Checkpointing And Tracking

Organization

VectorSpaceLab

checkpointing-and-tracking

Save and resume Accelerate training state, register checkpoint hooks and custom state, log safely across processes, use experiment trackers, profile runs, and clean up memory.

Overview

Publisher	VectorSpaceLab
Repository	`AREX-Skill`
Skill name	`checkpointing-and-tracking`
Stars	341
Forks	26
Bundled files	5
License	Apache-2.0
Links	Skill source Repository

Markdown instructions
A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.
Works with any LLM
AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.
5 bundled files
Scripts, templates, and references the model can read while it works. Files are read-only and never executed.
Open source
Published by VectorSpaceLab on GitHub. Read the source before you install it.

Installation

Install the Checkpointing And Tracking AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

Open the app and go to Plugins → Skills.
Choose "Install from GitHub".
Paste the skill folder URL below and confirm.
https://github.com/VectorSpaceLab/AREX-Skill/tree/main/skills/repositories/repo-skills/accelerate/sub-skills/checkpointing-and-tracking
Enable the skill in any chat where you want it available.

Store

Installed

Skills

Add skills

Import one or more skills from a public GitHub folder or a local .skill or .zip archive, or SKILL.md file.

From GitHub URLFrom .skill, .zip, or SKILL.md

GitHub repo or folder

https://github.com/VectorSpaceLab/AREX-Skill/tree/main/skills/repositories/repo-skills/accelerate/sub-skills/checkpointing-and-tracking

Points to a single skill folder, or any folder that contains multiple skills. Public repos only.

CancelContinue

Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills

git clone --depth 1 https://github.com/VectorSpaceLab/AREX-Skill.git /tmp/AREX-Skill
mkdir -p .claude/skills
cp -r /tmp/AREX-Skill/skills/repositories/repo-skills/accelerate/sub-skills/checkpointing-and-tracking .claude/skills/checkpointing-and-tracking

Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Checkpointing And Tracking in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Checkpointing And Tracking on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Checkpointing And Tracking is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

Loaded only when it is needed
The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.
Switch models mid-chat
Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Checkpointing and Tracking

Use this sub-skill when the task involves Accelerate checkpoint save/load, model export, tracker initialization/logging, distributed-safe logging, profiling, state/RNG handling, or memory cleanup around those workflows.

Route First

For the core forward/backward/optimizer loop, gradient accumulation, prepared dataloaders, and skip_first_batches, use ../training-loop-integration/.
For FSDP, DeepSpeed, Megatron-LM, or backend-specific checkpoint strategy and consolidation details, use ../distributed-training-backends/.
For large-model dispatch/offload checkpoint loading, route to the large-model or big-modeling sub-skill if present.

Fast Workflows

Save and resume same-script training state with Accelerator.save_state() and Accelerator.load_state(); see references/checkpointing.md.
Configure automatic checkpoint folders and retention with ProjectConfiguration; see references/checkpointing.md.
Register custom checkpointable objects and save/load pre-hooks for extra metadata; see references/checkpointing.md.
Initialize trackers with Accelerator(log_with=...), init_trackers(), log(), get_tracker(), and end_training(); see references/tracking-and-logging.md.
Use accelerate.logging.get_logger() for multiprocess-safe logs and ordered rank logs; see references/tracking-and-logging.md.
Profile CPU/GPU/XPU sections with ProfileKwargs and accelerator.profile(); see references/tracking-and-logging.md.
Diagnose common failures using references/troubleshooting.md before changing training code.

Bundled Helper

Run python scripts/checkpoint_tracker_smoke.py --help for usage.
Run python scripts/checkpoint_tracker_smoke.py to verify a CPU-only tiny save/load/custom-tracker/logging smoke test without any external tracker service.

Key Constraints

save_state()/load_state() are for resuming the same training script shape, not arbitrary model conversion between unrelated scripts.
Register custom objects only if they implement both state_dict() and load_state_dict().
For ordinary model artifact export, prefer accelerator.unwrap_model(), accelerator.get_state_dict(), accelerator.save(), or accelerator.save_model() rather than treating a training-state checkpoint as a deployment artifact.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Checkpointing And Tracking AI skill do?

Save and resume Accelerate training state, register checkpoint hooks and custom state, log safely across processes, use experiment trackers, profile runs, and clean up memory.

Why use Checkpointing And Tracking on TypingMind?

Because you install it once and use it with any model. Checkpointing And Tracking is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Checkpointing And Tracking in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/VectorSpaceLab/AREX-Skill/tree/main/skills/repositories/repo-skills/accelerate/sub-skills/checkpointing-and-tracking. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Checkpointing And Tracking?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Checkpointing And Tracking?

As many as you like. As long as a model supports skills, you can use Checkpointing And Tracking with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Checkpointing And Tracking AI skill free?

Yes. It is published on GitHub by VectorSpaceLab under the Apache-2.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Update Team Server

openclaw

OrganizationPopular

Update the operator-configured Team server through its canonical owner; trust native validation and acceptance without duplicate checks or schedulers.

Cloud Image Bake

Bake, select, prove, and safely retire a Cloud Worker image with crabbox and config one-liners.

Acp Router

Route plain-language requests for Claude Code, Cursor, Copilot, OpenClaw ACP, OpenCode, Gemini CLI, Qwen, Kiro, Kimi, iFlow, Factory Droid, Kilocode, or explicit ACP harness work into either OpenClaw…

Diffs

Use the diffs tool to produce real, shareable diffs (viewer URL, file artifact, or both) instead of manual edit summaries.

Overview

Installation

Install in TypingMind

Install in another agent

Use it in TypingMind

Works with any AI model

Skill instructions

Checkpointing and Tracking

Route First

Fast Workflows

Bundled Helper

Key Constraints

Bundled files

Frequently asked questions

Related AI Skills

Set up your own AI workspace now