Add function to compute cross_entropy for 2D image #802

wkentaro · 2017-02-20T21:24:57Z

This is used to compute loss for class segmentation task, for example.
Ignoring negative labels is required to use dataset with unlabeled region.

apaszke · 2017-02-21T00:35:56Z

Some time ago we've decided that we want to slowly start moving away from using Nd suffixes for loss and activation functions. Can you add that as a special condition to the cross_entropy function? Just use an if with an input.dim() condition.

torch/nn/functional.py

+        log_p = F.log_softmax(input)  # (N, C, H, W)
+        log_p = log_p.transpose(1, 2).transpose(2, 3).contiguous().view(-1, C)
+        log_p = log_p[target.view(N, H, W, 1).repeat(1, 1, 1, C) >= 0]
+        log_p = log_p.view(-1, C)  # (M, C) where M = (target >= 0).sum()


wkentaro · 2017-02-21T01:02:46Z

Thank you for the comment. I updated the commit.

torch/nn/functional.py

+        log_p = log_p.transpose(1, 2).transpose(2, 3).contiguous().view(-1, C)
+        log_p = log_p[target.view(N, H, W, 1).repeat(1, 1, 1, C) >= 0]
+        log_p = log_p.view(-1, C)  # (M, C) where M = (target >= 0).sum()
+        mask = target >= 0  # (M,)


torch/nn/functional.py

+        if size_average:
+            loss /= mask.sum().data[0]
+        else:
+            loss /= N


wkentaro · 2017-03-16T10:30:27Z

The total_weight argument seems the rescaling factor in element-wise.
Is this correct?

apaszke · 2017-03-16T11:08:08Z

I don't think it's always the case and I'd rather not expose this buffer. Can you remove the additional arg?

wkentaro · 2017-03-16T12:13:32Z

Ok, I removed it.

apaszke · 2017-03-16T16:34:09Z

Thanks!

wkentaro · 2017-03-19T00:29:34Z

@apaszke BTW, how can we apply masking in computation of cross_entropy? For example, I'd like to ignore -1 in the target.

apaszke · 2017-03-19T07:42:22Z

You can shift all targets one index forward and set the weight of the first class to 0.

fmassa · 2017-03-19T10:53:09Z

@apaszke This is a solution, but maybe not ideal as it forces the user to have an extra class for classification (the ignored class), and then at test time it needs to narrow all but the selected class before applying softmax (or we could end up having ignored classes as our prediction).
It's possible, but maybe not super convenient.

wkentaro · 2017-04-07T08:45:30Z

Are there any examples of using gpu kernel?
I think I can try implementing that functionality with it.

fmassa · 2017-04-07T09:53:55Z

The GPU kernels are implemented in https://github.com/pytorch/pytorch/blob/master/torch/lib/THCUNN/SpatialClassNLLCriterion.cu and the C function calling it in https://github.com/pytorch/pytorch/blob/master/torch/lib/THCUNN/generic/SpatialClassNLLCriterion.cu
But the same changes would need to be done in the corresponding CPU implementation, and then the python interface would need to be changed. But before doing it, we should see if it's something we would want to add to core or not (because the weighting can mimic the masking, but it's not super convenient for a few reasons mentioned above)

wkentaro · 2017-04-07T11:47:22Z

Thanks!
I see. Currently, I'm using advanced indexing of boolean array to extract values of interest (0 or positive label values in my case) from image, which returns 1d array, and then computing cross_entropy loss:
https://github.com/wkentaro/pytorch-fcn/blob/c5fe29ac8235befd186ef608fa4c6f5c51f56269/torchfcn/trainer.py#L25-L27
But I'm not sure if masking in kernel is better than that. Maybe it is faster than the boolean array indexing?

