Explore a comprehensive video comparing open-source AI models with a custom web search agent against Perplexity AI. Dive into benchmarking tests of Mistral AI's Codestral 22B and Mixtral, as well as GPT4o, to evaluate their performance. Learn about the testing methodology, agent schema, and detailed results for each model. Gain insights into building your own local "perplexity" with Ollama and setting up Llama 70b. Access resources for AI development, including a hands-on RAG app project and updates on AI, Data Science, and Large Language Models. Follow along with timestamped chapters covering the introduction, test questions, agent schema, model testing, and result analysis for Mixtral, Codestral, and GPT4o.
Comparing Open Source AI Agents to Perplexity AI - Testing Codestral, GPT4o, and Mixtral
Data Centric via YouTube
Overview
Syllabus
Introduction:
Test Questions and approach:
Agent Schema:
Testing Mixtral:
Results Mixtral:
Testing Codestral:
Results Codestral:
Live run GPT4o:
Results GPT4o:
Taught by
Data Centric
Related Courses
-
Comparing LLAMA 3, Phi 3, and GPT-3.5 Turbo AI Agents for Web Search Performance
-
Deploying Mixtral 8X7B - An Open AI Agent for Advanced NLP Tasks
-
Comparing Ollama Local WebSearch Agent with Llama 3 8B to Perplexity AI
-
Mistral Large vs GPT-4: Practical Benchmarking and LLM Evaluation
-
AI Agent - Open-Source Perplexity Alternative Using LangGraph and Advanced Prompt Engineering
