Tommy Yu 2b63b7a0a5 Support GPU version of Spatial Batch Norm (#11711) ...

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

2018-10-18 11:22:13 -07:00

..

docs

adapting caffe2 operator docs generator to pytorch url

2018-10-11 12:55:06 -07:00

examples

Add resnext model to OSS (#11468)

2018-09-12 15:59:20 -07:00

helpers

[Caffe2] Add full impl of GroupNorm (#7058)

2018-04-29 11:26:40 -07:00

ideep

Implementation MomentumSGD/MomentumSGDUpdate operators for mkl-dnn (#11686)

2018-09-27 13:39:59 -07:00

layers

Revert D10123465: Set the correct engine name for position weighted pooling when fp16 is used for training

2018-10-15 01:53:48 -07:00

mint

move flags to c10 (#12144)

2018-10-04 02:09:56 -07:00

mkl

Re-enable IDEEP graph rewrite test (#12661)

2018-10-15 15:50:28 -07:00

modeling

diagnose option: get_entry to print a whole row (#11308)

2018-09-06 21:26:30 -07:00

models

Add resnext model to OSS (#11468)

2018-09-12 15:59:20 -07:00

onnx

Enhance shape inference in ONNXIFI transformer (#12685)

2018-10-16 14:15:46 -07:00

operator_test

Support scales as inputs in ResizeNearest (#12720)

2018-10-17 23:08:53 -07:00

predictor

Rename cuda_gpu_id to device_id in DeviceOption (#12456)

2018-10-09 15:54:04 -07:00

rnn

Fix a typo

2018-03-30 21:00:44 -07:00

serialized_test

Use tempfile during serialized test comparison (#12021)

2018-09-25 20:55:45 -07:00

test

Update from Facebook (#8887)

2018-06-26 14:55:48 -07:00

trt

ONNXIFI transform (#9569)

2018-07-20 15:09:59 -07:00

__init__.py

caffe2::DeviceType -> at::DeviceType (#11254)

2018-09-05 16:28:09 -07:00

_import_c_extension.py

Completely remove build_aten and use_aten (#10469)

2018-08-20 20:26:42 -07:00

allcompare_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

attention.py

[Caffe2] Update elementwise ops to support numpy style boradcast (#8070)

2018-06-05 15:49:16 -07:00

benchmark_generator.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

binarysize.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

brew_test.py

Move tanh function to math (#9328)

2018-07-11 13:59:50 -07:00

brew.py

[Caffe2] Changes done inside Facebook (#6378)

2018-04-10 21:11:43 -07:00

build.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

cached_reader.py

Update from Facebook (#8887)

2018-06-26 14:55:48 -07:00

caffe_translator_test.py

Remove caffe1 specific proto (#10380)

2018-08-10 11:10:26 -07:00

caffe_translator.py

Fix bug in caffe_translator tool (#10056)

2018-10-11 13:13:12 -07:00

checkpoint_test.py

Revert D9566744: [New Checkpoint] Kill the dummy TaskOutput when task.get_step() (#11164)

2018-08-31 22:25:57 -07:00

checkpoint.py

Create class constant for string literal 'blob_names'

2018-08-24 22:11:43 -07:00

CMakeLists.txt

Add nomnigraph bindings

2018-08-28 12:40:16 -07:00

cnn.py

Rename cuda_gpu_id to device_id in DeviceOption (#12456)

2018-10-09 15:54:04 -07:00

compatibility.py

migrating deprecated calls without abc module for containers (#11515)

2018-09-13 15:09:22 -07:00

context_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

context.py

Resolve name conflict of ContextManager (#7244)

2018-06-22 00:41:51 -04:00

control_ops_grad.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

control_ops_util.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

control_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

control.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

convert_test.py

New serialization format (#12384)

2018-10-16 16:36:58 -07:00

convert.py

New serialization format (#12384)

2018-10-16 16:36:58 -07:00

convnet_benchmarks_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

convnet_benchmarks.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

core_gradients_test.py

[Caffe2] Update elementwise ops to support numpy style boradcast (#8070)

2018-06-05 15:49:16 -07:00

core_test.py

Rename cuda_gpu_id to device_id in DeviceOption (#12456)

2018-10-09 15:54:04 -07:00

core.py

Rename cuda_gpu_id to device_id in DeviceOption (#12456)

2018-10-09 15:54:04 -07:00

crf_predict.py

Move crf in caffe2 from fb to oss (#12200)

2018-10-01 18:31:41 -07:00

crf_viterbi_test.py

Move crf in caffe2 from fb to oss (#12200)

