Allow other ranks to profile as well. Doesn't seem to conflict

Signed-off-by: Rich Porter <rich.porter@uber.com>

horovod/spark/keras/remote.py

@@ -152,10 +152,6 @@ def train(serialized_model, train_rows, val_rows, avg_row_size):
                 hvd.callbacks.MetricAverageCallback(),
             ]
 
-            if hvd.local_rank() != 0:
-                # The TB callback appears to conflict across processes on the same machine
-                user_callbacks[:] = [c for c in user_callbacks if not isinstance(c, k.callbacks.TensorBoard)]
-
             callbacks += user_callbacks
 
             # Horovod: save checkpoints only on the first worker to prevent other workers from