balon

Referanslar:

adjunct_island

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın:

ds = tfds.load('huggingface:blimp/adjunct_island')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

anaphor_gender_anlaşma

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın:

ds = tfds.load('huggingface:blimp/anaphor_gender_agreement')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

anaphor_number_anlaşma

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın:

ds = tfds.load('huggingface:blimp/anaphor_number_agreement')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

animate_subject_passive

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/animate_subject_passive')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

animate_subject_trans

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/animate_subject_trans')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

nedensel

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/causative')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

karmaşık_NP_island

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/complex_NP_island')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

koordinat_yapısı_constraint_complex_left_branch

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/coordinate_structure_constraint_complex_left_branch')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

koordinat_yapısı_constraint_object_extraction

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/coordinate_structure_constraint_object_extraction')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

belirleyici_isim_anlaşma_1

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/determiner_noun_agreement_1')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

belirleyici_isim_anlaşma_2

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/determiner_noun_agreement_2')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

belirleyici_isim_anlaşma_irregular_1

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/determiner_noun_agreement_irregular_1')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

belirleyici_isim_anlaşma_irregular_2

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/determiner_noun_agreement_irregular_2')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

belirleyici_isim_agreement_with_adj_2

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/determiner_noun_agreement_with_adj_2')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

belirleyici_isim_agreement_with_adj_irregular_1

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/determiner_noun_agreement_with_adj_irregular_1')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

belirleyici_isim_agreement_with_adj_irregular_2

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/determiner_noun_agreement_with_adj_irregular_2')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

belirleyici_isim_anlaşma_with_adjective_1

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/determiner_noun_agreement_with_adjective_1')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

dikkat dağıtıcı_anlaşma_ilişkisel_isim

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/distractor_agreement_relational_noun')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

distractor_agreement_relative_clause

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/distractor_agreement_relative_clause')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

drop_argument

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/drop_argument')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

ellipsis_n_bar_1

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/ellipsis_n_bar_1')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

ellipsis_n_bar_2

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/ellipsis_n_bar_2')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

varoluşsal_ora_object_raising

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/existential_there_object_raising')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

varoluşsal_orada_niteleyiciler_1

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/existential_there_quantifiers_1')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

varoluşsal_orada_niteleyiciler_2

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/existential_there_quantifiers_2')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

varoluşsal_there_subject_raising

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/existential_there_subject_raising')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

expletive_it_object_raising

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/expletive_it_object_raising')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

yenilikçi

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/inchoative')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

geçişsiz

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/intransitive')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

düzensiz_geçmiş_participle_adjectives

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/irregular_past_participle_adjectives')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

düzensiz_geçmiş_participle_verbs

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/irregular_past_participle_verbs')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

düzensiz_plural_subject_verb_agreement_1

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/irregular_plural_subject_verb_agreement_1')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

düzensiz_plural_subject_verb_agreement_2

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/irregular_plural_subject_verb_agreement_2')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

left_branch_island_echo_question

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/left_branch_island_echo_question')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

left_branch_island_simple_question

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/left_branch_island_simple_question')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

matris_sorusu_npi_licensor_current

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/matrix_question_npi_licensor_present')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

npi_current_1

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/npi_present_1')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

npi_current_2

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/npi_present_2')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

only_npi_licensor_current

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/only_npi_licensor_present')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

only_npi_scope

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/only_npi_scope')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

pasif_1

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/passive_1')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

pasif_2

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/passive_2')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

prensip_A_c_komut

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/principle_A_c_command')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

prensip_A_case_1

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/principle_A_case_1')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

prensip_A_case_2

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/principle_A_case_2')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

prensip_A_domain_1

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/principle_A_domain_1')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

prensip_A_domain_2

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/principle_A_domain_2')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

prensip_A_domain_3

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/principle_A_domain_3')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

prensip_A_yeniden yapılanma

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/principle_A_reconstruction')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

normal_plural_subject_verb_agreement_1

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/regular_plural_subject_verb_agreement_1')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

normal_plural_subject_verb_agreement_2

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/regular_plural_subject_verb_agreement_2')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

sentential_negation_npi_licensor_current

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/sentential_negation_npi_licensor_present')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

sentential_negation_npi_scope

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/sentential_negation_npi_scope')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

sentential_subject_island

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/sentential_subject_island')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

superlative_quantifiers_1

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/superlative_quantifiers_1')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

superlative_quantifiers_2

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/superlative_quantifiers_2')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

zorlu_vs_raising_1

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/tough_vs_raising_1')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

zorlu_vs_raising_2

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/tough_vs_raising_2')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

geçişli

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/transitive')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

ne_adası

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/wh_island')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

wh_questions_object_gap

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/wh_questions_object_gap')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

wh_questions_subject_gap

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/wh_questions_subject_gap')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

wh_questions_subject_gap_long_distance

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/wh_questions_subject_gap_long_distance')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

wh_vs_that_no_gap

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/wh_vs_that_no_gap')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

wh_vs_that_no_gap_long_distance

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/wh_vs_that_no_gap_long_distance')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

wh_vs_that_with_gap

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/wh_vs_that_with_gap')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

wh_vs_that_with_gap_long_distance

Bu veri kümesini TFDS'ye yüklemek için aşağıdaki komutu kullanın: 

ds = tfds.load('huggingface:blimp/wh_vs_that_with_gap_long_distance')
  • Tanım : 
BLiMP is a challenge set for evaluating what language models (LMs) know about
major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each
containing 1000 minimal pairs isolating specific contrasts in syntax,
morphology, or semantics. The data is automatically generated according to
expert-crafted grammars.
  • Lisans : Bilinen lisans yok
  • Sürüm : 0.1.0
  • Bölünmeler :
Bölmek Örnekler
'train' 1000
  • Özellikler : 
{
    "sentence_good": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_bad": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "field": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "linguistics_term": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "UID": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "simple_LM_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "one_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "two_prefix_method": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "lexically_identical": {
        "dtype": "bool",
        "id": null,
        "_type": "Value"
    },
    "pair_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}