Summary:
Hi guys,
I'd like to build Caffe2 with more supported options in Windows with Microsoft Visual Studios.
This is the first pull request.
Running scripts/build_windows_shared.bat is able to build Caffe2 with both CMAKE_BUILD_TYPE=Debug and CMAKE_BUILD_TYPE=Release with Visual Studio 14 2015.
CUDA is 9.0, cudnn is 7.0.5, glog, gflags and lmdb are supported on my system.
Python is 3.5, Detectron works from python interface as well.
It was even possible to debug detectron code and step into caffe2_gpu.dll with pdbs built.
What is disappointing, that c10/experimental ops don't build with this Visual Studio generator, I added special option INCLUDE_EXPERIMENTAL_C10_OPS (default ON) to deal with it in build_windows_shared.bat.
After this pull request the next step is to add Visual Studio 2017 support in the script.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/13550
Reviewed By: ezyang
Differential Revision: D13042597
Pulled By: orionr
fbshipit-source-id: f313f909f599cd582a1d000eff766eef3a9fc4fc
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/13002
Batch dim wasn't handled in the CPU impl (will fail for inputs with N > 1).
Fixing that here.
Differential Revision: D10515159
fbshipit-source-id: ee7e4f489d2d4de793f550b31db7c0e2ba3651e8
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12714
This is a short change to enable c10 namespace in caffe2. We did not enable
it before due to gflags global variable confusion, but it should have been
mostly cleaned now. Right now, the plan on record is that namespace caffe2 and
namespace aten will fully be supersets of namespace c10.
Most of the diff is codemod, and only two places of non-codemod is in caffe2/core/common.h, where
```
using namespace c10;
```
is added, and in Flags.h, where instead of creating aliasing variables in c10 namespace, we directly put it in the global namespace to match gflags (and same behavior if gflags is not being built with).
Reviewed By: dzhulgakov
Differential Revision: D10390486
fbshipit-source-id: 5e2df730e28e29a052f513bddc558d9f78a23b9b
Summary:
This is for Caffe2 optimization.
WIth this optimization, the following two ops can boost a lot. (Test with MaskRCNN, on SKX8180 one socket)
BatchPermutation op: reduced from 8.296387 ms to 1.4501984 ms.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12153
Differential Revision: D10362823
Pulled By: ezyang
fbshipit-source-id: 04d1486f6c7db49270992cd8cde41092154e62ee
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12180
I had to fix a lot of call sites, because a lot of places assume that
you can actually get a const vector&, and if the internal representation
of sizes in a tensor is NOT a vector, it's not possible to fulfill
this API contract.
Framework changes:
- I deleted TensorImpl::dims(); caffe2::Tensor::dims() just forwards to
sizes() now.
- De-templatized SetDims; now it is an explicit list of ArrayRef and
variadic overloads. This makes implicit conversions work again,
so I don't need to explicitly list the std::vector cases too.
- As a knock-on effect, this causes Reset() to accept at::IntList as well as
const std::vector<int64_t>&
- Edited variadic overloads of SetDims to all forward to the underlying
arbitrary-dim implementation, reducing code duplication. (It's probably
marginally less efficient in the new world.)
- Replace Tensor constructor accepting const std::vector<int64_t>& with at::IntList
- Make MKLTensor accept ArrayRef along with vector in constructor and
Reset (unfortunately, no implicit conversions here, since it's templated on
index type.)
- There are a few other places, like cudnn, where I changed functions
that previously took const std::vector<int64_t>& to take at::IntList
instead.
Classification of call site changes:
- 'const std::vector<int64_t>& x_dims = x.dims()' ==>
'at::IntList x_dims = x.dims()'
- 'std::vector<int64_t> x_dims = x.dims()' ==>
'std::vector<int64_t> x_dims = x.dims().vec()' (we need a copy!)
Usually this is because we're about to mutably modify the vector
to compute some new dimension. However, it also very commonly occurs in the
form: 'x_dims_ = x.dims()' because we frequently cache sizes in operators.
- Instead of constructing std::vector<int64_t>{blah, blah}, construct an
at::IntList directly
ArrayRef changes:
- cbegin()/cend() iterators, they operate the same aas begin()/end() because
everything on ArrayRef is const.
- Moved operator<< into ArrayRef.h, so that it's always available when
working with ArrayRef. I also templated it, so it now works on an
ArrayRef of any type.
- Add operator== overload for ArrayRef, and also add variants to permit
comparison of ArrayRef with std::vector, a very common operation.
