Unable to launch the OpenAI API [Vicuna-7B]. Error log: Using pad_token, but it is not set yet.

Nobody has claimed this yet.

Would you suggest me some ways to debug that?
By the way, this model can be successfully inferenced with fastchat cli.

Error Log:
2023-06-13 15:19:44 | INFO | model_worker | Loading the model vicuna-7b on worker 005f53 ...
Loading checkpoint shards: 0%| | 0/2 [00:00 │
2023-06-13 15:20:03 | ERROR | stderr | │ │
2023-06-13 15:20:03 | ERROR | stderr | │ 411 │ │ │ ) │
2023-06-13 15:20:03 | ERROR | stderr | │ 412 │ │ os.environ["CUDA_VISIBLE_DEVICES"] = args.gpus │
2023-06-13 15:20:03 | ERROR | stderr | │ 413 │ │
2023-06-13 15:20:03 | ERROR | stderr | │ ❱ 414 │ worker = ModelWorker( │
2023-06-13 15:20:03 | ERROR | stderr | │ 415 │ │ args.controller_address, │
2023-06-13 15:20:03 | ERROR | stderr | │ 416 │ │ args.worker_address, │
2023-06-13 15:20:03 | ERROR | stderr | │ 417 │ │ worker_id, │
2023-06-13 15:20:03 | ERROR | stderr | │ │
2023-06-13 15:20:03 | ERROR | stderr | │ /mnt/lustre/duanhaodong/anaconda3/envs/mm2/lib/python3.8/site-packages/fastchat/serve/model_work │
2023-06-13 15:20:03 | ERROR | stderr | │ er.py:103 in __init__ │
2023-06-13 15:20:03 | ERROR | stderr | │ │
2023-06-13 15:20:03 | ERROR | stderr | │ 100 │ │ │ self.generate_stream_func = generate_stream │
2023-06-13 15:20:03 | ERROR | stderr | │ 101 │ │ │
2023-06-13 15:20:03 | ERROR | stderr | │ 102 │ │ if not no_register: │
2023-06-13 15:20:03 | ERROR | stderr | │ ❱ 103 │ │ │ self.register_to_controller() │
2023-06-13 15:20:03 | ERROR | stderr | │ 104 │ │ │ self.heart_beat_thread = threading.Thread( │
2023-06-13 15:20:03 | ERROR | stderr | │ 105 │ │ │ │ target=heart_beat_worker, args=(self,) │
2023-06-13 15:20:03 | ERROR | stderr | │ 106 │ │ │ ) │
2023-06-13 15:20:03 | ERROR | stderr | │ │
2023-06-13 15:20:03 | ERROR | stderr | │ /mnt/lustre/duanhaodong/anaconda3/envs/mm2/lib/python3.8/site-packages/fastchat/serve/model_work │
2023-06-13 15:20:03 | ERROR | stderr | │ er.py:119 in register_to_controller │
2023-06-13 15:20:03 | ERROR | stderr | │ │
2023-06-13 15:20:03 | ERROR | stderr | │ 116 │ │ │ "worker_status": self.get_status(), │
2023-06-13 15:20:03 | ERROR | stderr | │ 117 │ │ } │
2023-06-13 15:20:03 | ERROR | stderr | │ 118 │ │ r = requests.post(url, json=data) │
2023-06-13 15:20:03 | ERROR | stderr | │ ❱ 119 │ │ assert r.status_code == 200 │
2023-06-13 15:20:03 | ERROR | stderr | │ 120 │ │
2023-06-13 15:20:03 | ERROR | stderr | │ 121 │ def send_heart_beat(self): │
2023-06-13 15:20:03 | ERROR | stderr | │ 122 │ │ logger.info( │
2023-06-13 15:20:03 | ERROR | stderr | ╰────────────────────────────────────────────────────�
��─────────────────────────────────────────────╯
2023-06-13 15:20:03 | ERROR | stderr | AssertionError

Start with fastchat/serve/model_worker.py, especially ModelWorker initialization and register_to_controller, then compare the Vicuna-7B launch path with the fastchat CLI path. Reproduce the reported pad_token warning and registration AssertionError; done means the OpenAI API worker launches and registers successfully for Vicuna-7B.

Written by the indexing model from the issue text.