NSMC

ข้อมูลอ้างอิง:

ใช้คำสั่งต่อไปนี้เพื่อโหลดชุดข้อมูลนี้ใน TFDS:

ds = tfds.load('huggingface:nsmc')
  • คำอธิบาย :
This is a movie review dataset in the Korean language. Reviews were scraped from Naver movies. The dataset construction is based on the method noted in Large movie review dataset from Maas et al., 2011.
  • ใบอนุญาต : CC0 1.0 สากล (CC0 1.0)
  • เวอร์ชัน : 1.1.0
  • แยก :
แยก ตัวอย่าง
'test' 50000
'train' 150000
  • คุณสมบัติ :
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "document": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 2,
        "names": [
            "negative",
            "positive"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}