system HF staff commited on
Commit
99c0835
1 Parent(s): 4aa2c15

Update files from the datasets library (from 1.16.0)

Browse files

Release notes: https://github.com/huggingface/datasets/releases/tag/1.16.0

Files changed (4) hide show
  1. README.md +1 -0
  2. dataset_infos.json +1 -1
  3. dummy/trex/1.1.0/dummy_data.zip +2 -2
  4. lama.py +112 -106
README.md CHANGED
@@ -1,4 +1,5 @@
1
  ---
 
2
  annotations_creators:
3
  - crowdsourced
4
  - expert-generated
1
  ---
2
+ pretty_name: "LAMA: LAnguage Model Analysis"
3
  annotations_creators:
4
  - crowdsourced
5
  - expert-generated
dataset_infos.json CHANGED
@@ -1 +1 @@
1
- {"trex": {"description": "LAMA is a dataset used to probe and analyze the factual and commonsense knowledge contained in pretrained language models. See https://github.com/facebookresearch/LAMA.\n", "citation": "@inproceedings{petroni2019language,\n title={Language Models as Knowledge Bases?},\n author={F. Petroni, T. Rockt{\"{a}}schel, A. H. Miller, P. Lewis, A. Bakhtin, Y. Wu and S. Riedel},\n booktitle={In: Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2019},\n year={2019}\n}\n@inproceedings{petroni2020how,\n title={How Context Affects Language Models' Factual Predictions},\n author={Fabio Petroni and Patrick Lewis and Aleksandra Piktus and Tim Rockt{\"a}schel and Yuxiang Wu and Alexander H. Miller and Sebastian Riedel},\n booktitle={Automated Knowledge Base Construction},\n year={2020},\n url={https://openreview.net/forum?id=025X0zPfn}\n}\n", "homepage": "https://github.com/facebookresearch/LAMA", "license": "The Creative Commons Attribution-Noncommercial 4.0 International License. see https://github.com/facebookresearch/LAMA/blob/master/LICENSE", "features": {"uuid": {"dtype": "string", "id": null, "_type": "Value"}, "obj_uri": {"dtype": "string", "id": null, "_type": "Value"}, "obj_label": {"dtype": "string", "id": null, "_type": "Value"}, "sub_uri": {"dtype": "string", "id": null, "_type": "Value"}, "sub_label": {"dtype": "string", "id": null, "_type": "Value"}, "predicate_id": {"dtype": "string", "id": null, "_type": "Value"}, "sub_surface": {"dtype": "string", "id": null, "_type": "Value"}, "obj_surface": {"dtype": "string", "id": null, "_type": "Value"}, "masked_sentence": {"dtype": "string", "id": null, "_type": "Value"}, "template": {"dtype": "string", "id": null, "_type": "Value"}, "template_negated": {"dtype": "string", "id": null, "_type": "Value"}, "label": {"dtype": "string", "id": null, "_type": "Value"}, "description": {"dtype": "string", "id": null, "_type": "Value"}, "type": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "builder_name": "lama", "config_name": "trex", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 656913189, "num_examples": 1304391, "dataset_name": "lama"}}, "download_checksums": {"https://dl.fbaipublicfiles.com/LAMA/negated_data.tar.gz": {"num_bytes": 74639115, "checksum": "1a151058e6608e47983ea4c99c50bb69248c1c0763a04a3793b0a0b657aa0b61"}}, "download_size": 74639115, "post_processing_size": null, "dataset_size": 656913189, "size_in_bytes": 731552304}, "squad": {"description": "LAMA is a dataset used to probe and analyze the factual and commonsense knowledge contained in pretrained language models. See https://github.com/facebookresearch/LAMA.\n", "citation": "@inproceedings{petroni2019language,\n title={Language Models as Knowledge Bases?