Nemo mbridge perf activation recompute vs Nemo mbridge perf cpu offloading
Nemo mbridge perf activation recompute is an open-source agent skill for ai models and llm tooling. Nemo mbridge perf cpu offloading is an open-source agent skill for ai models and llm tooling.
- Nemo mbridge perf activation recompute is best for fix a gpu out-of-memory error during megatron bridge pretraining.
- Nemo mbridge perf cpu offloading is best for reduce gpu memory use when training a large model with optimizer offload.
- What it is
- Nemo mbridge perf activation recompute is an open-source agent skill for ai models and llm tooling.Nemo mbridge perf cpu offloading is an open-source agent skill for ai models and llm tooling.
- Best for
- Fix a GPU out-of-memory error during Megatron Bridge pretrainingReduce GPU memory use when training a large model with optimizer offload
- Kind
- SkillSkill
- Pricing
- Open sourceOpen source
- Platforms
- Claude Code, Codex, Cursor, Gemini CLI, GitHub Copilot, Windsurf, GooseClaude Code, Codex, Cursor, Gemini CLI, GitHub Copilot, Windsurf, Goose
- GitHub stars
- 4K4K
- Last commit
- Nov 2026Nov 2026
- License
- Apache-2.0Apache-2.0
- Open source
- Good for
- Fix a GPU out-of-memory error during Megatron Bridge pretraining; Pick which selective recompute modules to enable for an MoE model; Compare memory and step time with and without activation recomputeReduce GPU memory use when training a large model with optimizer offload; Choose between activation and optimizer offloading for a given model size; Debug an out-of-memory error after a config change to CPU offloading
What Nemo mbridge perf activation recompute does
Selective module choice
Explains which recompute labels target attention, MLA, MoE, dense MLP and norm activations.
Full-layer recompute settings
Covers uniform and block placement with the number of layers to checkpoint.
Measurement protocol
Defines the matched control, per-rank memory fields and timing windows to record.
Compatibility checks
Lists restrictions involving MoE overlap, FP8 scaling, context parallelism and CUDA graphs.
Failure diagnosis table
Maps symptoms like persistent OOMs or allocator caching to likely causes and next steps.
Reference benchmark tables
Provides matched H100 runs for specific Megatron Bridge recipes as context, not rankings.
What Nemo mbridge perf cpu offloading does
Activation offloading setup
Explains how to offload per-layer activations to CPU and the settings it requires, such as PP=1 and no recomputation.
Fractional optimizer offload
Covers moving a chosen fraction of Adam optimizer states to CPU through HybridDeviceOptimizer.
Config compatibility rules
Lists which offloading, recompute, CUDA graph and parallelism options conflict with one another.
Failure diagnosis table
Maps common offloading errors and OOM symptoms to causes, checks and fixes.
Verification steps
Provides unit test commands and criteria for confirming offloaded training matches the baseline.
Nemo mbridge perf activation recompute vs Nemo mbridge perf cpu offloading: questions
Is Nemo mbridge perf activation recompute or Nemo mbridge perf cpu offloading cheaper?
Both are open source. Compare current plans on each official site.
What is the difference between Nemo mbridge perf activation recompute and Nemo mbridge perf cpu offloading?
Nemo mbridge perf activation recompute is an open-source agent skill for ai models and llm tooling. Nemo mbridge perf cpu offloading is an open-source agent skill for ai models and llm tooling.
