README.md

---
dataset_info:
  features:
  - name: inputs
    dtype: string
  - name: targets
    dtype: string
  splits:
  - name: train
    num_bytes: 500257169
    num_examples: 801402
  - name: validation
    num_bytes: 61671631
    num_examples: 88927
  - name: test
    num_bytes: 61225544
    num_examples: 88931
  download_size: 212800258
  dataset_size: 623154344
license: cc-by-sa-4.0
task_categories:
- text2text-generation
- text-generation
language:
- th
size_categories:
- 100K<n<1M
---
# Dataset Card for "scb_mt_2020_th2en_prompt"


This dataset made from [scb_mt_enth_2020](https://huggingface.co/datasets/scb_mt_enth_2020) that removed nus_sms and paracrawl from source.

Source code for create dataset: [https://github.com/PyThaiNLP/support-aya-datasets/blob/main/translation/scb_mt.ipynb](https://github.com/PyThaiNLP/support-aya-datasets/blob/main/translation/scb_mt.ipynb)

## Template
```
Inputs: แปลประโยคหรือย่อหน้าต่อไปนี้จากภาษาไทยเป็นภาษาอังกฤษ:\n{th}

Targets: English sentence
```