id_panl_bppt

อ้างอิง:

id_panl_bppt

ใช้คำสั่งต่อไปนี้เพื่อโหลดชุดข้อมูลนี้ใน TFDS:

ds = tfds.load('huggingface:id_panl_bppt/id_panl_bppt')
  • คำอธิบาย :
Parallel Text Corpora for Multi-Domain Translation System created by BPPT (Indonesian Agency for the Assessment and
Application of Technology) for PAN Localization Project (A Regional Initiative to Develop Local Language Computing
Capacity in Asia). The dataset contains around 24K sentences divided in 4 difference topics (Economic, international,
Science and Technology and Sport).
  • ใบอนุญาต : ไม่มีใบอนุญาตที่รู้จัก
  • เวอร์ชัน : 1.0.0
  • แยก :
แยก ตัวอย่าง
'train' 24021
  • คุณสมบัติ :
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "translation": {
        "languages": [
            "en",
            "id"
        ],
        "id": null,
        "_type": "Translation"
    },
    "topic": {
        "num_classes": 4,
        "names": [
            "Economy",
            "International",
            "Science",
            "Sport"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}