ReALM

Parameters

3B

Published

Mar 29, 2024

ReALM is an AI model developed by Apple (United States), first published in March 2024. It works in the language domain, on tasks such as named entity recognition (ner), language modeling and part-of-speech tagging.

Epoch AI has no training-compute estimate for this model. The model has 3,000,000,000 parameters. It was trained on roughly 134B datapoints.

Access: Unreleased. Its weights are not openly released. It is built on top of Flan-T5 11B. Epoch AI rates the confidence of this record as confident.

Full record

Organization: Apple
Country of organization: United States
Domain: Language
Task: Named entity recognition (NER), Language modeling, Part-of-speech tagging
Parameters: 3,000,000,000
Dataset size: 134B
Model accessibility: Unreleased
Open weights: No
Base model: Flan-T5 11B
Epoch confidence: Confident