Skip to content

docs(retrieval): add fine-tuning guide#2306

Open
oliverholworthy wants to merge 26 commits into
mainfrom
oholworthy/docs/retrieval-finetuning-guide
Open

docs(retrieval): add fine-tuning guide#2306
oliverholworthy wants to merge 26 commits into
mainfrom
oholworthy/docs/retrieval-finetuning-guide

Conversation

@oliverholworthy

Copy link
Copy Markdown
Contributor

What does this PR do ?

Adds a retrieval fine-tuning guide and supporting retrieval utilities/tests for bi-encoder and cross-encoder workflows, including custom data formats, hard-negative mining, export/reload handoff, cache validation, and operational safety checks.

Changelog

  • Add a retrieval fine-tuning guide covering bi-encoder and cross-encoder configs, custom data, validation, mining, retraining, export, and troubleshooting.
  • Expand retrieval dataset docs and examples for corpus-backed JSON, inline JSONL, Hugging Face sources, and mined outputs.
  • Harden hard-negative mining cache validation, relative corpus path handling, atomic output writes, distributed cache reuse, and saved encoder loading.
  • Add retrieval data utility coverage for materialization, mined-negative audits, unrolled positives, encoder metadata export, and mining cache safety.

Before your PR is "Ready for review"

Pre checks:

  • Make sure you read and followed Contributor guidelines
  • Did you write any new necessary tests?
  • Did you add or update any necessary documentation?

Validated locally:

  • uv run ruff format ...
  • uv run ruff check --fix ...
  • uv run pytest tests/unit_tests/recipes/test_mine_hard_negatives.py tests/unit_tests/datasets/llm/test_materialize_hf_retrieval_subset.py -q
  • uv run pytest tests/unit_tests/_transformers/test_retrieval.py tests/unit_tests/recipes/test_mine_hard_negatives.py tests/unit_tests/datasets/llm/test_audit_mined_negatives.py tests/unit_tests/datasets/llm/test_materialize_hf_retrieval_subset.py tests/unit_tests/models/bi_encoder/test_llama_bidirectional_model.py -q
  • uv run --group docs sphinx-build -b dummy docs /tmp/automodel-docs-review-final-final -W --keep-going
  • git diff --check

Additional Information

  • Opened as draft while reviewers take a look.
  • Final bounded persona review found no confirmed P0/P1 blockers.
  • CI may require /ok to test <commit-sha> depending on branch trust/signature policy.

rnyak and others added 25 commits May 22, 2026 22:07
Signed-off-by: Ronay Ak <ronaya@nvidia.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Signed-off-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
@copy-pr-bot

copy-pr-bot Bot commented May 22, 2026

Copy link
Copy Markdown

Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually.

Contributors can view more details about this message here.

@akoumpa akoumpa marked this pull request as ready for review June 1, 2026 22:11

@jgerh jgerh left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

Completed tech pubs review of .md files and provided some copyedits and suggested text revisions.

Comment thread docs/guides/dataset-overview.md Outdated
Comment thread docs/guides/dataset-overview.md Outdated
Comment thread docs/guides/dataset-overview.md Outdated
Comment thread docs/guides/dataset-overview.md Outdated
Comment thread docs/guides/dataset-overview.md Outdated
Comment thread examples/retrieval/bi_encoder/llama_embed_nemotron_8b/README.md Outdated
Comment thread examples/retrieval/bi_encoder/llama_embed_nemotron_8b/README.md Outdated
Comment thread examples/retrieval/bi_encoder/llama_embed_nemotron_8b/README.md Outdated
Comment thread examples/retrieval/bi_encoder/llama_embed_nemotron_8b/README.md Outdated
Comment thread examples/retrieval/bi_encoder/llama_embed_nemotron_8b/README.md Outdated

@jgerh jgerh left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

Completed tech pubs review of .md files and provided some copyedits and suggested text revisions.

Co-authored-by: jgerh <163925524+jgerh@users.noreply.github.com>
lbliii added a commit that referenced this pull request Jun 8, 2026
Ports Oliver Holworthy's runnable bi-encoder/cross-encoder retrieval
fine-tuning guide (open PR #2306, authored in pre-migration Sphinx
format) into the Fern MDX structure as
docs/guides/llm/retrieval-finetune.mdx, registered under Recipes with
slug retrieval-finetune.

