Created by: yufengwhy

我用的是另外一个广告数据集kaggle avazu。因为只用了类别特征，去掉了连续值特征对应的输入，运行几分钟后报错。谢谢大家的建议~

(paddle) ➜ deep_fm CUDA_VISIBLE_DEVICES=2 python train.py --train_data_path data/train.txt --test_data_path data/valid.txt I0109 11:21:08.664938 31376 Util.cpp:166] commandline: --use_gpu=1 --trainer_count=1 I0109 11:21:14.895238 31376 GradientMachine.cpp:94] Initing parameters.. I0109 11:21:14.985219 31376 GradientMachine.cpp:101] Init parameters done. F0109 11:28:41.559394 31376 Matrix.cpp:653] Not supported

* Check failure stack trace: *

@ 0x7fd3475852fd google::LogMessage::Fail() @ 0x7fd347588dac google::LogMessage::SendToLog() @ 0x7fd347584e23 google::LogMessage::Flush() @ 0x7fd34758a2be google::LogMessageFatal::~LogMessageFatal() @ 0x7fd3473a6246 paddle::GpuMatrix::mul() @ 0x7fd34720e8ef paddle::FullyConnectedLayer::forward() @ 0x7fd34728af8d paddle::NeuralNetwork::forward() @ 0x7fd34728bca3 paddle::GradientMachine::forwardBackward() @ 0x7fd347561004 GradientMachine::forwardBackward() @ 0x7fd3470fbad9 _wrap_GradientMachine_forwardBackward @ 0x7fd3a543cbad PyEval_EvalFrameEx @ 0x7fd3a543dc3e PyEval_EvalCodeEx @ 0x7fd3a543d1f7 PyEval_EvalFrameEx @ 0x7fd3a543dc3e PyEval_EvalCodeEx @ 0x7fd3a543d1f7 PyEval_EvalFrameEx @ 0x7fd3a543dc3e PyEval_EvalCodeEx @ 0x7fd3a543d1f7 PyEval_EvalFrameEx @ 0x7fd3a543dc3e PyEval_EvalCodeEx @ 0x7fd3a543d1f7 PyEval_EvalFrameEx @ 0x7fd3a543dc3e PyEval_EvalCodeEx @ 0x7fd3a543dd52 PyEval_EvalCode @ 0x7fd3a545e450 PyRun_FileExFlags @ 0x7fd3a545e62f PyRun_SimpleFileExFlags @ 0x7fd3a5473fd4 Py_Main @ 0x7fd3a466cf45 __libc_start_main @ 0x400729 (unknown) [1] 31376 abort (core dumped) CUDA_VISIBLE_DEVICES=2 python train.py --train_data_path data/train.txt