How to Make Friends and Influence Users with Huggingface
summary
Shows a working offline chatbot that combines Blenderbot-400M (distilled) for generation and distilbert sentiment via transformers pipeline for real-time sentiment detection, all run with PyTorch tokenization and generation calls. Adds offline speech with pyttsx3 and a Vosk fork for recognition so voice input/output happens locally. Code uses Hugging Face AutoTokenizer and pipeline, treats negative sentiment as a trigger to stop the session, and demonstrates a live dialogue where the system responds empathetically and the demo successfully runs.
AI-written summary · about
chapters (6)
- 0:00Start
- 2:13Using Pipelines from Transformers
- 3:30Using Blenderbot
- 4:46Adding Text to Speech & Speech Recognition to the chatbot
- 5:12Recognizing speech using the speech_recognition package.
- 5:35Code Walkthrough
related talks
28mUnlock the Power of LLM: Build a HuggingFace Agent
31mHugging Face ecosystem for Local AI/ ML
02h 10mIntro to Hugging Face: Fine-tuning BERT for NLP tasks
14mKeynote: Building AI with Open Source & Hugging Face
27mA primer to write a conversational chatbot
01h 09mReinventing Machine Learning with Transformers and Hugging Face
33mMachine Learning 2.0 with Hugging Face
19mBuilding Machine Learning Applications using Hugging Face