md_gender_bias

Tài liệu tham khảo:

• Mã số
• ôm mặt

giới tính_words

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

ds = tfds.load('huggingface:md_gender_bias/gendered_words')

• Sự miêu tả :

Machine learning models are trained to find patterns in data.
NLP models can inadvertently learn socially undesirable patterns when training on gender biased text.
In this work, we propose a general framework that decomposes gender bias in text along several pragmatic and semantic dimensions:
bias from the gender of the person being spoken about, bias from the gender of the person being spoken to, and bias from the gender of the speaker.
Using this fine-grained framework, we automatically annotate eight large scale datasets with gender information.
In addition, we collect a novel, crowdsourced evaluation benchmark of utterance-level gender rewrites.
Distinguishing between gender bias along multiple dimensions is important, as it enables us to train finer-grained gender bias classifiers.
We show our classifiers prove valuable for a variety of important applications, such as controlling for gender bias in generative models,
detecting gender bias in arbitrary text, and shed light on offensive language in terms of genderedness.

• Giấy phép : Giấy phép MIT
• Phiên bản : 1.0.0
• Chia tách :

• Đặc trưng :

{
    "word_masculine": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "word_feminine": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

tên_giới tính

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

ds = tfds.load('huggingface:md_gender_bias/name_genders')

• Sự miêu tả :

Machine learning models are trained to find patterns in data.
NLP models can inadvertently learn socially undesirable patterns when training on gender biased text.
In this work, we propose a general framework that decomposes gender bias in text along several pragmatic and semantic dimensions:
bias from the gender of the person being spoken about, bias from the gender of the person being spoken to, and bias from the gender of the speaker.
Using this fine-grained framework, we automatically annotate eight large scale datasets with gender information.
In addition, we collect a novel, crowdsourced evaluation benchmark of utterance-level gender rewrites.
Distinguishing between gender bias along multiple dimensions is important, as it enables us to train finer-grained gender bias classifiers.
We show our classifiers prove valuable for a variety of important applications, such as controlling for gender bias in generative models,
detecting gender bias in arbitrary text, and shed light on offensive language in terms of genderedness.

• Giấy phép : Giấy phép MIT
• Phiên bản : 1.0.0
• Chia tách :

• Đặc trưng :

{
    "name": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "assigned_gender": {
        "num_classes": 2,
        "names": [
            "M",
            "F"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "count": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

dữ liệu mới

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

ds = tfds.load('huggingface:md_gender_bias/new_data')

• Sự miêu tả :

Machine learning models are trained to find patterns in data.
NLP models can inadvertently learn socially undesirable patterns when training on gender biased text.
In this work, we propose a general framework that decomposes gender bias in text along several pragmatic and semantic dimensions:
bias from the gender of the person being spoken about, bias from the gender of the person being spoken to, and bias from the gender of the speaker.
Using this fine-grained framework, we automatically annotate eight large scale datasets with gender information.
In addition, we collect a novel, crowdsourced evaluation benchmark of utterance-level gender rewrites.
Distinguishing between gender bias along multiple dimensions is important, as it enables us to train finer-grained gender bias classifiers.
We show our classifiers prove valuable for a variety of important applications, such as controlling for gender bias in generative models,
detecting gender bias in arbitrary text, and shed light on offensive language in terms of genderedness.

• Giấy phép : Giấy phép MIT
• Phiên bản : 1.0.0
• Chia tách :

• Đặc trưng :

{
    "text": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "original": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "labels": [
        {
            "num_classes": 6,
            "names": [
                "ABOUT:female",
                "ABOUT:male",
                "PARTNER:female",
                "PARTNER:male",
                "SELF:female",
                "SELF:male"
            ],
            "names_file": null,
            "id": null,
            "_type": "ClassLabel"
        }
    ],
    "class_type": {
        "num_classes": 3,
        "names": [
            "about",
            "partner",
            "self"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "turker_gender": {
        "num_classes": 5,
        "names": [
            "man",
            "woman",
            "nonbinary",
            "prefer not to say",
            "no answer"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "episode_done": {
        "dtype": "bool_",
        "id": null,
        "_type": "Value"
    },
    "confidence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

funpedia

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

ds = tfds.load('huggingface:md_gender_bias/funpedia')

• Sự miêu tả :

Machine learning models are trained to find patterns in data.
NLP models can inadvertently learn socially undesirable patterns when training on gender biased text.
In this work, we propose a general framework that decomposes gender bias in text along several pragmatic and semantic dimensions:
bias from the gender of the person being spoken about, bias from the gender of the person being spoken to, and bias from the gender of the speaker.
Using this fine-grained framework, we automatically annotate eight large scale datasets with gender information.
In addition, we collect a novel, crowdsourced evaluation benchmark of utterance-level gender rewrites.
Distinguishing between gender bias along multiple dimensions is important, as it enables us to train finer-grained gender bias classifiers.
We show our classifiers prove valuable for a variety of important applications, such as controlling for gender bias in generative models,
detecting gender bias in arbitrary text, and shed light on offensive language in terms of genderedness.

