[Lack parameter] javacpp-pytorch Transformer Encoder Decoder lack parameter

[mullerhai]

Hi, @saudet , compare with python pytorch offical doc ,javacpp pytorch Transformer Encoder Decoder lack some parameters ,

python need full four parameters : [ layer_norm_eps， batch_first，norm_first， bias, ] [enable_nested_tensor]
but the javacpp Transformer Encoder Decoder options dan't have ,I not know if cpp libtorch have these parameters or not.

https://pytorch.org/docs/2.3/generated/torch.nn.TransformerEncoderLayer.html
https://pytorch.org/docs/2.3/generated/torch.nn.TransformerDecoderLayer.html
https://pytorch.org/docs/2.3/generated/torch.nn.TransformerEncoder.html
https://pytorch.org/docs/2.3/generated/torch.nn.TransformerDecoder.html

layer_norm_eps ([float](https://docs.python.org/3/library/functions.html#float)) – the eps value in layer normalization components (default=1e-5).

batch_first ([bool](https://docs.python.org/3/library/functions.html#bool)) – If True, then the input and output tensors are provided as (batch, seq, feature). Default: False (seq, batch, feature).

norm_first ([bool](https://docs.python.org/3/library/functions.html#bool)) – if True, layer norm is done prior to attention and feedforward operations, respectively. Otherwise it’s done after. Default: False (after).

bias ([bool](https://docs.python.org/3/library/functions.html#bool)) – If set to False, Linear and LayerNorm layers will not learn an additive bias. Default: True.

·```
enable_nested_tensor

(bool) – if True, input will automatically convert to nested tensor (and convert back on output). This will improve the overall performance of TransformerEncoder when padding rate is high. Default: True (enabled).

[saudet]

The C++ API doesn't have all the features of the Python API, that's probably normal.

[mullerhai]

The C++ API doesn't have all the features of the Python API, that's probably normal.

maybe need myself coding