onnx/onnx@485b787

…orch#802) onnx/onnx@485b787

…27efbc Previous import was 403ccfbd0161c38f0834413d790bad0874afbf9a Included changes: - **[69894f2](onnx/onnx@69894f2)**: Use op schema.all tensor types in random like definitions (pytorch#865) <Scott McKay> - **[b9d6b90](onnx/onnx@b9d6b90)**: Clarify random like operators (pytorch#846) <Scott McKay> - **[fc6b5fb](onnx/onnx@fc6b5fb)**: Refactor shape inference implementation (pytorch#855) <anderspapitto> - **[b7d8dc8](onnx/onnx@b7d8dc8)**: fix cmake warning message (pytorch#863) <Eric S. Yu> - **[f585c5d](onnx/onnx@f585c5d)**: add pytorch-operator test for tile (pytorch#831) <Wenhao Hu> - **[993fe70](onnx/onnx@993fe70)**: add install step (pytorch#832) <Eric S. Yu> - **[68bc26c](onnx/onnx@68bc26c)**: add type inference for traditional ml ops except classifier ops. (pytorch#857) <Ke Zhang> - **[9cc0cda](onnx/onnx@9cc0cda)**: fix string representation of scalar types (pytorch#858) <G. Ramalingam> - **[1078925](onnx/onnx@1078925)**: fix y in pow test case to scalar (pytorch#852) <Wenhao Hu> - **[c66fb6f](onnx/onnx@c66fb6f)**: Add some math function shape inference (pytorch#845) <anderspapitto> - **[ff667d1](onnx/onnx@ff667d1)**: Refactor return type and docs for ONNXIFI_BACKEND_DIRECTX_ID (pytorch#853) <Marat Dukhan> - **[11c6876](onnx/onnx@11c6876)**: clear initializer names when clear initializer (pytorch#849) <Wenhao Hu> - **[73c34ae](onnx/onnx@73c34ae)**: Clarify FeatureVectorizer description. (pytorch#843) <Scott McKay> - **[1befb9b](onnx/onnx@1befb9b)**: Remove useless text in docs (pytorch#850) <Lu Fang> - **[e84788f](onnx/onnx@e84788f)**: Fix SELU attributes' default values (pytorch#839) <Lu Fang> - **[ebac046](onnx/onnx@ebac046)**: Add tile test case (pytorch#823) <Wenhao Hu> - **[8b7a925](onnx/onnx@8b7a925)**: a few more shape inference functions (pytorch#772) <anderspapitto> - **[9718f42](onnx/onnx@9718f42)**: Make the coefficient non optional for LinearClassifier (pytorch#836) <Jaliya Ekanayake> - **[ef083d0](onnx/onnx@ef083d0)**: Add save_tensor and load_tensor functions for Protos (pytorch#770) <Lu Fang> - **[45ceb55](onnx/onnx@45ceb55)**: Check if CMAKE_BUILD_TYPE set before project(). (pytorch#812) <Sergii Dymchenko> - **[4b3d2b0](onnx/onnx@4b3d2b0)**: [WIP] reenable shape inference tests (pytorch#834) <anderspapitto> - **[22d17ee](onnx/onnx@22d17ee)**: RNN tests: LSTM, GRU, SimpleRNN (pytorch#739) <Peyman Manikashani> - **[de65b95](onnx/onnx@de65b95)**: dimension denotation (pytorch#443) <Tian Jin> - **[eccc76e](onnx/onnx@eccc76e)**: fix field number issue in onnx operator proto and enable its build (pytorch#829) <Ke Zhang> - **[d582beb](onnx/onnx@d582beb)**: disable shape inference test to unbreak ci (pytorch#830) <Lu Fang> - **[485b787](onnx/onnx@485b787)**: function proto for composite op. (pytorch#802) <Ke Zhang> - **[cd58928](onnx/onnx@cd58928)**: specify defaults for attributes of Affine op (pytorch#820) <G. Ramalingam> - **[7ee2cf9](onnx/onnx@7ee2cf9)**: merge the dummy backend back into the main one (pytorch#743) <anderspapitto> - **[1c03a5a](onnx/onnx@1c03a5a)**: [Proposal] ONNX Interface for Framework Integration (previously ONNX Backend API) header and docs (pytorch#551) <Marat Dukhan> - **[3769a98](onnx/onnx@3769a98)**: Rename real model test case from VGG-16 to ZFNet (pytorch#821) <Lu Fang>

