Thomas Capelle

This is the way, glad you liked it!

it's a really nice place, I agree!

They basically jailbreak gpt-4o

Same vibes git commit -m "pbar"

happy to take a look on a call =)

Can you share a workspace?

This was a team effort from @morgymcg.bsky.social , Soumik, @parambharat.bsky.social , Agata Mlynarczyk, @ayshthkr.bsky.social and many others!

Local Weave Scorers | W&B Weave Weave's local scorers are a suite of small language models that run locally on your machine with minimal latency. These models evaluate the safety and quality of your AI system’s inputs, context, and ...

I'm excited to see how the community uses these tools, and I'm looking forward to more innovations in safe and reproducible AI!

Check the scorers and Weave here:

👉 wandb.me/weave_scorers

📚 A colab: wandb.me/scorers_colab

A personal highlight was working on the Fluency Scorer powered by AnswerDotAI ModernBERT-base; we hope to move all DeBerta-powered scorers to ModernBert in the next release so we can benefit from the longer context length and training speed!

As part of this initiative, we also created comprehensive evaluation datasets, drawing on invaluable contributions from the open-source community. Being a reproducibility-first company, we’ve made the full recipe public, including the scorers, model weights, and the training and evaluation datasets

We designed these non-LLM powered scorers to leverage state-of-the-art open source models – from the PleIAI/Celadon toxicity detector to the Vectara hallucination scorer – ensuring that our AI systems are evaluated across multiple dimensions.

Local Weave Scorers | W&B Weave Weave's local scorers are a suite of small language models that run locally on your machine with minimal latency. These models evaluate the safety and quality of your AI system’s inputs, context, and ...

Over the past few months, my team at Weights & Biases has been hard at work launching Weave Scorers and guardrails.

wandb.me/weave_scorers

👇

We are cooking here...

Same vibes, PR submitted, PR merged.

- new MacBook pro 😍
- french keyboard layout 😭

Many people have asked me about the France Action Summit.

I think a summit is typically most valuable as a catalyst, not as a solution in itself.

But, will share some observations.

It could have been called Gulf of North America

This is my favorite kind of Yoga

I just built a CI to run an Eval of some custom LLM scorers on top of @modal-labs.bsky.social
- Great to test against different GPUs
- No custom runner neded on github
- Fast and nice console outputs =)

Butternut Soup and kimchi side

Échalotes, tomates sèches et moules.

Samedi tu fais moules et frites, dimanche tu finis les moules dans un rissoto aux moules.

Why is the USA the only country in the world with bird flu H5N1 ripping through cattle herds?

Because in the United States, it’s legal to feed chicken shit to cattle.

That’s why. That’s literally the reason

www.telegraph.co.uk/global-healt...

Pancakes morning with the arrival of the @vendeeglobe.bsky.social

C'est super bon ça !

Don't miss Stacey in Paris!

We raised this internally! thanks for the info.

This budget forcing is really smart. We could do that we prefill on API models no?

This is getting out of hands...