File size: 1,727 Bytes
e4a3281
 
 
 
 
 
4a0e502
e4a3281
 
 
 
4a0e502
e4a3281
 
 
 
4a0e502
e4a3281
 
 
 
 
 
 
 
 
 
 
4a0e502
 
 
 
 
e4a3281
 
 
 
 
 
 
4a0e502
 
 
e4a3281
 
4a0e502
e4a3281
 
 
4a0e502
e4a3281
 
 
 
 
 
4a0e502
 
e4a3281
 
 
 
 
 
 
 
 
 
 
 
 
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
---
language:
- code
license: other
task_categories:
- text-generation
pretty_name: Luau Stack HQ ChatML
size_categories:
- 10K<n<100K
---

# Dataset Card for YCWTG/luau-stack-hq-ChatML

## Summary

This dataset is a pretraining-oriented repackaging of `khtsly/luau-stack-hq`.
It keeps only the source code content and stores it in a ChatML-style `messages` field.

- Rows: 21,704
- File format: JSONL
- Split: `train`
- Main file: `train.jsonl`

## Data Schema

Each row has one field:

```json
{
  "messages": [
    {"role": "assistant", "content": "<luau source code>"}
  ]
}
```

## Processing

The source file `/luau-stack.jsonl` was transformed with the following rule:

1. Read each JSON line.
2. Extract the original `code` field as `text`.
3. Wrap `text` into `messages` as:
   `{"messages": [{"role": "assistant", "content": text}]}`.
4. Drop all other fields.

No filtering, deduplication, or normalization was applied beyond field selection/renaming/wrapping.

## Intended Use

This dataset is intended for code language model pretraining or continued pretraining where a ChatML-style assistant message format is required.

## Load with `datasets`

```python
from datasets import load_dataset

ds = load_dataset("YCWTG/luau-stack-hq-ChatML", split="train")
print(ds[0]["messages"][0]["content"][:200])
```

## Source and Attribution

- Original source repository: `khtsly/luau-stack-hq`
- This dataset is a format-converted repackaging by `YCWTG`.

Please cite or acknowledge the original source repository when using this data.

## License

This repackaging does not change the licensing terms of the upstream data.
Please review and comply with the original source license and any file-level license obligations.