redvault-ai/llama_proxy_man/config.yaml

26 lines
626 B
YAML

hardware:
ram: 64G
vram: 8G
models:
- port: 18080
internal_port: 28080
env:
CUDA_VISIBLE_DEVICES: 0
HSA_OVERRIDE_GFX_VERSION: '11.0.0'
args:
model: /home/tristand/Downloads/models/Phi-3.5-mini-instruct-Q6_K_L.gguf
gpu-layers: 9999
ctx-size: 4096
vram_usage: 6G
ram_usage: 500M
- port: 18081
internal_port: 28081
env:
CUDA_VISIBLE_DEVICES: 0
HSA_OVERRIDE_GFX_VERSION: '11.0.0'
args:
model: /home/tristand/Downloads/models/Phi-3.5-mini-instruct-Q6_K_L.gguf
gpu-layers: 9999
ctx-size: 4096
vram_usage: 6G
ram_usage: 500M