diff --git a/index.html b/index.html index 9349f38..c1da49f 100644 --- a/index.html +++ b/index.html @@ -40,7 +40,7 @@

How's GPT-4 with Vision Doing?

You can contribute your own tests, too! See the GitHub README for contributing instructions.

-

Tests are run every day at 1am PT. Last updated December 27, 2023.

+

Tests are run every day at 1am PT. Last updated December 28, 2023.

Made with ❤️ by the team at Roboflow.

@@ -58,12 +58,12 @@

How's GPT-4 with Vision Doing?

Response Time

-

Today, the average response time to receive results from our tests was 5.52 seconds per request.

+

Today, the average response time to receive results from our tests was 5.42 seconds per request.

This number only accounts for requests made by this application.

-

5.52 s

+

5.42 s

@@ -122,7 +122,7 @@

Prompt

Image

Image of the input into GPT-4

Result

- 
9
+ 
7

Test submitted by Roboflow

@@ -162,7 +162,7 @@

Object Detection

Of the last 7 tests, conducted daily, this test has passed 0% of the time.

-

Today's request cost $0.009

+

Today's request cost $0.01

@@ -176,7 +176,7 @@

Prompt

Image

Image of the input into GPT-4

Result

- 
{'x': 0.34, 'y': 0.23, 'width': 0.16, 'height': 0.46}
+ 
{'x': 0.25, 'y': 0.26, 'width': 0.22, 'height': 0.52}

Test submitted by Roboflow

@@ -216,7 +216,7 @@

Graph Understanding

Of the last 7 tests, conducted daily, this test has passed 0% of the time.

-

Today's request cost $0.01

+

Today's request cost $0.011

@@ -232,10 +232,22 @@

Image

Result

```json
 {
-  "A": {"quantity": 10, "price": 15},
-  "B": {"quantity": 20, "price": 25},
-  "C": {"quantity": 30, "price": 35},
-  "D": {"quantity": 40, "price": 45}
+  "A": {
+    "quantity": 15,
+    "price": 5
+  },
+  "B": {
+    "quantity": 27,
+    "price": 18
+  },
+  "C": {
+    "quantity": 33,
+    "price": 30
+  },
+  "D": {
+    "quantity": 35,
+    "price": 40
+  }
 }
 ```

Test submitted by Roboflow

@@ -411,8 +423,8 @@

Image

Result

```json
 {
-  "length": 3.0,
-  "width": 3.0
+  "length": 2.5,
+  "width": 2.5
 }
 ```

Test submitted by Roboflow

@@ -639,7 +651,7 @@

Prompt

Image

Image of the input into GPT-4

Result

- 
[{'name': 'MARY THOMAS', 'time_per_day': 1, 'medication': 'ATENOLOL', 'dosage': 100, 'rx_number': '1234567-12345'}]
+ 
[{'name': 'Mary Thomas', 'time_per_day': 1, 'medication': 'Atenolol', 'dosage': 100, 'rx_number': '1234567-12345'}]

Test submitted by Roboflow

diff --git a/results/2023-12-28.json b/results/2023-12-28.json new file mode 100644 index 0000000..b8b6650 --- /dev/null +++ b/results/2023-12-28.json @@ -0,0 +1,90 @@ +{ + "zero_shot_classification": { + "score": 1, + "success": true, + "price": 0.00481, + "pass_fail": "Pass", + "response_time": 2.7964532375335693, + "result": "Toyota Camry" + }, + "count_fruit": { + "score": 0, + "success": false, + "price": 0.007870000000000002, + "pass_fail": "Fail", + "response_time": 1.9929437637329102, + "result": "7" + }, + "document_ocr": { + "score": 1, + "success": true, + "price": 0.00857, + "pass_fail": "Pass", + "response_time": 2.5470097064971924, + "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, + "success": true, + "price": 0.008730000000000002, + "pass_fail": "Pass", + "response_time": 4.106041669845581, + "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, + "success": true, + "price": 0.00719, + "pass_fail": "Pass", + "response_time": 3.9414350986480713, + "result": "[{'name': 'Mary Thomas', 'time_per_day': 1, 'medication': 'Atenolol', 'dosage': 100, 'rx_number': '1234567-12345'}]" + }, + "math_ocr": { + "score": 1.0, + "success": true, + "price": 0.01528, + "pass_fail": "Pass", + "response_time": 3.2695353031158447, + "result": "3x^2-6x+2" + }, + "object_detection": { + "score": 0.10008628127696285, + "success": false, + "price": 0.010480000000000001, + "pass_fail": "Fail", + "response_time": 7.456548690795898, + "result": "{'x': 0.25, 'y': 0.26, 'width': 0.22, 'height': 0.52}" + }, + "graph_understanding": { + "score": 0.9149999999999999, + "success": false, + "price": 0.01079, + "pass_fail": "Fail", + "response_time": 4.398157358169556, + "result": "```json\n{\n \"A\": {\n \"quantity\": 15,\n \"price\": 5\n },\n \"B\": {\n \"quantity\": 27,\n \"price\": 18\n },\n \"C\": {\n \"quantity\": 33,\n \"price\": 30\n },\n \"D\": {\n \"quantity\": 35,\n \"price\": 40\n }\n}\n```" + }, + "color_recognition": { + "score": 0.8941176470588236, + "success": false, + "price": 0.008870000000000001, + "pass_fail": "Fail", + "response_time": 2.8069121837615967, + "result": "```json\n{\n \"R\": 128,\n \"G\": 0,\n \"B\": 128\n}\n```" + }, + "annotation_qa": { + "score": 0.33333333333333337, + "success": false, + "price": 0.015300000000000001, + "pass_fail": "Fail", + "response_time": 3.415327548980713, + "result": "```json\n{\n \"missing\": 1\n}\n```" + }, + "measurement": { + "score": 0.7142857142857143, + "success": false, + "price": 0.00877, + "pass_fail": "Fail", + "response_time": 3.6652400493621826, + "result": "```json\n{\n \"length\": 2.5,\n \"width\": 2.5\n}\n```" + } +} \ No newline at end of file