Skip to content

fix: replace deprecated Nemotron Nano model - #166

Merged
rapids-bot[bot] merged 1 commit into
NVIDIA:release/0.1from
yczhang-nv:fix/update-nemotron-nano-model
Jul 31, 2026
Merged

fix: replace deprecated Nemotron Nano model#166
rapids-bot[bot] merged 1 commit into
NVIDIA:release/0.1from
yczhang-nv:fix/update-nemotron-nano-model

Conversation

@yczhang-nv

@yczhang-nv yczhang-nv commented Jul 31, 2026

Copy link
Copy Markdown
Contributor

Overview

Replace the deprecated nvidia/nemotron-3-nano-30b-a3b model with nvidia/nemotron-3-nano-omni-30b-a3b-reasoning across the maintained runtime presets and every dependent repository surface. This keeps the Hermes and Deep Agents NVIDIA API Catalog paths usable without changing their provider, credential, endpoint, or harness configuration.

  • I confirm this contribution is my own work, or I have the right to submit it under this project's license.
  • I searched existing issues and open pull requests, and this does not duplicate existing work.

Details

  • Update the Hermes and Deep Agents CLI preset defaults.
  • Keep generated Python and Rust scaffolds aligned with the updated preset.
  • Update README, Fern documentation, code-review and Harbor examples, and both onboarding notebooks.
  • Update Deep Agents, Hermes, Harbor, ATOF, and ATIF expectations to assert the replacement model.
  • Remove all 28 tracked references to the deprecated identifier.

There are no breaking API or configuration-shape changes.

Validation

  • cargo fmt --all -- --check
  • just test-rust — passed
  • just test-python — 574 passed, 15 skipped
  • npx --prefix docs --no-install fern check --warnings — 0 errors; redirect check skipped because Fern authentication was unavailable
  • Verified nemo-fabric plan --preset hermes and nemo-fabric plan --preset deepagents resolve the replacement model
  • Manually ran the Hermes code-review example with Relay and confirmed ATOF/ATIF model metadata
  • Manually ran the Harbor calculator Hermes Relay job — 1 trial, 0 exceptions, reward 1.0, telemetry validation succeeded
  • Manually ran the Harbor SWE-bench django__django-13741 Hermes job — 1 trial, 0 exceptions, reward 1.0

Where should the reviewer start?

Start in crates/fabric-cli/src/presets.rs, where the maintained Hermes and Deep Agents defaults are defined, then review the corresponding model expectations in crates/fabric-cli/src/scaffold.rs and tests/e2e/test_hermes_e2e.py.

Related Issues: (use one of the action keywords Closes / Fixes / Resolves / Relates to)

  • Relates to: none

Summary by CodeRabbit

  • Documentation

    • Updated quick-start guides, SDK examples, preset references, notebooks, and sample projects to use nvidia/nemotron-3-nano-omni-30b-a3b-reasoning.
    • Refreshed Hermes and Deep Agents preset documentation and example commands with the new default model.
  • Configuration

    • Updated default model selections across generated configurations and example agents.
  • Tests

    • Updated validation and integration coverage to reflect the new model identifier.

Signed-off-by: Yuchen Zhang <yuchenz@nvidia.com>
@coderabbitai

coderabbitai Bot commented Jul 31, 2026

Copy link
Copy Markdown

Review Change Stack

Walkthrough

The default NVIDIA model changes from nvidia/nemotron-3-nano-30b-a3b to nvidia/nemotron-3-nano-omni-30b-a3b-reasoning across CLI presets, documentation, examples, notebooks, and test expectations.

Changes

Nemotron reasoning model update

Layer / File(s) Summary
CLI defaults and scaffold expectations
crates/fabric-cli/src/presets.rs, crates/fabric-cli/src/scaffold.rs
Hermes and Deep Agents presets, along with scaffold assertions, use the updated model identifier.
Documentation and example configurations
README.md, docs/..., examples/...
Quickstarts, SDK examples, Harbor commands, notebooks, and the code review agent configuration use the reasoning model.
Adapter, E2E, and integration expectations
tests/adapters/*, tests/e2e/*, tests/integrations/*
Test inputs and expected generated configuration, metadata, normalized output, and telemetry records use the updated identifier.

Estimated code review effort: 1 (Trivial) | ~5 minutes

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name Status Explanation Resolution
Docstring Coverage ⚠️ Warning Docstring coverage is 20.00% which is insufficient. The required threshold is 80.00%. Write docstrings for the functions missing them to satisfy the coverage threshold.
✅ Passed checks (4 passed)
Check name Status Explanation
Linked Issues check ✅ Passed Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check ✅ Passed Check skipped because no linked issues were found for this pull request.
Title check ✅ Passed The title follows Conventional Commits format, uses an allowed lowercase type, is imperative and concise, and is under 72 characters.
Description check ✅ Passed The description covers the required overview, reviewer starting point, related issue notation, confirmations, validation, and scope of changes.
✨ Finishing Touches
🧪 Generate unit tests (beta)
  • Create PR with unit tests

Comment @coderabbitai help to get the list of available commands.

@yczhang-nv
yczhang-nv marked this pull request as ready for review July 31, 2026 00:49
@yczhang-nv
yczhang-nv requested review from a team as code owners July 31, 2026 00:49
@yczhang-nv yczhang-nv self-assigned this Jul 31, 2026
@github-actions

Copy link
Copy Markdown

@AnuradhaKaruppiah

Copy link
Copy Markdown
Collaborator

/merge

@rapids-bot
rapids-bot Bot merged commit e627734 into NVIDIA:release/0.1 Jul 31, 2026
31 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants