microsoft/onnxruntime

Pull Requests Commits

Bugfix for BiasSoftmax Fusion (#12517)

Lafi7e committed 3 years ago

Verified 0c6037b5

Eager Mode - Support Concatenation via aten::cat.out (#12527)

msftlincoln committed 3 years ago

Verified 0d9a02e6

Python module for dumping activation tensors when running an ONNX model (#12474)

chenfucn committed 3 years ago

Verified 47b787c2

Load checkpoint in cpp (#12352)

AdamLouly committed 3 years ago

Verified 2681648f

Add codeowners for requirement files (#12512)

faxu committed 3 years ago

Verified ee3b7574

[CUDA] SoftmaxCrossEntropy Kernels Refactor (#12482)

Lafi7e committed 3 years ago

Verified 2bed0d4a

[CUDA] Mod Op Kernel (#12499)

Lafi7e committed 3 years ago

Verified cfa09d16

Improve the compilation speed when compiling for multiple architectures. (#12490)

pengwa committed 3 years ago

Verified a2dc3e9e

Incrementally free initializers while saving to OrtValue instances (#12485)

skottmckay committed 3 years ago

Verified 56bd96a3

remove the link the comments (#12510)

HectorSVC committed 3 years ago

Verified 730240d2

[Java] JNI refactor for ONNX Tensor (#12281)

Craigacp committed 3 years ago

Verified 8a86b346

new quantized operators split (#12495)

jchen351 committed 3 years ago

Verified 8c5c2834

Fix various warning in kernel explorer (#12501)

cloudhan committed 3 years ago

Verified 9c055770

set zero point to 0 if all value are 0.0 (#12470)

yufenglee committed 3 years ago

Verified bdd6b00c

Avoid false-positive dependent name lookup error by not depending on auto keyword (#12483)

cloudhan committed 3 years ago

Verified ddea1e48

DML EP fix training build error (#12461)

fdwr committed 3 years ago

Verified eb90b52a

Update ORTModule Default Opset Version to 15 (#12419)

Lafi7e committed 3 years ago

Verified e85e31ee

CUDA kernel for ClipGradNorm for TensorSeq gradients (#12412)

baijumeswani committed 3 years ago

Verified a7d62907

[DELETE] delete python package rocm4.3.1 (#12480)

PeixuanZuo committed 3 years ago

Verified 3e1b0ac4

Fix Python Packaging CI (Rocm) (#12477)

ytaous committed 3 years ago

Verified b879dca5

Rework parts of Graph::Resolve to reduce memory usage (#12176)

skottmckay committed 3 years ago

Verified 8d830adf

Add composable kernel GEMM baseline for kernel explorer (#12364)

cloudhan committed 3 years ago

Verified f39354d7

[CUDA] BiasSoftmax Supporting New Pattern (#12361)

Lafi7e committed 3 years ago

Verified 37995a72

Lironkesem/unsqueeze_and_squeeze (#12421)

LironKesem committed 3 years ago

Verified d452462b

Remove dynamic allocation for ThreadPool ParallelSection (#12429)

yuslepukhin committed 3 years ago

Verified a4ef0e7f

Enable quant op to share quantization parameter between input and ouput (#12408)

yufenglee committed 3 years ago

Verified ac10f33d

Minor doc fixes (#12388)

RyanUnderhill committed 3 years ago

Verified 52d46997

Refactor InferenceSession Load member functions. (#12430)

edgchen1 committed 3 years ago

Verified 3efd9a73

dev notes for layout transformer (#12396)

askhade committed 3 years ago

Verified 97268e02

Update script to find optimizers that potentially need supported opset updates (#12330)

skottmckay committed 3 years ago

Verified a3de1bbf

Newer Older