added model

Dockerfile

@@ -10,7 +10,7 @@ COPY requirements.txt .
 RUN pip install --no-cache-dir -r requirements.txt
 
 # Copy app code
-COPY . .
+COPY ../backend .
 
 # Expose port
 EXPOSE 7860

README.md

@@ -1,10 +0,0 @@
----
-title: Derma Scan
-emoji: 👁
-colorFrom: yellow
-colorTo: green
-sdk: docker
-pinned: false
----
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -28,11 +28,8 @@ app.add_middleware(
 )
 
 # Load model and processor
-processor = AutoImageProcessor.from_pretrained(
-    "Anwarkh1/Skin_Cancer-Image_Classification",
-    use_fast=True
-)
-model = AutoModelForImageClassification.from_pretrained("Anwarkh1/Skin_Cancer-Image_Classification")
+processor = AutoImageProcessor.from_pretrained("model")
+model = AutoModelForImageClassification.from_pretrained("model")
 
 
 @app.get("/")

model/config.json

@@ -0,0 +1,42 @@
+{
+  "architectures": [
+    "ViTForImageClassification"
+  ],
+  "attention_probs_dropout_prob": 0.0,
+  "encoder_stride": 16,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.0,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "benign_keratosis-like_lesions",
+    "1": "basal_cell_carcinoma",
+    "2": "actinic_keratoses",
+    "3": "vascular_lesions",
+    "4": "melanocytic_Nevi",
+    "5": "melanoma",
+    "6": "dermatofibroma"
+  },
+  "image_size": 224,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "actinic_keratoses": 2,
+    "basal_cell_carcinoma": 1,
+    "benign_keratosis-like_lesions": 0,
+    "dermatofibroma": 6,
+    "melanocytic_Nevi": 4,
+    "melanoma": 5,
+    "vascular_lesions": 3
+  },
+  "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,
+  "qkv_bias": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.51.3"
+}

model/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b8136bbbdd085cc97f69bdef01bd0c4c54262cf3ba44f5a94ebc9c8c54e85d60
+size 343239356

model/preprocessor_config.json

@@ -0,0 +1,62 @@
+{
+  "_name_or_path": "google/vit-base-patch16-224-in21k",
+  "architectures": [
+    "ViTForImageClassification"
+  ],
+  "attention_probs_dropout_prob": 0.0,
+  "do_convert_rgb": null,
+  "do_normalize": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "encoder_stride": 16,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.0,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "benign_keratosis-like_lesions",
+    "1": "basal_cell_carcinoma",
+    "2": "actinic_keratoses",
+    "3": "vascular_lesions",
+    "4": "melanocytic_Nevi",
+    "5": "melanoma",
+    "6": "dermatofibroma"
+  },
+  "image_mean": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "image_processor_type": "ViTImageProcessor",
+  "image_size": 224,
+  "image_std": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "actinic_keratoses": 2,
+    "basal_cell_carcinoma": 1,
+    "benign_keratosis-like_lesions": 0,
+    "dermatofibroma": 6,
+    "melanocytic_Nevi": 4,
+    "melanoma": 5,
+    "vascular_lesions": 3
+  },
+  "layer_norm_eps": 1e-12,
+  "model_type": "vit",
+  "num_attention_heads": 12,
+  "num_channels": 3,
+  "num_hidden_layers": 12,
+  "patch_size": 16,
+  "qkv_bias": true,
+  "resample": 2,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "height": 224,
+    "width": 224
+  },
+  "torch_dtype": "float32",
+  "transformers_version": "4.38.2"
+}

test_huggingface.py

@@ -1,35 +0,0 @@
-from transformers import AutoImageProcessor, AutoModelForImageClassification
-from PIL import Image
-import torch
-import time
-import torch.nn.functional as F
-from pprint import pprint
-
-# Load model and processor
-processor = AutoImageProcessor.from_pretrained("Anwarkh1/Skin_Cancer-Image_Classification")
-model = AutoModelForImageClassification.from_pretrained("Anwarkh1/Skin_Cancer-Image_Classification")
-
-# Load and preprocess the image
-image = Image.open("skin_cancer_test.jpeg").convert("RGB")
-inputs = processor(images=image, return_tensors="pt")
-
-start = time.time()
-
-# Run inference
-with torch.no_grad():
-    outputs = model(**inputs)
-
-end = time.time()
-
-# Get predicted class
-# Convert logits to probabilities
-logits = outputs.logits
-probs = F.softmax(logits, dim=-1)[0]  # Shape: [num_classes]
-
-# Map class labels to probabilities
-id2label = model.config.id2label
-result = {id2label[i]: probs[i].item() for i in range(len(probs))}
-result = {"probabilities": result}
-# Print the result
-pprint(result)
-print("Elapsed time: ", end - start)

test_main.py

@@ -1,59 +0,0 @@
-import unittest
-import requests
-
-BACKEND_URL = "https://dermadetect.onrender.com"   # Change this to the backend url when testing
-TEST_IMG_PATH = "skin_cancer_test.jpeg"
-search_header = {"Content-Type": "application/json"}
-
-
-class MyTestCase(unittest.TestCase):
-
-    def test_root(self):
-        expected_response = {"Hello": "World"}
-        expected_response_code = 200
-        response = requests.get(BACKEND_URL)
-
-        self.assertEqual(response.status_code, expected_response_code)
-        self.assertEqual(response.json(), expected_response)
-
-    def test_search_doc(self):
-        expected_response_code = 200
-        search_endpoint = BACKEND_URL + "/findDoc"
-        sample_payload = {
-            "query": "melanoma",
-            "page": 0,
-            "getRankingInfo": True,
-            "aroundRadius": 100000,
-            "aroundLatLng": "37.6436195373535, -121.867645263672"
-        }
-        response = requests.post(search_endpoint, json=sample_payload, headers=search_header)
-        self.assertEqual(response.status_code, expected_response_code)
-        self.assertTrue("hits" in response.json())
-
-    def test_search_doc_invalid_type(self):
-        expected_response_code = 422
-        search_endpoint = BACKEND_URL + "/findDoc"
-        sample_payload = {
-            "query": "melanoma",
-            "page": 0,
-            "getRankingInfo": True,
-            "aroundRadius": 100000,
-            "aroundLatLng": 123
-        }
-        response = requests.post(search_endpoint, json=sample_payload, headers=search_header)
-        self.assertEqual(response.status_code, expected_response_code)
-
-    def test_get_predictions(self):
-        prediction_endpoint = BACKEND_URL + "/predict"
-        expected_response_code = 200
-        with open(TEST_IMG_PATH, "rb") as file:
-            file_name = TEST_IMG_PATH.split("/")[-1]
-            files = {"file": (file_name, file, "image/jpeg")}
-            response = requests.post(prediction_endpoint, files=files)
-
-            self.assertEqual(expected_response_code, response.status_code)
-            self.assertTrue("probabilities" in response.json())
-
-
-if __name__ == '__main__':
-    unittest.main()