File size: 1,197 Bytes
b796ff0 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 |
---
tags:
- image-classification
- pytorch
---
### Description
This model with transfer learning is trained to recognise Sign Language.
The model was trained on the basis of ResNet50.
The dataset was used for training https://huggingface.co/datasets/aliciiavs/sign_language_image_dataset
In `model.py` file you can find class of custom model.
For using model you need download `model.py`, import from it class `CustomResNetModel`,
create an example of this class and load and apply model weight from file
`sign_language_resnet50.pth`
Model's metrics: `loss: 0.5739429252889922` `accuracy: 0.901717`
#### Example of code
```
import torch
from huggingface_hub import hf_hub_download
from model import CustomResNetModel
device = "cuda" if torch.cuda.is_available() else "cpu"
model = CustomResNetModel(num_classes=24)
weight_path = hf_hub_download(
repo_id="Irgenija/sign_language_resnet50",
filename="sign_language_resnet50.pth",
)
if cls.device == "cpu":
checkpoint = torch.load(
weight_path, map_location=torch.device("cpu")
)
else:
checkpoint = torch.load(weight_path)
model.load_state_dict(checkpoint)
model = model.to(device) # Optional
model.eval()
``` |