ImageBind

Parameters

932M

Published

May 9, 2023

ImageBind is an AI model developed by Meta AI (United States), first published in May 2023. It works in the multimodal, vision, audio and 3 more domain, on tasks such as image classification, speech recognition (asr), image generation and language modeling/generation.

Epoch AI has no training-compute estimate for this model. The model has 932,000,000 parameters. Training ran on NVIDIA V100,NVIDIA A100.

Access: Open weights (non-commercial). Its weights are openly available. It is built on top of ViT-Huge/14. The reference paper has 1,515 citations. Epoch AI rates the confidence of this record as likely.

Full record

Organization: Meta AI
Country of organization: United States
Domain: Multimodal, Vision, Audio, Language, Image generation, Speech
Task: Image classification, Speech recognition (ASR), Image generation, Language modeling/generation
Parameters: 932,000,000
Training hardware: NVIDIA V100, NVIDIA A100
Model accessibility: Open weights (non-commercial)
Open weights: Yes
Base model: ViT-Huge/14
Citations: 1,515
Epoch confidence: Likely