facebookresearch/fairseq

Auto create commandline args from yamls with fairseq-hydra-train

Open

#5,035 opened on Mar 22, 2023

View on GitHub
 (0 comments) (0 reactions) (0 assignees)Python (6,224 forks)batch import
enhancementhelp wantedneeds triage

Repository metrics

Stars
 (29,107 stars)
PR merge metrics
 (No merged PRs in 30d)

Description

🚀 Feature Request

Though Fairseq is such a powerful framework, I got some problems when I used fairseq-hydra-train. If I want to define an arg in the config yaml file, I have to define it in some dataclass first. This is cumbersome and slow compared with the hydra's ability of auto-creating commandline args. Furthermore, this makes it impossible to use hydra's ability of multi-level nested parameters. How can I access the full ability of hydra with fairseq?

Contributor guide

Research direction
Investigate how to use Hydra's compose API to dynamically create command line arguments from YAML configs without pre defining dataclasses. Look into fairseq hydra train's config handling and see if you can override or extend it to support arbitrary YAML parameters.
Tech stack
python
Domain
cli
Issue type
Feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
3
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
Half day
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Clear
Prerequisites
Pythonfairseqhydra
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
35