Search

Unified LLM connectors for OpenAI, Google Gemini, Ollama, and LM Studio.

LangChain and LangGraph observability for Tuner SDKs

Download model files from local paths, HTTP, S3-compatible storage, GCS, and Git LFS.

Enterprise-grade resilient vLLM client network and preflight validation engine.

Seren: A Self-Refining Decoder-Only LLM with GQA, RoPE, Flash Attention, and a novel two-pass self-refinement mechanism.

Simple, tunable sentiment analysis from sentence embeddings — TensorFlow-free. Python sibling of the R package sentiment.ai.

Fast HEALPix center-in-polygon coverage for convex spherical polygons.

CLI utilities for listing and downloading Apple IPSW firmware.

Intelligent modular memory management system for LLMs

Tiny, drop-in Hugging Face KV cache compression using KIVI-style low-bit KV storage.

Simple OpenAI helper functions

Ordered SQL migrations with a version table for lakehouse storage

Reusable Manim components for creating LLM and Transformer explanation videos.

A small single-session chat client for OpenAI-compatible Responses API endpoints.

A lightweight Python framework for portable, versioned, reusable LLM extraction tasks.

Publication-grade figures for scientific papers — reproducible matplotlib & TikZ, with a quality bar, composition paradigms, and journal-grade craft. A self-contained Claude skill.

Daily Leetcode masterclass with Claude code DSA mentor

Claude Code skill for the full PSADT v4.x → Intune Win32 packaging lifecycle: guided build, packaging, testing, and optional direct Graph upload.

AI 模型训练数据标注规则文档撰写 skill：覆盖文本对话（单/多轮·CoT·RAG·Reward·Agent）与多模态（文生图/视频·VQA·ASR·TTS）共 11 种标注类型

Restore the Ask Gemini / Gemini in Chrome sidebar button after a macOS Chrome update flips the local Variations region seed to a non-US code. Bash tool + Claude Code skill.