(The non-templated version of operator== can get these automatically
via implicit conversion, but with templates C++ refuses to do
any explicit conversions.)
I'm planning to audit all dims() call sites to make sure they don't
expect 'auto x = t.dims()' to give you an x whose lifetime can validly
outlive the tensor.
I opted not to do a dims() to sizes() rename, because dims() also matches
the protobufs accessor. Bad news!
Reviewed By: jerryzh168
Differential Revision: D10111759
fbshipit-source-id: a2a81dc4b92c22ad4b3b8ef4077a7e97b6479452
Summary:
TSIA. Right now we should basically use C10_EXPORT and C10_IMPORT for explicitly marking dllexport and dllimport, as a continued effort of the C10 unification.
This is a codemod by mechanically doing the following change:
CAFFE2_{EXPORT,IMPORT} -> C10_{EXPORT,IMPORT}
AT_CORE_{EXPORT,IMPORT} -> C10_{EXPORT,IMPORT}
Pull Request resolved: https://github.com/pytorch/pytorch/pull/12019
Reviewed By: ezyang, teng-li
Differential Revision: D10016276
Pulled By: Yangqing
fbshipit-source-id: a420d62c43d1110105fc88f9e9076e28a3203164
Summary:
This completely removes BUILD_CAFFE2 from CMake. There is still a little bit of "full build" stuff in setup.py that enables USE_CUDNN and BUILD_PYTHON, but otherwise everything should be enabled for PyTorch as well as Caffe2. This gets us a lot closer to full unification.
cc mingzhe09088, pjh5, ezyang, smessmer, Yangqing
Pull Request resolved: https://github.com/pytorch/pytorch/pull/8338
Reviewed By: mingzhe09088
Differential Revision: D9600513
Pulled By: orionr
fbshipit-source-id: 9f6ca49df35b920d3439dcec56e7b26ad4768b7d
Summary:
1. Support ops needed for inference of Faster-RCNN/Mask-RCNN needed in Detectron, mostly direct fallbacks.
2. Use CPU device to hold 0-dim tensors and integer tensors in both fallback op and blob feeder, needed by Detectron models.
3. Ignore 0-dim tensor in MKL-DNN concat operator.
4. Generate dynamic library of Detectron module for CPU device.
This PR obsoletes #9164.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/10157
Differential Revision: D9276837
Pulled By: yinghai
fbshipit-source-id: dc364932ae4a2e7fcefdee70b5fce3c0cee91b6f
Summary:
Breaking out of #8338
This PR is a workaround for a bug with CUDA9.2 + GCC7.
Here is the error this PR fixed:
.../pytorch/caffe2/operators/elementwise_ops.h: In constructor ‘caffe2::BinaryElementwiseWithArgsOp<InputTypes, Context, Functor, OutputTypeMap>::BinaryElementwiseWithArgsOp(const caffe2::OperatorDef&, caffe2::Workspace*)’:
.../pytorch/caffe2/operators/elementwise_ops.h:106:189: error: ‘GetSingleArgument<bool>’ is not a member of ‘caffe2::BinaryElementwiseWithArgsOp<InputTypes, Context, Functor, OutputTypeMap>’
BinaryElementwiseWithArgsOp(const OperatorDef& operator_def, Workspace* ws)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/10510
Reviewed By: orionr
Differential Revision: D9319742
Pulled By: mingzhe09088
fbshipit-source-id: ce59e3db14539f071f3c20301e77ca36a6fc3f81
Summary:
On Windows, the FindRocksDB script doesn't detect rocksdb installation built by cmake.
And it doesn't include/link the RocksDB dependencies either, like:
* `Snappy`
* `Shlwapi.lib`
* `Rpcrt4.lib`
This PR try to detect in config mode first before using private find module.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/7315
Differential Revision: D9287587
Pulled By: Yangqing
fbshipit-source-id: 314a36a14bfe04aa45013349c5537163fb4c5c00
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/9939
Pull Request resolved: https://github.com/facebookresearch/weakly-supervised-action-detection/pull/13
Pull Request resolved: https://github.com/pytorch/translate/pull/166
Pull Request resolved: https://github.com/pytorch/pytorch/pull/9125
Closes https://github.com/pytorch/pytorch/pull/9125
Use inheritance for polymorphism, and remove template parameter
This is to change the templating in call sites, the core implementations will change later
Before Caffe2 Tensor class was compile-time fixed to bind to a particular device/context. With this change, we're making it a runtime property (stored inside the tensor), but preserve the same semantics. For example, one has to specify device type in order to create a Tensor - there are no uninitialized tensors. More specifically the changes are:
1. We added an extra argument *DeviceType* to most of the constructors of the tensor, e.g. (Tensor(DeviceType type)),
2. Semantics of constructor Tensor(const Tensor<SrcContext>& src, ContextForCopy* context); is changed, in this constructor, the second context is passed in to enable us to call the templated Copy function, it could be in a different context as source and target previously, now we'll enforce that the context should have same device type as src, if it is provided.
