Home / nvidia / skills · skills/nemo-mbridge-perf-activation-recompute/SKILL.md · GitHub

nemo-mbridge-perf-activation-recompute skillA

nemo-mbridge-perf-activation-recompute is agent-read markdown (skill) from nvidia/skills: Validate and use selective and full activation recompute in Megatron Bridge to reduce GPU memory usage at the cost of extra compute. Use for activation memory OOMs or regressions involving recompute_granularity, recompute_num_layers, recompute_modules, recompute_method, selective recompute, full recompute, or activation checkpointing..

Indexed from public GitHub and served as immutable, content-addressed versions. Install it pinned to an exact SHA-256 with the mdr CLI, and every file is verified against the hash recorded here before it reaches your agent. The deterministic audit below grades the latest version, and the same file always earns the same grade.

What the file says

# Activation Recompute

Stable docs: @docs/training/activation-recomputation.md
Card: @skills/nemo-mbridge-perf-activation-recompute/card.yaml

<!-- Guidance refreshed: 2026-08-12. -->

Activation recompute (activation checkpointing) trades additional forward work during backward for lower retained-activation memory. The useful checkpoint boundary depends on the model architecture, attention backend, parallelism, and the tensor that actually drives the per-rank peak.

## Quick Decision Guide

1. Confirm the pressure is real allocation, not allocator fragmentation. Compare `max_memory_allocated()` with `max_memory_reserved()` on every rank.
2. Keep an explicit no-recompute control when the workload fits. Under selective granularity, `recompute_modules=[]` is valid and useful for this comparison.
3. Select the first boundary from the architecture and observed peak:
   - **Standard attention:** `core_attn` is the common first candidate. It is strongest when unfused attention materializes score/probability tensors. With Transformer Engine fused or Flash Attention, compare it against `[]` because those backends already rematerialize attention internals.
…

Read the whole file at its exact version.

How to install

Latest version
mdr add nvidia/skills/nemo-mbridge-perf-activation-recompute@git:20260817.d36ff70
Exact content
mdr add nvidia/skills/nemo-mbridge-perf-activation-recompute@sha256:03379723143dfcd3

Pin to a label to follow the author's releases, or to a sha256 to freeze the exact bytes forever. Either way the resolved hash is written to mdr.lock, and mdr install reproduces it on any machine.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_a25z4swu5g26bxlh.svg)](https://markdownregistry.com/a/art_a25z4swu5g26bxlh)

1 badge views in 30 days

Versions

versioncommittedcommitsizeaudit
git:20260817.d36ff70 latest2026-08-17 d36ff70 18,915 BA view · diff
git:20260616.84918882026-06-16 8491888 9,197 BA view · diff
git:20260602.e695a832026-06-02 e695a83 10,151 BA view · diff
git:20260529.7d99fda2026-05-29 7d99fda 9,295 BA view

Audit of the latest version

A  17 of 17 checks passed. Deterministic, no model, same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (18915 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

nvidia/skills · 3,415 stars · license Apache-2.0 · pushed 2026-09-23 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_a25z4swu5g26bxlh
GET https://markdownregistry.com/api/v1/resolve?ref=nvidia/skills/nemo-mbridge-perf-activation-recompute
GET https://markdownregistry.com/api/v1/blob/03379723143dfcd3cce4338c148fecc2aa515649ab85d3624e43a29b80f03e4e

Your agent does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

More from nvidia/skills

nvidia-skill-finder skill
nvidia/skills · plugins/nvidia-skills/skills/nvidia-skill-finder/SKILL.md · Use for NVIDIA-related requests where an NVIDIA skill might help, even if the user did not ask for a skill. Trigger on…
git:20260717.ab94cb3 · audit A · 3,415 stars
accelerated-computing-cudf skill
nvidia/skills · skills/accelerated-computing-cudf/SKILL.md · Official NVIDIA-authored guidance for NVIDIA cuDF GPU DataFrames, pandas acceleration, dask-cuDF, ETL, joins, groupby…
git:20260915.2cd3507 · audit A · 3,415 stars
aiq-deploy skill
nvidia/skills · skills/aiq-deploy/SKILL.md · Use when asked to install, deploy, run, validate, troubleshoot, or stop NVIDIA AI-Q Blueprint infrastructure.
v2.1.0 · audit A · 3,415 stars
aiq-research skill
nvidia/skills · skills/aiq-research/SKILL.md · Use when asked to run deep research or AI-Q research through a reachable NVIDIA AI-Q Blueprint backend.
v2.1.0 · audit A · 3,415 stars
amc-run-rtsp-calibration skill
nvidia/skills · skills/amc-run-rtsp-calibration/SKILL.md · Calibrate a new dataset from live RTSP camera streams via the AutoMagicCalib REST API. Use when the user provides RTSP…
v1.0.0 · audit B · 3,415 stars
amc-run-sample-calibration skill
nvidia/skills · skills/amc-run-sample-calibration/SKILL.md · Run end-to-end calibration on the shipped sample dataset (sdg_08_2_sample_data_010926.zip) against a running AMC…
v1.0.0 · audit A · 3,415 stars
amc-run-video-calibration skill
nvidia/skills · skills/amc-run-video-calibration/SKILL.md · Calibrates pre-recorded `cam_*.mp4` datasets through the AutoMagicCalib REST API. Use for user-supplied local MP4s…
v1.0.0 · audit A · 3,415 stars
amc-setup-calibration-stack skill
nvidia/skills · skills/amc-setup-calibration-stack/SKILL.md · Launch AutoMagicCalib microservice and web UI from NGC release images via Docker Compose. Use when user says 'deploy…
v1.0.0 · audit A · 3,415 stars
kermt-add-cmim-pretrain skill
nvidia/skills · skills/bionemo-kermt-add-cmim-pretrain/SKILL.md · Convert a grover_base checkpoint (encoder-only or encoder + vocab heads) into a hybrid checkpoint by adding a…
git:20260915.2cd3507 · audit A · 3,415 stars
kermt-continue-pretrain skill
nvidia/skills · skills/bionemo-kermt-continue-pretrain/SKILL.md · Continue KERMT pretraining on a custom SMILES corpus with a grover_base, cmim, or hybrid checkpoint. Use a local…
git:20260915.2cd3507 · audit A · 3,415 stars
kermt-embed skill
nvidia/skills · skills/bionemo-kermt-embed/SKILL.md · Extract per-molecule embeddings from any encoder-bearing KERMT checkpoint. Use a local checkpoint or optionally…
git:20260915.2cd3507 · audit A · 3,415 stars
kermt-finetune skill
nvidia/skills · skills/bionemo-kermt-finetune/SKILL.md · Finetune a pretrained KERMT encoder on a labeled CSV. Validate the checkpoint and data, prepare features, and run…
git:20260915.2cd3507 · audit A · 3,415 stars

Every file in nvidia/skills

Browse by kind, by grade A, or by owner.