Training in progress epoch 4

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: 0.0507
-- Train Dice Coef: 0.5971
-- Train Iou: 0.4430
-- Validation Loss: 0.0459
-- Validation Dice Coef: 0.7780
-- Validation Iou: 0.6420
+- Train Loss: 0.0443
+- Train Dice Coef: 0.6379
+- Train Iou: 0.4837
+- Validation Loss: 0.0393
+- Validation Dice Coef: 0.8008
+- Validation Iou: 0.6709
 - Train Lr: 1e-04
-- Epoch: 3
+- Epoch: 4
 
 ## Model description
 
@@ -51,6 +51,7 @@ The following hyperparameters were used during training:
 | 0.0805     | 0.4647          | 0.3202    | 0.0690          | 0.6569               | 0.4990         | 1e-04    | 1     |
 | 0.0582     | 0.5595          | 0.4069    | 0.0509          | 0.7745               | 0.6369         | 1e-04    | 2     |
 | 0.0507     | 0.5971          | 0.4430    | 0.0459          | 0.7780               | 0.6420         | 1e-04    | 3     |
+| 0.0443     | 0.6379          | 0.4837    | 0.0393          | 0.8008               | 0.6709         | 1e-04    | 4     |
 
 
 ### Framework versions

tf_model.h5

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e91bf132428c2bde09c71febd571048bb52a621c62e397e5d83644663b9f44aa
+oid sha256:dfd1b9f1b2df0fda28303aaf27a2953b90aaa8168920d59fe828c0b2b050442b
 size 15133664