Philippe Modard retweeted
Here's what I see as a likely AGI trajectory over the next decade.
I claim that later parts of the path present the biggest alignment risks/challenges. The alignment world has been focusing a lot on the lower left corner lately, which I'm worried is somewhat of a Maginot line.
Philippe Modard retweeted
We've released a Gemini 1.5 Pro API integration w/ Kaggle Models so now you can use it seamlessly in Notebooks 😃
Check out this starter notebook showcasing the 1M token context window + image understanding capabilities kaggle.com/code/prathameshba…
Philippe Modard retweeted
Very cool project -- a Gemma model fine-tuned to turn text into mind map summaries in dotfile format: kaggle.com/models/toshik/min…
Philippe Modard retweeted
We're announcing TacticAI: an AI assistant capable of offering insights to football experts on corner kicks. ⚽
Developed with @LFC, it can help teams sample alternative player setups to evaluate possible outcomes, and achieves state-of-the-art results. 🧵 dpmd.ai/49PGq1b
Philippe Modard retweeted
This is huge for us. Using Gemma-2B sliced our inference bill by 66% (we were using Mistral-7B). And a nice bonus that Gemma-2B model performance is also stronger.
Big kudos @SubhamDe2021 for figuring this out.
I've seen many reports of people struggling to finetune Gemma. We are using Alpaca-style instruction formats over conversational style and we are now getting superior performance out of Gemma-2b compared with Mistral-7b for our NER task (see screenshot to see our task).
The approach was inspired by @maximelabonne xitter.cf/maximelabonne/status/1…
Philippe Modard retweeted
New top-tier open model just dropped: Gemma.
ai.google.dev/gemma
2B and 7B sizes. Base model and instruction-tuned versions available for each.
Fast inference thanks to JAX backend in Keras 3. Built-in LoRA fine-tuning capability.
Philippe Modard retweeted
✨ We’re SO excited to welcome @Google’s Gemma to Kaggle!
Gemma is Google’s new family of OPEN models built from the same research and tech used to create Gemini. The 2B and 7B versions are available now on Kaggle w/ implementations in 8 different frameworks & amazing code samples.
Congratulations to the Gemma team, amazing work!
👉goo.gle/3UJ9fYf
Google@Google21 Feb 2024United States
United StatesConnected via WebAccount-level information, not a live location or per-post device.
Gemma is a new family of open models that help developers and researchers build AI. Along with the lightweight models, we’re launching tooling that encourages collaboration and a guide to responsible use of these models. Learn more → goo.gle/49IOArw
Philippe Modard retweeted

Google@Google8 Feb 2024
United States
United StatesConnected via WebAccount-level information, not a live location or per-post device.
Today, we’re entering the next chapter of our Gemini era by bringing our #GeminiAI models to more of our products, starting with Bard — which will now be called Gemini. goo.gle/4bxymmT
Philippe Modard retweeted
Ten months ago, we launched the Vesuvius Challenge to solve the ancient problem of the Herculaneum Papyri, a library of scrolls that were flash-fried by the eruption of Mount Vesuvius in 79 AD.
Today we are overjoyed to announce that our crazy project has succeeded. After 2000 years, we can finally read the scrolls:
This image was produced by @Youssef_M_Nader, @LukeFarritor, and @JuliSchillij, who have now won the Vesuvius Challenge Grand Prize of $700,000. Congratulations!!
These fifteen columns come from the very end of the first scroll we have been able to read and contain new text from the ancient world that has never been seen before. The author – probably Epicurean philosopher Philodemus – writes here about music, food, and how to enjoy life's pleasures. In the closing section, he throws shade at unnamed ideological adversaries – perhaps the stoics? – who "have nothing to say about pleasure, either in general or in particular."
This year, the Vesuvius Challenge continues. The text that we revealed so far represents just 5% of one scroll.
In 2024, our goal is to from reading a few passages of text to entire scrolls, and we're announcing a new $100,000 grand prize for the first team that is able to read at least 90% of all four scrolls that we have scanned.
The scrolls stored in Naples that remain to be read represent more than 16 megabytes of ancient text. But the villa where the scrolls were found was only partially excavated, and scholars tell us that there may be thousands more scrolls underground. Our hope is that the success of the Vesuvius Challenge catalyzes the excavation of the villa, that the main library is discovered, and that whatever we find there rewrites history and inspires all of us.
It's been a great joy to work on this strange and amazing project. Thanks to Brent Seales for laying the foundation for this work over so many years, thanks to the friends and Twitter users whose donations powered our effort, and thanks to the many contestants whose contributions have made the Vesuvius Challenge successful!
Read more in our announcement: scrollprize.org/grandprize
Philippe Modard retweeted
🎥 Check out the details of our recent Keras integration with @fchollet and @MeganRisdal!
Philippe Modard retweeted
[Arena Update]
We've collected over 6000 and 1500 votes for Mixtral-8x7B and Gemini Pro. Both show strong performance against GPT-3.5-Turbo.
Big congrats again on the release! @MistralAI @GoogleDeepMind
Full leaderboard: huggingface.co/spaces/lmsys/…
♊️Gemini is now in the Arena. Excited to see its ranking with human evals!
Meanwhile, our server just hit the highest traffic since May, with 10,000 votes in just 2 days. How incredible!
Huge thanks to @karpathy and the amazing community😂Let's vote at chat.lmsys.org
Philippe Modard retweeted
Google's Gemini Pro and Gemini Pro Vision is now available on Kaggle! Check out this notebook to learn how to get started at no cost with an API key.
👉kaggle.com/code/prathameshba…
We can't wait to see what the community creates with this.
Philippe Modard retweeted
We’re excited to announce 𝗚𝗲𝗺𝗶𝗻𝗶: @Google’s largest and most capable AI model.
Built to be natively multimodal, it can understand and operate across text, code, audio, image and video - and achieves state-of-the-art performance across many tasks. 🧵 dpmd.ai/announcing-gemini
Philippe Modard retweeted
Big news: we just released Keras 3.0!
▶ Run Keras on top of JAX, TensorFlow, and PyTorch
▶ Train faster with XLA compilation
▶ Unlock training runs with any number of devices & hosts via the new Keras distribution API
It's live on PyPI now! 🚀
Philippe Modard retweeted
TensorFlow Hub is moving to @kaggle Models! 🥳
We're excited to join the Kaggle community and see TF models available for experimentation with:
🔸 Model usage stats
🔸 A broad model collection
🔸 Diverse data for real-world stress testing
🔸 And more ↓ goo.gle/47GE9nt
Philippe Modard retweeted
🎉 We upgraded @kaggle GPU Notebooks to offer 29 GB RAM (up from 13GB) & 4 CPU cores (up from 2).
Have fun go nuts share cool stuff! kaggle.com/discussions/produ…
Philippe Modard retweeted
📣Kaggle's 2023 AI Report is out right now! 🎉 The report is a collection of 20+ essays written by Kaggle community members covering the latest AI advancements & salient topics in modern ML.
👉 Download: kaggle.com/AI-Report-2023
🎬 youtube.com/shorts/DfWpXuwSF…
Philippe Modard retweeted
We're launching Keras Core, a new library that brings the Keras API to JAX and PyTorch in addition to TensorFlow.
It enables you to write cross-framework deep learning components and to benefit from the best that each framework has to offer.
Read more: keras.io/keras_core/announce…



























