xierhacker / learningtensorflow Goto Github PK

Hello, I found a performance issue in the definition of test, Project/NeuralMachineTranslation/model/test.py, tf.expand_dims(input=[1],axis=0) will be called repeatedly during program execution, resulting in reduced efficiency. I think it should be created before the loop.

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 /Project: 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:

• /NeuralMachineTranslation/model/test.py: parsed_dataset.batch(1)(here) should be called before dataset.map(map_func=_parse_data)(here).
• /LanguageModel/dataset/ptb_process.py: parsed_dataset.batch(2)(here) should be called before dataset.map(map_func=_parse_data)(here).
• /LanguageModel/model/test.py: parsed_dataset.batch(parameter.BATCH_SIZE)(here) should be called before dataset.map(map_func=_parse_data)(here).
• /LanguageModel/model/train.py: .batch(parameter.BATCH_SIZE)(here) should be called before dataset.map(map_func=_parse_data)(here).

Besides, you need to check the function called in map()(e.g., _parse_data called in dataset.map(map_func=_parse_data)) whether to be affected or not to make the changed code work properly. For example, if _parse_data 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 found a performance issue in 11.TFRecord/read_cifar10.py ,
dataset=dataset.map(map_func=_parse_data) was called without num_parallel_calls.
I think it will increase the efficiency of your program if you add this.

The same issues also exist in dataset=dataset.map(map_func=_parse_data) , parsed_dataset = dataset.map, parsed_dataset = dataset.map,parsed_dataset = dataset.map ,parsed_dataset = dataset.map and dataset=dataset.map

Here is the documemtation of tensorflow to support this thing.

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