TLDRocket
Sign in

Hugging Face

127 summarised stories about Hugging Face, each linking back to the original source. Browse all topics →

Wednesday, 16 March 2022

Image search with 🤗 datasets

Hugging Face Blog 4 years ago

Hugging Face's datasets library added support for image search by combining its ImageFolder loader with CLIP embeddings and FAISS indexing, allowing users to search collections of images using text queries. The demonstration used a 10,000-image dataset of embellishments from digitised British Library books, encoded with the CLIP-ViT-B-32 model to create searchable embeddings. Users can now load image datasets, embed them with sentence_transformers, index them with FAISS, and retrieve results by querying with natural language descriptions.

Read at Hugging Face Blog
Hugging Face

Accelerate BERT inference with Hugging Face Transformers and AWS Inferentia

Hugging Face Blog 4 years ago

Hugging Face Transformers and AWS released a tutorial for deploying BERT models on AWS Inferentia chips, custom hardware designed to accelerate inference workloads. AWS Inferentia delivers up to 80% lower cost per inference and 2.3X higher throughput than comparable GPU instances, with the tutorial achieving 5-6ms latency for sequence length 128. Companies moving BERT to production can reduce inference costs and increase throughput by using Inferentia instead of GPUs or CPUs.

Read at Hugging Face Blog
Hugging Face Inference Optimization

The daily briefing

Every AI story that matters, in your inbox by 8am.

TLDRocket reads 60+ sources, removes duplicate coverage, and summarises the day in two minutes. Free, no spam, unsubscribe anytime.