Checkpoints

We provide links for you to download our checkpoints, including pretrained and finetuned models on different tasks.

Pretraining

ONE-PEACE can be disassembled into different branches to handle different tasks. In addition to the full model, we also provide the vision-branch of ONE-PEACE, which can be used to perform vision tasks.

• Pretrained checkpoint
• ONE-PEACE Vision Branch

Vision

• Finetuned checkpoint for ImageNet Classification 384 resolution
• Finetuned checkpoint for ImageNet Classification 512 resolution
• Finetuned checkpoint for COCO Object Detection and Instance Segmentation
• Finetuned checkpoint for ADE20K Semantic Segmentation
• Finetuned checkpoint for K400 Action Recognition

Audio

• Finetuned checkpoint for FSD50K
• Finetuned checkpoint for Vggsound
• Finetuned checkpoint for Audio-Language Retrieval (AudioCaps & Clotho)
• Finetuned checkpoint for AQA

Vision-Language

• Finetuned checkpoint for COCO Image-Text Retrieval
• Finetuned checkpoint for Flickr30K Image-Text Retrieval
• Finetuned checkpoint for RefCOCO
• Finetuned checkpoint for RefCOCO+
• Finetuned checkpoint for RefCOCOg
• Finetuned checkpoint for NLVR2
• Finetuned checkpoint for VQA