Data agents

transformers safetensors gguf gemma2 text-generation conversational

transformers gguf llama text-generation en de

transformers gguf llama text-generation en de

docker region:us

gguf text-generation en dataset:SicariusSicariiStuff/UBW_Tapestries base_model:SicariusSicariiStuff/Assistant_Pepe_70B base_model:quantized:SicariusSicariiStuff

Automate QQ and WeChat farm games with multi-account support, real-time logs, data analysis, and a web control panel using Node.js.

working on something and tweaking somethings to just get a better clarity on the internal changes in lang graph, crew Ai, tool calling and setting up multi agent using RAG: Primary focus HEALTHCARE, Mostly Synth data

Semantic gravity database in Rust: related records live together on disk, so graph traversal is a range scan and vector search is exact — never approximate. Runs in 256 MB, with a built-in MCP server for agents.

AI-powered repository intelligence platform that analyzes GitHub repositories, explains beginner-friendly issues, generates contribution roadmaps, and produces structured PDF guides for open-source contributors.

Spinosa is a local framework for working with LLMs. It features a set of sub-agents coherent with modern SKILL and AGENTS protocols to streamline research and build better practices.

Self-hosted agentic AI for strategy & business intelligence — market research, competitive analysis, and executive-level recommendations, reachable via Telegram, Discord, WhatsApp, or Slack. Model-provider agnostic, deployable 24/7.

Go SDK for Dakera AI agent memory — self-hosted, 88.2% LoCoMo. Vectors, hybrid search, knowledge graphs, sessions.

Open-source LLM cost monitoring & optimization platform. Track spend across OpenAI, Anthropic, DeepSeek & OpenRouter in real-time.

Product-strategy workflow you run with an AI agent — idea → analysis → strategy → sprint plan, with living hypothesis/risk/metric registers. Nothing invented; every claim sourced.

🤖 Tajan – An intelligent bilingual (Persian/English) intent recognition chatbot powered by BiLSTM deep learning. Features: offline Persian/English lemmatizer, smart data augmentation (1.8x), GUI interface, and 419 intent classes with ~66% accuracy. Fully customizable, open-source, and runs without internet. Built with TensorFlow, Hazm, and NLTK.

🎓 AI-Powered Timetable Generation System for Educational Institutions | Genetic Algorithms | React + Node.js + MongoDB | Real-time Analytics | Gemini AI Chatbot

A native macOS reader that treats a PDF like a book — realistic page turns, 9 reading themes, and AI author personas grounded in your own library. SwiftUI + AppKit, BYOK.

Tử Vi Horoscope MCP Server - a Python MCP server for generating, analyzing, and managing Vietnamese Tử Vi astrology charts with clean JSON output for LLM agents.

四大师并行投研系统 — 段永平/巴菲特/芒格/李录 多视角 AI 投研，叠加 V10 TextGrad 自进化引擎(计算图+文本梯度反传)、18 Skills、确定性金融工具链。面向 OpenClaw/QwenPaw 类 Agent 运行时。

An open-ended, agent-driven research loop: proposes ideas, runs real experiments, and writes up the paper — end to end, unattended.

开源 AI-Native ERP 参考实现：自然语言驱动企业运营，7 领域 Agent + 28+ 受控 Tool + RAG 知识库，.NET 10 + React，带完整治理链路与 Demo 数据。Open-source AI-Native ERP reference: natural language drives business ops with 7 domain agents, 28+ governed tools, and RAG — .NET 10 + React, full governance pipeline & demo data.

Why the LLM Wiki pattern breaks past 1,000 files, and the five invariants that fix it. A quote is a pointer, not a copy. Paper + full decision record. Free to read noncommercially with attribution; commercial use licensed.

The original data contains 500k conversations, but this is just a sample. You can download the 10k dataset with personal and sensitive information removed.

Governed graph-native agent memory: knowledge extraction, fusion, hybrid RAG, scoped access tokens. A push for data sovereignty.

Local-first, AI-native desktop workbench for product decisions: evidence, competitors, exact unit economics, and an evidence-linked report in one project folder you own. AI drafts and waits; nothing acts on its own. Read-only MCP server included. No cloud, no accounts, no telemetry.

Self-hosted agentic AI for strategy & business intelligence — market research, competitive analysis, and executive-level recommendations, reachable via Telegram, Discord, WhatsApp, or Slack. Model-provider agnostic, deployable 24/7.

Long-running Telegram agent on MTProto: durable runs that resume after a crash, markdown/JSONL canon instead of a database, and a subscription-backed LLM relay in Rust

companion.ai: Craft and converse with personalized AI companions for engaging chats and insightful discussions.