* [bootcamp] Improve "Shape" operator to support axes specification To improve .shape operator of Caffe2 to support x.shape(tensor, axes), which takes an optional int array "axes" as input. For example, x.shape(tensor, [1, 0]) will return the dimension for axis 1 and 0 following the specified order. For current version, "axes" input allows duplications and can have arbitrary length. * Back out "Add barrier net that runs before training nets" Original commit changeset: b373fdc9c30f. Need additional changes to some callers to support barrier failures. * Change warning to verbose log to reduce log spam The `LOG(WARNING)` was a bit spammy for regular use so lets just make it a `VLOG`. * Extract the shared code from different caffe2_benchmark binaries The OSS benchmark and Internal benchmark will share most functions in the benchmark. * Support MFR in sequence training As titled. * Make knowledge distillation work with using logged prediction feature as teacher label. 1) Add loading raw dense feature as teacher label. 2) Optional calibration function for teacher label 3) Add teacher label into generic unit test 4) Deprecated TTSN workflow version using feature_options to config teacher label * [C2/CUDA]: unjoined cross entropy sigmoid as desc * Add async_scheduling executor into deferrable_net_exec_test Add async_scheduling into tests and fix some exception cases * Fix Event disabled error When disabling event in RNN ops make sure we don't call Finish on disabled event from op's RunAsync * cuda ensure cpu output op can handle both TensorCPU and TensorCUDA as desc. * [C2 Core] Infer input device option in C2 hypothesis_test checkers Improve how we default input blob device options. Previously it defaults as where op lives but it is not necessarily the case. For example: CopyCPUToGPU * [C2 Op]SplitByLengthsOp CPU/GPU implementation [C2 Op]SplitByLengthsOp CPU/GPU implementation * fix undefined symbol error not sure why we're getting undefined symbol even with link_whole = True Need to figure out why but need this workaround for now * Add tools in DAIPlayground platform to help debugging models Add additional tools to allow Plauground override individual method defined in AnyExp. This will allow user to create module that specificly change certain default method behavior. An example included in this diff is deactivating test model and checkpointing. When debugging any model problems, switching off components helps me quickly narrow down the location of the bug. The technique is extensively used in task T27038712 (Steady memory increase in EDPM, eventually resulting in gloo/cuda.cu:34: out of memory) * add shape and type inference for int8 conversion operator * Fix flaky test for group_norm Fix flaky test for group_norm * Fix group_norm_op_test flaky Fix group_norm_op_test flaky * Implementation of composite learning rate policy In many state-of-the-arts deep learning works, people use a simple trick to schedule the learning rate: use a fixed learning rate until error plateaus and then switch to a different fixed learning rate, and so on. In this diff, we implemented a simple version of the composite learning rate. The user gives a set of learning rates policies and corresponding iteration nums, and the optimizer will change the learning rate policy based on the number of iterations so far. For example, the user give two learning rate policies, one is FixedLearningRate and PolyLearningRate, with an iteration number of 1k. Then the first 1k iteration, we use FixedLearningRate. For the following iterations, we use PolyLearningRate. * Split two use cases of CachedReader into two classes, DBFileReader and CachedReader # Use Cases: 1). input: DB file -> output: DatasetReader. Use DBFileReader. 2). input: Reader -> build cache DB file -> output: DatasetReader. Use CachedReader. # Changes to CachedReader: 1). Move db_path to the constructor. Because in mock reader. cache will always be built ahead. # Changes to tests: 1). Make a separate TestCase class for CachedReader and DBFileReader. 