Python script, which empowers people with no programming background to generate robust leads on a mass scale. This repo will be compiled of various versatile techniques used in lead generation.
An Asynchronous Reinforcement Learning Engine for Omni-Modal Post-Training at Scale
Toolkit for large-scale whole-slide image processing.
TiDB database documentation. TiDB is an open-source, cloud-native, distributed, MySQL-Compatible database for elastic scale and real-time analytics.
framework for large-scale SAR satellite data processing
Open Data Cube analyses continental scale Earth Observation data through time
A research-oriented framework for building optimization-driven decision intelligence systems, integrating causal inference, risk modeling, and AI pipelines for large-scale socio-technical systems
Pax is a Jax-based machine learning framework for training large scale models. Pax allows for advanced and fully configurable experimentation and parallelization, and has demonstrated industry leading model flop utilization rates.
Quickly search, compare, and analyze genomic and metagenomic data sets.
SWE-Bench Pro: Can AI Agents Solve Long-Horizon Software Engineering Tasks?
FlagScale is a large model toolkit based on open-sourced projects.
Describe Anything, Anywhere, at Any Moment (DAAAM), a novel approach to real-time, large-scale, spatio-temporal memory
Infrahub is a graph-based data management platform with built-in version control, CI workflows, peer review, and API access. It’s purpose-built to power reliable infrastructure automation at scale.
TrueNAS SCALE Build System
🐉 Loong: Synthesize Long CoTs at Scale through Verifiers.
The ai agent workflow tool that scales from small, simple projects to large, complex codebases.
A 1:1 scale streamed Earth for BeamNG.drive: real elevation and satellite imagery, floating origin, LOD streaming from a companion server
Creating realistic RL environments requires collaboration between researchers, engineers, and domain experts across many dimensions: artifacts, environments tools, dynamism of the environment, reproducibility, and more. There is no open source framework for building these environments effectively. Until now.
[MobiSys 2026] A large-scale, multimodal dataset and benchmark for Human Action Recognition, Understanding and Reasoning
Let your agents talk: messages, data, files and wake-ups over Tailscale.
MiniMaxH3_LatentUpscaler for ComfyUI
Local-first, auditable multi-agent control room for coding, review, evidence, and private remote work.
Upscale long, high-resolution MiniMax H3 video on a VRAM-limited GPU in a single node.
Self Gradient Forcing (SGF) recovers the missing context-gradient path for self-generated causal memory through a bounded two-pass replay, enabling models trained with only a 5-second window to extrapolate to minute-scale videos with stronger identity, layout, and temporal stability.
A general-purpose engine for large-scale, repeated information gathering by a fleet of workers.
A large-scale synchronized vision–touch–audio dataset of robotic sliding contact
Designed for managing vulnerability discovery at scale with agentic AI and integrates with HackerOne.
Decky Loader plugin for Steam Deck and other SteamOS/Bazzite handhelds: download, build and install HelixSR (a DLSS-based FSR 3.1 replacement) per game, directly or through OptiScaler, and check that it's working.
UrbanGround: From Local Perception to Spatial Agency in a Real-Scale City
Memory-efficient CTC loss with exact loss and first-order gradient equivalence under vocabulary reduction, and activation memory that no longer scales linearly with vocabulary size.
[NeurIPS 26] Official implementation of the paper "Urbex: Agentic Spatial Grounding for City-Scale 3D Scenes"
Billion-scale embedded vector database built entirely on Parquet and Arrow.
OceanScale — GPU-native ocean simulator for underwater robotics (source alpha)
Physics-guided augmentation and multi-scale Transformer learning for low-resource spectroscopy; CNMM-MSST XRF implementation optimized based on the JAAS paper.
:dog: Beagle is designed to be the unified platform for the broader Recursive Self-Improvement (RSI) ecosystem. Think of it as Hugging Face Trainer for agent evolution. Its efficient rollout engine enables evaluation and evolution at scale across different agents, datasets, benchmarks, and evolution methods.
RSI Bench is an ongoing effort to evaluate whether AI agents can develop the capabilities required to advance AI R&D.
GPU-native NVIDIA video enhancement, restoration, temporal analysis, and experimental DLSS 5 Neural Rendering nodes for ComfyUI.
[arXiv'26] AMB3R-SLAM: Kilometer-scale SLAM with Hierarchical Backend
🏠 家庭 DNS 广告拦截(红果短剧 / 番茄小说等)—— 在家、出门(Tailscale)都能用。85 条精确规则 + 928 条上游社区规则,零误伤。Home DNS ad-blocker for Chinese apps.
CVE-2026-8452 PreAuth RCE
Neighborhood-scale surface heat research pipeline and interactive Los Angeles atlas
Self-hosted live dashboard for the Xiaomi Mi Body Composition Scale 2 over Bluetooth Low Energy: profiles, body composition, history
Official Paper Implementation of ScaleResfusion: Residual Rectified Flow Based on Residual Vector Field
MiniMax-H3 on single DGX Spark: keys Sol-Engine ports + Heretic TE + Spectrum/Sage speed + 0.5MP upscaler finish. Formal bench included.
Automated data preparation for machine learning. Load a dataset, clean, transform, encode, scale and split it, explore it with built-in analysis, and export model-ready data with a full report. Built with Python and Streamlit.
LAtent super-Resolution high-Performance image upscaler
Upscale a finished MiniMax H3 clip in ComfyUI: H3 latent upscaler + one guided re-draw in tiles, with the h3upscale x2 LoRA (on Hugging Face).
A living guide of the systems that train, serve, and scale LLMs
Tideline-Memory is a long-term memory system built for AI Agents. Most agent memory: You ask, it finds. Tideline: The agent wakes up already knowing who he is, not querying "who am I?" every session. Achieving accurate memory hits while also preventing memory from expanding at scale.No compression, no forgetting.
Decky Loader plugin to install, configure and monitor OptiScaler per game from the Game Mode on SteamOS/Bazzite/Etc. With live in-game frame generation and upscaler control.
Removes the artifacts that GPT Image 2 leaves in its output — excessive sharpening, random bright specks, unnatural edges, and scale-like texture patterns — while leaving the image itself intact.
A Large-Scale High-Quality Dataset for Compositional Instruction-Guided Video Editing
A mobile control plane for persistent AI agents over Tailscale, Mosh, Herdr and ntfy.
Local OCR, metadata automation and RAG document chat for Paperless-ngx using PaddleOCR and Ollama, designed for CPU-only self-hosted servers
An open-source tool for running controlled behavioral experiments on AI systems at scale.
—
An all-in-one director console for MiniMax H3 long-form video in ComfyUI. It chains shots into one continuous video by anchoring each previous segment's tail latent to the next segment's head, adds a latent-upscale second pass built on the minimax_h3_latent_upscaler model, and ships experimental improvements researched on top of both.
Learned 2x latent upscaling for Qwen Image 2.1 in ComfyUI.
Awomo-0.5: World Action Model Pre-trained at Scale
The Upscaling Workflow and Nodeset of your Dreams