lift is a structured extraction model from Datalab that pulls structured JSON out of PDFs and images.

AI-powered open-source tool that turns long videos into viral-ready shorts with automatic highlights, subtitles, hooks, cropping, translation and voiceover

An intuitive console agent to fine-tune AI models with little to no experience required!

A multi-language text rewriting tool designed to bypass such ai detection services as GPTZero, Turnitin and more

Free open-source extractor for AI coding assistant chat histories. Supports Claude Code, Cursor, Windsurf, Aider, Cline/Roo Code, and more.

오철칙(Iron Laws) - AI가 만든 코드의 보안·구조·유지보수 문제를 짚어 주는 소스코드 점검 CLI

CapCut Commerce Pro — video editor for e-commerce & marketing. 77 MB · Info page and download link. Third-party software; all rights belong to the original authors.

A Claude Code skill that designs a brand identity as one system: logo, typography and colour palette built together from one brief, measured, compared as identity cards, and delivered as a brand guidelines kit.

A curated list of plugins, apps, tools and resources for Tern, Stencil's Rust-native neoterminal

prism方案解决chatgpt降智

Shotcut v26.9.27 — Shotcut is a free, open source, cross-platform video editor. Info page and download link. Third-party software; all rights belong to the original authors.

<video controls width="864" height="480"> <source src="https://huggingface.

12 电商出图 Agent Skills for Claude Code / Codex / Cursor · GPT Image 2.5 (Flare/Sunburst) · Amazon 白底主图, A+, 场景图, 模特图, 小红书封面, 淘宝详情页

AlphaProtein Novo

**Choose the target model from the download links in this card or use the Winnow Quickstart.

本仓库包含一个基于 **Qwen/Qwen2.5-3B-Instruct** 的微调版本，使用 **LoRA（低秩适配）** 技术，在自定义的越狱数据集上进行训练。目标是用于实验性研究，特别是理解大语言模型的安全性和对齐行为。 ---

Using <a href="https://github.

Community Core ML runtime, Swift SDK, and iOS/macOS samples for Irodori TTS v4.1 Small MF.

<Gallery /> LoRAs for **Krea-2-Raw** that perform inpainting by masked editing: you provide an image with the region to (re)generate filled in black, plus a text prompt describing what should go...

Qwen image 2.1 Enhancement Nodes

**Code on GitHub** · **Demo** · **Benchmark: razvilka** · **Colab quickstart** · **vLLM server** **FRIDA-Decisions** makes structured decisions over Russian text in a single encoder pass: pick one...

Bulk-provision TokenMix accounts with disposable temp-mail.io inboxes and auto-generate API keys with random names. Turnstile-aware, Playwright-driven, JSON/CSV output.

Give the agent "this" and he gets a real PC.

Multi-chain MEME radar with deterministic evidence, ARC Mainnet monitoring, and isolated execution boundaries

MiniMax H3 (video + audio) on TensorFold 0.6.5 with an H3 family, Mac Studio M5 Ultra: 5 s 864x480 clip in ~50 s, 40 s at 768x448 — the fastest H3 we have measured on Apple Silicon; int8 tensor-unit kernels + Turbo adapter; one-shot + GHCR carrier

Research and writing can be automated. Authorship cannot. Evidence list + agent skills for spotting AI-produced papers no human stood behind.

Native red, blue, and purple security-agent harness for CYBER-FROST

Switchless RoCE collectives, NCCL profiles and fabric tooling for DGX Spark inference recipes

Portable Lean 4 autoformalization skills for Codex and Claude Code

Free Claude skill that audits, builds and improves other skills against Anthropic's skill-writing rules, including what changed for Opus 5.5, Sonnet 5.5 and Fable 5.

最大化你的 Codex 额度利用率！

The Arabic AI ecosystem as a map, a list, and a skill your agent can install.

Multi-market equity research workspace for US, Shanghai and Hong Kong stocks, featuring editable Word/PDF reports, transparent DCF valuation and DeepSeek-assisted assumption analysis.

