AWeirdDev commited on
Commit
fc7d412
β€’
1 Parent(s): 171e20a

added cleaning info

Browse files
Files changed (1) hide show
  1. README.md +37 -0
README.md CHANGED
@@ -49,6 +49,17 @@ All questions Bill Wurtz answers on [billwurtz.com/questions](https://billwurtz.
49
  - πŸ• For tasks: `text-generation`, `question-answering`, + more
50
  - πŸ“œ Rows: `129,362` (129k)
51
 
 
 
 
 
 
 
 
 
 
 
 
52
  Download with [πŸ€— Datasets](https://pypi.org/project/datasets):
53
 
54
  ```python
@@ -58,3 +69,29 @@ dataset = load_dataset("AWeirdDev/bill-wurtz")
58
  dataset["train"][0]
59
  # => { "link": "...", "question": "your opinion on ceilings?", "answer": "incredible" }
60
  ```
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
49
  - πŸ• For tasks: `text-generation`, `question-answering`, + more
50
  - πŸ“œ Rows: `129,362` (129k)
51
 
52
+ ```python
53
+ DatasetDict({
54
+ train: Dataset({
55
+ features: ['link', 'question', 'answer'],
56
+ num_rows: 129362
57
+ })
58
+ })
59
+ ```
60
+
61
+ ## Use This Dataset
62
+
63
  Download with [πŸ€— Datasets](https://pypi.org/project/datasets):
64
 
65
  ```python
 
69
  dataset["train"][0]
70
  # => { "link": "...", "question": "your opinion on ceilings?", "answer": "incredible" }
71
  ```
72
+
73
+ <details>
74
+ <summary><b>🧹 Cleaning the dataset</b></summary>
75
+ <p>
76
+
77
+ Some questions/answers may be blank. Clean the dataset before you use it.
78
+
79
+ ```python
80
+ from datasets import Dataset
81
+
82
+ raw_dataset = dataset["train"].to_list()
83
+
84
+ for i, d in enumerate(raw_dataset):
85
+ if not d['question'].strip() or not d['answer'].strip():
86
+ del raw_dataset[i]
87
+
88
+ raw_dataset = Dataset.from_list(raw_dataset)
89
+ raw_dataset
90
+ # Dataset({
91
+ # features: ['link', 'question', 'answer'],
92
+ # num_rows: 123922
93
+ # })
94
+ ```
95
+
96
+ </p>
97
+ </details>