Add Application File

Dockerfile

@@ -0,0 +1,21 @@
+FROM python:3.10-slim
+
+# Avoid Python .pyc files
+ENV PYTHONDONTWRITEBYTECODE=1
+ENV PYTHONUNBUFFERED=1
+
+# Set up user and workspace
+RUN useradd -m -u 1000 user
+USER user
+WORKDIR /app
+ENV PATH="/home/user/.local/bin:$PATH"
+
+# Copy and install dependencies
+COPY --chown=user requirements.txt .
+RUN pip install --no-cache-dir --upgrade -r requirements.txt
+
+# Copy all app code
+COPY --chown=user . .
+
+# Run the app
+CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]

app.py

@@ -0,0 +1,33 @@
+from fastapi import FastAPI, UploadFile, File
+from transformers import AutoProcessor, AutoConfig, AutoModelForImageClassification
+from PIL import Image
+import torch
+import io
+
+app = FastAPI()
+
+# Load processor and config
+processor = AutoProcessor.from_pretrained("quantized_model")
+config = AutoConfig.from_pretrained("dima806/deepfake_vs_real_image_detection")
+
+# Load model architecture and quantized weights
+model = AutoModelForImageClassification.from_config(config)
+model_quantized = torch.quantization.quantize_dynamic(
+    model, {torch.nn.Linear}, dtype=torch.qint8
+)
+model_quantized.load_state_dict(torch.load("quantized_model/model_quantized.pt", map_location="cpu"))
+model_quantized.eval()
+
+@app.post("/predict")
+async def predict(file: UploadFile = File(...)):
+    contents = await file.read()
+    image = Image.open(io.BytesIO(contents)).convert("RGB")
+
+    inputs = processor(images=image, return_tensors="pt")
+    with torch.no_grad():
+        logits = model_quantized(**inputs).logits
+        predicted_idx = logits.argmax(-1).item()
+        confidence = logits.softmax(-1)[0][predicted_idx].item()
+        label = model_quantized.config.id2label[predicted_idx]
+
+    return {"label": label, "confidence": confidence}

quantized_model/config.json

@@ -0,0 +1,34 @@
+{
+  "_attn_implementation_autoset": true,
+  "architectures": [
+    "ViTForImageClassification"
+  ],
+  "attention_probs_dropout_prob": 0.0,
+  "encoder_stride": 16,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.0,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "Real",
+    "1": "Fake"
+  },
+  "image_size": 224,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "Fake": 1,
+    "Real": 0
+  },
+  "layer_norm_eps": 1e-12,
+  "model_type": "vit",
+  "num_attention_heads": 12,
+  "num_channels": 3,
+  "num_hidden_layers": 12,
+  "patch_size": 16,
+  "pooler_act": "tanh",
+  "pooler_output_size": 768,
+  "problem_type": "single_label_classification",
+  "qkv_bias": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.51.3"
+}

quantized_model/model_quantized.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:326c5407df0c791de2246f53db8c03081ced7628f42d8b74afff4d6d3ef32aa0
+size 88529646

quantized_model/preprocessor_config.json

@@ -0,0 +1,23 @@
+{
+  "do_convert_rgb": null,
+  "do_normalize": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "image_processor_type": "ViTImageProcessor",
+  "image_std": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "resample": 2,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "height": 224,
+    "width": 224
+  }
+}

requirements.txt

@@ -0,0 +1,6 @@
+transformers
+torch
+fastapi
+uvicorn
+pillow
+python-multipart