fluid 集群训练 pserver报错RuntimeError: all distributed lookup_table_ops should have only one table
Created by: CruiseSun
Traceback (most recent call last): File "train.py", line 182, in main(use_cuda, is_local) File "train.py", line 174, in main train(use_cuda=use_cuda, save_dirname=save_dirname, is_local=is_local) File "train.py", line 93, in train optimizer.minimize(avg_cost) File "/home/disk1/normandy/maybach/app-user-20181228115951-2365/workspace/python27-gcc482/lib/python2.7/site-packages/paddle/fluid/optimizer.py", line 310, in minimize self._process_distribute_lookuptable(params_grads, loss, startup_program) File "/home/disk1/normandy/maybach/app-user-20181228115951-2365/workspace/python27-gcc482/lib/python2.7/site-packages/paddle/fluid/optimizer.py", line 262, in _process_distribute_lookuptable table_name = find_distributed_lookup_table(program) File "/home/disk1/normandy/maybach/app-user-20181228115951-2365/workspace/python27-gcc482/lib/python2.7/site-packages/paddle/fluid/distribute_lookup_table.py", line 33, in find_distributed_lookup_table raise RuntimeError("all distributed lookup_table_ops" RuntimeError: all distributed lookup_table_ops should have only one table