-
I analyzed 23 pathways, including 156 reaction to reaction 'nextStep' links. The conversion process captures 120 (77%) of these links as causal relations, losing 36 that cross pathway boundaries.
…
goodb updated
4 years ago
-
For English, both [EWT](/UniversalDependencies/UD_English-EWT) and [GUM](/UniversalDependencies/UD_English-GUM) have many examples of copulas tagged as `VERB`.
- [EWT query](http://bionlp-www.utu.f…
-
**Describe the bug**
A very high CPU usage caused by activating only this ltex extension (caused by java-9-oracle)
**Steps to reproduce**
1. Open latex files
2. Activate ltex extension
3. Obs…
-
This is a suggestion, mostly because I can do it.
Instead of having 1 giant sentences.csv file that people need to frequently download to update, you can add git-tracking and split the sentences.cs…
-
The deepset.ai German pretrained BERT model with the German word pieces works well but has the common problem of all of the original transformer models in that it uses a lot of memory and is not too f…
-
The `try()` proposal #32437 has met with significant complaint, especially about
a) lack of support for decorating/wrapping returned errors
b) lack of support for widespread alternatives to returnin…
-
Link: https://www.aclweb.org/anthology/D19-1572.pdf
Short Description:
> The authors propose Multi-lingual language model Fine-Tuning (MultiFiT) to enable people to train and fine-tune languag…
-
Hello, first of all thank you for your contributions and sharing the code.
I'm having general issues running it, for the data, I'm using the development (`dv50_vcc2020_24kHz`) set to train (as I coul…
-
I don't believe the spec covers how to shape vowel/standalone/broken/symbol syllables.
What HarfBuzz [appears to do](https://github.com/harfbuzz/harfbuzz/blob/master/src/hb-ot-shape-complex-indic.h…
-
an example from the glosstag corpus:
> a prolonged dogfight over their rival bids for the contract
We don't have a sense for 'rival' as an adjective.