Sebastian Gehrmann commited on
Commit
3fb328f
1 Parent(s): 2c1e531
Files changed (2) hide show
  1. conversational_weather.py +9 -4
  2. dataset_infos.json +1 -1
conversational_weather.py CHANGED
@@ -75,9 +75,9 @@ _URLs = {
75
  }
76
 
77
  class ConversationalWeather(datasets.GeneratorBasedBuilder):
78
- """The Conversational Weather dataset is designed for generation of responses to weather queries
79
- based on a structured input data. The input allows specifying data attributes such as dates, times,
80
- locations, weather conditions, and errors, and also offers control over structure of response through
81
  discourse relations such as join, contrast, and justification."""
82
 
83
  VERSION = datasets.Version("1.1.0")
@@ -98,6 +98,9 @@ class ConversationalWeather(datasets.GeneratorBasedBuilder):
98
  "user_query": datasets.Value("string"),
99
  "tree_str_mr": datasets.Value("string"),
100
  "response": datasets.Value("string"),
 
 
 
101
  }
102
  )
103
 
@@ -105,7 +108,7 @@ class ConversationalWeather(datasets.GeneratorBasedBuilder):
105
  # This is the description that will appear on the datasets page.
106
  description=_DESCRIPTION,
107
  # This defines the different columns of the dataset and their types
108
- features=features,
109
  # If there's a common (input, target) tuple from the features,
110
  # specify them here. They'll be used if as_supervised=True in
111
  # builder.as_dataset.
@@ -206,4 +209,6 @@ class ConversationalWeather(datasets.GeneratorBasedBuilder):
206
  "user_query": row[1],
207
  "tree_str_mr": row[2],
208
  "response": row[3],
 
 
209
  }
75
  }
76
 
77
  class ConversationalWeather(datasets.GeneratorBasedBuilder):
78
+ """The Conversational Weather dataset is designed for generation of responses to weather queries
79
+ based on a structured input data. The input allows specifying data attributes such as dates, times,
80
+ locations, weather conditions, and errors, and also offers control over structure of response through
81
  discourse relations such as join, contrast, and justification."""
82
 
83
  VERSION = datasets.Version("1.1.0")
98
  "user_query": datasets.Value("string"),
99
  "tree_str_mr": datasets.Value("string"),
100
  "response": datasets.Value("string"),
101
+ "target": datasets.Value("string"),
102
+ "references": [datasets.Value("string")],
103
+
104
  }
105
  )
106
 
108
  # This is the description that will appear on the datasets page.
109
  description=_DESCRIPTION,
110
  # This defines the different columns of the dataset and their types
111
+ features=features,
112
  # If there's a common (input, target) tuple from the features,
113
  # specify them here. They'll be used if as_supervised=True in
114
  # builder.as_dataset.
209
  "user_query": row[1],
210
  "tree_str_mr": row[2],
211
  "response": row[3],
212
+ "target": row[3],
213
+ "references": [row[3]],
214
  }
dataset_infos.json CHANGED
@@ -1 +1 @@
1
- {"default": {"description": "The Conversational Weather dataset is designed for generation of responses to weather queries based on a structured input data. The input allows specifying data attributes such as dates, times, locations, weather conditions, and errors, and also offers control over structure of response through discourse relations such as join, contrast, and justification.\n", "citation": "@inproceedings{balakrishnan-etal-2019-constrained,\n title = \"Constrained Decoding for Neural {NLG} from Compositional Representations in Task-Oriented Dialogue\",\n author = \"Balakrishnan, Anusha and\n Rao, Jinfeng and\n Upasani, Kartikeya and\n White, Michael and\n Subba, Rajen\",\n booktitle = \"Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2019\",\n address = \"Florence, Italy\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/P19-1080\",\n doi = \"10.18653/v1/P19-1080\",\n pages = \"831--844\"\n}\n", "homepage": "https://github.com/facebookresearch/TreeNLG", "license": "CC-BY-NC-4.0", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "data_id": {"dtype": "string", "id": null, "_type": "Value"}, "user_query": {"dtype": "string", "id": null, "_type": "Value"}, "tree_str_mr": {"dtype": "string", "id": null, "_type": "Value"}, "response": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "conversational_weather", "config_name": "default", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 19860574, "num_examples": 25390, "dataset_name": "conversational_weather"}, "test": {"name": "test", "num_bytes": 2467176, "num_examples": 3121, "dataset_name": "conversational_weather"}, "validation": {"name": "validation", "num_bytes": 2390007, "num_examples": 3078, "dataset_name": "conversational_weather"}}, "download_checksums": {"https://raw.githubusercontent.com/facebookresearch/TreeNLG/master/data/weather/train.tsv": {"num_bytes": 18982974, "checksum": "5d3d27566eb2ac6c8d5295a78072209a78670c8d8ab8e20443d678e9bd2501db"}, "https://raw.githubusercontent.com/facebookresearch/TreeNLG/master/data/weather/val.tsv": {"num_bytes": 2292601, "checksum": "7e50b746de247c5ac9bae1f53381e8230856edc43ec9460601fc2577fbe77286"}, "https://raw.githubusercontent.com/facebookresearch/TreeNLG/master/data/weather/test.tsv": {"num_bytes": 2365273, "checksum": "f2a36007698e510e3308d7fa4a23c2b492321047d849ed5e04409cb773b3b1ca"}}, "download_size": 23640848, "post_processing_size": null, "dataset_size": 24717757, "size_in_bytes": 48358605}, "challenge": {"description": "The Conversational Weather dataset is designed for generation of responses to weather queries based on a structured input data. The input allows specifying data attributes such as dates, times, locations, weather conditions, and errors, and also offers control over structure of response through discourse relations such as join, contrast, and justification.\n", "citation": "@inproceedings{balakrishnan-etal-2019-constrained,\n title = \"Constrained Decoding for Neural {NLG} from Compositional Representations in Task-Oriented Dialogue\",\n author = \"Balakrishnan, Anusha and\n Rao, Jinfeng and\n Upasani, Kartikeya and\n White, Michael and\n Subba, Rajen\",\n booktitle = \"Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2019\",\n address = \"Florence, Italy\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/P19-1080\",\n doi = \"10.18653/v1/P19-1080\",\n pages = \"831--844\"\n}\n", "homepage": "https://github.com/facebookresearch/TreeNLG", "license": "CC-BY-NC-4.0", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "data_id": {"dtype": "string", "id": null, "_type": "Value"}, "user_query": {"dtype": "string", "id": null, "_type": "Value"}, "tree_str_mr": {"dtype": "string", "id": null, "_type": "Value"}, "response": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "conversational_weather", "config_name": "challenge", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"disc_test": {"name": "disc_test", "num_bytes": 250926, "num_examples": 263, "dataset_name": "conversational_weather"}, "disc_test_freq_0": {"name": "disc_test_freq_0", "num_bytes": 1127210, "num_examples": 1685, "dataset_name": "conversational_weather"}, "disc_test_freq_1": {"name": "disc_test_freq_1", "num_bytes": 1110690, "num_examples": 1208, "dataset_name": "conversational_weather"}, "disc_test_freq_2": {"name": "disc_test_freq_2", "num_bytes": 