This resolves the 5 dangling /recipes-e2e-examples/retrieval-finetune
links in the embedding/reranker model-coverage pages (added by #2392),
which referenced a guide that was never merged to main. Converted MyST
constructs to Fern: :::{warning} -> <Warning>, {download} roles -> GitHub
source links, and ../*.md links -> version-agnostic Fern paths.

#2306 can be closed as superseded once this lands.

Co-Authored-By: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Signed-off-by: Lawrence Lane <llane@nvidia.com>
akoumpa pushed a commit that referenced this pull request Jun 11, 2026
…2391)

* docs(fern): relocate Fern under docs/ and remove legacy Sphinx tree

Move the Fern build infrastructure from the top-level fern/ directory into
docs/fern/ (config, theme, components, version navs, and the frozen 0.4.0
snapshot), delete the legacy Sphinx scaffolding (conf.py, autodoc parser,
project.json, versions1.json), and retire the Sphinx publish workflows in
favor of Fern CI. The bleeding-edge docs/ markdown pages are intentionally
left as .md here; they are converted to .mdx in the follow-up PR so the
conversion lands as a per-file rename+edit diff. docs.yml ships the latest +
v0.4 trains (both mount the frozen 0.4.0 pages); the nightly train is enabled
in the follow-up PR alongside the converted content.

Signed-off-by: Lawrence Lane <llane@nvidia.com>
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* docs(fern): convert nightly docs from Markdown to MDX under docs/ (#2392)

Rename each docs/*.md page to docs/*.mdx and apply the Sphinx/MyST -> Fern
MDX conversion (YAML frontmatter, version-agnostic links, Fern components),
so every page shows as a single rename+edit diff that git log --follow can
trace. Enable the nightly version train (nightly.yml + docs.yml versions)
mounting the converted top-level docs/ tree. Page content matches origin/main,
including #2289's checkpoint-export updates.

Signed-off-by: Lawrence Lane <llane@nvidia.com>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>

* ci(fern): source frozen doc versions from a docs-archive branch at build time (#2400)

Keep only the nightly tree (top-level docs/) on main and move the frozen
v0.4 GA snapshot off main onto the long-lived `docs-archive` branch. A new
`stitch-fern-versions` composite action restores the archived pages/ trees
into the working copy before every Fern build, so the multi-version site
still publishes intact.

Why a build-time stitch is required: Fern has no native cross-ref sourcing
for a version train's prose — `fern generate --docs` reads the single local
working tree, and publish is a full-site snapshot (a train missing from the
tree is unpublished). `fern check` likewise fatals on any referenced path
that is absent. So the frozen pages must be physically present at build time.

- Add .github/actions/stitch-fern-versions: ref-agnostic restore via
  `git fetch --depth=1 origin <ref>` + `git restore --source=FETCH_HEAD`.
  The registry value is an opaque git ref, so a branch (default) or an
  immutable docs/v* tag both work; hard-fails if the ref/subtree is missing.
- Run the stitch in publish-fern-docs.yml, fern-docs-ci.yml, and
  fern-docs-preview-build.yml (before fern check / generate / artifact
  upload). docs.yml and the nav YAMLs always come from the live checkout.
- Remove docs/fern/versions/v0.4/pages (143 files) from main; preserved on
  the docs-archive branch. Gitignore the restore path.
- Add `make docs-stitch` (ARCHIVE_REF overridable) and make docs / docs-check
  / docs-preview depend on it for local dev.
- Update the fern-docs SKILL.md and docs/fern/README.md: archived-versions
  flow, cutting-a-new-train steps, CI diagram, debugging, and references.

Scope: Fern fern/ pipeline only. The legacy S3/Akamai Sphinx publish
(release-docs.yml, release-nightly-docs.yml) is intentionally untouched.

Stacked on #2392 (convert-docs-to-mdx); the migration stack must merge first.