Claude Code plugin: turn one still image into a ~5 s seamless looping animation (MP4, GIF, WebP, Live Photo). 把一张图做成约 5 秒的无缝循环动图。

偏航 · 让小机去旅行。Let your AI companion drift alone to a real corner of the world and come home with a travelogue they wrote and one thing they chose. Self-hosted MCP + web roadbook.

Reddit OSINT by username. Discover indexed posts, comments, deleted and live content, activity patterns, exposed identifiers, and an overall exposure score.

Machine Learning in Bend 2: Dependent Types, Shapes at the Type Level, and Proven Laws

Python 3.12+ 的 HeyGem 核心对口型引擎重实现。输入视频 + 语音,输出对口型视频。

Run with https://llama.

12 个电商短视频 Agent Skills：3秒钩子、卖点分镜、图生视频提示词、edge-tts 配音、字幕花字、ffmpeg 自动成片、抖音/TikTok Shop/视频号导出、直播切片、多语言配音、封面A/B · Claude Code / Codex / Cursor

Run with https://llama.app ```bash llama serve -hf ggml-org/Laya-GGUF ``` This is a decision model, to be used via `/v1/systemone` API. See https://github.com/ggml-org/llama.cpp/pull/29818

Agent skill for designing, mocking, building and auditing terminal UIs

High-speed parallel stream engine & local web dashboard for streaming HTTP/HTTPS downloads directly to PlayStation 5 FTP storage. Pure Python standard library, zero pip packages, macOS & Windows.

Ψηφιακή κάρτα εργασίας για το ΕΡΓΑΝΗ · Digital work card for Ergani

Google Calendar-style weekly view of your Claude Code sessions — local, live, read-only

DeepSeek-V4.1-Flash on 2x NVIDIA DGX Spark with the TensorFold engine: 1.8-1.9x code, 2.5x multi-stream, 1.7-1.95x prefill vs the vLLM kit, exact speculative decoding. Work in progress.

**ForgePlex-M2-9M** is a ~9.

Pre-submission integrity gate for LaTeX papers: every reference verified against real scholarly records. No LLM guessing.

A software-defined transceiver from one ESP32-S3 board: rtl_tcp receiver for SDR++ and FM/SSB voice transmitter for 13 cm (experimental)

QEMU harness for AI agents: boot aarch64/arm/riscv64/x86_64 kernels in seconds, exec via guest agent, fast rootfs rebuilds, IoT emulation

Build your own personal AI agent, one mission at a time. Mission 01: your AI learns who it's working for.

Native PC port of God of War (PS2) via static recompilation with PS2Recomp

**A first-party Blackfrost-AI mixed-precision NVFP4 model for security professionals conducting authorized research, assessment, engineering, and response work.**

Bộ skill dành cho thiết kế vi mạch

Trace changed experiment results to the numbers, claims and figures in your LaTeX paper. Offline bilingual review, safe numeric patches and read-only MCP tools.

Claude Code skill: make anything better and prove it before it ships — blind two-order judging, hard checks, reversible rollout. From the method behind harborseo.ai.

Skills and generators for agents converting Rails applications to compatible Rust applications.

Unofficial feature mod for the original 1010music Blackbox (patch only; bring your own stock 3.1.9 firmware). Not affiliated with 1010music.

Personal AIOps engineering project with durable workflows, scoped hybrid retrieval, reliable event delivery, and verified API/browser contracts.

我了个豆：本地拼豆出图、库存检查与报价 MVP。非商业用途源码公开，商业使用需另行授权。

The gap in the next 2 years (source - Anthropic Life Science Lead), will be the orchestration layer of taking experiments described in natural language and orchestrating them in the real world in a reproducible way.

Expose any AI agent worker as an OpenAI-compatible API — plug it into 9Router (or any OpenAI client) as a custom provider.

Campari Mirko - campari.mirko@gmail.com

Computational ledger for two room-temperature Luttinger-compensated magnet candidates, YBaMnFeO5 and KV[Cr(CN)6]: raw QE inputs/outputs, checker, re-runs

Paper (arXiv) · Project page This repository holds the **LoRA adapter** of a 4-step (4 NFE) student distilled from MiniMax-H3 with **Projected Distribution Matching Distillation (PDMD)**.

