tiagopavan

3 ms·

1. 1.▲
   by tiagopavan 1y ago

   dspy approach seems rather similar to that: https://dspy.ai/tutorials/gepa_ai_program/
2. 2.▲
   by tiagopavan 3y ago

   That's a great outcome. I was the producer of "The Edge of Democracy" - line 14047 of the Excel file with 200,000 hours viewed. Although we were nominated for an Oscar and became a Netflix originals, they've never disclo
3. 3.▲
   AI Is Already a Commodity(twitter.com)

   23 points by tiagopavan 3y ago|6 comments
4. 4.▲
   by tiagopavan 4y ago

   There's an interesting stanford webinar covering this topic. It's what they call "Retrieval augmented in-context learning": https://www.youtube.com/watch?v=-lnHHWRCDGk
5. 5.▲
   by tiagopavan 4y ago

   Thank you for your comment! I hadn't come across Qdrant before, but I will definitely check it out. Although I've been experimenting with milvus.io lately, these vector databases weren't on my radar when I first started explo
6. 6.▲
   From Zendesk to ChatGPT: Leveraging Embeddings(colab.research.google.com)

   3 points by tiagopavan 4y ago|3 comments
7. 7.▲
   by tiagopavan 4y ago

   Since the release of the gpt-3.5-turbo model by OpenAI, I have been wondering how companies are dealing with the token limit to gather insights from a large corpus of text without having to retrain the model. To understand how it works, I c