Merge pull request #4 from roboflow/add-tests

Added the fruits SOM image and logic fixes

index.html

@@ -55,7 +55,7 @@ <h2>Response Time</h2>
                         </div>
                         <div class="chart">
                             <div class="chart_box chart_box_green">
-                                <p>4.12 ms</p>
+                                <p>4.12 s</p>
                             </div>
                         </div>
                     </div>
@@ -80,8 +80,7 @@ <h2>Zero-Shot Classification</h2>
                                 <h3><span class="explainer_icon far fa-comment-dots"></span>Prompt</h3>
                                 <pre>
                             What is in the image? Return the class of the object in the image. Here are the classes: Toyota Camry, Tesla Model 3. You can only return one class from that list.
-                        </pre
-                                >
+                        </pre>
                                 <h3><span class="explainer_icon far fa-image"></span>Image</h3>
                                 <img class="test_image" src="images/car.jpeg" />
                                 <h3><span class="explainer_icon far fa-sparkles"></span>Result</h3>

results/2023-11-17.json

@@ -1 +1 @@
-{"zero_shot_classification": {"score": 1, "response_time": 1.5038928985595703, "result": "Toyota Camry"}, "count_fruit": {"score": 0, "response_time": 2.0013718605041504, "result": "9"}, "document_ocr": {"score": 1, "response_time": 10.280664682388306, "result": "I was thinking earlier today that I have gone through, to use the lingo, eras of listening to each of Swift's Eras. Meta indeed. I started listening to Ms. Swift's music after hearing the Midnights album. A few weeks after hearing the album for the first time, I found myself playing various songs on repeat. I listened to the album in order multiple times."}, "handwriting_ocr": {"score": 1, "response_time": 8.720427989959717, "result": "The words of songs on the album have been echoing in my head all week. \"Fades into the grey of my day old tea.\""}, "extraction_ocr": {"score": 1.0, "response_time": 4.255826950073242, "result": "[{'name': 'MARY THOMAS', 'time_per_day': 1, 'medication': 'ATENOLOL', 'dosage': 100, 'rx_number': '1234567-12345'}]"}, "math_ocr": {"score": 1.0, "response_time": 2.5475921630859375, "result": "3x^2-6x+2"}, "object_detection": {"score": 0.18220793140407288, "response_time": 5.249211311340332, "result": "{'x': 0.4, 'y': 0.2, 'width': 0.2, 'height': 0.4}"}}
+{"zero_shot_classification": {"score": 1, "response_time": 1.8139240741729736, "result": "Toyota Camry"}, "count_fruit": {"score": 0, "response_time": 2.2961227893829346, "result": "9"}, "document_ocr": {"score": 1, "response_time": 3.707613945007324, "result": "I was thinking earlier today that I have gone through, to use the lingo, eras of listening to each of Swift's Eras. Meta indeed. I started listening to Ms. Swift's music after hearing the Midnights album. A few weeks after hearing the album for the first time, I found myself playing various songs on repeat. I listened to the album in order multiple times."}, "handwriting_ocr": {"score": 1, "response_time": 7.602664232254028, "result": "The words of songs on the album have been echoing in my head all week. \"Fades into the grey of my day old tea.\""}, "extraction_ocr": {"score": 1.0, "response_time": 3.052698850631714, "result": "[{'name': 'MARY THOMAS', 'time_per_day': 1, 'medication': 'ATENOLOL', 'dosage': 100, 'rx_number': '1234567-12345'}]"}, "math_ocr": {"score": 0.9, "response_time": 2.2950241565704346, "result": "3x^{2}-6x+2"}, "object_detection": {"score": 0, "response_time": 2.439383029937744, "result": "Failed to produce a valid JSON output"}, "set_of_mark": {"score": 0.4782608695652174, "response_time": 4.843470096588135, "result": "```json\n[\n  10, \n  11, \n  12, \n  13, \n  14, \n  15, \n  16, \n  17, \n  18, \n  19, \n  20, \n  21, \n  22, \n  23, \n  24, \n  25, \n  26, \n  27\n]\n```"}}

