File size: 311 Bytes
197cde1
 
 
 
1
2
3
4
5
# Copy to env.sh and fill in. generate.py reads these from the environment.
export VLLM_BASE_URL="http://localhost:8001/v1"
export VLLM_MODEL="gemma-4-31B-it-fp8-mtp"   # the --served-model-name of your vLLM server
export VLLM_API_KEY="..."                     # only if your server was started with an API key