File size: 2,653 Bytes
eecccf7
7cae844
eecccf7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7cae844
 
 
 
eecccf7
7cae844
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
---
size_categories: n<1K
dataset_info:
  features:
  - name: _id
    dtype: string
  - name: url
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  - name: score
    dtype: float64
  - name: views
    dtype: float64
  - name: model_name
    dtype: string
  - name: query
    dtype: string
  splits:
  - name: train
    num_bytes: 903807
    num_examples: 1500
  download_size: 618051
  dataset_size: 903807
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
tags:
- synthetic
- distilabel
- rlaif
---

<p align="left">
  <a href="https://github.com/argilla-io/distilabel">
    <img src="https://raw.githubusercontent.com/argilla-io/distilabel/main/docs/assets/distilabel-badge-light.png" alt="Built with Distilabel" width="200" height="32"/>
  </a>
</p>

# Dataset Card for cohere-wikipedia-2023-11-cs-queries

This dataset has been created with [distilabel](https://distilabel.argilla.io/).

## Dataset Summary

This dataset contains a `pipeline.yaml` which can be used to reproduce the pipeline that generated it in distilabel using the `distilabel` CLI:

```console
distilabel pipeline run --config "https://huggingface.co/datasets/rasdani/cohere-wikipedia-2023-11-cs-queries/raw/main/pipeline.yaml"
```

or explore the configuration:

```console
distilabel pipeline info --config "https://huggingface.co/datasets/rasdani/cohere-wikipedia-2023-11-cs-queries/raw/main/pipeline.yaml"
```

## Dataset structure

The examples have the following structure per configuration:


<details><summary> Configuration: default </summary><hr>

```json
{
    "_id": "20231101.cs_19_9",
    "model_name": "gpt-4o",
    "query": "Kdo objevil rentgenov\u00e9 z\u00e1\u0159en\u00ed a kdy k tomu do\u0161lo?",
    "score": 1.0,
    "text": "Roku 1895 Wilhelm Conrad R\u00f6ntgen objevuje \u201epaprsky X\u201c (rentgenov\u00e9 z\u00e1\u0159en\u00ed), o rok pozd\u011bji Henri Becquerel objevuje radioaktivitu, o dal\u0161\u00ed rok pozd\u011bji Joseph John Thomson objevuje elektron. Pierre Curie a Marie Curie-Sk\u0142odowsk\u00e1 studuj\u00ed j\u00e1chymovsk\u00e9 radium. Vznik\u00e1 tak jadern\u00e1 fyzika.",
    "title": "Fyzika",
    "url": "https://cs.wikipedia.org/wiki/Fyzika",
    "views": 4458.917356449113
}
```

This subset can be loaded as:

```python
from datasets import load_dataset

ds = load_dataset("rasdani/cohere-wikipedia-2023-11-cs-queries", "default")
```

Or simply as it follows, since there's only one configuration and is named `default`: 

```python
from datasets import load_dataset

ds = load_dataset("rasdani/cohere-wikipedia-2023-11-cs-queries")
```


</details>