Kevin Lee

32 posts

Kevin Lee banner
Kevin Lee

Kevin Lee

@KLee4422

Member of Recruiting Staff at @LiquidAI my opinions are my own

Katılım Mart 2024
231 Takip Edilen123 Takipçiler
Kevin Lee
Kevin Lee@KLee4422·
Check out our newest Encoder Models! These allow users to do zero-shot prompt routing. Try it for yourself on Hugging Face 💧
Liquid AI@liquidai

Today we release LFM2.5-Encoder-230M and LFM2.5-Encoder-350M: bidirectional encoders that stay fast at long context, even on CPU. > LFM2.5-Encoder-230M: about 3.7x faster than ModernBERT-base on CPU at 8,192 tokens. Under 30s per forward pass, versus over a minute and a half. > LFM2.5-Encoder-350M: 4th of 14 models on GLUE, SuperGLUE, and multilingual classification, behind only three larger models, one of them nearly 10x its size. 🧵

English
0
2
12
1.2K
Kevin Lee retweetledi
Liquid AI
Liquid AI@liquidai·
Today we release LFM2.5-Encoder-230M and LFM2.5-Encoder-350M: bidirectional encoders that stay fast at long context, even on CPU. > LFM2.5-Encoder-230M: about 3.7x faster than ModernBERT-base on CPU at 8,192 tokens. Under 30s per forward pass, versus over a minute and a half. > LFM2.5-Encoder-350M: 4th of 14 models on GLUE, SuperGLUE, and multilingual classification, behind only three larger models, one of them nearly 10x its size. 🧵
English
34
111
912
189.5K
Kevin Lee retweetledi
Ramin
Ramin@ramin_m_h·
fireside chat Ion Stoica (@istoica05) and I had with Bloomberg's @rachelmetz on open source AI, and what comes after today's static models, and how we measure whether AI is actually getting better! This was the opening panel at the @RaiseSummit about 3 weeks ago, in Paris. enjoy! youtu.be/mxFmC5-SFWc?is…
YouTube video
YouTube
English
1
5
22
3K
Kevin Lee retweetledi
Liquid AI
Liquid AI@liquidai·
In strong support of open-weight AI and American AI leadership, we are proud contributors to the open-source community and excited to announce that Liquid Foundation Models (LFMs) have surpassed 40 million downloads by the community! Going forward, we remain committed to accelerating the open-weight release of the next generation of lightweight, powerful LFMs to the world. excited to see what you build with them! liquid.ai/models
Liquid AI tweet media
San Francisco, CA 🇺🇸 English
23
68
685
40K
Kevin Lee retweetledi
Ramin
Ramin@ramin_m_h·
Open source is core to what we do at Liquid. If you zoom out a bit, as humanity, how else can we accelerate our understanding of the universe, solving our thoughest challenges, and satisfy our own curiocity, while enjoying the hell out of it? I do not know of any other way. Also, cogratulations to my unparalleled team at @liquidai. Achiving this degree of distribution of value (1.4M downloads/week), with 1000x less resources compared to other frontier labs is an insane accomplishment! very well done So proud and humbled by this
Liquid AI@liquidai

In strong support of open-weight AI and American AI leadership, we are proud contributors to the open-source community and excited to announce that Liquid Foundation Models (LFMs) have surpassed 40 million downloads by the community! Going forward, we remain committed to accelerating the open-weight release of the next generation of lightweight, powerful LFMs to the world. excited to see what you build with them! liquid.ai/models

San Francisco, CA 🇺🇸 English
4
12
61
5K
Kyle Wong
Kyle Wong@ewveggies·
Yesterday was my 23rd birthday. As a present, I got laid off from Amazon AGI! I’m actively looking for new opportunities, particularly related to computer use agents and post-training. Since October 2024, I've been working on GUI agents non-stop: computer use at @SimularAI, mobile use @Apple, and browser use @AmazonScience. With recent benchmarks like OSWorld 2.0, ALE, etc., it's clear CUA is far from solved, and would love to continue pushing. Please reach out if you have any openings!
Polymarket@Polymarket

JUST IN: Amazon cuts jobs from its artificial general intelligence team.

English
76
17
961
194.1K
Yuxin Tang 🦉
Yuxin Tang 🦉@yuxin_tang_·
Looking for new oppotunities! Most members in our MoE pre-training team got impacted today 😅 Feel free to ask me about what happened inside of Amazon AGI foundation team and I will share my brief thoughts with my tenure period. Also, if you know there is any place that needs people to help/optimize LLM pretraining, please share that with me! I am open to new opportunities in the next few month! #amazon #agi #moe #llm #pretraining
English
47
39
849
410.2K
Kevin Lee
Kevin Lee@KLee4422·
@bhaumikk9 Hey Bhaumik! So nice to hear from you. I’ll message you on LinkedIn to set up a time to chat.
English
1
0
2
229
Kevin Lee
Kevin Lee@KLee4422·
So sad to hear about the layoffs of the Amazon AGI team. I personally know several people over there who are quite brilliant. If anyone there is looking for their next adventure, my teams at Liquid AI are currently hiring. Please feel free to reach out via DM or on LinkedIn (link in my profile) 🙏
English
3
11
42
4.6K
Kevin Lee
Kevin Lee@KLee4422·
@miao_xiong_cs Hi Miao! Our Pre-Training Data team at Liquid AI is currently hiring. Just sent you a DM
English
0
0
5
1.2K
Miao Xiong
Miao Xiong@miao_xiong_cs·
I was laid off by Amazon AGI today, along with many of my colleagues. My job was deciding which data points matter for pretraining. Turns out I was the one that got filtered out. At 1pm, the reminder for our weekly project sync still went off — for a project that no longer has a team. We could only exchange a smile that needed no words. More seriously: everything I've worked on has been about data, data curation pipeline, data reordering, and data uncertainty. I’m actively looking for new opportunities, including pretraining, trustworthy LLMs, safety, and LLM efficiency — please reach out if you have any openings! To my manager and teammates: thank you for all the help and support you've given me along the way — I learned so much, and I loved the work we did together. It was a great journey, and I know even better ones are ahead for each of you.
English
145
130
2.6K
478.8K
Kevin Lee retweetledi
Jimmy Smith
Jimmy Smith@jimmysmith1919·
New report + blog: we increased LFM2.5-8B-A1B's tokenizer vocab from 65K to 128K to fix languages the original split too finely. Result: 2.2–3.7× faster per-character decoding on-device for Hindi, Thai, & Vietnamese. Report: arxiv.org/abs/2607.15232 Blog: liquid.ai/blog/tokenizer…
Liquid AI@liquidai

We doubled LFM2.5-8B-A1B's tokenizer from 65K to 128K to fix the languages it split too finely. Today we're sharing the recipe for upgrading a pretrained model's tokenizer in place. > Thai now takes 4.0× fewer tokens, Vietnamese 2.6×, Hindi 2.4× > Est. 2.2 to 3.7× faster per-character decoding on-device for these languages 🧵

English
0
5
24
3.5K
Kevin Lee retweetledi
Mathias Lechner
Mathias Lechner@mlech26l·
For LFM2.5-8B we expanded our tokenizer in-place. So we don't have to pretrain a new model but have much better non-English language support. Why not a fat tokenizer (>200k vocab) in the first place? -> because it tanks decode speed on edge devices
Mathias Lechner tweet media
English
1
7
28
1.8K
Kevin Lee retweetledi
Dave Blundin
Dave Blundin@DaveBlundin·
Incredible to have Co-Founder and CEO of @liquidai, Ramin Hasani (@ramin_m_h) on the pod last week! Liquid is the only foundation model company, that I know of, that reached unicorn status after coming out of @MIT. They're building post-transformer AI, with their roots right here in Cambridge! So, @miramurati just shipped Inkling, a 975B open-weight model. @OpenAI started open source and went closed. @Meta did the same with Llama. If you throw your model open source, what's your revenue model? We’re about to find out! Also, @demishassabis proposed a FINRA-style regulator for frontier AI. The problem is that nobody at the top labs is going to carve out years of their life to serve in a regulatory body. And FINRA has no mechanism for the real unlock: AI regulating AI. The only we’ll be able to keep up.
English
5
9
55
2.7K

Keşfet

@istoica05 @rachelmetz @RaiseSummit @k44yej @liquidai @ewveggies @SimularAI @Apple