AI Briefing#transformer
The latest AI and developer news about #transformer, with the original source and a short summary.
The latest AI and developer news about #transformer, with the original source and a short summary.
Hugging Face now supports AWS Inferentia2 optimization, improving inference performance and cost efficiency for Transformers models.
Explains how to implement graph classification using the Transformers library and the Graphormer model.
A technical tutorial on implementing federated learning for language models by combining Hugging Face with the Flower framework.
This covers a case where machine learning was used during the Turkey earthquake to structure rescue request information from social media and support rescue activities.
Fetch built its own AI models using Hugging Face and AWS to improve receipt processing efficiency and shorten development time.
BLIP-2 introduces efficient vision-language learning through a Q-Former that connects a frozen image encoder and an LLM.
SpeechT5, a unified speech processing model, has been added to the Hugging Face Transformers library.
It covers how to optimize inference performance of PyTorch Transformer models using Intel Sapphire Rapids CPUs and the Optimum Intel library.
This explains the core training strategies and architecture of Vision-Language Models, which combine visual and language information.
Hugging Face is expanding its computer vision ecosystem, supporting 8 core vision tasks and over 3,000 models.
Introducing a universal model that performs instance, semantic, and panoptic segmentation in an integrated way through Mask2Former and OneFormer.
Introduces how to accelerate PyTorch Transformer training using the AMX technology of Intel Sapphire Rapids.
The Bumblebee library has been released, enabling Hugging Face models such as GPT2 and Stable Diffusion to run in the Elixir environment.
It explains a method for performing probabilistic time series forecasting that incorporates uncertainty by leveraging the Encoder-Decoder structure of the Transformer.
Introduces how to perform Transformer-based sentiment analysis on encrypted data using Concrete-ML.
Unveiled **OFA**, a multimodal pretrained model that unifies tasks, modalities, and architecture into one.
Contrastive Search, which solves the text repetition problem, has been introduced to the Transformers library.
This guide explains how to fine-tune the Whisper model on multilingual automatic speech recognition (ASR) datasets using Hugging Face Transformers.
HuggingFace explains the technical process of optimizing the BLOOM model's inference server, improving latency by 5x and throughput by 50x.
Introduces the concept and training method of Decision Transformer, which solves reinforcement learning as sequence modeling.
It explains an 8-bit quantization technique that cuts memory usage of large language models by 2x while preserving performance, and its integration into the Hugging Face library.
We introduce Nyströmformer, which uses the Nyström method to reduce the complexity of Transformer's self-attention from $O(n^2)$ to $O(n)$.
By combining TensorFlow and XLA in Hugging Face's `transformers`, text generation speed can be increased by up to 100x.
Explains how to convert Transformers models to ONNX format and optimize inference performance using Hugging Face Optimum.