NVIDIA / NVIDIA/apex

Set AmpState.verbosity to prevent unnecessary prints

Open
#803 0 comments 1 reaction 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
9k
Forks
1.5k
Avg merge
2d 4h
Merged PRs (30d)
3

Description

Currently each time amp.init is called it prints out lines like the following polluting the stdout despite passing verbose=False in amp.init. This is because when AmpState is created and it sets hardcoded verbosity of 1 and has no way in API to turn this off. I think amp.init should set AmpState.verbosity to 0 if verbous=False.

enabled                : True
opt_level              : O2
cast_model_type        : torch.float16
patch_torch_functions  : False
keep_batchnorm_fp32    : True
master_weights         : True
loss_scale             : dynamic
Processing user overrides (additional kwargs that are not None)...
After processing overrides, optimization options are:
enabled                : True
opt_level              : O2
cast_model_type        : torch.float16
patch_torch_functions  : False
keep_batchnorm_fp32    : True
master_weights         : True
loss_scale             : dynamic

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Search the Python implementation of amp.init and AmpState, focusing on how the verbose argument is handled when AmpState is created. Reproduce initialization with verbose=False and verify that the optimization settings are no longer printed while verbose=True retains the expected output.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
machine-learning
Issue type
Bug
Difficulty
2/5
Estimated time
1-3 hours
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
50/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.