Add model card
Browse files
README.md
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: pytorch
|
| 3 |
+
license: mit
|
| 4 |
+
pipeline_tag: image-classification
|
| 5 |
+
tags:
|
| 6 |
+
- vision-transformer
|
| 7 |
+
- walkability
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
# ViT-Walkability
|
| 11 |
+
|
| 12 |
+
This repository contains a fine-tuned Vision Transformer model for walkability-related image classification tasks.
|
| 13 |
+
|
| 14 |
+
## Checkpoint details
|
| 15 |
+
- Source file: `best.pt`
|
| 16 |
+
- Top-level type: `dict`
|
| 17 |
+
- Base model: `google/vit-base-patch16-224`
|
| 18 |
+
- Num labels: `5`
|
| 19 |
+
- Labels: `['LABEL_0', 'LABEL_1', 'LABEL_2', 'LABEL_3', 'LABEL_4']`
|
| 20 |
+
- Keys: `['model_state_dict', 'epoch']`
|
| 21 |
+
|
| 22 |
+
## Loading notes
|
| 23 |
+
The uploaded files are in standard Hugging Face Transformers format and can be loaded directly with `from_pretrained`.
|