An AI-powered journaling application that organizes and analyzes daily thoughts automatically. 🧠✨

中英双语知识库，为推荐系统与计算广告 auto research 提供可精确引用的技术资料 — 召回/排序/重排、生成式推荐（语义 ID、HSTU、OneRec、EGA、OneRec-Think、RecZero、扩散模型）、计算广告（GFP/GSP/VCG、智能出价、预估校准、WAND/ANN 检索、反作弊）。

In-flight circuit breaker & loop recovery for LangGraph, LangChain, OpenAI & Anthropic agents. Catches reasoning loops, prevents runaway LLM costs with budget hard caps, and auto-injects Reflect & Pivot prompts.

A 4-step multi-agent RAG system (preference understanding, semantic alignment, contextual summarization, ranking) leveraging user watch history, mood, and movie metadata; deployed via Streamlit UI and evaluated using NDCG@5 and HIT@5.

AI-powered B2B lead finder: discovery on licensed lead data, LLM fit qualification with the reason written out, CSV out. The finder core, extracted from OpenOutreach.

Lehman-Yao B-link tree concurrent index enabling lock-free read operations across dynamic node splits with high-key right links.

Multiversion Timestamp Ordering (MVTO) concurrency control engine supporting non-blocking reads and Thomas Write Rule serialization.

Strict Two-Phase Locking (SS2PL) concurrency manager with shared/exclusive lock upgrades and DFS wait-for graph cycle deadlock detection.

Datar-Gionis-Indyk-Motwani (DGIM) streaming algorithm for estimating frequency and bit counts over massive sliding windows in logarithmic memory.

Production-grade genpark-aries-wal-crash-recovery-engine-skill skill for AI agents

Production-grade genpark-percolator-distributed-snapshot-transactions-skill skill for AI agents

Log-Structured Merge Tree (LSM-Tree) engine with in-memory skip-list memtable, immutable SSTable flushes, and leveled compaction.

Observed-Removed Set (OR-Set) CRDT resolving concurrent add/remove conflicts with unique element tagging and add-wins convergence.

Production-grade genpark-lsm-tree-compaction-memtable-engine-skill skill for AI agents

Linux Read-Copy Update (RCU) synchronization model supporting zero-overhead lockless read sections and deferred memory reclamation.

Production-grade genpark-lsm-tree-compaction-memtable-engine-skill skill for AI agents

Production-grade genpark-daubechies-d4-wavelet-transform-skill skill for AI agents

Kung-Robinson Optimistic Concurrency Control (OCC) engine with backward validation phase for read-heavy transaction processing.

Cooley-Tukey Radix-2 Fast Fourier Transform (FFT) frequency analyzer converting discrete time-domain signals into spectral complex frequency bins.

Inverted File Index with Flat Quantization (IVF-Flat) vector engine performing clustered Voronoi cell partitioning and fast top-K Euclidean ranking.

Cooley-Tukey Radix-2 Fast Fourier Transform (FFT) frequency analyzer converting discrete time-domain signals into spectral complex frequency bins.

Positive-Negative Counter (PN-Counter) Conflict-Free Replicated Data Type ensuring eventual consistency across partitioned distributed nodes with 0 pip.

Strict Two-Phase Locking (SS2PL) concurrency manager with shared/exclusive lock upgrades and DFS wait-for graph cycle deadlock detection.

Mel-scale triangular filterbank spectral analyzer mapping linear frequency spectra to perceptual log-mel energies for speech AI.

Conflict-free replicated data types (CRDT) implementing state-based PN-Counters and Last-Write-Wins registers.

Production-grade genpark-hnsw-hierarchical-navigable-small-world-skill skill for AI agents

Topological Data Analysis (TDA) Vietoris-Rips filtration generator tracking metric space topological features across expanding radii.

Linux Read-Copy Update (RCU) synchronization model supporting zero-overhead lockless read sections and deferred memory reclamation.

Reservoir Sampling Algorithm R selecting a uniform random sample of k items from an unbounded, single-pass data stream in O(N).

Production-grade genpark-hilbert-transform-analytic-signal-skill skill for AI agents

Abstract Interpretation static analyzer using the numeric Interval Domain to prove buffer bounds safety and absence of division-by-zero errors.

Lehman-Yao B-link tree concurrent index enabling lock-free read operations across dynamic node splits with high-key right links.

Raft log compaction and state machine snapshotting engine discarding committed entries and preserving consistent cluster metadata.

Doubly-Connected Edge List (DCEL) / Half-Edge polygon mesh structure supporting topological face circulation and Euler characteristic auditing.

Production-grade genpark-aries-wal-crash-recovery-engine-skill skill for AI agents

Production-grade genpark-cooley-tukey-radix2-fft-engine-skill skill for AI agents

Production-grade genpark-cooley-tukey-radix2-fft-engine-skill skill for AI agents

Legal Intelligence Terminal — Hybrid RAG for UAE & Indian Law, Taxation & Compliance. Built by Evolucent AI, engineered by Eulogik.

