Harsh Trivedi
Harsh Trivedi
@thomwolf Can you take a look at this PR? I am trying to use `neuralcoref` on a large corpus but w/o GPU it's taking too much of time.
I could get it to work on GPU using this fix, thanks @dirkgr!
Hi Chandresh, It's fixed now. Thanks for letting me know!
Hi Chandresh, Thanks for letting me know. It's fixed now. Please let me know if you find any more issue! Harsh On Thu, Jul 27, 2017 at 2:30 AM, Chandresh...
Tag auto-completion in the connector would be very useful.
+1 to this feature request.
Sorry for the extremely late response here. No, the models are not compatible with HF. They were made using AllenNLP library.
The contexts/paragraphs were taken from the original source datasets. However, I did apply [ftfy](https://ftfy.readthedocs.io/en/latest/) at runtime. See `commaqa/inference/dataset_readers.py` for example. You might want to give it a try.
I am not sure I understand the question. Can you please elaborate?
Hi @RufengSama, thank you for your interest! > What is the value of max_interactions that you suggest? Around 50 should be enough. Our thresholds were higher as we wanted to...