PrototypeCommon
OpenBMB/CPM.cu
CPM.cu is a lightweight, high-performance CUDA implementation for LLMs, optimized for end-device inference and featuring cutting-edge techniques in sparse architecture, speculative sampling and quantization.
- Owner
- OpenBMB (organization)
- Stars
- Stars: 243
- Forks
- Forks: 27
- Language
- Cuda
- License
- Apache-2.0
- Created
- 2025
- Tier
- Prototype (100 to 249 ★) What this tier means
More repos by OpenBMB
All 50 repos by OpenBMB- OpenBMB/VoxCPMStars: 40,960PythonLTS
- OpenBMB/ChatDevStars: 34,639PythonLTS
- OpenBMB/MiniCPM-VStars: 27,047PythonLTS
- OpenBMB/MiniCPMStars: 11,780Jupyter NotebookStable
- OpenBMB/XAgentStars: 8,550PythonStable
- OpenBMB/ToolBenchStars: 5,763PythonStable
Nearby in Cuda
All Cuda repos- ZaidQureshi/bamStars: 249CudaPrototype
- AlibabaResearch/flash-llmStars: 246CudaPrototype
- rahul-goel/fused-ssimStars: 245CudaPrototype
- vernamlab/cuFHEStars: 240CudaPrototype
- HenryNdubuaku/cuda-tutorialsStars: 239CudaPrototype
- layumi/Pedestrian_AlignmentStars: 238CudaPrototype