Created by: ming5432ming

之前1.7.2一直可以使用，现在重新bcloud local了一下，发现不能用了，报错如下：

data/fluid_models/face_segmentation/face_segmentation.conf start init ! WARNING: Logging before InitGoogleLogging() is written to STDERR I0917 18:46:43.960003 25002 paddle_helper.h:246] init_paddlepredictor I0917 18:46:43.960059 25002 paddle_helper.h:187] [date:Sep 17 2020][time:18:46:18][version:1.9.8.7][provider:vis-arch] I0917 18:46:43.960070 25002 paddle_helper.h:200] paddle version: 1.8.0 commit: 0231f58e branch: HEAD I0917 18:46:43.960119 25002 paddle_helper.h:203] [wiki: https://dwz.cn/jLej3OTv] I0917 18:46:43.960135 25002 conf_parser.cpp:38] read_text_file! I0917 18:46:43.960202 25002 paddle_helper.h:146] Predictor::init_model() W0917 18:46:43.960275 25002 paddle_helper.h:294] fail to get specify_input_name from conf, set as false I0917 18:46:43.960291 25002 paddle_helper.h:311] use gpu! I0917 18:46:43.960307 25002 paddle_helper.h:98] Predictor::init_gpu() W0917 18:46:43.960323 25002 conf_parser.cpp:167] fail to get: test_fluid_multiclass_gpu_memory_mb W0917 18:46:43.960335 25002 paddle_helper.h:103] fail to get gpu_memory_mb from conf, set as 1024 W0917 18:46:43.960367 25002 conf_parser.cpp:202] fail to get: test_fluid_multiclass_infer_engine W0917 18:46:43.960379 25002 paddle_helper.h:111] disable infer engine W0917 18:46:43.960391 25002 conf_parser.cpp:202] fail to get: test_fluid_multiclass_delete_pass W0917 18:46:43.960400 25002 conf_parser.cpp:202] fail to get: test_fluid_multiclass_append_pass terminate called after throwing an instance of 'paddle::platform::EnforceNotMet' what():

C++ Call Stacks (More useful to developers):

0 std::string paddle::platform::GetTraceBackString<char const*>(char const*&&, char const*, int) 1 paddle::platform::EnforceNotMet::EnforceNotMet(std::__exception_ptr::exception_ptr, char const*, int) 2 std::unique_ptr<paddle::PaddlePredictor, std::default_deletepaddle::PaddlePredictor > paddle::CreatePaddlePredictorpaddle::AnalysisConfig(paddle::AnalysisConfig const&)

Error Message Summary:

Error: Note: Each config can only be used for one predictor. at (/home/scmbuild/workspaces_cluster.dev/baidu.lib.paddlepaddle/baidu/lib/paddlepaddle/Paddle/paddle/fluid/inference/api/analysis_predictor.cc:505)