Update README.md

README.md

@@ -16,20 +16,20 @@ should 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 the andrewljohnson/magic_cards dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.8470
-- Mean Iou: 0.5709
-- Mean Accuracy: 0.8685
-- Overall Accuracy: 0.8676
+- Loss: 0.059
+- Mean Iou: 0.654
+- Mean Accuracy: 0.981
+- Overall Accuracy: 0.983
 - Accuracy Unlabeled: nan
-- Accuracy Front: 0.8772
-- Accuracy Back: 0.8598
+- Accuracy Front: 0.964
+- Accuracy Back: 0.998
 - Iou Unlabeled: 0.0
-- Iou Front: 0.8598
-- Iou Back: 0.8530
+- Iou Front: 0.965
+- Iou Back: 0.998
 
 ## Model description
 
-More information needed
+Segment magic card front and backs.
 
 ## Intended uses & limitations
 
@@ -50,14 +50,10 @@ The following hyperparameters were used during training:
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- num_epochs: 1
+- num_epochs: 50
 
 ### Training results
 
-| Training Loss | Epoch | Step | Validation Loss | Mean Iou | Mean Accuracy | Overall Accuracy | Accuracy Unlabeled | Accuracy Front | Accuracy Back | Iou Unlabeled | Iou Front | Iou Back |
-|:-------------:|:-----:|:----:|:---------------:|:--------:|:-------------:|:----------------:|:------------------:|:--------------:|:-------------:|:-------------:|:---------:|:--------:|
-| 0.7118        | 0.91  | 20   | 0.8470          | 0.5709   | 0.8685        | 0.8676           | nan                | 0.8772         | 0.8598        | 0.0           | 0.8598    | 0.8530   |
-
 
 ### Framework versions