issues
search
openlangrid
/
mlgrid-services
A framework to serve various generative AIs and other machine learning functions via WebSocket and JSONRPC.
Apache License 2.0
2
stars
0
forks
source link
GPUを使用するメモリ量に基づいて割り当てられるようにする
#23
Open
takawitter
opened
9 months ago
takawitter
commented
9 months ago
現状1プロセス1GPUで割り当てているが、使用メモリ量が足りるのであれば、複数プロセスを同じGPUに割り当てられるようにする。
[x] #24
[x] #26
[ ] 過去の使用メモリ量を元にGPUを割り当てられるようにする
現状1プロセス1GPUで割り当てているが、使用メモリ量が足りるのであれば、複数プロセスを同じGPUに割り当てられるようにする。