Fp4

A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit and 4-bit floating point (FP8 and FP4) precision on Hopper, Ada and Blackwel…

Model compression toolkit engineered for enhanced usability, comprehensiveness, and efficiency.

Fully uncensored, capability-enhanced abliteration of Qwen3.6-27B. NVFP4 + z-lab DFlash speculative decoding (n=12) on the unified ghcr.io/aeon-7/aeon-vllm-ulti…

From-scratch C++/CUDA inference engine for the NVIDIA RTX 5090 (sm_120a) — the best single-GPU backend for agentic AI: tool calling, long-context loops, reasoni…