Arize AI Phoenix: Open-Source Tracing & Evaluation for AI (LLM/RAG/Agent)
HTML-код
- Опубликовано: 22 авг 2024
- Welcome to my tutorial on using Phoenix by Arize AI, the open-source AI observability platform that's revolutionizing experimentation, evaluation, and troubleshooting in AI applications. In this video, I’ll walk you through the powerful features of Phoenix, including tracing, evaluation, and inference analysis.
What You'll Learn:
1. Tracing: Understand how to trace your LLM application’s runtime using OpenTelemetry-based instrumentation.
2. Evaluation: Discover how to leverage LLMs to benchmark your application’s performance using response and retrieval evaluations.
3. Inference Analysis: Learn to visualize inferences and embeddings with dimensionality reduction and clustering to identify drift and performance degradation.
Why Phoenix?
Phoenix is vendor and language agnostic with out-of-the-box support for popular frameworks like 🦙 LlamaIndex, 🦜⛓ LangChain, and 🧩 DSPy, and LLM providers like OpenAI and Bedrock. Whether you’re working on a Jupyter notebook, local machine, containerized deployment, or in the cloud, Phoenix has you covered.
Don’t forget to like, comment, and subscribe for more tutorials on AI and machine learning tools. Your support helps me create more content to help you on your AI journey!
🔔 Hit the bell icon to get notified whenever I post a new video.
Join Discord: / discord
Join this channel to get access to perks:
/ @aianytime
To further support the channel, you can contribute via the following methods:
Bitcoin Address: 32zhmo5T9jvu8gJDGW3LTuKBM1KPMHoCsW
UPI: sonu1000raw@ybl
GitHub: github.com/AIA...
#arize #ai #aiagents
Nxt level
Thank you for your wonderful efforts Mr. Sonu ❤
Evaluation all you need, thanks for sharing
Glad it was helpful!
@@AIAnytime I have sent you a connection request on LinkedIn please accept, I don't have any connection note left, thanks
Nice...............Evaluation is highly needed
Just tried Langfuse... run standalone in a Docker container...not in detail though. This looks cool with the 'evaluator' for QA correctness and hallucinations. Tried with mini as the eval model... think it missed one... 4o seems good. But the groq llama 70B (or the faiss embeddings and retreiver) seems iffy. Come on llama 400B next week!... and the groq version since i don't have a 1TB GPU ;)
Wow this is what I need
Hi Sonu. a good tutorial like always. In your opinion which is the best LLM evaluator?