mirror of
https://github.com/saymrwulf/pytorch.git
synced 2026-05-14 20:57:59 +00:00
d549ddfb14
Summary: Use a logger to control output to console. This is useful for hiding out debug/detail messages from the console v/s showing everything together.
Test Plan:
Ran `torchfrtrace` with various switches.
The `-v` verbose swtch
```
torchfrtrace --prefix "trace_" /tmp/ -v
loaded 2 files in 0.2567298412322998s
built groups, memberships
Not all ranks joining collective 3 at entry 2
group info: 0:default_pg
collective: nccl:all_reduce
missing ranks: {1}
input sizes: [[4, 5]]
output sizes: [[4, 5]]
expected ranks: 2
collective state: scheduled
collective stack trace:
<module> at /home/cpio/test/c.py:66
appending a non-matching collective
built collectives, nccl_calls
Groups
id desc size
-------------------- ---------- ------
09000494312501845833 default_pg 2
Memberships
group_id global_rank
-------------------- -------------
09000494312501845833 0
09000494312501845833 1
Collectives
id group_id
---- ----------
0 0
1 0
NCCLCalls
id collective_id group_id global_rank traceback_id collective_type sizes
---- --------------- ---------- ------------- -------------- ----------------- --------
0 0 0 0 0 nccl:all_reduce [[3, 4]]
1 0 0 1 0 nccl:all_reduce [[3, 4]]
2 1 0 0 0 nccl:all_reduce [[3, 4]]
3 1 0 1 0 nccl:all_reduce [[3, 4]]
4 0 0 0 nccl:all_reduce [[4, 5]]
```
Without the verbose switch
```
❯ torchfrtrace --prefix "trace_" /tmp/
Not all ranks joining collective 3 at entry 2
group info: 0:default_pg
collective: nccl:all_reduce
missing ranks: {1}
input sizes: [[4, 5]]
output sizes: [[4, 5]]
expected ranks: 2
collective state: scheduled
collective stack trace:
<module> at /home/cpio/test/c.py:66
```
With the `-j` switch:
```
❯ torchfrtrace --prefix "trace_" /tmp/ -j
Rank 0 Rank 1
------------------------------------------------- -------------------------------------------------
all_reduce(input_sizes=[[3, 4]], state=completed) all_reduce(input_sizes=[[3, 4]], state=completed)
all_reduce(input_sizes=[[3, 4]], state=completed) all_reduce(input_sizes=[[3, 4]], state=completed)
all_reduce(input_sizes=[[4, 5]], state=scheduled)
```
Differential Revision: D65438520
Pull Request resolved: https://github.com/pytorch/pytorch/pull/139656
Approved by: https://github.com/fduwjj
|
||
|---|---|---|
| .. | ||
| alerts | ||
| amd_build | ||
| autograd | ||
| bazel_tools | ||
| build/bazel | ||
| build_defs | ||
| code_analyzer | ||
| code_coverage | ||
| config | ||
| coverage_plugins_package | ||
| dynamo | ||
| flight_recorder | ||
| gdb | ||
| github | ||
| iwyu | ||
| jit | ||
| linter | ||
| lite_interpreter | ||
| lldb | ||
| onnx | ||
| pyi | ||
| rules | ||
| rules_cc | ||
| setup_helpers | ||
| shared | ||
| stats | ||
| test | ||
| testing | ||
| __init__.py | ||
| bazel.bzl | ||
| BUCK.bzl | ||
| BUCK.oss | ||
| build_libtorch.py | ||
| build_pytorch_libs.py | ||
| build_with_debinfo.py | ||
| download_mnist.py | ||
| extract_scripts.py | ||
| gen_flatbuffers.sh | ||
| gen_vulkan_spv.py | ||
| generate_torch_version.py | ||
| generated_dirs.txt | ||
| git_add_generated_dirs.sh | ||
| git_reset_generated_dirs.sh | ||
| nightly.py | ||
| nightly_hotpatch.py | ||
| nvcc_fix_deps.py | ||
| README.md | ||
| render_junit.py | ||
| substitute.py | ||
| update_masked_docs.py | ||
| vscode_settings.py | ||
This folder contains a number of scripts which are used as
part of the PyTorch build process. This directory also doubles
as a Python module hierarchy (thus the __init__.py).
Overview
Modern infrastructure:
- autograd - Code generation for autograd. This includes definitions of all our derivatives.
- jit - Code generation for JIT
- shared - Generic infrastructure that scripts in
tools may find useful.
- module_loader.py - Makes it easier to import arbitrary Python files in a script, without having to add them to the PYTHONPATH first.
Build system pieces:
- setup_helpers - Helper code for searching for third-party dependencies on the user system.
- build_pytorch_libs.py - cross-platform script that builds all of the constituent libraries of PyTorch, but not the PyTorch Python extension itself.
- build_libtorch.py - Script for building libtorch, a standalone C++ library without Python support. This build script is tested in CI.
Developer tools which you might find useful:
- git_add_generated_dirs.sh and git_reset_generated_dirs.sh - Use this to force add generated files to your Git index, so that you can conveniently run diffs on them when working on code-generation. (See also generated_dirs.txt which specifies the list of directories with generated files.)
Important if you want to run on AMD GPU:
- amd_build - HIPify scripts, for transpiling CUDA
into AMD HIP. Right now, PyTorch and Caffe2 share logic for how to
do this transpilation, but have separate entry-points for transpiling
either PyTorch or Caffe2 code.
- build_amd.py - Top-level entry point for HIPifying our codebase.
Tools which are only situationally useful:
- docker - Dockerfile for running (but not developing) PyTorch, using the official conda binary distribution. Context: https://github.com/pytorch/pytorch/issues/1619
- download_mnist.py - Download the MNIST dataset; this is necessary if you want to run the C++ API tests.