Skip to content

Address follow-up review comments on custom training containers docs - #5420

Merged
Eliza Farley (elizafarley) merged 4 commits into
mainfrom
eliza/custom-containers-followups
Oct 15, 2026
Merged

Eliza Farley (elizafarley) merged 4 commits into
mainfrom
eliza/custom-containers-followups

Conversation

@elizafarley

@elizafarley Eliza Farley (elizafarley) commented Oct 14, 2026 •

Copy link
Copy Markdown
Contributor

Summary

Follow-up to #5407, addressing Lillian Hwang-Geddes (@lvhg)'s review comments that landed after merge.

  • docs/cli/reference.md: the training-script test-local summary line now lists the optional --dataset-file, --model-output-directory, and --custom-args flags instead of [...named args]. The section example also shows --dataset-file. (comment)
  • docs/cli/datasets-and-training.md: notes that --org-id is optional for viam train containers list and that without it the command lists only Viam-managed containers. The test-local example shows --dataset-file. (comment)
  • docs/train/custom-containers.md: clarifies that jobs skip installing dependencies with pip only when the container includes every dependency the training script needs (comment), and adds --dataset-file to the test-local example (comment).

Also, from review on this PR:

  • docs/cli/datasets-and-training.md: the test-local example now shows --model-output-directory, like the other examples (comment).
  • docs/train/custom-containers.md: removed the base image guidance and the example Dockerfile, which used an unverified nvidia/cuda tag. Any image that meets the page's listed requirements works, so the section now says to build an image that meets them, starting from any base image.

Verified against code

rdk cli/app.go on main: test-local takes --dataset-file (default dataset.jsonl), --model-output-directory (default .), and --custom-args, all optional. MLListContainers in cli/ml_training.go passes --org-id straight through with no default-org fallback, so omitting it returns only Viam-managed containers.

Checks

prettier 3.2.5, markdownlint, vale, frontmatter_rules.py, and make build-prod all pass.

🤖 Generated with Claude Code

- Add the optional --dataset-file, --model-output-directory, and
  --custom-args flags to the training-script test-local summary line.
- Show --dataset-file in test-local examples.
- Note that --org-id is optional for train containers list.
- Clarify that jobs skip dependency installs only when the container
  includes every dependency the script needs.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@netlify

netlify Bot commented Oct 14, 2026 •

Copy link
Copy Markdown

✅ Deploy Preview for viam-docs ready!

Name Link
🔨 Latest commit 32fb99e
🔍 Latest deploy log https://app.netlify.com/projects/viam-docs/deploys/6acfcecdb3e8b797e1338c9c
😎 Deploy Preview https://deploy-preview-5420--viam-docs.netlify.app
📱 Preview on mobile
Toggle QR Code...

QR Code

Use your smartphone camera to open QR code link.
Lighthouse
Lighthouse
1 paths audited
Performance: 38 (🔴 down 4 from production)
Accessibility: 100 (no change from production)
Best Practices: 100 (no change from production)
SEO: 92 (no change from production)
PWA: 60 (no change from production)
View the detailed breakdown and full score reports

To edit notification comments on pull requests, go to your Netlify project configuration.

@viambot viambot added the safe to build This pull request is marked safe to build from a trusted zone label Oct 14, 2026
Comment thread docs/cli/datasets-and-training.md Outdated
--dataset-root=./my-dataset/ \
--training-script-directory=./my-training-script/ \
--dataset-file=dataset.jsonl \
--container-id=<container-id>

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is there a reason you left out model-output-directory here?

…UDA tag

- Show --model-output-directory in the datasets-and-training test-local
  example, matching the other test-local examples.
- Replace the specific nvidia/cuda tag in the example Dockerfile with a
  placeholder, link the nvidia/cuda Docker Hub page, and tell readers to
  pick a tag with the CUDA version their libraries need.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Any image that meets the listed requirements works, so don't prescribe
a CUDA base image or show an unverified example Dockerfile.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@elizafarley
Eliza Farley (elizafarley) merged commit fe7312d into main Oct 15, 2026
14 checks passed
@elizafarley
Eliza Farley (elizafarley) deleted the eliza/custom-containers-followups branch October 15, 2026 13:51
@github-actions

Copy link
Copy Markdown

🔎💬 Inkeep AI search and chat service is syncing content for source 'Viam Docs'

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

safe to build This pull request is marked safe to build from a trusted zone

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants