Update README.md
Browse files
README.md
CHANGED
|
@@ -16,7 +16,7 @@ tags:
|
|
| 16 |
This is a demo version of ImageQuality-R1 which is trained on the combination of KADID-10K, TID2013, and KONIQ-10K. The base model of ImageQuality-R1 is Qwen2.5-VL-7B-Instruct.
|
| 17 |
|
| 18 |
## Quick Start
|
| 19 |
-
```
|
| 20 |
from transformers import Qwen2_5_VLForConditionalGeneration, AutoTokenizer, AutoProcessor
|
| 21 |
from qwen_vl_utils import process_vision_info
|
| 22 |
|
|
@@ -97,11 +97,11 @@ random.seed(42)
|
|
| 97 |
device = torch.device("cuda:0") if torch.cuda.is_available() else torch.device("cpu")
|
| 98 |
|
| 99 |
### Modify here
|
| 100 |
-
|
| 101 |
image_path = ""
|
| 102 |
|
| 103 |
reasoning, score = score_image(
|
| 104 |
-
model_path=
|
| 105 |
image_path=image_path
|
| 106 |
)
|
| 107 |
|
|
|
|
| 16 |
This is a demo version of ImageQuality-R1 which is trained on the combination of KADID-10K, TID2013, and KONIQ-10K. The base model of ImageQuality-R1 is Qwen2.5-VL-7B-Instruct.
|
| 17 |
|
| 18 |
## Quick Start
|
| 19 |
+
```python
|
| 20 |
from transformers import Qwen2_5_VLForConditionalGeneration, AutoTokenizer, AutoProcessor
|
| 21 |
from qwen_vl_utils import process_vision_info
|
| 22 |
|
|
|
|
| 97 |
device = torch.device("cuda:0") if torch.cuda.is_available() else torch.device("cpu")
|
| 98 |
|
| 99 |
### Modify here
|
| 100 |
+
model_path = ""
|
| 101 |
image_path = ""
|
| 102 |
|
| 103 |
reasoning, score = score_image(
|
| 104 |
+
model_path=model_path,
|
| 105 |
image_path=image_path
|
| 106 |
)
|
| 107 |
|