Qwen2.5推理报错

[zsy-v]

python3 pipeline.py --model_path qwen2.5-1.5b_int4_seq512_1dev.bmodel --tokenizer_path ../support/token_config/ --devid 0 --generation_mode greedy
Load ../support/token_config/ ...
Device [ 0 ] loading ....
Model[qwen2.5-1.5b_int4_seq512_1dev.bmodel] loading ....
Done!

Load Time: 21.452 s

=================================================================

1. If you want to quit, please enter one of [q, quit, exit]
2. To create a new chat session, please enter one of [clear, new]
   =================================================================

Question: nihao

Answer: [bmlib_runtime][error] bmdev handle sync api failed, ioclt ret = -1 455
python3: /data/zhousy26/hello/python_demo/../support/include/utils.h:437: void empty(bm_context*&, bm_device_mem_t&): Assertion `BM_SUCCESS == ret' failed.
Aborted

这个是发生了什么问题？

[chuxiaoyi2023]

bm-smi康康呢，像是环境问题