BetaRare
0xSero/turboquant
TurboQuant: Near-optimal KV cache quantization for LLM inference (3-bit keys, 2-bit values) with Triton kernels + vLLM integration
- Owner
- 0xSero (personal account)
- Stars
- Stars: 1,814
- Forks
- Forks: 199
- Language
- Python
- License
- GPL-3.0
- Created
- 2026
- Tier
- Beta (700 to 1,999 ★) What this tier means
More repos by 0xSero
All 14 repos by 0xSero- 0xSero/litterStars: 2,539RustRC
- 0xSero/parchiStars: 546TypeScriptAlpha
- 0xSero/orchestraStars: 281TypeScriptAlpha
- 0xSero/ascii-artStars: 247JavaScriptPrototype
- 0xSero/pi-brainStars: 184TypeScriptPrototype
- 0xSero/moe-compressStars: 183PythonPrototype
Nearby in Python
All Python repos- kohya-ss/sd-webui-additional-networksStars: 1,814PythonBeta
- triple-mu/YOLOv8-TensorRTStars: 1,814PythonBeta
- zi-yue-1129/DATAGENStars: 1,814PythonBeta
- erikwijmans/Pointnet2_PyTorchStars: 1,814PythonBeta
- google/compare_ganStars: 1,814PythonBeta
- josh0xA/darkdumpStars: 1,814PythonBeta