beelzeebuub commited on
Commit
c462d23
·
1 Parent(s): 173c796
Files changed (1) hide show
  1. app.py +0 -40
app.py CHANGED
@@ -1,43 +1,3 @@
1
  import gradio as gr
2
 
3
- class OrdinalRegressionMetric(Metric):
4
- def __init__(self):
5
- super().__init__()
6
- self.total = 0
7
- self.count = 0
8
-
9
- def accumulate(self, learn):
10
- # Get predictions and targets
11
- preds, targs = learn.pred, learn.y
12
-
13
- # Your custom logic to convert predictions and targets to numeric values
14
- preds_numeric = torch.argmax(preds, dim=1)
15
- targs_numeric = targs
16
-
17
- #print("preds_numeric: ",preds_numeric)
18
- #print("targs_numeric: ",targs_numeric)
19
-
20
- # Calculate the metric (modify this based on your specific needs)
21
- squared_diff = torch.sum(torch.sqrt((preds_numeric - targs_numeric)**2))
22
-
23
- # Normalize by the maximum possible difference
24
- max_diff = torch.sqrt((torch.max(targs_numeric) - torch.min(targs_numeric))**2)
25
-
26
- #print("squared_diff: ",squared_diff)
27
- #print("max_diff: ",max_diff)
28
-
29
- # Update the metric value
30
- self.total += squared_diff
31
- #print("self.total: ",self.total)
32
- self.count += max_diff
33
- #print("self.count: ",self.count)
34
- @property
35
- def value(self):
36
- if self.count == 0:
37
- return 0.0 # or handle this case appropriately
38
- #print("self.total / self.count: ", (self.total / self.count))
39
- # Calculate the normalized metric value
40
- metric_value = 1/(self.total / self.count)
41
- return metric_value
42
-
43
  gr.load("models/beelzeebuub/FJModel").launch()
 
1
  import gradio as gr
2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3
  gr.load("models/beelzeebuub/FJModel").launch()