{
    "agnews": {
        "features": {
            "text": {
                "_type": "Value",
                "dtype": "string"
            },
            "label": {
                "_type": "ClassLabel",
                "names": [
                    "World",
                    "Sports",
                    "Business",
                    "Sci/Tech"
                ]
            }
        },
        "description": "Multi-choice, Topic classification task (Agnews)."
    },
    "amazon-reviews": {
        "features": {
            "text": {
                "_type": "Value",
                "dtype": "string"
            },
            "label": {
                "_type": "Value",
                "dtype": "int8"
            }
        },
        "description": "Multi-choice, Sentiment classification task with numerical labels (Amazon-Review)."
    },
    "imdb": {
        "features": {
            "text": {
                "_type": "Value",
                "dtype": "string"
            },
            "label": {
                "_type": "ClassLabel",
                "names": [
                    "Negative",
                    "Positive"
                ]
            }
        },
        "description": "Binary sentiment classification (IMDB)."
    },
    "commonsenseqa": {
        "features": {
            "text": {
                "_type": "Value",
                "dtype": "string"
            },
            "options": {
                "feature": {
                    "_type": "Value",
                    "dtype": "string"
                },
                "length": 5,
                "_type": "Sequence"
            },
            "label": {
                "_type": "Value",
                "dtype": "int8"
            }
        },
        "description": "Multi-choice, Commense knowledge-answering datasets (CommensenseQA)."
    },
    "fever": {
        "features": {
            "text": {
                "_type": "Value",
                "dtype": "string"
            },
            "label": {
                "_type": "ClassLabel",
                "names": [
                    "Incorrect",
                    "Correct"
                ]
            }
        },
        "description": "Binary classification, fact-checking (FEVER)."
    },
    "myriadlama": {
        "features": {
            "text": {
                "_type": "Value",
                "dtype": "string"
            },
            "options": {
                "feature": {
                    "dtype": "string",
                    "id": null,
                    "_type": "Value"
                },
                "length": 4,
                "_type": "Sequence"
            },
            "label": {
                "_type": "Value",
                "dtype": "int8"
            }
        },
        "description": "Multi-choice, open-domain knowledge question-answering (MyriadLAMA)."
    },
    "templama": {
        "features": {
            "text": {
                "_type": "Value",
                "dtype": "string"
            },
            "options": {
                "feature": {
                    "dtype": "string",
                    "id": null,
                    "_type": "Value"
                },
                "length": 4,
                "_type": "Sequence"
            },
            "label": {
                "_type": "Value",
                "dtype": "int8"
            }
        },
        "description": "Multi-choice, open-domain temporary knowledge QA (TempLAMA)."
    },
    "chunking": {
        "features": {
            "text": {
                "_type": "Value",
                "dtype": "string"
            },
            "chunks": {
                "feature": {
                    "_type": "Sequence",
                    "feature": {
                        "_type": "Value",
                        "dtype": "string"
                    }
                },
                "_type": "Sequence"
            },
            "target_chunk": {
                "_type": "Sequence",
                "feature": {
                    "_type": "Value",
                    "dtype": "string"
                }
            },
            "options": {
                "feature": {
                    "dtype": "string",
                    "id": null,
                    "_type": "Value"
                },
                "length": 4,
                "_type": "Sequence"
            },
            "label": {
                "_type": "Value",
                "dtype": "int8"
            }
        },
        "description": "Multi-choice, Phrase chunking task based on CoNLL 2000 Dataset."
    },
    "ner": {
        "features": {
            "text": {
                "_type": "Value",
                "dtype": "string"
            },
            "chunks": {
                "feature": {
                    "_type": "Sequence",
                    "feature": {
                        "_type": "Value",
                        "dtype": "string"
                    }
                },
                "_type": "Sequence"
            },
            "target_chunk": {
                "_type": "Sequence",
                "feature": {
                    "_type": "Value",
                    "dtype": "string"
                }
            },
            "options": {
                "feature": {
                    "dtype": "string",
                    "id": null,
                    "_type": "Value"
                },
                "length": 4,
                "_type": "Sequence"
            },
            "label": {
                "_type": "Value",
                "dtype": "int8"
            }
        },
        "description": "Multi-choice, named-entity-recognition tagging based on CoNLL 2000 Dataset."
    },
    "postag": {
        "features": {
            "text": {
                "_type": "Value",
                "dtype": "string"
            },
            "tokens": {
                "feature": {
                    "_type": "Value",
                    "dtype": "string"
                },
                "_type": "Sequence"
            },
            "target_token": {
                "_type": "Value",
                "dtype": "string"
            },
            "options": {
                "feature": {
                    "dtype": "string",
                    "id": null,
                    "_type": "Value"
                },
                "length": 4,
                "_type": "Sequence"
            },
            "label": {
                "_type": "Value",
                "dtype": "int8"
            }
        },
        "description": "Multi-choice, part-of-speech tagging based on CoNLL 2003 Dataset."