3. To preserve 'get-or-construct' semantics of Blob, we added specialized getter Blob::GetMutableTensor that verifies both that Blob contains a Tensor and that it's of a correct type
4. Specifically, Tensor type is not default-constructible any more (as we don't have unknown device tensors) and thus some of the code handling STL containers needs to change
Note: Some changes are postponed just to keep this diff a bit smaller. Please see `TODO`s.
Reviewed By: ezyang, houseroad
Differential Revision: D9024330
fbshipit-source-id: e0b8295d2dc6ebe2963383ded5af799ad17164ba
Summary:
Pull Request resolved: https://github.com/facebookresearch/weakly-supervised-action-detection/pull/13
Pull Request resolved: https://github.com/pytorch/translate/pull/166
Pull Request resolved: https://github.com/pytorch/pytorch/pull/9125
Closes https://github.com/pytorch/pytorch/pull/9125
Use inheritance for polymorphism, and remove template parameter
This is to change the templating in call sites, the core implementations will change later
Before Caffe2 Tensor class was compile-time fixed to bind to a particular device/context. With this change, we're making it a runtime property (stored inside the tensor), but preserve the same semantics. For example, one has to specify device type in order to create a Tensor - there are no uninitialized tensors. More specifically the changes are:
1. We added an extra argument *DeviceType* to most of the constructors of the tensor, e.g. (Tensor(DeviceType type)),
2. Semantics of constructor Tensor(const Tensor<SrcContext>& src, ContextForCopy* context); is changed, in this constructor, the second context is passed in to enable us to call the templated Copy function, it could be in a different context as source and target previously, now we'll enforce that the context should have same device type as src, if it is provided.
3. To preserve 'get-or-construct' semantics of Blob, we added specialized getter Blob::GetMutableTensor that verifies both that Blob contains a Tensor and that it's of a correct type
4. Specifically, Tensor type is not default-constructible any more (as we don't have unknown device tensors) and thus some of the code handling STL containers needs to change
Note: Some changes are postponed just to keep this diff a bit smaller. Please see `TODO`s.
Reviewed By: xw285cornell
Differential Revision: D8121878
fbshipit-source-id: 4a5e9a677ba4ac82095df959851a054c81eccf81
Summary:
This is to simplify the data format during benchmarking. After this change, we can use the same benchmarking harness data conversion method to parse data from multiple binaries.
This change should be coordinated with the PR: https://github.com/facebook/FAI-PEP/pull/63
Pull Request resolved: https://github.com/pytorch/pytorch/pull/9555
Reviewed By: pjh5
Differential Revision: D8903024
Pulled By: sf-wind
fbshipit-source-id: 61cabcff99f0873729142ec6cb6dc230c685d13a
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/9153
Closes https://github.com/pytorch/pytorch/pull/9153
Modified the values reported by the benchmarking platform to include tensor_shape and op_args. These values have a different naming scheme to values like flops and latency.
Reviewed By: sf-wind
Differential Revision: D8729791
fbshipit-source-id: f050200be01c6d0794bf5faaa6e8cef12a00affe
Summary:
Closes https://github.com/pytorch/pytorch/pull/9199
The input shapes are not logged correctly in production because `PerfNetObserver::Stop()` only gets called after the inference is done for the net and in the mobile models, it's common practice to reuse the blobs as much as possible to save memory. And the shapes of the blobs keep changing during inference. By the time you you query `InputTensorShapes()` in `PerfNetObserver::Stop()`, you only get the final shape of the blobs.
To fix this bug, I moved the 'InputTensorShapes()' query from `PerfNetObserver::Stop()` to `PerfOperatorObserver::Stop()`. The latter gets called at the end of operator->run() whereas `PerfNetObserver::Stop()` gets called at the end of net->run().
