run a 27b local ai on your gaming gpu, rtx 3060 and up, radeon rx 6600 and up in beta, 8gb or more, linux and windows, one line to install
-
Updated
Oct 16, 2026 - Python
run a 27b local ai on your gaming gpu, rtx 3060 and up, radeon rx 6600 and up in beta, 8gb or more, linux and windows, one line to install
Q3_PLE quantization and learned MTP research for Qwen3.8-Flash-Next on an RTX 5070 12GB.
CUDA stress tool that checks whether your PSU can handle NVIDIA GPU power transients. It runs sustained loads, 1-200 Hz square waves and bursts from idle (optionally with a full CPU load), logs power, clocks and temperature through NVML to CSV, and plots the results. Windows and Linux.
Pre-built onnxruntime-gpu 1.24.1 with Blackwell sm_120 CUDA kernels (RTX 5090/5080/5070)
To associate your repository with the rtx-5070 topic, visit your repo's landing page and select "manage topics."