Anyscale, the company behind API platform Anyscale Endpoints, has announced its collaboration with Meta to advance the Llama 2 ecosystem. Presumably, this collaboration will hinge upon the work on LLM fine-tuning showing that, when fine-tuned, even the smallest Llama model can outperform GPT-4.
Microsoft’s protein-generating diffusion model, EvoDiff, is now open source. The groundbreaking model generates protein sequences rather than 3D structures, allowing its output to be more general, flexible, and realistic than that of other similar models.
The UAE’s Technology Innovation Institute has unveiled the newest iteration in its Falcon family of large language models. Currently at the top of the Hugging Face Leaderboard for pretrained LLMs, Falcon 180B is the largest LLM openly available for research and commercial use.
Join the Data Phoenix webinar in which Dmytro Spodarets and guest Shishir Patil (Ph.D. student at UC Berkeley) will introduce the Gorilla project, designed to connect LLMs with various services and applications exposed through APIs.
Join the workshop of our good friends (Greg Loughnane and Chris Alexiuk from AI Makerspace) to learn how to build your first LLM application using the OpenAI API, Chainlit, and Hugging Face Spaces.
Join the Data Phoenix webinar in which Dmytro Spodarets, together with guest Gloria Felicia (CEO and co-founder of InsightCircle AI), will explore potential risks, break down advanced defense methods, and share the latest on keeping Large Language Models safe.
Join DataTalks.Club course to get started with machine learning engineering.
Learn from AI Makerspace team how to leverage the Llama 2 with LangChain to build the most popular type of LLM application — a Retrieval Augmented Generation (RAG) or Retrieval Augmented Question Answering (RAQA) system.
Arthur released Arthur Bench, an open-source tool to help users find the best LLM for a particular data set by evaluating how different LLMs will perform in real-world scenarios.
This talk will share some reference architectures and demos for emerging Generative AI applications and use cases.
Join the Data Phoenix webinar in which Dmytro Spodarets, together with guest Zenodia Charpy (Senior Deep Learning Data Scientist, NVIDIA), will dive into foundation and ChatGPT-style models, generative AI and LLM technology at NVIDIA, shortcomings, and proposed guardrails, and the road ahead.
Segmind, a platform specializing in open-source serverless APIs for generative models, announced the release of its distillation of SDXL 1.0. SSD-1B is half the size and 60% faster than its base model. It can be fine-tuned to custom specifications and called using Segmind's serverless API.