Add dataset loading script; Readme skeleton; metadata json

README.md

@@ -0,0 +1,182 @@
+---
+annotations_creators:
+- expert-generated
+language_creators:
+- found
+- machine-generated
+- expert-generated
+languages:
+- bn-hi
+- cs-en
+- de-en
+- de-fr
+- en-cs
+- en-de
+- en-ha
+- en-is
+- en-ja
+- en-ru
+- en-zh
+- fr-de
+- ha-en
+- hi-bn
+- is-en
+- ja-en
+- ru-en
+- xh-zh
+- zh-en
+- zu-xh
+licenses:
+- unknown
+multilinguality:
+- translation
+paperswithcode_id: null
+pretty_name: WMT21 Metrics Shared Task
+size_categories:
+- 100K<n<1M
+source_datasets: []
+task_categories:
+- translation
+task_ids: []
+---
+
+# Dataset Card for WMT21 Metrics Task
+
+## Table of Contents
+- [Table of Contents](#table-of-contents)
+- [Dataset Description](#dataset-description)
+  - [Dataset Summary](#dataset-summary)
+  - [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
+  - [Languages](#languages)
+- [Dataset Structure](#dataset-structure)
+  - [Data Instances](#data-instances)
+  - [Data Fields](#data-fields)
+  - [Data Splits](#data-splits)
+- [Dataset Creation](#dataset-creation)
+  - [Curation Rationale](#curation-rationale)
+  - [Source Data](#source-data)
+  - [Annotations](#annotations)
+  - [Personal and Sensitive Information](#personal-and-sensitive-information)
+- [Considerations for Using the Data](#considerations-for-using-the-data)
+  - [Social Impact of Dataset](#social-impact-of-dataset)
+  - [Discussion of Biases](#discussion-of-biases)
+  - [Other Known Limitations](#other-known-limitations)
+- [Additional Information](#additional-information)
+  - [Dataset Curators](#dataset-curators)
+  - [Licensing Information](#licensing-information)
+  - [Citation Information](#citation-information)
+  - [Contributions](#contributions)
+
+## Dataset Description
+
+- **Homepage:** [WMT21 Metrics Shared Task](https://www.statmt.org/wmt21/metrics-task.html)
+- **Repository:** [MT Metrics Eval Github Repository](https://github.com/google-research/mt-metrics-eval)
+- **Paper:** [Paper](https://aclanthology.org/2021.wmt-1.73/)
+
+### Dataset Summary
+
+[More Information Needed]
+
+### Supported Tasks and Leaderboards
+
+[More Information Needed]
+
+### Languages
+
+The dataset comprises twenty language pairs:
+- Bengali-Hindi (`bn-hi`)
+- Czech-English (`cs-en`)
+- German-English (`de-en`)
+- German-French (`de-fr`)
+- English-Czech (`en-cs`)
+- English-German (`en-de`)
+- English-Hausa (`en-ha`)
+- English-Icelandic (`en-is`)
+- English-Japanese (`en-ja`)
+- English-Russian (`en-ru`)
+- English-Chinese (`en-zh`)
+- French-German (`fr-de`)
+- Hausa-English (`ha-en`)
+- Hindi-Bengali (`hi-bn`)
+- Icelandic-English (`is-en`)
+- Japenese-English (`ja-en`)
+- Russian-English (`ru-en`)
+- Xhosa-Zulu (`xh-zu`)
+- Chinese-English (`zh-en`)
+- Zulu-Xhosa (`zu-xh`)
+
+## Dataset Structure
+
+### Data Instances
+
+[More Information Needed]
+
+### Data Fields
+
+[More Information Needed]
+
+### Data Splits
+
+[More Information Needed]
+
+## Dataset Creation
+
+### Curation Rationale
+
+[More Information Needed]
+
+### Source Data
+
+#### Initial Data Collection and Normalization
+
+[More Information Needed]
+
+#### Who are the source language producers?
+
+[More Information Needed]
+
+### Annotations
+
+#### Annotation process
+
+[More Information Needed]
+
+#### Who are the annotators?
+
+[More Information Needed]
+
+### Personal and Sensitive Information
+
+[More Information Needed]
+
+## Considerations for Using the Data
+
+### Social Impact of Dataset
+
+[More Information Needed]
+
+### Discussion of Biases
+
+[More Information Needed]
+
+### Other Known Limitations
+
+[More Information Needed]
+
+## Additional Information
+
+### Dataset Curators
+
+[More Information Needed]
+
+### Licensing Information
+
+[More Information Needed]
+
+### Citation Information
+
+[More Information Needed]
+
+### Contributions
+
+Thanks to [@github-username](https://github.com/<github-username>) for adding this dataset.

dataset_infos.json

@@ -0,0 +1,2222 @@
+{
+    "bn-hi": {
+        "description": "This shared task will examine automatic evaluation metrics for machine translation. We will \nprovide you with MT system outputs along with source text and the human reference translations. \nWe are looking for automatic metric scores for translations at the system-level, and segment-level. \nWe will calculate the system-level, and segment-level correlations of your scores with human judgements.\n\nWe invite submissions of reference-free metrics in addition to reference-based metrics.\n",
+        "citation": "@inproceedings{freitag-etal-2021-results,\n    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},\n    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}\",\n    booktitle = {Proceedings of the Sixth Conference on Machine Translation},\n    month = {nov},\n    year = {2021},\n    address = {Online},\n    publisher = {Association for Computational Linguistics},\n    url = {https://aclanthology.org/2021.wmt-1.73},\n    pages = {733--774}\n}\n",
+        "homepage": "https://www.statmt.org/wmt21/metrics-task.html",
+        "license": "Unknown",
+        "features": {
+            "translation": {
+                "languages": [
+                    "bn",
+                    "hi"
+                ],
+                "id": null,
+                "_type": "Translation"
+            },
+            "mt_system": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "mqm": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-raw": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-z": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "pair": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "dataset": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "sent_id": {
+                "dtype": "int32",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_name": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_ref": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-A": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-B": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-C": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-D": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            }
+        },
+        "post_processed": null,
+        "supervised_keys": null,
+        "task_templates": null,
+        "builder_name": "wmt21_metrics_task",
+        "config_name": "bn-hi",
+        "version": {
+            "version_str": "1.1.0",
+            "description": null,
+            "major": 1,
+            "minor": 1,
+            "patch": 0
+        },
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_bytes": 5003705,
+                "num_examples": 4527,
+                "dataset_name": "wmt21_metrics_task"
+            }
+        },
+        "download_checksums": {
+            "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/bn-hi/train.csv": {
+                "num_bytes": 4910374,
+                "checksum": "760560318837ee72176059bb127f8438ee440164d1ead355ddcdb00bbcc73dde"
+            }
+        },
+        "download_size": 4910374,
+        "post_processing_size": null,
+        "dataset_size": 5003705,
+        "size_in_bytes": 9914079
+    },
+    "hi-bn": {
+        "description": "This shared task will examine automatic evaluation metrics for machine translation. We will \nprovide you with MT system outputs along with source text and the human reference translations. \nWe are looking for automatic metric scores for translations at the system-level, and segment-level. \nWe will calculate the system-level, and segment-level correlations of your scores with human judgements.\n\nWe invite submissions of reference-free metrics in addition to reference-based metrics.\n",
+        "citation": "@inproceedings{freitag-etal-2021-results,\n    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},\n    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}\",\n    booktitle = {Proceedings of the Sixth Conference on Machine Translation},\n    month = {nov},\n    year = {2021},\n    address = {Online},\n    publisher = {Association for Computational Linguistics},\n    url = {https://aclanthology.org/2021.wmt-1.73},\n    pages = {733--774}\n}\n",
+        "homepage": "https://www.statmt.org/wmt21/metrics-task.html",
+        "license": "Unknown",
+        "features": {
+            "translation": {
+                "languages": [
+                    "hi",
+                    "bn"
+                ],
+                "id": null,
+                "_type": "Translation"
+            },
+            "mt_system": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "mqm": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-raw": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-z": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "pair": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "dataset": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "sent_id": {
+                "dtype": "int32",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_name": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_ref": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-A": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-B": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-C": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-D": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            }
