skills.rest
learningmatter-mit
learningmatter-mitOFFICIAL
Explore all skills by this author →
WebsiteGitHub

ml-fairchem-finetune

Fine-tune Fairchem interatomic potentials on labeled structure datasets.

144|21|Updated Jan 8, 2026
Search Tags:#fairchem#model benchmarking#mlip fine-tuning#hydra training#lmdb datasets#materials modeling#energy forces stress
One-click install
npx skills add https://github.com/learningmatter-mit/AtomisticSkills --skill ml-fairchem-finetune
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ml-fairchem-finetune
Source: https://github.com/learningmatter-mit/AtomisticSkills/tree/main/.agents/skills/ml-fairchem-finetune
Command: npx skills add https://github.com/learningmatter-mit/AtomisticSkills --skill ml-fairchem-finetune

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Fine-tuning a foundation Fairchem machine learning interatomic potential to accurately match a specific chemical system or target property without manually wiring an end-to-end training pipeline.

Core Features & Use Cases

  • Dataset preparation from JSON labels: Converts structure dictionaries into extxyz and builds Fairchem-compatible LMDB datasets (including generating training configuration templates).
  • Foundation model fine-tuning: Runs Fairchem training for UMA/ESEN-style foundation potentials using a generated Hydra configuration and optional backbone freezing.
  • Benchmarking and training validation: Benchmarks against a foundation model via the referenced ml-mlip-benchmark skill, then parses logs to confirm convergence and training curves.
  • Model registration for reuse: Registers the newly fine-tuned checkpoint into a local registry so future research tasks can discover and reuse it.

Quick Start

Run the data preparation script for your labeled JSON dataset, specifying your Fairchem base model and output directory, so the Skill generates the LMDBs and the Fairchem fine-tuning configuration for training.

Frequently Asked Questions about ml-fairchem-finetune

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

FAQPage Schema
How do I fine-tune Fairchem interatomic potentials on a custom dataset?▼

Fine-tune Fairchem potentials by preparing LMDB datasets from your labeled structures and generating a Hydra-compatible configuration via provided scripts, then running training to adapt UMA or ESEN foundation models to your specific chemical systems.

What do I need to prepare before fine-tuning a machine learning interatomic potential?▼

Before fine-tuning a machine learning interatomic potential, you need labeled structure datasets formatted as JSON dictionaries, which scripts convert into Fairchem-compatible LMDB datasets and Hydra training configuration templates for energy, forces, and stress supervised learning.

Can I freeze the backbone when fine-tuning foundation models for materials modeling?▼

Yes, you can freeze the backbone when fine-tuning foundation models for materials modeling; the Skill supports optional backbone freezing during Hydra training to efficiently adapt UMA or ESEN potentials while preserving learned representations.

How do I benchmark a fine-tuned model against the original foundation potential?▼

Benchmark a fine-tuned model against the original foundation potential by referencing the ml-mlip-benchmark skill, then parse training logs to validate convergence and compare prediction accuracy across your target materials tasks.

Does this workflow handle stress tensor calculations for atomistic simulations?▼

Yes, the workflow handles stress tensor calculations for atomistic simulations by ensuring unit-consistent stress handling during supervised training of energy, forces, and stress predictions on your custom chemical systems.

How are fine-tuned checkpoints managed after training completes?▼

Fine-tuned checkpoints are registered into a local registry after training completes, enabling future research tasks to discover and reuse the adapted interatomic potentials for specific materials modeling applications.

Related Skills

Explore More Education & Research Skills→
learningmatter-mitlearningmatter-mit
144

ml-mace-finetune

Fine-tune MACE interatomic potentials on labeled atomistic datasets.

Official
Advanced
automatedigitalautomatedigital
1

unsloth

Fine-tune large language models with LoRA and QLoRA optimization.

Official
Intermediate
codeApe-7codeApe-7
17

fine-tuning-expert

Guide dataset validation, hyperparameter optimization, and deployment of fine-tuned large language models.

Community
Advanced
learningmatter-mitlearningmatter-mit
144

ml-matgl-finetune

Fine-tune MatGL interatomic potentials on labeled atomistic datasets with PyTorch Lightning.

Official
Advanced