micro-llmapi/tasks.md

check out the image size

origin: ghcr.io/ggerganov/llama.cpp:server-cuda

custom: artifactory.turkcell.com.tr/local-docker-dist-dev/com/turkcell/sensai/llmapi/llamacpp-base:0.0.1

upload to artifactory

pull -> no VPN, push -> open VPN

prepare volume mounts

locate source model gguf

#  llama 8b 32fp indir citrix ttech'e indir oradan oc volume'a gönder

write template env file to use in oc apply as paramfile

pvc kapasite al

fp32 modeli upload et

modelname güncelle deployment