+        },
+        "post_processed": null,
+        "supervised_keys": null,
+        "task_templates": null,
+        "builder_name": "wmt21_metrics_task",
+        "config_name": "hi-bn",
+        "version": {
+            "version_str": "1.1.0",
+            "description": null,
+            "major": 1,
+            "minor": 1,
+            "patch": 0
+        },
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_bytes": 4994664,
+                "num_examples": 4581,
+                "dataset_name": "wmt21_metrics_task"
+            }
+        },
+        "download_checksums": {
+            "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/hi-bn/train.csv": {
+                "num_bytes": 4896905,
+                "checksum": "16c1ddb44d4244b2aa18720000ab0e1583a7dd0d5d7f1b4e145bf7cd5015679a"
+            }
+        },
+        "download_size": 4896905,
+        "post_processing_size": null,
+        "dataset_size": 4994664,
+        "size_in_bytes": 9891569
+    },
+    "xh-zu": {
+        "description": "This shared task will examine automatic evaluation metrics for machine translation. We will \nprovide you with MT system outputs along with source text and the human reference translations. \nWe are looking for automatic metric scores for translations at the system-level, and segment-level. \nWe will calculate the system-level, and segment-level correlations of your scores with human judgements.\n\nWe invite submissions of reference-free metrics in addition to reference-based metrics.\n",
+        "citation": "@inproceedings{freitag-etal-2021-results,\n    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},\n    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}\",\n    booktitle = {Proceedings of the Sixth Conference on Machine Translation},\n    month = {nov},\n    year = {2021},\n    address = {Online},\n    publisher = {Association for Computational Linguistics},\n    url = {https://aclanthology.org/2021.wmt-1.73},\n    pages = {733--774}\n}\n",
+        "homepage": "https://www.statmt.org/wmt21/metrics-task.html",
+        "license": "Unknown",
+        "features": {
+            "translation": {
+                "languages": [
+                    "xh",
+                    "zu"
+                ],
+                "id": null,
+                "_type": "Translation"
+            },
+            "mt_system": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "mqm": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-raw": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-z": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "pair": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "dataset": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "sent_id": {
+                "dtype": "int32",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_name": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_ref": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-A": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-B": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-C": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-D": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            }
+        },
+        "post_processed": null,
+        "supervised_keys": null,
+        "task_templates": null,
+        "builder_name": "wmt21_metrics_task",
+        "config_name": "xh-zu",
+        "version": {
+            "version_str": "1.1.0",
+            "description": null,
+            "major": 1,
+            "minor": 1,
+            "patch": 0
+        },
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_bytes": 1566816,
+                "num_examples": 3018,
+                "dataset_name": "wmt21_metrics_task"
+            }
+        },
+        "download_checksums": {
+            "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/xh-zu/train.csv": {
+                "num_bytes": 1501982,
+                "checksum": "6c0f0b065c9ae6fccc5c501af7de8b53a295b1928dee5dd711504ce2bbb0907d"
+            }
+        },
+        "download_size": 1501982,
+        "post_processing_size": null,
+        "dataset_size": 1566816,
+        "size_in_bytes": 3068798
+    },
+    "zu-xh": {
+        "description": "This shared task will examine automatic evaluation metrics for machine translation. We will \nprovide you with MT system outputs along with source text and the human reference translations. \nWe are looking for automatic metric scores for translations at the system-level, and segment-level. \nWe will calculate the system-level, and segment-level correlations of your scores with human judgements.\n\nWe invite submissions of reference-free metrics in addition to reference-based metrics.\n",
+        "citation": "@inproceedings{freitag-etal-2021-results,\n    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},\n    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}\",\n    booktitle = {Proceedings of the Sixth Conference on Machine Translation},\n    month = {nov},\n    year = {2021},\n    address = {Online},\n    publisher = {Association for Computational Linguistics},\n    url = {https://aclanthology.org/2021.wmt-1.73},\n    pages = {733--774}\n}\n",
+        "homepage": "https://www.statmt.org/wmt21/metrics-task.html",
+        "license": "Unknown",
+        "features": {
+            "translation": {
+                "languages": [
+                    "zu",
+                    "xh"
+                ],
+                "id": null,
+                "_type": "Translation"
+            },
+            "mt_system": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "mqm": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-raw": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-z": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "pair": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "dataset": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "sent_id": {
+                "dtype": "int32",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_name": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_ref": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-A": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-B": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-C": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-D": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            }
+        },
+        "post_processed": null,
+        "supervised_keys": null,
+        "task_templates": null,
+        "builder_name": "wmt21_metrics_task",
+        "config_name": "zu-xh",
+        "version": {
+            "version_str": "1.1.0",
+            "description": null,
+            "major": 1,
+            "minor": 1,
+            "patch": 0
+        },
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_bytes": 1326809,
+                "num_examples": 2545,
+                "dataset_name": "wmt21_metrics_task"
+            }
+        },
+        "download_checksums": {
+            "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/zu-xh/train.csv": {
+                "num_bytes": 1271100,
+                "checksum": "a48331a0414e34eb41b8b27d0f24f937005ff627e4dc75897960141bbd273fe9"
+            }
+        },
+        "download_size": 1271100,
+        "post_processing_size": null,
+        "dataset_size": 1326809,
+        "size_in_bytes": 2597909
+    },
+    "cs-en": {
+        "description": "This shared task will examine automatic evaluation metrics for machine translation. We will \nprovide you with MT system outputs along with source text and the human reference translations. \nWe are looking for automatic metric scores for translations at the system-level, and segment-level. \nWe will calculate the system-level, and segment-level correlations of your scores with human judgements.\n\nWe invite submissions of reference-free metrics in addition to reference-based metrics.\n",
+        "citation": "@inproceedings{freitag-etal-2021-results,\n    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},\n    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}\",\n    booktitle = {Proceedings of the Sixth Conference on Machine Translation},\n    month = {nov},\n    year = {2021},\n    address = {Online},\n    publisher = {Association for Computational Linguistics},\n    url = {https://aclanthology.org/2021.wmt-1.73},\n    pages = {733--774}\n}\n",
+        "homepage": "https://www.statmt.org/wmt21/metrics-task.html",
+        "license": "Unknown",
+        "features": {
+            "translation": {
+                "languages": [
+                    "cs",
+                    "en"
+                ],
+                "id": null,
+                "_type": "Translation"
+            },
+            "mt_system": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "mqm": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-raw": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-z": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "pair": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "dataset": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "sent_id": {
+                "dtype": "int32",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_name": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_ref": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-A": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-B": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-C": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-D": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            }
+        },
+        "post_processed": null,
+        "supervised_keys": null,
+        "task_templates": null,
+        "builder_name": "wmt21_metrics_task",
+        "config_name": "cs-en",
+        "version": {
+            "version_str": "1.1.0",
+            "description": null,
+            "major": 1,
+            "minor": 1,
+            "patch": 0
+        },
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_bytes": 5147689,
+                "num_examples": 9000,
+                "dataset_name": "wmt21_metrics_task"
+            }
+        },
+        "download_checksums": {
+            "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/cs-en/train.csv": {
+                "num_bytes": 4954537,
+                "checksum": "a6a11d32f62f49f8b09dd9b8ddda32732ab5f1c7e5855503412393ee42e1ddd2"
+            }
+        },
+        "download_size": 4954537,
+        "post_processing_size": null,
+        "dataset_size": 5147689,
+        "size_in_bytes": 10102226
+    },
+    "de-en": {
+        "description": "This shared task will examine automatic evaluation metrics for machine translation. We will \nprovide you with MT system outputs along with source text and the human reference translations. \nWe are looking for automatic metric scores for translations at the system-level, and segment-level. \nWe will calculate the system-level, and segment-level correlations of your scores with human judgements.\n\nWe invite submissions of reference-free metrics in addition to reference-based metrics.\n",
+        "citation": "@inproceedings{freitag-etal-2021-results,\n    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},\n    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}\",\n    booktitle = {Proceedings of the Sixth Conference on Machine Translation},\n    month = {nov},\n    year = {2021},\n    address = {Online},\n    publisher = {Association for Computational Linguistics},\n    url = {https://aclanthology.org/2021.wmt-1.73},\n    pages = {733--774}\n}\n",
+        "homepage": "https://www.statmt.org/wmt21/metrics-task.html",
+        "license": "Unknown",
+        "features": {
+            "translation": {
+                "languages": [
+                    "de",
+                    "en"
+                ],
+                "id": null,
+                "_type": "Translation"
+            },
+            "mt_system": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "mqm": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-raw": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-z": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "pair": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "dataset": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "sent_id": {
+                "dtype": "int32",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_name": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_ref": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-A": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-B": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-C": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-D": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            }
+        },
+        "post_processed": null,
+        "supervised_keys": null,
+        "task_templates": null,
+        "builder_name": "wmt21_metrics_task",
+        "config_name": "de-en",
+        "version": {
+            "version_str": "1.1.0",
+            "description": null,
+            "major": 1,
+            "minor": 1,
+            "patch": 0
+        },
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_bytes": 10895095,
+                "num_examples": 20000,
+                "dataset_name": "wmt21_metrics_task"
+            }
+        },
+        "download_checksums": {
+            "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/de-en/train.csv": {
+                "num_bytes": 10486849,
+                "checksum": "cb7218ebb5bc39545eb340d11476449e15eb2bbb454debecf7533326ef6993eb"
+            }
+        },
+        "download_size": 10486849,
+        "post_processing_size": null,
+        "dataset_size": 10895095,
+        "size_in_bytes": 21381944
+    },
+    "de-fr": {
+        "description": "This shared task will examine automatic evaluation metrics for machine translation. We will \nprovide you with MT system outputs along with source text and the human reference translations. \nWe are looking for automatic metric scores for translations at the system-level, and segment-level. \nWe will calculate the system-level, and segment-level correlations of your scores with human judgements.\n\nWe invite submissions of reference-free metrics in addition to reference-based metrics.\n",
+        "citation": "@inproceedings{freitag-etal-2021-results,\n    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},\n    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}\",\n    booktitle = {Proceedings of the Sixth Conference on Machine Translation},\n    month = {nov},\n    year = {2021},\n    address = {Online},\n    publisher = {Association for Computational Linguistics},\n    url = {https://aclanthology.org/2021.wmt-1.73},\n    pages = {733--774}\n}\n",
+        "homepage": "https://www.statmt.org/wmt21/metrics-task.html",
+        "license": "Unknown",
+        "features": {
+            "translation": {
+                "languages": [
+                    "de",
+                    "fr"
+                ],
+                "id": null,
+                "_type": "Translation"
+            },
+            "mt_system": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "mqm": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-raw": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-z": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "pair": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "dataset": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "sent_id": {
+                "dtype": "int32",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_name": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_ref": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-A": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-B": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-C": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-D": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            }
+        },
+        "post_processed": null,
+        "supervised_keys": null,
+        "task_templates": null,
+        "builder_name": "wmt21_metrics_task",
+        "config_name": "de-fr",
+        "version": {
+            "version_str": "1.1.0",
+            "description": null,
+            "major": 1,
+            "minor": 1,
+            "patch": 0
+        },
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_bytes": 5004293,
+                "num_examples": 10000,
+                "dataset_name": "wmt21_metrics_task"
+            }
+        },
+        "download_checksums": {
+            "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/de-fr/train.csv": {
+                "num_bytes": 4803809,
+                "checksum": "6476bbd760a91181b688961009eb2de0228d199e46f353032ed215d297d4ff22"
+            }
+        },
+        "download_size": 4803809,
+        "post_processing_size": null,
+        "dataset_size": 5004293,
+        "size_in_bytes": 9808102
+    },
+    "en-cs": {
+        "description": "This shared task will examine automatic evaluation metrics for machine translation. We will \nprovide you with MT system outputs along with source text and the human reference translations. \nWe are looking for automatic metric scores for translations at the system-level, and segment-level. \nWe will calculate the system-level, and segment-level correlations of your scores with human judgements.\n\nWe invite submissions of reference-free metrics in addition to reference-based metrics.\n",
+        "citation": "@inproceedings{freitag-etal-2021-results,\n    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},\n    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}\",\n    booktitle = {Proceedings of the Sixth Conference on Machine Translation},\n    month = {nov},\n    year = {2021},\n    address = {Online},\n    publisher = {Association for Computational Linguistics},\n    url = {https://aclanthology.org/2021.wmt-1.73},\n    pages = {733--774}\n}\n",
+        "homepage": "https://www.statmt.org/wmt21/metrics-task.html",
+        "license": "Unknown",
+        "features": {
+            "translation": {
+                "languages": [
