Add a new optimizer

fix: add EinsumGrad

fix: optimize some APIs

Add pad preprocessing for `imdb` dataset

fix: inconsistent shape error while training embedding layer

fix: A bug fix in KerasObjectLoader.cs

I added `ToArray()` so that there is no "The collection has changed" error after `_delete_tracking`.

fix: fix the bug of repeated progress bar in Model.fit()

fix: revise earlystopping callback's min_delta parameter and add GRUCell layer

feat: Add UpSampling1D layer and test.

refactor: Standardize TensorFlowNET.Keras/Losses/*

fix: type converting errors when loading imdb dataset

Improve the API of  LayerNormalization

fix: error pf training RNN on Linux

feat: Support training of RNN and LSTM.

Fix my bad:

Bad handling between test_function and test_step_multi_inputs_function.

IModel.Dictionary<string, float> evaluate(NDArray, NDArray, ...) is now IModel.Dictionary<string, float> evaluate(Tensor, Tensor, ...)
Merge Model.Evaluate.test_step_multi_inputs_function(...) and Model.Evaluate.test_function(...)

Note: An internal function need to add an explicit cast in Tensor