},\n author={F. Petroni, T. Rockt{\"{a}}schel, A. H. Miller, P. Lewis, A. Bakhtin, Y. Wu and S. Riedel},\n booktitle={In: Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2019},\n year={2019}\n}\n@inproceedings{petroni2020how,\n title={How Context Affects Language Models' Factual Predictions},\n author={Fabio Petroni and Patrick Lewis and Aleksandra Piktus and Tim Rockt{\"a}schel and Yuxiang Wu and Alexander H. Miller and Sebastian Riedel},\n booktitle={Automated Knowledge Base Construction},\n year={2020},\n url={https://openreview.net/forum?id=025X0zPfn}\n}\n", "homepage": "https://github.com/facebookresearch/LAMA", "license": "The Creative Commons Attribution-Noncommercial 4.0 International License. see https://github.com/facebookresearch/LAMA/blob/master/LICENSE", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "sub_label": {"dtype": "string", "id": null, "_type": "Value"}, "obj_label": {"dtype": "string", "id": null, "_type": "Value"}, "negated": {"dtype": "string", "id": null, "_type": "Value"}, "masked_sentence": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "builder_name": "lama", "config_name": "squad", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 57188, "num_examples": 305, "dataset_name": "lama"}}, "download_checksums": {"https://dl.fbaipublicfiles.com/LAMA/negated_data.tar.gz": {"num_bytes": 74639115, "checksum": "1a151058e6608e47983ea4c99c50bb69248c1c0763a04a3793b0a0b657aa0b61"}}, "download_size": 74639115, "post_processing_size": null, "dataset_size": 57188, "size_in_bytes": 74696303}, "google_re": {"description": "LAMA is a dataset used to probe and analyze the factual and commonsense knowledge contained in pretrained language models. See https://github.com/facebookresearch/LAMA.\n", "citation": "@inproceedings{petroni2019language,\n title={Language Models as Knowledge Bases?},\n author={F. Petroni, T. Rockt{\"{a}}schel, A. H. Miller, P. Lewis, A. Bakhtin, Y. Wu and S. Riedel},\n booktitle={In: Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2019},\n year={2019}\n}\n@inproceedings{petroni2020how,\n title={How Context Affects Language Models' Factual Predictions},\n author={Fabio Petroni and Patrick Lewis and Aleksandra Piktus and Tim Rockt{\"a}schel and Yuxiang Wu and Alexander H. Miller and Sebastian Riedel},\n booktitle={Automated Knowledge Base Construction},\n year={2020},\n url={https://openreview.net/forum?id=025X0zPfn}\n}\n", "homepage": "https://github.com/facebookresearch/LAMA", "license": "The Creative Commons Attribution-Noncommercial 4.0 International License. see https://github.com/facebookresearch/LAMA/blob/master/LICENSE", "features": {"pred": {"dtype": "string", "id": null, "_type": "Value"}, "sub": {"dtype": "string", "id": null, "_type": "Value"}, "obj": {"dtype": "string", "id": null, "_type": "Value"}, "evidences": {"dtype": "string", "id": null, "_type": "Value"}, "judgments": {"dtype": "string", "id": null, "_type": "Value"}, "sub_w": {"dtype": "string", "id": null, "_type": "Value"}, "sub_label": {"dtype": "string", "id": null, "_type": "Value"}, "sub_aliases": {"dtype": "string", "id": null, "_type": "Value"}, "obj_w": {"dtype": "string", "id": null, "_type": "Value"}, "obj_label": {"dtype": "string", "id": null, "_type": "Value"}, "obj_aliases": {"dtype": "string", "id": null, "_type": "Value"}, "uuid": {"dtype": "string", "id": null, "_type": "Value"}, "masked_sentence": {"dtype": "string", "id": null, "_type": "Value"}, "template": {"dtype": "string", "id": null, "_type": "Value"}, "template_negated": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "builder_name": "lama", "config_name": "google_re", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 7638657, "num_examples": 6106, "dataset_name": "lama"}}, "download_checksums": {"https://dl.fbaipublicfiles.com/LAMA/negated_data.tar.gz": {"num_bytes": 74639115, "checksum": "1a151058e6608e47983ea4c99c50bb69248c1c0763a04a3793b0a0b657aa0b61"}}, "download_size": 74639115, "post_processing_size": null, "dataset_size": 7638657, "size_in_bytes": 82277772}, "conceptnet": {"description": "LAMA is a dataset used to probe and analyze the factual and commonsense knowledge contained in pretrained language models. See https://github.com/facebookresearch/LAMA.