NLP

Introduction Huggingface models can be quite large and require a lot of computational resources to train and run, which can be challenging for users who want to run these models on their local machines or in cloud-based environments. One solution to this problem is to use caching, which involves storing precomputed values so that they can be reused later without having to be recalculated. In the context of Hugging Face and transformer models, caching involves storing intermediate values that are generated during the processing of text data using a transformer model....

Introduction BERT (Bidirectional Encoder Representations from Transformers) and GPT (Generative Pre-trained Transformer) are both models for natural language processing (NLP) developed by Google and OpenAI respectively. The main difference between the two models lies in their approach to processing text. BERT is a bidirectional model, which means it processes text in both directions (from left to right and from right to left) to capture the context of the words. It is a transformer-based model that is pre-trained on a large corpus of text using two unsupervised learning tasks:...

Introduction Entity linking is a critical task in natural language processing that involves linking entities mentioned in text to their corresponding entries in a knowledge base. In this project, we implement a creative approach to entity linking that leverages the complementary capabilities of Semantic Role Labeling (SRL), Named Entity Recognition (NER), and BERT-based FAISS indexing. We use SRL and NER to extract relevant queries from the input text, which are then used to perform a search in our FAISS index....