Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/13206
Add has device option for checking if a node has a device option set
Reviewed By: bwasti
Differential Revision: D12815365
fbshipit-source-id: 58477df93777f470cfb30cd75f02a659a7017b7c
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/13132
Expose more of the C++ API to python
Reviewed By: duc0
Differential Revision: D10855086
fbshipit-source-id: 98cc89bc72ef91ed1c59c1a19688e047765cf90b
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/13203
Minor changes in the test workflow to run the model on CPUs
Reviewed By: stephenyan1231
Differential Revision: D9925797
fbshipit-source-id: b7b1fb2658ab68b1ffc2b1f7b314958ea4732b32
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/13004
Implement BucketWeighted model layer, which learns a weight for each possible score in an IdScoreList. Here, we assume that the scores in the IdScoreList have already been converted into the appropriate 'buckets'. If this is not done, then essentially each score represents its own bucket.
We assume that the scores/buckets are integers, and if max_score is not set, we assume that the maximum cardinality of the score is less than or equal to the cardinality of the ids.
Reviewed By: chonglinsun
Differential Revision: D10413186
fbshipit-source-id: 743e643a1b36adf124502a8b6b29976158cdb130
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12843
This adds a cuda implementation for the UpsampleBilinearOp and UpsampleBilinearGradientOp.
The CUDA code is based off of the corresponding ResizeNearest operators but with bilinear interpolation logic taken from the CPU implementation.
Reviewed By: houseroad
Differential Revision: D10453776
fbshipit-source-id: b29ac330b72465974ddb27c0587bca590773fdec
Summary:
This is mostly for reusing all the cudnn test cases in our python operator_tests.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12278
Differential Revision: D10842592
Pulled By: bddppq
fbshipit-source-id: 4b3ed91fca64ff02060837b3270393bc2f9a9898
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/13007
No reason to use the hook if it's set, this helps fbcode traces.
This slightly pessimizes the stack trace for ATen functions,
because we are no longer skipping all of the frames we should.
This is probably OK.
Reviewed By: Yangqing
Differential Revision: D10518499
fbshipit-source-id: be54e490df3c3fde7ff894b5b1473442ffc7ded3
Summary:
TSIA - we want to deprecate numba in fbcode when moving to new compiler tiers.
Converted the old test to a non-numba regular python op test.
Reviewed By: xw285cornell
Differential Revision: D10519910
fbshipit-source-id: 0e9188a6d0fc159100f0db704b106fbfde3c5833
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12848
Updated all non-test uses of protobuf::MessageLite::SerializeAsString to call
SerializeAsString_EnforceCheck so that the return value is checked and can
throw an exception if failing.
Most of the affected code was called from classes derived from BlobSerializeBase.
Didn't touch most tests and ENFORCE calls because they usually do checks
anyway.
Original commit changeset: c0760e73ecc7
Reviewed By: dzhulgakov
Differential Revision: D10453456
fbshipit-source-id: d2f2b7b4578e721924354149f08f627c7e3bf070
Summary:
- exhaustive_search attribute will be blacklisted so it
will be discarded from the coverted onnx model. At present
it throws error while verifying the onnx model
Signed-off-by: Parth Raichura <parth.raichura@softnautics.com>
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12805
Differential Revision: D10502374
Pulled By: ezyang
fbshipit-source-id: 0926dfa3237a8a431184e7f7250146e5b0cbfb85
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12900
Workspace sometimes will be populated with input tensors for shape inference but net.external_input() is not a reliable way to tell weights from input in the workspace. We say in some usecases where net.external_input() is empty. In this case, we need to give user an option to provide input hint.
Reviewed By: bddppq
Differential Revision: D10476822
fbshipit-source-id: 1a3fa2df69b959d5b952a7824eba9e6c713f4f07
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12897
UnsafeCoalesce Op is used during memonger days when we try to coalesce operators
for better efficienct computation kernels. It creates a little bit of an unsafe
underlying memory storage pattern.
With the new tensor unification I am not sure if it is still safe for us to do
so, so I propose we delete it for the sake of safety.
Reviewed By: bddppq, ilia-cher
Differential Revision: D10475980
fbshipit-source-id: b1a838c9f47d681c309ee8e2f961b432236e157e
Summary:
This test flushes out the issue that IDEEP cannot handle tensor with dims like (0, 2), which is a valid tensor shape.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/8459
Differential Revision: D10419328
Pulled By: yinghai
fbshipit-source-id: c5efcd152364a544180a8305c47a2a2d126ab070
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12736
This updates UpsampleBilinearOp and UpsampleBilinearGradientOp to support scales to bring it inline with ResizeNearestOp https://github.com/pytorch/pytorch/pull/12720.
