Performance issues in /examples/train_embedding.py (by P3)

[DLPerf]

Hello! I've found a performance issue in /examples/train_embedding.py: batch() should be called before map(), which could make your program more efficient. Here is the tensorflow document to support it.

Detailed description is listed below:

• ds_train = ds_train.batch(BATCH_SIZE)(line 33) should be called before ds_train = ds_train.map(transform)(line 32).
• ds_test = ds_test.batch(BATCH_SIZE)(line 37) should be called before ds_test = ds_test.map(transform)(line 36).

Besides, you need to check the function called in map()(e.g., transform called in ds_test.map(transform)) whether to be affected or not to make the changed code work properly. For example, if transform needs data with shape (x, y, z) as its input before fix, it would require data with shape (batch_size, x, y, z).

Looking forward to your reply. Btw, I am very glad to create a PR to fix it if you are too busy.