+                    "en",
+                    "cs"
+                ],
+                "id": null,
+                "_type": "Translation"
+            },
+            "mt_system": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "mqm": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-raw": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-z": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "pair": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "dataset": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "sent_id": {
+                "dtype": "int32",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_name": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_ref": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-A": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-B": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-C": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-D": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            }
+        },
+        "post_processed": null,
+        "supervised_keys": null,
+        "task_templates": null,
+        "builder_name": "wmt21_metrics_task",
+        "config_name": "en-cs",
+        "version": {
+            "version_str": "1.1.0",
+            "description": null,
+            "major": 1,
+            "minor": 1,
+            "patch": 0
+        },
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_bytes": 6339403,
+                "num_examples": 11022,
+                "dataset_name": "wmt21_metrics_task"
+            }
+        },
+        "download_checksums": {
+            "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/en-cs/train.csv": {
+                "num_bytes": 6166368,
+                "checksum": "bf5ba4da7b1431f0fb4f0bb36e0ed3ea7de3e28ae68c2569ca9f6ce69d963c1c"
+            }
+        },
+        "download_size": 6166368,
+        "post_processing_size": null,
+        "dataset_size": 6339403,
+        "size_in_bytes": 12505771
+    },
+    "en-de": {
+        "description": "This shared task will examine automatic evaluation metrics for machine translation. We will \nprovide you with MT system outputs along with source text and the human reference translations. \nWe are looking for automatic metric scores for translations at the system-level, and segment-level. \nWe will calculate the system-level, and segment-level correlations of your scores with human judgements.\n\nWe invite submissions of reference-free metrics in addition to reference-based metrics.\n",
+        "citation": "@inproceedings{freitag-etal-2021-results,\n    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},\n    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}\",\n    booktitle = {Proceedings of the Sixth Conference on Machine Translation},\n    month = {nov},\n    year = {2021},\n    address = {Online},\n    publisher = {Association for Computational Linguistics},\n    url = {https://aclanthology.org/2021.wmt-1.73},\n    pages = {733--774}\n}\n",
+        "homepage": "https://www.statmt.org/wmt21/metrics-task.html",
+        "license": "Unknown",
+        "features": {
+            "translation": {
+                "languages": [
+                    "en",
+                    "de"
+                ],
+                "id": null,
+                "_type": "Translation"
+            },
+            "mt_system": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "mqm": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-raw": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-z": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "pair": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "dataset": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "sent_id": {
+                "dtype": "int32",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_name": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_ref": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-A": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-B": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-C": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-D": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            }
+        },
+        "post_processed": null,
+        "supervised_keys": null,
+        "task_templates": null,
+        "builder_name": "wmt21_metrics_task",
+        "config_name": "en-de",
+        "version": {
+            "version_str": "1.1.0",
+            "description": null,
+            "major": 1,
+            "minor": 1,
+            "patch": 0
+        },
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_bytes": 36195641,
+                "num_examples": 37542,
+                "dataset_name": "wmt21_metrics_task"
+            }
+        },
+        "download_checksums": {
+            "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/en-de/train.csv": {
+                "num_bytes": 35044309,
+                "checksum": "e1e10a46c9a059494ef588f64d7eaf14f18bb1406582e77d49c0365a0bf7d609"
+            }
+        },
+        "download_size": 35044309,
+        "post_processing_size": null,
+        "dataset_size": 36195641,
+        "size_in_bytes": 71239950
+    },
+    "en-ha": {
+        "description": "This shared task will examine automatic evaluation metrics for machine translation. We will \nprovide you with MT system outputs along with source text and the human reference translations. \nWe are looking for automatic metric scores for translations at the system-level, and segment-level. \nWe will calculate the system-level, and segment-level correlations of your scores with human judgements.\n\nWe invite submissions of reference-free metrics in addition to reference-based metrics.\n",
+        "citation": "@inproceedings{freitag-etal-2021-results,\n    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},\n    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}\",\n    booktitle = {Proceedings of the Sixth Conference on Machine Translation},\n    month = {nov},\n    year = {2021},\n    address = {Online},\n    publisher = {Association for Computational Linguistics},\n    url = {https://aclanthology.org/2021.wmt-1.73},\n    pages = {733--774}\n}\n",
+        "homepage": "https://www.statmt.org/wmt21/metrics-task.html",
+        "license": "Unknown",
+        "features": {
+            "translation": {
+                "languages": [
+                    "en",
+                    "ha"
+                ],
+                "id": null,
+                "_type": "Translation"
+            },
+            "mt_system": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "mqm": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-raw": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-z": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "pair": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "dataset": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "sent_id": {
+                "dtype": "int32",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_name": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_ref": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-A": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-B": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-C": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-D": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            }
+        },
+        "post_processed": null,
+        "supervised_keys": null,
+        "task_templates": null,
+        "builder_name": "wmt21_metrics_task",
+        "config_name": "en-ha",
+        "version": {
+            "version_str": "1.1.0",
+            "description": null,
+            "major": 1,
+            "minor": 1,
+            "patch": 0
+        },
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_bytes": 8466827,
+                "num_examples": 15000,
+                "dataset_name": "wmt21_metrics_task"
+            }
+        },
+        "download_checksums": {
+            "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/en-ha/train.csv": {
+                "num_bytes": 8124624,
+                "checksum": "12222a116e8b73a009c885c5c831d856aca0e95e643f334d3f9f9b0dd6b19710"
+            }
+        },
+        "download_size": 8124624,
+        "post_processing_size": null,
+        "dataset_size": 8466827,
+        "size_in_bytes": 16591451
+    },
+    "en-is": {
+        "description": "This shared task will examine automatic evaluation metrics for machine translation. We will \nprovide you with MT system outputs along with source text and the human reference translations. \nWe are looking for automatic metric scores for translations at the system-level, and segment-level. \nWe will calculate the system-level, and segment-level correlations of your scores with human judgements.\n\nWe invite submissions of reference-free metrics in addition to reference-based metrics.\n",
+        "citation": "@inproceedings{freitag-etal-2021-results,\n    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},\n    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}\",\n    booktitle = {Proceedings of the Sixth Conference on Machine Translation},\n    month = {nov},\n    year = {2021},\n    address = {Online},\n    publisher = {Association for Computational Linguistics},\n    url = {https://aclanthology.org/2021.wmt-1.73},\n    pages = {733--774}\n}\n",
