docs: add "What's Hot" section to README#1992
Conversation
Greptile SummaryThis PR adds a "What's Hot" section to the top of
Confidence Score: 5/5Documentation-only change with no code modifications; safe to merge once the three linked doc URLs are manually confirmed to resolve. The change is limited to inserting a Markdown table and blockquote in README.md. No logic, configuration, or runtime behaviour is affected. The only open question is whether the three external documentation URLs resolve correctly, which the PR's own test plan flags as an outstanding check. No files require special attention beyond confirming the three external doc links are live before merge. Important Files Changed
Flowchart%%{init: {'theme': 'neutral'}}%%
flowchart TD
A[README.md] --> B[Badges & Header]
B --> C[NeMo suite tagline]
C --> D["## What's Hot (NEW)"]
D --> D1["Curator on Slurm → Slurm Deployment Guide"]
D --> D2["Audio Curation → Audio Guide"]
D --> D3["Inference Server → Inference Server docs"]
D --> E["## Updates (release log)"]
E --> F["## What You Can Build"]
F --> G[Rest of README...]
Reviews (3): Last reviewed commit: "Merge branch 'main' into lbliii/readme-w..." | Re-trigger Greptile |
Surfaces the latest capabilities developers should know about — Slurm, Audio, and Inference Server — at the top of the README. Addresses NVIDIA-NeMo#1552. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com> Signed-off-by: Lawrence Lane <llane@nvidia.com>
b03bfbd to
32f6c6b
Compare
|
|
||
| | Feature | What it unlocks | Read this | | ||
| |---------|-----------------|-----------| | ||
| | **Curator on Slurm** | Run multi-node Ray pipelines on HPC clusters — text, image, video, and audio workloads at scale | [Slurm Deployment Guide](https://docs.nvidia.com/nemo/curator/latest/admin/deployment/slurm/multi-node-ray.html) | |
There was a problem hiding this comment.
This link is broken for me.
| | Feature | What it unlocks | Read this | | ||
| |---------|-----------------|-----------| | ||
| | **Curator on Slurm** | Run multi-node Ray pipelines on HPC clusters — text, image, video, and audio workloads at scale | [Slurm Deployment Guide](https://docs.nvidia.com/nemo/curator/latest/admin/deployment/slurm/multi-node-ray.html) | | ||
| | **Audio Curation** | Curate speech datasets for ASR and multimodal training — transcription, WER filtering, and quality assessment | [Audio Guide](https://docs.nvidia.com/nemo/curator/latest/curate-audio/index.html) | |
There was a problem hiding this comment.
Agree that we should list audio curation but not sure this is the best description since transcription, WER filtering, and quality assessment were the first things added to audio. Since then several other pipelines have been added: https://github.com/NVIDIA-NeMo/Curator/commits/main/nemo_curator/stages/audio
Maybe most notably, ALM (Audio Language Model) data curation, quality filtering via AudioDataFilterStage, and audio tagging and speaker diarization. (Claude can probably summarize these concepts better than me.)
Summary
Context
Requested by Arham in #docs Slack thread — developers landing on the README should immediately see the latest capabilities worth trying. The three highlighted features were called out by Arham; cc @nemo-curator-leads if anything else should be added before merge (candidates: Cosmos-Embed1 video embeddings, Nemotron-CC recipe, simplified
ResourcesAPI).Test plan
🤖 Generated with Claude Code