पंजे-x

सन्दर्भ:

एन

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

ds = tfds.load('huggingface:paws-x/en')
  • विवरण :
PAWS-X, a multilingual version of PAWS (Paraphrase Adversaries from Word Scrambling) for six languages.

This dataset contains 23,659 human translated PAWS evaluation pairs and 296,406 machine
translated training pairs in six typologically distinct languages: French, Spanish, German,
Chinese, Japanese, and Korean. English language is available by default. All translated
pairs are sourced from examples in PAWS-Wiki.

For further details, see the accompanying paper: PAWS-X: A Cross-lingual Adversarial Dataset
for Paraphrase Identification (https://arxiv.org/abs/1908.11828)

Note: There might be some missing or wrong labels in the dataset and we have replaced them with -1.
  • लाइसेंस : डेटासेट का उपयोग किसी भी उद्देश्य के लिए स्वतंत्र रूप से किया जा सकता है, हालांकि डेटा स्रोत के रूप में Google LLC ("Google") की स्वीकृति की सराहना की जाएगी। डेटासेट बिना किसी व्यक्त या निहित वारंटी के "जैसा है वैसा" प्रदान किया जाता है। Google डेटासेट के उपयोग के परिणामस्वरूप प्रत्यक्ष या अप्रत्यक्ष रूप से होने वाली किसी भी क्षति के लिए सभी दायित्वों से इनकार करता है।
  • संस्करण : 1.1.0
  • विभाजन :
विभाजित करना उदाहरण
'test' 2000
'train' 49401
'validation' 2000
  • विशेषताएँ :
{
    "id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "sentence1": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence2": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 2,
        "names": [
            "0",
            "1"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}

डे

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

ds = tfds.load('huggingface:paws-x/de')
  • विवरण :
PAWS-X, a multilingual version of PAWS (Paraphrase Adversaries from Word Scrambling) for six languages.

This dataset contains 23,659 human translated PAWS evaluation pairs and 296,406 machine
translated training pairs in six typologically distinct languages: French, Spanish, German,
Chinese, Japanese, and Korean. English language is available by default. All translated
pairs are sourced from examples in PAWS-Wiki.

For further details, see the accompanying paper: PAWS-X: A Cross-lingual Adversarial Dataset
for Paraphrase Identification (https://arxiv.org/abs/1908.11828)

Note: There might be some missing or wrong labels in the dataset and we have replaced them with -1.
  • लाइसेंस : डेटासेट का उपयोग किसी भी उद्देश्य के लिए स्वतंत्र रूप से किया जा सकता है, हालांकि डेटा स्रोत के रूप में Google LLC ("Google") की स्वीकृति की सराहना की जाएगी। डेटासेट बिना किसी व्यक्त या निहित वारंटी के "जैसा है वैसा" प्रदान किया जाता है। Google डेटासेट के उपयोग के परिणामस्वरूप प्रत्यक्ष या अप्रत्यक्ष रूप से होने वाली किसी भी क्षति के लिए सभी दायित्वों से इनकार करता है।
  • संस्करण : 1.1.0
  • विभाजन :
विभाजित करना उदाहरण
'test' 2000
'train' 49401
'validation' 2000
  • विशेषताएँ :
{
    "id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "sentence1": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence2": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 2,
        "names": [
            "0",
            "1"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}

तों

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

ds = tfds.load('huggingface:paws-x/es')
  • विवरण :
PAWS-X, a multilingual version of PAWS (Paraphrase Adversaries from Word Scrambling) for six languages.

This dataset contains 23,659 human translated PAWS evaluation pairs and 296,406 machine
translated training pairs in six typologically distinct languages: French, Spanish, German,
Chinese, Japanese, and Korean. English language is available by default. All translated
pairs are sourced from examples in PAWS-Wiki.

For further details, see the accompanying paper: PAWS-X: A Cross-lingual Adversarial Dataset
for Paraphrase Identification (https://arxiv.org/abs/1908.11828)

Note: There might be some missing or wrong labels in the dataset and we have replaced them with -1.
  • लाइसेंस : डेटासेट का उपयोग किसी भी उद्देश्य के लिए स्वतंत्र रूप से किया जा सकता है, हालांकि डेटा स्रोत के रूप में Google LLC ("Google") की स्वीकृति की सराहना की जाएगी। डेटासेट बिना किसी व्यक्त या निहित वारंटी के "जैसा है वैसा" प्रदान किया जाता है। Google डेटासेट के उपयोग के परिणामस्वरूप प्रत्यक्ष या अप्रत्यक्ष रूप से होने वाली किसी भी क्षति के लिए सभी दायित्वों से इनकार करता है।
  • संस्करण : 1.1.0
  • विभाजन :
विभाजित करना उदाहरण
'test' 2000
'train' 49401
'validation' 2000
  • विशेषताएँ :
{
    "id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "sentence1": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence2": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 2,
        "names": [
            "0",
            "1"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}

फादर

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

ds = tfds.load('huggingface:paws-x/fr')
  • विवरण :
PAWS-X, a multilingual version of PAWS (Paraphrase Adversaries from Word Scrambling) for six languages.

This dataset contains 23,659 human translated PAWS evaluation pairs and 296,406 machine
translated training pairs in six typologically distinct languages: French, Spanish, German,
Chinese, Japanese, and Korean. English language is available by default. All translated
pairs are sourced from examples in PAWS-Wiki.

