Training in progress epoch 17

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: 0.2985
-- Validation Loss: 0.6752
-- Validation Mean Iou: 0.3106
-- Validation Mean Accuracy: 0.3773
-- Validation Overall Accuracy: 0.8330
-- Validation Per Category Iou: [0.         0.70132108 0.84614259 0.46333382 0.56218727 0.24626879
-        nan 0.40366902 0.47740097 0.0250144  0.80566818 0.
- 0.                nan 0.         0.53413689 0.         0.
- 0.65726308 0.08036975 0.36642674 0.26539948 0.                nan
- 0.         0.35749813 0.13576539 0.         0.84624546 0.80479301
- 0.90656978 0.00419358 0.21944385 0.22915907 0.        ]
-- Validation Per Category Accuracy: [0.         0.75132765 0.94954979 0.66569367 0.76524644 0.28265659
-        nan 0.5278041  0.66034967 0.0295519  0.90881823 0.
- 0.                nan 0.         0.62968235 0.         0.
- 0.91281232 0.08130926 0.483296   0.29302424 0.                nan
- 0.         0.44332482 0.15759113 0.         0.93912247 0.92471332
- 0.96940512 0.0103905  0.42560363 0.26159798 0.        ]
-- Epoch: 16
+- Train Loss: 0.2833
+- Validation Loss: 0.6446
+- Validation Mean Iou: 0.3261
+- Validation Mean Accuracy: 0.3949
+- Validation Overall Accuracy: 0.8447
+- Validation Per Category Iou: [0.00000000e+00 7.08995850e-01 8.49245624e-01 5.02729624e-01
+ 5.44178924e-01 2.80874020e-01            nan 4.09531940e-01
+ 4.68642701e-01 5.72334850e-02 7.86049035e-01 4.15253651e-04
+ 0.00000000e+00            nan 0.00000000e+00 5.20689462e-01
+ 0.00000000e+00 0.00000000e+00 7.19140232e-01 1.09126891e-01
+ 4.66887817e-01 3.77114972e-01 0.00000000e+00            nan
+ 0.00000000e+00 3.57675941e-01 1.49472418e-01 0.00000000e+00
+ 8.48995042e-01 8.23287514e-01 9.17986999e-01 4.86177781e-03
+ 2.08704959e-01 3.22665861e-01 0.00000000e+00]
+- Validation Per Category Accuracy: [0.00000000e+00 8.00869189e-01 9.55886409e-01 6.05400396e-01
+ 6.79327160e-01 3.35582853e-01            nan 5.64117225e-01
+ 7.19032298e-01 7.06182643e-02 9.12828794e-01 4.70698988e-04
+ 0.00000000e+00            nan 0.00000000e+00 6.96671162e-01
+ 0.00000000e+00 0.00000000e+00 8.85781041e-01 1.13699831e-01
+ 6.05324120e-01 4.37527674e-01 0.00000000e+00            nan
+ 0.00000000e+00 4.41345654e-01 1.73312203e-01 0.00000000e+00
+ 9.54319708e-01 8.94089174e-01 9.60961557e-01 6.44308877e-03
+ 4.22542478e-01 3.99729132e-01 0.00000000e+00]
+- Epoch: 17
 
 ## Model description
 
@@ -262,6 +268,23 @@ The following hyperparameters were used during training:
  0.91281232 0.08130926 0.483296   0.29302424 0.                nan
  0.         0.44332482 0.15759113 0.         0.93912247 0.92471332
  0.96940512 0.0103905  0.42560363 0.26159798 0.        ]                                                                                                                                                                | 16    |
+| 0.2833     | 0.6446          | 0.3261              | 0.3949                   | 0.8447                      | [0.00000000e+00 7.08995850e-01 8.49245624e-01 5.02729624e-01
+ 5.44178924e-01 2.80874020e-01            nan 4.09531940e-01
+ 4.68642701e-01 5.72334850e-02 7.86049035e-01 4.15253651e-04
+ 0.00000000e+00            nan 0.00000000e+00 5.20689462e-01
+ 0.00000000e+00 0.00000000e+00 7.19140232e-01 1.09126891e-01
+ 4.66887817e-01 3.77114972e-01 0.00000000e+00            nan
+ 0.00000000e+00 3.57675941e-01 1.49472418e-01 0.00000000e+00
+ 8.48995042e-01 8.23287514e-01 9.17986999e-01 4.86177781e-03
+ 2.08704959e-01 3.22665861e-01 0.00000000e+00] | [0.00000000e+00 8.00869189e-01 9.55886409e-01 6.05400396e-01
+ 6.79327160e-01 3.35582853e-01            nan 5.64117225e-01
+ 7.19032298e-01 7.06182643e-02 9.12828794e-01 4.70698988e-04
+ 0.00000000e+00            nan 0.00000000e+00 6.96671162e-01
+ 0.00000000e+00 0.00000000e+00 8.85781041e-01 1.13699831e-01
+ 6.05324120e-01 4.37527674e-01 0.00000000e+00            nan
+ 0.00000000e+00 4.41345654e-01 1.73312203e-01 0.00000000e+00
+ 9.54319708e-01 8.94089174e-01 9.60961557e-01 6.44308877e-03
+ 4.22542478e-01 3.99729132e-01 0.00000000e+00] | 17    |
 
 
 ### Framework versions

tf_model.h5

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5bb4d112c49fff6b92e7be400fe9d308eec7243baf9c4cd8f32c8c1eb0509294
+oid sha256:f4f04e5cd78114c77130df86a4314b9347c65685acd01b9c3e0bbbac8cb576f4
 size 15167588