\n", "citation": "@inproceedings{petroni2019language,\n title={Language Models as Knowledge Bases?},\n author={F. Petroni, T. Rockt{\"{a}}schel, A. H. Miller, P. Lewis, A. Bakhtin, Y. Wu and S. Riedel},\n booktitle={In: Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2019},\n year={2019}\n}\n@inproceedings{petroni2020how,\n title={How Context Affects Language Models' Factual Predictions},\n author={Fabio Petroni and Patrick Lewis and Aleksandra Piktus and Tim Rockt{\"a}schel and Yuxiang Wu and Alexander H. Miller and Sebastian Riedel},\n booktitle={Automated Knowledge Base Construction},\n year={2020},\n url={https://openreview.net/forum?id=025X0zPfn}\n}\n", "homepage": "https://github.com/facebookresearch/LAMA", "license": "The Creative Commons Attribution-Noncommercial 4.0 International License. see https://github.com/facebookresearch/LAMA/blob/master/LICENSE", "features": {"uuid": {"dtype": "string", "id": null, "_type": "Value"}, "sub": {"dtype": "string", "id": null, "_type": "Value"}, "obj": {"dtype": "string", "id": null, "_type": "Value"}, "pred": {"dtype": "string", "id": null, "_type": "Value"}, "obj_label": {"dtype": "string", "id": null, "_type": "Value"}, "masked_sentence": {"dtype": "string", "id": null, "_type": "Value"}, "negated": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "builder_name": "lama", "config_name": "conceptnet", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 4130000, "num_examples": 29774, "dataset_name": "lama"}}, "download_checksums": {"https://dl.fbaipublicfiles.com/LAMA/negated_data.tar.gz": {"num_bytes": 74639115, "checksum": "1a151058e6608e47983ea4c99c50bb69248c1c0763a04a3793b0a0b657aa0b61"}}, "download_size": 74639115, "post_processing_size": null, "dataset_size": 4130000, "size_in_bytes": 78769115}}
1
+ {"trex": {"description": "LAMA is a dataset used to probe and analyze the factual and commonsense knowledge contained in pretrained language models. See https://github.com/facebookresearch/LAMA.\n", "citation": "@inproceedings{petroni2019language,\n title={Language Models as Knowledge Bases?},\n author={F. Petroni, T. Rockt{\"{a}}schel, A. H. Miller, P. Lewis, A. Bakhtin, Y. Wu and S. Riedel},\n booktitle={In: Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2019},\n year={2019}\n}\n@inproceedings{petroni2020how,\n title={How Context Affects Language Models' Factual Predictions},\n author={Fabio Petroni and Patrick Lewis and Aleksandra Piktus and Tim Rockt{\"a}schel and Yuxiang Wu and Alexander H. Miller and Sebastian Riedel},\n booktitle={Automated Knowledge Base Construction},\n year={2020},\n url={https://openreview.net/forum?id=025X0zPfn}\n}\n", "homepage": "https://github.com/facebookresearch/LAMA", "license": "The Creative Commons Attribution-Noncommercial 4.0 International License. see https://github.com/facebookresearch/LAMA/blob/master/LICENSE", "features": {"uuid": {"dtype": "string", "id": null, "_type": "Value"}, "obj_uri": {"dtype": "string", "id": null, "_type": "Value"}, "obj_label": {"dtype": "string", "id": null, "_type": "Value"}, "sub_uri": {"dtype": "string", "id": null, "_type": "Value"}, "sub_label": {"dtype": "string", "id": null, "_type": "Value"}, "predicate_id": {"dtype": "string", "id": null, "_type": "Value"}, "sub_surface": {"dtype": "string", "id": null, "_type": "Value"}, "obj_surface": {"dtype": "string", "id": null, "_type": "Value"}, "masked_sentence": {"dtype": "string", "id": null, "_type": "Value"}, "template": {"dtype": "string", "id": null, "_type": "Value"}, "template_negated": {"dtype": "string", "id": null, "_type": "Value"}, "label": {"dtype": "string", "id": null, "_type": "Value"}, "description": {"dtype": "string", "id": null, "_type": "Value"}, "type": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "lama", "config_name": "trex", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 