Aim is a best-in-class AI metadata tracking tool. Experiment tracking, AI Agents tracing repo: https://nitter.cf/t.co/aP02BhLUuc discord: https://nitter.cf/t.co/CHb7o635Ve
Berkeley, CA
Joined October 2020
- Tweets506
- Following37
- Followers909
- Likes128
AimStack retweeted
What I love most about @aimstackio Reports:
Running snippets and getting charts in under 2 seconds, add my analyses, conclusions
- all in one place.
It’s a major upgrade from switching between training runs, Google Docs, exporting charts, and taking endless screenshots.
AimStack retweeted
My personal preference for experiment tracking is also open-source:
github.com/aimhubio/aim
AimStack retweeted
Aim was one of the first observability platforms we ever supported! You can find our docs on how to use it (it's really simple!) here: docs.llamaindex.ai/en/stable…
👀 Imagine being able to see everything that happens in your @llama_index environment, from the moment you input a query to the final response.
🔎LlamaIndex Observer is one of the logging apps built in AimOS. It logs metadata like retrieval nodes, queries and responses, embeddings chunks, and provides end-to-end overview of prompts, agent traces.
🧵Now, let’s see how that works:
AimStack retweeted
why is nobody talking about AIM as their experiment tracking platform? Fully OS weights and biases alternative, super fast.
AimStack retweeted
Does there exist an experiment-tracking solution that would read your mind? 😂
I am getting the weirdest experience with @aimstackio
I think of something that I need and with only a couple of clicks it is there 🤯
+ 0 waiting for stuff to load
I can't believe this is real 😁
AimStack retweeted
What if there existed a way to track your ML experiments that would:
• be super easy to integrate into your code
• didn't hit you with a ton of useless data just for fun
• ran locally
• was blazingly fast
@aimstackio, I am in love 🥰🥰🥰
AimStack retweeted
Aim is objectively the best experiment tracker. We have been using it for the last year or so with fantastic success.
Would highly recommend for everyone to try.
LLMs are transforming industries! Training LLMs is extremely challenging.
Here is a behind-the-scenes of one of the first LLMs, @BigscienceW Bloom, with @aimstackio . 🧵
AimStack retweeted
🤙 Callbacks
Callbacks are intercepters between each step in the chain, allowing for a variety of logging observability
Top integrations include @arizeai @WhyLabs @argilla_io @promptlayer @streamlit @aimstackio @Cometml @itsArthurAI
Page: integrations.langchain.com/c…
AimStack retweeted
And we're live!! 🚀
FastTrackML says Hello World!
fasttrackml.io/
#experimenttracking #machinelearning
@aimstackio @MLflow
AimStack retweeted
I've prompted the agent to:
1️⃣ List 5 supernova facts
2️⃣ Find the latest explosion
3️⃣ Calculate years since the last 🌠
🔗Agent built with @langchain 🦜, executions traced by @aimstackio ✨
🔍Observability for LLMs: tracking inputs,outputs, agents' actions, and chains results
AimStack retweeted
LlamaIndex now integrates with @aimstackio! 🔌✨
@aimstackio is an awesome tool for tracking metadata for your LLM app (e.g. experiments 🧪, prompts 📝, and more).
It’s the perfect tool for tracing within LlamaIndex, which manages multiple LLM calls on top of your data.
AimStack retweeted
Unlock the full potential of your AI apps with @langchain – now supercharged with open-source AI metadata tracking from @aimstackio! Experience ultimate LLM composability – check out the demo here: github.com/hwchase17/langcha…
🎯 AIM Tracking
@aimstackio is an OS framework for logging runs and any AI Metadata, and give you a beautiful UI to compare & observe them
Thanks to @tmynn__ and @gor__ar you can now easily use this in LangChain
Docs: python.langchain.com/en/late…
AimStack retweeted
🙌 This is a fun and useful tutorial to go through, I highly recommend it ↓
🚀 Building Multi task AI agent with @langchain and using @aimstackio to trace and visualize the executions
Let’s play..
We’ll prompt the agent to discover who Leonardo DiCaprio’s girlfriend is and calculate her current age raised to the power of 0.43🧵
1/7
AimStack retweeted
AI progress has skyrocketed with the release of #ChatGPT 🚀
As AI systems get complex, the ability to effectively debug and monitor them becomes crucial.
Check out my latest article on how to use @aimstackio to monitor AI systems built with @langchain
bit.ly/3Kk7Xwl
AimStack retweeted
Now you can trace and visualize your @langchain chains with @aimstackio Aim! 🎇
It takes just a few lines to configure the Aim callback - all your inputs and outputs will be automagically captured and visualized.
#LLMs #GPT4 #ChatGPT
🎯 AIM Tracking
@aimstackio is an OS framework for logging runs and any AI Metadata, and give you a beautiful UI to compare & observe them
Thanks to @tmynn__ and @gor__ar you can now easily use this in LangChain
Docs: python.langchain.com/en/late…
AimStack retweeted
Major thanks to @mattierialgirl for helping integrate @aimstackio Aim with the mighty lightning-hydra-template repo.
PR: github.com/ashleve/lightning…
Aim: github.com/aimhubio/aim
AimStack retweeted
Check out my latest blogpost on how to combine @huggingface Datasets and @aimstackio Aim together to make experiments traceable, reproducible, and easier to compare!
bit.ly/3YQy8jE
Aim: github.com/aimhubio/aim
Datasets: github.com/huggingface/datas…
#MachineLearning #MLOps
AimStack retweeted
Excited to announce Aim v3.16 is out! 🚀
Highlights:
- Run messages tab in UI
- Real-time sync with TensorBoard logs
- Support for Python 3.11
- Integrations with @huggingface Datasets, @DeepMind acme, #StableBaselines3
The blog: bit.ly/3Ea4eiv
AimStack retweeted
“Exploring MLflow experiments with a powerful UI” by Gor Arakelyan
link.medium.com/MweQVqeqUwb
AimStack retweeted
@aimstackio helps conduct large-scale trainings (such as #OPT, #GPT), saving human-hours spent on monitoring runs and compute hours.
A quote from “Scaling Laws for Generative Mixed-Modal Language Models” (Aghajanyan et al., 2023, arxiv.org/abs/2301.03728).
#MLOps #AI #ML
