pipeline

Automate post-training workflows by syncing checkpoints and running evaluations.

Updated Aug 27, 2026

Search Tags:#automation #evaluation #pipeline #checkpoint #ml #multicluster

Ready to run

npx skills add https://github.com/kang-jaehyun/Mac-the-researcher --skill pipeline-kang-jaehyun

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: pipeline
Source: https://github.com/kang-jaehyun/Mac-the-researcher/tree/main/.claude/skills/pipeline
Command: npx skills add https://github.com/kang-jaehyun/Mac-the-researcher --skill pipeline-kang-jaehyun

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Automates end-to-end post-training workflows by detecting completion, syncing checkpoints to the evaluation cluster, and running evaluations with result collection.

Core Features & Use Cases

Automated detection of training completion and checkpoint synchronization to the skt cluster.
Submission and execution of evaluations on L40S, with metrics parsing and results aggregation.
Update of the experiment store and optional sharing of results to external hubs for collaboration and traceability.

Quick Start

To start automated post-training pipelines, run run_pipeline with source_cluster, job_id, eval_task_name, and dest_cluster.

Frequently Asked Questions about pipeline

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I automate checkpoint syncing and evaluation across multiple ML clusters?▼

To automate checkpoint syncing across ML clusters, this pipeline detects training completion, syncs checkpoints to the evaluation cluster, runs evaluations on L40S hardware, and parses the resulting metrics automatically.

What is an end-to-end post-training pipeline for machine learning experiments?▼

An end-to-end post-training pipeline automates ML workflows by detecting when training finishes, migrating checkpoints to a destination cluster, executing evaluations, and aggregating the results into an experiment store.

Can I run evaluations on L40S hardware after syncing checkpoints from another cluster?▼

Yes, evaluations can be run on L40S hardware after syncing checkpoints from a source cluster. The pipeline submits evaluation jobs to the destination cluster and parses the output metrics for aggregation.

How do I start an automated post-training workflow using sbatch outputs?▼

You can start the automated post-training workflow by running the pipeline with the source cluster, job ID, evaluation task name, and destination cluster, relying on sbatch outputs and cluster availability assumptions.

Does the pipeline require access to an experiment store for evaluation result collection?▼

Yes, the pipeline requires access to an experiment store to update and optionally share evaluation results to external hubs, ensuring collaboration and traceability for your ML experiments.

Why does my multicluster ML evaluation fail when checkpoints are not synced?▼

Multicluster ML evaluations fail without checkpoint syncing because the pipeline relies on the find_checkpoint_path and sync_checkpoint tooling to migrate model states to the destination cluster before running evaluations.