2018-10-01 18:31:41 -07:00

crf.py

Productionize CRF layer in PyText (#10362)

2018-08-22 00:25:26 -07:00

data_parallel_model_test.py

Support GPU version of Spatial Batch Norm (#11711)

2018-10-18 11:22:13 -07:00

data_parallel_model.py

Support GPU version of Spatial Batch Norm (#11711)

2018-10-18 11:22:13 -07:00

data_workers_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

data_workers.py

Fixed log message (#10874)

2018-09-05 09:55:52 -07:00

dataio_test.py

Update from Facebook (#8887)

2018-06-26 14:55:48 -07:00

dataio.py

Fixing stop condition on composite reader (#9888)

2018-08-20 03:02:20 -07:00

dataset.py

Update from facebook (#7855)

2018-05-29 11:38:02 -07:00

db_file_reader.py

Update from Facebook (#8887)

2018-06-26 14:55:48 -07:00

db_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

device_checker.py

Update from facebook (#7451)

2018-05-10 23:14:27 -07:00

dlpack.h

Support for DLPack in Python op

2017-12-21 17:02:16 -08:00

dyndep.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

embedding_generation_benchmark.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

experiment_util.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

extension_loader.py

Completely remove build_aten and use_aten (#10469)

2018-08-20 20:26:42 -07:00

functional_test.py

Add support for specifying device_option in Functional (#9619)

2018-07-24 14:41:59 -07:00

functional.py

Caffe2 Functional enforcing inplace output (#10797)

2018-08-23 22:42:47 -07:00

fused_8bit_rowwise_conversion_ops_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

gradient_check_test.py

[Caffe2] Fix gradient_check on in-place ops (#8828)

2018-06-25 15:25:56 -07:00

gradient_checker.py

make the variable declaration closer to usage

2018-10-12 12:07:08 -07:00

gru_cell.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

hsm_util.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

hypothesis_test_util.py

Rename cuda_gpu_id to device_id in DeviceOption (#12456)

2018-10-09 15:54:04 -07:00

hypothesis_test.py

Enable CircleCI for Linux jobs (#12389)

2018-10-08 17:09:37 -07:00

ideep_test_util.py

[feature request] [Caffe2] Enable MKLDNN support for inference (#6699)

2018-04-22 21:58:14 -07:00

layer_model_helper.py

parallize the dense part in event models

2018-08-22 22:40:07 -07:00

layer_model_instantiator.py

[caffe2] Fbcode to GitHub sync (#6208)

2018-04-02 16:35:27 -07:00

layer_parameter_sharing_test.py

[layer model] remove duplicated init ops

2018-03-27 18:10:39 -07:00

layer_test_util.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

layers_test.py

Allow use substitute ops for LayerNorm (#12177)

2018-10-11 17:36:10 -07:00

lengths_reducer_fused_8bit_rowwise_ops_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

lengths_reducer_rowwise_8bit_ops_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

lstm_benchmark.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

memonger_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

memonger.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

mkl_test_util.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

model_device_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

model_helper_test.py

keep net type info when generating model complete net (#11032)

2018-09-04 21:10:06 -07:00

model_helper.py

Rename cuda_gpu_id to device_id in DeviceOption (#12456)

2018-10-09 15:54:04 -07:00

modifier_context.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

mpi_python.cc

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

muji_test.py

[Caffe2] Support non peer access in muji and fix bug when reduced_affix is empty (#6896)

2018-06-05 03:14:43 +08:00

muji.py

Rename cuda_gpu_id to device_id in DeviceOption (#12456)

2018-10-09 15:54:04 -07:00

net_builder_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

net_builder.py

Update from Facebook (#6692)

2018-04-17 23:36:40 -07:00

net_drawer.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

net_printer_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

net_printer.py

Rename cuda_gpu_id to device_id in DeviceOption (#12456)

2018-10-09 15:54:04 -07:00

nomnigraph_test.py

Improve Python API with the addition of pythonic setters/getters

2018-10-16 00:57:54 -07:00

nomnigraph.py

Enable mapping from operatordef to converted node for debugging

2018-10-16 21:03:28 -07:00

normalizer_context.py

Update from Facebook (#8887)

2018-06-26 14:55:48 -07:00

normalizer_test.py

Update from Facebook (#8887)

2018-06-26 14:55:48 -07:00

normalizer.py

Enable alternative LayerNorm impl in FisherGan (#12178)

2018-10-11 17:36:11 -07:00

numa_benchmark.py

Update from facebook (#6871)

