README.md

---
language:
- en
license: other
tags:
- pretraining
- text-generation
- language-modeling
task_categories:
- text-generation
task_ids:
- language-modeling
size_categories:
- 10K<n<100K
pretty_name: fp_pt_split_hub
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*.parquet
  - split: validation
    path: data/validation-*.parquet
  - split: test
    path: data/test-*.parquet
train-eval-index:
- config: default
  task: text-generation
  task_id: language-modeling
  splits:
    train_split: train
    eval_split: validation
  col_mapping:
    text: text
---

# Dataset Card for fp_pt_split_hub

## Dataset Summary

This dataset contains text records converted from local JSONL files. Each source
record had a text field, which is exposed as the dataset's `text` column.

Conversion cleanup:

- HTML character references such as `&#39;`, `&amp;`, `&lt;`, and `&quot;` are decoded.
- Conservative text cleanup removes transcript stage tags, common YouTube boilerplate, control/replacement characters, PDF percent-space artifacts, and raw LaTeX `tabular` wrappers.
- Punctuation spacing is normalized for common extraction artifacts such as `A.Connella`, `ISSN:2153`, and `text.Next`.

## How to Load

```python
from datasets import load_dataset

ds = load_dataset("fp_pt_split_hub")
```

For a local checkout, use:

```python
from datasets import load_dataset

ds = load_dataset("/path/to/fp_pt_split_hub")
```

## Dataset Stats

| Split | Rows |
| --- | ---: |
| `train` | 14,522 |
| `validation` | 147 |
| `test` | 75 |

## Data Schema

Columns:

- `text`

## Licensing Information

License: other