mamba-architecture
davila7State-space model with O(n) complexity vs Transformers' O(n²). 5× faster inference, million-token sequences, no KV cache. Selective SSM with hardware-aware design. Mamba-1 (d_state=16) and Mamba-2 (d_state=128, multi-head). Models 130M-2.8B on HuggingFace.
evaluating-llms-harness
davila7Evaluates LLMs across 60+ academic benchmarks (MMLU, HumanEval, GSM8K, TruthfulQA, HellaSwag). Use when benchmarking model quality, comparing models, reporting academic results, or tracking training progress. Industry standard used by EleutherAI, HuggingFace, and major labs. Supports HuggingFace, vLLM, APIs.
huggingface-tokenizers
davila7Fast 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.
huggingface-accelerate
davila7Simplest 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.
stable-diffusion-image-generation
davila7State-of-the-art text-to-image generation with Stable Diffusion models via HuggingFace Diffusers. Use when generating images from text prompts, performing image-to-image translation, inpainting, or building custom diffusion pipelines.
hqq-quantization
davila7Half-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.
moe-training
davila7Train Mixture of Experts (MoE) models using DeepSpeed or HuggingFace. Use when training large-scale models with limited compute (5× cost reduction vs dense models), implementing sparse architectures like Mixtral 8x7B or DeepSeek-V3, or scaling model capacity without proportional compute increase. Covers MoE architectures, routing mechanisms, load balancing, expert parallelism, and inference optimization.
peft-fine-tuning
davila7Parameter-efficient fine-tuning for LLMs using LoRA, QLoRA, and 25+ methods. Use when fine-tuning large models (7B-70B) with limited GPU memory, when you need to train less than 1% of parameters with minimal accuracy loss, or for multi-adapter serving. HuggingFace's official library integrated with transformers ecosystem.
quantizing-models-bitsandbytes
davila7Quantizes LLMs to 8-bit or 4-bit for 50-75% memory reduction with minimal accuracy loss. Use when GPU memory is limited, need to fit larger models, or want faster inference. Supports INT8, NF4, FP4 formats, QLoRA training, and 8-bit optimizers. Works with HuggingFace Transformers.
llamaguard
davila7Meta's 7-8B specialized moderation model for LLM input/output filtering. 6 safety categories - violence/hate, sexual content, weapons, substances, self-harm, criminal planning. 94-95% accuracy. Deploy with vLLM, HuggingFace, Sagemaker. Integrates with NeMo Guardrails.
llama-factory
davila7Expert guidance for fine-tuning LLMs with LLaMA-Factory - WebUI no-code, 100+ models, 2/3/4/5/6/8-bit QLoRA, multimodal support
evaluating-code-models
davila7Evaluates code generation models across HumanEval, MBPP, MultiPL-E, and 15+ benchmarks with pass@k metrics. Use when benchmarking code models, comparing coding abilities, testing multi-language support, or measuring code generation quality. Industry standard from BigCode Project used by HuggingFace leaderboards.
weights-and-biases
davila7Track ML experiments with automatic logging, visualize training in real-time, optimize hyperparameters with sweeps, and manage model registry with W&B - collaborative MLOps platform
mlflow
davila7Track ML experiments, manage model registry with versioning, deploy models to production, and reproduce experiments with MLflow - framework-agnostic ML lifecycle platform
fine-tuning-with-trl
davila7Fine-tune LLMs using reinforcement learning with TRL - SFT for instruction tuning, DPO for preference alignment, PPO/GRPO for reward optimization, and reward model training. Use when need RLHF, align model with preferences, or train from human feedback. Works with HuggingFace Transformers.
axolotl
davila7Expert guidance for fine-tuning LLMs with Axolotl - YAML configs, 100+ models, LoRA/QLoRA, DPO/KTO/ORPO/GRPO, multimodal support
add-model-descriptions
huggingfaceAdd descriptions for new models from the HuggingFace router to chat-ui configuration. Use when new models are released on the router and need descriptions added to prod.yaml and dev.yaml. Triggers on requests like "add new model descriptions", "update models from router", "sync models", or when explicitly invoking /add-model-descriptions.
hugging-face-tool-builder
huggingfaceUse this skill when the user wants to build tool/scripts or achieve a task where using data from the Hugging Face API would help. This is especially useful when chaining or combining API calls or the task will be repeated/automated. This Skill creates a reusable script to fetch, enrich or process data.
hf-mcp
huggingfaceUse Hugging Face Hub via MCP server tools. Search models, datasets, Spaces, papers. Get repo details, fetch documentation, run compute jobs, and use Gradio Spaces as AI tools. Available when connected to the HF MCP server.
hugging-face-evaluation
huggingfaceAdd 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.
hugging-face-model-trainer
huggingfaceThis 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.
hugging-face-jobs
huggingfaceThis 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.
hugging-face-paper-publisher
huggingfacePublish 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.
hugging-face-trackio
huggingfaceTrack and visualize ML training experiments with Trackio. Use when logging metrics during training (Python API) or retrieving/analyzing logged metrics (CLI). Supports real-time dashboard visualization, HF Space syncing, and JSON output for automation.
hugging-face-datasets
huggingfaceCreate and manage datasets on the Hugging Face Hub. Supports initializing repositories, defining configs and system prompts, streaming row updates, and SQL-based dataset querying and transformation. Designed to work alongside the HF MCP server for comprehensive dataset workflows.
hugging-face-cli
huggingfaceExecute 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.
comfy-cli
openclawInstall, manage, and run ComfyUI instances. Use when setting up ComfyUI, launching servers, installing/updating/debugging custom nodes, downloading models from CivitAI/HuggingFace, managing workspaces, running API workflows, or troubleshooting node conflicts with bisect.
