added dedicated model

requirements.txt

@@ -0,0 +1 @@
+transformers

utils/inference.py

@@ -1,11 +1,39 @@
-import requests
+# import requests
 
-API_URL = "https://api-inference.huggingface.co/models/trpakov/vit-face-expression"
-headers = {"Authorization": "Bearer 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"}
+# API_URL = "https://api-inference.huggingface.co/models/trpakov/vit-face-expression"
+# headers = {"Authorization": "Bearer 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"}
+
+
+# def query(filename):
+#     with open(filename, "rb") as f:
+#         data = f.read()
+#     response = requests.post(API_URL, headers=headers, data=data)
+#     return response.json()
+
+
+from PIL import Image
+from transformers import CLIPProcessor, CLIPModel
+
+model = CLIPModel.from_pretrained("openai/clip-vit-large-patch14")
+processor = CLIPProcessor.from_pretrained("openai/clip-vit-large-patch14")
 
 
 def query(filename):
-    with open(filename, "rb") as f:
-        data = f.read()
-    response = requests.post(API_URL, headers=headers, data=data)
-    return response.json()
+    image = Image.open(filename)
+    inputs = processor(
+        text=["Happy", "Sad", "Surprised", "Angry", "Disgusted", "Neutral", "Fearful"],
+        images=image,
+        return_tensors="pt",
+        padding=True,
+    )
+    outputs = model(**inputs)
+    logits_per_image = outputs.logits_per_image
+    probs = logits_per_image.softmax(dim=1)
+    output = {
+        label: prob.item()
+        for label, prob in zip(
+            ["Happy", "Sad", "Surprised", "Angry", "Disgusted", "Neutral", "Fearful"],
+            probs[0],
+        )
+    }
+    return output