2). Make it possible to add more test functions by adding setUp, tearDown and _make_temp_path. 3). Make delete db_path more general. `db_path` could be a file for `log_file_db`, but could also be a directory for `leveldb`. * Back out "On Mobile phones, call GlobalInit with no arguments in predictor in case we need to perform initialization" Original commit changeset: 4489c6133f11 * Fix LARS bug Fixed a bug in the LARS implementation which caused all subsequent blobs not using LARS to have the LARS learning rate multiplier applied to them. * [tum] support sparse init & add uniformFill option as title * Propagate exception for async nets Capture the exception when an exception is thrown in async nets and re-throw it after wait(). This allows exceptions to be propagated up to the caller. This diff was a part of D7752068. We split the diff so that C2 core files changes are in a separate diff. * Automatic update of fbcode/onnx to 69894f207dfcd72d1e70497d387201cec327efbc Previous import was 403ccfbd0161c38f0834413d790bad0874afbf9a Included changes: - **[69894f2](onnx/onnx@69894f2)**: Use op schema.all tensor types in random like definitions (#865) <Scott McKay> - **[b9d6b90](onnx/onnx@b9d6b90)**: Clarify random like operators (#846) <Scott McKay> - **[fc6b5fb](onnx/onnx@fc6b5fb)**: Refactor shape inference implementation (#855) <anderspapitto> - **[b7d8dc8](onnx/onnx@b7d8dc8)**: fix cmake warning message (#863) <Eric S. Yu> - **[f585c5d](onnx/onnx@f585c5d)**: add pytorch-operator test for tile (#831) <Wenhao Hu> - **[993fe70](onnx/onnx@993fe70)**: add install step (#832) <Eric S. Yu> - **[68bc26c](onnx/onnx@68bc26c)**: add type inference for traditional ml ops except classifier ops. (#857) <Ke Zhang> - **[9cc0cda](onnx/onnx@9cc0cda)**: fix string representation of scalar types (#858) <G. Ramalingam> - **[1078925](onnx/onnx@1078925)**: fix y in pow test case to scalar (#852) <Wenhao Hu> - **[c66fb6f](onnx/onnx@c66fb6f)**: Add some math function shape inference (#845) <anderspapitto> - **[ff667d1](onnx/onnx@ff667d1)**: Refactor return type and docs for ONNXIFI_BACKEND_DIRECTX_ID (#853) <Marat Dukhan> - **[11c6876](onnx/onnx@11c6876)**: clear initializer names when clear initializer (#849) <Wenhao Hu> - **[73c34ae](onnx/onnx@73c34ae)**: Clarify FeatureVectorizer description. (#843) <Scott McKay> - **[1befb9b](onnx/onnx@1befb9b)**: Remove useless text in docs (#850) <Lu Fang> - **[e84788f](onnx/onnx@e84788f)**: Fix SELU attributes' default values (#839) <Lu Fang> - **[ebac046](onnx/onnx@ebac046)**: Add tile test case (#823) <Wenhao Hu> - **[8b7a925](onnx/onnx@8b7a925)**: a few more shape inference functions (#772) <anderspapitto> - **[9718f42](onnx/onnx@9718f42)**: Make the coefficient non optional for LinearClassifier (#836) <Jaliya Ekanayake> - **[ef083d0](onnx/onnx@ef083d0)**: Add save_tensor and load_tensor functions for Protos (#770) <Lu Fang> - **[45ceb55](onnx/onnx@45ceb55)**: Check if CMAKE_BUILD_TYPE set before project(). (#812) <Sergii Dymchenko> - **[4b3d2b0](onnx/onnx@4b3d2b0)**: [WIP] reenable shape inference tests (#834) <anderspapitto> - **[22d17ee](onnx/onnx@22d17ee)**: RNN tests: LSTM, GRU, SimpleRNN (#739) <Peyman Manikashani> - **[de65b95](onnx/onnx@de65b95)**: dimension denotation (#443) <Tian Jin> - **[eccc76e](onnx/onnx@eccc76e)**: fix field number issue in onnx operator proto and enable its build (#829) <Ke Zhang> - **[d582beb](onnx/onnx@d582beb)**: disable shape inference test to unbreak ci (#830) <Lu Fang> - **[485b787](onnx/onnx@485b787)**: function proto for composite op. (#802) <Ke Zhang> - **[cd58928](onnx/onnx@cd58928)**: specify defaults for attributes of Affine op (#820) <G. Ramalingam> - **[7ee2cf9](onnx/onnx@7ee2cf9)**: merge the dummy backend back into the main one (#743) <anderspapitto> - **[1c03a5a](onnx/onnx@1c03a5a)**: [Proposal] ONNX Interface for Framework Integration (previously ONNX Backend API) header and docs (#551) <Marat Dukhan> - **[3769a98](onnx/onnx@3769a98)**: Rename real model test case from VGG-16 to ZFNet (#821) <Lu Fang> * [C2]ReluN Op relu n op. tf reference: https://www.tensorflow.org/api_docs/python/tf/nn/relu6 * Call destructor when assigning a blob value * Add executor overrides Add executor overrides flag to enable migration to async_scheduling executor * Add barrier net that runs before training nets - attempt #2 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. This change was landed previously but caused errors for some EDPM workflows - See https://fb.facebook.com/groups/1426530000692545/permalink/1906766366002237/ - because EDPM assumes any call to CreateOrCloneCommonWorld and Gloo ops are wrapped in exception handlers but in this case exception thrown in the barrier init net is not handled. To address this issue, we add _CreateOrCloneCommonWorld to the param_init_net instead of a new barrier init net. Since errors for param_init_net run is handled gracefully and re-rendezvous, it should fixes the problem. * Handle empty nets in async_scheduling Make sure we don't get stuck on empty nets * use CUDA_ARCH for conditional compile * [C2 fix] infer function for ensure_cpu_output_op * Update group_norm test to reduce flaky test * Fix lr_multiplier for GPU

