From the course: The AI Ecosystem for Developers: Models, Datasets, and APIs

Unlock this course with a free trial

Join today to access over 24,600 courses taught by industry experts.

NLP architectures: RNNs and transformers

NLP architectures: RNNs and transformers

From the course: The AI Ecosystem for Developers: Models, Datasets, and APIs

Start my 1-month free trial Buy for my team

NLP architectures: RNNs and transformers

“

- [Instructor] Natural language processing, NLP, is the subfield of AI focused on sequences like text and speech. AI architecture in this domain are designed to model sequential dependencies, making them essential for tasks like language modeling, translation, and text generation. Some of the most popular model architecture used in NLP include recurrent neural networks, RNNs. RNNs are a class of neural networks designed for sequential data, where previous inputs influence future predictions. The components of RNNs include the input layer, which processes sequences like words or full names. Hidden state stores memory of previous inputs, allowing the model to retain context. Output layer produces predictions such as the next word in a sentence. Recurrent connections connects the hidden layer to itself, allowing information to persist across time steps. RNNs are used for language modeling, predicting the next word in a sequence, machine translation, translating text from one language to…

Contents