GPT-4 vs Open Source LLMs: Epic Rap Battles Test Creativity with AutoGen

Explore the creative capabilities of open-source language models in an epic rap battle showdown against GPT-4-Turbo. Witness Wizard LM 70B, Orca-2 13B, and Mistral 7B showcase their linguistic prowess in freestyle rap competitions. Delve into the performance analysis of each model against the advanced GPT-4-Turbo benchmark. Access complementary resources including a detailed blog post, GitHub repository with code, and links to related tutorials and courses. Learn about deploying open-source models using AutoGen and Runpod, and gain insights into vLLM documentation. Follow the structured chapters covering introduction, script run-through, individual rap battles, and concluding thoughts on the creative potential of large language models in this engaging 30-minute video from Data Centric.

Intro
Script Run-through
: Rap Battle - Mistral 7b vs GPT 4 -Turbo
Rap Battle - Orca 2 13B vs. GPT 4 Turbo
Rap Battle - Wizard 70 B vs. GPT 4 Turbo
- Final thoughts