656913189, "num_examples": 1304391, "dataset_name": "lama"}}, "download_checksums": {"https://dl.fbaipublicfiles.com/LAMA/negated_data.tar.gz": {"num_bytes": 74639115, "checksum": "1a151058e6608e47983ea4c99c50bb69248c1c0763a04a3793b0a0b657aa0b61"}, "https://s3.amazonaws.com/datasets.huggingface.co/lama/relations.jsonl": {"num_bytes": 13086, "checksum": "154be499a67d5a681bdeaff3bce578a64064c6ce73e471523c6423071e3e5298"}}, "download_size": 74652201, "post_processing_size": null, "dataset_size": 656913189, "size_in_bytes": 731565390}, "squad": {"description": "LAMA is a dataset used to probe and analyze the factual and commonsense knowledge contained in pretrained language models. See https://github.com/facebookresearch/LAMA.\n", "citation": "@inproceedings{petroni2019language,\n title={Language Models as Knowledge Bases?},\n author={F. Petroni, T. Rockt{\"{a}}schel, A. H. Miller, P. Lewis, A. Bakhtin, Y. Wu and S. Riedel},\n booktitle={In: Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2019},\n year={2019}\n}\n@inproceedings{petroni2020how,\n title={How Context Affects Language Models' Factual Predictions},\n author={Fabio Petroni and Patrick Lewis and Aleksandra Piktus and Tim Rockt{\"a}schel and Yuxiang Wu and Alexander H. Miller and Sebastian Riedel},\n booktitle={Automated Knowledge Base Construction},\n year={2020},\n url={https://openreview.net/forum?id=025X0zPfn}\n}\n", "homepage": "https://github.com/facebookresearch/LAMA", "license": "The Creative Commons Attribution-Noncommercial 4.0 International License. see https://github.com/facebookresearch/LAMA/blob/master/LICENSE", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "sub_label": {"dtype": "string", "id": null, "_type": "Value"}, "obj_label": {"dtype": "string", "id": null, "_type": "Value"}, "negated": {"dtype": "string", "id": null, "_type": "Value"}, "masked_sentence": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "lama", "config_name": "squad", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 57188, "num_examples": 305, "dataset_name": "lama"}}, "download_checksums": {"https://dl.fbaipublicfiles.com/LAMA/negated_data.tar.gz": {"num_bytes": 74639115, "checksum": "1a151058e6608e47983ea4c99c50bb69248c1c0763a04a3793b0a0b657aa0b61"}}, "download_size": 74639115, "post_processing_size": null, "dataset_size": 57188, "size_in_bytes": 74696303}, "google_re": {"description": "LAMA is a dataset used to probe and analyze the factual and commonsense knowledge contained in pretrained language models. See https://github.com/facebookresearch/LAMA.\n", "citation": "@inproceedings{petroni2019language,\n title={Language Models as Knowledge Bases?},\n author={F. Petroni, T. Rockt{\"{a}}schel, A. H. Miller, P. Lewis, A. Bakhtin, Y. Wu and S. Riedel},\n booktitle={In: Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2019},\n year={2019}\n}\n@inproceedings{petroni2020how,\n title={How Context Affects Language Models' Factual Predictions},\n author={Fabio Petroni and Patrick Lewis and Aleksandra Piktus and Tim Rockt{\"a}schel and Yuxiang Wu and Alexander H. Miller and Sebastian Riedel},\n booktitle={Automated Knowledge Base Construction},\n year={2020},\n url={https://openreview.net/forum?id=025X0zPfn}\n}\n", "homepage": "https://github.com/facebookresearch/LAMA", "license": "The Creative Commons Attribution-Noncommercial 4.0 International License. see