AI-Code-Mastery (Episode 5): Zero-Shot document question answering with Flan-ULv2

[@cinematheque20]

Thanks for uploading this. It is quite well explained.
It looks like you are running on this your local machine. What is the computation power of your machine? I am getting a time out error (Error raised by inference API: Model google/flan-ul2 time out) when I run the last cell in the notebook. Any idea about this?

[@amriteshmadhur]

Nice Video
There is no repository link in the description , if you could please add it will be much helpful Cheers and Thanks

[@AIology2022]

Note: Token is restarted. So I did not mask it :)

[@yashvirsangha9757]

Amazing video!

[@WillMcCartneyAI]

nice video man

[@IshmeetSinghahuja]

awesome can you link your notebook?

[@wilfredomartel7781]

Huggingface by default use "sentence-transformers/all-MiniLM-L6-v2". If you have non english language this must be changed.

[@wilfredomartel7781]

Excellent!

[@wilfredomartel7781]

I have a doubt. How does lang change generate the embeddings ? Or it need to connect to openAi to generate embeddings with the free token? Please explain me this part. Thank you

[@AIology2022]

We are using flan-ul2 from huggingface for tokenization and inference. If you replace that we OpenAI(), it should still work.

[@Ruzzeem]

great !

[@jackmartin1146]

Can you share the code/jupyter notebook?

[@AIology2022]

Thanks! Just added to description. I will update that repo with new methods soon :)

[@jackmartin1146]

@@AIology2022 Don't see the Colab or Jupyter notebook/Github in the description, planning to share it later?

[@yashvirsangha9757]

@@AIology2022 Hey, I don't think the repo/notebook got added to the description, would really appreciate the code, thank you!

[@wilfredomartel7781]

Can you share your colab or any github bro?

[@AIology2022]

Thanks! Just added to description. I will update that repo with new methods soon :)

[@mahdirasouli6005]

آقا خسته نباشید. اون وسط ها چندتا کات می‌دادید بهتر بود