How a Data Organisation Structure should be. Google Cloud Singapore #GoogleCloudSummit #GoogleCloudSummitSG

Dataplex is the new thing in Data Governance coming from Google Cloud. Google Cloud Singapore #GoogleCloudSummit

Apigee is the platform to monetize and manage your APIs. Google Cloud Singapore #GoogleCloudSummit #GoogleSummitSG

Vertex AI from Google Cloud has introduced their new Generative AI platform where they're allowing us to train & fine-tune our own LLM. Google Cloud Singapore #GoogleCloudSummit #GoogleCloudSG

Google Cloud's Alloy DB is a fully managed PostgreSQL-compatible database service for your most demanding enterprise workloads. AlloyDB combines the best of Google with PostgreSQL, for superior performance, scale, and availability. Google Cloud Singapore #GoogleCloudSummit

Google Cloud presents Gen AI App Builder. It's interesting to see how easily we can build our custom Gen AI (conversational AI) models with our private knowledge base. Google Cloud Singapore #GoogleCloudSummit #GoogleCloudSummitSG

We will see that a lot of weird behaviors and problems of LLMs actually trace back to tokenization. We'll go through a number of these issues, discuss why tokenization is at fault, and why someone out there ideally finds a way to delete this stage entirely.

Hugging Face You can learn more about the model from below resources (h/t to Niels Rogge), thanks a lot for reading 🤗 Docs: huggingface.co/docs/transform… 📚 Checkpoints: huggingface.co/collections/mi… Demo notebooks: github.com/NielsRogge/Tra… 📕

Large Language Model Engineering Map 1. Data Collection and Preparation 1.1 Web Scraping 1.1.1 Crawling websites 1.1.2 Extracting text data 1.1.3 Handling different file formats (HTML, PDF, etc.) 1.2 Corpus Creation 1.2.1 Combining data from

ARR needs your help! We received 5800+ submissions for the June cycle, but with our current capacity, we can handle only half of these submissions. We can't start the review process without significant help. (1/2)

Pistis-RAG: A Scalable Cascading Framework Towards Trustworthy Retrieval-Augmented Generation Presents a multi-stage framework for RAG systems, with a content-centric approach that aligns retrieved information with LLM preferences and user feedback. 📝arxiv.org/abs/2407.00072

Huge congrats to AI at Meta on the Llama 3.1 release! Few notes: Today, with the 405B model release, is the first time that a frontier-capability LLM is available to everyone to work with and build on. The model appears to be GPT-4 / Claude 3.5 Sonnet grade and the weights are

Leveraging FourierKAN Classification Head for Pre-Trained Transformer-based Text Classification: arxiv.org/abs/2408.08803

[Self Attention] by Hand ✍️ - Spreadsheet 👉 by-hand.ai/s/self I made a customizable spreadsheet version of the self-attention by hand exercise I posted a while ago. Now you can finally use your own numbers and see the calculations. In this version, I introduced two new

I just added the new Llama 3.2 1B and 3B models to LitGPT, the open-source LLM library I help develop (focused on efficiency and code readability). LitGPT allows you to fine-tune and use these models on the cloud or a laptop. So, if you are looking for something to play with this

FourierKAN outperforms MLP on Text Classification Head Fine-tuning aRxiv: arxiv.org/abs/2408.08803 #KAN #AI #NLP

Whisper Turbo already runs locally on macOS with mlx_whisper. Transcribes 12 minutes in 14 seconds on an M2 Ultra (~50X faster than real time). pip install mlx_whisper Example:

Ilya Sutskever had a conversation last year with Jensen Huang but still, that made me rethink the idea of 'Predicting the Next Word' Listen to what Ilya said on this.

This is the Syllabus of "world's obviously best AI course" built by Andrej Karpathy

This is Dario Amodei. He's the CEO behind Claude, one of the world's most advanced AIs. Yesterday, in a 5.5 hour conversation with Lex Fridman, he revealed our timeline to superintelligence. Let me save you 5 hours: 🧵