[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7B5W7HXSJ22C3DAC4BER52C":3,"skill-same:01M7B5W7HXSJ22C3DAC4BER52C":81},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":29,"ownerHtmlUrl":30,"ownerBlogUrl":31,"ownerDescription":6,"updatedAt":32,"install":33,"resources":46},"01M7B5W7HXSJ22C3DAC4BER52C","modelopt-launching-evals",null,"Run, monitor, analyze, and debug LLM evaluations via nemo-evaluator-launcher. Covers running evaluations, checking status and live progress, debugging failed runs, exporting artifacts and logs, and analyzing results. ALWAYS triggers on mentions of running evaluations, checking progress, debugging failed evals, analyzing or analysing runs or results, run directories or artifact paths on clusters, Slurm job issues, invocation IDs, or inspecting logs (client logs, server logs, SSH to cluster, tail logs, grep logs). Do NOT use for creating or modifying evaluation configs.","NVIDIA\u002FModel-Optimizer","NVIDIA","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F1728152?v=4","01M6BXXZYTRNBXSTRXNNBB356H","NVIDIA Corporation","01M6BXY02X3PRVHNNATDHA2D0W",5604,"Apache-2.0","https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FModel-Optimizer\u002Fblob\u002Fmain\u002Fskills\u002Fmodelopt-launching-evals\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: modelopt-launching-evals\ndescription: Run, monitor, analyze, and debug LLM evaluations via nemo-evaluator-launcher. Covers running evaluations, checking status and live progress, debugging failed runs, exporting artifacts and logs, and analyzing results. ALWAYS triggers on mentions of running evaluations, checking progress, debugging failed evals, analyzing or analysing runs or results, run directories or artifact paths on clusters, Slurm job issues, invocation IDs, or inspecting logs (client logs, server logs, SSH to cluster, tail logs, grep logs). Do NOT use for creating or modifying evaluation configs.\nlicense: Apache-2.0\n# Vendored verbatim from NVIDIA NeMo Evaluator (commit 8fa16b2)\n# https:\u002F\u002Fgithub.com\u002FNVIDIA-NeMo\u002FEvaluator\u002Ftree\u002F8fa16b237d11e213ea665d5bad6b44d393762317\u002Fpackages\u002Fnemo-evaluator-launcher\u002F.claude\u002Fskills\u002Flaunching-evals\n# To re-sync: .claude\u002Fscripts\u002Fsync-upstream-skills.sh\n---\n\n# NeMo Evaluator Skill\n\n## Quick Reference\n\n### nemo-evaluator-launcher CLI\n\n```bash\n# Run evaluation\nuv run nemo-evaluator-launcher run --config \u003Cpath.yaml>\nuv run nemo-evaluator-launcher run --config \u003Cpath.yaml> -t \u003Ca_single_task_to_be_run_by_name>\nuv run nemo-evaluator-launcher run --config \u003Cpath.yaml> -t \u003Ctask_name_1> -t \u003Ctask_name_2> ...\nuv run nemo-evaluator-launcher run --config \u003Cpath.yaml> -o evaluation.nemo_evaluator_config.config.params.limit_samples=10 ...\n\n# Preview the resolved config and the sbatch script without running the evaluation\nuv run nemo-evaluator-launcher run --config \u003Cpath.yaml> --dry-run\n\n# Check status (--json for machine-readable output)\nuv run nemo-evaluator-launcher status \u003Cinvocation_id> --json\n\n# Get evaluation run info (output paths, slurm job IDs, cluster hostname, etc.)\nuv run nemo-evaluator-launcher info \u003Cinvocation_id>\n\n# Copy just the logs (quick — good for debugging)\nuv run nemo-evaluator-launcher info \u003Cinvocation_id> --copy-logs .\u002Fevaluation-results\u002F\n\n# For artifacts: use `nel info` to discover paths. If remote, SSH to explore and rsync what you need.\n# If local, just read directly from the paths shown by `nel info`.\n# ssh \u003Cuser>@\u003Chostname> \"ls \u003Cartifacts_path>\u002F\"\n# rsync -avzP \u003Cuser>@\u003Chostname>:\u003Cartifacts_path>\u002F{results.yml,eval_factory_metrics.json,config.yml} .\u002Fevaluation-results\u002F\u003Cinvocation_id>.\u003Cjob_index>\u002Fartifacts\u002F\n\n# Resume a failed\u002Finterrupted run (re-sbatches existing run.sub in the original run directory)\nuv run nemo-evaluator-launcher resume \u003Cinvocation_id>\n\n# List past runs\nuv run nemo-evaluator-launcher ls runs --since 1d   \n\n# List available evaluation tasks (by default, only shows tasks from the latest released containers)\nuv run nemo-evaluator-launcher ls tasks\nuv run nemo-evaluator-launcher ls tasks --from_container nvcr.io\u002Fnvidia\u002Feval-factory\u002Fsimple-evals:26.03\n```\n\n## Workflow\n\nThe complete evaluation workflow is divided into the following steps you should follow IN ORDER.