+        "homepage": "https://www.statmt.org/wmt21/metrics-task.html",
+        "license": "Unknown",
+        "features": {
+            "translation": {
+                "languages": [
+                    "en",
+                    "is"
+                ],
+                "id": null,
+                "_type": "Translation"
+            },
+            "mt_system": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "mqm": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-raw": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-z": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "pair": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "dataset": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "sent_id": {
+                "dtype": "int32",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_name": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_ref": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-A": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-B": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-C": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-D": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            }
+        },
+        "post_processed": null,
+        "supervised_keys": null,
+        "task_templates": null,
+        "builder_name": "wmt21_metrics_task",
+        "config_name": "en-is",
+        "version": {
+            "version_str": "1.1.0",
+            "description": null,
+            "major": 1,
+            "minor": 1,
+            "patch": 0
+        },
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_bytes": 6797138,
+                "num_examples": 12000,
+                "dataset_name": "wmt21_metrics_task"
+            }
+        },
+        "download_checksums": {
+            "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/en-is/train.csv": {
+                "num_bytes": 6550772,
+                "checksum": "53e0297c3791bb696d7dd654bd55fcc9ccb9f5086702fdb654e9ba37b063c5c8"
+            }
+        },
+        "download_size": 6550772,
+        "post_processing_size": null,
+        "dataset_size": 6797138,
+        "size_in_bytes": 13347910
+    },
+    "en-ja": {
+        "description": "This shared task will examine automatic evaluation metrics for machine translation. We will \nprovide you with MT system outputs along with source text and the human reference translations. \nWe are looking for automatic metric scores for translations at the system-level, and segment-level. \nWe will calculate the system-level, and segment-level correlations of your scores with human judgements.\n\nWe invite submissions of reference-free metrics in addition to reference-based metrics.\n",
+        "citation": "@inproceedings{freitag-etal-2021-results,\n    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},\n    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}\",\n    booktitle = {Proceedings of the Sixth Conference on Machine Translation},\n    month = {nov},\n    year = {2021},\n    address = {Online},\n    publisher = {Association for Computational Linguistics},\n    url = {https://aclanthology.org/2021.wmt-1.73},\n    pages = {733--774}\n}\n",
+        "homepage": "https://www.statmt.org/wmt21/metrics-task.html",
+        "license": "Unknown",
+        "features": {
+            "translation": {
+                "languages": [
+                    "en",
+                    "ja"
+                ],
+                "id": null,
+                "_type": "Translation"
+            },
+            "mt_system": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "mqm": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-raw": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-z": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "pair": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "dataset": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "sent_id": {
+                "dtype": "int32",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_name": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_ref": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-A": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-B": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-C": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-D": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            }
+        },
+        "post_processed": null,
+        "supervised_keys": null,
+        "task_templates": null,
+        "builder_name": "wmt21_metrics_task",
+        "config_name": "en-ja",
+        "version": {
+            "version_str": "1.1.0",
+            "description": null,
+            "major": 1,
+            "minor": 1,
+            "patch": 0
+        },
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_bytes": 9670276,
+                "num_examples": 16000,
+                "dataset_name": "wmt21_metrics_task"
+            }
+        },
+        "download_checksums": {
+            "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/en-ja/train.csv": {
+                "num_bytes": 9388263,
+                "checksum": "1d24e564ad74a9a85f1402bfb9b7ccd9e4c582b6ef3b84ba0b76083c37ede2d2"
+            }
+        },
+        "download_size": 9388263,
+        "post_processing_size": null,
+        "dataset_size": 9670276,
+        "size_in_bytes": 19058539
+    },
+    "en-ru": {
+        "description": "This shared task will examine automatic evaluation metrics for machine translation. We will \nprovide you with MT system outputs along with source text and the human reference translations. \nWe are looking for automatic metric scores for translations at the system-level, and segment-level. \nWe will calculate the system-level, and segment-level correlations of your scores with human judgements.\n\nWe invite submissions of reference-free metrics in addition to reference-based metrics.\n",
+        "citation": "@inproceedings{freitag-etal-2021-results,\n    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},\n    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}\",\n    booktitle = {Proceedings of the Sixth Conference on Machine Translation},\n    month = {nov},\n    year = {2021},\n    address = {Online},\n    publisher = {Association for Computational Linguistics},\n    url = {https://aclanthology.org/2021.wmt-1.73},\n    pages = {733--774}\n}\n",
+        "homepage": "https://www.statmt.org/wmt21/metrics-task.html",
+        "license": "Unknown",
+        "features": {
+            "translation": {
+                "languages": [
+                    "en",
+                    "ru"
+                ],
+                "id": null,
+                "_type": "Translation"
+            },
+            "mt_system": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "mqm": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-raw": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-z": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "pair": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "dataset": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "sent_id": {
+                "dtype": "int32",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_name": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_ref": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-A": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-B": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-C": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-D": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            }
+        },
+        "post_processed": null,
+        "supervised_keys": null,
+        "task_templates": null,
+        "builder_name": "wmt21_metrics_task",
+        "config_name": "en-ru",
+        "version": {
+            "version_str": "1.1.0",
+            "description": null,
+            "major": 1,
+            "minor": 1,
+            "patch": 0
+        },
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_bytes": 22135859,
+                "num_examples": 23712,
+                "dataset_name": "wmt21_metrics_task"
+            }
+        },
+        "download_checksums": {
+            "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/en-ru/train.csv": {
+                "num_bytes": 21552005,
+                "checksum": "dcbe28e2ea368e5ca41f8957c0da8fcce23b5e81fec5ca9352b54af52b0e9c2d"
+            }
+        },
+        "download_size": 21552005,
+        "post_processing_size": null,
+        "dataset_size": 22135859,
+        "size_in_bytes": 43687864
+    },
+    "en-zh": {
+        "description": "This shared task will examine automatic evaluation metrics for machine translation. We will \nprovide you with MT system outputs along with source text and the human reference translations. \nWe are looking for automatic metric scores for translations at the system-level, and segment-level. \nWe will calculate the system-level, and segment-level correlations of your scores with human judgements.\n\nWe invite submissions of reference-free metrics in addition to reference-based metrics.\n",
