Vittorio Perera

NLP engineer by day, MTG player by night (and weekends)

[Vittorio Perera @erpasd]

In the past I posted about the work my colleague at @Tenyx_AI did on fine-tuning LLMs. Today we announce Tenyx Voice, check out https://t.co/ojQNM4j6zm for all the details!

[Robert Taylor 🔥 @fireshoes]

Current publicly disclosed Hall of Fame votes. We're at 9 out of 40 voters. 24 votes is needed to be inducted in the HoF. Let me know if I missed any or can't do basic math! 😆 #mtg #mtghof

[Matteo Silvestri @Silver22_2k]

Reaching San Francisco! 🇺🇸

❓️"How contamination can be detected within LLMs on tabular data?"
➡️ https://t.co/6w8exTdG7n

#colm26 #datacontamination #hercolelab

[Vittorio Perera @erpasd]

Proud to share EMBL LIBRARIAN, our first milestone at EMBL Rome and a huge shoutout to @luigi_sigillo who led the project from start to finish. Try it out, take it for a spin, and let us know what you think!

[Luigi Sigillo @luigi_sigillo]

🧬 Introducing 𝐄𝐌𝐁𝐋 𝐀𝐈 𝐋𝐈𝐁𝐑𝐀𝐑𝐈𝐀𝐍.
A knowledge layer for life-science AI agents.
Natural-language question in. Evidence snippet out.
No vector database. No separate literature index. Always up to date.
Open source ↓

[Luigi Sigillo @luigi_sigillo]

🧬 Introducing 𝐄𝐌𝐁𝐋 𝐀𝐈 𝐋𝐈𝐁𝐑𝐀𝐑𝐈𝐀𝐍.
A knowledge layer for life-science AI agents.
Natural-language question in. Evidence snippet out.
No vector database. No separate literature index. Always up to date.
Open source ↓

[Vittorio Perera @erpasd]

and if you are extra curious you can also check https://t.co/51QJ7WDBGI

[Vittorio Perera @erpasd]

In the past I posted about the work my colleague at @Tenyx_AI did on fine-tuning LLMs. Today we announce Tenyx Voice, check out https://t.co/ojQNM4j6zm for all the details!

[Tenyx @Tenyx_AI]

Today we announced Tenyx Voice to help enterprises combat #CX deficiencies by reducing costs, and improving customer satisfaction with accurate call routing, clear communication, low latency and continuously learning #AI. Read more from @craigss in @Forbes 👇 https://t.co/4gwE4RJf0g

[Vittorio Perera @erpasd]

and if you really like the model and want to deploy it, check out the tutorial from @Inferless_
https://t.co/7Y6FoJqzeN

[Inferless @Inferless_]

🎉 Exciting news from our Inferless community 🎉

🌟 We're thrilled to support our user @Tenyx_AI 's incredible success. Their Tenyx_Chat-7B-v1 model is now leading the MT-Bench Leaderboard, surpassing giants like Chat GPT! We are a proud compute partner!

🚀 To help you get started, we've created a deployment tutorial.

Detailed Tutorial Here: https://t.co/AjudP9hNgW
Hugging Face Model Link: https://t.co/3ngQjbX60W

[Vittorio Perera @erpasd]

At @Tenyx_AI we just released TenyxChat-7B-v1. It’s a finetuned LLM that outperform all the 7B models available on the Huggingface MT-bench leaderboard. To get the full picture check out https://t.co/Wx8HcDaIaI

[Vittorio Perera @erpasd]

@Tenyx_AI If you want to play around with the model you can use this huggingface space:
https://t.co/cVaaLUIa9E

[Tenyx @Tenyx_AI]

We are thrilled to share that, powered by our cutting-edge fine-tuning technology, #Tenyx has achieved a remarkable milestone: our open-source chat LLM (Tenyx_Chat-7B-v1) ranks as the top-performing small model on the MT-Bench Leaderboard outperforming much larger models like Chat GPT!

Check it out here: https://t.co/HhoTi6LBD4

[Vittorio Perera @erpasd]

At
@Tenyx_AI
we have been working on a solution to mitigate catastrophic forgetting when fine-tuning LLMs. Check out our blog post (https://t.co/chLOV2xjl9…) #NLP #finetuning #LLMs

[Tenyx @Tenyx_AI]

Today, we’re announcing our fine-tuning solution to address catastrophic forgetting in LLMs. From increased proficiency to stronger knowledge retention, enterprises can now harness the full potential of LLMs without compromising core capabilities via @VentureBeat: https://t.co/ycobcfkjai

[Billy Perrigo @billyperrigo]

🚨Exclusive: OpenAI used outsourced Kenyan workers earning less than $2 per hour to make ChatGPT less toxic, my investigation found

(Thread)

https://t.co/302G0z7vy3

[Vittorio Perera @erpasd]

@jacobandreas This was a really interesting read. Any chance you will release the dataset used in Section 2?

[Dileep George @dileeplearning]

You know English. You don’t know Malayalam. You decide to learn it by listening to podcasts that are purely in Malayalam. What will happen?

A thread about language models, meaning, understanding etc…. Plus, you get to learn Malayalam :-) 1/N

[Ben Phillips @benphillips76]

A clip from Don’t Look Up, and then a real TV interview that just happened

[Gavin Buckingham @DrGBuckingham]

Look closely.

There are actually 16 circles in this image.

And once you see them, the image appears changed forever.

The wonderful coffer illusion, by Anthony Norcia

[@emilymbender.bsky.social @emilymbender]

On what we should be demanding of journalistic coverage of so-called "AI" systems:

https://t.co/7YYD3QxI7R