Good morning fellow cerebrals 🧠!
Are you curious about tools you can create with small language models, running only on CPUs?
Check out Bullet Points, a project I just launched! Bullet Points is your research assistant for any topic in the news. Stay on top of key updates for your industry niche, political issues, research topics, and more. You can subscribe to daily or weekly briefs for any topic, delivered to your inbox. Bullet Points scans thousands of news sources 24/7.
🛠️ For those interested in implementation, I built this project with open-source models only:
• Instructor for embeddings (https://huggingface.co/hkunlp/instructor-large)
• BART, fine-tuned on the CNN Daily Mail dataset, for summarization (https://huggingface.co/facebook/bart-large-cnn)
• FLAN T-5 for curation (https://huggingface.co/google/flan-t5-base)
And I’m using pgvector for embedding search.
If you like this project, try subscribing to a brief and please consider upvoting it on futuretools.io! 🙌 (this is the AI tool directory that’s driven the most traffic for me, highly recommend!)