Skill UI

This skill provides comprehensive guidance on training Mixture of Experts (MoE) models, including architectures like Mixtral and DeepSeek-V3. It details techniques for achieving large-scale model capacity with significantly reduced computational costs (e.g., 5x reduction vs. dense models). Topics covered include top-k routing, load balancing, expert parallelism, and integration with frameworks like DeepSpeed and HuggingFace for efficient, resource-constrained training.

This guide details advanced architectural patterns for building complex AI systems using multiple Language Model instances (agents). It addresses the single-agent context bottleneck by distributing work, enabling parallelism, and facilitating specialization across different subtasks. Key patterns include Supervisor/Orchestrator, Peer-to-Peer/Swarm, and Hierarchical designs, providing guidelines for context isolation and effective coordination protocols.

nanoGPT is a simplified, minimalist implementation designed specifically for educational purposes. It reproduces the core architecture of GPT-2 (124M) using clean, hackable code, allowing users to understand the entire transformer pipeline from scratch. The framework supports full workflows, including data preparation on various datasets (e.g., Shakespeare, OpenWebText), model training, and text generation, making it ideal for students and researchers learning NLP and deep learning principles.

This comprehensive skill profile models an expert network engineer specializing in modern cloud networking, security architectures, and performance optimization. It covers multi-cloud environments (AWS, Azure, GCP), advanced concepts like Zero Trust, Service Mesh (Istio), and high-availability design. Key capabilities include implementing IaC (Terraform), load balancing, DNS management, and deep network troubleshooting. Ideal for complex DevOps, SRE, or infrastructure design roles.

Guides researchers through nnsight’s tracing and activation APIs so you can inspect or modify PyTorch internals locally or via NDIF-powered remote runs on 70B+ foundation models without local GPUs.

Expert guidance on designing highly scalable, distributed NoSQL databases, specifically covering Apache Cassandra and Amazon DynamoDB. This skill emphasizes the 'Query-First' modeling approach, teaching best practices like partitioning, denormalization, single-table design, and avoiding common anti-patterns (e.g., hot partitions, joins) to ensure optimal read/write performance at massive scale.

This skill automates the setup and configuration of Ollama, allowing users to deploy Large Language Models (LLMs) locally. It performs system assessment (OS, GPU, RAM), selects optimal models based on hardware availability, and handles installation across macOS, Linux, and Docker. Use this when you require free, self-hosted AI alternatives, need offline inference capabilities, or wish to eliminate reliance on costly hosted API services.

Automate OpenAI API workflows through the Composio MCP integration, covering text and multimodal responses with structured outputs, embeddings, image generation, and model discovery.

This bundle collects comprehensive diagnostic evidence from the OpenEvidence clinical decision support API. It validates API credentials, checks query endpoint health, retrieves evidence citations, and captures model metadata. Use it to diagnose complex issues like query failures, missing references, or high latency without needing direct dashboard access.

This comprehensive reference architecture guides the integration of OpenEvidence into clinical decision support systems (CDS). It is designed for healthcare platforms needing evidence-based query processing, deterministic citation tracking, and full audit logging. Key considerations include HIPAA compliance, optimizing for sub-second response times in point-of-care workflows, and establishing robust data models and event pipelines.

This SDK provides comprehensive, production-ready patterns for interacting with the OpenEvidence clinical evidence API. It enables developers to efficiently query medical literature, retrieve clinical guidelines, and generate evidence-based recommendations. Key features include a singleton client for consistent authentication, robust error wrappers (handling retries, invalid keys), and structured data models designed to ensure audit compliance and detailed citation provenance in complex healthcare applications.

High-performance RLHF framework that combines Ray, vLLM, and ZeRO-3 to accelerate PPO/GRPO/RLOO/DPO training for 7B-70B+ models on distributed GPU clusters, providing a single-stack workflow for reward models and policy optimization.