Run with https://llama.app ```bash llama serve -hf ggml-org/Clef-GGUF ``` This is a decision model, to be used via `/v1/systemone` API. Requires https://github.com/ggml-org/llama.cpp/pull/29831

Qwen3.

Website · Docs · Run tasks · Blog · GitHub

方寸 FANGCUN｜互动 HTML 演示制作 Skill。非商业源码公开，未经另行书面授权不得商用。含粒子开场、三条锦鲤动态背景及内容互动玩法。

Cross-platform amateur radio RF exposure (FCC MPE) calculator

数字联盟 (cn.shuzilm) 设备标识协议 DUID 逆向 —— 协议分析 / 抓取链路 / 锚点机制 / 可运行工具

Use command line to control your device via iPhone MIrroring.app

Real-time chessboard piece detection, tracking, and Stockfish analysis from a camera feed, using a fine-tuned YOLO11-seg model

A small, hand-constructed Spotlight transformer that runs Python: the examples from Can LLMs Grow Their Own Capabilities?.

[ACCV 2026] TRACE: Time-Adaptive Residual Attention Control with Content–Style Decomposition for Training-Free Diffusion Style Transfer

Fixed WD14 / PixAI tagger nodes for ComfyUI

personal pentesting assistant CLI tool

Portable offline AI assistant for SAP/ABAP learning, code drafting, debugging, and development planning. Runs locally with lightweight models, streaming responses, personal note retrieval, and optional GPU acceleration.

Small decision models (jev style) for latency critical scenarios

Claude. But less.

The Frob, An Apple II Atari 2600 Development System, re-made as an RP2354 based device.

One front end to search, browse, request and play everything on your self-hosted media server (Calibre, Audiobookshelf, Jellyfin, Sonarr, Radarr, Komga, RomM…)

Office laser-print + MFP-scan simulator for OCR test data (physical chain, no generative models)

Setting yourself up to use local models inside of Codex/ChatGPT

CCU (Claude Computer Use): let Claude Code use your Mac via ChatGPT's Computer Use, set up in one sentence. 让 Claude Code 动手用你的 Mac。

Unofficial GTFS-Realtime feed (VehiclePositions + TripUpdates) for Athens OASA buses and trolleys, built from the OASA telematics API and the official static GTFS

GLM-5.3 EXL3 on DGX Sparks with TensorFold

Faber-style trend-following asset allocation across 7 ETFs: a research backtest (2007–2026) with no look-ahead, transaction costs and robustness/regime analysis, plus a live Alpaca paper-trading bot and a Streamlit dashboard.

Use NetEase UU to access your Ubuntu desktop, with 4K presets, input fixes and a separate management window.

Local Humble Bundle key manager with Choice claiming and Steam library matching. Derived from gfargo/humble-bundle-keys (MIT).

Native SDR++ source module for ESP-SDR on the ESP32-S3: real IQ over USB, no bridge needed

Local ComfyUI workflow for watermark, mosaic and irregular overlay removal with Qwen Image 2.1 and SAM

git diff for CAD assemblies — two STEP files in, a picture of what changed plus a machine-readable change list out.

A simple WeGame laucncher for Linux.

Transforms UI workflow action sequences into structured video demo chapters, narration voiceover scripts, and timing metadata

Normalizes, routes, and thread-correlates multi-agent notifications and interactive prompts across team collaboration channels

Parses unified git diffs, detects security and performance anomalies, generates inline code reviews, and synthesizes unified patches

Universal Autonomous Agent Runtime Self-Healing & Diagnostics Engine. Detects configuration drift, corrupted execution state, and missing runtime resources, executing automated root-cause classification and snapshot rollback.

Universal Autonomous Agent Runtime Self-Healing & Diagnostics Engine. Detects configuration drift, corrupted execution state, and missing runtime resources, executing automated root-cause classification and snapshot rollback.

Software for a Raspberry Pi garden time-lapse camera: daylight capture every minute, automatic sync to your Mac, and one command to make a 4K video. Free and open source. Build guide coming soon.

