-
Look at different methods of search.
- https://github.com/google-research/google-research/tree/master/scann
- https://github.com/facebookresearch/faiss
- https://github.com/nmslib/hnswlib
- http…
-
Update the beam search to not allow beams with the same caption to exist (if a beam is encountered with the same caption and higher log prob, replace the existing beam with that caption.)
-
Without a decoding method one cannot actually uses the trained network to translate... The greedy decoding requires a very good network.
-
Hi,
Thank you providing this code.
Could you explain your code for backtracking in beam search.
In particular how do you handle the dropped sequences that have seen EOS earlier during forwa…
-
As beam search is widely used in machine translation, could you support beam search sampling and evaluate the model?
-
Though I can find you have implemented this function in your another repository 'video-caption-openNMT.pytorch', it's hard to comprehend it. Would you please make it available in this repository? Than…
-
A next step for better generation is to implement a beam search for the generation. An example of it can be seen on the huggingface repo [here](https://github.com/huggingface/transformers/blob/9911729…
-
In most text generating architecture, beam search provide a quality improvement by generating more natural text.
**Is it useful to use beam search with XLNet ?**
---
As far as I understand, s…
-
Hi,
I have conducted experiments based on the InterCTC model on the AISHELL-1 dataset, with auxiliary CTC losses calculated at layers [3, 6, 9, 12, 15]. I currently have the following three questi…
-
[1] https://arxiv.org/pdf/1904.04479.pdf
[2] https://arxiv.org/pdf/1712.09444.pdf
[3] https://arxiv.org/pdf/1702.01806.pdf
[4] https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.75.6306&…