pphuman LANE_SEG模型加上 run_mode=trt_fp16报错

[truthsun22]

问题确认 Search before asking

• [X] 我已经查询历史issue，没有发现相似的bug。I have searched the issues and found no similar bug report.

Bug组件 Bug Component

No response

Bug描述 Describe the Bug

在运行pphuman车辆检测加车牌识别时，加上了 --run_mode=trt_fp16后报如下错误 Traceback (most recent call last): File "D:\ai\paddle\pddetect\pipeline\pipeline.py", line 1849, in main() File "D:\ai\paddle\pddetect\pipeline\pipeline.py", line 1828, in main pipeline = Pipeline(FLAGS, cfg) File "D:\ai\paddle\pddetect\pipeline\pipeline.py", line 103, in init self.predictor = PipePredictor(args, cfg, self.is_video) File "D:\ai\paddle\pddetect\pipeline\pipeline.py", line 393, in init self.vehicleplate_detector = PlateRecognizer(args, vehicleplate_cfg) File "D:\ai\paddle\pddetect\pipeline\ppvehicle\vehicle_plate.py", line 245, in init self.textrecognizer = TextRecognizer(args, cfg, use_gpu=use_gpu) File "D:\ai\paddle\pddetect\pipeline\ppvehicle\vehicle_plate.py", line 163, in init create_predictor(args, cfg, 'rec') File "D:\ai\paddle\pddetect\pipeline\ppvehicle\vehicle_plateutils.py", line 178, in create_predictor predictor = inference.create_predictor(config) RuntimeError: (NotFound) Operator (preln_residual_bias) is not registered. [Hint: op_info_ptr should not be null.] (at ..\paddle/fluid/framework/op_info.h:156)

复现环境 Environment

paddle2.4

Bug描述确认 Bug description confirmation

• [X] 我确认已经提供了Bug复现步骤、代码改动说明、以及环境信息，确认问题是可以复现的。I confirm that the bug replication steps, code change instructions, and environment information have been provided, and the problem can be reproduced.

是否愿意提交PR？ Are you willing to submit a PR?

• [ ] 我愿意提交PR！I'd like to help by submitting a PR!

[truthsun22]

标题描述不准确，不是LANE_SEG模型模型的问题，是车牌检测或识别模型在trt16下报错

[lyuwenyu]

导出onnx然后转fp16试一下