Hugging Face
The AI community building the future.
Start new thread

Deep learning chat - Chatting with a deep learning chatbot

by•

Add a comment

Replies

Best
Seems human to me 🙃
Maker
😂 😂 😂
Maker
Thanks for hunting ! Guys, this is a demo of chatting with a Deep learning chatbot trained through Neuralconvo, a Torch library that implements Sequence to Sequence Learning with Neural Networks, reproducing the results in the Neural Conversational Model paper (aka the famous Google chatbot paper). The Google “Neural conversational model” chatbot was discussed at length by Wired, Motherboard and more, but there wasn't an easy way to actually try it online. Now you can 👋 Let us know what you think here in the comments!
Very solid technical explanations of the process by here “Chatting with a Deep learning brain”
This is not what I was expecting... haha :)
Likewise :) It be would nice to have a unified thread of conversation screenshots.
Maker
let's do a contest of the funniest screenshots hahaha
Great job here, congrats guys 👍 I've been testing it and it's been interesting and quite fun too. I'll be curious to follow the auto-improvement of it 👌 Would you keep it conversational only or is it possible to add any external integration like automating task management for example?
Maker
Thanks ! Our subject is conversation (we're building a fun messaging AI) so we'll prob focus on this.
Thanks for your answer , looking forward to seeing its future development 🤖
Even this chatbot flirts with me... 😂😂😂
Maker
I can literally spend hours playing with it. Would be perfect to be able to play with it offline in the subway. Great job, congrats guys !
Maker
Thanks Shubham. You'll love our real product then! It's 100 times better.
Maker
FYI, we've passed the 20,000 messages shared with the AI in just a few hours. Feel free to share your best screenshots!
This bot smoke too much weed 😂
I tried: - Because i've never been here before. - where have you been? - Oh, i'm not going anywhere. - but you do - Yes. - wtf - It's a secret between us
Maker
hahaha, the answers are definitely all over the place right now but it's interesting to be able to try it at large scale (15k messages already just in a few hours). We'll probably try with a different dataset.
haha it was already worth trying it out, I had fun with it even though it's a little dumb as of now 😀
12
Next
Last