chunyuan 8b11d81058 [Re-landing 68111] Add JIT graph fuser for oneDNN Graph API (Preview4.1) ... Re-landing https://github.com/pytorch/pytorch/pull/68111 ## Description Preview4 PR of this [RFC](https://github.com/pytorch/pytorch/issues/49444). On the basis of https://github.com/pytorch/pytorch/pull/50256, the below improvements are included: - The [preview4 release branch](https://github.com/oneapi-src/oneDNN/releases/tag/graph-v0.4.1) of the oneDNN Graph API is used - The fuser now works with the profiling graph executor. We have inserted type check nodes to guard the profiled tensor properties. ### User API: The optimization pass is disabled by default. Users could enable it by: ``` torch.jit.enable_onednn_fusion(True) ``` ### Performance: [pytorch/benchmark](https://github.com/pytorch/benchmark) tool is used to compare the performance: - SkyLake 8180 (1 socket of 28 cores):  - SkyLake 8180 (single thread):  \* By mapping hardswish to oneDNN Graph, it’s 8% faster than PyTorch JIT (NNC + OFI) \** We expect performance gain after mapping transpose, contiguous & view to oneDNN graph ops ### Directory structure of the integration code Fuser-related code are placed under: ``` torch/csrc/jit/codegen/onednn/ ``` Optimization pass registration is done in: ``` torch/csrc/jit/passes/onednn_graph_fuser.h ``` CMake for the integration code is: ``` caffe2/CMakeLists.txt ``` ## Limitations - In this PR, we have only supported the optimization on Linux platform. The support on Windows and MacOS will be enabled as the next step. - We have only optimized the inference use case. Pull Request resolved: https://github.com/pytorch/pytorch/pull/74596 Approved by: https://github.com/malfet		2022-04-29 01:01:33 +00:00
..
init.cpp	[Re-landing 68111] Add JIT graph fuser for oneDNN Graph API (Preview4.1)	2022-04-29 01:01:33 +00:00
init.h
module_python.h
pybind_utils.cpp	Introducing SymInt to Pytorch (for tracing size arithmetic) (master rebase) (#74861)	2022-03-31 21:59:59 +00:00
pybind_utils.h	Get rid of TorchScript sparse tensor is experimental warning. (#73874)	2022-03-09 15:45:24 +00:00
pybind.h
python_arg_flatten.cpp
python_arg_flatten.h
python_custom_class.cpp
python_custom_class.h
python_dict.cpp
python_dict.h
python_interpreter.cpp
python_ir.cpp	Introducing SymInt to Pytorch (for tracing size arithmetic) (master rebase) (#74861)	2022-03-31 21:59:59 +00:00
python_ir.h
python_ivalue.h
python_list.cpp	[BE] Fix pybind deprecation warnings (#72376)	2022-02-07 18:33:32 +00:00
python_list.h	Fix sign-compare violations in python_list.h	2022-04-01 19:15:51 +00:00
python_sugared_value.cpp	[JIT] support parameterlist iteration	2022-04-21 18:51:27 +00:00
python_sugared_value.h	sup torch script parameterlist	2022-04-20 20:53:07 +00:00
python_tracer.cpp
python_tracer.h
python_tree_views.cpp	Reland "Make debug_pkl smaller by only emitting unique traces." (#73368)	2022-04-18 22:34:21 +00:00
python_tree_views.h
script_init.cpp	sup torch script parameterlist	2022-04-20 20:53:07 +00:00
script_init.h
update_graph_executor_opt.cpp
update_graph_executor_opt.h