Upload README.md with huggingface_hub

README.md

@@ -1,57 +1,31 @@
-# SageAttention 2.2.0 — Pre-built Wheel
+# K2Horizon llama.cpp CUDA Binary
 
-Pre-compiled SageAttention 2.2.0 wheel for **Linux x86_64** with CUDA 13 and C++20.
+Pre-built llama.cpp binary with CUDA support for Linux x86_64.
 
-## Compatibility
+## Repository
 
-| Component | Version |
-|-----------|---------|
-| Python | 3.12 |
-| PyTorch | 2.15.0 nightly (cu132) |
-| CUDA Toolkit | 13.3 |
-| GPU | Ada Lovelace (sm_89) — RTX 4050/4060/4070/4080/4090 |
-| SageAttention | 2.2.0 |
+- Source: https://github.com/MBZUAI-IFM/llama.cpp
+- Branch: model/K2Horizon
+- Version: 0.3.0-dev
+- Build: 10671
+- Commit: 35999d101
 
-## Install
-
-```bash
-pip install sageattention-2.2.0-cp312-cp312-linux_x86_64.whl --no-deps
-```
-
-## Build from Source (if needed)
+## Requirements
 
-### Requirements
-- CUDA Toolkit (nvcc)
-- PyTorch with CUDA support
-- 16GB+ RAM or adequate swap
+- Linux x86_64
+- NVIDIA GPU with CUDA support
+- CUDA 12.8+
 
-### Steps
+## Files
 
-```bash
-git clone --depth 1 https://github.com/thu-ml/SageAttention.git /tmp/SageAttention
-cd /tmp/SageAttention
-
-# Patch: PyTorch 2.15 nightly requires C++20
-sed -i 's/-std=c++17/-std=c++20/g' setup.py
-
-# Build
-CUDA_HOME=/usr/local/cuda-13.3 \
-TORCH_CUDA_ARCH_LIST="8.9" \
-MAX_JOBS=1 \
-NVCC_APPEND_FLAGS="--threads 8" \
-python setup.py install
-```
+| File | Description |
+|------|-------------|
+| `llama-k2horizon-cuda-linux-x64.tar.gz` | K2Horizon llama.cpp CUDA binary |
+| `llama.cpp-b9628-cuda-12.8-amd64.tar.gz` | llama.cpp build b9628 for CUDA 12.8 |
+| `sageattention-2.2.0-cp312-cp312-linux_x86_64.whl` | SageAttention 2.2.0 pre-built wheel |
 
-### Create Wheel
+## Install
 
 ```bash
-python setup.py bdist_wheel
-cp dist/*.whl /path/to/wheels/
+tar xzf llama-k2horizon-cuda-linux-x64.tar.gz
 ```
-
-## Notes
-
-- Built on Ubuntu 24.04 with 14GB RAM + 19GB swap
-- `MAX_JOBS=1` used to avoid OOM during CUDA kernel compilation
-- PyTorch 2.15 nightly requires C++20; upstream SageAttention ships with C++17 which must be patched
-- Wheel includes sm_80 (Ampere) and sm_89 (Ada) kernels