remove transformers dependency, added script for endpoint testing

endpoint_tester.py

@@ -0,0 +1,37 @@
+import json
+from typing import List
+import requests as r
+import base64
+from PIL import Image
+from io import BytesIO
+
+ENDPOINT_URL = "https://j0thokqylseue22z.us-east-1.aws.endpoints.huggingface.cloud" # your endpoint url
+HF_TOKEN = "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" # your huggingface token `hf_xxx`
+
+# helper image utils
+def encode_image(image_path):
+  with open(image_path, "rb") as i:
+    b64 = base64.b64encode(i.read())
+  return b64.decode("utf-8")
+
+def predict(image):
+    image = encode_image(image)
+
+    # prepare sample payload
+    payload = {"image": image, "inputs": ""}
+    # headers
+    headers = {
+        "Authorization": f"Bearer {HF_TOKEN}",
+        "Content-Type": "application/json",
+    }
+
+    response = r.post(ENDPOINT_URL, headers=headers, json=payload)
+    if response.status_code != 200:
+        print(response.text)
+        raise Exception("Prediction failed")
+    return response.json()
+
+prediction = predict(
+  image = "test.png"
+)
+

handler.py

@@ -79,6 +79,7 @@ class EndpointHandler():
             reverse_norm_std = torch.tensor([0.26862954, 0.26130258, 0.27577711])[:, None, None]
             image_tensor = image_tensor * reverse_norm_std + reverse_norm_mean
             pil_img = T.ToPILImage()(image_tensor)
+            a
             image_h = pil_img.height
             image_w = pil_img.width
             image = np.array(pil_img)[:, :, [2, 1, 0]]

requirements.txt

@@ -1,2 +1 @@
 opencv-python
-transformers