jetson-speculative-decoding

von nvidia

Füge EAGLE-3 oder Draft-Model-Spekulativ-Decodierung zu einem Jetson vLLM-Server hinzu, wenn TPOT der Engpass ist.

npx skills add https://github.com/nvidia/skills --skill jetson-speculative-decoding

Jetson Speculative Decoding (vLLM)

Speculative decoding lets a small "draft" model propose tokens that the target model verifies in a single forward pass, reducing per-token latency. On Jetson, the win/loss is dominated by VRAM headroom, not by the draft quality. This skill encodes the parts an LLM won't already know.

Purpose

Tune an existing Jetson vLLM deployment for faster token generation by appending the right --speculative-config and validating whether it improves single-stream decode speed.

When to use

  • TPOT/ITL is the bottleneck (TTFT is fine, output is just slow).
  • Workload is single-stream or low-concurrency (≤2). Speculation usually loses at high concurrency.
  • Jetson family is Thor or AGX Orin. Do not suggest EAGLE-3 on Orin Nano/NX — there is rarely enough VRAM headroom to host both target and draft, and you'll OOM at startup.

When NOT to use

  • High-concurrency serving (≥8): batched decode usually beats speculation; the draft model just steals VRAM.
  • Models without a published EAGLE-3 head — do not train one ad-hoc as a "fix".
  • After applying jetson-inference-mem-tune flags that already pushed --gpu-memory-utilization near the ceiling. Free at least ~2 GB first.

Prerequisites

  • A working vLLM server recipe from jetson-llm-serve.
  • Enough memory headroom for the draft model or EAGLE-3 head in addition to the target model.
  • A benchmark baseline from jetson-llm-benchmark before enabling speculation.
  • A target model with a compatible EAGLE-3 head, or a small same-family draft model for the fallback path.

Instructions

Append --speculative-config to the vllm serve command shown in jetson-llm-serve.

EAGLE-3 (preferred when a head is published for the target model):

--speculative-config '{
  "method": "eagle3",
  "model": "<eagle3-head-repo-id>",
  "num_speculative_tokens": 5,
  "draft_tensor_parallel_size": 1
}'

Draft-model (fallback — pair a small same-family model):

--speculative-config '{
  "method": "draft_model",
  "model": "<small-draft-model-repo-id>",
  "num_speculative_tokens": 4,
  "draft_tensor_parallel_size": 1
}'

Jetson-specific tuning rules

  • num_speculative_tokens: start at 5 on Thor, 3 on AGX Orin. Higher values pay off only if the draft acceptance rate is >0.6.
  • Always pair with the same vLLM runtime path used by jetson-llm-serve: upstream vLLM 0.20+ (vllm/vllm-openai:latest) or validated native vLLM 0.20+ on Thor, upstream vLLM 0.20+ on Orin JetPack 7.2 / L4T r39+, or the NVIDIA-AI-IOT vLLM image on older Orin. Do not use an Orin NVIDIA-AI-IOT vLLM image on Thor. Older runtimes may lack EAGLE-3 or the current --speculative-config shape.
  • Drop --gpu-memory-utilization by ~0.05 vs the non-speculative baseline to give the draft model headroom.

How to verify it actually helped

  1. Run jetson-llm-benchmark (vLLM path) at --concurrency 1 before and after enabling speculation.
  2. Acceptance: target ≥30% improvement in throughput_tok_s and ≥20% drop in tpot_ms_p50 at concurrency 1.
  3. If improvement is <10%, or throughput_tok_s regresses at concurrency 8, disable speculation. The draft model is costing more than it returns.

Limitations

  • Speculative decoding improves decode-heavy workloads; it does not reduce TTFT-dominated latency.
  • High concurrency can erase the benefit because continuous batching already keeps the GPU busy.
  • Orin Nano/NX usually lack enough memory headroom for both target and draft models.
  • Acceptance rate and draft overhead are model-specific, so benchmark before and after instead of assuming a speedup.

Error handling

  • If vLLM rejects --speculative-config, verify that Thor and Orin JetPack 7.2 / L4T r39+ are using vLLM 0.20+ and that older Orin is using a JetPack-matched NVIDIA-AI-IOT vLLM image; then switch back to the non-speculative serving command if the runtime still rejects it.
  • If startup OOMs, lower --gpu-memory-utilization, use a smaller draft, or disable speculation and hand off to jetson-inference-mem-tune.
  • If benchmark throughput regresses, remove --speculative-config; a bad draft path is worse than no speculation.

Hand off to

  • jetson-llm-benchmark to quantify the change.
  • jetson-inference-mem-tune if startup OOMs after enabling speculation.

Source

vLLM speculative decoding docs and the Jetson AI Lab GenAI tutorial.

Mehr Skills von nvidia

compileiq-debug
nvidia
Verwenden, wenn etwas nicht stimmt: Search() hängt, alle Evaluierungen geben INVALID_SCORE zurück, Scores verbessern sich nicht, jede Konfiguration liefert dieselbe Zahl, ptxas-Fehler…
create-github-pr
nvidia
Erstelle GitHub-Pull-Requests mit der gh CLI. Verwende, wenn der Benutzer einen neuen PR erstellen, Code zur Überprüfung einreichen oder einen Pull-Request öffnen möchte. Auslöser-Schlüsselwörter -…
nemoclaw-maintainer-cross-issue-sweep
nvidia
Scannt andere offene Issues, um solche zu finden, die ein bestimmter PR möglicherweise ebenfalls behebt oder versehentlich kaputt macht. Gibt benachbarte Fix-Möglichkeiten und Widerspruchsrisiken mit Datei:Zeile… aus.
fhir-basics
nvidia
Bringt Agenten bei, wie FHIR R4 APIs funktionieren, welche Ressourcen verfügbar sind, wie man sie mit Suchparametern abfragt und wie man alle Antwortformate korrekt parst…
compileiq-validate-result
nvidia
Verwende NACH Abschluss einer Suche und VOR dem Einfordern eines Speedups oder dem Versand eines ACF. Lädt die dump_results CSV, extrahiert Top-K-Kandidaten (Einzelziel)...
changelog-audit
nvidia
Auditiere die CHANGELOG.md vor einem Release: stelle verlorene Einträge wieder her, sortiere nach Benutzerauswirkung, verfeinere die Sprache der Einträge, führe Zeilenumbrüche durch und (im Release-Branch-Modus) erhöhe die Vergleichsnummer…
maintain-dynamic-plugins
nvidia
Verwalte NeMo Relay dynamische Plugin-Lader, Manifeste, Rust native SDKs, gRPC Worker-Protokoll, Python Worker-SDK, Dokumentation, Tests und Abdeckung des Release-Workflows
dgx-diagnose
nvidia
Diagnostizieren Sie häufige DGX Station GB300-Probleme – CUDA-Abstürze, falsche GPU-Zuweisung, vLLM/SGLang-Container-Fehler, MIG-Status-Probleme, NVLink/Fabric-Manager-Fehler,…