+        "citation": "@inproceedings{freitag-etal-2021-results,\n    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},\n    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}\",\n    booktitle = {Proceedings of the Sixth Conference on Machine Translation},\n    month = {nov},\n    year = {2021},\n    address = {Online},\n    publisher = {Association for Computational Linguistics},\n    url = {https://aclanthology.org/2021.wmt-1.73},\n    pages = {733--774}\n}\n",
+        "homepage": "https://www.statmt.org/wmt21/metrics-task.html",
+        "license": "Unknown",
+        "features": {
+            "translation": {
+                "languages": [
+                    "en",
+                    "zh"
+                ],
+                "id": null,
+                "_type": "Translation"
+            },
+            "mt_system": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "mqm": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-raw": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-z": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "pair": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "dataset": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "sent_id": {
+                "dtype": "int32",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_name": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_ref": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-A": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-B": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-C": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-D": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            }
+        },
+        "post_processed": null,
+        "supervised_keys": null,
+        "task_templates": null,
+        "builder_name": "wmt21_metrics_task",
+        "config_name": "en-zh",
+        "version": {
+            "version_str": "1.1.0",
+            "description": null,
+            "major": 1,
+            "minor": 1,
+            "patch": 0
+        },
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_bytes": 20183942,
+                "num_examples": 31062,
+                "dataset_name": "wmt21_metrics_task"
+            }
+        },
+        "download_checksums": {
+            "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/en-zh/train.csv": {
+                "num_bytes": 19614722,
+                "checksum": "9cdabeca45c58359a6e604fbb1c20967ad7b78581289f2ad48fd710a69810445"
+            }
+        },
+        "download_size": 19614722,
+        "post_processing_size": null,
+        "dataset_size": 20183942,
+        "size_in_bytes": 39798664
+    },
+    "fr-de": {
+        "description": "This shared task will examine automatic evaluation metrics for machine translation. We will \nprovide you with MT system outputs along with source text and the human reference translations. \nWe are looking for automatic metric scores for translations at the system-level, and segment-level. \nWe will calculate the system-level, and segment-level correlations of your scores with human judgements.\n\nWe invite submissions of reference-free metrics in addition to reference-based metrics.\n",
+        "citation": "@inproceedings{freitag-etal-2021-results,\n    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},\n    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}\",\n    booktitle = {Proceedings of the Sixth Conference on Machine Translation},\n    month = {nov},\n    year = {2021},\n    address = {Online},\n    publisher = {Association for Computational Linguistics},\n    url = {https://aclanthology.org/2021.wmt-1.73},\n    pages = {733--774}\n}\n",
+        "homepage": "https://www.statmt.org/wmt21/metrics-task.html",
+        "license": "Unknown",
+        "features": {
+            "translation": {
+                "languages": [
+                    "fr",
+                    "de"
+                ],
+                "id": null,
+                "_type": "Translation"
+            },
+            "mt_system": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "mqm": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-raw": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-z": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "pair": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "dataset": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "sent_id": {
+                "dtype": "int32",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_name": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_ref": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-A": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-B": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-C": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-D": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            }
+        },
+        "post_processed": null,
+        "supervised_keys": null,
+        "task_templates": null,
+        "builder_name": "wmt21_metrics_task",
+        "config_name": "fr-de",
+        "version": {
+            "version_str": "1.1.0",
+            "description": null,
+            "major": 1,
+            "minor": 1,
+            "patch": 0
+        },
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_bytes": 6058153,
+                "num_examples": 10260,
+                "dataset_name": "wmt21_metrics_task"
+            }
+        },
+        "download_checksums": {
+            "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/fr-de/train.csv": {
+                "num_bytes": 5811674,
+                "checksum": "45042e130d477362d5352a6be0708afd3bdc8641e20f8615f8a523c4a7200e2a"
+            }
+        },
+        "download_size": 5811674,
+        "post_processing_size": null,
+        "dataset_size": 6058153,
+        "size_in_bytes": 11869827
+    },
+    "ha-en": {
+        "description": "This shared task will examine automatic evaluation metrics for machine translation. We will \nprovide you with MT system outputs along with source text and the human reference translations. \nWe are looking for automatic metric scores for translations at the system-level, and segment-level. \nWe will calculate the system-level, and segment-level correlations of your scores with human judgements.\n\nWe invite submissions of reference-free metrics in addition to reference-based metrics.\n",
+        "citation": "@inproceedings{freitag-etal-2021-results,\n    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},\n    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}\",\n    booktitle = {Proceedings of the Sixth Conference on Machine Translation},\n    month = {nov},\n    year = {2021},\n    address = {Online},\n    publisher = {Association for Computational Linguistics},\n    url = {https://aclanthology.org/2021.wmt-1.73},\n    pages = {733--774}\n}\n",
+        "homepage": "https://www.statmt.org/wmt21/metrics-task.html",
+        "license": "Unknown",
+        "features": {
+            "translation": {
+                "languages": [
+                    "ha",
+                    "en"
+                ],
+                "id": null,
+                "_type": "Translation"
+            },
+            "mt_system": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "mqm": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-raw": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-z": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "pair": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "dataset": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "sent_id": {
+                "dtype": "int32",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_name": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_ref": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-A": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-B": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-C": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-D": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            }
+        },
+        "post_processed": null,
+        "supervised_keys": null,
+        "task_templates": null,
+        "builder_name": "wmt21_metrics_task",
+        "config_name": "ha-en",
+        "version": {
+            "version_str": "1.1.0",
+            "description": null,
+            "major": 1,
+            "minor": 1,
+            "patch": 0
+        },
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_bytes": 7778578,
+                "num_examples": 13958,
+                "dataset_name": "wmt21_metrics_task"
+            }
+        },
+        "download_checksums": {
+            "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/ha-en/train.csv": {
+                "num_bytes": 7465098,
+                "checksum": "7975c117792f628074c6d6296e8e4a50ba86fdbc9c3e3097d849cd777ee40883"
+            }
+        },
+        "download_size": 7465098,
+        "post_processing_size": null,
+        "dataset_size": 7778578,
+        "size_in_bytes": 15243676