2018-04-23 15:01:56 -07:00

numa_test.py

Rename cuda_gpu_id to device_id in DeviceOption (#12456)

2018-10-09 15:54:04 -07:00

observer_test.py

Fix RNN scoping situation

2018-02-07 17:35:29 -08:00

optimizer_context.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

optimizer_test_util.py

Implementation of Wngrad optimizer caffe2 python wrapper and unit test on least square regression (#9001)

2018-07-13 18:54:52 -07:00

optimizer_test.py

Add Adadelta optimizer to caffe2 (#9088)

2018-07-24 20:09:21 -07:00

optimizer.py

Rename cuda_gpu_id to device_id in DeviceOption (#12456)

2018-10-09 15:54:04 -07:00

parallel_workers_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

parallel_workers.py

Update from facebook (#7696)

2018-05-19 23:10:48 -07:00

parallelize_bmuf_distributed_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

pipeline_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

pipeline.py

SNNTest with Data Preproc Service (#11707)

2018-09-17 21:25:49 -07:00

predictor_constants.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

pybind_state_dlpack.cc

codemod: caffe::float16 -> at::Half (#11785)

2018-09-20 18:55:19 -07:00

pybind_state_dlpack.h

Rename cuda_gpu_id to device_id in DeviceOption (#12456)

2018-10-09 15:54:04 -07:00

pybind_state_gpu.cc

Remove many caffe2::TIndex and replace them with int64_t (#11943)

2018-09-22 18:11:04 -07:00

pybind_state_hip.cc

Remove many caffe2::TIndex and replace them with int64_t (#11943)

2018-09-22 18:11:04 -07:00

pybind_state_ideep.cc

Guard NumPy usage using USE_NUMPY (#11798)

2018-10-04 12:11:02 -07:00

pybind_state_int8.cc

Guard NumPy usage using USE_NUMPY (#11798)

2018-10-04 12:11:02 -07:00

pybind_state_nomni.cc

Enable mapping from operatordef to converted node for debugging

2018-10-16 21:03:28 -07:00

pybind_state_registry.cc

Move registry fully to c10 (#12077)

2018-09-27 03:09:54 -07:00

pybind_state_registry.h

Move registry fully to c10 (#12077)

2018-09-27 03:09:54 -07:00

pybind_state.cc

New serialization format (#12384)

2018-10-16 16:36:58 -07:00

pybind_state.h

Remove StaticContext (#12547)

2018-10-10 19:41:03 -07:00

python_op_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

queue_util.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

record_queue.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

recurrent.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

regularizer_context.py

Update from Facebook (#8887)

2018-06-26 14:55:48 -07:00

regularizer_test.py

Add GroupL1Norm regularizer (#9115)

2018-07-06 13:26:09 -07:00

regularizer.py

Add GroupL1Norm regularizer (#9115)

2018-07-06 13:26:09 -07:00

rnn_cell.py

Rewrite http://pytorch.org -> https://pytorch.org throughout project (#12636)

2018-10-15 13:03:27 -07:00

schema_test.py

Add util function from core type to dtype (#10716)

2018-08-21 10:55:19 -07:00

schema.py

Add util function from core type to dtype (#10716)

2018-08-21 10:55:19 -07:00

scope_test.py

Update from Facebook (#8887)

2018-06-26 14:55:48 -07:00

scope.py

Update from Facebook (#8887)

2018-06-26 14:55:48 -07:00

session_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

session.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

sparse_to_dense_mask_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

sparse_to_dense_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

task.py

Revert D9566744: [New Checkpoint] Kill the dummy TaskOutput when task.get_step() (#11164)

2018-08-31 22:25:57 -07:00

test_util.py

Enable CircleCI for Linux jobs (#12389)

2018-10-08 17:09:37 -07:00

text_file_reader.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

timeout_guard.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

toy_regression_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

transformations_test.py

nomnigraph - easy - some code cleanup for transformations_test (#12101)

2018-10-01 11:31:08 -07:00

transformations.py

Enable Conv fusion optimizations in optimizeForIdeep (#9255)

2018-07-16 21:28:50 -07:00

tt_core_test.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

tt_core.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

utils.py

migrating deprecated calls without abc module for containers (#11515)

2018-09-13 15:09:22 -07:00

visualize.py

Remove Apache headers from source.

2018-03-27 13:10:18 -07:00

workspace_test.py

Add workspace.RunPlanInBackground (#9637)

2018-07-20 14:56:12 -07:00

workspace.py

New serialization format (#12384)

2018-10-16 16:36:58 -07:00