mw66

Experienced the same problem: with ``` x = layers.GlobalAveragePooling1D(data_format="channels_last")(x) ``` the number of parameters are (only show the last few rows that differ): ``` ... global_average_pooling1d (Glob (None, 1) 0...

Hi , @fchollet https://github.com/keras-team/keras-io/blob/8ec940273419cd5a61f056f462753ba4ac0c9590/examples/timeseries/timeseries_classification_transformer.py#L115 This line need to be changed to `channel_first` to see the good result on: https://keras.io/examples/timeseries/timeseries_classification_transformer/ Can you make this change? can also add explanation why `channel_first`...

BTW, using `auto` won't compile in D: ``` auto myfunc = (auto x, auto y) => (x + y); ``` with error: ``` tests/build/lambda.d(5): Error: variable `lambda.show.myfunc` - type `void`...

@jmfernandes @noLeash fix : #507 continuation of: #517

@jmfernandes can you take a look of this simple PR?

@jmfernandes can you take a look of this simple PR? It contains important fix. Thanks

@jmfernandes can you merge this simple and important PR?

For people who need this fix, please install from here: ``` pip3 install -U git+https://github.com/mw66/robin_stocks.git@fix_2 ```

@PBbanana can you create a PR? Thanks.

https://stackoverflow.com/a/79438138/873275 """ Experiencing the same problem. I noticed that with pytorch backend the GPU memory is ~10x smaller, so I increased the batch size to be 16x, then the training...