Training in progress epoch 12

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: 0.3545
-- Validation Loss: 0.6001
-- Validation Mean Iou: 0.3183
-- Validation Mean Accuracy: 0.3876
-- Validation Overall Accuracy: 0.8403
-- Validation Per Category Iou: [0.         0.72446673 0.85906725 0.50307184 0.59402909 0.33935205
-        nan 0.36653095 0.49281565 0.         0.79716382 0.
- 0.                nan 0.         0.54502138 0.         0.
- 0.67058988 0.01229056 0.33404111 0.35615386 0.                nan
- 0.         0.3402916  0.13274148 0.         0.84876007 0.80426728
- 0.90568008 0.00767061 0.27079805 0.280819   0.        ]
-- Validation Per Category Accuracy: [0.         0.80252235 0.93835903 0.64483513 0.75981034 0.47343152
-        nan 0.51602703 0.68908551 0.         0.91483973 0.
- 0.                nan 0.         0.72268005 0.         0.
- 0.90438608 0.01229056 0.42963846 0.43749539 0.                nan
- 0.         0.40012652 0.15828843 0.         0.92988659 0.92495156
- 0.96015873 0.01578965 0.43451194 0.3352385  0.        ]
-- Epoch: 11
+- Train Loss: 0.3397
+- Validation Loss: 0.6115
+- Validation Mean Iou: 0.3192
+- Validation Mean Accuracy: 0.3868
+- Validation Overall Accuracy: 0.8430
+- Validation Per Category Iou: [0.00000000e+00 7.41164914e-01 8.53475354e-01 4.90058493e-01
+ 5.38510608e-01 3.20539250e-01            nan 3.99845020e-01
+ 4.92502734e-01 2.05511936e-03 7.89583806e-01 0.00000000e+00
+ 0.00000000e+00            nan 0.00000000e+00 5.24630954e-01
+ 0.00000000e+00 0.00000000e+00 6.89126355e-01 1.00217065e-01
+ 3.82209653e-01 2.80744319e-01 0.00000000e+00            nan
+ 0.00000000e+00 3.32851489e-01 1.26552665e-01 0.00000000e+00
+ 8.47240409e-01 8.22332088e-01 9.10638012e-01 2.90160652e-05
+ 2.21796079e-01 3.46893446e-01 0.00000000e+00]
+- Validation Per Category Accuracy: [0.00000000e+00 8.48318759e-01 9.46569249e-01 6.63363278e-01
+ 6.38141209e-01 3.85953320e-01            nan 5.53582721e-01
+ 7.04615765e-01 2.21213840e-03 9.22610901e-01 0.00000000e+00
+ 0.00000000e+00            nan 0.00000000e+00 6.66079812e-01
+ 0.00000000e+00 0.00000000e+00 8.81123024e-01 1.01945058e-01
+ 5.26802988e-01 3.23168426e-01 0.00000000e+00            nan
+ 0.00000000e+00 3.84975486e-01 1.44025357e-01 0.00000000e+00
+ 9.40049850e-01 9.07373703e-01 9.60336520e-01 4.89348514e-05
+ 4.45552728e-01 4.32057650e-01 0.00000000e+00]
+- Epoch: 12
 
 ## Model description
 
@@ -201,6 +207,23 @@ The following hyperparameters were used during training:
  0.90438608 0.01229056 0.42963846 0.43749539 0.                nan
  0.         0.40012652 0.15828843 0.         0.92988659 0.92495156
  0.96015873 0.01578965 0.43451194 0.3352385  0.        ]                                                                                                                                                                | 11    |
+| 0.3397     | 0.6115          | 0.3192              | 0.3868                   | 0.8430                      | [0.00000000e+00 7.41164914e-01 8.53475354e-01 4.90058493e-01
+ 5.38510608e-01 3.20539250e-01            nan 3.99845020e-01
+ 4.92502734e-01 2.05511936e-03 7.89583806e-01 0.00000000e+00
+ 0.00000000e+00            nan 0.00000000e+00 5.24630954e-01
+ 0.00000000e+00 0.00000000e+00 6.89126355e-01 1.00217065e-01
+ 3.82209653e-01 2.80744319e-01 0.00000000e+00            nan
+ 0.00000000e+00 3.32851489e-01 1.26552665e-01 0.00000000e+00
+ 8.47240409e-01 8.22332088e-01 9.10638012e-01 2.90160652e-05
+ 2.21796079e-01 3.46893446e-01 0.00000000e+00] | [0.00000000e+00 8.48318759e-01 9.46569249e-01 6.63363278e-01
+ 6.38141209e-01 3.85953320e-01            nan 5.53582721e-01
+ 7.04615765e-01 2.21213840e-03 9.22610901e-01 0.00000000e+00
+ 0.00000000e+00            nan 0.00000000e+00 6.66079812e-01
+ 0.00000000e+00 0.00000000e+00 8.81123024e-01 1.01945058e-01
+ 5.26802988e-01 3.23168426e-01 0.00000000e+00            nan
+ 0.00000000e+00 3.84975486e-01 1.44025357e-01 0.00000000e+00
+ 9.40049850e-01 9.07373703e-01 9.60336520e-01 4.89348514e-05
+ 4.45552728e-01 4.32057650e-01 0.00000000e+00] | 12    |
 
 
 ### Framework versions

tf_model.h5

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1934ef2c5222ea1994aceba5c5053b979f33f806f73a003c460397da7346686f
+oid sha256:557beb67bfe2ae54872212b19d61fca1983c80e9b557073cb55fb2312bc350c9
 size 15167588