Woww

gpt2-chatbot was amazing!
I had it side by side with Claude 3 in the arena and asked them to create a basic rust interpreter in python. gpt2-chatbot nailed it first try while Claude 3 opus failed. And the rust interpreter worked! Pure magic!

Introducing ExecuTorch Alpha ⚡

ExecuTorch Alpha is focused on deploying large language models and large ML models to the edge, stabilizing the API surface, and improving installation processes.

Learn more in our latest blog: hubs.la/Q02vzrrW0

🤯 𝐒𝐭𝐨𝐫𝐲𝐃𝐢𝐟𝐟𝐮𝐬𝐢𝐨𝐧 - Generates high-quality videos using consistent generated images or with user's input-images as a condition.

More examples and links below👇

Andrew Tan 💯. My mantra this year is to be a better specialist in all my generalist categories. 😄

Apple presents CatLIP

CLIP-level Visual Recognition Accuracy with 2.7x Faster Pre-training on Web-scale Image-Text Data

Contrastive learning has emerged as a transformative method for learning effective visual representations through the alignment of image and text

Yesterday, we open sourced the Cohere Toolkit. We think this will be a major accelerant for getting LLMs into production within enterprise. github.com/cohere-ai/cohe…

Wow

Apple is joining the public AI game with 4 new models on the Hugging Face hub! huggingface.co/collections/ap…

Apple presents OpenELM

An Efficient Language Model Family with Open-source Training and Inference Framework

The reproducibility and transparency of large language models are crucial for advancing open research, ensuring the trustworthiness of results, and

Two new AI releases by Apple today:

🧚‍♀️ OpenELM, a set of small (270M-3B) efficient language models. Weights on the Hub:
Pretrained: huggingface.co/collections/ap…
Instruct: huggingface.co/collections/ap…

👷‍♀️ CoreNet, a training library used to train OpenELM: github.com/apple/corenet

🔥🔥🔥🔥

Amazing work from the Phi Team, pushing the boundaries of llm efficiency beyond limits!

Microsoft announces Phi-3

A Highly Capable Language Model Locally on Your Phone

We introduce phi-3-mini, a 3.8 billion parameter language model trained on 3.3 trillion tokens, whose overall performance, as measured by both academic benchmarks and internal testing,

🦙 We’re excited to support AI at Meta Llama 3 models in NVIDIA TensorRT-LLM, accelerating and optimizing Llama 3 inference performance on our GPUs. #LLM

Get started ➡️ nvda.ws/4b57w4z

In a clear, step-by-step tutorial, Ayoola Olafenwa walks us through the process of deploying quantized LLMs, and shares annotated code snippets in case you want to experiment with this workflow on your own. buff.ly/3U3vyGo

We have just released 🍷 FineWeb: 15 trillion tokens of high quality web data.
We filtered and deduplicated all CommonCrawl between 2013 and 2024.
Models trained on FineWeb outperform RefinedWeb, C4, DolmaV1.6, The Pile and SlimPajama!

Keeping this here! 🥰🥰

💃 Visited Action Tutoring ‘s Fivefields x+why office on 17 April 2024!

🎉 Met the charity’s most hospitable team! 💯

💯 Many thanks for the opportunity to volunteer with the education charity, helping me become a passionate award-winning tutor!✍️

Llama 3 is officially the fastest model from release to #1 trending on Hugging Face - in just a few hours.

30,000 new models have been released based on llama 1 & 2 so I can't wait to see the impact that the third and most powerful version will have on the ecosystem! 🚀🚀🚀

In addition to Llama 3, today we’re also publishing a new paper: Imagine Flash: Accelerating Emu Diffusion Models with Backward Distillation ➡️ go.fb.me/g4r584

This work from GenAI researchers is enabling new image generation features in Meta AI on WhatsApp & web.