dkoterwa commited on
Commit
2a78aea
1 Parent(s): acdcc09

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +27 -0
README.md CHANGED
@@ -21,3 +21,30 @@ configs:
21
  - split: train
22
  path: data/train-*
23
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
21
  - split: train
22
  path: data/train-*
23
  ---
24
+
25
+ # mkqa filtered version
26
+
27
+ For a better dataset description, please visit the official site of the source dataset: [LINK](https://huggingface.co/datasets/mkqa) <br>
28
+ <br>
29
+ **This dataset was prepared by converting mkqa dataset**.
30
+
31
+ **I additionaly share the code which I used to convert the original dataset to make everything more clear**
32
+ ```
33
+ mkqa = load_dataset("mkqa", split="train").to_pandas()
34
+ needed_langs = ["en", "ar", "de", "es", "vi", "zh_cn"]
35
+ rows = []
36
+ for i, row in tqdm(mkqa.iterrows(), total=mkqa.shape[0]):
37
+ for lang in needed_langs:
38
+ rows.append([lang, row["example_id"], row["queries"][lang], row["answers"][lang][0]["text"]])
39
+
40
+ filtered_dataset = pd.DataFrame(rows, columns=["lang", "example_id", "query", "answer"])
41
+ filtered_dataset.dropna(inplace=True)
42
+ filtered_dataset.reset_index(drop=True, inplace=True)
43
+ ```
44
+
45
+ **How to download**
46
+
47
+ ```
48
+ from datasets import load_dataset
49
+ data = load_dataset("dkoterwa/oasst1_filtered_retrieval")
50
+ ```