\n\n1. Create or modify a config using the `nel-assistant` skill. If the user provides a past run, use its `config.yml` artifact as a starting point.\n2. Run the evaluation. See `references\u002Frun-evaluation.md` when executing this step.\n3. **Monitor progress (MANDATORY after every `nel run`)**: poll status repeatedly until SUCCESS\u002FFAILED. See `references\u002Fcheck-progress.md`.\n4. Post-run actions (when terminal state reached):\n   1. When the evaluation status is `SUCCESS`, analyze the results. See `references\u002Fanalyze-results.md` when executing this step.\n   2. When the evaluation status is `FAILED`, debug the failed run. See `references\u002Fdebug-failed-runs.md` when executing this step.\n\n# Key Facts\n\n- Benchmark-specific info learned during launching\u002Fanalyzing evals should be added to `references\u002Fbenchmarks\u002F`\n- **PPP** = Slurm account \u002F project portfolio code (the `account` field in cluster_config.yaml). When the user says \"change PPP to X\", update the account value (e.g., `\u003Cold_account>` → `\u003Cnew_account>`).\n- **Slurm job pairs**: NEL (nemo-evaluator-launcher) submits paired Slurm jobs — a RUNNING job + a PENDING restart job (for when the 4h walltime expires). Never cancel the pending restart jobs — they are expected and necessary.\n- **HF cache requirement**: For configs with `HF_HUB_OFFLINE=1`, models must be pre-downloaded to the HF cache on each cluster before launching. **Before running a model on a new cluster, always ask the user if the model is already cached there.** If not, on the cluster login node: `python3 -m venv hf_cli && source hf_cli\u002Fbin\u002Factivate && pip install huggingface_hub` then `HF_HOME=\u003Cyour_hf_cache_path> hf download \u003Cmodel>` (on lustre-style HPC clusters this is typically under `\u002Flustre\u002F...\u002F\u003Cgroup>\u002Fusers\u002F\u003Cusername>\u002Fcache\u002Fhuggingface`). Without this, vLLM will fail with `LocalEntryNotFoundError`.\n- **`data_parallel_size` is per node**: `dp_size=1` with `num_nodes=8` means 8 model instances total (one per node), load-balanced by haproxy. Do NOT interpret `dp_size` as the global replica count.\n- **`payload_modifier` interceptor**: The `params_to_remove` list (e.g. `[max_tokens, max_completion_tokens]`) strips those fields from the outgoing payload, intentionally lifting output length limits so reasoning models can think as long as they need.\n- **Auto-export git workaround**: The export container (`python:3.12-slim`) lacks `git`. When installing the launcher from a git URL, set `auto_export.launcher_install_cmd` to install git first (e.g., `apt-get update -qq && apt-get install -qq -y git && pip install \"nemo-evaluator-launcher[all] @ git+...#subdirectory=packages\u002Fnemo-evaluator-launcher\"`).\n- **Do NOT use `nemo-evaluator-launcher export --dest local`** — it only writes a summary JSON (`processed_results.json`), it does NOT copy actual logs or artifacts despite accepting `--copy_logs` and `--copy-artifacts` flags. `nel info --copy-artifacts` works but copies everything (very slow for large benchmarks). Preferred approach: use `nel info` to discover paths — if local, read directly; if remote, SSH to explore and rsync only what you need. Note that `nel info` prints standard artifacts but benchmarks produce additional artifacts in subdirs — explore to find them.\n","# NeMo Evaluator Skill\n\n## Quick Reference\n\n### nemo-evaluator-launcher CLI\n\n```bash\n# Run evaluation\nuv run nemo-evaluator-launcher run --config \u003Cpath.yaml>\nuv run nemo-evaluator-launcher run --config \u003Cpath.yaml> -t \u003Ca_single_task_to_be_run_by_name>\nuv run nemo-evaluator-launcher run --config \u003Cpath.yaml> -t \u003Ctask_name_1> -t \u003Ctask_name_2> ...\nuv run nemo-evaluator-launcher run --config \u003Cpath.yaml> -o evaluation.nemo_evaluator_config.config.params.limit_samples=10 ...\n\n# Preview the resolved config and the sbatch script without running the evaluation\nuv run nemo-evaluator-launcher run --config \u003Cpath.yaml> --dry-run\n\n# Check status (--json for machine-readable output)\nuv run nemo-evaluator-launcher status \u003Cinvocation_id> --json\n\n# Get evaluation run info (output paths, slurm job IDs, cluster hostname, etc.)