Hello! I've found a performance issue in /ctr_of_recommendation/Wide&Deep_Demo: 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:
- /wide_component.py:
dataset.batch(batch_size)(here) should be called before .map(parse_csv, num_parallel_calls=5)(here).
- /wide_deep.py:
dataset.batch(batch_size)(here) should be called before .map(parse_csv, num_parallel_calls=5)(here).
Besides, you need to check the function called in map()(e.g., parse_csv called in .map(parse_csv, num_parallel_calls=5)) whether to be affected or not to make the changed code work properly. For example, if parse_csv 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.
Hello! I've found a performance issue in /ctr_of_recommendation/Wide&Deep_Demo:
batch()should be called beforemap(), which could make your program more efficient. Here is the tensorflow document to support it.Detailed description is listed below:
dataset.batch(batch_size)(here) should be called before.map(parse_csv, num_parallel_calls=5)(here).dataset.batch(batch_size)(here) should be called before.map(parse_csv, num_parallel_calls=5)(here).Besides, you need to check the function called in
map()(e.g.,parse_csvcalled in.map(parse_csv, num_parallel_calls=5)) whether to be affected or not to make the changed code work properly. For example, ifparse_csvneeds 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.