Explore >> Select a destination
|
You are here 
|
blog.lmorchard.com | ||
| | | | |
blog.miguelgrinberg.com
|
|
| | | | | miguelgrinberg.com | |
| | | | |
isthisit.nz
|
|
| | | | | August 2024 Update: Now a solved problem. Use Structured Outputs. Large language models (LLMs) return unstructured output. When we prompt them they respond with one large string. This is fine for applications such as ChatGPT, but in others where we want the LLM to return structured data such as lists or key value pairs, a parseable response is needed. In Building A ChatGPT-enhanced Python REPL I used a technique to prompt the LLM to return output in a text format I could parse. | |
| | | | |
blog.pythonanywhere.com
|
|
| | | | | [AI summary] The text describes a process of building an AI-powered website to answer questions about PythonAnywhere using Retrieval Augmented Generation (RAG). It explains how to: 1) Use OpenAI's API to get embeddings of help pages, 2) Find the most relevant help page for a user's query using cosine similarity, 3) Use the found help page as context for the AI to generate answers, and 4) Implement this in a Flask-based website. The text also discusses challenges like hallucination and suggests using GPT-4 for better results. | |
| | | | |
www.laptopmag.com
|
|
| | | OpenAI rolled out GPT-4o image generation this week, and it went viral, with people using the AI to generate Studio Ghibli-style animation. It went so viral that it started melting OpenAI's GPUs. | ||