\nuv run nemo-evaluator-launcher info \u003Cinvocation_id>\n\n# Copy just the logs (quick — good for debugging)\nuv run nemo-evaluator-launcher info \u003Cinvocation_id> --copy-logs .\u002Fevaluation-results\u002F\n\n# For artifacts: use `nel info` to discover paths. If remote, SSH to explore and rsync what you need.\n# If local, just read directly from the paths shown by `nel info`.\n# ssh \u003Cuser>@\u003Chostname> \"ls \u003Cartifacts_path>\u002F\"\n# rsync -avzP \u003Cuser>@\u003Chostname>:\u003Cartifacts_path>\u002F{results.yml,eval_factory_metrics.json,config.yml} .\u002Fevaluation-results\u002F\u003Cinvocation_id>.\u003Cjob_index>\u002Fartifacts\u002F\n\n# Resume a failed\u002Finterrupted run (re-sbatches existing run.sub in the original run directory)\nuv run nemo-evaluator-launcher resume \u003Cinvocation_id>\n\n# List past runs\nuv run nemo-evaluator-launcher ls runs --since 1d   \n\n# List available evaluation tasks (by default, only shows tasks from the latest released containers)\nuv run nemo-evaluator-launcher ls tasks\nuv run nemo-evaluator-launcher ls tasks --from_container nvcr.io\u002Fnvidia\u002Feval-factory\u002Fsimple-evals:26.03\n```\n\n## Workflow\n\nThe complete evaluation workflow is divided into the following steps you should follow IN ORDER.\n\n1. Create or modify a config using the `nel-assistant` skill. If the user provides a past run, use its `config.yml` artifact as a starting point.\n2. Run the evaluation. See `references\u002Frun-evaluation.md` when executing this step.\n3. **Monitor progress (MANDATORY after every `nel run`)**: poll status repeatedly until SUCCESS\u002FFAILED. See `references\u002Fcheck-progress.md`.\n4. Post-run actions (when terminal state reached):\n   1. When the evaluation status is `SUCCESS`, analyze the results. See `references\u002Fanalyze-results.md` when executing this step.\n   2. When the evaluation status is `FAILED`, debug the failed run. See `references\u002Fdebug-failed-runs.md` when executing this step.\n\n# Key Facts\n\n- Benchmark-specific info learned during launching\u002Fanalyzing evals should be added to `references\u002Fbenchmarks\u002F`\n- **PPP** = Slurm account \u002F project portfolio code (the `account` field in cluster_config.yaml). When the user says \"change PPP to X\", update the account value (e.g., `\u003Cold_account>` → `\u003Cnew_account>`).\n- **Slurm job pairs**: NEL (nemo-evaluator-launcher) submits paired Slurm jobs — a RUNNING job + a PENDING restart job (for when the 4h walltime expires). Never cancel the pending restart jobs — they are expected and necessary.\n- **HF cache requirement**: For configs with `HF_HUB_OFFLINE=1`, models must be pre-downloaded to the HF cache on each cluster before launching. **Before running a model on a new cluster, always ask the user if the model is already cached there.** If not, on the cluster login node: `python3 -m venv hf_cli && source hf_cli\u002Fbin\u002Factivate && pip install huggingface_hub` then `HF_HOME=\u003Cyour_hf_cache_path> hf download \u003Cmodel>` (on lustre-style HPC clusters this is typically under `\u002Flustre\u002F...\u002F\u003Cgroup>\u002Fusers\u002F\u003Cusername>\u002Fcache\u002Fhuggingface`). Without this, vLLM will fail with `LocalEntryNotFoundError`.\n- **`data_parallel_size` is per node**: `dp_size=1` with `num_nodes=8` means 8 model instances total (one per node), load-balanced by haproxy. Do NOT interpret `dp_size` as the global replica count.\n- **`payload_modifier` interceptor**: The `params_to_remove` list (e.g. `[max_tokens, max_completion_tokens]`) strips those fields from the outgoing payload, intentionally lifting output length limits so reasoning models can think as long as they need.\n- **Auto-export git workaround**: The export container (`python:3.12-slim`) lacks `git`. When installing the launcher from a git URL, set `auto_export.launcher_install_cmd` to install git first (e.g., `apt-get update -qq && apt-get install -qq -y git && pip install \"nemo-evaluator-launcher[all] @ git+...#subdirectory=packages\u002Fnemo-evaluator-launcher\"`).