…orch#802) onnx/onnx@485b787

* [bootcamp] Improve "Shape" operator to support axes specification To improve .shape operator of Caffe2 to support x.shape(tensor, axes), which takes an optional int array "axes" as input. For example, x.shape(tensor, [1, 0]) will return the dimension for axis 1 and 0 following the specified order. For current version, "axes" input allows duplications and can have arbitrary length. * Back out "Add barrier net that runs before training nets" Original commit changeset: b373fdc9c30f. Need additional changes to some callers to support barrier failures. * Change warning to verbose log to reduce log spam The `LOG(WARNING)` was a bit spammy for regular use so lets just make it a `VLOG`. * Extract the shared code from different caffe2_benchmark binaries The OSS benchmark and Internal benchmark will share most functions in the benchmark. * Support MFR in sequence training As titled. * Make knowledge distillation work with using logged prediction feature as teacher label. 1) Add loading raw dense feature as teacher label. 2) Optional calibration function for teacher label 3) Add teacher label into generic unit test 4) Deprecated TTSN workflow version using feature_options to config teacher label * [C2/CUDA]: unjoined cross entropy sigmoid as desc * Add async_scheduling executor into deferrable_net_exec_test Add async_scheduling into tests and fix some exception cases * Fix Event disabled error When disabling event in RNN ops make sure we don't call Finish on disabled event from op's RunAsync * cuda ensure cpu output op can handle both TensorCPU and TensorCUDA as desc. * [C2 Core] Infer input device option in C2 hypothesis_test checkers Improve how we default input blob device options. Previously it defaults as where op lives but it is not necessarily the case. For example: CopyCPUToGPU * [C2 Op]SplitByLengthsOp CPU/GPU implementation [C2 Op]SplitByLengthsOp CPU/GPU implementation * fix undefined symbol error not sure why we're getting undefined symbol even with link_whole = True Need to figure out why but need this workaround for now * Add tools in DAIPlayground platform to help debugging models Add additional tools to allow Plauground override individual method defined in AnyExp. This will allow user to create module that specificly change certain default method behavior. An example included in this diff is deactivating test model and checkpointing. When debugging any model problems, switching off components helps me quickly narrow down the location of the bug. The technique is extensively used in task T27038712 (Steady memory increase in EDPM, eventually resulting in gloo/cuda.cu:34: out of memory) * add shape and type inference for int8 conversion operator * Fix flaky test for group_norm Fix flaky test for group_norm * Fix group_norm_op_test flaky Fix group_norm_op_test flaky * Implementation of composite learning rate policy In many state-of-the-arts deep learning works, people use a simple trick to schedule the learning rate: use a fixed learning rate until error plateaus and then switch to a different fixed learning rate, and so on. In this diff, we implemented a simple version of