CVE-2026-103956 - Loom for AWS - Critical - Auth bypass - unauthenticated super-admin when no IdP is configured

Lullian - structure-first engineering skill for software projects: requirements, architecture, and governed tool/MCP admission. Fail-closed by design.

Static recompilation of Rings of Power for Linux and Windows, with widescreen, smooth scrolling, external fonts, mouse controls, and saves.

Analyzes digital content for AI search engine citation visibility, factual definitions, and structured metrics

Translates high-level natural language phrases into deterministic SaaS API schema payloads

Decky plugin for the JSAUX PIXEL faceplate on the Steam Machine, no vendor software needed

Multi-Agent Speculative Execution & Early Cancellation Coordinator. Spawns speculative candidate execution paths concurrently, tracks milestone validation scores, and broadcasts early cancellation signals to inferior paths to slash inference overhead.

Autonomous Commerce Dynamic Bundling & Cross-Sell Recommender (inspired by Meta Muse, Shopify & Instacart Smart Carts). Analyzes cart items and intent, computes product complementary affinity scores, respects merchant margin constraints, and synthesizes personalized discount bundles.

Analyzes digital content for AI search engine citation visibility, factual definitions, and structured metrics

Autonomous Commerce Dynamic Bundling & Cross-Sell Recommender (inspired by Meta Muse, Shopify & Instacart Smart Carts). Analyzes cart items and intent, computes product complementary affinity scores, respects merchant margin constraints, and synthesizes personalized discount bundles.

Curated, link-only index of community game ports that run natively on Apple Vision Pro (visionOS). Bring your own game files.

Dynamic Few-Shot Agentic Prompt Compression & Token Sieve. Compresses multi-turn conversation history, strips verbose boilerplate and filler tokens, retains critical tool parameters and named entities, slashing inference costs by 40-60% deterministically.

Contains a payload and all the necessary files for a Raspberry-Pi Pico W to use keystroke injection to open up any LinkedIn URL, or any URL!

Web UI to design, test and export AI agents with a provider-pluggable Python harness ai-agents, agent-harness, agent-builder, python, fastapi, llm, anthropic, openai, ollama

Manages MCP tool execution rate limits and compresses bloated tool response context payloads

Observes telemetry errors, maps root causes, and generates safe executable runbooks for production incident mitigation

Stores agent episodic memories with semantic tag indexes and time-decay relevance calculation

Modular 14-Agent Production Pipeline for Autonomous Open Cinema

Conducts conversational qualification, dynamically tailoring step-by-step product walkthrough paths based on prospect pain points

Observes telemetry errors, maps root causes, and generates safe executable runbooks for production incident mitigation

Translates high-level natural language phrases into deterministic SaaS API schema payloads

Enterprise Cross-System Identity & Token Exchange Bridge (inspired by WeChat Work, Tencent Cloud OAuth2 & WorkBuddy Delegated Execution). Safely manages on-behalf-of token exchange, scoped permission delegation, token lifetime refresh, and zero-leakage ephemeral credentials.

Dynamic Agent Persona & Contextual Communication Style Adapter (inspired by Personal AI Copilots). Adapts communication formality, brevity, tone, and technical depth based on channel context, user cognitive load, and sentiment feedback.

Multi-Modal Visual Diagram-to-Code Architecture Synthesizer. Translates geometric bounding boxes, flowchart node connections, and architectural sketches into clean Mermaid graphs, JSON topology matrices, and boilerplate code scaffolds.

Enterprise Cross-System Identity & Token Exchange Bridge (inspired by WeChat Work, Tencent Cloud OAuth2 & WorkBuddy Delegated Execution). Safely manages on-behalf-of token exchange, scoped permission delegation, token lifetime refresh, and zero-leakage ephemeral credentials.

Conducts conversational qualification, dynamically tailoring step-by-step product walkthrough paths based on prospect pain points

GA100 (CMP 170HX) fuse architecture reverse-engineering: FF-pool decode with verified anchors, RIR macro verification, corrected burn lists, 96GB append recipe, raw dumps

面向 Codex 的本地 PDF 转 Markdown 技能，适用于扫描书、学术论文与文献引用整理，保留公式图表和原页定位。

Universal Agentic Webhook Signature Verifier & Ingress Multiplexer. Authenticates multi-provider webhooks (Shopify, Stripe, GitHub, Slack, PayPal) with HMAC-SHA256 constant-time verification, timestamp replay window validation, and multiplexed agent event dispatch.

