Created by: fourseven471

Paddle_baseline_KDD2019 : https://github.com/PaddlePaddle/models/tree/develop/PaddleRec/ctr/Paddle_baseline_KDD2019

The above baseline mode can run in python2 + cpu environment. But when i run 'local_train.py' in python3 environment, the errors occur as follow. Can you determine whether it is a code problem or an environment of python3 problem？

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

EnforceNotMet Traceback (most recent call last) ~/mnt/py36/baseline/local_train.py in 81 82 if name == 'main': ---> 83 train()

~/mnt/py36/baseline/local_train.py in train() 37 print("ready to optimize") 38 optimizer = fluid.optimizer.SGD(learning_rate=1e-4) ---> 39 optimizer.minimize(loss) 40 #single machine CPU training. more options on trainig please visit PaddlePaddle site 41 exe = fluid.Executor(fluid.CPUPlace())

/usr/local/cuda-9.2/conda/envs/py36-paddle/lib/python3.6/site-packages/paddle/fluid/optimizer.py in minimize(self, loss, startup_program, parameter_list, no_grad_set) 496 startup_program=startup_program, 497 parameter_list=parameter_list, --> 498 no_grad_set=no_grad_set) 499 optimize_ops = self.apply_optimize( 500 loss, startup_program=startup_program, params_grads=params_grads)

/usr/local/cuda-9.2/conda/envs/py36-paddle/lib/python3.6/site-packages/paddle/fluid/optimizer.py in backward(self, loss, startup_program, parameter_list, no_grad_set, callbacks) 401 with program_guard(program, startup_program): 402 params_grads = append_backward(loss, parameter_list, --> 403 no_grad_set, callbacks) 404 # Note: since we can't use all_reduce_op now, 405 # dgc_op should be the last op of one grad.

/usr/local/cuda-9.2/conda/envs/py36-paddle/lib/python3.6/site-packages/paddle/fluid/backward.py in append_backward(loss, parameter_list, no_grad_set, callbacks) 509 510 append_backward_ops(root_block, op_path, root_block, no_grad_dict, --> 511 grad_to_var, callbacks) 512 513 # Because calc_gradient may be called multiple times,

/usr/local/cuda-9.2/conda/envs/py36-paddle/lib/python3.6/site-packages/paddle/fluid/backward.py in append_backward_ops(block, ops, target_block, no_grad_dict, grad_to_var, callbacks) 295 # Getting op's corresponding grad_op 296 grad_op_desc, op_grad_to_var = core.get_grad_op_desc( --> 297 op.desc, cpt.to_text(no_grad_dict[block.idx]), grad_sub_block_list) 298 299 grad_op_descs.extend(grad_op_desc)

EnforceNotMet: grad_op_maker_ should not be null Operator GradOpMaker has not been registered. at [/paddle/paddle/fluid/framework/op_info.h:64] PaddlePaddle Call Stacks: 0 0x7f0395b9aa28p void paddle::platform::EnforceNotMet::Initstd::string(std::string, char const*, int) + 360 1 0x7f0395b9ad77p paddle::platform::EnforceNotMet::EnforceNotMet(std::string const&, char const*, int) + 87 2 0x7f0395b9ba7cp paddle::framework::OpInfo::GradOpMaker() const + 108 3 0x7f0395b92f3ep 4 0x7f0395bccfeep 5 0x7f04001b8ccdp _PyCFunction_FastCallDict + 557 6 0x7f0400251c5ap 7 0x7f0400256911p _PyEval_EvalFrameDefault + 17649 8 0x7f040025185ep 9 0x7f0400251b72p 10 0x7f0400256911p _PyEval_EvalFrameDefault + 17649