Artificial Intelligence #186

Hey, in this issue: open challenges in LLM research; GPT-3.5 Turbo is open for custom tuning; AI raises complicated questions about authorship; SeamlessM4T: a foundational multilingual and multitask model that seamlessly translates and transcribes across speech and text; llama-gpt: a self-hosted, offline, ChatGPT-like chatbot; and more.

The sponsors of this issue are Cerebrium and Mission. To become a sponsor, please reach out to?[email protected].

• Open challenges in LLM research
• OpenAI opens GPT-3.5 Turbo up for custom tuning
• IDEFICS: the first open SOTA visual language model at the 80B scale
• AI raises complicated questions about authorship
• [Sponsored] Try Cerebrium For Free

Cerebrium: Seamless ML Model deployment, <1s cold start, major frameworks supported (Pytorch, Onnx, XGBoost), 18+ pre-built models, fine-tuning (FlantT5, GPT-Neo, Stable Diffusion). See why Cerebrium is trusted by users in organizations such as Twilio, Ramp, and Writesonic! $10 Free Credit When You Sign Up.

• [Ars Technica] How ChatGPT turned generative AI into an “anything tool”
• [Meta] SeamlessM4T: a foundational multilingual and multitask model that seamlessly translates and transcribes across speech and text
• [Sponsored] [Ebook] | Designing Large Language Model Applications | Learn the tools and playbooks for building valuable products
• [Project] llama-gpt: a self-hosted, offline, ChatGPT-like chatbot. Powered by Llama 2. 100% private, with no data leaving your device
• [Nvidia] Curating trillion-token datasets: Introducing NVIDIA NeMo Data Curator

More than 690,000 subscribers are reading this newsletter. If you are building an AI or a data product or service, you can become a sponsor of one of the future newsletter issues and get your business featured in the newsletter. Feel free to contact?[email protected]?for more details on sponsorships.

Enjoy the newsletter? Help us make it bigger and better by sharing it with colleagues and friends.