Universal Agentic Webhook Signature Verifier & Ingress Multiplexer. Authenticates multi-provider webhooks (Shopify, Stripe, GitHub, Slack, PayPal) with HMAC-SHA256 constant-time verification, timestamp replay window validation, and multiplexed agent event dispatch.

Parses unified git diffs, detects security and performance anomalies, generates inline code reviews, and synthesizes unified patches

Multi-Modal Visual Diagram-to-Code Architecture Synthesizer. Translates geometric bounding boxes, flowchart node connections, and architectural sketches into clean Mermaid graphs, JSON topology matrices, and boilerplate code scaffolds.

Manages MCP tool execution rate limits and compresses bloated tool response context payloads

Dynamic Few-Shot Agentic Prompt Compression & Token Sieve. Compresses multi-turn conversation history, strips verbose boilerplate and filler tokens, retains critical tool parameters and named entities, slashing inference costs by 40-60% deterministically.

Normalizes, routes, and thread-correlates multi-agent notifications and interactive prompts across team collaboration channels

Autonomous Agentic Multi-Turn Negotiator & Counter-Offer Strategist (inspired by Consumer Commerce & Contract Negotiations). Calculates Zone of Possible Agreement (ZOPA), concession decay curves, reservation price thresholds, and structured bargaining counter-offers.

受控中文技术写作技能：ASD-STE100 的中文对应物，含 stdlib-only 确定性 linter。Controlled Chinese technical writing skill for agents.

Autonomous Agentic Multi-Turn Negotiator & Counter-Offer Strategist (inspired by Consumer Commerce & Contract Negotiations). Calculates Zone of Possible Agreement (ZOPA), concession decay curves, reservation price thresholds, and structured bargaining counter-offers.

Dynamic Agent Persona & Contextual Communication Style Adapter (inspired by Personal AI Copilots). Adapts communication formality, brevity, tone, and technical depth based on channel context, user cognitive load, and sentiment feedback.

Typed Python bindings for fframes video rendering

Stores agent episodic memories with semantic tag indexes and time-decay relevance calculation

Sketch Parallax · 手绘视界 — 将人物、动物与物件图片制作成透明彩铅手绘分层 HTML，支持电脑端 3D 混合穿梭与逐页独立手稿。Turn images into interactive pencil-sketch 3D spaces, with shared fly-throughs and independent sketch pages.

Scores outbound lead profiles, evaluates purchase intent, and synthesizes personalized multi-channel outreach messages

Transforms UI workflow action sequences into structured video demo chapters, narration voiceover scripts, and timing metadata

Scores outbound lead profiles, evaluates purchase intent, and synthesizes personalized multi-channel outreach messages

a desktop config for niri & noctaliaV5

Multi-Agent Speculative Execution & Early Cancellation Coordinator. Spawns speculative candidate execution paths concurrently, tracks milestone validation scores, and broadcasts early cancellation signals to inferior paths to slash inference overhead.

Enterprise Multi-Speaker Meeting Action Item Extractor & WeChat Work Task Dispatcher. Analyzes conversational meeting transcripts from Tencent Meeting, Zoom, and Teams, extracts explicit commitments, dates, and assignees, performs Eisenhower urgency-importance matrix prioritization, and formats collaborative task cards.

Autonomous Multi-Channel Inventory Restock Predictive Balancer & PO Generator. Forecasts SKU sales velocity using exponential moving averages (EMA), calculates safety stock buffers ($SS = z \cdot \sigma_d \sqrt{L}$), dynamically computes reorder points (ROP), and synthesizes supplier Purchase Orders.

Autonomous Shipment Exception Detection & Delivery Rescheduler. Ingests FedEx, UPS, DHL, and SF Express tracking webhooks, detects failed delivery attempts and customs delays, auto-negotiates delivery windows, locker hold orders, and neighbor safe-drop instructions.

Autonomous Post-Purchase Price Drop Monitor & Retailer Refund Claim Arbiter. Tracks online purchases across Shopify, Walmart, Best Buy, and Target, detects price drops within the merchant price-match guarantee window, verifies eligibility conditions, and synthesizes automated refund claims.

Enterprise Multi-Modal Contract Clause Risk Auditor & Redline Synthesizer. Scans enterprise MSAs, NDAs, and Vendor SOWs in Tencent Docs, detects high-risk clauses (unlimited indemnity, one-sided IP assignment, perpetual non-compete, evergreen auto-renewal traps), calculates composite risk scores, and generates precise legal redlines.

Autonomous Multi-Channel Inventory Restock Predictive Balancer & PO Generator. Forecasts SKU sales velocity using exponential moving averages (EMA), calculates safety stock buffers ($SS = z \cdot \sigma_d \sqrt{L}$), dynamically computes reorder points (ROP), and synthesizes supplier Purchase Orders.

Enterprise Multi-Modal Contract Clause Risk Auditor & Redline Synthesizer. Scans enterprise MSAs, NDAs, and Vendor SOWs in Tencent Docs, detects high-risk clauses (unlimited indemnity, one-sided IP assignment, perpetual non-compete, evergreen auto-renewal traps), calculates composite risk scores, and generates precise legal redlines.

Autonomous Post-Purchase Price Drop Monitor & Retailer Refund Claim Arbiter. Tracks online purchases across Shopify, Walmart, Best Buy, and Target, detects price drops within the merchant price-match guarantee window, verifies eligibility conditions, and synthesizes automated refund claims.

Autonomous Shipment Exception Detection & Delivery Rescheduler. Ingests FedEx, UPS, DHL, and SF Express tracking webhooks, detects failed delivery attempts and customs delays, auto-negotiates delivery windows, locker hold orders, and neighbor safe-drop instructions.

Enterprise Multi-Speaker Meeting Action Item Extractor & WeChat Work Task Dispatcher. Analyzes conversational meeting transcripts from Tencent Meeting, Zoom, and Teams, extracts explicit commitments, dates, and assignees, performs Eisenhower urgency-importance matrix prioritization, and formats collaborative task cards.

GGUF quantizations of Aleph-Alpha/Kolibri-1, a 78B-parameter Mixture-of-Experts reasoning model (3.

Desktop app for Windows, macOS and Linux that reads text from images and PDFs on your own computer. Pick an OCR model (PaddleOCR, Tesseract, EasyOCR, Ollama) and DocBox sets it up in one click.

Source-cited Obsidian knowledge base about Claude Code mods: how they work, how to build them, and how to check one before you install it.

4-bit (**AWQ W4A16, routed-experts-only**) quantization of **MiniMax-M3**, packaged to serve on **NVIDIA A100 (SM80)** under vLLM — text + vision, 1M-context, at TP-8 (or 2× TP-4).

> **Run with Rapid-MLX:** `rapid-mlx serve ling-3.

## SUB-TITLE: Turbo-Brilliance System V1.0 ---

Five ICE tiers of **huihui-ai's abliterated Ornith-1.

Repackaged and uncensored models files for **ComfyUI** use.

Decisions like routing a ticket, picking a tool or approving an action have a fixed set of answers.

Paper (arXiv) · Project page Full transformer weights (`MiniMaxH3Transformer3DModel`, bf16) of a 4-step (4 NFE) student distilled from MiniMax-H3 with **Projected Distribution Matching Distillation...

Paper (arXiv) · Project page This repository holds the **LoRA adapter** of a 2-step (2 NFE) student distilled from MiniMax-H3 with **Projected Distribution Matching Distillation (PDMD)**.

Multi-task speech classifier based on the **Whisper-base encoder** that predicts a speaker's **gender** and **language** from a short audio clip.

Fine-tuned YOLO11m object detector on the **VisDrone-DET** benchmark dataset, trained and evaluated as part of DetectionBench -- a framework for reproducibly benchmarking modern object detectors with...

