microsoft/onnxruntime

Pull Requests Commits

fix attention mask

yufenglee committed 2 years ago

249fcc1e

comments and docs

aciddelgado committed 2 years ago

96b02705

change seqlen logic

aciddelgado committed 2 years ago

bcd126ae

aciddelgado committed 2 years ago

b5b62cdc

aciddelgado committed 2 years ago

6d754f68

undo cmake/external/onnx change

aciddelgado committed 2 years ago

66f1600d

aciddelgado committed 2 years ago

791621b0

address comments

aciddelgado committed 2 years ago

2fab38e7

remove kv share flag and bnsh flag

aciddelgado committed 2 years ago

9e2fae76

aciddelgado committed 2 years ago

d89995f0

fix warning and lint

aciddelgado committed 2 years ago

e2eadab7

aciddelgado committed 2 years ago

40f6e3b5

disable memory efficient and pipeline test

aciddelgado committed 2 years ago

90f23c0d

[TensorRT EP] Properly set CUDA_INCLUDE_DIR for onnx-tensorrt (#18274)

chilo-ms committed 2 years ago

Verified dfafcb58

flash attention works no buffer

aciddelgado committed 2 years ago

4c5a32ab

attention mask for flash attention with cache

aciddelgado committed 2 years ago

afa8ea01

Remove internal enforce for IO binding inputs (#18266)

kunal-vaishnavi committed 2 years ago

Verified 08eaa1c5

[TensorRT EP] Fix bug for shape tensor input (#18253)

chilo-ms committed 2 years ago

Verified 84bdf04b

Block-wise 4b quantization matmul operator change (#18172)

chenfucn committed 2 years ago

Verified 26b39641

Make MlasTestFixture::mlas_tester an inline variable. (#18263)

edgchen1 committed 2 years ago

Verified 2ec1f94b

Change a bitwise logical xor to logical wise (#18246)

Changming Sun committed 2 years ago

Verified 4c4d79a6

Fix Signed Mismatch (#18258)

nums11 committed 2 years ago

Verified 192caee8

[JS/Web] Added Unifroms support to unary ops. (#18223)

satyajandhyala committed 2 years ago

Verified e207060a

[QNN EP] Fix Pad UT (#17982)

winskuo-quic committed 2 years ago

Verified 90f205e7

Rework/cleanup the C# build infrastructure for nuget packages. (#18127)

skottmckay committed 2 years ago

Verified c352e9b1

Update XNNPACK to latest version (#18038)

skottmckay committed 2 years ago

Verified 4f2096be

Introduce new optimizer Pad + Conv/MaxPool (#18136)

sumitsays committed 2 years ago

Verified e36d0037

Pre-link when creating static library for apple framework (#18241)

skottmckay committed 2 years ago

Verified 016b7526

Partially disable QGemm tests for float 8 types (#18196)

xadupre committed 2 years ago

Verified 1439da36

Rerun the flaky ort-web tests automatically (#18187)

mszhanyi committed 2 years ago

Verified 9f5a6856

Older