template.html

@@ -55,7 +55,7 @@ <h2>Response Time</h2>
                         </div>
                         <div class="chart">
                             <div class="chart_box chart_box_green">
-                                <p>{{results['avg_time']}} ms</p>
+                                <p>{{results['avg_time']}} s</p>
                             </div>
                         </div>
                     </div>

web.py

@@ -159,7 +159,7 @@ def extraction_ocr():
 
     code_regex = r'```[a-zA-Z]*\n(.*?)\n```'
     code_blocks = re.findall(code_regex,result, re.DOTALL)
-
+    if(len(code_blocks) == 0): return 0, inference_time, "Failed to produce a valid JSON output"
     answer_array = json.loads(code_blocks[0])
 
     correct_array = [
@@ -190,6 +190,7 @@ def math_ocr():
 
     code_regex = r'```[a-zA-Z]*\n(.*?)\n```'
     code_blocks = re.findall(code_regex,result, re.DOTALL)
+    if(len(code_blocks) == 0): return 0, inference_time, "Failed to produce a valid JSON output"
     answer_array = json.loads(code_blocks[0])
     answer_equation = answer_array[0].replace(" ", "")
 
@@ -213,6 +214,7 @@ def object_detection():
 
     code_regex = r'```[a-zA-Z]*\n(.*?)\n```'
     code_blocks = re.findall(code_regex, result, re.DOTALL)
+    if(len(code_blocks) == 0): return 0, inference_time, "Failed to produce a valid JSON output"
     answer = json.loads(code_blocks[0])
 
     correct = {'x': 0.465, 'y': 0.42, 'width': 0.37, 'height': 0.38}
@@ -232,31 +234,32 @@ def object_detection():
     return iou, inference_time, str(answer)
 
 def set_of_mark():
-  base_model = GPT4V(
-      ontology=CaptionOntology({"none": "none"}),
-      api_key=os.environ["OPENAI_API_KEY"],
-  )
+    base_model = GPT4V(
+        ontology=CaptionOntology({"none": "none"}),
+        api_key=os.environ["OPENAI_API_KEY"],
+    )
 
-  result, inference_time = base_model.predict(
-      "images/fruits_som.png",
-      classes=[],
-      result_serialization="text",
-      prompt="Find all the fruits in this image and return a JSON array of all the applicable numbers.",
-  )
+    result, inference_time = base_model.predict(
+        "images/fruits_som.png",
+        classes=[],
+        result_serialization="text",
+        prompt="Find all the fruits in this image and return a JSON array of all the applicable numbers.",
+    )
 
-  code_regex = r'```[a-zA-Z]*\n(.*?)\n```'
-  code_blocks = re.findall(code_regex,result, re.DOTALL)
-  answer = json.loads(code_blocks[0])
+    code_regex = r'```[a-zA-Z]*\n(.*?)\n```'
+    code_blocks = re.findall(code_regex, result, re.DOTALL)
+    if(len(code_blocks) == 0): return 0, inference_time, "Failed to produce a valid JSON output"
+    answer = json.loads(code_blocks[0])
 
-  correct = [35,40,26,2,13,17,29,21,10,42,8,43,0,11,7,4,12,27,37,39,22,15,25]
+    correct = [35,40,26,2,13,17,29,21,10,42,8,43,0,11,7,4,12,27,37,39,22,15,25]
 
-  score = 0
-  for guess in answer:
-      if guess in correct: score += 1
+    score = 0
+    for guess in answer:
+        if guess in correct: score += 1
 
-  accuracy = score/len(correct)
+    accuracy = score/len(correct)
 
-  return accuracy, inference_time, result
+    return accuracy, inference_time, result
 
 
 tests = [