Dev & Engineering ✓ NVIDIA · Official moe-vlm-trainingmegatron-bridgefsdp3d-parallelismqwen3-vlcuda-graphsactivation-recompute

NeMo MBridge MoE VLM Training Guide

Guides FSDP and 3D-parallel choices and tuning for MoE vision-language model training.

FollowSkills review · FSRS-2.0
Not recommended
57/ 100 5-point scale 2.9 / 5
Trust18 / 25 · 3.6/5

This is read-only-style training guidance with no request for credentials, uploads, or destructive actions; scope, target hardware, some risks, Apache-2.0 licensing, NVIDIA ownership, and a security disclosure channel are visible. Deductions apply because user confirmation, permission boundaries, rollback, sensitive-data handling, and dependency-security guidance are absent; repository license metadata is still NOASSERTION.

Reliability8 / 20 · 2.0/5

The FSDP/3D-parallel decision logic, configuration families, and pitfalls are internally consistent, and the supplied benchmark reports passed static validation with limited task evaluation. Deductions apply because there are no scripts, installation requirements, compatibility matrix, reproducible commands, or abnormal-input/failure feedback; the evaluation has one positive task and no independently reviewable execution artifacts. Static review caps this dimension at 10.

Adaptability10 / 15 · 3.3/5

The audience, trigger terms, FSDP versus 3D-parallel scenarios, Qwen3-VL/Qwen3.5 scope, and GB200/B200 focus are reasonably clear, with limitations for FP8, long context, and architecture coverage. Deductions apply because input/output contracts, non-fit boundaries, and semantic trigger rules remain underspecified; Chinese-language support is not addressed, and the guidance depends heavily on specific hardware and Megatron Bridge environments.

Convention9 / 15 · 3.0/5

The material is readable and layered, with a decision guide, tuning knobs, configuration families, compatibility table, pitfalls, and known limitations; the skill card supplies NVIDIA ownership, Apache 2.0, validation date, and version clues. Deductions apply because Instructions and Examples are missing, as are author/tags metadata, installation/dependency notes, FAQ, explicit changelog, and a clear maintenance/update path; the skill-card version and evaluation revision are not fully aligned.

Effectiveness7 / 15 · 2.3/5

The skill can directly guide comparison of FSDP-first and 3D-parallel paths and provides useful TP/CP/PP/EP/ETP, HybridEP, MBS, recompute, CUDA-graph, and real-image-data advice; the supplied evaluation reports favorable results on its limited task. Deductions apply because it provides no runnable training commands or complete configuration, performance claims are rounded empirical summaries, scope is narrow, and substantial environment-specific adaptation and validation remain. Static review caps this dimension at 7.

Verifiability5 / 10 · 2.5/5

The skill card lists measured configurations, hardware scales, and named internal performance trackers, while BENCHMARK.md gives limited evaluation methodology, task details, and results, providing some audit trail. Deductions apply because raw experiment records, externally checkable citations, committed key-path tests, and corroborating evidence are absent; one evaluation task cannot independently reproduce the main performance claims. Static review caps this dimension at 5.

Evidence confidence:Low Reviewed Jul 20, 2026 Reviewed revision 55f18499943e
The upstream repository has new commits since this review. The score still applies to the reviewed revision shown and may not cover the latest changes.
Before you use it
  • Do not treat mock-data performance as real VLM throughput; use real or realistic image payloads and normalize comparisons by useful tokens and workload shape.
  • The recommendations target specific Qwen architectures, Megatron Bridge conditions, and GB200/B200 topologies; validate on other models, encoders, GPUs, or layouts with a small controlled run first.
  • The skill lacks executable configurations, dependency installation guidance, failure diagnosis, and rollback steps; users must supply environment checks and recovery procedures before execution.
See the full review method →

What does this skill do, and when should you use it?

This skill is for engineers training mixture-of-experts vision-language models with Megatron Bridge. It compares FSDP with 3D parallelism and recommends establishing a reliable run with FSDP before pursuing higher throughput through deeper 3D-parallel tuning. The guidance covers real multimodal input, recompute, micro-batch size, CUDA Graph scope, HybridEP, and ETP. It also applies when investigating a change that caused MoE VLM training failure or out-of-memory errors.

Compares FSDP and 3D parallelism by bring-up simplicity, memory behavior, pipeline boundaries, and throughput ceiling; recommends a path based on model stability and tuning capacity; explains why mock image-free data is an unreliable VLM performance proxy; guides vision-stack freezing, aggressive MBS sweeps, selective recompute, CUDA Graph scope selection, and conditional ETP use; provides representative GB200 FSDP-first and 3D-parallel configuration families; and advises profiling the vision encoder, projector, and decoder separately.

  1. An engineer bringing up a new MoE VLM who needs the simplest path to a reliable first run.
  2. A team tuning Qwen3-VL-class training on GB200 or B200 systems and investigating memory or throughput sensitivity.
  3. A performance engineer replacing misleading mock-data measurements with real or realistic image payloads.
  4. A developer tracing a commit that caused MoE VLM training failure or OOM.
  5. A team with a stable FSDP run deciding whether 3D parallelism is worth the additional tuning work.

What are this skill's strengths and limitations?

Pros
  • Provides a clear FSDP-first progression toward 3D-parallel optimization.
  • Addresses the practical interaction between real multimodal input, memory fit, and steady-state throughput.
  • Names concrete tuning levers including HybridEP, MBS, recompute, CUDA Graphs, and ETP.
  • Includes experience-based guidance for GB200 and B200 environments and common pitfalls.
Limitations
  • It is guidance rather than an executable training script or end-to-end experiment workflow.
  • Configuration families are high-level and omit exact model, batch-size, and topology values.
  • The supplied material does not provide an independent benchmark dataset, test suite, or broad platform matrix.
  • Its findings are based on named Qwen3-VL, Qwen3-Next, and related experiments and are not guarantees for every model or system.