Signed-off-by: Lawrence Lane <llane@nvidia.com>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>

* fix(docs): drop unrelated source churn from the relocation

The relocation branch was committed on a stale base (predating main's
save_consolidated migration), so its diff carried 121 unrelated files under
nemo_automodel/, tests/, examples/, and tools/ that would have reverted main
on merge. Restore them to main's current versions. The docs relocation, MDX
conversion, and archive/stitch content are unchanged.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Signed-off-by: Lawrence Lane <llane@nvidia.com>

* docs(fern): drop stray distributed-training skill edit

Signed-off-by: Lawrence Lane <llane@nvidia.com>

* docs(fern): render landing quick links as a Cards grid

Replace the inline pipe-separated emoji link list on the docs landing
page with a Fern <Cards> grid (body-less cards), matching the page's
existing navigation idiom. Labels and hrefs are unchanged.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Signed-off-by: Lawrence Lane <llane@nvidia.com>

* docs(fern): render landing quick links as small Button row

Swap the Cards grid for a row of Fern <Button small outlined> links —
slimmer, native component, labels and hrefs unchanged.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Signed-off-by: Lawrence Lane <llane@nvidia.com>

* docs(fern): add Quick Links header above the landing button row

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Signed-off-by: Lawrence Lane <llane@nvidia.com>

* docs(fern): pin Recipes section slug to recipes-e2e-examples

The "Recipes & E2E Examples" section had no explicit slug, so Fern
auto-derived it as "recipes-e-2-e-examples" (the slugifier splits the
"E2E" digit). Every internal link — the landing quick-link buttons and
the "Recipes & Guides" table — targets /recipes-e2e-examples/..., so the
whole section 404'd. Pin the slug across nightly, v0.4, and latest navs
so the intended URL resolves on every version train.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Signed-off-by: Lawrence Lane <llane@nvidia.com>

* docs(fern): fix broken internal links surfaced by full link audit

- nemotron-3-ultra: /launcher/slurm -> /job-launchers/slurm-cluster
- vlm model coverage: step-3-7 -> step-3-7-flash (actual page slug)
- convert GitHub links to removed Sphinx docs/*.md into internal links:
  docs/launcher/slurm.md -> /job-launchers/slurm-cluster (8x),
  docs/guides/llm/dataset.md -> /datasets/text-dataset,
  docs/guides/dataset-overview.md -> /datasets/overview,
  docs/guides/vlm/dataset.md -> /datasets/multi-modal-dataset (2x)

Audited all 142 internal link targets against the nav route set; only
the (not-yet-written) Embedding/Reranking guide remains, tracked
separately.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Signed-off-by: Lawrence Lane <llane@nvidia.com>

* docs(fern): add retrieval fine-tuning guide (port of #2306)

Ports Oliver Holworthy's runnable bi-encoder/cross-encoder retrieval
fine-tuning guide (open PR #2306, authored in pre-migration Sphinx
format) into the Fern MDX structure as
docs/guides/llm/retrieval-finetune.mdx, registered under Recipes with
slug retrieval-finetune.

This resolves the 5 dangling /recipes-e2e-examples/retrieval-finetune
links in the embedding/reranker model-coverage pages (added by #2392),
which referenced a guide that was never merged to main. Converted MyST
constructs to Fern: :::{warning} -> <Warning>, {download} roles -> GitHub
source links, and ../*.md links -> version-agnostic Fern paths.

#2306 can be closed as superseded once this lands.

Co-Authored-By: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Signed-off-by: Lawrence Lane <llane@nvidia.com>

* docs(readme): restore save_consolidated: final in checkpoint sketch

Same stale-base leak as the distributed-training SKILL.md fix: the
relocation commit snapshotted a pre-#2289 copy where save_consolidated
was still `true`, diverging from main's `final`. Restore `final` to match
main (flagged in review on #2391).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Signed-off-by: Lawrence Lane <llane@nvidia.com>

* test: scan docs/model-coverage for .mdx after fern relocation

Signed-off-by: Dong Hyuk Chang <9426164+thomasdhc@users.noreply.github.com>
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

