Commit Graph

6 Commits

Author SHA1 Message Date
Han Qi
3d37f5b052 Make debug_pkl smaller by only emitting unique traces. (#72596)
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/72596

debug_pkl file inside of pytorch's .pt file consists of a list of SourceRanges. Each SourceRange points to a Source which is a stack track, filename, and start, end numbers. Those are emitted in debug_pkl file as strings.

Since many SourceRange shares the same source, the string for trace can be deduped.

The newer format saves a set of unique traces in a tuple, then each SourceRange will save the offset of it's trace w.r.t. position in that tuple. (i.e. manually applying dictionary compression).

The above helps with smaller file size. On loading, if we copy each trace to Source as string the runtime memory would still blowup.
To mitigate this, we use SourceView directly instead of source which will take the reference of string inside of Deserializer and make that into string_view. This is safe because Deserializer is hold by Unpickler by shared_ptr, and Unpickler is also hold by shared_ptr by another Source object. That Source object will be alive during the model construction.

Test Plan:
unit test

Took original file (312271638_930.predictor.disagg.local); loaded with `torch.jit.load` save again with `torch.jit.save`. Unzip both, look at contents:
```
[qihan@devvm5585.vll0 ~]$ du archive -h
4.0K    archive/xl_model_weights
3.7M    archive/extra
8.0K    archive/code/__torch__/caffe2/torch/fb/model_transform/splitting
8.0K    archive/code/__torch__/caffe2/torch/fb/model_transform
8.0K    archive/code/__torch__/caffe2/torch/fb
8.0K    archive/code/__torch__/caffe2/torch
8.0K    archive/code/__torch__/caffe2
20M     archive/code/__torch__/torch/fx/graph_module
20M     archive/code/__torch__/torch/fx
8.0K    archive/code/__torch__/torch/classes
20M     archive/code/__torch__/torch
20M     archive/code/__torch__
20M     archive/code
2.7M    archive/constants
35M     archive
[qihan@devvm5585.vll0 ~]$ du resaved -h
4.0K    resaved/extra
8.0K    resaved/code/__torch__/caffe2/torch/fb/model_transform/splitting
8.0K    resaved/code/__torch__/caffe2/torch/fb/model_transform
8.0K    resaved/code/__torch__/caffe2/torch/fb
8.0K    resaved/code/__torch__/caffe2/torch
8.0K    resaved/code/__torch__/caffe2
1.3M    resaved/code/__torch__/torch/fx/graph_module
1.3M    resaved/code/__torch__/torch/fx
8.0K    resaved/code/__torch__/torch/classes
1.4M    resaved/code/__torch__/torch
1.4M    resaved/code/__torch__
1.4M    resaved/code
2.7M    resaved/constants
13M     resaved
[qihan@devvm5585.vll0 ~]$
```

Reviewed By: JasonHanwen

Differential Revision: D33994011

fbshipit-source-id: 8e6224c6e942e91c3403f686c8f0937d1002ed41
(cherry picked from commit a7014dd4029308c95007f362a57c31796d686647)
2022-02-24 09:31:16 +00:00
gmagogsfm
147f7559b1 Add SourceView which doesn't own source text as base class of Source (#65309)
Summary:
This would save the cost copying text from stack to heap in some cases (like
parsing function schema during loading phase of libtorch.so)

Pull Request resolved: https://github.com/pytorch/pytorch/pull/65309

Reviewed By: swolchok

Differential Revision: D31060315

Pulled By: gmagogsfm

fbshipit-source-id: 0caf7a688b40df52bb4388c5191d1a42351d6f1a
2021-10-18 23:17:22 -07:00
Zhengxu Chen
2b0ec9c3cf Reapply "[jit] Implement ScriptProfile to collect instruction profiles." (#58783)
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/58783

This reverts commit fc804b5def.

Test Plan: Imported from OSS

Reviewed By: gmagogsfm

Differential Revision: D28617037

Pulled By: zhxchen17

fbshipit-source-id: 645de2ede20500a5c218d6ec3c7faae94de37a14
2021-05-24 18:23:21 -07:00
Edward Yang
fc804b5def Revert D28133579: [jit] Implement ScriptProfile to collect instruction profiles.
Test Plan: revert-hammer

Differential Revision:
D28133579 (034a238bab)

Original commit changeset: e7e30e961513

fbshipit-source-id: 5a7756468b4f2eeed24d2abb7b52ab46d081a95e
2021-05-21 08:18:40 -07:00
Zhengxu Chen
034a238bab [jit] Implement ScriptProfile to collect instruction profiles. (#57397)
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/57397

Introduces two main classes in C++ runtime:

ScriptProfile is the implementation for enalbing and disabling interpreter
profiling in C++. This should be only used from Python, and we will add
corresponding Python API in the next diff.

InstructionSpan is a utility class to instrument execution of each single
instruction. A start timestamp is recorded in the consturctor, and an end
timestamp is recorded in the destructor. During destruction, this will send
runtime data to all enabled ScriptProfile instances.

Test Plan:
build/bin/test_jit --gtest_filter='ScriptProfileTest.Basic'

Imported from OSS

Reviewed By: gmagogsfm

Differential Revision: D28133579

fbshipit-source-id: e7e30e96151367022793ab3ad323f01c51ad4a3b
2021-05-20 14:11:03 -07:00
Zhengxu Chen
3a3959d253 [jit] Add a utility class SourceRef to represent Source as keys (#57396)
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/57396

A new type SourceRef is introduced to represent a unique identifier to source
text. The type holds refcount to underlying source, and supports comparators
and hash functions, such that it can be used in C++ and Python maps. In later
diffs we will use this to aggregate and print profiling information.

Test Plan: Imported from OSS

Reviewed By: nikithamalgifb

Differential Revision: D28133578

fbshipit-source-id: c3d5199a8269c5006c85a145b281bcaaf3e2dc1c
2021-05-18 18:20:53 -07:00