258120, "num_examples": 220, "dataset_name": "conversational_weather"}, "disc_test_freq_3": {"name": "disc_test_freq_3", "num_bytes": 13446, "num_examples": 8, "dataset_name": "conversational_weather"}, "dial_test_freq_1": {"name": "dial_test_freq_1", "num_bytes": 324181, "num_examples": 608, "dataset_name": "conversational_weather"}, "dial_test_freq_2": {"name": "dial_test_freq_2", "num_bytes": 781478, "num_examples": 1138, "dataset_name": "conversational_weather"}, "dial_test_freq_3": {"name": "dial_test_freq_3", "num_bytes": 840694, "num_examples": 923, "dataset_name": "conversational_weather"}, "dial_test_freq_4": {"name": "dial_test_freq_4", "num_bytes": 468256, "num_examples": 396, "dataset_name": "conversational_weather"}, "dial_test_freq_5": {"name": "dial_test_freq_5", "num_bytes": 93962, "num_examples": 56, "dataset_name": "conversational_weather"}}, "download_checksums": {"./data/challenge_sets/disc_test.tsv": {"num_bytes": 240759, "checksum": "142608abfb4cae43b760cd7cd5788ce9c599f8c43671ea011a0d16c1dca40b58"}, "./data/challenge_sets/dial_test_freq_1.tsv": {"num_bytes": 296303, "checksum": "3057b7c34ce66ebaadcb22ff5834158b80af5a1a5f3ce92c5b2d237c31d84feb"}, "./data/challenge_sets/dial_test_freq_2.tsv": {"num_bytes": 729082, "checksum": "af3b7471015b87af6fb118ff5d97b12c07070cdd6256498105a0f7be060fc4e5"}, "./data/challenge_sets/dial_test_freq_3.tsv": {"num_bytes": 798326, "checksum": "f2999eca340417ecbecead1719df3a9ae8fe0d8afc6624212fa2dc465d18357a"}, "./data/challenge_sets/dial_test_freq_4.tsv": {"num_bytes": 450130, "checksum": "678041916baa5f9cd5d415638689b85c0e8e3fdc9a9fe66273a1342cfd4ba0d4"}, "./data/challenge_sets/dial_test_freq_5.tsv": {"num_bytes": 91432, "checksum": "10bf51037283c8322d3aaea64e6b1ba5e3ae6c45ed621e327eabdf791ce6c1b2"}, "./data/challenge_sets/disc_test_freq_0.tsv": {"num_bytes": 1049105, "checksum": "e70ca7cacfd43b424fbd375ac8fc3e7bde24538caccdcbd19c056b8825437422"}, "./data/challenge_sets/disc_test_freq_1.tsv": {"num_bytes": 1055004, "checksum": "0f9ab3179e19ee7d48e9d4ffd81ea0d37f08ff17213d3e6c5c02a3a3d2cc4921"}, "./data/challenge_sets/disc_test_freq_2.tsv": {"num_bytes": 248090, "checksum": "3c912b019fa5d1f39c7d83e404f79250e6b54ef127aff201cab3faeac186995a"}, "./data/challenge_sets/disc_test_freq_3.tsv": {"num_bytes": 13074, "checksum": "e99097dd34607d0f6c03a17cb35125f37f15571d5dfe57d4218f4db56d45c40b"}}, "download_size": 4971305, "post_processing_size": null, "dataset_size": 5268963, "size_in_bytes": 10240268}}
1
+ {"default": {"description": "The Conversational Weather dataset is designed for generation of responses to weather queries based on a structured input data. The input allows specifying data attributes such as dates, times, locations, weather conditions, and errors, and also offers control over structure of response through discourse relations such as join, contrast, and justification.\n", "citation": "@inproceedings{balakrishnan-etal-2019-constrained,\n title = \"Constrained Decoding for Neural {NLG} from Compositional Representations in Task-Oriented Dialogue\",\n author = \"Balakrishnan, Anusha and\n Rao, Jinfeng and\n Upasani, Kartikeya and\n White, Michael and\n Subba, Rajen\",\n booktitle = \"Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2019\",\n address = \"Florence, Italy\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/P19-1080\",\n doi = \"10.18653/v1/P19-1080\",\n pages = \"831--844\"\n}\n", "homepage": "https://github.com/facebookresearch/TreeNLG", "license": "CC-BY-NC-4.0", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "data_id": {"dtype": "string", "id": null, "_type": "Value"}, "user_query": {"dtype": "string", "id": null, "_type": "Value"}, "tree_str_mr": {"dtype": "string", "id": null, "_type": "Value"}, "response": {"dtype": "string", "id": null, "_type": "Value"}, "target": {"dtype": "string", "id": null, "_type": "Value"}, "references": [{"dtype": "string", "id": null, "_type": "Value"}]}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "conversational_weather", "config_name": "default", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 36514596, "num_examples": 25390, "dataset_name": "conversational_weather"}, "test": {"name": "test", "num_bytes": 4544810, "num_examples": 3121, "dataset_name": "conversational_weather"}, "validation": {"name": "validation", "num_bytes": 4399453, "num_examples": 3078, "dataset_name": "conversational_weather"}}, "download_checksums": {"https://raw.githubusercontent.com/facebookresearch/TreeNLG/master/data/weather/train.tsv": {"num_bytes": 18982974, "checksum": "5d3d27566eb2ac6c8d5295a78072209a78670c8d8ab8e20443d678e9bd2501db"}, "https://raw.githubusercontent.com/facebookresearch/TreeNLG/master/data/weather/val.tsv": {"num_bytes": 2292601, "checksum": "7e50b746de247c5ac9bae1f53381e8230856edc43ec9460601fc2577fbe77286"}, "https://raw.githubusercontent.com/facebookresearch/TreeNLG/master/data/weather/test.tsv": {"num_bytes": 2365273, "checksum": "f2a36007698e510e3308d7fa4a23c2b492321047d849ed5e04409cb773b3b1ca"}}, "download_size": 23640848, "post_processing_size": null, "dataset_size": 45458859, "size_in_bytes": 69099707}, "challenge": {"description": "The Conversational Weather dataset is designed for generation of responses to weather queries based on a structured input data. The input allows specifying data attributes such as dates, times, locations, weather conditions, and errors, and also offers control over structure of response through discourse relations such as join, contrast, and justification.