Download Latest Version
katago-v1.6.0-windows-x64.zip (4.8 MB)
Get an email when there's a new version of Kata_speed
| Name | Modified | Size | Downloads / Week |
|---|---|---|---|
| Parent folder | |||
| katago-v1.14.0-trt8.6.1-cuda12.1-windows-x64.zip | 2024-05-26 | 4.5 MB | |
| katago-v1.14.1-opencl-windows-x64.zip | 2024-05-26 | 4.7 MB | |
| AI.cfg | 2024-05-26 | 31.7 kB | |
| default_gtp.cfg | 2024-05-26 | 31.7 kB | |
| lizgoban.cfg | 2024-05-26 | 31.7 kB | |
| doudoubot.cfg | 2024-05-26 | 31.7 kB | |
| README.md | 2024-05-26 | 1.8 kB | |
| v1.1.1 source code.tar.gz | 2024-05-26 | 9.2 MB | |
| v1.1.1 source code.zip | 2024-05-26 | 9.2 MB | |
| Totals: 9 Items | 27.7 MB | 1 | |
What's Changed
- created network file by @changcheng967 in https://github.com/changcheng967/Kata_speed/pull/1
- Merge pull request [#1] from changcheng967/main by @changcheng967 in https://github.com/changcheng967/Kata_speed/pull/2
- Added Tensor versions by @changcheng967 in https://github.com/changcheng967/Kata_speed/pull/3
New Contributors
- @changcheng967 made their first contribution in https://github.com/changcheng967/Kata_speed/pull/1
Full Changelog: https://github.com/changcheng967/Kata_speed/compare/v1.0.0...v1.1.1
Added TensorRT and CUDA Support: Integration of TensorRT 8.6.1.6 and CUDA 12.1. Improved Calculations: Enhanced inference performance leveraging TensorRT for optimized neural network computations. Powerful Computation Capabilities: Utilizing CUDA 12.1 for accelerated GPU processing, enabling more complex and faster calculations. cuDNN 9.1: Integration of cuDNN 9.1 to enhance deep learning operations with optimized implementations for standard routines. New Networks: Added support for new networks b28 and b60. b28 Network: Designed for medium complexity tasks with optimized performance. b60 Network: Designed for high complexity tasks, providing powerful computational capabilities for advanced operations. Changed Longer Load Times: Initial load times may be longer due to the integration of TensorRT and CUDA. This is a trade-off for significantly better runtime performance during inference. Fixed Compatibility Updates: Ensured compatibility with the latest CUDA and cuDNN versions for smoother operation and support. Known Issues CUDA 12.4 Bug: Running the application with CUDA 12.4 causes it to exit without sending output. Users are advised to use CUDA 12.1 for stable operation until this issue is resolved.
