Latest hashcat docker for CUDA, OpenCL, and POCL. Deployed on Vast.ai

An Agent Skill for the DL experiment lifecycle: RUN (a GPU you own or rent) → VERIFY the number is real → DELIVER reproducible, single-source figures and tables.

Stable Diffusion WebUI Forge docker images for use in GPU cloud and local environments. Includes AI-Dock base for authentication and improved user experience.

Kohya's GUI docker images for use in GPU cloud and local environments. Includes AI-Dock base for authentication and improved user experience.

A collection of Stable Diffusion Templates that will allow you to use best UIs with the best checkpoints and significantly reduce the time needed to get started.

Terraform setup for deploying a private coding LLM on Vast.ai with vLLM, Qwen3 Coder, and OpenCode.

Provably fair Asteroids on Stellar: deterministic gameplay tapes are verified with RISC Zero proofs before KALIEN rewards are minted on-chain.

Real-time TUI for monitoring cloud GPU training instances

A TUI dashboard for renting vast.ai GPUs, running llama.cpp models, and wiring them into your coding agent (opencode).

This package provides a Python client for the Vast.ai GPU rent service, that allows to list, create, destroy GPU instances programmaticaly, using Python.

MLE kit MCP server: Tools for working with files and code locally and with a remote GPU

🛠️ Useful Scripts/Tools to use ComfyUI as a dev. metadata cleaners, workflow analyzers, cloud deployment optimizations

This is a unofficial java api for Vast.ai

Desktop GUI app for automating deep learning training on Vast.ai cloud GPUs — classification & regression with one click

Build a template on the fly for deployment to Vast.ai

A Claude Code skill to rent, watch, and tear down Vast.ai GPU instances safely — one CLI for search/up/status/dashboard/run/ssh/sync/down, with auto-destroy deadlines and a watchdog so nothing bills silently.

Design study and reference architecture for an autonomous control loop managing GPU compute on Vast.ai — boundary-first, fail-closed, absence-tolerant. Research artifact, not a production system.

Nix flake — Vast.ai GPU-template provisioning toolkit for macOS

Personal GPU-rental harness — provision a vLLM pod on Vast.ai or RunPod, expose it over Tailscale, terminate when the daily budget is spent.

GPU inference in one command. Auto-picks a cheap Vast.ai GPU, loads any GGUF model, gives you an OpenAI-compatible endpoint.