Update requirements.txt

Update README.md

Use 10 beams for decoding

Switch to 3 beams for faster inference

Update title

Update title

Change title

Change title to Automated Counterspeech

Remove r-word example

Move system description to below

Move system description to the bottom

Update examples

Update disabled hate speech example

Update examples to be more diverse

Update example

Update model to use doubly-tuned DGPT on Gab then MultiCONAN

Switch to 5-gram blocking

Change to minimum response length of 20

Minor wording change and link to counterspeech definition

Change wording of hate speech input

Update disclaimer to say researchers not developers

Remove quotes around submit

Improve disclaimer

Add 'click submit'

Don't cache examples

Add disclaimer note about inappropriate responses

Add examples

Change link to MLMI only

Add link to MLMI page

Fix spacing

Update title

Update description with details about the project

Change wording in description

Change wording in description

Change wording in description

Add title, description and article

Add app using dialoGPT-finetuned beam10 no-minlen

initial commit