Training in progress epoch 3

README.md

@@ -14,14 +14,14 @@ probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [nvidia/mit-b0](https://huggingface.co/nvidia/mit-b0) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Train Loss: 0.0500
-- Train Dice Coef: 0.5831
-- Train Iou: 0.4322
-- Validation Loss: 0.0806
-- Validation Dice Coef: 0.5067
-- Validation Iou: 0.3659
+- Train Loss: 0.0432
+- Train Dice Coef: 0.6204
+- Train Iou: 0.4699
+- Validation Loss: 0.1085
+- Validation Dice Coef: 0.3757
+- Validation Iou: 0.2479
 - Train Lr: 1e-04
-- Epoch: 2
+- Epoch: 3
 
 ## Model description
 
@@ -50,6 +50,7 @@ The following hyperparameters were used during training:
 | 0.1460     | 0.3657          | 0.2410    | 0.0908          | 0.4603               | 0.3168         | 1e-04    | 0     |
 | 0.0610     | 0.5251          | 0.3760    | 0.1773          | 0.1542               | 0.0892         | 1e-04    | 1     |
 | 0.0500     | 0.5831          | 0.4322    | 0.0806          | 0.5067               | 0.3659         | 1e-04    | 2     |
+| 0.0432     | 0.6204          | 0.4699    | 0.1085          | 0.3757               | 0.2479         | 1e-04    | 3     |
 
 
 ### Framework versions

tf_model.h5

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9532e982b4a5cc4a148099a96bcc904f2150fe8d384c68da2db06d88eb8724bd
+oid sha256:cdec0b0e5d5f9e6638bbae56a4daa75f850ed4186463a30160c8a5a40ffbad6f
 size 15133664