AIC++★ 298⑂ 41 forksApache-2.0
Wallawalla47/Infernix
Infernix GPU Inference
C++/CUDA single-GPU inference engine for Qwen3.8 models, including Qwen3.8-Flash-Next with offloaded experts, optimized for RTX 5090; derived from NInfer.
// KEY FEATURES
- Loads Qwen3.8 and Qwen3.8‑Flash‑Next models, offloading experts to GPU memory.
- Executes single‑GPU inference on RTX 5090 using custom CUDA kernels.
- Delivers low‑latency, high‑throughput chat completions for Qwen3.8‑style prompts.
New repositories every 30 minutes
REDDYX AI scans GitHub 24/7 and ships the best AI/ML/Web3 projects to Telegram.
Join on Telegram