\n", "citation": "@inproceedings{balakrishnan-etal-2019-constrained,\n title = \"Constrained Decoding for Neural {NLG} from Compositional Representations in Task-Oriented Dialogue\",\n author = \"Balakrishnan, Anusha and\n Rao, Jinfeng and\n Upasani, Kartikeya and\n White, Michael and\n Subba, Rajen\",\n booktitle = \"Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2019\",\n address = \"Florence, Italy\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/P19-1080\",\n doi = \"10.18653/v1/P19-1080\",\n pages = \"831--844\"\n}\n", "homepage": "https://github.com/facebookresearch/TreeNLG", "license": "CC-BY-NC-4.0", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "data_id": {"dtype": "string", "id": null, "_type": "Value"}, "user_query": {"dtype": "string", "id": null, "_type": "Value"}, "tree_str_mr": {"dtype": "string", "id": null, "_type": "Value"}, "response": {"dtype": "string", "id": null, "_type": "Value"}, "target": {"dtype": "string", "id": null, "_type": "Value"}, "references": [{"dtype": "string", "id": null, "_type": "Value"}]}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "conversational_weather", "config_name": "challenge", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"disc_test": {"name": "disc_test", "num_bytes": 460468, "num_examples": 263, "dataset_name": "conversational_weather"}, "disc_test_freq_0": {"name": "disc_test_freq_0", "num_bytes": 2071610, "num_examples": 1685, "dataset_name": "conversational_weather"}, "disc_test_freq_1": {"name": "disc_test_freq_1", "num_bytes": 2025862, "num_examples": 1208, "dataset_name": "conversational_weather"}, "disc_test_freq_2": {"name": "disc_test_freq_2", "num_bytes": 464858, "num_examples": 220, "dataset_name": "conversational_weather"}, "disc_test_freq_3": {"name": "disc_test_freq_3", "num_bytes": 24806, "num_examples": 8, "dataset_name": "conversational_weather"}, "dial_test_freq_1": {"name": "dial_test_freq_1", "num_bytes": 619371, "num_examples": 608, "dataset_name": "conversational_weather"}, "dial_test_freq_2": {"name": "dial_test_freq_2", "num_bytes": 1435082, "num_examples": 1138, "dataset_name": "conversational_weather"}, "dial_test_freq_3": {"name": "dial_test_freq_3", "num_bytes": 1517612, "num_examples": 923, "dataset_name": "conversational_weather"}, "dial_test_freq_4": {"name": "dial_test_freq_4", "num_bytes": 844282, "num_examples": 396, "dataset_name": "conversational_weather"}, "dial_test_freq_5": {"name": "dial_test_freq_5", "num_bytes": 169906, "num_examples": 56, "dataset_name": "conversational_weather"}}, "download_checksums": {"./data/challenge_sets/disc_test.tsv": {"num_bytes": 240759, "checksum": "142608abfb4cae43b760cd7cd5788ce9c599f8c43671ea011a0d16c1dca40b58"}, "./data/challenge_sets/dial_test_freq_1.tsv": {"num_bytes": 296303, "checksum": "3057b7c34ce66ebaadcb22ff5834158b80af5a1a5f3ce92c5b2d237c31d84feb"}, "./data/challenge_sets/dial_test_freq_2.tsv": {"num_bytes": 729082, "checksum": "af3b7471015b87af6fb118ff5d97b12c07070cdd6256498105a0f7be060fc4e5"}, "./data/challenge_sets/dial_test_freq_3.tsv": {"num_bytes": 798326, "checksum": "f2999eca340417ecbecead1719df3a9ae8fe0d8afc6624212fa2dc465d18357a"}, "./data/challenge_sets/dial_test_freq_4.tsv": {"num_bytes": 450130, "checksum": "678041916baa5f9cd5d415638689b85c0e8e3fdc9a9fe66273a1342cfd4ba0d4"}, "./data/challenge_sets/dial_test_freq_5.tsv": {"num_bytes": 91432, "checksum": "10bf51037283c8322d3aaea64e6b1ba5e3ae6c45ed621e327eabdf791ce6c1b2"}, "./data/challenge_sets/disc_test_freq_0.tsv": {"num_bytes": 1049105, "checksum": "e70ca7cacfd43b424fbd375ac8fc3e7bde24538caccdcbd19c056b8825437422"}, "./data/challenge_sets/disc_test_freq_1.tsv": {"num_bytes": 1055004, "checksum": "0f9ab3179e19ee7d48e9d4ffd81ea0d37f08ff17213d3e6c5c02a3a3d2cc4921"}, "./data/challenge_sets/disc_test_freq_2.tsv": {"num_bytes": 248090, "checksum": "3c912b019fa5d1f39c7d83e404f79250e6b54ef127aff201cab3faeac186995a"}, "./data/challenge_sets/disc_test_freq_3.tsv": {"num_bytes": 13074, "checksum": "e99097dd34607d0f6c03a17cb35125f37f15571d5dfe57d4218f4db56d45c40b"}}, "download_size": 4971305, "post_processing_size": null, "dataset_size": 9633857, "size_in_bytes": 14605162}}