https://github.com/facebookresearch/LAMA/blob/master/LICENSE", "features": {"pred": {"dtype": "string", "id": null, "_type": "Value"}, "sub": {"dtype": "string", "id": null, "_type": "Value"}, "obj": {"dtype": "string", "id": null, "_type": "Value"}, "evidences": {"dtype": "string", "id": null, "_type": "Value"}, "judgments": {"dtype": "string", "id": null, "_type": "Value"}, "sub_w": {"dtype": "string", "id": null, "_type": "Value"}, "sub_label": {"dtype": "string", "id": null, "_type": "Value"}, "sub_aliases": {"dtype": "string", "id": null, "_type": "Value"}, "obj_w": {"dtype": "string", "id": null, "_type": "Value"}, "obj_label": {"dtype": "string", "id": null, "_type": "Value"}, "obj_aliases": {"dtype": "string", "id": null, "_type": "Value"}, "uuid": {"dtype": "string", "id": null, "_type": "Value"}, "masked_sentence": {"dtype": "string", "id": null, "_type": "Value"}, "template": {"dtype": "string", "id": null, "_type": "Value"}, "template_negated": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "lama", "config_name": "google_re", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 7638657, "num_examples": 6106, "dataset_name": "lama"}}, "download_checksums": {"https://dl.fbaipublicfiles.com/LAMA/negated_data.tar.gz": {"num_bytes": 74639115, "checksum": "1a151058e6608e47983ea4c99c50bb69248c1c0763a04a3793b0a0b657aa0b61"}}, "download_size": 74639115, "post_processing_size": null, "dataset_size": 7638657, "size_in_bytes": 82277772}, "conceptnet": {"description": "LAMA is a dataset used to probe and analyze the factual and commonsense knowledge contained in pretrained language models. See https://github.com/facebookresearch/LAMA.\n", "citation": "@inproceedings{petroni2019language,\n title={Language Models as Knowledge Bases?},\n author={F. Petroni, T. Rockt{\"{a}}schel, A. H. Miller, P. Lewis, A. Bakhtin, Y. Wu and S. Riedel},\n booktitle={In: Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2019},\n year={2019}\n}\n@inproceedings{petroni2020how,\n title={How Context Affects Language Models' Factual Predictions},\n author={Fabio Petroni and Patrick Lewis and Aleksandra Piktus and Tim Rockt{\"a}schel and Yuxiang Wu and Alexander H. Miller and Sebastian Riedel},\n booktitle={Automated Knowledge Base Construction},\n year={2020},\n url={https://openreview.net/forum?id=025X0zPfn}\n}\n", "homepage": "https://github.com/facebookresearch/LAMA", "license": "The Creative Commons Attribution-Noncommercial 4.0 International License. see https://github.com/facebookresearch/LAMA/blob/master/LICENSE", "features": {"uuid": {"dtype": "string", "id": null, "_type": "Value"}, "sub": {"dtype": "string", "id": null, "_type": "Value"}, "obj": {"dtype": "string", "id": null, "_type": "Value"}, "pred": {"dtype": "string", "id": null, "_type": "Value"}, "obj_label": {"dtype": "string", "id": null, "_type": "Value"}, "masked_sentence": {"dtype": "string", "id": null, "_type": "Value"}, "negated": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "lama", "config_name": "conceptnet", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 4130000, "num_examples": 29774, "dataset_name": "lama"}}, "download_checksums": {"https://dl.fbaipublicfiles.com/LAMA/negated_data.tar.gz": {"num_bytes": 74639115, "checksum": "1a151058e6608e47983ea4c99c50bb69248c1c0763a04a3793b0a0b657aa0b61"}}, "download_size": 74639115, "post_processing_size": null, "dataset_size": 4130000, "size_in_bytes": 78769115}}
dummy/trex/1.1.0/dummy_data.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a27203dba9d33e498c47575e0166d08a2c8f7fa3a83f680ab835d0bfa7073fe4
3
- size 2167
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:642f602212b312c023ca121b8031dd305401a26dbfd77b5fca624f4c5dd4467a
3
+ size 4229
lama.py CHANGED
@@ -15,9 +15,8 @@
15
  """The LAMA Dataset"""
16
 
