Directory
Published directory entries matching your search.
GreenSSH offers premium SSH and VPN accounts hosted on what it describes as eco-friendly infrastructure. It lists SSH tunneling, SSH over SSL/TLS, SlowDNS, WebSocket, UDP, OpenVPN and ShadowSocks options, and states that accounts are valid for 7 days with high-speed performance. A feature request form is also available.
HandBrake is an open-source, GPL-licensed, multiplatform, multithreaded video transcoder built by volunteers. It converts video from nearly any format to widely supported codecs, with built-in device presets and options for basic and advanced tuning. It runs on Windows, Mac and Linux and is free.
An interactive sound project that lets users play any combination of home-related sounds and tunes. Users first select a tune, then choose sounds and sound combinations such as outside, inside, pets, family, partner or roommate, and neighbors in an apartment or house. Sound design and web development are credited to Hansdale Hsu.
Hong Kong Transit 3D is an independent project presenting a real-time 3D schedule simulation of Hong Kong's trains, trams, buses, and ferries over real terrain. Vehicles move according to published frequencies and the Hong Kong clock, not GPS tracking. The map includes depth-true tunnels, a per-line underground cross-section view, and trilingual support.
Materials for the Hugging Face Diffusion Models Course, a free course on diffusion models. It covers theory, generating images and audio with Diffusers, training and fine-tuning models, and conditional generation. Prerequisites include Python skills and basics in deep learning and PyTorch.
Hyperopt is an open-source Python library available on GitHub, designed to provide distributed asynchronous hyperparameter optimization functionality for tuning machine learning models in development workflows, production environments, as well as research settings.
Hyprland is a dynamic tiling Wayland compositor. It provides Wayland features, dynamic tiling, animations, plugins and multiple fine-tuneable layouts. It offers live reloading configuration and socket-based IPC. Installation is available for Arch, NixOS, FreeBSD and openSUSE.
Hysteria is an open-source, MIT-licensed proxy built on a customized QUIC protocol, designed for performance over unreliable and lossy networks. It supports modes including SOCKS5, HTTP Proxy, TCP/UDP Forwarding, Linux TProxy and TUN, and the protocol masquerades as standard HTTP/3 traffic. Builds exist for major platforms and architectures.
IBM Watson is enterprise AI technology from IBM. IBM Research began work in 2007 on a computer system to compete at Jeopardy!, and in 2011 the Watson system defeated two top-ranked players. Watson technology became available as a cloud development platform in 2013. In 2023 IBM announced the watsonx AI portfolio for training, tuning and distributing models.
Kiln AI is a platform and open-source library for building, evaluating, and optimizing AI systems. It offers tools for working with retrieval-augmented generation, AI agents, synthetic data, model fine-tuning, and prompt optimization, with a free downloadable application available for macOS, Windows, and Linux.
Kubeflow is described as the cloud native AI platform, composed of modular open source projects forming a Kubernetes-native stack for data and AI workloads. Its subprojects include Kubeflow Pipelines for building and deploying portable machine learning workflows, Kubeflow Trainer for distributed AI training and LLM fine-tuning, Kubeflow Katib for automated machine learning, Kubeflow Notebooks for interactive development environments, and Kubeflow Hub for indexing models and metadata. It is a Cloud Native Computing Foundation project.
Lambda offers hourly rental of NVIDIA H100, H200, B200, and A100 GPUs through a self-serve interface, with instances launchable via UI, API, or CLI. Users can train, fine-tune, and serve models on one to eight GPU instances and pay by the minute, with no egress fees.
LightlyTrain is an open-source GitHub repository offering all-in-one training capabilities for vision models, including YOLO, ViTs, RT-DETR, and DINOv3. It supports pretraining, fine-tuning, and distillation workflows for these model types.
Liquorix is an enthusiast Linux kernel optimized for low latency. It provides tuning for responsiveness at the cost of throughput and power usage, a PDS process scheduler for gaming and multimedia, and a 1000hz tick rate. Binary builds are produced for Debian, Ubuntu, Fedora and Arch, installed via an install script.
LLama Cpp Agent is a framework designed for easy interaction with large language models (LLMs). It allows users to chat with LLMs, execute structured function calls, and retrieve structured output. It works with models not fine-tuned for JSON output or function calls. The project is hosted as a public repository on GitHub.
UI tool for fine-tuning and testing your own LoRA models base on LLaMA, GPT-J and more. One-click run on Google Colab. + A Gradio ChatGPT-like Chat UI to demonstrate your language models.
No-code batch compute platform for LLM evaluation and tuning workloads.
A directory of abstracts and links to academic papers focused on local language models, current as of the end of March 2026. The listing compiles research published between 2017 and 2023, including foundational works such as Attention Is All You Need, BERT, T5, and PaLM, alongside studies on efficient transformer inference, multimodal models, and instruction tuning.
Lockdown Privacy offers a firewall and a Secure Tunnel VPN for blocking trackers, ads and badware in apps. The firewall is described as open source and on-device. The site states the products are Openly Operated and provide transparency into data handling. The listing covers iPhone, iPad and Mac.
Lora is a repository applying low-rank adaptation to fine-tune diffusion models. It offers fine-tuning that produces small end results of 1MB to 6MB, compatibility with diffusers, support for inpainting, merging checkpoints, and CLIP, Unet and token fine-tuning pipelines.
Ludwig is a low-code framework for building custom large language models, neural networks, and other AI models. The project's repository includes introductory documentation, release notes for version 0.16, installation guides, quick start materials, and resources for instruction fine-tuning of LLMs.
Chinese LLM, Based on LLaMA and fine tune by Stanford Alpaca, Alpaca LoRA, Japanese-Alpaca-LoRA.
Mamba-Chat is a chat language model built on a state-space model architecture rather than a transformer. The repository provides training and fine-tuning code based on a modified Huggingface Trainer class. The model is based on Mamba-2.8B and was fine-tuned on 16,000 samples of the HuggingFaceH4/ultrachat_200k dataset. Code is provided for testing and fine-tuning, including a command-line chatbot and a Gradio app.
mljar-supervised is an Automated Machine Learning Python package for tabular data. It automates data preprocessing, model construction and hyperparameter tuning, and produces a Markdown report for each model. It offers Explain, Perform and Compete modes, along with model explanations and generated web apps for predictions.
