băng ghế lớn

Tài liệu tham khảo:

tóm tắt_tường thuật_hiểu

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/abstract_narrative_understanding')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 3000
'train' 2400
'validation' 600
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

lỗi thời

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/anachronisms')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 230
'train' 184
'validation' 46
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

tương tự_tương tự

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/analogical_similarity')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 323
'train' 259
'validation' 64
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

phân tích_entailment

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/analytic_entailment')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 70
'train' 54
'validation' 16
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

số học

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/arithmetic')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 15023
'train' 12019
'validation' 3004
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

ascii_word_recognition

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/ascii_word_recognition')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 5000
'train' 4000
'validation' 1000
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

quyền tác giả_verification

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/authorship_verification')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 880
'train' 704
'validation' 176
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

auto_categorization

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/auto_categorization')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 328
'train' 263
'validation' 65
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

auto_debugging

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/auto_debugging')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 34
'train' 18
'validation' 16
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

bbq_lite_json

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/bbq_lite_json')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 16076
'train' 12866
'validation' 3210
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

bắc cầu_anaphora_độ phân giải_barqa

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/bridging_anaphora_resolution_barqa')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 648
'train' 519
'validation' 129
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

nhân_quả

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/causal_judgment')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 190
'train' 152
'validation' 38
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

nguyên nhân và kết quả

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/cause_and_effect')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 153
'train' 123
'validation' 30
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

chiếu tướng_in_one

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/checkmate_in_one')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 3498
'train' 2799
'validation' 699
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

cờ vua_state_tracking

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/chess_state_tracking')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 6000
'train' 4800
'validation' 1200
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

Chinese_remainder_theorem

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/chinese_remainder_theorem')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 500
'train' 400
'validation' 100
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

cifar10_classification

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/cifar10_classification')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 20000
'train' 16000
'validation' 4000
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

code_line_description

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/code_line_description')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 60
'train' 44
'validation' 16
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

mật danh

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/codenames')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 85
'train' 68
'validation' 17
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

màu sắc

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/color')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 4000
'train' 3200
'validation' 800
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

chung_morpheme

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/common_morpheme')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 50
'train' 34
'validation' 16
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

sự kết hợp khái niệm

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/conceptual_combinations')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 103
'train' 84
'validation' 19
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

conlang_translation

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/conlang_translation')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 164
'train' 132
'validation' 32
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

contextual_parametric_know_conflicts

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/contextual_parametric_knowledge_conflicts')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 17528
'train' 14023
'validation' 3505
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

tai nạn_hoa

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/crash_blossom')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 38
'train' 22
'validation' 16
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

crass_ai

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/crass_ai')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 44
'train' 28
'validation' 16
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

cryobiology_tiếng Tây Ban Nha

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/cryobiology_spanish')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 146
'train' 117
'validation' 29
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

tiền điện tử

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/cryptonite')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 26157
'train' 20926
'validation' 5231
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

thuật toán cs_

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/cs_algorithms')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 1320
'train' 1056
'validation' 264
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

bóng tối_hài hước_phát hiện

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/dark_humor_detection')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 80
'train' 64
'validation' 16
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

ngày_hiểu

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/date_understanding')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 369
'train' 296
'validation' 73
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

định hướng_qa

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/disambiguation_qa')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 258
'train' 207
'validation' 51
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

diễn ngôn_marker_prediction

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/discourse_marker_prediction')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 857
'train' 686
'validation' 171
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

disfl_qa

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/disfl_qa')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 8000
'train' 6400
'validation' 1600
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

dyck_ngôn ngữ

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/dyck_languages')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 1000
'train' 800
'validation' 200
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

sơ cấp_math_qa

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/elementary_math_qa')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 38160
'train' 30531
'validation' 7629
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

biểu tượng cảm xúc_phim

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/emoji_movie')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 100
'train' 80
'validation' 20
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

biểu tượng cảm xúc_cảm xúc_prediction

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/emojis_emotion_prediction')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 131
'train' 105
'validation' 26
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

phán đoán thực nghiệm

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/empirical_judgments')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 99
'train' 80
'validation' 19
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

câu tục ngữ tiếng anh

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/english_proverbs')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 34
'train' 18
'validation' 16
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

tiếng anh_russian_tục ngữ

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/english_russian_proverbs')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 80
'train' 64
'validation' 16
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

đòi hỏi_phân cực

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/entailed_polarity')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 148
'train' 119
'validation' 29
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

đòi hỏi_phân cực_hindi

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/entailed_polarity_hindi')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 138
'train' 111
'validation' 27
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

nhận thức_lý luận

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/epistemic_reasoning')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 2000
'train' 1600
'validation' 400
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

đánh giá_thông tin_thiết yếu

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/evaluating_information_essentiality')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 68
'train' 52
'validation' 16
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

người kiểm tra sự thật

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:bigbench/fact_checker')
  • Sự miêu tả :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • Giấy phép : Giấy phép Apache 2.0
  • Phiên bản : 0.0.0
  • Chia tách :
Tách ra Ví dụ
'default' 7154
'train' 5724
'validation' 14h30
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id":