    },
    "clang8": {
        "features": {
            "false": {
                "_type": "Value",
                "dtype": "string"
            },
            "true": {
                "_type": "Value",
                "dtype": "string"
            }
        },
        "description": "Binary classification, Grammatic error detection (Clang8)."
    },
    "mnli": {
        "features": {
            "premise": {
                "_type": "Value",
                "dtype": "string"
            },
            "hypothesis": {
                "_type": "Value",
                "dtype": "string"
            },
            "label": {
                "_type": "ClassLabel",
                "names": [
                    "Contradiction",
                    "Neutral",
                    "Entailment"
                ]
            }
        },
        "description": "Multi-choice, Entailment inference task (MNLI)."
    },
    "paws": {
        "features": {
            "sentence1": {
                "_type": "Value",
                "dtype": "string"
            },
            "sentence2": {
                "_type": "Value",
                "dtype": "string"
            },
            "label": {
                "_type": "ClassLabel",
                "names": [
                    "False",
                    "True"
                ]
            }
        },
        "description": "Multi-choice, Paraphrase detection (PAWS)."
    },
    "swag": {
        "features": {
            "text": {
                "_type": "Value",
                "dtype": "string"
            },
            "options": {
                "feature": {
                    "_type": "Value",
                    "dtype": "string"
                },
                "length": 4,
                "_type": "Sequence"
            },
            "label": {
                "_type": "Value",
                "dtype": "int8"
            }
        },
        "description": "Multi-choice, Situations With Adversarial Generations task (SWAG)."
    },
    "halueval": {
        "features": {
            "dialogue": {
                "_type": "Value",
                "dtype": "string"
            },
            "response": {
                "_type": "Value",
                "dtype": "string"
            },
            "label": {
                "_type": "ClassLabel",
                "names": [
                    "False",
                    "True"
                ]
            }
        },
        "description": "Binary classification, Hallucination detection (HaluEval-dialogue)."
    },
    "stereoset": {
        "features": {
            "context": {
                "_type": "Value",
                "dtype": "string"
            },
            "options": {
                "feature": {
                    "_type": "Value",
                    "dtype": "string"
                },
                "length": 3,
                "_type": "Sequence"
            },
            "labels": {
                "feature": {
                    "_type": "Value",
                    "dtype": "int8"
                },
                "length": 3,
                "_type": "Sequence"
            }
        },
        "description": "Stereotype detection task"
    },
    "toxicity": {
        "features": {
            "text": {
                "_type": "Value",
                "dtype": "string"
            },
            "label": {
                "_type": "ClassLabel",
                "names": [
                    "False",
                    "True"
                ]
            }
        },
        "description": "Toxicity detection task."
    },
    "amazon-review-multi": {
        "features": {
            "text": {
                "_type": "Value",
                "dtype": "string"
            },
            "label": {
                "_type": "Value",
                "dtype": "int8"
            },
            "lang": {
                "_type": "Value",
                "dtype": "string"
            }
        },
        "description": "Multi-choice, multilingual Amazon reviews task."
    },
    "lti": {
        "features": {
            "text": {
                "_type": "Value",
                "dtype": "string"
            },
            "lang": {
                "_type": "Value",
                "dtype": "string"
            }
        },
        "description": "Language and text identifier task (LTI)."
    },
    "mlama": {
        "features": {
            "uuid": {
                "_type": "Value",
                "dtype": "string"
            },
            "text": {
                "_type": "Value",
                "dtype": "string"
            },
            "options": {
                "feature": {
                    "_type": "Value",
                    "dtype": "string"
                },
                "length": 4,
                "_type": "Sequence"
            },
            "label": {
                "_type": "Value",
                "dtype": "int8"
            },
            "lang": {
                "_type": "Value",
                "dtype": "string"
            }
        },
        "description": "Multi-choice, Multilingual factual knowledge qa (mLAMA)."
    },
    "mpostag": {
        "features": {
            "text": {
                "_type": "Value",
                "dtype": "string"
            },
            "tokens": {
                "feature": {
                    "_type": "Value",
                    "dtype": "string"
                },
                "_type": "Sequence"
            },
            "target_token": {
                "_type": "Value",
                "dtype": "string"
            },
            "options": {
                "feature": {
                    "dtype": "string",
                    "id": null,
                    "_type": "Value"
                },
                "length": 4,
                "_type": "Sequence"
            },
            "label": {
                "_type": "Value",
                "dtype": "int8"
            },
            "lang": {
                "_type": "Value",
                "dtype": "string"
            }
        },
        "description": "Multi-choice, multilingual part-of-speech tagging from Universal Treebank corpora."
    },
    "xnli": {
        "features": {
            "index": {
                "_type": "Value",
                "dtype": "string"
            },
            "premise": {
                "_type": "Value",
                "dtype": "string"
            },
            "hypothesis": {
                "_type": "Value",
                "dtype": "string"
            },
            "label": {
                "_type": "ClassLabel",
                "names": [
                    "Entailment",
                    "Neutral",
                    "Contradiction"
                ]
            },
            "lang": {
                "_type": "Value",
                "dtype": "string"
            }
        },
        "description": "Cross-lingual natural language inference task (XNLI)."
    }
}