Uploaded February 2020 | Updated September 2026, 5 hours ago
In this video we work on the first part of making an AI that can answer questions. We use DistilBERT, transformer based NLP model pretrained on the Stanford SQUAD dataset to answer questions. In part 2 we work on all the functions to read and process text from online to use as our context. We then finish the project and try it out with a few questions.
Relevant Links:
GitHub code: github.com/ejmejm/AskMeAnything
Google Colab: colab.research.google.com/drive/1DbOL_kNsMct1_9qCrMfnaiyTDGNfBkq3
Simple Transformers: github.com/ThilinaRajapakse/simpletransformers
DistilBERT Paper: arxiv.org/abs/1910.01108
In this video we work on the first part of making an AI that can answer questions. We use DistilBERT, transformer based NLP model pretrained on the Stanford SQUAD dataset to answer questions. In part 2 we work on all the functions to read and process text from online to use as our context. We then finish the project and try it out with a few questions.
Relevant Links:
GitHub code: github.com/ejmejm/AskMeAnything
Google Colab: colab.research.google.com/drive/1DbOL_kNsMct1_9qCrMfnaiyTDGNfBkq3
Simple Transformers: github.com/ThilinaRajapakse/simpletransformers
DistilBERT Paper: arxiv.org/abs/1910.01108