• Giấy phép : Giấy phép MIT
• Phiên bản : 1.0.0
• Chia tách :

• Đặc trưng :

{
    "text": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "title": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "persona": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "gender": {
        "num_classes": 3,
        "names": [
            "gender-neutral",
            "female",
            "male"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}

hình ảnh_chat

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

ds = tfds.load('huggingface:md_gender_bias/image_chat')

• Sự miêu tả :

Machine learning models are trained to find patterns in data.
NLP models can inadvertently learn socially undesirable patterns when training on gender biased text.
In this work, we propose a general framework that decomposes gender bias in text along several pragmatic and semantic dimensions:
bias from the gender of the person being spoken about, bias from the gender of the person being spoken to, and bias from the gender of the speaker.
Using this fine-grained framework, we automatically annotate eight large scale datasets with gender information.
In addition, we collect a novel, crowdsourced evaluation benchmark of utterance-level gender rewrites.
Distinguishing between gender bias along multiple dimensions is important, as it enables us to train finer-grained gender bias classifiers.
We show our classifiers prove valuable for a variety of important applications, such as controlling for gender bias in generative models,
detecting gender bias in arbitrary text, and shed light on offensive language in terms of genderedness.

• Giấy phép : Giấy phép MIT
• Phiên bản : 1.0.0
• Chia tách :

• Đặc trưng :

{
    "caption": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "male": {
        "dtype": "bool_",
        "id": null,
        "_type": "Value"
    },
    "female": {
        "dtype": "bool_",
        "id": null,
        "_type": "Value"
    }
}

thuật sĩ

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

ds = tfds.load('huggingface:md_gender_bias/wizard')

• Sự miêu tả :

Machine learning models are trained to find patterns in data.
NLP models can inadvertently learn socially undesirable patterns when training on gender biased text.
In this work, we propose a general framework that decomposes gender bias in text along several pragmatic and semantic dimensions:
bias from the gender of the person being spoken about, bias from the gender of the person being spoken to, and bias from the gender of the speaker.
Using this fine-grained framework, we automatically annotate eight large scale datasets with gender information.
In addition, we collect a novel, crowdsourced evaluation benchmark of utterance-level gender rewrites.
Distinguishing between gender bias along multiple dimensions is important, as it enables us to train finer-grained gender bias classifiers.
We show our classifiers prove valuable for a variety of important applications, such as controlling for gender bias in generative models,
detecting gender bias in arbitrary text, and shed light on offensive language in terms of genderedness.

• Giấy phép : Giấy phép MIT
• Phiên bản : 1.0.0
• Chia tách :

• Đặc trưng :

{
    "text": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "chosen_topic": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "gender": {
        "num_classes": 3,
        "names": [
            "gender-neutral",
            "female",
            "male"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}

convai2_inferred

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

ds = tfds.load('huggingface:md_gender_bias/convai2_inferred')

• Sự miêu tả :

Machine learning models are trained to find patterns in data.
NLP models can inadvertently learn socially undesirable patterns when training on gender biased text.
In this work, we propose a general framework that decomposes gender bias in text along several pragmatic and semantic dimensions:
bias from the gender of the person being spoken about, bias from the gender of the person being spoken to, and bias from the gender of the speaker.
Using this fine-grained framework, we automatically annotate eight large scale datasets with gender information.
In addition, we collect a novel, crowdsourced evaluation benchmark of utterance-level gender rewrites.
Distinguishing between gender bias along multiple dimensions is important, as it enables us to train finer-grained gender bias classifiers.
We show our classifiers prove valuable for a variety of important applications, such as controlling for gender bias in generative models,
detecting gender bias in arbitrary text, and shed light on offensive language in terms of genderedness.