---------

Signed-off-by: Lawrence Lane <llane@nvidia.com>
Signed-off-by: Dong Hyuk Chang <9426164+thomasdhc@users.noreply.github.com>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Co-authored-by: Dong Hyuk Chang <9426164+thomasdhc@users.noreply.github.com>
HuiyingLi pushed a commit to khazic/Automodel_lao that referenced this pull request Jun 13, 2026
…VIDIA-NeMo#2391)

* docs(fern): relocate Fern under docs/ and remove legacy Sphinx tree

Move the Fern build infrastructure from the top-level fern/ directory into
docs/fern/ (config, theme, components, version navs, and the frozen 0.4.0
snapshot), delete the legacy Sphinx scaffolding (conf.py, autodoc parser,
project.json, versions1.json), and retire the Sphinx publish workflows in
favor of Fern CI. The bleeding-edge docs/ markdown pages are intentionally
left as .md here; they are converted to .mdx in the follow-up PR so the
conversion lands as a per-file rename+edit diff. docs.yml ships the latest +
v0.4 trains (both mount the frozen 0.4.0 pages); the nightly train is enabled
in the follow-up PR alongside the converted content.

Signed-off-by: Lawrence Lane <llane@nvidia.com>
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* docs(fern): convert nightly docs from Markdown to MDX under docs/ (NVIDIA-NeMo#2392)

Rename each docs/*.md page to docs/*.mdx and apply the Sphinx/MyST -> Fern
MDX conversion (YAML frontmatter, version-agnostic links, Fern components),
so every page shows as a single rename+edit diff that git log --follow can
trace. Enable the nightly version train (nightly.yml + docs.yml versions)
mounting the converted top-level docs/ tree. Page content matches origin/main,
including NVIDIA-NeMo#2289's checkpoint-export updates.

Signed-off-by: Lawrence Lane <llane@nvidia.com>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>

* ci(fern): source frozen doc versions from a docs-archive branch at build time (NVIDIA-NeMo#2400)

Keep only the nightly tree (top-level docs/) on main and move the frozen
v0.4 GA snapshot off main onto the long-lived `docs-archive` branch. A new
`stitch-fern-versions` composite action restores the archived pages/ trees
into the working copy before every Fern build, so the multi-version site
still publishes intact.

Why a build-time stitch is required: Fern has no native cross-ref sourcing
for a version train's prose — `fern generate --docs` reads the single local
working tree, and publish is a full-site snapshot (a train missing from the
tree is unpublished). `fern check` likewise fatals on any referenced path
that is absent. So the frozen pages must be physically present at build time.

- Add .github/actions/stitch-fern-versions: ref-agnostic restore via
  `git fetch --depth=1 origin <ref>` + `git restore --source=FETCH_HEAD`.
  The registry value is an opaque git ref, so a branch (default) or an
  immutable docs/v* tag both work; hard-fails if the ref/subtree is missing.
- Run the stitch in publish-fern-docs.yml, fern-docs-ci.yml, and
  fern-docs-preview-build.yml (before fern check / generate / artifact
  upload). docs.yml and the nav YAMLs always come from the live checkout.
- Remove docs/fern/versions/v0.4/pages (143 files) from main; preserved on
  the docs-archive branch. Gitignore the restore path.
- Add `make docs-stitch` (ARCHIVE_REF overridable) and make docs / docs-check
  / docs-preview depend on it for local dev.
- Update the fern-docs SKILL.md and docs/fern/README.md: archived-versions
  flow, cutting-a-new-train steps, CI diagram, debugging, and references.

Scope: Fern fern/ pipeline only. The legacy S3/Akamai Sphinx publish
(release-docs.yml, release-nightly-docs.yml) is intentionally untouched.

Stacked on NVIDIA-NeMo#2392 (convert-docs-to-mdx); the migration stack must merge first.