the composite learning rate. The user gives a set of learning rates policies and corresponding iteration nums, and the optimizer will change the learning rate policy based on the number of iterations so far. For example, the user give two learning rate policies, one is FixedLearningRate and PolyLearningRate, with an iteration number of 1k. Then the first 1k iteration, we use FixedLearningRate. For the following iterations, we use PolyLearningRate. * Split two use cases of CachedReader into two classes, DBFileReader and CachedReader # Use Cases: 1). input: DB file -> output: DatasetReader. Use DBFileReader. 2). input: Reader -> build cache DB file -> output: DatasetReader. Use CachedReader. # Changes to CachedReader: 1). Move db_path to the constructor. Because in mock reader. cache will always be built ahead. # Changes to tests: 1). Make a separate TestCase class for CachedReader and DBFileReader. 2). Make it possible to add more test functions by adding setUp, tearDown and _make_temp_path. 3). Make delete db_path more general. `db_path` could be a file for `log_file_db`, but could also be a directory for `leveldb`. * Back out "On Mobile phones, call GlobalInit with no arguments in predictor in case we need to perform initialization" Original commit changeset: 4489c6133f11 * Fix LARS bug Fixed a bug in the LARS implementation which caused all subsequent blobs not using LARS to have the LARS learning rate multiplier applied to them. * [tum] support sparse init & add uniformFill option as title * Propagate exception for async nets Capture the exception when an exception is thrown in async nets and re-throw it after wait(). This allows exceptions to be propagated up to the caller. This diff was a part of D7752068. We split the diff so that C2 core files changes are in a separate diff. * Automatic update of fbcode/onnx to 69894f207dfcd72d1e70497d387201cec327efbc Previous import was 403ccfbd0161c38f0834413d790bad0874afbf9a Included changes: - **[69894f2](onnx/onnx@69894f2)**: Use op schema.all tensor types in random like definitions (pytorch#865) <Scott McKay> - **[b9d6b90](onnx/onnx@b9d6b90)**: Clarify random like operators (pytorch#846) <Scott McKay> - **[fc6b5fb](onnx/onnx@fc6b5fb)**: Refactor shape inference implementation (pytorch#855) <anderspapitto> - **[b7d8dc8](onnx/onnx@b7d8dc8)**: fix cmake warning message (pytorch#863) <Eric S. Yu> - **[f585c5d](onnx/onnx@f585c5d)**: add pytorch-operator test for tile (pytorch#831) <Wenhao Hu> - **[993fe70](onnx/onnx@993fe70)**: add install step (pytorch#832) <Eric S. Yu> - **[68bc26c](onnx/onnx@68bc26c)**: add type inference for traditional ml ops except classifier ops. (pytorch#857) <Ke Zhang> - **[9cc0cda](onnx/onnx@9cc0cda)**: fix string representation of scalar types (pytorch#858) <G. Ramalingam> - **[1078925](onnx/onnx@1078925)**: fix y in pow test case to scalar (pytorch#852) <Wenhao Hu> - **[c66fb6f](onnx/onnx@c66fb6f)**: Add some math function shape inference (pytorch#845) <anderspapitto> - **[ff667d1](onnx/onnx@ff667d1)**: Refactor return type and docs for ONNXIFI_BACKEND_DIRECTX_ID (pytorch#853) <Marat Dukhan> - **[11c6876](onnx/onnx@11c6876)**: clear initializer