Fine-Tune Transformer Models For Question Answering On Custom Data

📹✨ Fine-Tune Transformer Models For Question Answering On Custom Data! 🤖💡
In this exciting tutorial, we'll dive into the world of fine-tuning the Hugging Face RoBERTa QA Model on your very own custom data. 🚀✨ Get ready for significant performance boosts and unleash the power of transformers!
📚 Learn step-by-step how to fine-tune transformer-based models on your unique data, unlocking their full potential. 📈💪 Whether you're a data scientist, researcher, or just curious about NLP, this video is for you!
🔗 Check out the blog post for more insights and details: 👉 towardsdatasci...
🔥 Don't miss out on the opportunity to explore the code on GitHub: 👉 github.com/ska...
🗂️ Want to experiment with the SubjQA dataset? Here's the link: 👉 github.com/meg...
🔔🎥 Don't forget to like, comment, and subscribe for more incredible AI tutorials! Let's embark on this transformative journey together! 🌟🤖

[@gursehajbirsingh8956]

You haven't exaplined the code what is the preprocessing code doing how is it working ? Its a request to make a new video on proper explanaation of the code.

[@betsonwilliam9470]

Thankyou sir.Please contribute more

[@scienceineverydaylife3596]

Thanks - definitely, stay tuned for more videos on applying AI, data science to real-world cases :)

[@mrsilver8151]

hi sir thanks for your great work, my question is , is this considered as generative Q&A or extractive Q&A.

[@DescolaDev]

Extremely helpful! Thanks

[@scienceineverydaylife3596]

Glad it was helpful!

[@felipeekeziarosa4270]

i'm trying a qa in portuguese, but the models I found on HF just returns short answers. Would you suggets any model?

[@kiranhipparagi543]

If I want to fine tune for RFP’s document, what is the best way? Please help.

[@pablogonzalezrobles803]

same concept in sage maker please

[@shivacharan7431]

I'm currently trying to use some random Wikipedia raw data in any context and I want my model to train on it and generate answers as per the questions on the same context, can you suggest me...what model to use..

[@scienceineverydaylife3596]

Try BERT or distilBERT

[@williamguan5962]

Thanks for your video. But I am curious what’s such training meaning in comparing of using ChatGPT with contextual prompts if we ignore the openai api cost.

[@scienceineverydaylife3596]

Good question! On smaller tasks, fine-tuned transformers may perform almost as good as ChatGPT - but cheaper, faster, and more private (hosted on a private cloud instance). But smaller models would perform poorly in generative applications like chat bots. It's a question of evaluating choices based on needs.

[@terrencegeorge3730]

👉 Promo SM

[@echofloripa]

How do I do it if I just have questions and answer, without the context?

[@scienceineverydaylife3596]

In this case there is no context to "extract" from - so it becomes generative Q&A - which is a different beast to tackle. You might look at instruction fine-tuning/RLHF for that

[@echofloripa]

@@scienceineverydaylife3596 I'm doing a prototype with the input to the fine-tuning as: ### Category: FEDERAÇÕES ### Question: 5. - Como é feito o registro de candidaturas de partidos que integrem federações? ### Answer: 5. - Como a federação age como se fosse um só partido, o número de candidatos a ser indicados por uma federação é igual ao número de um partido isolado. Por exemplo, se o número de candidatos por partido for 10 serão também indicados...