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.
A new benchmark leaderboard has been released to evaluate the performance of LLMs specialized in the financial domain.
BenCzechMark, a benchmark that comprehensively evaluates LLMs' Czech language understanding and reasoning abilities, has been released.
HuggingFace has released FineVideo, a dataset of 43,000 videos with rich metadata.
A method and models for fine-tuning existing LLMs into the 1.58-bit BitNet architecture to maximize computational efficiency have been released.
Hugging Face has launched a SQL Console feature that lets users query datasets directly in the browser.
A new feature has been released that lets users connect Hugging Face Spaces directly as tools in HuggingChat, extending the model's capabilities.
Hugging Face's distributed training framework Accelerate has announced the release of version 1.0.0.
Hugging Face has partnered with TruffleHog to introduce scanning that prevents API key and token leaks within repositories.
Hugging Face has released the LeRobotDataset format, which leverages video encoding to improve storage efficiency and loading speed for robotics datasets.
Hugging Face introduced packing functionality compatible with Flash Attention 2, boosting LLM training efficiency by up to 2x.
A Hugging Face guide walks through deploying Llama 3.1 405B on Google Cloud Vertex AI.
Hugging Face has released a unified Tool Use API that is compatible across various model families.
Hugging Face has acquired XetHub to optimize the management of large-scale data and models.
It introduces a multimodal data augmentation technique that simultaneously transforms text and images in document images to improve VLM training efficiency.
Hugging Face summarized key security features—including fine-grained permission tokens, 2FA, and commit signing—designed to protect Hub users' assets.
Meta's Llama 3.1 model series has been released, offering 128K context and multilingual support.
Hugging Face's TGI has launched a feature that dynamically serves multiple LoRA adapters from a single base model.
HuggingFace has released Docmatix, a DocVQA dataset that is 240 times larger than existing ones, containing 2.4 million images and 9.5 million Q/A pairs.
HuggingFace released the SmolLM series of small language models in 135M, 360M, and 1.7B sizes, trained on high-quality data.
HuggingFace's TRL library now supports Direct Preference Optimization (DPO) for Vision Language Models (VLMs).
Hugging Face is experimenting with a report feature that uses Presidio to flag whether datasets contain personally identifiable information (PII).
Transformers models from Hugging Face can now be loaded and used directly in KerasHub.
Google Cloud TPU v5e is now available in Hugging Face's Inference Endpoints and Spaces.
Hugging Face has introduced dataset search capabilities by modality, size, format, and library.