17
 
18
- import glob
19
  import json
20
- import os
21
 
22
  import datasets
23
 
@@ -45,12 +44,9 @@ _HOMEPAGE = "https://github.com/facebookresearch/LAMA"
45
 
46
  _LICENSE = "The Creative Commons Attribution-Noncommercial 4.0 International License. see https://github.com/facebookresearch/LAMA/blob/master/LICENSE"
47
 
48
- _URLs = {
49
- "trex": "https://dl.fbaipublicfiles.com/LAMA/negated_data.tar.gz",
50
- "squad": "https://dl.fbaipublicfiles.com/LAMA/negated_data.tar.gz",
51
- "google_re": "https://dl.fbaipublicfiles.com/LAMA/negated_data.tar.gz",
52
- "conceptnet": "https://dl.fbaipublicfiles.com/LAMA/negated_data.tar.gz",
53
- }
54
 
55
 
56
  class Lama(datasets.GeneratorBasedBuilder):
@@ -168,16 +164,16 @@ class Lama(datasets.GeneratorBasedBuilder):
168
 
169
  def _split_generators(self, dl_manager):
170
  """Returns SplitGenerators."""
171
- my_urls = _URLs[self.config.name]
172
- data_dir = dl_manager.download_and_extract(my_urls)
173
  if self.config.name == "trex":
 
174
  return [
175
  datasets.SplitGenerator(
176
  name=datasets.Split.TRAIN,
177
  gen_kwargs={
178
- "filepath": [os.path.join(data_dir, "relations.jsonl")]
179
- + list(glob.glob(os.path.join(data_dir, "TREx", "*"))),
180
- "split": "train",
181
  },
182
  ),
183
  ]
@@ -186,15 +182,12 @@ class Lama(datasets.GeneratorBasedBuilder):
186
  datasets.SplitGenerator(
187
  name=datasets.Split.TRAIN,
188
  gen_kwargs={
189
- "filepath": [
190
- os.path.join(data_dir, *f.split("/"))
191
- for f in [
192
- "Google_RE/date_of_birth_test.jsonl",
193
- "Google_RE/place_of_birth_test.jsonl",
194
- "Google_RE/place_of_death_test.jsonl",
195
- ]
196
  ],
197
- "split": "train",
198
  },
199
  ),
200
  ]
@@ -203,8 +196,8 @@ class Lama(datasets.GeneratorBasedBuilder):
203
  datasets.SplitGenerator(
204
  name=datasets.Split.TRAIN,
205
  gen_kwargs={
206
- "filepath": os.path.join(data_dir, "ConceptNet", "test.jsonl"),
207
- "split": "train",
208
  },
209
  ),
210
  ]
@@ -213,26 +206,26 @@ class Lama(datasets.GeneratorBasedBuilder):
213
  datasets.SplitGenerator(
214
  name=datasets.Split.TRAIN,
215
  gen_kwargs={
216
- "filepath": os.path.join(data_dir, "Squad", "test.jsonl"),
217
- "split": "train",
218
  },
219
  ),
220
  ]
221
 
222
- def _generate_examples(self, filepath, split):
223
  """Yields examples from the LAMA dataset."""
 
