Implementing Keras v2.x instead of using Keras v1.x
Implementing tf.Keras which is updated in Tensorflow r1.4
Issue reason:
TypeError: The `mode` argument of `BatchNormalization` no longer exists. `mode=1` and `mode=2` are no longer supported.
UserWarning: Update your `Model` call to the Keras 2 API: `Model(inputs=Tensor("ac..., outputs=Tensor("ti...)`
UserWarning: Update your `Conv1D` call to the Keras 2 API: `Conv1D(1000, 11, strides=2, padding="valid", name="conv1d", kernel_initializer="glorot_uniform", activation="relu")`
UserWarning: Update your `GRU` call to the Keras 2 API: `GRU(1000, return_sequences=True, name="rnn_1", activation="relu", kernel_initializer="glorot_uniform")`
UserWarning: Update your `Dense` call to the Keras 2 API: `Dense(29, name="dense", activation="linear", kernel_initializer="glorot_uniform")`
Issue explained at https://github.com/baidu-research/ba-dls-deepspeech/issues/34
Brief explanation:
Issue reason:
TypeError: The `mode` argument of `BatchNormalization` no longer exists. `mode=1` and `mode=2` are no longer supported.
UserWarning: Update your `Model` call to the Keras 2 API: `Model(inputs=Tensor("ac..., outputs=Tensor("ti...)`
UserWarning: Update your `Conv1D` call to the Keras 2 API: `Conv1D(1000, 11, strides=2, padding="valid", name="conv1d", kernel_initializer="glorot_uniform", activation="relu")`
UserWarning: Update your `GRU` call to the Keras 2 API: `GRU(1000, return_sequences=True, name="rnn_1", activation="relu", kernel_initializer="glorot_uniform")`
UserWarning: Update your `Dense` call to the Keras 2 API: `Dense(29, name="dense", activation="linear", kernel_initializer="glorot_uniform")`