update the API to fluid v1.1

[iclementine]

update the API to fluid v1.1

1. Use ParallelExecutor instead of deprecated fluid.layers.get_places and fluid.layers.ParallelDo.
2. Use paddle.batch instead of paddle.v2.batch, which drops the only batch by default even when the number of examples is less than the batch size.
3. add train.sh, 'eval.sh', 'infer.sh' for out-of-box usages.