names when clear initializer (pytorch#849) <Wenhao Hu> - **[73c34ae](onnx/onnx@73c34ae)**: Clarify FeatureVectorizer description. (pytorch#843) <Scott McKay> - **[1befb9b](onnx/onnx@1befb9b)**: Remove useless text in docs (pytorch#850) <Lu Fang> - **[e84788f](onnx/onnx@e84788f)**: Fix SELU attributes' default values (pytorch#839) <Lu Fang> - **[ebac046](onnx/onnx@ebac046)**: Add tile test case (pytorch#823) <Wenhao Hu> - **[8b7a925](onnx/onnx@8b7a925)**: a few more shape inference functions (pytorch#772) <anderspapitto> - **[9718f42](onnx/onnx@9718f42)**: Make the coefficient non optional for LinearClassifier (pytorch#836) <Jaliya Ekanayake> - **[ef083d0](onnx/onnx@ef083d0)**: Add save_tensor and load_tensor functions for Protos (pytorch#770) <Lu Fang> - **[45ceb55](onnx/onnx@45ceb55)**: Check if CMAKE_BUILD_TYPE set before project(). (pytorch#812) <Sergii Dymchenko> - **[4b3d2b0](onnx/onnx@4b3d2b0)**: [WIP] reenable shape inference tests (pytorch#834) <anderspapitto> - **[22d17ee](onnx/onnx@22d17ee)**: RNN tests: LSTM, GRU, SimpleRNN (pytorch#739) <Peyman Manikashani> - **[de65b95](onnx/onnx@de65b95)**: dimension denotation (pytorch#443) <Tian Jin> - **[eccc76e](onnx/onnx@eccc76e)**: fix field number issue in onnx operator proto and enable its build (pytorch#829) <Ke Zhang> - **[d582beb](onnx/onnx@d582beb)**: disable shape inference test to unbreak ci (pytorch#830) <Lu Fang> - **[485b787](onnx/onnx@485b787)**: function proto for composite op. (pytorch#802) <Ke Zhang> - **[cd58928](onnx/onnx@cd58928)**: specify defaults for attributes of Affine op (pytorch#820) <G. Ramalingam> - **[7ee2cf9](onnx/onnx@7ee2cf9)**: merge the dummy backend back into the main one (pytorch#743) <anderspapitto> - **[1c03a5a](onnx/onnx@1c03a5a)**: [Proposal] ONNX Interface for Framework Integration (previously ONNX Backend API) header and docs (pytorch#551) <Marat Dukhan> - **[3769a98](onnx/onnx@3769a98)**: Rename real model test case from VGG-16 to ZFNet (pytorch#821) <Lu Fang> * [C2]ReluN Op relu n op. tf reference: https://www.tensorflow.org/api_docs/python/tf/nn/relu6 * Call destructor when assigning a blob value * Add executor overrides Add executor overrides flag to enable migration to async_scheduling executor * Add barrier net that runs before training nets - attempt pytorch#2 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. This change was landed previously but caused errors for some EDPM workflows - See https://fb.facebook.com/groups/1426530000692545/permalink/1906766366002237/ - because EDPM assumes any call to CreateOrCloneCommonWorld and Gloo ops are wrapped in exception handlers but in this case exception thrown in the barrier init net is not handled. To address this issue, we add _CreateOrCloneCommonWorld to the param_init_net instead of a new barrier init net. Since errors for param_init_net run is handled gracefully and re-rendezvous, it should fixes the problem. * Handle empty nets in async_scheduling Make sure we don't get stuck on empty nets * use CUDA_ARCH for conditional compile * [C2 fix] infer function for ensure_cpu_output_op * Update group_norm test to reduce flaky test * Fix lr_multiplier for GPU

