openscience

Post-training 4-bit quantization for LLMs with minimal accuracy loss. Use for deploying large models (70B, 405B) on consumer GPUs, when you need 4× memory reduction with <2% perplexity degradation, or for faster inference (3-4× speedup) vs FP16. Integrates with transformers and PEFT for QLoRA fine-tuning.

Ultra-fast LLM inference on custom LPU hardware. OpenAI-compatible API at api.groq.com. Lowest latency in the industry (500-1000+ tok/s). Supports chat completions, vision, audio (Whisper STT + TTS), tool calling, JSON mode, and streaming. Free tier available. Inference only — no training.

Expert guidance for GRPO/RL fine-tuning with TRL for reasoning and task-specific model training

High-performance toolkit for genomic interval analysis in Rust with Python bindings. Use when working with genomic regions, BED files, coverage tracks, overlap detection, tokenization for ML models, or fragment analysis in computational genomics and machine learning applications.

Control LLM output with regex and grammars, guarantee valid JSON/XML/code generation, enforce structured formats, and build multi-step workflows with Guidance - Microsoft Research's constrained generation framework

Query NHGRI-EBI GWAS Catalog for SNP-trait associations. Search variants by rs ID, disease/trait, gene, retrieve p-values and summary statistics, for genetic epidemiology and polygenic risk scores.

Hamiltonian mechanics — symplectic integrators (leapfrog, Yoshida), Hamilton's equations, Poisson brackets, canonical transformations, action-angle variables, and KAM theory analysis. Use for energy-conserving long-time integration of conservative systems.

Create a self-contained continuation packet for another agent. Use when the user invokes /handoff, asks to transfer work, or needs the current context saved before another agent continues.

Patterns for loading PDE simulation datasets (PDEBench, PhiFlow, JAX-CFD) from HDF5 files. Handles layout detection (single tensor vs separate variables), spatial/temporal downsampling, multi-variable systems, HuggingFace and DaRUS data sources, and efficient PyTorch DataLoader creation. Use when preparing PDE data for neural operator training.

Lightweight WSI tile extraction and preprocessing. Use for basic slide processing tissue detection, tile extraction, stain normalization for H&E images. Best for simple pipelines, dataset preparation, quick tile-based analysis. For advanced spatial proteomics, multiplexed imaging, or deep learning pipelines use pathml.

Access Human Metabolome Database (220K+ metabolites). Search by name/ID/structure, retrieve chemical properties, biomarker data, NMR/MS spectra, pathways, for metabolomics and identification.

Half-Quadratic Quantization for LLMs without calibration data. Use when quantizing models to 4/3/2-bit precision without needing calibration datasets, for fast quantization workflows, or when deploying with vLLM or HuggingFace Transformers.

Simplest distributed training API. 4 lines to add distributed support to any PyTorch script. Unified API for DeepSpeed/FSDP/Megatron/DDP. Automatic device placement, mixed precision (FP16/BF16/FP8). Interactive config, single launch command. HuggingFace ecosystem standard.

Execute Hugging Face Hub operations using the `hf` CLI. Use when the user needs to download models/datasets/spaces, upload files to Hub repositories, create repos, manage local cache, or run compute jobs on HF infrastructure. Covers authentication, file transfers, repository creation, cache operations, and cloud compute.

Create and manage datasets on Hugging Face Hub. Supports initializing repos, defining configs/system prompts, streaming row updates, and SQL-based dataset querying/transformation. Designed to work alongside HF MCP server for comprehensive dataset workflows.

Add and manage evaluation results in Hugging Face model cards. Supports extracting eval tables from README content, importing scores from Artificial Analysis API, and running custom model evaluations with vLLM/lighteval. Works with the model-index metadata format.

This skill should be used when users want to run any workload on Hugging Face Jobs infrastructure. Covers UV scripts, Docker-based jobs, hardware selection, cost estimation, authentication with tokens, secrets management, timeout configuration, and result persistence. Designed for general-purpose compute workloads including data processing, inference, experiments, batch jobs, and any Python-based tasks. Should be invoked for tasks involving cloud compute, GPU workloads, or when users mention running jobs on Hugging Face infrastructure without local setup.

This skill should be used when users want to train or fine-tune language models using TRL (Transformer Reinforcement Learning) on Hugging Face Jobs infrastructure. Covers SFT, DPO, GRPO and reward modeling training methods, plus GGUF conversion for local deployment. Includes guidance on the TRL Jobs package, UV scripts with PEP 723 format, dataset preparation and validation, hardware selection, cost estimation, Trackio monitoring, Hub authentication, and model persistence. Should be invoked for tasks involving cloud GPU training, GGUF conversion, or when users mention training on Hugging Face Jobs without local GPU setup.

Publish and manage research papers on Hugging Face Hub. Supports creating paper pages, linking papers to models/datasets, claiming authorship, and generating professional markdown-based research articles.

Fast tokenizers optimized for research and production. Rust-based implementation tokenizes 1GB in <20 seconds. Supports BPE, WordPiece, and Unigram algorithms. Train custom vocabularies, track alignments, handle padding/truncation. Integrates seamlessly with transformers. Use when you need high-performance tokenization or custom tokenizer training.