gauthambalraj07@gmail.com commited on
Commit
3dc870c
·
1 Parent(s): 707dfbd

fix yolo call

Browse files
Files changed (4) hide show
  1. main.py +3 -0
  2. requirements.txt +1 -1
  3. test.ipynb +4 -4
  4. utils.py +7 -12
main.py CHANGED
@@ -36,6 +36,9 @@ try:
36
  except:
37
  yolo_model = torch.load("weights/icon_detect/best.pt", map_location="cpu", weights_only=False)["model"]
38
 
 
 
 
39
  from transformers import AutoProcessor, AutoModelForCausalLM
40
  import torch
41
 
 
36
  except:
37
  yolo_model = torch.load("weights/icon_detect/best.pt", map_location="cpu", weights_only=False)["model"]
38
 
39
+
40
+ print(f"YOLO model type: {type(yolo_model)}")
41
+
42
  from transformers import AutoProcessor, AutoModelForCausalLM
43
  import torch
44
 
requirements.txt CHANGED
@@ -1,4 +1,4 @@
1
- torch==2.0.1
2
  easyocr
3
  torchvision
4
  supervision==0.18.0
 
1
+ torch==2.2.2
2
  easyocr
3
  torchvision
4
  supervision==0.18.0
test.ipynb CHANGED
@@ -2,7 +2,7 @@
2
  "cells": [
3
  {
4
  "cell_type": "code",
5
- "execution_count": 3,
6
  "metadata": {},
7
  "outputs": [
8
  {
@@ -11,8 +11,8 @@
11
  "text": [
12
  "Processing imscdr_ac_in___8.png...\n",
13
  "Status Code: 500\n",
14
- "Response Text: {\"detail\":\"BaseModel.predict() got an unexpected keyword argument 'source'\"}\n",
15
- "Finished processing imscdr_ac_in___8.png in 43.56 seconds\n",
16
  "Processing imscdr_ac_in___9.png...\n",
17
  "Status Code: 500\n",
18
  "Response Text: <!DOCTYPE html>\n",
@@ -124,7 +124,7 @@
124
  "</body>\n",
125
  "</html>\n",
126
  "\n",
127
- "Finished processing imscdr_ac_in___9.png in 9.98 seconds\n",
128
  "{}\n"
129
  ]
130
  }
 
2
  "cells": [
3
  {
4
  "cell_type": "code",
5
+ "execution_count": 4,
6
  "metadata": {},
7
  "outputs": [
8
  {
 
11
  "text": [
12
  "Processing imscdr_ac_in___8.png...\n",
13
  "Status Code: 500\n",
14
+ "Response Text: {\"detail\":\"BaseModel.predict() got an unexpected keyword argument 'conf'\"}\n",
15
+ "Finished processing imscdr_ac_in___8.png in 10.38 seconds\n",
16
  "Processing imscdr_ac_in___9.png...\n",
17
  "Status Code: 500\n",
18
  "Response Text: <!DOCTYPE html>\n",
 
124
  "</body>\n",
125
  "</html>\n",
126
  "\n",
127
+ "Finished processing imscdr_ac_in___9.png in 10.32 seconds\n",
128
  "{}\n"
129
  ]
130
  }
utils.py CHANGED
@@ -378,19 +378,14 @@ def predict_yolo(model, image_path, box_threshold, imgsz, scale_img, iou_thresho
378
  """ Use huggingface model to replace the original model
379
  """
380
  # model = model['model']
 
 
 
 
 
381
  if scale_img:
382
- result = model(
383
- image_path,
384
- conf=box_threshold,
385
- imgsz=imgsz,
386
- iou=iou_threshold, # default 0.7
387
- )
388
- else:
389
- result = model(
390
- image_path,
391
- conf=box_threshold,
392
- iou=iou_threshold, # default 0.7
393
- )
394
  boxes = result[0].boxes.xyxy#.tolist() # in pixel space
395
  conf = result[0].boxes.conf
396
  phrases = [str(i) for i in range(len(boxes))]
 
378
  """ Use huggingface model to replace the original model
379
  """
380
  # model = model['model']
381
+
382
+ kwargs = {
383
+ 'conf': box_threshold,
384
+ 'iou': iou_threshold,
385
+ }
386
  if scale_img:
387
+ kwargs['imgsz'] = imgsz
388
+ result = model(image_path, **kwargs)
 
 
 
 
 
 
 
 
 
 
389
  boxes = result[0].boxes.xyxy#.tolist() # in pixel space
390
  conf = result[0].boxes.conf
391
  phrases = [str(i) for i in range(len(boxes))]