For further details, see the accompanying paper: PAWS-X: A Cross-lingual Adversarial Dataset
for Paraphrase Identification (https://arxiv.org/abs/1908.11828)

Note: There might be some missing or wrong labels in the dataset and we have replaced them with -1.
  • लाइसेंस : डेटासेट का उपयोग किसी भी उद्देश्य के लिए स्वतंत्र रूप से किया जा सकता है, हालांकि डेटा स्रोत के रूप में Google LLC ("Google") की स्वीकृति की सराहना की जाएगी। डेटासेट बिना किसी व्यक्त या निहित वारंटी के "जैसा है वैसा" प्रदान किया जाता है। Google डेटासेट के उपयोग के परिणामस्वरूप प्रत्यक्ष या अप्रत्यक्ष रूप से होने वाली किसी भी क्षति के लिए सभी दायित्वों से इनकार करता है।
  • संस्करण : 1.1.0
  • विभाजन :
विभाजित करना उदाहरण
'test' 2000
'train' 49401
'validation' 2000
  • विशेषताएँ :
{
    "id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "sentence1": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence2": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 2,
        "names": [
            "0",
            "1"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}

जा

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

ds = tfds.load('huggingface:paws-x/ja')
  • विवरण :
PAWS-X, a multilingual version of PAWS (Paraphrase Adversaries from Word Scrambling) for six languages.

This dataset contains 23,659 human translated PAWS evaluation pairs and 296,406 machine
translated training pairs in six typologically distinct languages: French, Spanish, German,
Chinese, Japanese, and Korean. English language is available by default. All translated
pairs are sourced from examples in PAWS-Wiki.

For further details, see the accompanying paper: PAWS-X: A Cross-lingual Adversarial Dataset
for Paraphrase Identification (https://arxiv.org/abs/1908.11828)

Note: There might be some missing or wrong labels in the dataset and we have replaced them with -1.
  • लाइसेंस : डेटासेट का उपयोग किसी भी उद्देश्य के लिए स्वतंत्र रूप से किया जा सकता है, हालांकि डेटा स्रोत के रूप में Google LLC ("Google") की स्वीकृति की सराहना की जाएगी। डेटासेट बिना किसी व्यक्त या निहित वारंटी के "जैसा है वैसा" प्रदान किया जाता है। Google डेटासेट के उपयोग के परिणामस्वरूप प्रत्यक्ष या अप्रत्यक्ष रूप से होने वाली किसी भी क्षति के लिए सभी दायित्वों से इनकार करता है।
  • संस्करण : 1.1.0
  • विभाजन :
विभाजित करना उदाहरण
'test' 2000
'train' 49401
'validation' 2000
  • विशेषताएँ :
{
    "id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "sentence1": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence2": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 2,
        "names": [
            "0",
            "1"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}

को

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

ds = tfds.load('huggingface:paws-x/ko')
  • विवरण :
PAWS-X, a multilingual version of PAWS (Paraphrase Adversaries from Word Scrambling) for six languages.

This dataset contains 23,659 human translated PAWS evaluation pairs and 296,406 machine
translated training pairs in six typologically distinct languages: French, Spanish, German,
Chinese, Japanese, and Korean. English language is available by default. All translated
pairs are sourced from examples in PAWS-Wiki.

For further details, see the accompanying paper: PAWS-X: A Cross-lingual Adversarial Dataset
for Paraphrase Identification (https://arxiv.org/abs/1908.11828)

Note: There might be some missing or wrong labels in the dataset and we have replaced them with -1.
  • लाइसेंस : डेटासेट का उपयोग किसी भी उद्देश्य के लिए स्वतंत्र रूप से किया जा सकता है, हालांकि डेटा स्रोत के रूप में Google LLC ("Google") की स्वीकृति की सराहना की जाएगी। डेटासेट बिना किसी व्यक्त या निहित वारंटी के "जैसा है वैसा" प्रदान किया जाता है। Google डेटासेट के उपयोग के परिणामस्वरूप प्रत्यक्ष या अप्रत्यक्ष रूप से होने वाली किसी भी क्षति के लिए सभी दायित्वों से इनकार करता है।
  • संस्करण : 1.1.0
  • विभाजन :
विभाजित करना उदाहरण
'test' 2000
'train' 49401
'validation' 2000
  • विशेषताएँ :
{
    "id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "sentence1": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence2": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 2,
        "names": [
            "0",
            "1"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}

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

ds = tfds.load('huggingface:paws-x/zh')
  • विवरण :
PAWS-X, a multilingual version of PAWS (Paraphrase Adversaries from Word Scrambling) for six languages.

This dataset contains 23,659 human translated PAWS evaluation pairs and 296,406 machine
translated training pairs in six typologically distinct languages: French, Spanish, German,
Chinese, Japanese, and Korean. English language is available by default. All translated
pairs are sourced from examples in PAWS-Wiki.

For further details, see the accompanying paper: PAWS-X: A Cross-lingual Adversarial Dataset
for Paraphrase Identification (https://arxiv.org/abs/1908.11828)

Note: There might be some missing or wrong labels in the dataset and we have replaced them with -1.
  • लाइसेंस : डेटासेट का उपयोग किसी भी उद्देश्य के लिए स्वतंत्र रूप से किया जा सकता है, हालांकि डेटा स्रोत के रूप में Google LLC ("Google") की स्वीकृति की सराहना की जाएगी। डेटासेट बिना किसी व्यक्त या निहित वारंटी के "जैसा है वैसा" प्रदान किया जाता है। Google डेटासेट के उपयोग के परिणामस्वरूप प्रत्यक्ष या अप्रत्यक्ष रूप से होने वाली किसी भी क्षति के लिए सभी दायित्वों से इनकार करता है।
  • संस्करण : 1.1.0
  • विभाजन :
विभाजित करना उदाहरण
'test' 2000
'train' 49401
'validation' 2000
  • विशेषताएँ :
{
    "id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "sentence1": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence2": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 2,
        "names": [
            "0",
            "1"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}