added audio_dataset.py

[hazemessamm]

No description provided.

[fchollet]

Thanks for the PR. This is great!

Please add corresponding unit tests in audio_dataset_test.py.

[fchollet]

Thanks for the update! The tests look good.

One last thing: please add support for the subset="both" API like in this PR: https://github.com/keras-team/keras/pull/16390/files

[fchollet]

LGTM, thank you!

[fchollet]

LGTM, thank you for the great contribution! 👍

[fchollet]

It seems there are test failures, please take a look. https://source.cloud.google.com/results/invocations/44f7b725-0b1e-4576-b30f-a88bbddfbdbc/targets/keras%2Fgithub%2Fubuntu%2Fcpu%2Fpresubmit/log

[hazemessamm]

It seems there are test failures, please take a look. https://source.cloud.google.com/results/invocations/44f7b725-0b1e-4576-b30f-a88bbddfbdbc/targets/keras%2Fgithub%2Fubuntu%2Fcpu%2Fpresubmit/log

The test failures are because of the imports. ModuleNotFoundError: No module named 'keras.utils.audio_dataset' . I thought that I should import it by adding keras.utils but it seems that it cannot find it. Is there any other file I should edit to include audio_dataset ?

[hazemessamm]

It seems there are test failures, please take a look. https://source.cloud.google.com/results/invocations/44f7b725-0b1e-4576-b30f-a88bbddfbdbc/targets/keras%2Fgithub%2Fubuntu%2Fcpu%2Fpresubmit/log

The test failures are because of the imports. ModuleNotFoundError: No module named 'keras.utils.audio_dataset' . I thought that I should import it by adding keras.utils but it seems that it cannot find it. Is there any other file I should edit to include audio_dataset ?

I think the problem was that I did not add edit the BUILD file and I edited it.

[fchollet]

I think you will also need to list audio_dataset in the utils library in the BUILD file, here:

py_library(
    name = "utils",
    srcs = [
        "__init__.py",
    ],
    srcs_version = "PY3",
    deps = [
        ":data_utils",
        ":generic_utils",
        ":image_dataset",
        ":image_utils",
        ":layer_utils",
        ":np_utils",
        ":text_dataset",
        ":timeseries_dataset",
        ":vis_utils",
    ],
)

[hazemessamm]

I think you will also need to list audio_dataset in the utils library in the BUILD file, here:

py_library(
    name = "utils",
    srcs = [
        "__init__.py",
    ],
    srcs_version = "PY3",
    deps = [
        ":data_utils",
        ":generic_utils",
        ":image_dataset",
        ":image_utils",
        ":layer_utils",
        ":np_utils",
        ":text_dataset",
        ":timeseries_dataset",
        ":vis_utils",
    ],
)

I Added it.

[hazemessamm]

LGTM, thank you for the great contribution! 👍

You are welcome 🙏🏻

[fchollet]

To confirm, your email address registered with your GitHub account is hazemessam199@gmail.com? It seems your commits aren't signed with an email address, which can lead to loss of authorship tracking in the git history.

[hazemessamm]

To confirm, your email address registered with your GitHub account is hazemessam199@gmail.com? It seems your commits aren't signed with an email address, which can lead to loss of authorship tracking in the git history.

Yes, this email address is registered with my GitHub account but it's not the primary one, I don't know why my commits are not signed and I will check it.

[fchollet]

I've set the merge commits to be signed with your registered email, hopefully that will work.