summaryrefslogtreecommitdiff
path: root/torch/onnx
AgeCommit message (Expand)AuthorFilesLines
2018-12-06Add LogSigmoid support in ONNX symbolic (#14830)zrphercule1-0/+6
2018-12-05Add torch.nn.RReLU support in symbolic (#14781)zrphercule1-1/+8
2018-12-03Implement Device as a type in the script (#14666)Zachary DeVito1-0/+2
2018-11-29Switch import/export to python printing (#14400)Zachary DeVito1-3/+3
2018-11-27Speed-up "advanced" indexing operations (#13420)Sam Gross1-2/+2
2018-11-27Trace in-place ops (#14254)Michael Suo1-2/+2
2018-11-08Tracer now records Python variable names (#13441)James Reed1-64/+0
2018-11-05Support new upsample in symbolic, caffe2 backend & caffe2 frontend (#13272)zrphercule1-4/+7
2018-11-02Fix EraseListConstruct pass during ONNX export (#13195)Wanchao Liang1-10/+3
2018-11-02add augmented assignment ops (#13364)Michael Suo1-0/+1
2018-10-31Back out "[pytorch][PR] Support upsample" (#13413)Rui Zhu1-7/+4
2018-10-30Fix aten::to symbolic + add expand_as (#13325)James Reed1-0/+10
2018-10-30Fix "CUDA Tensor __rsub__ breaks when device is not 0" (#12956)Will Feng1-0/+4
2018-10-29ArgumentStash for int64_t arguments (#12939)James Reed1-1/+1
2018-10-29Support upsample (#13152)zrphercule1-4/+7
2018-10-26Move ConstantPadNd into ATen (#10885)William Horton1-0/+8
2018-10-25Add c10::optional to type syntax (#12582)Wanchao Liang2-8/+9
2018-10-25Add randn in onnx symbolic (#12880)zrphercule1-0/+6
2018-10-18Move torch.onnx.operators functions into ATen (#12803)James Reed2-12/+13
2018-10-15Fix the symbolic for pixel shuffle (#12192)Lu Fang1-0/+14
2018-10-15Fix ONNX upsample mode (#12648)ahirner1-1/+1
2018-10-15Add AdaptiveAvgPool2d and AdaptiveMaxPool2d to ONNX.symbolic (#9711)Konstantin Semianov1-0/+12
2018-10-12Fix aten::to export in ONNXJames Reed1-5/+5
2018-10-11Implement aten::contiguous (#12541)James Reed1-0/+4
2018-10-11Enable addmm fusion for ONNX export only (#12538)James Reed1-2/+2
2018-10-08Fix issues with ATenOp handling methods where `self` is not the first arg (#1...James Reed1-0/+4
2018-10-03add torch factory methods (zeros/ones) to onnx symbolicWanchao Liang1-12/+16
2018-09-27Fix some symbolics for ReduceSum, GE, LE (#12123)Wanchao Liang1-2/+2
2018-09-26Fix ONNX bug, add symbolic for fullAdam Paszke1-1/+12
2018-09-21Minor JIT improvements (#11654)Adam Paszke1-1/+10
2018-09-19Fix the symbolic for embedding_bag in ONNX_ATEN_FALLBACK (#11840)Lu Fang1-1/+1
2018-09-18allow no-alpha add/sub in onnx symbolic (#10972)Wanchao Liang1-10/+7
2018-09-13migrating deprecated calls without abc module for containers (#11515)Jeff Smith1-2/+2
2018-09-13End-to-end dynamic slicing with ONNX DynamicSlice experimental operator (#11255)James Reed1-3/+13
2018-09-13move instance_norm to aten (#10792)Roy Li1-16/+16
2018-09-12Fused weightnorm for ATen (#10842)Michael Carilli1-0/+5
2018-09-11Make .to() methods native functions (to fix JIT tracing)Adam Paszke1-0/+18
2018-09-10Improve support for tracing sizes, add more tracer warnings (#11288)Adam Paszke1-0/+19
2018-09-08fix #10838 incorrect bidirectional output format (#11368)Yifei Teng1-3/+3
2018-09-05Handling for py2/py3 division differences (#11016)Richard Zou1-0/+2
2018-09-05Fix the batchnorm onnx exporting when affine=FalseLu Fang1-0/+10
2018-09-05Port PackedSequences functions to C++ (#11224)Adam Paszke1-0/+28
2018-09-04Report an error when trying to record a mutable operator when (#11129)Zachary DeVito1-1/+5
2018-08-31Traceable detach (#11038)James Reed1-0/+5
2018-08-30Don't flatten output lists in the JIT IR (#10949)Adam Paszke2-7/+20
2018-08-30Change torch.jit.trace to no longer be a decorator (#11069)Zachary DeVito1-1/+1
2018-08-29Record tensor factory functions in trace (#10935)Zachary DeVito1-13/+22
2018-08-28Enable fusing aten::expand on GT, LT, EQ (#10845)Lu Fang1-3/+4
2018-08-28Fix bugs in handling of negative slice + gather indices (#10973)James Reed1-3/+10
2018-08-27Tuples/Lists can now be inputs/outputs to script and other simple fixes. (#10...Zachary DeVito1-0/+2
generated by cgit v1.2.3 (git 2.25.1) at 2024-11-29 22:19:50 +0000