Downloads · 30 days
0
alibi_vit_imagenet100
aadex/alibi_vit_imagenet100
alibi_vit_imagenet100 is a image classification model from aadex. Use it when you need a label for an image. The card lists the license as apache-2.0.
This model was trained using the vit-analysis framework.
Access
Public
Updated Dec 28, 2025
Repo size
66.3 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.pth66.3 MB · 100%
At a glance
- Task
- Image Classification
- License
- apache-2.0
- Access
- Public
- Created
- Dec 28, 2025
- Updated
- Dec 28, 2025
- SHA
- 529eb1c2
Model card
From the Hugging Face model README
Alibi Vit - IMAGENET100
This model was trained using the vit-analysis framework.
Model Details
- Model Type: ALIBI Vision Transformer
- Dataset: imagenet100
- Best Accuracy: 72.96%
- Image Size: 224
- Patch Size: 16
- Hidden Dim: 192
- Depth: 12
- Num Heads: 3
- MLP Dim: 768
- Num Classes: 100
Training Configuration
- Epochs: 120
- Batch Size: 512
- Learning Rate: 0.004
- Weight Decay: 0.05
- Label Smoothing: 0.1
Usage
import torch
from models import ALiBiSimpleVisionTransformer
# Load checkpoint
checkpoint = torch.load('alibi_vit_imagenet100_best.pth')
model = ... # Initialize model with same config
model.load_state_dict(checkpoint['state_dict'])
- Task
- Image Classification
- License
- apache-2.0
- Created
- Dec 28, 2025
- Updated
- Dec 28, 2025