\n- **Do NOT use `nemo-evaluator-launcher export --dest local`** — it only writes a summary JSON (`processed_results.json`), it does NOT copy actual logs or artifacts despite accepting `--copy_logs` and `--copy-artifacts` flags. `nel info --copy-artifacts` works but copies everything (very slow for large benchmarks). Preferred approach: use `nel info` to discover paths — if local, read directly; if remote, SSH to explore and rsync only what you need. Note that `nel info` prints standard artifacts but benchmarks produce additional artifacts in subdirs — explore to find them.\n",{"name":5,"license":15,"description":7},"main","e169b4c2bc62a32c45aaef793cde7599dee2c1b2","skills\u002Fmodelopt-launching-evals",806,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FModel-Optimizer","https:\u002F\u002Fnvidia.github.io\u002FModel-Optimizer\u002F","https:\u002F\u002Fgithub.com\u002FNVIDIA","https:\u002F\u002Fnvidia.com","2026-11-15 02:45:59",{"source":8,"skillName":5,"commands":34,"prompt":44,"zipAvailable":45},[35,38,41],{"tool":36,"command":37},"npx","npx skills add NVIDIA\u002FModel-Optimizer --skill modelopt-launching-evals",{"tool":39,"command":40},"bunx","bunx skills add NVIDIA\u002FModel-Optimizer --skill modelopt-launching-evals",{"tool":42,"command":43},"pnpm","pnpm dlx skills add NVIDIA\u002FModel-Optimizer --skill modelopt-launching-evals","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FModel-Optimizer\u002Fblob\u002Fmain\u002Fskills\u002Fmodelopt-launching-evals\u002FSKILL.md",true,[47,53,57,61,65,69,73,77],{"path":48,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":51,"sourceHtmlUrl":52},"references\u002Fanalyze-results.md","REFERENCE","参考文档",4617,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FModel-Optimizer\u002Fblob\u002Fmain\u002Fskills\u002Fmodelopt-launching-evals\u002Freferences\u002Fanalyze-results.md",{"path":54,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":55,"sourceHtmlUrl":56},"references\u002Fbenchmarks\u002Fswebench-general-info.md",11726,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FModel-Optimizer\u002Fblob\u002Fmain\u002Fskills\u002Fmodelopt-launching-evals\u002Freferences\u002Fbenchmarks\u002Fswebench-general-info.md",{"path":58,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":59,"sourceHtmlUrl":60},"references\u002Fbenchmarks\u002Fterminal-bench-general-info.md",7908,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FModel-Optimizer\u002Fblob\u002Fmain\u002Fskills\u002Fmodelopt-launching-evals\u002Freferences\u002Fbenchmarks\u002Fterminal-bench-general-info.md",{"path":62,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":63,"sourceHtmlUrl":64},"references\u002Fbenchmarks\u002Fterminal-bench-trace-analysis.md",5866,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FModel-Optimizer\u002Fblob\u002Fmain\u002Fskills\u002Fmodelopt-launching-evals\u002Freferences\u002Fbenchmarks\u002Fterminal-bench-trace-analysis.md",{"path":66,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":67,"sourceHtmlUrl":68},"references\u002Fcheck-progress.md",1234,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FModel-Optimizer\u002Fblob\u002Fmain\u002Fskills\u002Fmodelopt-launching-evals\u002Freferences\u002Fcheck-progress.md",{"path":70,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":71,"sourceHtmlUrl":72},"references\u002Fdebug-failed-runs.md",6883,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FModel-Optimizer\u002Fblob\u002Fmain\u002Fskills\u002Fmodelopt-launching-evals\u002Freferences\u002Fdebug-failed-runs.md",{"path":74,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":75,"sourceHtmlUrl":76},"references\u002Frun-evaluation.md",1122,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FModel-Optimizer\u002Fblob\u002Fmain\u002Fskills\u002Fmodelopt-launching-evals\u002Freferences\u002Frun-evaluation.md",{"path":78,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":79,"sourceHtmlUrl":80},"tests.json",2779,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FModel-Optimizer\u002Fblob\u002Fmain\u002Fskills\u002Fmodelopt-launching-evals\u002Ftests.json",[82,89,96,103,110,117],{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":19},"01M7B5XD9M965H2TAV4ZJMPF6N","modelopt-speculative-decoding","Train, debug, and validate a speculative decoding draft model (EAGLE3, DFlash, DSpark, Domino) through the ModelOpt launcher pipeline. Use when the user wants to add a new model to a draft-training pipeline, asks why a pipeline run failed, wants experiment logs