[Question]: ValueError: Tokenizer class TokenizersBackend does not exist or is not currently imported.

(vllm) root@fdf:~/CoPaw-Flash-9B# LD_LIBRARY_PATH=/root/miniforge3/envs/vllm/lib:/usr/local/cuda-12.1/lib64:$LD_LIBRARY_PATH vllm serve /root/CoPaw-Flash-9B/CoPaw-Flash-9B --port 8000 --tensor-parallel-size 1 --max-model-len 262144 --reasoning-parser qwen3 --enable-auto-tool-choice --tool-call-parser qwen3_xml --trust_remote_code
(APIServer pid=8919) INFO 04-08 05:31:30 [utils.py:299]
(APIServer pid=8919) INFO 04-08 05:31:30 [utils.py:299] █ █ █▄ ▄█
(APIServer pid=8919) INFO 04-08 05:31:30 [utils.py:299] ▄▄ ▄█ █ █ █ ▀▄▀ █ version 0.19.0
(APIServer pid=8919) INFO 04-08 05:31:30 [utils.py:299] █▄█▀ █ █ █ █ model /root/CoPaw-Flash-9B/CoPaw-Flash-9B
(APIServer pid=8919) INFO 04-08 05:31:30 [utils.py:299] ▀▀ ▀▀▀▀▀ ▀▀▀▀▀ ▀ ▀
(APIServer pid=8919) INFO 04-08 05:31:30 [utils.py:299]
(APIServer pid=8919) INFO 04-08 05:31:30 [utils.py:233] non-default args: {'model_tag': '/root/CoPaw-Flash-9B/CoPaw-Flash-9B', 'enable_auto_tool_choice': True, 'tool_call_parser': 'qwen3_xml', 'model': '/root/CoPaw-Flash-9B/CoPaw-Flash-9B', 'trust_remote_code': True, 'max_model_len': 262144, 'reasoning_parser': 'qwen3'}
(APIServer pid=8919) INFO 04-08 05:31:30 [model.py:549] Resolved architecture: Qwen3_5ForConditionalGeneration
(APIServer pid=8919) INFO 04-08 05:31:30 [model.py:2013] Downcasting torch.float32 to torch.bfloat16.
(APIServer pid=8919) INFO 04-08 05:31:30 [model.py:1678] Using max model len 262144
(APIServer pid=8919) INFO 04-08 05:31:30 [config.py:281] Setting attention block size to 528 tokens to ensure that attention page size is >= mamba page size.
(APIServer pid=8919) INFO 04-08 05:31:30 [config.py:312] Padding mamba page size by 0.76% to ensure that mamba page size and attention page size are exactly equal.
(APIServer pid=8919) INFO 04-08 05:31:30 [vllm.py:790] Asynchronous scheduling is enabled.
(APIServer pid=8919) Traceback (most recent call last):
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/bin/vllm", line 6, in
(APIServer pid=8919) sys.exit(main())
(APIServer pid=8919) ^^^^^^
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/lib/python3.12/site-packages/vllm/entrypoints/cli/main.py", line 75, in main
(APIServer pid=8919) args.dispatch_function(args)
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/lib/python3.12/site-packages/vllm/entrypoints/cli/serve.py", line 122, in cmd
(APIServer pid=8919) uvloop.run(run_server(args))
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/lib/python3.12/site-packages/uvloop/__init__.py", line 96, in run
(APIServer pid=8919) return __asyncio.run(
(APIServer pid=8919) ^^^^^^^^^^^^^^
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/lib/python3.12/asyncio/runners.py", line 195, in run
(APIServer pid=8919) return runner.run(main)
(APIServer pid=8919) ^^^^^^^^^^^^^^^^
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/lib/python3.12/asyncio/runners.py", line 118, in run
(APIServer pid=8919) return self._loop.run_until_complete(task)
(APIServer pid=8919) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=8919) File "uvloop/loop.pyx", line 1518, in uvloop.loop.Loop.run_until_complete
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/lib/python3.12/site-packages/uvloop/__init__.py", line 48, in wrapper
(APIServer pid=8919) return await main
(APIServer pid=8919) ^^^^^^^^^^
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/lib/python3.12/site-packages/vllm/entrypoints/openai/api_server.py", line 670, in run_server
(APIServer pid=8919) await run_server_worker(listen_address, sock, args, **uvicorn_kwargs)
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/lib/python3.12/site-packages/vllm/entrypoints/openai/api_server.py", line 684, in run_server_worker
(APIServer pid=8919) async with build_async_engine_client(
(APIServer pid=8919) ^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/lib/python3.12/contextlib.py", line 210, in __aenter__
(APIServer pid=8919) return await anext(self.gen)
(APIServer pid=8919) ^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/lib/python3.12/site-packages/vllm/entrypoints/openai/api_server.py", line 100, in build_async_engine_client
(APIServer pid=8919) async with build_async_engine_client_from_engine_args(
(APIServer pid=8919) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/lib/python3.12/contextlib.py", line 210, in __aenter__
(APIServer pid=8919) return await anext(self.gen)
(APIServer pid=8919) ^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/lib/python3.12/site-packages/vllm/entrypoints/openai/api_server.py", line 136, in build_async_engine_client_from_engine_args
(APIServer pid=8919) async_llm = AsyncLLM.from_vllm_config(
(APIServer pid=8919) ^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/lib/python3.12/site-packages/vllm/v1/engine/async_llm.py", line 225, in from_vllm_config
(APIServer pid=8919) return cls(
(APIServer pid=8919) ^^^^
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/lib/python3.12/site-packages/vllm/v1/engine/async_llm.py", line 135, in __init__
(APIServer pid=8919) self.renderer = renderer = renderer_from_config(self.vllm_config)
(APIServer pid=8919) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/lib/python3.12/site-packages/vllm/renderers/registry.py", line 83, in renderer_from_config
(APIServer pid=8919) tokenizer = cached_tokenizer_from_config(model_config, **kwargs)
(APIServer pid=8919) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/lib/python3.12/site-packages/vllm/tokenizers/registry.py", line 227, in cached_tokenizer_from_config
(APIServer pid=8919) return cached_get_tokenizer(
(APIServer pid=8919) ^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/lib/python3.12/site-packages/vllm/tokenizers/registry.py", line 210, in get_tokenizer
(APIServer pid=8919) tokenizer = tokenizer_cls_.from_pretrained(tokenizer_name, *args, **kwargs)
(APIServer pid=8919) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/lib/python3.12/site-packages/vllm/tokenizers/hf.py", line 110, in from_pretrained
(APIServer pid=8919) raise e
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/lib/python3.12/site-packages/vllm/tokenizers/hf.py", line 85, in from_pretrained
(APIServer pid=8919) tokenizer = AutoTokenizer.from_pretrained(
(APIServer pid=8919) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=8919) File "/root/miniforge3/envs/vllm/lib/python3.12/site-packages/transformers/models/auto/tokenization_auto.py", line 1153, in from_pretrained
(APIServer pid=8919) raise ValueError(
(APIServer pid=8919) ValueError: Tokenizer class TokenizersBackend does not exist or is not currently imported.

这是什么原因？折腾半天了，模型是从模搭下载的，使用vllm的docker和虚拟环境都试了，都是这个问题。