Nathan Fradet commited on
Commit
23bb343
·
unverified ·
1 Parent(s): 5ec4e13

editing readme

Browse files
Files changed (1) hide show
  1. README.md +25 -2
README.md CHANGED
@@ -8,7 +8,7 @@ tags:
8
  - metric
9
  description: Levenshtein (edit) distance
10
  sdk: gradio
11
- sdk_version: 5.16.0
12
  app_file: app.py
13
  pinned: false
14
  ---
@@ -17,7 +17,7 @@ pinned: false
17
 
18
  ## Metric Description
19
 
20
- This metric computes the Levenshtein distance, also commonly called "edit distance". The Levenshtein distance measures the number of combined editions, deletions and additions to perform on a string so that it becomes identical to a second one. It is a popular metric for text similarity.
21
  This module directly calls the [Levenshtein package](https://github.com/rapidfuzz/Levenshtein) for fast execution speed.
22
 
23
  ## How to Use
@@ -35,6 +35,28 @@ Dictionary mapping to the average Levenshtein distance (lower is better) and the
35
 
36
  ### Examples
37
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
38
  ```Python
39
  import evaluate
40
 
@@ -46,6 +68,7 @@ results = levenshtein.compute(
46
  references=[
47
  "foo", "bar"
48
  ],
 
49
  )
50
  print(results)
51
  # {"levenshtein": 1, "levenshtein_ratio": 0.875}
 
8
  - metric
9
  description: Levenshtein (edit) distance
10
  sdk: gradio
11
+ sdk_version: 5.24.0
12
  app_file: app.py
13
  pinned: false
14
  ---
 
17
 
18
  ## Metric Description
19
 
20
+ This metric computes the Levenshtein distance, also commonly called "edit distance". The Levenshtein distance measures the number of combined insertions, deletions and substitutions operations (one per character) to perform on a string so that it becomes identical to a second one. It is a popular metric for text similarity.
21
  This module directly calls the [Levenshtein package](https://github.com/rapidfuzz/Levenshtein) for fast execution speed.
22
 
23
  ## How to Use
 
35
 
36
  ### Examples
37
 
38
+ #### Levenshtein distance
39
+
40
+ ```Python
41
+ import evaluate
42
+
43
+ levenshtein = evaluate.load("Natooz/Levenshtein")
44
+ results = levenshtein.compute(
45
+ predictions=[
46
+ "foo", "baroo" # 0 and 2 edits
47
+ ],
48
+ references=[
49
+ "foo", "bar"
50
+ ],
51
+ )
52
+ print(results)
53
+ # {"levenshtein": 1, "levenshtein_ratio": 0.875}
54
+ ```
55
+
56
+ #### Indel (insertion-deletion) distance
57
+
58
+ The weight of each operation can be provided in order to customize the score. For example, the substitution score can be set to 2 to compute the "indel" distance, so that each substitution is counted as two operations (deletion + insertion).
59
+
60
  ```Python
61
  import evaluate
62
 
 
68
  references=[
69
  "foo", "bar"
70
  ],
71
+ weights=(1, 1, 2), # weight of 2 for substitutions
72
  )
73
  print(results)
74
  # {"levenshtein": 1, "levenshtein_ratio": 0.875}