How do you install this skill?

Install the NVIDIA skills collection with the README-supported command: npx skills add nvidia/skills. When the CLI prompts for a skill, select nemo-mbridge-perf-moe-vlm-training. The supplied source does not document a separate manual-copy procedure for this skill.

How do you use this skill?

After installation, invoke it with a relevant request such as: “Compare FSDP and 3D parallelism for training an MoE VLM and investigate a Qwen3-VL OOM.” Include hardware, real-versus-mock image input, MBS, recompute, and vision-freezing details when available.

How does this skill compare with similar options?

The skill directly compares FSDP and 3D parallelism. FSDP is positioned for first bring-up, memory-first tuning, and awkward pipeline boundaries; 3D parallelism is positioned for stable models with a clear pipeline layout and time for deeper sweeps when maximum steady-state throughput matters.

FAQ

Which approach should I start with?
Start with FSDP to obtain a reliable run and stabilize real-data input, recompute, and memory behavior. Move to 3D parallelism when the potential throughput gain justifies the extra tuning.
Can mock image data be used for throughput decisions?
Not reliably. The guidance reports that image-free mock runs can appear roughly twice as fast as runs with real multimodal input, so real or realistic image payloads should be used.
Should ETP be enabled by default?
No. Use ETP when expert parallelism alone is insufficient for fit or topology; it adds communication and tuning complexity.
Does this skill automatically run training or fix OOM errors?
The supplied SKILL.md provides configuration and troubleshooting guidance. It does not claim to include scripts that automatically run training or repair failures.

More skills from this repository

All from NVIDIA/skills

Dev & Engineering ✓ NVIDIA · Official

Megatron Bridge Activation Recompute

Trade extra compute for lower GPU activation memory in Megatron Bridge training.

Dev & Engineering ✓ NVIDIA · Official

NeMo MBridge Long-Context MoE Training

Practical Megatron Bridge guidance for sizing parallelism, managing memory, and preserving throughput in long-context MoE training.

Dev & Engineering ✓ NVIDIA · Official

NeMo MBridge GPU Memory Tuning

Diagnose GPU OOMs in Megatron Bridge training and reduce fragmentation, activation, and PEFT memory usage.

Dev & Engineering ✓ NVIDIA · Official

NeMo MBridge MoE Communication Overlap

Tune expert-parallel communication and computation overlap in Megatron Bridge.

Dev & Engineering ✓ NVIDIA · Official

NeMo MBridge CUDA Graphs

Configure, validate, and benchmark CUDA Graph capture in Megatron Bridge to reduce host-driver overhead.

Dev & Engineering ✓ NVIDIA · Official

NeMo MBridge MoE Performance Workflow

Systematically diagnose and improve MoE training throughput in Megatron Bridge.

Dev & Engineering ✓ NVIDIA · Official

NeMo MBridge MoE Expert-Parallel Overlap

Configure, validate, and troubleshoot MoE expert-parallel communication overlap in Megatron-Bridge.

Dev & Engineering ✓ NVIDIA · Official

NeMo MBridge Sequence Packing

Configure and validate sequence packing and long-context training in Megatron-Bridge.

Dev & Engineering ✓ NVIDIA · Official

Megatron FSDP Configuration and Performance Guide

Enable, troubleshoot, and verify Megatron FSDP in Megatron-Bridge.

Dev & Engineering ✓ NVIDIA · Official

NeMo MBridge Hierarchical Context Parallelism

Enable, troubleshoot, and verify hierarchical context parallelism in Megatron-Bridge.

Dev & Engineering ✓ NVIDIA · Official

NeMo MBridge Parallelism Strategy Guide

Choose, combine, and validate Megatron Bridge parallelism configurations.

Dev & Engineering ✓ NVIDIA · Official

NeMo MBridge CPU Offload

Configure, validate, and troubleshoot CPU offloading in Megatron Bridge to relieve GPU memory pressure.

Dev & Engineering ✓ NVIDIA · Official

Megatron-Bridge Communication Overlap Guide

Configure and verify TP, DP, and PP communication overlap in Megatron-Bridge training.

Dev & Engineering ✓ NVIDIA · Official

NeMo MBridge Recipe Recommender

Recommends adjustable Megatron Bridge recipes for your model, GPU budget, and training goal.

Dev & Engineering ✓ NVIDIA · Official

NeMo MBridge Training Resiliency

Configure fault recovery and hang detection for Megatron Bridge distributed training.

Dev & Engineering ✓ NVIDIA · Official

NeMo MBridge Training Guide

Run Megatron-LM and Megatron Bridge training and verify whether their loss curves match.

Data & Analysis ✓ NVIDIA · Official

Cosmos Reason Video QA Fine-Tuning

A Cosmos-RL playbook for FSDP-based SFT, LoRA, evaluation, and AutoML of Cosmos3-Nano video QA models.

Dev & Engineering ✓ NVIDIA · Official

MoE Hardware Configuration Reference

Plan MoE training layouts and throughput expectations across NVIDIA GPU platforms.

Dev & Engineering ✓ NVIDIA · Official

NeMo AutoModel Distributed Training

Choose and configure multi-GPU and multi-node training strategies for NeMo AutoModel.

Automation & Ops ✓ NVIDIA · Official

Megatron-LM on SLURM

Launch, monitor, and troubleshoot multi-node Megatron-LM training on SLURM GPU clusters.

Related skills