Training in progress epoch 22

README.md

@@ -14,24 +14,30 @@ 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: 0.2566
-- Validation Loss: 0.6314
-- Validation Mean Iou: 0.3271
-- Validation Mean Accuracy: 0.3947
-- Validation Overall Accuracy: 0.8466
-- Validation Per Category Iou: [0.         0.70998525 0.86678486 0.49194558 0.5537256  0.33629647
-        nan 0.41985347 0.5135536  0.04026674 0.81694759 0.
- 0.                nan 0.         0.57746551 0.         0.
- 0.69225859 0.14474648 0.39427342 0.33636012 0.                nan
- 0.         0.35794269 0.17788878 0.         0.85568704 0.8282232
- 0.91236012 0.01923597 0.18998213 0.23102913 0.        ]
-- Validation Per Category Accuracy: [0.         0.80619283 0.95245983 0.57379426 0.72315118 0.42797816
-        nan 0.57649234 0.7130394  0.04452638 0.92405685 0.
- 0.                nan 0.         0.70041704 0.         0.
- 0.89411357 0.17005455 0.48671835 0.37690491 0.                nan
- 0.         0.43708457 0.22123613 0.         0.95864813 0.92771201
- 0.96946207 0.0313183  0.43519983 0.27845097 0.        ]
-- Epoch: 21
+- Train Loss: 0.2332
+- Validation Loss: 0.6204
+- Validation Mean Iou: 0.3332
+- Validation Mean Accuracy: 0.4059
+- Validation Overall Accuracy: 0.8483
+- Validation Per Category Iou: [0.00000000e+00 7.35393964e-01 8.60690644e-01 5.19021773e-01
+ 5.30969013e-01 3.58918972e-01            nan 4.20497772e-01
+ 4.95716847e-01 1.37272913e-02 8.19133033e-01 5.46224225e-04
+ 0.00000000e+00            nan 0.00000000e+00 5.68465477e-01
+ 0.00000000e+00 0.00000000e+00 7.20011345e-01 1.42622295e-01
+ 4.17365004e-01 3.82529649e-01 0.00000000e+00            nan
+ 6.70870790e-05 3.60538230e-01 1.69147635e-01 0.00000000e+00
+ 8.55389638e-01 8.18918753e-01 9.17882158e-01 1.64044825e-02
+ 2.01049360e-01 3.37421484e-01 0.00000000e+00]
+- Validation Per Category Accuracy: [0.00000000e+00 8.46018595e-01 9.32405953e-01 6.36213716e-01
+ 6.56514609e-01 5.05878965e-01            nan 5.80474598e-01
+ 7.45257836e-01 1.52580828e-02 9.27215688e-01 6.27598651e-04
+ 0.00000000e+00            nan 0.00000000e+00 7.09744281e-01
+ 0.00000000e+00 0.00000000e+00 8.72681917e-01 1.56351474e-01
+ 5.64231108e-01 4.44759880e-01 0.00000000e+00            nan
+ 1.40173816e-04 4.49791012e-01 2.10332805e-01 0.00000000e+00
+ 9.49627606e-01 9.31113143e-01 9.68909254e-01 3.53635859e-02
+ 4.13840545e-01 4.36765484e-01 0.00000000e+00]
+- Epoch: 22
 
 ## Model description
 
@@ -323,6 +329,23 @@ The following hyperparameters were used during training:
  0.89411357 0.17005455 0.48671835 0.37690491 0.                nan
  0.         0.43708457 0.22123613 0.         0.95864813 0.92771201
  0.96946207 0.0313183  0.43519983 0.27845097 0.        ]                                                                                                                                                                | 21    |
+| 0.2332     | 0.6204          | 0.3332              | 0.4059                   | 0.8483                      | [0.00000000e+00 7.35393964e-01 8.60690644e-01 5.19021773e-01
+ 5.30969013e-01 3.58918972e-01            nan 4.20497772e-01
+ 4.95716847e-01 1.37272913e-02 8.19133033e-01 5.46224225e-04
+ 0.00000000e+00            nan 0.00000000e+00 5.68465477e-01
+ 0.00000000e+00 0.00000000e+00 7.20011345e-01 1.42622295e-01
+ 4.17365004e-01 3.82529649e-01 0.00000000e+00            nan
+ 6.70870790e-05 3.60538230e-01 1.69147635e-01 0.00000000e+00
+ 8.55389638e-01 8.18918753e-01 9.17882158e-01 1.64044825e-02
+ 2.01049360e-01 3.37421484e-01 0.00000000e+00] | [0.00000000e+00 8.46018595e-01 9.32405953e-01 6.36213716e-01
+ 6.56514609e-01 5.05878965e-01            nan 5.80474598e-01
+ 7.45257836e-01 1.52580828e-02 9.27215688e-01 6.27598651e-04
+ 0.00000000e+00            nan 0.00000000e+00 7.09744281e-01
+ 0.00000000e+00 0.00000000e+00 8.72681917e-01 1.56351474e-01
+ 5.64231108e-01 4.44759880e-01 0.00000000e+00            nan
+ 1.40173816e-04 4.49791012e-01 2.10332805e-01 0.00000000e+00
+ 9.49627606e-01 9.31113143e-01 9.68909254e-01 3.53635859e-02
+ 4.13840545e-01 4.36765484e-01 0.00000000e+00] | 22    |
 
 
 ### Framework versions

tf_model.h5

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4dca16d4e9e599544e5177e978461db45d4126d08dc7beefdf0cb9c38ad9ce33
+oid sha256:897c1a48620ba448828014cb6d4ead5d80a40480bdf98d35b15feba04bdc316f
 size 15167588