Also remove `PerfOperatorObserver::getAnalyticalCost()` since it's now done on the server side and no longer needed on mobile
Reviewed By: Maratyszcza
Differential Revision: D8743346
fbshipit-source-id: 5d2d0132e3f5e084be7d0173863e695e62a6b4a0
* Adding instance weight to batch distill loss
as title
* add bfloat 16-31
added bfloat 16-31 and their respective unit tests
* [CUDA9] Upgrade - fbcode
CUDA9 upgrade diff D5654023 has been out for a while thanks to Pieter. But with time growing it's becoming quite hard to rebase, because of the symlinks and auto-generated build/config files in tp2. Break D5654023 into two diffs, one touching tp2 config files, and another one touching fbcode TARGETS file (adding nvcc flag). These two should be a bit easier to rebase (for detailed procedure see "Test Plan").
This diff can only be committed if:
1. CUDA 9 rpm is rolled out fleet-wide (TBD)
2. NVidia driver 390.40 is rolled out fleet-wide (done)
3. Upgrade CUDA 9.1, cudnn 7.1, nccl 2.1 (done)
4. Make sure all dependents are built (done)
5. Test all C2 operators, PyTorch (see test plan)
* Share intermediate int32 buffer across Conv ops
Adding a known type
* [C2 fix] infer function for ensure_cpu_output_op
this is adding the missing device funtion for ensure_cpu_output_op
* [int8] Add blob serializer/deserializer for Int8TensorCPU
To export to logfiledb
* [nomnigraph] Add try catch block to optimization passes in predictor
This will catch failures that happen in the optimization pass.
* Caffe2: avoid static initialization order fiasco for CAFFE_ENFORCE
CAFFE_ENFORCE uses strack trace fetcher. Which is currently a
global static variable. If at static initialization time CAFFE_ENFORCE
is used, this is a SIOF. Recently CAFFE_ENFORCE was added into init
functions registration, so we started to see this.
Meyers singleton is going to provide safety here. If stacktrace
fetcher was not registered yet, it will just use a dummy one.
* NUMA support in SparseNN CPU benchmark
Adding support for NUMA in SparseNN CPU benchmark
* [mobile-roofline] Add logging needed for roofline model
This should be all that's needed
* Let the operators using the same input if the operators are not chained
or else, we have to change the input data dims
* fix null-pointer-use UBSAN errors in in reshape_op.h
* revert previous fix on input blob name
as title
* Adding flag to let MineHardNegative automatically extract single value from dict
Model exporter requires the output of the model to be a struct. This makes it convenient to use those models directly in MineHardNegative by allow automatic extraction of the single element of dict, which is a common use case.
* Reverting change that broke internal tests back to OSS compatible state
* Import/export observer symbols for DLL, which fixes the linking error in Visual Studio.
* Add support of all default cmake build types for release to cuda.
* [fix] Re-enable events in RNN ops
We have earlier added event disabling in RNN ops as back then we didn't use
events, with current use cases this is no longer true
(https://fburl.com/8vd0lp8y)
* use ops with cude impl
* Revert D7729695: [caffe2][fix] Re-enable events in RNN ops
This reverts commit 4b215c7496fb724656ff4c776933a15bdbbcde5e
@bypass-lint
An infra SEV is better than not reverting this diff.
If you copy this password, see you in SEV Review!
@cause_a_sev_many_files
* [observer] Clean up observer_config.h
#accept2ship
* [1/n] Refactor dataio_test.py
Replace code duplication with a common function
* Add barrier net that runs before training nets
Add a synchonize barrier net that is run before training nets. With this net, shards that are faster will wait for other shards before start training. This reduce chances of the faster shards timing out during GLOO AllReduce.
Removed explicit data_parallel_model.py.synchronize call in holmes workflow. Similar change in speech/asr_training workflow will come in another diff.
* Support the dnnlowp backend in caffe2_benchmark
This is for SHARE operator latency evaluation
* Migrate integral_image_op to main caffe2
migrate integral_image_op(GPU version) given by https://fburl.com/yvqezigi
to caffe2/caffe2/operators and implement its CPU version. Write up a test
using the hypothesis_test mechanism
* [pos_disc, fbcode] Implement unjoined lr loss
As explained in https://our.intern.facebook.com/intern/wiki/Model_Based_Calibration/, when the dataset is an joined data set, where labels might change later, we need to use unjoined logloss.
The implementation is almost the same as in Sigrid (https://fburl.com/1trngsls), where
loss = y (log(p) - log(1-p)) + (1-y)(log(1-p)) = xy - (1-y)x - (1-y)log(1+exp(-x))
For x < 0, to ensure stability and avoid overflow, we reformulate the above exp as
loss = xy - (1-y)x - (1-y)x + (1-y)log(1+exp(x)) = xy + (1-y)log(1+exp(x))
Then the final expression becomes
loss = xy + (y - 1) x (x >= 0) - (1 - y) log(1 + exp(x - 2 x (x >= 0)))
where y is the true label, x is the dot product and p = logistic(x).