+    },
+    "is-en": {
+        "description": "This shared task will examine automatic evaluation metrics for machine translation. We will \nprovide you with MT system outputs along with source text and the human reference translations. \nWe are looking for automatic metric scores for translations at the system-level, and segment-level. \nWe will calculate the system-level, and segment-level correlations of your scores with human judgements.\n\nWe invite submissions of reference-free metrics in addition to reference-based metrics.\n",
+        "citation": "@inproceedings{freitag-etal-2021-results,\n    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},\n    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}\",\n    booktitle = {Proceedings of the Sixth Conference on Machine Translation},\n    month = {nov},\n    year = {2021},\n    address = {Online},\n    publisher = {Association for Computational Linguistics},\n    url = {https://aclanthology.org/2021.wmt-1.73},\n    pages = {733--774}\n}\n",
+        "homepage": "https://www.statmt.org/wmt21/metrics-task.html",
+        "license": "Unknown",
+        "features": {
+            "translation": {
+                "languages": [
+                    "is",
+                    "en"
+                ],
+                "id": null,
+                "_type": "Translation"
+            },
+            "mt_system": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "mqm": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-raw": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-z": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "pair": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "dataset": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "sent_id": {
+                "dtype": "int32",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_name": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_ref": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-A": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-B": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-C": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-D": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            }
+        },
+        "post_processed": null,
+        "supervised_keys": null,
+        "task_templates": null,
+        "builder_name": "wmt21_metrics_task",
+        "config_name": "is-en",
+        "version": {
+            "version_str": "1.1.0",
+            "description": null,
+            "major": 1,
+            "minor": 1,
+            "patch": 0
+        },
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_bytes": 4510788,
+                "num_examples": 10000,
+                "dataset_name": "wmt21_metrics_task"
+            }
+        },
+        "download_checksums": {
+            "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/is-en/train.csv": {
+                "num_bytes": 4283179,
+                "checksum": "d576d3ac03d91275d9c877d1dab489ac27b35b84a8a74dec7bad4a0684979a8a"
+            }
+        },
+        "download_size": 4283179,
+        "post_processing_size": null,
+        "dataset_size": 4510788,
+        "size_in_bytes": 8793967
+    },
+    "ja-en": {
+        "description": "This shared task will examine automatic evaluation metrics for machine translation. We will \nprovide you with MT system outputs along with source text and the human reference translations. \nWe are looking for automatic metric scores for translations at the system-level, and segment-level. \nWe will calculate the system-level, and segment-level correlations of your scores with human judgements.\n\nWe invite submissions of reference-free metrics in addition to reference-based metrics.\n",
+        "citation": "@inproceedings{freitag-etal-2021-results,\n    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},\n    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}\",\n    booktitle = {Proceedings of the Sixth Conference on Machine Translation},\n    month = {nov},\n    year = {2021},\n    address = {Online},\n    publisher = {Association for Computational Linguistics},\n    url = {https://aclanthology.org/2021.wmt-1.73},\n    pages = {733--774}\n}\n",
+        "homepage": "https://www.statmt.org/wmt21/metrics-task.html",
+        "license": "Unknown",
+        "features": {
+            "translation": {
+                "languages": [
+                    "ja",
+                    "en"
+                ],
+                "id": null,
+                "_type": "Translation"
+            },
+            "mt_system": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "mqm": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-raw": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-z": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "pair": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "dataset": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "sent_id": {
+                "dtype": "int32",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_name": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_ref": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-A": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-B": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-C": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-D": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            }
+        },
+        "post_processed": null,
+        "supervised_keys": null,
+        "task_templates": null,
+        "builder_name": "wmt21_metrics_task",
+        "config_name": "ja-en",
+        "version": {
+            "version_str": "1.1.0",
+            "description": null,
+            "major": 1,
+            "minor": 1,
+            "patch": 0
+        },
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_bytes": 8479226,
+                "num_examples": 16080,
+                "dataset_name": "wmt21_metrics_task"
+            }
+        },
+        "download_checksums": {
+            "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/ja-en/train.csv": {
+                "num_bytes": 8124564,
+                "checksum": "727fe3f95e735306bb967bfcebb974382264e0fe64d2ab82bd9baf77a21998aa"
+            }
+        },
+        "download_size": 8124564,
+        "post_processing_size": null,
+        "dataset_size": 8479226,
+        "size_in_bytes": 16603790
+    },
+    "ru-en": {
+        "description": "This shared task will examine automatic evaluation metrics for machine translation. We will \nprovide you with MT system outputs along with source text and the human reference translations. \nWe are looking for automatic metric scores for translations at the system-level, and segment-level. \nWe will calculate the system-level, and segment-level correlations of your scores with human judgements.\n\nWe invite submissions of reference-free metrics in addition to reference-based metrics.\n",
+        "citation": "@inproceedings{freitag-etal-2021-results,\n    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},\n    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}\",\n    booktitle = {Proceedings of the Sixth Conference on Machine Translation},\n    month = {nov},\n    year = {2021},\n    address = {Online},\n    publisher = {Association for Computational Linguistics},\n    url = {https://aclanthology.org/2021.wmt-1.73},\n    pages = {733--774}\n}\n",
+        "homepage": "https://www.statmt.org/wmt21/metrics-task.html",
+        "license": "Unknown",
+        "features": {
+            "translation": {
+                "languages": [
+                    "ru",
+                    "en"
+                ],
+                "id": null,
+                "_type": "Translation"
+            },
+            "mt_system": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "mqm": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-raw": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-z": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "pair": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "dataset": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "sent_id": {
+                "dtype": "int32",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_name": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_ref": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-A": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-B": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-C": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-D": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            }
+        },
+        "post_processed": null,
+        "supervised_keys": null,
+        "task_templates": null,
+        "builder_name": "wmt21_metrics_task",
+        "config_name": "ru-en",
+        "version": {
+            "version_str": "1.1.0",
+            "description": null,
+            "major": 1,
+            "minor": 1,
+            "patch": 0
+        },
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_bytes": 6992095,
+                "num_examples": 11000,
+                "dataset_name": "wmt21_metrics_task"
+            }
+        },
+        "download_checksums": {