Fine-tuned YOLO26m object detector on the **VisDrone-DET** benchmark dataset, trained and evaluated as part of DetectionBench -- a framework for reproducibly benchmarking modern object detectors with...

Persian-IPA-to-Speech-F5 is a Persian speech synthesis checkpoint fine-tuned from F5-TTS v1 Base. It is conditioned directly on raw Unicode IPA sequences rather than ordinary Persian orthography.

An **all-NVFP4 abliterated** artifact of ukisai/Swift-1.

- **Website:** https://tokforge.

FP32 ONNX export of **finegrain/finegrain-box-segmenter** (an MVANet high-resolution background remover), ready to run in the browser, Node.

<Gallery />

**A first-party Blackfrost-AI FP8 deployment artifact for security professionals conducting authorized research, assessment, engineering, and response work.**

Abliterated DeepSeek v4.

A high-throughput, structurally distilled causal language model engineered for local deployment, low-latency execution, and reasoning density.

<!-- ### quantize_version: 2 --> <!-- ### output_tensor_quantised: 1 --> <!-- ### convert_type: hf --> <!-- ### vocab_type: --> <!-- ### tags: --> <!-- ### quants: x-f16 Q4_K_S Q2_K Q8_0 Q6_K...

**Original model by Naive AI.

Ok, so the model cards never get to the point.

<!-- ### quantize_version: 2 --> <!-- ### output_tensor_quantised: 1 --> <!-- ### convert_type: hf --> <!-- ### vocab_type: --> <!-- ### tags: nicoboss --> <!-- ### quants: Q2_K IQ3_M Q4_K_S...

<!-- ### quantize_version: 2 --> <!-- ### output_tensor_quantised: 1 --> <!-- ### convert_type: hf --> <!-- ### vocab_type: --> <!-- ### tags: nicoboss --> <!-- ### quants: Q2_K IQ3_M Q4_K_S...

> 📖 **Research White Paper:** **📄 Read Whitepaper (PDF)** &nbsp;|&nbsp; **🌐 View Online** &nbsp;|&nbsp; **📝 Full MI300X Serving Blog** High-efficiency, hardware-benchmarked GGUF releases of...

**Qwen3.

## Abstract Mephisto-1.

**Fast local decisions over dynamic candidate sets.

> [!Note] > 本仓库包含以 Hugging Face Transformers 格式提供的后训练模型的权重和配置文件。 > > 这些产物兼容 Hugging Face Transformers、vLLM、SGLang、KTransformers 等框架。 > > 如您需要在Ollama及llama.cpp上使用，请使用gguf量化版本

![Open In Colab](https://huggingface.

git clone https://github.

> **This repository is now considered a legacy experimental release and is no longer recommended for new downloads.

!Shingi — 審議 Shingi 27B is a local decision model.

!AEON Qwen — Supreme Being of the Digital Cosmos A **DFlash speculative-decoding drafter** for **`Qwen/Qwen3.

A **W4A4 + paired-4:8 sparse** compressed checkpoint of `Qwen/Qwen3-30B-A3B`, produced with **MoESQ**.

GGUF builds of OpenDecider-small-td, the 4B OpenDecider decision model, for **LM Studio**, **Ollama** and anything else built on llama.

**Research project for inference of large dense models on the M6 Apple Neural Engine.

A **100M-parameter** chat/instruct model that **beats SupraLabs/Supra2-100M-Instruct on 6/7 benchmarks** on the official EleutherAI LM-Eval Harness — *and* actually converses.

<!-- ### quantize_version: 2 --> <!-- ### output_tensor_quantised: 1 --> <!-- ### convert_type: hf --> <!-- ### vocab_type: --> <!-- ### tags: --> <!-- ### quants: x-f16 Q4_K_S Q2_K Q8_0 Q6_K...

> 🚧 **Status: In Development** > This model is actively evolving. Architecture, dataset and behavior may change in future versions. ---

English | 中文 Qwen/Qwen3.

!Reginutė — Lietuviškai.

<div align="center"> <img src="banner.png"> </div> ⚠️ Это учебный/исследовательский проект, а не модель для продакшена. Подробнее об ограничениях см. раздел «Качество».