This kind of implementation is align with the current implementation of the original cross entropy in
https://phabricator.intern.facebook.com/diffusion/FBS/browse/master/fbcode/caffe2/caffe2/operators/cross_entropy_op.cc;0bae3b5d0f825897c5e0dd0ff10f489d7271bf25$7-13
* Keep the array to fix the conflict
* [C2] Compute Adagrad effective LR
The AdagradWithLR op outputs an extra blob which is contains the average effective learning rate across all weights in this blob.
* Open-source extractMetaNetDef & runGlobalInitialization, add new Predictor constructor from db file, and add run_map_outputs
1. Open-source extractMetaNetDef and runGlobalInitialization, for use in
2. new Predictor constructor from db file.
3. Add new run function that returns outputs as TensorMap
* Disable eigen cpu
Disable eigen cpu in transpose and reduce
* Introduce request_only/object_only property of ModelLayer
by default this is False
* A simple TC Caffe2 benchmark
We can run tunner, get MappingOptions and then use them to
compare against cuBLAS
currently broken due to LLVM issues. How to run:
hg checkout eec1ab31b59c03b8deded1c755a9abaf8c45be01
add D7401202
add D7434625
add D7506031
add D7540728
buck run @mode/dev-nosan tc/tc/benchmarks_python:caffe2_benchmark
* Move Caffe2 feature_maps_ops to open source
Need feature maps operators in open source project facebookresearch/BlueWhale
* Manually fix the conflicts in channel shuffle op
* Fix the inconsistency between different gh and fbcode
* Skip Adagrad GPU Test (Because some gpu implementation is missing)
* Fix another test to make sure it won't run on gpu when implementation is not available yet
* Track checkpoint performance in scuba
As title.
* [C2/CUDA]: fix cross entropy sigmoid with logits
when adding log_d_trick, I forgot to add it to the cuda impl; this diff fixes
it.
* Back out "[caffe2] Unregister MKL fallbacks for NCHW conversions"
Original commit changeset: 8918dd40205a
Will land after @jongsoo's diff https://phabricator.intern.facebook.com/D7596315 lands
* [Easy][C2] Don't add blob to external outputs from output_record if it's already external output
As desc.
* On Mobile phones, call GlobalInit with no arguments in predictor in case we need to perform initialization
FACEBOOK:
The QPL logger needs the initialization code. In the past, the initialization code is put in the pipeline calling Caffe2. However, those places become obsolete quickly, as the product teams change places to call Caffe2 from time to time. We also need to track which teams use Caffe2 so that we can put the initialization code there.
With this diff, the initialization code is put in the predictor constructor, only enabled for mobile phones. This way, we can always enable QPL logging.
Once we do this, we can check how many times Caffe2 inference is called in production, and which models are more popular in production. This way, we can prioritize our effort supporting those models.
Will clean up the old code calling the init in the product in a separate diff.
* add padding op for sparse length tensor
to pad length-based sparse tensor with padding_value
* Add conv_op with cudaconvnet engine
Add conv_op with cudaconvnet engine
* [numa] Fix simple NUMA copy benchmark
Move XavierFill into init_net and also compute BW
* call roundf (device function) instead of round (host function)
* [caffe2_benchmark][observer] Make caffe2_benchmark use its own observer
1. Add ClearGlobalNetObservers()
2. Make caffe2_benchmark use its own observer and observer_reporter
* [detectron] Use roundf instead of round in the detectron module ops
* allow K larger than number of elements in top k op
one use case is to use this op together with PackSegments for sparse tensors, where the number of elements in each slice is not statistically defined.
* add ChannelShuffle DNNLOWP op
* fixup math_cpu.cc break
* Add support to TensorRT
* Removed License header
* Bind input/output by position
* Comments
* More comments
* Add benchmark
* Add warning for performance degradation on large batch
* Address comments
* comments
This diff is added to support the ProfileObserver in order to differentiate operators in the stepnet properly. Since copy() is only used in the context of RNNs, the name has been changed to reflect that.
Summary:
This will help releasing models that are using Caffe2 but have their own operator implementations and extensions. More detailed docs to arrive later. Let's see what contbuild says.
Closes https://github.com/caffe2/caffe2/pull/1378
Differential Revision: D6155045
Pulled By: Yangqing
fbshipit-source-id: 657a4c8de2f8e095bad5ed5db5b3e476b2a877e1
