-
Hi @spro, i've read your implementation of luong attention in pytorch seq2seq translation tutorial and in the context calculation step, you're using rnn_output as input when calculating attn_weights …
-
Description:
We are currently enabling the multi-node for mxnet sockeye and found that currently if the normalization type is valid the loss normalizer for softmax is not correct in distributed train…
-
## Issue description
I was testing the difference between LSTM and LSTMCell implementations, ideally for same input they should have same outputs, but the outputs are different, looks like something …
-
Dear all,
I'm living the dream ever since I discovered arraymancer, really good work, thank you.
I think ONNX and/or PMML support would be a very good addition to the library and will accelerate t…
-
Since issue https://github.com/Theano/Theano/issues/5679, we make Theano compile with it. We should do:
- [ ] Make a todo list of all the new features we need to wrap, to help do the follow up.
- …
nouiz updated
7 years ago
-
The docstring of `TransformerSenderReinforce` mentions that the `max_len` parameter includes the EOS token: https://github.com/facebookresearch/EGG/blob/main/egg/core/reinforce_wrappers.py#L687
How…
-
Hi,
I am facing challenges in explaining how unixcoder/clone-detection/java model makes its predictions.
I would appreciate any guidance or resources on how to perform model explainability for t…
-
Minor detail that has been bothering me:
- `plasma-python` GitHub repository name
- `plasma` Python module name
- `PPPLDeepLearning` GitHub Organization name
- "PPPL deep learning disruption predi…
-
Once we have an implementation of the Layer Class https://github.com/arrayfire/arrayfire_ml/issues/17 , the Optimizer class and the DataSet class we can go about creating RNN flavors. There are 3 mode…
-
On a high level this is an interface that would be nice to have for rnns:
```
(weights, state) = initrnn(input)
(output,state) = rnn(weights,input,state)
````
state can be used to encapsulate var…