Created by: PseudoProgrammer

when i add sparse_remote_update=True to input of hidden layer,it returns error my paddle version is 1.1.0.3 my config: Layer(type = "data", name = "input1", size = 806867) Layer(inputs = [Input("input1", parameter_name = "_layer1_1.w", sparse_remote_update = True)], name = "layer1_1", bias = Bias(parameter_name = "_layer1_1.bias"), active_type = "tanh", type = "fc", size = 80)

the error:

Thu Dec 8 21:44:02 2016[1,0]:+ ./paddle_trainer --num_gradient_servers=1 --trainer_id=0 --pservers=10.90.164.41 --rdma_tcp=tcp --nics=xgbe0 --saving_period=1 --port=7164 --ports_num=1 --local=0 --comment=job.14582.instances --dot_period=1000 --log_period=1000 --num_passes=100 --trainer_count=10 --ports_num_for_sparse=1 --use_sparse_updater=1 --use_old_updater=1 --config=conf/trainer_config.conf --save_dir=./output --python_path=./python-gcc345 --python_bin=python2.7 --use_gpu=0 Thu Dec 8 21:44:09 2016[1,0]:F1208 21:44:09.404992 9555 SparseRowMatrix.cpp:257] Check failed: blockSize % this->width == 0 Thu Dec 8 21:44:09 2016[1,0]:*** Check failure stack trace: *** Thu Dec 8 21:44:09 2016[1,0]: @ 0x8d7788 google::LogMessage::Fail() Thu Dec 8 21:44:09 2016[1,0]: @ 0x8d76e0 google::LogMessage::SendToLog() Thu Dec 8 21:44:09 2016[1,0]: @ 0x8d7175 google::LogMessage::Flush() Thu Dec 8 21:44:09 2016[1,0]: @ 0x8d9f36 google::LogMessageFatal::~LogMessageFatal() Thu Dec 8 21:44:09 2016[1,0]: @ 0x84aa89 paddle::SparsePrefetchRowCpuMatrix: :setBlockSize() Thu Dec 8 21:44:09 2016[1,0]: @ 0x76ef2d paddle::ParameterClient2::init() Thu Dec 8 21:44:09 2016[1,0]: @ 0x7553c5 paddle::SparseRemoteParameterUpdater::init() Thu Dec 8 21:44:09 2016[1,0]: @ 0x74b437 _ZNSt6thread5_ImplISt12_Bind_resultIvFZN6paddle14SyncThreadPool5startEvEUliE_mEEE6_M_runEv Thu Dec 8 21:44:09 2016[1,0]: @ 0x7f0d5d408462 execute_native_thread_routine Thu Dec 8 21:44:09 2016[1,0]: @ 0x7f0d5da91d30 start_thread Thu Dec 8 21:44:09 2016[1,0]: @ 0x7f0d5cc58afd clone