gorilla-openfunctions-v2模型本地GPU电脑运行模型生成和评估命令报错

[gaochunrong]

本地电脑下载了gorilla-openfunctions-v2模型，用export HF_HOME=XX指向该模型路径。
handler_map.py文件中api_inference_handler_map的"gorilla-openfunctions-v2": GorillaHandler去掉，在local_inference_handler_map中添加
"gorilla-openfunctions-v2": GorillaHandler ,
执行命令：bfcl generate --model gorilla-openfunctions-v2 --test-category simple --backend vllm --num-gpus 1 --gpu-memory-utilization 0.9

报如下错误：
提示：{"id": "simple_18", "result": "Error during inference: HTTPSConnectionPool(host='luigi.millennium.berkeley.edu', port=443): Max retries exceeded with url: /v1/chat/completions (Caused by SSLError(SSLCertVerificationError(1, '[SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: certificate has expired (_ssl.c:1007)')))"}

现在不是通过API访问gorilla-openfunctions-v2，请问如何解决以上问题？