nemotron-3-ultra-text2sql-lora
Run the Nemotron-3 Ultra Text2SQL LoRA fine-tuning tutorial (NeMo Megatron-Bridge) end-to-end for the user on their SLURM cluster: data prep, distributed checkpoint conversion, and packed LoRA fine-tuning of the 550B hybrid Mamba-Transformer MoE, ending at a saved adapter. Use when the user wants to run this cookbook, fine-tune Nemotron-3 Ultra with LoRA, or adapt the notebook to their own cluster.
70
85%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
View guide
Low
Low-risk findings worth noting
Low
Low-risk findings.
2 low severity findings. Worth noting, but not necessarily harmful.
W011: Third-party content exposure detected (indirect prompt injection risk).
What this means
The skill exposes the agent to untrusted, user-generated content from public third-party sources, creating a risk of indirect prompt injection. This includes browsing arbitrary URLs, reading social media posts or forum comments, and analyzing content from unknown websites.
Why it was flagged
Source text is ingested at runtime from the public Hugging Face datasets `xu3kev/BIRD-SQL-data-train` and `meowterspace45/bird-sql-train-with-reasoning` via `dataprep.py` → `DatasetBIRD._load_dataset()` / `DatasetBIRDReasoning._load_dataset()` using `datasets.load_dataset`, and then the dataset’s free-text fields (schema/question/evidence/SQL/reasoning_trace) are converted into LLM prompts via `apply_chat_template` and concatenated into `training.jsonl` without selecting a specific pre-vetted record.
W012: Unverifiable external dependency detected (runtime URL that controls agent).
What this means
The skill fetches instructions or code from an external URL at runtime, and the fetched content directly controls the agent’s prompts or executes code. This dynamic dependency allows the external source to modify the agent’s behavior without any changes to the skill itself.
Why it was flagged
The skill calls datasets from Hugging Face at runtime (e.g., load_dataset("xu3kev/BIRD-SQL-data-train") and load_dataset("meowterspace45/bird-sql-train-with-reasoning")), and those remote dataset contents (including reasoning traces) are injected into the constructed chat prompts used for training, so the fetched content directly controls prompts at runtime.
- Repository
- NVIDIA-NeMo/Nemotron
- Commit
5277655
- Audited
- Security analysis
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.