One Decision model for the request-time signals of vLLM Semantic Router: subject area, output modality and user feedback as Choice questions, and prompt attack, harmful request, twelve hazard...

**Open decision models that run on a laptop CPU, built by senior AI engineer Mert Kaya.

**The best open-source Turkish decision model in its class, built by senior AI engineer Mert Kaya.

All downstream numbers below come from fine-tuning this model (`albert-base-v2-sq`) end to end on each task and, for context, fine-tuning mBERT (`bert-base-multilingual-cased`) and XLM-R-base...

> **A ~30M parameter language model trained from scratch to translate Gen Z slang into clear, standard English.**

This is a locally assembled and tested quantization of `orcarouter/Qwen3.

Katagurun virallinen tuotantotason **GGUF-julkaisu** laivastolippulaivasta Kataguru Titan-v3.

STAR-XPC predicts how much of an X-ray beam passes through a polymer composite shield.

**A rank-16 Uno drafter for Gemma 4 26B A4B, the model BroadNet runs in production in English and Arabic.

Self-trained **DFlash2 block-diffusion speculative-decoding drafter** for GLM-5.

<!-- ### quantize_version: 2 --> <!-- ### output_tensor_quantised: 1 --> <!-- ### convert_type: hf --> <!-- ### vocab_type: --> <!-- ### tags: --> <!-- ### quants: x-f16 Q4_K_S Q2_K Q8_0 Q6_K...

Measured on the abliterated parent (GGUF quants inherit this behavior): | Metric | Result | |---|--:| | True refusal (harmful, n=300) | **0 / 300 = 0.

**29 control vectors for Gemma 4 31B, each an attempt to reproduce a released fine-tune.

Browser-compatible ONNX model assets used by Timeline Studio, an MIT-licensed, local-first AI video editor.

Use Linux, Python 3.

**Qwen3.

> !NOTE] > This is the conservative BF16-`lm_head` reference and rollback checkpoint.

JobBERT model fine-tuned on the Silver training/development sets (JobBERT) and evaluated on the 150-sentence human reference set. **Project · Data · Reproduction guide**

Ekko v1 Tiny is a compact, 110M-parameter Danish speech recognition model with word timestamps.

- **Website:** https://tokforge.

DecisionTune 1.

**Jev-like decision models that can also see images.

LoRA adapter finetune tu `Qwen/Qwen1.5-MoE-A2.7B` theo baseline **MidAlign**, Alternate Training: step chan = task step (LM loss), step le = contrastive align step.

> !WARNING] > This is an experimental text-serving checkpoint for Ascend 310P.

### *Ultra-Lightweight Multilingual Speech Recognition for Edge, Mobile & CPU Deployments*

*Published by PythiaFinance.

`jaredpalmer/kev-0.

* Base Model: Qwen-0.5B * Training Type: LoRA fine-tuning * Domain: Finance * Language: English * Training Precision: FP32

> **Run with Rapid-MLX:** `rapid-mlx serve qwen3.

> **Used by Rapid-MLX:** sidecar for the experimental DSpark mode of `rapid-mlx serve deepseek-v41-flash-reap-2bit`; not a standalone model.

ReDimNet2-B6 produces local speaker embeddings for comparing clean voice samples. It does not diarize audio or assign names by itself.

**MarkOne** is a high-precision, zero-shot NLI cross-encoder model designed for **dynamic intent routing and query classification** within modular AI architectures .

A **decision model**: one text (the *state*) and typed questions in (`Choice`, `Noul` yes/no, `Score`), a probability per option out, in one forward pass.

<style> .

**Model Name:** `Helcyon-Solara-2-14B-GGUF` **Version:** Series 7 **Owner:** HardWire **Base:** Ministral 3 14B **Quantized GGUFs:** IQ4_XS, Q4_K_M, Q5_K_M, Q6_K, Q8_0, f16 **Tags:**...

GGUF files of Ternary-Bonsai-2-27B-DFlash2-ft5, a DFlash 2 drafter fine-tuned against PrismML's Ternary-Bonsai-2-27B, for llama.

**Structured decisions from text and images.

**Structured decisions from text and images.