vllm部署模型服务的时候会遇到报错

[icey-zhang]

CUDA_CUDA_VISIBLE_DEVICES=1,2,3,4 vllm serve /root/workspace/models/IQuest-Coder-V1-40B-Instruct/ --port 8002 --served-model-name iquest --host 0.0.0.0 --trust-remote-code

加载模型服务的时候会遇到报错，请问这是权重问题么？缺少lm_head.weight

使用vllm版本
Name: vllm
Version: 0.8.2
使用transformers版本
Name: transformers
Version: 4.53.3

INFO 01-21 15:13:39 [loader.py:447] Loading weights took 33.29 seconds
ERROR 01-21 15:13:40 [core.py:343] EngineCore hit an exception: Traceback (most recent call last):
ERROR 01-21 15:13:40 [core.py:343] File "/root/env/ENTER/envs/arpo/lib/python3.10/site-packages/vllm/v1/engine/core.py", line 335, in run_engine_core
ERROR 01-21 15:13:40 [core.py:343] engine_core = EngineCoreProc(*args, **kwargs)
ERROR 01-21 15:13:40 [core.py:343] File "/root/env/ENTER/envs/arpo/lib/python3.10/site-packages/vllm/v1/engine/core.py", line 290, in init
ERROR 01-21 15:13:40 [core.py:343] super().init(vllm_config, executor_class, log_stats)
ERROR 01-21 15:13:40 [core.py:343] File "/root/env/ENTER/envs/arpo/lib/python3.10/site-packages/vllm/v1/engine/core.py", line 60, in init
ERROR 01-21 15:13:40 [core.py:343] self.model_executor = executor_class(vllm_config)
ERROR 01-21 15:13:40 [core.py:343] File "/root/env/ENTER/envs/arpo/lib/python3.10/site-packages/vllm/executor/executor_base.py", line 52, in init
ERROR 01-21 15:13:40 [core.py:343] self._init_executor()
ERROR 01-21 15:13:40 [core.py:343] File "/root/env/ENTER/envs/arpo/lib/python3.10/site-packages/vllm/executor/uniproc_executor.py", line 47, in _init_executor
ERROR 01-21 15:13:40 [core.py:343] self.collective_rpc("load_model")
ERROR 01-21 15:13:40 [core.py:343] File "/root/env/ENTER/envs/arpo/lib/python3.10/site-packages/vllm/executor/uniproc_executor.py", line 56, in collective_rpc
ERROR 01-21 15:13:40 [core.py:343] answer = run_method(self.driver_worker, method, args, kwargs)
ERROR 01-21 15:13:40 [core.py:343] File "/root/env/ENTER/envs/arpo/lib/python3.10/site-packages/vllm/utils.py", line 2255, in run_method
ERROR 01-21 15:13:40 [core.py:343] return func(*args, **kwargs)
ERROR 01-21 15:13:40 [core.py:343] File "/root/env/ENTER/envs/arpo/lib/python3.10/site-packages/vllm/v1/worker/gpu_worker.py", line 136, in load_model
ERROR 01-21 15:13:40 [core.py:343] self.model_runner.load_model()
ERROR 01-21 15:13:40 [core.py:343] File "/root/env/ENTER/envs/arpo/lib/python3.10/site-packages/vllm/v1/worker/gpu_model_runner.py", line 1177, in load_model
ERROR 01-21 15:13:40 [core.py:343] self.model = get_model(vllm_config=self.vllm_config)
ERROR 01-21 15:13:40 [core.py:343] File "/root/env/ENTER/envs/arpo/lib/python3.10/site-packages/vllm/model_executor/model_loader/init.py", line 14, in get_model
ERROR 01-21 15:13:40 [core.py:343] return loader.load_model(vllm_config=vllm_config)
ERROR 01-21 15:13:40 [core.py:343] File "/root/env/ENTER/envs/arpo/lib/python3.10/site-packages/vllm/model_executor/model_loader/loader.py", line 456, in load_model
ERROR 01-21 15:13:40 [core.py:343] raise ValueError(
ERROR 01-21 15:13:40 [core.py:343] ValueError: Following weights were not initialized from checkpoint: {'lm_head.weight'}
ERROR 01-21 15:13:40 [core.py:343]
CRITICAL 01-21 15:13:40 [core_client.py:269] Got fatal signal from worker processes, shutting down. See stack trace above for root cause issue.