Spaces:
Sleeping
Sleeping
debug
Browse files
app.py
CHANGED
|
@@ -13,8 +13,18 @@ from utils import check_ocr_box, get_yolo_model, get_caption_model_processor, ge
|
|
| 13 |
import torch
|
| 14 |
from PIL import Image
|
| 15 |
|
| 16 |
-
yolo_model = get_yolo_model(model_path='weights/icon_detect/best.pt')
|
| 17 |
-
caption_model_processor = get_caption_model_processor(model_name="florence2", model_name_or_path="weights/icon_caption_florence")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18 |
platform = 'pc'
|
| 19 |
if platform == 'pc':
|
| 20 |
draw_bbox_config = {
|
|
@@ -51,10 +61,10 @@ MARKDOWN = """
|
|
| 51 |
OmniParser is a screen parsing tool to convert general GUI screen to structured elements.
|
| 52 |
"""
|
| 53 |
|
| 54 |
-
DEVICE = torch.device('cuda')
|
| 55 |
|
| 56 |
# @spaces.GPU
|
| 57 |
-
|
| 58 |
# @torch.autocast(device_type="cuda", dtype=torch.bfloat16)
|
| 59 |
@spaces.GPU(duration=65)
|
| 60 |
def process(
|
|
|
|
| 13 |
import torch
|
| 14 |
from PIL import Image
|
| 15 |
|
| 16 |
+
# yolo_model = get_yolo_model(model_path='weights/icon_detect/best.pt')
|
| 17 |
+
# caption_model_processor = get_caption_model_processor(model_name="florence2", model_name_or_path="weights/icon_caption_florence")
|
| 18 |
+
|
| 19 |
+
from ultralytics import YOLO
|
| 20 |
+
yolo_model = YOLO('weights/icon_detect/best.pt').to('cuda')
|
| 21 |
+
from transformers import AutoProcessor, AutoModelForCausalLM
|
| 22 |
+
processor = AutoProcessor.from_pretrained("microsoft/Florence-2-base", trust_remote_code=True)
|
| 23 |
+
model = AutoModelForCausalLM.from_pretrained("weights/icon_caption_florence", torch_dtype=torch.float16, trust_remote_code=True).to('cuda')
|
| 24 |
+
caption_model_processor = {'processor': processor, 'model': model}
|
| 25 |
+
print('finish loading model!!!')
|
| 26 |
+
|
| 27 |
+
|
| 28 |
platform = 'pc'
|
| 29 |
if platform == 'pc':
|
| 30 |
draw_bbox_config = {
|
|
|
|
| 61 |
OmniParser is a screen parsing tool to convert general GUI screen to structured elements.
|
| 62 |
"""
|
| 63 |
|
| 64 |
+
# DEVICE = torch.device('cuda')
|
| 65 |
|
| 66 |
# @spaces.GPU
|
| 67 |
+
@torch.inference_mode()
|
| 68 |
# @torch.autocast(device_type="cuda", dtype=torch.bfloat16)
|
| 69 |
@spaces.GPU(duration=65)
|
| 70 |
def process(
|