Spaces:
Running
on
Zero
Running
on
Zero
Commit
•
3666985
1
Parent(s):
8fa6ab2
update app.py
Browse files- __pycache__/utils.cpython-310.pyc +0 -0
- app.py +4 -3
- weights/icon_detect/best.pt +3 -0
__pycache__/utils.cpython-310.pyc
ADDED
|
Binary file (14.3 kB). View file
|
|
|
app.py
CHANGED
|
@@ -6,7 +6,7 @@ import torch
|
|
| 6 |
from PIL import Image
|
| 7 |
import io
|
| 8 |
|
| 9 |
-
|
| 10 |
import base64, os
|
| 11 |
from utils import check_ocr_box, get_yolo_model, get_caption_model_processor, get_som_labeled_img
|
| 12 |
import torch
|
|
@@ -53,8 +53,9 @@ OmniParser is a screen parsing tool to convert general GUI screen to structured
|
|
| 53 |
DEVICE = torch.device('cuda')
|
| 54 |
|
| 55 |
# @spaces.GPU
|
| 56 |
-
|
| 57 |
-
|
|
|
|
| 58 |
def process(
|
| 59 |
image_input,
|
| 60 |
box_threshold,
|
|
|
|
| 6 |
from PIL import Image
|
| 7 |
import io
|
| 8 |
|
| 9 |
+
import spaces
|
| 10 |
import base64, os
|
| 11 |
from utils import check_ocr_box, get_yolo_model, get_caption_model_processor, get_som_labeled_img
|
| 12 |
import torch
|
|
|
|
| 53 |
DEVICE = torch.device('cuda')
|
| 54 |
|
| 55 |
# @spaces.GPU
|
| 56 |
+
@torch.inference_mode()
|
| 57 |
+
@torch.autocast(device_type="cuda", dtype=torch.bfloat16)
|
| 58 |
+
@spaces.GPU
|
| 59 |
def process(
|
| 60 |
image_input,
|
| 61 |
box_threshold,
|
weights/icon_detect/best.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:929baa2512022543b67485143ca20a2d556183043bd7ec9fff642b62ebf7723e
|
| 3 |
+
size 12224434
|