Memory infrastructure for AI agents: store, recall by meaning or association, ingest documents, and track beliefs in one Python library.

Autonomous web browser action planner mapping goals into resilient selector action sequences

A curated list of agentic environment synthesis, evolution, quality & scaling — built on three surveys (AEE, Environment Scaling, ACE lens)

Production-grade genpark-hnsw-hierarchical-navigable-small-world-skill skill for AI agents

Mel-scale triangular filterbank spectral analyzer mapping linear frequency spectra to perceptual log-mel energies for speech AI.

Anti-entropy Merkle tree range reconciliation protocol for fast replica divergence detection and state synchronization.

Production-grade genpark-tdigest-streaming-quantile-sketch-skill skill for AI agents

A domain-specific OFFLINE RAG system for intelligence analysis, built to retrieve and rank terrorism incident data from the Global Terrorism Database. Uses hybrid semantic + keyword retrieval, metadata scoring, and attack-type reranking, achieving 95.2% Precision@5 on evaluation queries.

M1 is a research project exploring large-scale music generation using diffusion transformers. This repository contains the implementation of our proposed architecture combining recent advances in diffusion models, transformer architectures, and music processing.

Production-grade RAG pipelines with evaluation baked in

High-performance vector database & RAG memory layer - hybrid search, embeddings, RAPTOR trees, BM25 fusion for AI systems.

McPortfolio: A Model Context Protocol server providing 9 specialized tools for LLM-driven portfolio optimization using natural language, covering mean-variance to machine learning approaches.

AIOS — AI Operating System. A 5-layer AI OS that wraps around any business: Context, Data, Intelligence, Automation, Build.

Your own personal AI interview assistant. Any OS. Any Platform.

Use a RAG Langchain agent that connects to your supermarket's inventory database and get information about product stocks and locations.

Universal AI Consciousness Genesis & Navigation System

Google Gemini Agentic AI Vision for Wildlife Analytics

🎶 Enhance audio quality with ComfyUI-AudioSR, a versatile tool for upscaling sounds to 48kHz for better clarity and listening experience.

Smart AI career assistant - resume parsing, skill gap analysis, and personalized guidance for students and professionals.

The LangGraph project implements a "Reflection Agent" designed to iteratively refine answers to user queries using a Large Language Model (LLM) and web search. It simulates a research process where an initial answer is generated, critiqued, and revised based on information gathered from web searches, all managed by a LangGraph workflow.

A simple implementation of the RAG architecture on an Investor database to give you a VC fund raising assistant

Siomi is a platform multi-agent AI to generate answers to our clients by messages, documents, graphics, report from big data and more

RockTalk is a private, locally running chatbot using Streamlit for the frontend, LangChain for the logic, and Amazon Bedrock as the backend. The webapp provides a user-friendly interface for interacting with various Language Models (LLMs) with advanced features for customization and data input, while keeping your session data local and private.

A RAG assistant using Ollama (Mistral), Qdrant vector DB, and Streamlit UI. Upload documents, scrape web pages, and interact with your data using real-time, session-isolated chat.

This is a document question answering app made with LangChain and deployed on Streamlit where you can upload a .txt or .docx file, ask questions based on the file and an LLM like Falcon-7B or Dolly-V2-3B answers it. ChromaDB is used as the vector database.

AI-powered book recommendations engine that fuses semantic understanding and emotion intelligence to deliver deeply relevant book suggestions. Powered by LangChain, OpenAI embeddings, and emotion-aware NLP across ~7,000 titles, with an interactive Gradio experience for intuitive exploration.

This repo mainly encapsulates an LLM model + front-end + back-end for generating stories for children. It was a personal project in April last year and is now open source.

Python-based system designed to transcribe audio files, split the transcripts into manageable chunks, create text embeddings using HuggingFace models, and employ advanced question-answering models for retrieval-based QA.

This AI-powered app enables users to ask questions to their Notion database using natural language.

RAG (Retrieval-Augmented Generation) app integrated with a voice assistant and knowledge base management system.

All CPU efficient GPU-less Financial Analysis RAG Model with Qdrant, Langchain and GPT4All x Mistral-7B, run RAG without any GPU support!

Scientific Research Assistant built with LLMs, Retrieval Augmented Generation, and Semantic Search.

AI Engineering Bootcamp for Programmers - A 19-module, 18-week immersive study plan designed for programmers looking to master AI engineering through real-world projects, foundational theory, and practical tools.

Upload PDFs, ask questions, and get answers with source references using LangChain, Gradio, and local document storage. An open-source AI-powered Q&A tool with source transparency.