Promote Sign in Sign up

Multimodal model

UGC VideoCaptioner

UGC VideoCaptioner is a 3B parameter omni model that writes detailed, paragraph style captions for short form user generated videos, combining visual, audio, and on screen text understanding into a single narrative description.

Memories.ai Multimodal Open weights mit
Visit model website ↗ Hugging Face
Type
Multimodal
Weights
Open
Downloads
13
Likes
2
Pipeline
video text to text
License
mit

Model hub

Memories-ai/UGC-VideoCaptioner

Repo
Memories-ai/UGC-VideoCaptioner
Library
transformers
Access
Public
Technical tags and languages
transformers safetensors qwen2_5_omni text-to-audio multimodal video-captioning audio-visual ugc video-text-to-text dataset:openinterx/UGC-VideoCap base_model:Qwen/Qwen2.5-Omni-3B base_model:finetune:Qwen/Qwen2.5-Omni-3B
Open on Hugging Face

Papers

1

More from Memories.ai

2

About Memories.ai

1

Memories.ai is an AI research company that builds a "Large Visual Memory Model" (video AI) enabling machines to index, search, and act on large-scale video/camera footage.