Hugging Face announces the Open Medical-LLM Leaderboard, a standardized platform for evaluating and comparing large language models on medical tasks using datasets like MedQA, MedMCQA, PubMedQA, and…
Hugging Face announces the release of Meta's Llama 3 open LLM family, including 8B and 70B parameter models (base and instruct variants) and Llama Guard 2, with integrations into the Hugging Face…
Ryght launched Ryght Preview, a generative AI platform for healthcare and life sciences, now publicly available, with support from Hugging Face's Expert Support Program.
Zama announces that pre-compiled FHE (Fully Homomorphic Encryption) models using Concrete ML are now available for deployment on Hugging Face Endpoints, enabling privacy-preserving inferences on…
Hugging Face introduces the LiveCodeBench leaderboard, a new benchmark for evaluating code LLMs that collects coding problems over time from various coding contest platforms to prevent contamination,…
This post explains Gradio's reload mode, which allows developers to see UI changes instantly without restarting the server, and demonstrates building a document analyzer app using the Hugging Face…
Hugging Face releases Idefics2, an 8B parameter vision-language model with Apache 2.0 license, improved OCR, and competitive performance on multimodal benchmarks.
LG AI Research's blog post highlights three time series forecasting models (FITS, ModernTCN, TimeMixer) presented at ICLR 2024, describing their methodologies and performance.
OpenAI states that LLMs are susceptible to prompt injections, jailbreaks, and other attacks that allow adversaries to overwrite a model's original instructions.
Replicate announces that Meta Llama 3 is now available to run via its API, with four variants (70B instruct, 8B instruct, 70B base, 8B base) and instructions for JavaScript, Python, cURL, and other…