reviewed, or wants to check whether a run's acceptance rate meets threshold. Triggers on \"EAGLE3\", \"DFlash\", \"DSpark\", \"Domino\", \"draft model\", \"acceptance rate\", \"speculative decoding pipeline\". Do NOT use for quantizing a model (use modelopt-ptq) or serving a checkpoint (use modelopt-deployment).","https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FModel-Optimizer\u002Fblob\u002Fmain\u002Fskills\u002Fmodelopt-speculative-decoding\u002FSKILL.md",[],[],{"skillNo":90,"name":91,"nameZh":6,"description":92,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":93,"categories":94,"keywords":95,"publishedAt":19},"01M7B5X9SG5FHW174VFK979DQ9","modelopt-run-agentx","Run the AgentX agentic serving benchmark with the SemiAnalysis harness. Use for AgentX setup, dataset downloads, concurrency sweeps, or serving latency and throughput comparisons.","https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FModel-Optimizer\u002Fblob\u002Fmain\u002Fskills\u002Fmodelopt-run-agentx\u002FSKILL.md",[],[],{"skillNo":97,"name":98,"nameZh":6,"description":99,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":100,"categories":101,"keywords":102,"publishedAt":19},"01M7B5X4C0XGVA8N0BXF9N4HCX","modelopt-release-cherry-pick","Audit merged bug-fix PRs and release NVBugs for missing cherry-pick labels, then cherry-pick labeled PRs into a release branch and open a PR. Use when asked to \"cherry-pick PRs for release\u002FX.Y.Z\", \"pick PRs to release branch\", \"verify cherry-pick labels\", or \"cherry-pick labeled PRs\".","https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FModel-Optimizer\u002Fblob\u002Fmain\u002Fskills\u002Fmodelopt-release-cherry-pick\u002FSKILL.md",[],[],{"skillNo":104,"name":105,"nameZh":6,"description":106,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":107,"categories":108,"keywords":109,"publishedAt":19},"01M7B5WXMDYXCK43YA5ZEF7ECY","modelopt-quant-recipe-search","Use when the user asks to find, search for, or optimize the best quantization recipe for a model, including direct requests like \"find the best quantization recipe and generate a PTQ checkpoint.\" Guides the multi-candidate loop: choose compute-vs-memory success metrics, select ModelOpt recipe baselines, design AutoQuant\u002Fmanual recipe deltas, interpret sensitivity, and decide next candidates. Do NOT use for a single known PTQ recipe run (use modelopt-ptq), serving (use modelopt-deployment), creating\u002Frunning evals (use modelopt-evaluation or modelopt-launching-evals), monitoring jobs (use modelopt-monitor), MLflow browsing (use modelopt-accessing-mlflow), or comparing completed baseline-vs-candidate scores only (use modelopt-compare-results).","https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FModel-Optimizer\u002Fblob\u002Fmain\u002Fskills\u002Fmodelopt-quant-recipe-search\u002FSKILL.md",[],[],{"skillNo":111,"name":112,"nameZh":6,"description":113,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":114,"categories":115,"keywords":116,"publishedAt":19},"01M7B5WRMPZE23YXCFMZBM8KJA","modelopt-qad","Run explicitly requested ModelOpt Quantization-Aware Distillation (QAD) through Megatron Bridge (recommended) or Hugging Face Transformers to recover a measured BF16-to-PTQ accuracy gap. Use only when the user explicitly asks for QAD, including its topology, data preparation, launch, resume, checkpoint export, or recovery decisions.","https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FModel-Optimizer\u002Fblob\u002Fmain\u002Fskills\u002Fmodelopt-qad\u002FSKILL.md",[],[],{"skillNo":118,"name":119,"nameZh":6,"description":120,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":121,"categories":122,"keywords":123,"publishedAt":19},"01M7B5WJHB4D787EZCY4RNN44T","modelopt-ptq","Use when the user asks to \"quantize a model\", \"run PTQ\", \"post-training quantization\", \"NVFP4 quantization\", \"FP8 quantization\", \"INT8 quantization\", \"INT4 AWQ\", \"quantize LLM\", \"quantize MoE\", \"quantize VLM\", or needs to produce a quantized HuggingFace checkpoint from a pretrained model using ModelOpt. Do NOT use for multi-candidate recipe exploration or optimization (use modelopt-quant-recipe-search).","https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FModel-Optimizer\u002Fblob\u002Fmain\u002Fskills\u002Fmodelopt-ptq\u002FSKILL.md",[],[]]