• Giấy phép : Giấy phép MIT
• Phiên bản : 1.0.0
• Chia tách :

• Đặc trưng :

{
    "text": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "binary_label": {
        "num_classes": 2,
        "names": [
            "ABOUT:female",
            "ABOUT:male"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "binary_score": {
        "dtype": "float32",
        "id": null,
        "_type": "Value"
    },
    "ternary_label": {
        "num_classes": 3,
        "names": [
            "ABOUT:female",
            "ABOUT:male",
            "ABOUT:gender-neutral"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "ternary_score": {
        "dtype": "float32",
        "id": null,
        "_type": "Value"
    }
}

light_inferred

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

ds = tfds.load('huggingface:md_gender_bias/light_inferred')

• Sự miêu tả :

Machine learning models are trained to find patterns in data.
NLP models can inadvertently learn socially undesirable patterns when training on gender biased text.
In this work, we propose a general framework that decomposes gender bias in text along several pragmatic and semantic dimensions:
bias from the gender of the person being spoken about, bias from the gender of the person being spoken to, and bias from the gender of the speaker.
Using this fine-grained framework, we automatically annotate eight large scale datasets with gender information.
In addition, we collect a novel, crowdsourced evaluation benchmark of utterance-level gender rewrites.
Distinguishing between gender bias along multiple dimensions is important, as it enables us to train finer-grained gender bias classifiers.
We show our classifiers prove valuable for a variety of important applications, such as controlling for gender bias in generative models,
detecting gender bias in arbitrary text, and shed light on offensive language in terms of genderedness.

• Giấy phép : Giấy phép MIT
• Phiên bản : 1.0.0
• Chia tách :

• Đặc trưng :

{
    "text": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "binary_label": {
        "num_classes": 2,
        "names": [
            "ABOUT:female",
            "ABOUT:male"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "binary_score": {
        "dtype": "float32",
        "id": null,
        "_type": "Value"
    },
    "ternary_label": {
        "num_classes": 3,
        "names": [
            "ABOUT:female",
            "ABOUT:male",
            "ABOUT:gender-neutral"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "ternary_score": {
        "dtype": "float32",
        "id": null,
        "_type": "Value"
    }
}

opensubtitles_inferred

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

ds = tfds.load('huggingface:md_gender_bias/opensubtitles_inferred')

• Sự miêu tả :

Machine learning models are trained to find patterns in data.
NLP models can inadvertently learn socially undesirable patterns when training on gender biased text.
In this work, we propose a general framework that decomposes gender bias in text along several pragmatic and semantic dimensions:
bias from the gender of the person being spoken about, bias from the gender of the person being spoken to, and bias from the gender of the speaker.
Using this fine-grained framework, we automatically annotate eight large scale datasets with gender information.
In addition, we collect a novel, crowdsourced evaluation benchmark of utterance-level gender rewrites.
Distinguishing between gender bias along multiple dimensions is important, as it enables us to train finer-grained gender bias classifiers.
We show our classifiers prove valuable for a variety of important applications, such as controlling for gender bias in generative models,
detecting gender bias in arbitrary text, and shed light on offensive language in terms of genderedness.

• Giấy phép : Giấy phép MIT
• Phiên bản : 1.0.0
• Chia tách :

• Đặc trưng :

{
    "text": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "binary_label": {
        "num_classes": 2,
        "names": [
            "ABOUT:female",
            "ABOUT:male"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "binary_score": {
        "dtype": "float32",
        "id": null,
        "_type": "Value"
    },
    "ternary_label": {
        "num_classes": 3,
        "names": [
            "ABOUT:female",
            "ABOUT:male",
            "ABOUT:gender-neutral"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "ternary_score": {
        "dtype": "float32",
        "id": null,
        "_type": "Value"
    }
}

yelp_inferred

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

ds = tfds.load('huggingface:md_gender_bias/yelp_inferred')

• Sự miêu tả :

Machine learning models are trained to find patterns in data.
NLP models can inadvertently learn socially undesirable patterns when training on gender biased text.
In this work, we propose a general framework that decomposes gender bias in text along several pragmatic and semantic dimensions:
bias from the gender of the person being spoken about, bias from the gender of the person being spoken to, and bias from the gender of the speaker.
Using this fine-grained framework, we automatically annotate eight large scale datasets with gender information.
In addition, we collect a novel, crowdsourced evaluation benchmark of utterance-level gender rewrites.
Distinguishing between gender bias along multiple dimensions is important, as it enables us to train finer-grained gender bias classifiers.
We show our classifiers prove valuable for a variety of important applications, such as controlling for gender bias in generative models,
detecting gender bias in arbitrary text, and shed light on offensive language in terms of genderedness.

• Giấy phép : Giấy phép MIT
• Phiên bản : 1.0.0
• Chia tách :

• Đặc trưng :

{
    "text": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "binary_label": {
        "num_classes": 2,
        "names": [
            "ABOUT:female",
            "ABOUT:male"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "binary_score": {
        "dtype": "float32",
        "id": null,
        "_type": "Value"
    }
}