firefox-translations-training
firefox-translations-training copied to clipboard
Publish training charts from Taskcluster
This includes publishing:
- live training logs to W&B dashboards
I assume we'll have separate publishing scripts for other things.
Let's use Taskgraph transforms not to pollute Taskcluster kinds with even more logic.
We can start working on this now.
I identified these next steps to discuss together:
- build a dedicated command in the tracking package to support specific tasks (training & evaluation) and use a Taskcluster secret to get the Weight & Biases token
- Setup the publication code through a Taskcluster transform
- Patch training shell scripts to capture output and publish from logs
- when an env variable is set
- scripts to patch
-
pipeline/train/train.sh
-
pipeline/train/spm-vocab.sh
-
- Identify other tasks to track & publish
After discussion with Evgeny, we can start to build the dedicated script to interact with Taskcluster secret & logs.
One small correction: pipeline/train/spm-vocab.sh
trains a vocab and we don't want to track this in W&B. So only train.sh
should be patched.
@evabardou will work on Taskcluster secret support.
We can start experimenting log parsing from CI using --in-stream
option too