microsoft/onnxruntime

Pull Requests Commits

Minor change to improve performance for operator Pad. (#5537)

xadupre committed 5 years ago

Verified 8a0f5c50

Remove most ROCm-specific element-wise code and reuse CUDA element-wise code.

jessebenson committed 5 years ago

7ccdfed1

Use TArray in AMD element-wise kernels, rather than manually copying memory to device.

jessebenson committed 5 years ago

52228a70

Fix a memory leak in test_inference.cc (#6201)

Changming Sun committed 5 years ago

Verified 1fc7f92f

Openvino ep 2021.2 (#6196)

sfatimar committed 5 years ago

Verified 73479969

Add ability to skip GPU tests based on GPU adapter name (#6198)

Ryan Lai committed 5 years ago

Verified 0494a0f9

Dockerfile to build onnxruntime with ROCm 4.0

jessebenson committed 5 years ago

c5629527

Implement comparing outputs that are sequence of maps of strings to floats (#6180)

Ryan Lai committed 5 years ago

Verified 21395f8e

Clean up checkpoint tests to use the new checkpoint functions (#6188)

baijumeswani committed 5 years ago

Verified a8b48268

Condition fix in Resize operator (#6193)

hariharans29 committed 5 years ago

Verified 04b3e0ef

Implement ScatterND for CUDA EP (#6184)

hariharans29 committed 5 years ago

Verified fc27074b

Lochi/quantization tool for trt (#6103)

chilo-ms committed 5 years ago

Verified 945fae8f

Add Status.csv to EP Perf Tool (#6167)

oliviajain committed 5 years ago

Verified 234e94b4

Tune fast Gelu to use exp(x) instead of tanh(x) on Rocm platform (#6174)

Suffian Khan committed 5 years ago

Verified 67ac6ae4

improve perf for softmax (#6128)

weixingzhang committed 5 years ago

Verified 53307a5f

Add usage details of unified MCR container image (#6182)

smkarlap committed 5 years ago

Verified ea9cfa55

Android coverage dashboard (#6163)

satyajandhyala committed 5 years ago

Verified 201d0dbb

Backend APIs for checkpointing (#5803)

jingyanwangms committed 5 years ago

Verified f874260b

Helper for compiling EP to generate deterministic unique ids for use in MetaDef names (#6156)

skottmckay committed 5 years ago

Verified 2da8060f

Update get_docker_image.py to enable use without image cache container registry. (#6177)

edgchen1 committed 5 years ago

Verified cd3a5acc

Fix typo in BERT pretraining script (#6175)

Derek Murray committed 5 years ago

Verified 11b0a540

[NNAPI EP] Enable per-channel quantization for QlinearConv (#6155)

guoyu-wang committed 5 years ago

Verified bbb52e92

aggregate model states only for the case when mixed precision was true (#6176)

baijumeswani committed 5 years ago

Verified 39aedbc9

Update documentation for contributing a PR and add deprecation notices for PyOp and ORT server. (#6172)

Pranav Sharma committed 5 years ago

Verified 86493e6d

Don't try to bind unused inputs in the Training frontend (#6166)

Sergii Dymchenko committed 5 years ago

Verified 824ef9a1

save_checkpoint, load_checkpoint and aggregate_checkpoints (#6136)

baijumeswani committed 5 years ago

Verified adc20710

Remove ignored build warnings for pybind on Mac (#6165)

guoyu-wang committed 5 years ago

Verified c339bb2d

Revert "Fuse MatMulIntegerToFloat only when scales are scalar (#6008)" (#6169)

yufenglee committed 5 years ago

Verified 98d8a3e3

Bugfix for topk cuda kernel (#6164)

duli2012 committed 5 years ago

Verified 34725ae5

Update TensorRT-ExecutionProvider.md (#6161)

jayrodge committed 5 years ago

Verified dec703b6

Newer Older