# vLLM version
VLLM_VERSION="v0.8.0"

# Model configuration
VLLM_MODEL="facebook/opt-125m"
VLLM_MAX_MODEL_LEN=2048
VLLM_GPU_MEMORY_UTIL=0.9

# Hugging Face token for model downloads
HF_TOKEN=""

# Port to bind to on the host machine
VLLM_PORT_OVERRIDE=8000