Signed-off-by: Lawrence Lane <llane@nvidia.com>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>

* fix(docs): drop unrelated source churn from the relocation

The relocation branch was committed on a stale base (predating main's
save_consolidated migration), so its diff carried 121 unrelated files under
nemo_automodel/, tests/, examples/, and tools/ that would have reverted main
on merge. Restore them to main's current versions. The docs relocation, MDX
conversion, and archive/stitch content are unchanged.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Signed-off-by: Lawrence Lane <llane@nvidia.com>

* docs(fern): drop stray distributed-training skill edit

Signed-off-by: Lawrence Lane <llane@nvidia.com>

* docs(fern): render landing quick links as a Cards grid

Replace the inline pipe-separated emoji link list on the docs landing
page with a Fern <Cards> grid (body-less cards), matching the page's
existing navigation idiom. Labels and hrefs are unchanged.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Signed-off-by: Lawrence Lane <llane@nvidia.com>

* docs(fern): render landing quick links as small Button row

Swap the Cards grid for a row of Fern <Button small outlined> links —
slimmer, native component, labels and hrefs unchanged.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Signed-off-by: Lawrence Lane <llane@nvidia.com>

* docs(fern): add Quick Links header above the landing button row

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Signed-off-by: Lawrence Lane <llane@nvidia.com>

* docs(fern): pin Recipes section slug to recipes-e2e-examples

The "Recipes & E2E Examples" section had no explicit slug, so Fern
auto-derived it as "recipes-e-2-e-examples" (the slugifier splits the
"E2E" digit). Every internal link — the landing quick-link buttons and
the "Recipes & Guides" table — targets /recipes-e2e-examples/..., so the
whole section 404'd. Pin the slug across nightly, v0.4, and latest navs
so the intended URL resolves on every version train.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Signed-off-by: Lawrence Lane <llane@nvidia.com>

* docs(fern): fix broken internal links surfaced by full link audit

- nemotron-3-ultra: /launcher/slurm -> /job-launchers/slurm-cluster
- vlm model coverage: step-3-7 -> step-3-7-flash (actual page slug)
- convert GitHub links to removed Sphinx docs/*.md into internal links:
  docs/launcher/slurm.md -> /job-launchers/slurm-cluster (8x),
  docs/guides/llm/dataset.md -> /datasets/text-dataset,
  docs/guides/dataset-overview.md -> /datasets/overview,
  docs/guides/vlm/dataset.md -> /datasets/multi-modal-dataset (2x)

Audited all 142 internal link targets against the nav route set; only
the (not-yet-written) Embedding/Reranking guide remains, tracked
separately.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Signed-off-by: Lawrence Lane <llane@nvidia.com>

* docs(fern): add retrieval fine-tuning guide (port of NVIDIA-NeMo#2306)

Ports Oliver Holworthy's runnable bi-encoder/cross-encoder retrieval
fine-tuning guide (open PR NVIDIA-NeMo#2306, authored in pre-migration Sphinx
format) into the Fern MDX structure as
docs/guides/llm/retrieval-finetune.mdx, registered under Recipes with
slug retrieval-finetune.

This resolves the 5 dangling /recipes-e2e-examples/retrieval-finetune
links in the embedding/reranker model-coverage pages (added by NVIDIA-NeMo#2392),
which referenced a guide that was never merged to main. Converted MyST
constructs to Fern: :::{warning} -> <Warning>, {download} roles -> GitHub
source links, and ../*.md links -> version-agnostic Fern paths.

NVIDIA-NeMo#2306 can be closed as superseded once this lands.

Co-Authored-By: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Signed-off-by: Lawrence Lane <llane@nvidia.com>

* docs(readme): restore save_consolidated: final in checkpoint sketch

Same stale-base leak as the distributed-training SKILL.md fix: the
relocation commit snapshotted a pre-NVIDIA-NeMo#2289 copy where save_consolidated
was still `true`, diverging from main's `final`. Restore `final` to match
main (flagged in review on NVIDIA-NeMo#2391).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Signed-off-by: Lawrence Lane <llane@nvidia.com>

* test: scan docs/model-coverage for .mdx after fern relocation

Signed-off-by: Dong Hyuk Chang <9426164+thomasdhc@users.noreply.github.com>
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

---------

Signed-off-by: Lawrence Lane <llane@nvidia.com>
Signed-off-by: Dong Hyuk Chang <9426164+thomasdhc@users.noreply.github.com>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: Oliver Holworthy <1216955+oliverholworthy@users.noreply.github.com>
Co-authored-by: Dong Hyuk Chang <9426164+thomasdhc@users.noreply.github.com>
Signed-off-by: HuiyingLi <willwin.lee@gmail.com>
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