README.md

---
library_name: transformers
tags:
- classification
- deepfake
base_model:
- facebook/convnextv2-tiny-1k-224
---

# Model Card for Model ID

<!-- Provide a quick summary of what the model is/does. -->



## Model Details

### Model Description

<!-- Provide a longer summary of what this model is. -->

This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.


### Model Sources [optional]

<!-- Provide the basic links for the model. -->

- **Repository:** [github](https://github.com/g25ait2119/mlops-pipeline-group30)
- **Paper:** [ConvNeXt V2](https://arxiv.org/abs/2301.00808)

## Model Description

The model is fine-tuned on ConvNext V2 model.




## Uses

This finetuned model can be used for text classification. It has been trained to classify real and fake images.





#### Hardware

- **Hardware Type:** GPU T4
- **Hours used:** ~ 14 Minutes
- **Cloud Provider:** Kaggle


### Inference code

```python

import os
import torch
from PIL import Image
from transformers import AutoImageProcessor, AutoModelForImageClassification


MODEL_ID = "computervisionpro/convnextv2-real-fake"


def predict(image_path, model_id=MODEL_ID):
    # device = "cuda" if torch.cuda.is_available() else "cpu"
    device = "cpu"
    # hf_token = os.getenv("HF_TOKEN") or None

    processor = AutoImageProcessor.from_pretrained(model_id)
    model = AutoModelForImageClassification.from_pretrained(model_id)
    model.to(device)
    model.eval()

    image = Image.open(image_path).convert("RGB")
    inputs = processor(images=image, return_tensors="pt")
    inputs = {key: value.to(device) for key, value in inputs.items()}

    with torch.inference_mode():
        outputs = model(**inputs)
        probs = torch.softmax(outputs.logits, dim=-1)[0]

    pred_id = int(torch.argmax(probs).item())
    label = model.config.id2label.get(pred_id, str(pred_id))
    confidence = float(probs[pred_id].item())

    return {
        "image": image_path,
        "model": model_id,
        "prediction": label,
        "confidence": confidence,
        "probabilities": {
            model.config.id2label.get(i, str(i)): float(prob.item())
            for i, prob in enumerate(probs)
        },
    }


result = predict("./dataset/test/fake/fake_1006.jpg")
print()
print(result)
```


### Results

- [WandB](https://wandb.ai/computervisionpro-na/mlops-assignment3)



## Important Links


- [Data](https://www.kaggle.com/datasets/manjilkarki/deepfake-and-real-images)
- [Kaggle Notebook](https://www.kaggle.com/code/computervisionpro/group30-mlops-a3)