noamrot commited on
Commit
aaefaa5
1 Parent(s): 3fa9cd1

demo almost done

Browse files
BLIP/models/__pycache__/__init__.cpython-37.pyc ADDED
Binary file (170 Bytes). View file
 
BLIP/models/__pycache__/blip.cpython-37.pyc ADDED
Binary file (7.18 kB). View file
 
BLIP/models/__pycache__/med.cpython-37.pyc ADDED
Binary file (28.2 kB). View file
 
BLIP/models/__pycache__/vit.cpython-37.pyc ADDED
Binary file (12.3 kB). View file
 
BLIP/models/blip.py CHANGED
@@ -8,8 +8,8 @@
8
  import warnings
9
  warnings.filterwarnings("ignore")
10
 
11
- from models.vit import VisionTransformer, interpolate_pos_embed
12
- from models.med import BertConfig, BertModel, BertLMHeadModel
13
  from transformers import BertTokenizer
14
 
15
  import torch
@@ -22,7 +22,7 @@ from timm.models.hub import download_cached_file
22
 
23
  class BLIP_Base(nn.Module):
24
  def __init__(self,
25
- med_config = 'configs/med_config.json',
26
  image_size = 224,
27
  vit = 'base',
28
  vit_grad_ckpt = False,
@@ -77,7 +77,7 @@ class BLIP_Base(nn.Module):
77
 
78
  class BLIP_Decoder(nn.Module):
79
  def __init__(self,
80
- med_config = 'configs/med_config.json',
81
  image_size = 384,
82
  vit = 'base',
83
  vit_grad_ckpt = False,
 
8
  import warnings
9
  warnings.filterwarnings("ignore")
10
 
11
+ from BLIP.models.vit import VisionTransformer, interpolate_pos_embed
12
+ from BLIP.models.med import BertConfig, BertModel, BertLMHeadModel
13
  from transformers import BertTokenizer
14
 
15
  import torch
 
22
 
23
  class BLIP_Base(nn.Module):
24
  def __init__(self,
25
+ med_config = 'BLIP/configs/med_config.json',
26
  image_size = 224,
27
  vit = 'base',
28
  vit_grad_ckpt = False,
 
77
 
78
  class BLIP_Decoder(nn.Module):
79
  def __init__(self,
80
+ med_config = 'BLIP/configs/med_config.json',
81
  image_size = 384,
82
  vit = 'base',
83
  vit_grad_ckpt = False,
app.py CHANGED
@@ -6,23 +6,19 @@
6
  # iface = gr.Interface(fn=greet, inputs="image", outputs="text")
7
  # iface.launch()
8
 
9
-
10
  import sys
11
- from IPython.display import display, HTML
12
  from BLIP.models.blip import blip_decoder
13
- from google_drive_downloader import GoogleDriveDownloader as gdd
14
  from PIL import Image
15
  import requests
16
  import torch
17
  from torchvision import transforms
18
  from torchvision.transforms.functional import InterpolationMode
19
  from urllib.parse import urlparse
20
- from google_drive_downloader import GoogleDriveDownloader as gdd
21
-
22
 
23
  device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
24
 
25
-
26
  transform = transforms.Compose([
27
  transforms.Resize((image_size,image_size),interpolation=InterpolationMode.BICUBIC),
28
  transforms.ToTensor(),
 
6
  # iface = gr.Interface(fn=greet, inputs="image", outputs="text")
7
  # iface.launch()
8
 
9
+ import gradio as gr
10
  import sys
 
11
  from BLIP.models.blip import blip_decoder
 
12
  from PIL import Image
13
  import requests
14
  import torch
15
  from torchvision import transforms
16
  from torchvision.transforms.functional import InterpolationMode
17
  from urllib.parse import urlparse
 
 
18
 
19
  device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
20
 
21
+ image_size = 384
22
  transform = transforms.Compose([
23
  transforms.Resize((image_size,image_size),interpolation=InterpolationMode.BICUBIC),
24
  transforms.ToTensor(),
flagged/log.csv ADDED
@@ -0,0 +1,2 @@
 
 
 
1
+ raw_image,Caption,flag,username,timestamp
2
+ /Users/snoamr/Documents/superCap/internet_page/demo/FuseCap/flagged/raw_image/tmpw95d_cla.jpg,a white dog celebrates its birthday with a colorful array of balloons and a cake on a white table the dog wears a pink hat and has an open mouth as it enjoys the festivities,,,2023-05-25 12:59:13.544370
flagged/raw_image/tmpw95d_cla.jpg ADDED
requirements.py CHANGED
@@ -3,5 +3,4 @@ transformers==4.15.0
3
  fairscale==0.4.4
4
  torch
5
  torchvision
6
- Pillow
7
- google_drive_downloader
 
3
  fairscale==0.4.4
4
  torch
5
  torchvision
6
+ Pillow