Hugging Face's logo

asi
/
albert-act-base

Transformers
PyTorch
TensorFlow
TensorBoard
English
albert_act
Eval Results
Inference Endpoints
Model card Files Metrics Community
asi commited on
Commit
52eebd2
•
1 Parent(s): 1a3ee82

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +83 -0
README.md CHANGED
@@ -40,6 +40,89 @@ model-index:
40
  - type: f1
41
  value: 86.5
42
  name: F1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
43
  ---
44
 
45
 
 
40
  - type: f1
41
  value: 86.5
42
  name: F1
43
+ - task:
44
+ type: text-similarity
45
+ name: STS-B
46
+ dataset:
47
+ type: glue
48
+ name: STS-B # Semantic Textual Similarity Benchmark
49
+ split: stsb
50
+ metrics:
51
+ - type: spearmanr
52
+ value: 83.0
53
+ name: Spearman Corr
54
+ - type: pearsonr
55
+ value: 84.2
56
+ name: Pearson Corr
57
+ - task:
58
+ type: text-classification
59
+ name: QQP
60
+ dataset:
61
+ type: glue
62
+ name: QQP # Quora Question Pairs
63
+ split: qqp
64
+ metrics:
65
+ - type: f1
66
+ value: 68.5
67
+ name: F1
68
+ - type: accuracy
69
+ value: 87.7
70
+ name: Accuracy
71
+ - task:
72
+ type: text-classification
73
+ name: MNLI-m
74
+ dataset:
75
+ type: glue
76
+ name: MNLI-m # MultiNLI Matched
77
+ split: mnli_matched
78
+ metrics:
79
+ - type: accuracy
80
+ value: 79.9
81
+ name: Accuracy
82
+ - task:
83
+ type: text-classification
84
+ name: MNLI-mm
85
+ dataset:
86
+ type: glue
87
+ name: MNLI-mm # MultiNLI Matched
88
+ split: mnli_mismatched
89
+ metrics:
90
+ - type: accuracy
91
+ value: 79.2
92
+ name: Accuracy
93
+ - task:
94
+ type: text-classification
95
+ name: QNLI
96
+ dataset:
97
+ type: glue
98
+ name: QNLI # Question NLI
99
+ split: qnli
100
+ metrics:
101
+ - type: accuracy
102
+ value: 89.0
103
+ name: Accuracy
104
+ - task:
105
+ type: text-classification
106
+ name: RTE
107
+ dataset:
108
+ type: glue
109
+ name: RTE # Recognizing Textual Entailment
110
+ split: rte
111
+ metrics:
112
+ - type: accuracy
113
+ value: 63.0
114
+ name: Accuracy
115
+ - task:
116
+ type: text-classification
117
+ name: WNLI
118
+ dataset:
119
+ type: glue
120
+ name: WNLI # Winograd NLI
121
+ split: wnli
122
+ metrics:
123
+ - type: accuracy
124
+ value: 65.1
125
+ name: Accuracy
126
  ---
127
 
128