Trained on *correctly* labeled 448*224 data.

README.md

@@ -15,7 +15,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [microsoft/kosmos-2-patch14-224](https://huggingface.co/microsoft/kosmos-2-patch14-224) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.0708
+- Loss: 0.0458
 
 ## Model description
 
@@ -49,12 +49,12 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
-| No log        | 0.0558 | 200  | 0.1314          |
-| 1.0995        | 0.1117 | 400  | 0.1034          |
-| 0.1154        | 0.1675 | 600  | 0.0912          |
-| 0.097         | 0.2234 | 800  | 0.0785          |
-| 0.0849        | 0.2792 | 1000 | 0.0728          |
-| 0.0849        | 0.3351 | 1200 | 0.0708          |
+| No log        | 0.0497 | 200  | 0.1075          |
+| 1.077         | 0.0993 | 400  | 0.0816          |
+| 0.0909        | 0.1490 | 600  | 0.0634          |
+| 0.0731        | 0.1986 | 800  | 0.0531          |
+| 0.0626        | 0.2483 | 1000 | 0.0482          |
+| 0.0626        | 0.2979 | 1200 | 0.0458          |
 
 
 ### Framework versions

model-00001-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8f52463b357ccd7b0855caf74ceed0b625099f929e62896a087a59db8b01b823
+oid sha256:c008b367d97d64d45bc7f97ffc1a6376a7d9b4933c12fd33e34fbf9387004925
 size 4999738624

model-00002-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9cf86183b70a488d46fe8915b128fe662dce93d9f74f26110f5a5e6a165f37cf
+oid sha256:bd2f1b1b7c7cc4cd69cbabab6a1944e764e4ebfba704b32ec3d5246e610d5b28
 size 1658313704