

Verified f0015b95

[TensorRT] Fix DDS output bug during engine update (#26272)

toothache committed 127 days ago

Verified 654137fa

Fix shape inference failure with in-memory external data (#26263)

Changming Sun committed 128 days ago

Verified aafdb3a5

[WebNN] Fix some issues in reduction ops (#26289)

Honry committed 128 days ago

Verified 04ed484f

[MIGraphX EP ] Add support for QLinearAveragePool and QLinearGlobalAveragePool (#26162)

TedThemistokleous committed 128 days ago

Verified 5d786dfd

[webgpu] Enable indirect dispatch for flash attention (#26207)

qjia7 committed 128 days ago

Verified cd4ac494

Bump onnx to 1.19.1 (#26202)

titaiwangms committed 128 days ago

Verified 94de31fa

Fix build.sh --parallel 1 incorrectly triggering parallel build (#26264)

XXXXRT666 committed 128 days ago

Verified 4c7fc439

[MIGraphX EP] Link FP4 types between OnnxRT and MIGraphX APIs (#26231)

TedThemistokleous committed 128 days ago

Verified 24d7aee6

[CUDA] Upgrade build pipelines to use CUDA 12.8 + cuDNN 9.8 (#26267)

tianleiwu committed 128 days ago

Verified 311b4a64

Reworked sgemm_kleidi memory allocations to reuse memory buffers (#26166)

JonathanC-ARM committed 128 days ago

Verified 96f56ba0

Add windows server to supported OS (#26275)

tianleiwu committed 129 days ago

Verified 826a1e62

Fix Memory Issue sparse_attention Rotary (#26278)

jiafatom committed 129 days ago

Verified 41b238ed

Bump vite from 6.3.5 to 6.3.6 in /js/web/test/e2e/exports/testcases/vite-default (#26000)

dependabot[bot] committed 131 days ago

Verified dde2fefe

fix np.testing argument order (#26128)

xiaofeihan1 committed 131 days ago

Verified 1442fe00

Older