lewtun HF staff commited on
Commit
55f272a
1 Parent(s): 765031d

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +24 -2
README.md CHANGED
@@ -100,9 +100,31 @@ Using the official [question answering notebook](https://github.com/huggingface/
100
  'exact': 79.92082877116145,
101
  'f1': 82.91749890730917,
102
  'total': 11873}
103
- ```
 
 
104
 
105
- which is consistent with the officially reported results.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
106
 
107
  ## Authors
108
  **Branden Chan:** branden.chan@deepset.ai
100
  'exact': 79.92082877116145,
101
  'f1': 82.91749890730917,
102
  'total': 11873}
103
+ ```
104
+
105
+ which is consistent with the officially reported results. Using the question answering `Evaluator` from `evaluate` gives:
106
 
107
+ ```
108
+ {'HasAns_exact': 77.91835357624831,
109
+ 'HasAns_f1': 84.07820736158186,
110
+ 'HasAns_total': 5928,
111
+ 'NoAns_exact': 81.91757779646763,
112
+ 'NoAns_f1': 81.91757779646763,
113
+ 'NoAns_total': 5945,
114
+ 'best_exact': 79.92082877116145,
115
+ 'best_exact_thresh': 0.996823787689209,
116
+ 'best_f1': 82.99634576260925,
117
+ 'best_f1_thresh': 0.996823787689209,
118
+ 'exact': 79.92082877116145,
119
+ 'f1': 82.9963457626089,
120
+ 'latency_in_seconds': 0.016523243643392558,
121
+ 'samples_per_second': 60.52080460605492,
122
+ 'total': 11873,
123
+ 'total_time_in_seconds': 196.18047177799986}
124
+ ```
125
+
126
+ which is also consistent with the officially reported results.
127
+
128
 
129
  ## Authors
130
  **Branden Chan:** branden.chan@deepset.ai