skills.rest
yo-steven
yo-stevenCOMMUNITY
Explore all skills by this author →
GitHub

nemo-gym-debugging

Classify Nemo Gym rollout failures and guide evidence-first debugging.

Updated May 23, 2026
Search Tags:#debugging#nemo gym#ray slurm#jsonl schema#vllm readiness#tool-call validation#cache resume
One-click install
npx skills add https://github.com/yo-steven/skills-exploration-20260522 --skill nemo-gym-debugging
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: nemo-gym-debugging
Source: https://github.com/yo-steven/skills-exploration-20260522/tree/main/skills/NeMo-Gym/nemo-gym-debugging
Command: npx skills add https://github.com/yo-steven/skills-exploration-20260522 --skill nemo-gym-debugging

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires argparse, json, sys, collections, pathlib, typing, and includes scripts (resource) and references (resource) components.

What problem does it solve?

Nemo Gym runs and reward profiling jobs can fail in ways that look similar at the top level (empty outputs, nested 500s, schema/verifier errors, stalled serving, or stale cache), but the root cause lives in a specific layer.

Core Features & Use Cases

  • Layered failure classification: Guides you to distinguish infra (Slurm/Ray/container/filesystem), model serving (vLLM readiness/throughput), config, data/schema, verifier/runtime, cache/resume issues, and throughput bottlenecks.
  • Deterministic debug order: Provides a step-by-step escalation ladder that prioritizes high-signal checks (job state/logs, vLLM readiness, server readiness, tool/sandbox readiness, materialized vs source timestamps, output/profiling counts, verifier exceptions).
  • Targeted offline tooling for tool-call schema: Offers a bundled static checker script for Nemo Gym tool-call JSONL to catch vLLM/Outlines grammar/schema incompatibilities before expensive rollouts.
  • References for copy-paste diagnostics: Includes request-boundary visibility and reusable log/JSONL inspection snippets to recover identity and evidence when nested errors hide the true provider/verifier response.

Quick Start

Run the tool-call JSONL static sanity check to quickly rule out schema/grammar incompatibilities before re-running your Nemo Gym job.

Frequently Asked Questions about nemo-gym-debugging

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I debug Nemo Gym rollout collection failures?▼

Debug Nemo Gym rollout failures by following an ordered workflow that checks job state, vLLM readiness, materialized inputs, output shape, and verifier exceptions to isolate the failing layer.

Why does my Nemo Gym reward profiling job stall with empty outputs?▼

Empty outputs often stem from infra instability, vLLM readiness issues, or stale cache artifacts; classify the failure layer by checking job logs and server readiness first to isolate the root cause.

How do I validate Nemo Gym tool-call JSONL schemas offline?▼

Validate tool-call JSONL schemas offline using the bundled static checker script to catch vLLM and Outlines grammar incompatibilities before running expensive rollouts.

Does this debugging workflow support Slurm and Ray job instability?▼

Yes, the workflow supports Slurm and Ray job instability by guiding evidence-first investigation of job states, container issues, and filesystem problems to distinguish infra failures from model serving errors.

What is the best way to diagnose vLLM model-server readiness issues in Nemo Gym?▼

Diagnose vLLM readiness issues by checking server readiness and read-throughput as high-signal checks early in the debug ladder, ensuring the model server is responsive before deeper analysis.

How do I fix Hydra config miscomposition causing nested 500 errors in Nemo Gym?▼

Fix Hydra config miscomposition by inspecting request-boundary visibility and verifier exceptions to recover the true provider response, then validate schema compatibility using bundled offline references.

Related Skills

Explore More Software Engineering Skills→
NVIDIANVIDIA
2.8k

add-benchmark

Scaffold NeMo-Gym benchmarks with resources servers, data pipelines, and YAML configs.

Official
Advanced
yo-stevenyo-steven

nemo-gym-reward-profiling

Join materialized inputs with rollout results for Nemo Gym reward profiling.

Community
Intermediate
niznik-devniznik-dev
16

Run Experiment

Orchestrate SLURM fine-tuning jobs with torchtune and evaluate with inspect-ai.

Community
Advanced
SignmanalSignmanal
1

rest-graphql-debug

Diagnose REST and GraphQL API failures by isolating the failing request layer.

Community
Advanced