YusuphaJuwara commited on
Commit
e804229
·
verified ·
1 Parent(s): d52f9f2

Upload roberta_fever_plus_adv_samples/README.md with huggingface_hub

Browse files
roberta_fever_plus_adv_samples/README.md CHANGED
@@ -11,19 +11,40 @@ license: unknown
11
  metrics:
12
  epoch:
13
  - 0
 
 
 
14
  train_loss:
15
  - 1.1077719926834106
 
 
 
16
  val_loss:
17
  - 1.116754412651062
 
 
 
18
  train_acc:
19
  - 0.25
 
 
 
20
  val_acc:
21
  - 0.34375
 
 
 
22
  train_f1_score:
23
  - 0.25
 
 
 
24
  val_f1_score:
25
  - 0.34375
26
- best_metric: 1.116754412651062
 
 
 
27
  model-index:
28
  - name: nli-fever
29
  results:
@@ -35,7 +56,7 @@ model-index:
35
  type: fever
36
  metrics:
37
  - type: acc
38
- value: '0.34'
39
  name: Accuracy
40
  verified: false
41
  ---
@@ -60,10 +81,10 @@ The model was trained on the FEVER (Fact Extraction and VERification) dataset.
60
 
61
  ## Training procedure
62
 
63
- The model was trained for [0] epochs
64
- with a final loss of 1.116754412651062, an
65
- accuracy of 0.34375, and
66
- F1 score of 0.34375.
67
 
68
  ## How to use
69
 
 
11
  metrics:
12
  epoch:
13
  - 0
14
+ - 0
15
+ - 0
16
+ - 0
17
  train_loss:
18
  - 1.1077719926834106
19
+ - 1.0808662176132202
20
+ - 1.1135263442993164
21
+ - 1.0963691473007202
22
  val_loss:
23
  - 1.116754412651062
24
+ - 1.1231188774108887
25
+ - 1.1370257139205933
26
+ - 1.1022729873657227
27
  train_acc:
28
  - 0.25
29
+ - 0.65625
30
+ - 0.625
31
+ - 0.6538462042808533
32
  val_acc:
33
  - 0.34375
34
+ - 0.375
35
+ - 0.3854166567325592
36
+ - 0.35882866382598877
37
  train_f1_score:
38
  - 0.25
39
+ - 0.65625
40
+ - 0.625
41
+ - 0.6538461446762085
42
  val_f1_score:
43
  - 0.34375
44
+ - 0.375
45
+ - 0.3854166567325592
46
+ - 0.35882866382598877
47
+ best_metric: 1.1022729873657227
48
  model-index:
49
  - name: nli-fever
50
  results:
 
56
  type: fever
57
  metrics:
58
  - type: acc
59
+ value: '0.36'
60
  name: Accuracy
61
  verified: false
62
  ---
 
81
 
82
  ## Training procedure
83
 
84
+ The model was trained for [0, 0, 0, 0] epochs
85
+ with a final loss of 1.1022729873657227, an
86
+ accuracy of 0.35882866382598877, and
87
+ F1 score of 0.35882866382598877.
88
 
89
  ## How to use
90