Nia
openclawIndex and search code repositories, documentation, research papers, HuggingFace datasets, local folders, and packages with Nia AI. Includes Oracle autonomous research, dependency analysis, context sharing, and code advisor.
common-skills
llama-farmBest practices for the Common utilities package in LlamaFarm. Covers HuggingFace Hub integration, GGUF model management, and shared utilities.
daggr
Build DAG-based AI pipelines connecting Gradio Spaces, HuggingFace models, and Python functions into visual workflows. Use when asked to create a workflow, build a pipeline, connect AI models, chain Gradio Spaces, create a daggr app, build multi-step AI applications, or orchestrate ML models. Triggers on: "build a workflow", "create a pipeline", "connect models", "daggr", "chain Spaces", "AI pipeline".
huggingface-classifier
a5c-aiHugging Face transformer model fine-tuning and inference for intent classification
evaluating-code-models
sangrokjungEvaluates code generation models across HumanEval, MBPP, MultiPL-E, and 15+ benchmarks with pass@k metrics. Use when benchmarking code models, comparing coding abilities, testing multi-language support, or measuring code generation quality. Industry standard from BigCode Project used by HuggingFace leaderboards.
evaluating-llms-harness
sangrokjungEvaluates LLMs across 60+ academic benchmarks (MMLU, HumanEval, GSM8K, TruthfulQA, HellaSwag). Use when benchmarking model quality, comparing models, reporting academic results, or tracking training progress. Industry standard used by EleutherAI, HuggingFace, and major labs. Supports HuggingFace, vLLM, APIs.
discover-ml
randAutomatically discover machine learning and AI skills when working with machine learning, PyTorch, training, inference, RAG, embeddings, fine-tuning, LLM, DSPy, HuggingFace, or diffusion models. Activates for ML development tasks.
count-dataset-tokens
letta-aiThis skill provides guidance for counting tokens in datasets using specific tokenizers. It should be used when tasks involve tokenizing dataset content, filtering data by domain or category, and aggregating token counts. Common triggers include requests to count tokens in HuggingFace datasets, filter datasets by specific fields, or use particular tokenizers (e.g., Qwen, DeepSeek, GPT).
mteb-leaderboard
letta-aiThis skill provides guidance for retrieving and verifying information from dynamic ML leaderboards (MTEB, Scandinavian Embedding Benchmark, HuggingFace leaderboards, etc.) with specific temporal requirements. It should be used when tasks involve finding top-performing models, rankings, or benchmark results as of a specific date, especially when the data source is frequently updated.
hf-model-inference
letta-aiGuidance for deploying HuggingFace models as inference APIs/services. This skill applies when tasks involve downloading pre-trained models from HuggingFace Hub, creating REST APIs for model inference, building Flask/FastAPI services around ML models, or setting up sentiment analysis, text classification, or other NLP inference endpoints.
import
the-crypt-keeperImport model evaluation results into ReasonScape m12x dataset. Use when importing new models, quantization variants, or adding evaluation data. Keywords - model, import, evaluate, dataset, results, eval, hf, huggingface.
huggingface-transformers
applied-artificial-intelligenceHugging Face Transformers best practices including model loading, tokenization, fine-tuning workflows, and inference optimization. Use when working with transformer models, fine-tuning LLMs, implementing NLP tasks, or optimizing transformer inference.
huggingface_transformers
vuralserhat86Hugging Face Transformers best practices including model loading, tokenization, fine-tuning workflows, and inference optimization. Use when working with transformer models, fine-tuning LLMs, implementing NLP tasks, or optimizing transformer inference.
model_finetuning
vuralserhat86Fine-tune LLMs using reinforcement learning with TRL - SFT for instruction tuning, DPO for preference alignment, PPO/GRPO for reward optimization, and reward model training. Use when need RLHF, align model with preferences, or train from human feedback. Works with HuggingFace Transformers.
transformers-huggingface
MindrallyExpert guidance for working with Hugging Face Transformers library for NLP, computer vision, and multimodal AI tasks.
policyengine-district-analysis
PolicyEngineAnalyze policy impacts for congressional districts and representatives' constituents. Use when the user mentions a specific district (NY-17, CA-52), a representative's name, or asks about geographic policy impacts at district level. Provides HuggingFace district datasets.
huggingface
TerminalSkillsWork with Hugging Face's ecosystem for machine learning — transformers library, model hub, tokenizers, inference pipelines, and fine-tuning. Covers downloading pre-trained models, running inference, training custom models, and publishing to the Hub.
model-card
andikarachmanGenerate standardized model documentation following HuggingFace Model Card and NVIDIA Model Card++ formats. Use when preparing a model for deployment or handoff.
processing-data-pipelines
gitwalterDataset preparation for ML, HuggingFace datasets, custom loaders, data validation with Great Expectations, feature engineering, data versioning with DVC, preprocessing pipelines
dataset-processing-multiprocessing
anhvthAdvanced pattern for processing large HuggingFace datasets in parallel using speedy_utils.multi_process, with sharding, worker initialization, temporary file management, and safe cleanup.
onnx-webgpu-converter
jakerainsConvert HuggingFace transformer models to ONNX format for browser inference with Transformers.js and WebGPU. Use when given a HuggingFace model link to convert to ONNX, when setting up optimum-cli for ONNX export, when quantizing models (fp16, q8, q4) for web deployment, when configuring Transformers.js with WebGPU acceleration, or when troubleshooting ONNX conversion errors. Triggers on mentions of ONNX conversion, Transformers.js, WebGPU inference, optimum export, model quantization for browser, or running ML models in the browser.