2
Repos
4
Stars
0
Forks
0
Watchers
Repositories (2)
qwen3.8-27b-mi50-cpp-engine
bespokeontology/qwen3.8-27b-mi50-cpp-engine
C++
C++/HIP Qwen3.8-27B on 3x MI50 32GB (gfx906): 69.17 native / 110.66 committed MTP tok/s; qualified through 262K.
2
0
openPangu-2.0-Flash-CUDA-ROCm
bespokeontology/openPangu-2.0-Flash-CUDA-ROCm
HIP
openPangu Flash92 — C++/CUDA & C++/HIP ROCm Engines. Fully resident, Blackwell tensor cores, no Python. Powered by openPangu.
2
0