wkentaro commented Feb 21, 2017

View reviewed changes

wkentaro force-pushed the cross_entropy2d branch from d2e40bd to 0b61aa5 Compare February 21, 2017 01:05

apaszke reviewed Feb 22, 2017

View reviewed changes

torch/nn/functional.py Outdated

if size_average:

loss /= mask.sum().data[0]

else:

loss /= N

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

wkentaro closed this Mar 16, 2017

wkentaro deleted the cross_entropy2d branch March 16, 2017 09:47

wkentaro restored the cross_entropy2d branch March 16, 2017 09:47

wkentaro reopened this Mar 16, 2017

wkentaro added 3 commits March 16, 2017 19:18

Add function to compute cross_entropy for 2D image

8b9a0b1

Support cross_entropy for 2D in cross_entropy func

43a710f

Fix method name: ndim -> dim

184bf07

wkentaro force-pushed the cross_entropy2d branch from 0b61aa5 to 0bfcb9c Compare March 16, 2017 10:23

Add element-wise weight to compute nll_loss

d417728

wkentaro force-pushed the cross_entropy2d branch from 0bfcb9c to d417728 Compare March 16, 2017 10:26

Remove total_weight arg

b9ed14e

apaszke approved these changes Mar 16, 2017

View reviewed changes

apaszke merged commit 7654b3f into pytorch:master Mar 16, 2017

wkentaro deleted the cross_entropy2d branch March 16, 2017 16:36

onnxbot added a commit that referenced this pull request Apr 26, 2018

[auto] Update onnx to 485b787 - function proto for composite op. (#802)

eb4154a

onnx/onnx@485b787

Jorghi12 pushed a commit to wsttiger/pytorch that referenced this pull request May 10, 2018

[auto] Update onnx to 485b787 - function proto for composite op. (pyt…

7376e39

…orch#802) onnx/onnx@485b787

weiyangfb pushed a commit to weiyangfb/pytorch that referenced this pull request Jun 11, 2018

[auto] Update onnx to 485b787 - function proto for composite op. (pyt…

401fa5d

…orch#802) onnx/onnx@485b787

ezyang added the open source label Jun 24, 2019

jjsjann123 pushed a commit to jjsjann123/pytorch that referenced this pull request Apr 11, 2021

Make sure Stmt is visited once (pytorch#802)

8e713ac

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add function to compute cross_entropy for 2D image #802

Add function to compute cross_entropy for 2D image #802

wkentaro commented Feb 20, 2017 •

edited

apaszke commented Feb 21, 2017

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

wkentaro commented Feb 21, 2017

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

wkentaro commented Mar 16, 2017

apaszke commented Mar 16, 2017

wkentaro commented Mar 16, 2017

apaszke commented Mar 16, 2017

wkentaro commented Mar 19, 2017

apaszke commented Mar 19, 2017

fmassa commented Mar 19, 2017

wkentaro commented Apr 7, 2017

fmassa commented Apr 7, 2017

wkentaro commented Apr 7, 2017 •

edited

Add function to compute cross_entropy for 2D image #802

Add function to compute cross_entropy for 2D image #802

Conversation

wkentaro commented Feb 20, 2017 • edited

apaszke commented Feb 21, 2017

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

wkentaro commented Feb 21, 2017

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

wkentaro commented Mar 16, 2017

apaszke commented Mar 16, 2017

wkentaro commented Mar 16, 2017

apaszke commented Mar 16, 2017

wkentaro commented Mar 19, 2017

apaszke commented Mar 19, 2017

fmassa commented Mar 19, 2017

wkentaro commented Apr 7, 2017

fmassa commented Apr 7, 2017

wkentaro commented Apr 7, 2017 • edited

wkentaro commented Feb 20, 2017 •

edited

wkentaro commented Apr 7, 2017 •

edited