updated recall val
Browse files
README.md
CHANGED
|
@@ -23,7 +23,9 @@ This model aligns medical images (X-rays, CTs, MRIs) with their textual descript
|
|
| 23 |
## Performance (Test Set)
|
| 24 |
- **Batch-wise Recall@1:** 70.83% (State-of-the-art for T4 fine-tuning)
|
| 25 |
- **Batch-wise Recall@5:** 96.99%
|
| 26 |
-
- **Global Retrieval Recall@
|
|
|
|
|
|
|
| 27 |
|
| 28 |
## Usage
|
| 29 |
|
|
|
|
| 23 |
## Performance (Test Set)
|
| 24 |
- **Batch-wise Recall@1:** 70.83% (State-of-the-art for T4 fine-tuning)
|
| 25 |
- **Batch-wise Recall@5:** 96.99%
|
| 26 |
+
- **Global Retrieval Recall@1:** ~6% (500x better than random chance)
|
| 27 |
+
- **Global Retrieval Recall@5:** ~16%
|
| 28 |
+
Though a lot of work need to be done on this as the recall is still quite low. It will be updated with newer version
|
| 29 |
|
| 30 |
## Usage
|
| 31 |
|