Spaces:
Runtime error
Runtime error
Update app.py
Browse files
app.py
CHANGED
|
@@ -3,6 +3,8 @@ from transformers import AutoProcessor, AutoTokenizer, AutoImageProcessor, AutoM
|
|
| 3 |
import torch
|
| 4 |
import open_clip
|
| 5 |
|
|
|
|
|
|
|
| 6 |
torch.hub.download_url_to_file('http://images.cocodataset.org/val2017/000000039769.jpg', 'cats.jpg')
|
| 7 |
torch.hub.download_url_to_file('https://huggingface.co/datasets/nielsr/textcaps-sample/resolve/main/stop_sign.png', 'stop_sign.png')
|
| 8 |
torch.hub.download_url_to_file('https://cdn.openai.com/dall-e-2/demos/text2im/astronaut/horse/photo/0.jpg', 'astronaut.jpg')
|
|
@@ -23,9 +25,11 @@ vitgpt_processor = AutoImageProcessor.from_pretrained("nlpconnect/vit-gpt2-image
|
|
| 23 |
vitgpt_model = VisionEncoderDecoderModel.from_pretrained("nlpconnect/vit-gpt2-image-captioning")
|
| 24 |
vitgpt_tokenizer = AutoTokenizer.from_pretrained("nlpconnect/vit-gpt2-image-captioning")
|
| 25 |
|
|
|
|
|
|
|
| 26 |
coca_model, _, coca_transform = open_clip.create_model_and_transforms(
|
| 27 |
"coca_ViT-L-14",
|
| 28 |
-
pretrained=
|
| 29 |
)
|
| 30 |
|
| 31 |
device = "cuda" if torch.cuda.is_available() else "cpu"
|
|
|
|
| 3 |
import torch
|
| 4 |
import open_clip
|
| 5 |
|
| 6 |
+
from huggingface_hub import hf_hub_download
|
| 7 |
+
|
| 8 |
torch.hub.download_url_to_file('http://images.cocodataset.org/val2017/000000039769.jpg', 'cats.jpg')
|
| 9 |
torch.hub.download_url_to_file('https://huggingface.co/datasets/nielsr/textcaps-sample/resolve/main/stop_sign.png', 'stop_sign.png')
|
| 10 |
torch.hub.download_url_to_file('https://cdn.openai.com/dall-e-2/demos/text2im/astronaut/horse/photo/0.jpg', 'astronaut.jpg')
|
|
|
|
| 25 |
vitgpt_model = VisionEncoderDecoderModel.from_pretrained("nlpconnect/vit-gpt2-image-captioning")
|
| 26 |
vitgpt_tokenizer = AutoTokenizer.from_pretrained("nlpconnect/vit-gpt2-image-captioning")
|
| 27 |
|
| 28 |
+
filepath = hf_hub_download(repo_id="gpucce/CoCa", filename=""laion2B-s13B-b90k-mscoco-2014.pt")
|
| 29 |
+
|
| 30 |
coca_model, _, coca_transform = open_clip.create_model_and_transforms(
|
| 31 |
"coca_ViT-L-14",
|
| 32 |
+
pretrained=filepath,
|
| 33 |
)
|
| 34 |
|
| 35 |
device = "cuda" if torch.cuda.is_available() else "cpu"
|