类库
› SenseNova-Vision
OpenSenseNova/SenseNova-Vision
SenseNova-Vision是一个统一的视觉多模态生成模型，支持图像理解与生成。它通过单一架构处理多种视觉任务，提供7B参数模型权重、50M训练数据集及推理代码，旨在简化多模态AI应用的开发与评估流程。
技术栈
根目录 python
查看全部依赖 (44)
依赖
NumPy
OpenEXR
Pandas
Pillow
PyYAML
Requests
accelerate
bitsandbytes
decord
einops
gradio
h5py
huggingface_hub
matplotlib
ninja
omegaconf
open3d
opencv_python
openjdk
8
packaging
pip
25.1
projectaria-tools
protobuf
psutil
pyarrow
pycocoevalcap
pycocotools
python
3.10.16
regex
safetensors
scipy
sentencepiece
setuptools
78.1.1
shapely
tabulate
torch
torchvision
tqdm
transformers
trimesh
triton
wandb
wheel
0.45.1
xlsxwriter
