इंडोन्लि

सन्दर्भ:

indonli

इस डेटासेट को TFDS में लोड करने के लिए निम्नलिखित कमांड का उपयोग करें:

ds = tfds.load('huggingface:indonli/indonli')
  • विवरण :
IndoNLI is the first human-elicited Natural Language Inference (NLI) dataset for Indonesian.
  IndoNLI is annotated by both crowd workers and experts. The expert-annotated data is used exclusively as a test set.
  It is designed to provide a challenging test-bed for Indonesian NLI by explicitly incorporating various linguistic phenomena such as numerical reasoning, structural changes, idioms, or temporal and spatial reasoning.
  • लाइसेंस : CC BY-SA 4.0

    एट्रिब्यूशन - आपको उचित क्रेडिट देना होगा, लाइसेंस का लिंक प्रदान करना होगा और यह बताना होगा कि क्या परिवर्तन किए गए हैं। आप ऐसा किसी भी उचित तरीके से कर सकते हैं, लेकिन किसी भी तरह से नहीं जिससे यह लगे कि लाइसेंसकर्ता आपको या आपके उपयोग का समर्थन करता है।

    शेयरएलाइक - यदि आप सामग्री का रीमिक्स, रूपांतरण या निर्माण करते हैं, तो आपको अपना योगदान मूल के समान लाइसेंस के तहत वितरित करना होगा।

    कोई अतिरिक्त प्रतिबंध नहीं - आप कानूनी शर्तों या तकनीकी उपायों को लागू नहीं कर सकते हैं जो दूसरों को लाइसेंस द्वारा अनुमति दी गई किसी भी चीज़ को करने से कानूनी रूप से प्रतिबंधित करते हैं।

  • संस्करण : 1.1.0

  • विभाजन :

विभाजित करना उदाहरण
'test_expert' 2984
'test_lay' 2201
'train' 10330
'validation' 2197
  • विशेषताएँ :
{
    "premise": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "hypothesis": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 3,
        "names": [
            "entailment",
            "neutral",
            "contradiction"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}