+            "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/ru-en/train.csv": {
+                "num_bytes": 6757022,
+                "checksum": "35a8c93ba72de8b93febba1933a1f07289053e6ab9a377d934cbd5b428008c5c"
+            }
+        },
+        "download_size": 6757022,
+        "post_processing_size": null,
+        "dataset_size": 6992095,
+        "size_in_bytes": 13749117
+    },
+    "zh-en": {
+        "description": "This shared task will examine automatic evaluation metrics for machine translation. We will \nprovide you with MT system outputs along with source text and the human reference translations. \nWe are looking for automatic metric scores for translations at the system-level, and segment-level. \nWe will calculate the system-level, and segment-level correlations of your scores with human judgements.\n\nWe invite submissions of reference-free metrics in addition to reference-based metrics.\n",
+        "citation": "@inproceedings{freitag-etal-2021-results,\n    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},\n    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}\",\n    booktitle = {Proceedings of the Sixth Conference on Machine Translation},\n    month = {nov},\n    year = {2021},\n    address = {Online},\n    publisher = {Association for Computational Linguistics},\n    url = {https://aclanthology.org/2021.wmt-1.73},\n    pages = {733--774}\n}\n",
+        "homepage": "https://www.statmt.org/wmt21/metrics-task.html",
+        "license": "Unknown",
+        "features": {
+            "translation": {
+                "languages": [
+                    "zh",
+                    "en"
+                ],
+                "id": null,
+                "_type": "Translation"
+            },
+            "mt_system": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "mqm": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-raw": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "wmt-z": {
+                "dtype": "float32",
+                "id": null,
+                "_type": "Value"
+            },
+            "pair": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "dataset": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "sent_id": {
+                "dtype": "int32",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_name": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "doc_ref": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-A": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-B": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-C": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            },
+            "ref-D": {
+                "dtype": "string",
+                "id": null,
+                "_type": "Value"
+            }
+        },
+        "post_processed": null,
+        "supervised_keys": null,
+        "task_templates": null,
+        "builder_name": "wmt21_metrics_task",
+        "config_name": "zh-en",
+        "version": {
+            "version_str": "1.1.0",
+            "description": null,
+            "major": 1,
+            "minor": 1,
+            "patch": 0
+        },
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_bytes": 57822701,
+                "num_examples": 71085,
+                "dataset_name": "wmt21_metrics_task"
+            }
+        },
+        "download_checksums": {
+            "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/zh-en/train.csv": {
+                "num_bytes": 55812435,
+                "checksum": "07ece5ccc8a3083b2f34ceae2bf8f26b87a85e6fddc0989f11a5d575a8f0332d"
+            }
+        },
+        "download_size": 55812435,
+        "post_processing_size": null,
+        "dataset_size": 57822701,
+        "size_in_bytes": 113635136
+    }
+}

wmt21_metrics_task.py

@@ -0,0 +1,138 @@
+# coding=utf-8
+# Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""WMT21 Metrics Shared Task: Segment-Level Data"""
+
+import datasets
+import csv
+
+_CITATION = """\
+@inproceedings{freitag-etal-2021-results,
+    title = {Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain},
+    author = {Freitag, Markus and Rei, Ricardo  and Mathur, Nitika  and Lo, Chi-kiu  and Stewart, Craig  and Foster, George  and Lavie, Alon and Bojar, Ondrej}",
+    booktitle = {Proceedings of the Sixth Conference on Machine Translation},
+    month = {nov},
+    year = {2021},
+    address = {Online},
+    publisher = {Association for Computational Linguistics},
+    url = {https://aclanthology.org/2021.wmt-1.73},
+    pages = {733--774}
+}
+"""
+
+_DESCRIPTION = """\
+This shared task will examine automatic evaluation metrics for machine translation. We will 
+provide you with MT system outputs along with source text and the human reference translations. 
+We are looking for automatic metric scores for translations at the system-level, and segment-level. 
+We will calculate the system-level, and segment-level correlations of your scores with human judgements.
+
+We invite submissions of reference-free metrics in addition to reference-based metrics.
+"""
+
+_HOMEPAGE = "https://www.statmt.org/wmt21/metrics-task.html"
+
+_LICENSE = "Unknown"
+
+
+_LANGUAGE_PAIRS = [('bn', 'hi'), ('hi', 'bn'), ('xh', 'zu'), ('zu', 'xh'), ('cs', 'en'), ('de', 'en'), ('de', 'fr'), ('en', 'cs'), ('en', 'de'), ('en', 'ha'), 
+('en', 'is'), ('en', 'ja'), ('en', 'ru'), ('en', 'zh'), ('fr', 'de'), ('ha', 'en'), ('is', 'en'), ('ja', 'en'), ('ru', 'en'), ('zh', 'en')]
+
+_URL_BASE = "https://huggingface.co/datasets/muibk/wmt21_metrics_task/resolve/main/"
+_URLs = {f"{src_lg}-{trg_lg}": f"{_URL_BASE}{src_lg}-{trg_lg}/train.csv" for src_lg, trg_lg in _LANGUAGE_PAIRS}
+
+
+class WmtMetricsTaskConfig(datasets.BuilderConfig):
+    """BuilderConfig for WMT Metrics Shared Task."""
+    def __init__(self, src_lg, tgt_lg, **kwargs):
+        super(WmtMetricsTaskConfig, self).__init__(**kwargs)
+        self.src_lg = src_lg
+        self.tgt_lg = tgt_lg
+
+
+class Wmt21MetricsTask(datasets.GeneratorBasedBuilder):
+    """WMT Metrics Shared Task."""
+
+    BUILDER_CONFIGS = [
+        WmtMetricsTaskConfig(
+            name=f"{src_lg}-{tgt_lg}",
+            version=datasets.Version("1.1.0"),
+            description=f"WMT 2021 Metrics Task: {src_lg} - {tgt_lg}",
+            src_lg=src_lg,
+            tgt_lg=tgt_lg,
+        )
+        for (src_lg, tgt_lg) in _LANGUAGE_PAIRS
+    ]
+    BUILDER_CONFIG_CLASS = WmtMetricsTaskConfig
+
+    def _info(self):
+        # define feature types
+        features = datasets.Features(
+            {
+                #'source' : datasets.Value("string"),
+                #'system_output':datasets.Value("string"),
+                'translation': datasets.Translation(languages=(self.config.src_lg, self.config.tgt_lg)),
+                'mt_system':datasets.Value("string"),
+                'mqm':datasets.Value("float32"),
+                'wmt-raw':datasets.Value("float32"),
+                'wmt-z':datasets.Value("float32"),
+                'pair':datasets.Value("string"),
+                'dataset':datasets.Value("string"),
+                'sent_id':datasets.Value("int32"),
+                'doc_name':datasets.Value("string"),
+                'doc_ref':datasets.Value("string"),
+                'ref-A':datasets.Value("string"),
+                'ref-B':datasets.Value("string"),
+                'ref-C':datasets.Value("string"),
+                'ref-D':datasets.Value("string")
+            }
+        )
+
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=features,
+            supervised_keys=None,
+            homepage=_HOMEPAGE,
+            license=_LICENSE,
+            citation=_CITATION,
+        )
+
+    def _split_generators(self, dl_manager):
+        """Returns SplitGenerators."""
+        pair = f"{self.config.src_lg}-{self.config.tgt_lg}" # string identifier for language pair
+        url = _URLs[pair] # url for download of pair-specific train.csv
+        data_file = dl_manager.download_and_extract(url) # extract downloaded data and store path in data_file
+
+        return [
+            datasets.SplitGenerator(
+                name=datasets.Split.TRAIN,
+                gen_kwargs={
+                    "filepath": data_file,
+                    "source_lg": self.config.src_lg,
+                    "target_lg": self.config.tgt_lg,
+                }
+            )
+        ]
+    
+    def _generate_examples(self, filepath, source_lg, target_lg):
+        with open(filepath, encoding="utf-8") as f:
+            reader = csv.DictReader(f, delimiter=";") # read each line into dict
+            for id_, row in enumerate(reader):
+                row["translation"] = {source_lg : row["source"], target_lg: row["system_output"]} # create translation json
+                for key in ["source", "system_output"]: # remove obsolete columns
+                    row.pop(key)
+                row = {k: None if not v else v for k, v in row.items()} # replace empty values
+                yield id_, row
+
+# to test the script, go to the root folder of the repo (wmt21_metrics_task) and run:
+# datasets-cli test wmt21_metrics_task --save_infos --all_configs