Reviewed By: houseroad
Differential Revision: D10416228
fbshipit-source-id: f339b7e06979c9c566afb4cee64a2d939b352957
Summary: Added 2 years ago in D3665603, never used, kill it.
Reviewed By: ezyang
Differential Revision: D10421336
fbshipit-source-id: 1b027a9ef2b71d0dd2c572cd4338bc8e046320d8
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12799
Updated all non-test uses of protobuf::MessageLite::SerializeAsString to call
SerializeAsString_EnforceCheck so that the return value is checked and can
throw an exception if failing.
Most of the affected code was called from classes derived from BlobSerializeBase.
Didn't touch most tests and ENFORCE calls because they usually do checks
anyway.
Reviewed By: ezyang
Differential Revision: D10416438
fbshipit-source-id: cb842e3e26b0918829d71267a375d4dd40600d58
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12681
- Get rid of NodeMatchCriteria as a template parameter, which was too generic. So MatchNode<NodeMatchCriteria> becomes MatchNode<GraphType>, and MatchStore stores the predicate on GraphType::NodeRef.
- Similarly, get rid of NNNodeMatchCriteria
Now one can just pass in a function pointer NodeRef -> bool to NNMatchNode constructor directly like this
mg.createNode(is<Relu>)
- Merge static utilities in SubgraphMatcher class into MatchGraph class
- Rename MatchNode to MatchPredicate
Change use cases and tests to make it work
Reviewed By: ZolotukhinM
Differential Revision: D10386907
fbshipit-source-id: 43874bd154e3d7c29ce07b4b74eca8a7a9f3078a
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/11711
Added GPU support for spatial batch normalization. This functions by reducing values from GPUs onto a CPU and broadcasting those results back to each GPU. We have run several experiments, and found these results to be better than those without spatial bn: https://fb.quip.com/fr7HAeDliPB8
Reviewed By: enosair
Differential Revision: D9547420
fbshipit-source-id: ccbd2937efd6cfd61182fff2f098fb7c5ae8aeb1
Summary: Add a mapping for conversion -- this will help with debugging as well but is directly used by the TUI stacked on top of this
Reviewed By: duc0
Differential Revision: D10396130
fbshipit-source-id: cdd39278f0ed563bb828b1aebbbd228f486d89c8
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12685
In this diff, we push the fake run of the net into the ONNXIFI transformer, because
1. We cannot do shape inference for every op
2. Since the net has been SSA rewritten, we cannot use shape info from outer workspace directly.
In addition, this diff adds input shape info when querying the `onnxBackendCompatibility` function.
Reviewed By: bddppq
Differential Revision: D10390164
fbshipit-source-id: 80475444da2170c814678ed0ed3298e28a1fba92
Summary:
Simple additions that make it vastly easier to use nomnigraph in
python
Reviewed By: duc0
Differential Revision: D10383027
fbshipit-source-id: 441a883b84d4c53cca4f9c6fcc70e58692b8f782
Summary:
This seems to be a typo that never got caught - no actual functionality changes.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12688
Differential Revision: D10391704
Pulled By: Yangqing
fbshipit-source-id: ce633776957628c4881956c5423bfab78294d512
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12661
Was disabled since workspace.has_mkldnn is now set to false
Reviewed By: yinghai
Differential Revision: D10383913
fbshipit-source-id: ad6dc705f0606b3711e8b450dc384ad3ebb87686
Summary:
The pytorch.org site redirects all of the http:// requests to the https:// site anyway, so the comments and error messages might as well refer directly to the https:// site. The GitHub project description should also be updated to point to https://pytorch.org
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12636
Differential Revision: D10377099
Pulled By: soumith
fbshipit-source-id: f47eaba1dd3eecc5dbe62afaf7022573dc3fd039
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12382
implement fp16-> (uint8 + scale and bias in fp32)
this is similar to fp32 rowwise quantization
we could have done scale and bias in fp16 but not too motivated since we are not saving much and those datatypes have to be converted to fp32 to process since x86 doesn't support half float operations anyways
Reviewed By: csummersea
Differential Revision: D10220463
fbshipit-source-id: 6c382026de881f03798c2e5fc43abfc80f84ea1f
