AI Briefing#huggingface
The latest AI and developer news about #huggingface, with the original source and a short summary.
The latest AI and developer news about #huggingface, with the original source and a short summary.
Artificial Analysis has introduced an LLM performance leaderboard on Hugging Face that comprehensively evaluates price, speed, and quality.
It introduces how to implement a custom handler that combines Whisper and Pyannote in Hugging Face Inference Endpoints to support speaker diarization and speculative decoding.
A study has found that minor formatting changes in prompts significantly affect LLM benchmark scores and model rankings.
A code generation model that trained itself using StarCoder2-15B, without human intervention or paid LLMs, has been released.
HuggingFace has released JAT, a general-purpose agent capable of performing various tasks, along with its dataset.
The Open Medical-LLM Leaderboard has been released to evaluate the performance of medical-domain LLMs in a standardized way.
Meta has unveiled Llama 3, a new open LLM available in 8B and 70B parameter sizes.
Zama's Concrete ML enables Fully Homomorphic Encryption (FHE)-based privacy-preserving inference on Hugging Face Endpoints.
Gradio's new Reload Mode instantly reflects UI changes without reloading the model.
Hugging Face has released Idefics2, an 8B-scale open source multimodal model with strong OCR performance.
A guide summarizing the principles and key models of Vision Language Models (VLMs), which process images and text simultaneously.
An integration has launched that lets developers easily deploy Hugging Face models to Google Cloud's Vertex AI and GKE.
Google has released the CodeGemma model series, optimized for code generation and mathematical reasoning.
Hugging Face has organized and published how it develops its AI policy and the key policy response materials it has submitted to U.S. government agencies.
It introduces Text-to-SQL technology that combines the DuckDB-NSQL-7B model with the Hugging Face API to query data using natural language.
Hugging Face has partnered with cloud security company Wiz to strengthen platform security.
AI21 has released Jamba, a Mamba-based model supporting a 256K context window.
Embedding quantization techniques are presented as a way to reduce the memory usage and cost of vector search while increasing speed.
Lighthouz AI and Hugging Face have launched the Chatbot Guardrails Arena to test LLMs' privacy protection guardrails.
HuggingFace has released Cosmopedia, a large-scale open synthetic dataset for LLM pretraining.
Hugging Face has released Quanto, a general-purpose PyTorch quantization backend for the Optimum library.
A large-scale synthetic dataset called WebSight and a model called Sightseer, which convert screenshots into HTML code, have been released.
Hugging Face and Argilla have unveiled a program that supports building high-quality datasets through community collaboration.
BigCode has released StarCoder2, a next-generation open-source code LLM, along with The Stack v2, a large-scale code dataset.