#nlp #zeroshotlearning #huggingface
Hello, everyone!
In this video, I cover the very basics of Natural Language Processing, NLP, and zero-shot learning using Python.
I also introduce Hugging Face ([ Ссылка ]) and Flair, two popular platforms and packages, respectively, that aid you with text classification with zero-shot learning.
Lastly, I tell you a little bit about few-shot learning (an upcoming video), a natural evolution of zero-shot learning.
I have demonstrated zero-shot learning with the help of three of the most commonly downloaded transformer models from Hugging Face; one developed by the researchers at Facebook, one cross-encoder, and one evolution of the BART-large model trained on the MNLI dataset. We see that all of these models have excellent performance in a variety of classification scenarios. In the end, I compare the performance of these three models with the TARS classifier available in the Flair Python package.
