Generative Chatbot with tensorflow
summary
Shows how to build a generative chatbot using TensorFlow, Deep Learning and NLP by training an LSTM dual-encoder model on a movie-dialogues dataset. Covers text preprocessing steps, contrasts generative vs. retrieval chatbots (generative models produce replies word-by-word rather than selecting canned responses), and discusses adaptations to apply the same architecture to Portuguese datasets.
AI-written summary · about
chapters (13)
- 0:00Introduction and motivation
- 3:42Chatbot types
- 7:37Generative chatbots
- 12:32Recurrent neural networks
- 14:43Long Short-Term Memory (LSTM)
- 20:40LSTM encoder-decoder
- 23:49First generative chatbot
- 26:58Challenges and lessons
- 31:11Data preprocessing
- 35:24Chatbot architecture
- 38:13Training and results
- 43:09Improvements and next steps
- 45:00Closing
related talks
41mConversational Agents at Scale: Retrieval and Generative approaches
31mMachine Learning-Enabled Chatbots
24mA Journey through Generative AI's Frontier Python Libraries
33mBuilding the next-generation Conversational AI with Python and Deep Learning
30mConversational AI: Building clever chatbots
01h 38mConversational AI with Rasa Core & NLU
11mTalking to my past self (without introducing temporal paradoxes!)
53mGenerative AI: Overview and use cases