224
  if self.config.name == "trex":
225
- paths = filepath
226
- relations_path = paths[0]
227
- paths = paths[1:]
228
  all_rels = {}
229
  with open(relations_path, encoding="utf-8") as f:
230
  for row in f:
231
  data = json.loads(row)
232
  all_rels[data["relation"]] = data
233
  id_ = -1
234
- for filepath in paths:
235
- with open(filepath, encoding="utf-8") as f:
 
 
236
  for row in f:
237
  data = json.loads(row)
238
  pred = all_rels.get(data["predicate_id"], {})
@@ -254,91 +247,104 @@ class Lama(datasets.GeneratorBasedBuilder):
254
  "description": str(pred.get("description", "")),
255
  "type": str(pred.get("type", "")),
256
  }
 
 
257
  elif self.config.name == "conceptnet":
258
  id_ = -1
259
- with open(filepath, encoding="utf-8") as f:
260
- for row in f:
261
- data = json.loads(row)
262
- if data.get("negated") is not None:
263
- for masked_sentence, negated in zip(data["masked_sentences"], data["negated"]):
264
- id_ += 1
265
- yield id_, {
266
- "uuid": str(data["uuid"]),
267
- "sub": str(data.get("sub", "")),
268
- "obj": str(data.get("obj", "")),
269
- "pred": str(data["pred"]),
270
- "obj_label": str(data["obj_label"]),
271
- "masked_sentence": str(masked_sentence),
272
- "negated": str(negated),
273
- }
274
- else:
275
- for masked_sentence in data["masked_sentences"]:
276
- id_ += 1
277
- yield id_, {
278
- "uuid": str(data["uuid"]),
279
- "sub": str(data.get("sub", "")),
280
- "obj": str(data.get("obj", "")),
281
- "pred": str(data["pred"]),
282
- "obj_label": str(data["obj_label"]),
283
- "masked_sentence": str(masked_sentence),
284
- "negated": str(""),
285
- }
 
 
 
 
286
  elif self.config.name == "squad":
287
  id_ = -1
288
- with open(filepath, encoding="utf-8") as f:
289
- for row in f:
290
- data = json.loads(row)
291
- for masked_sentence in data["masked_sentences"]:
292
- id_ += 1
293
- yield id_, {
294
- "id": str(data["id"]),
295
- "sub_label": str(data["sub_label"]),
296
- "obj_label": str(data["obj_label"]),
297
- "negated": str(data.get("negated", "")),
298
- "masked_sentence": str(masked_sentence),
299
- }
300
- elif self.config.name == "google_re":
301
- id_ = -1
302
- paths = filepath
303
- for filepath in paths:
304
- # from https://github.com/facebookresearch/LAMA/blob/master/scripts/run_experiments.py
305
- if "place_of_birth" in filepath:
306
- pred = {
307
- "relation": "place_of_birth",
308
- "template": "[X] was born in [Y] .",
309
- "template_negated": "[X] was not born in [Y] .",
310
- }
311
- elif "date_of_birth" in filepath:
312
- pred = {
313
- "relation": "date_of_birth",
314
- "template": "[X] (born [Y]).",
315
- "template_negated": "[X] (not born [Y]).",
316
- }
317
- else:
318
- pred = {
319
- "relation": "place_of_death",
320
- "template": "[X] died in [Y] .",
321
- "template_negated": "[X] did not die in [Y] .",
322
- }
323
- with open(filepath, encoding="utf-8") as f:
324
  for row in f:
325
  data = json.loads(row)
326
  for masked_sentence in data["masked_sentences"]:
327
  id_ += 1
328
  yield id_, {
329
- "pred": str(data["pred"]),
330
- "sub": str(data["sub"]),
331
- "obj": str(data["obj"]),
332
- "evidences": str(data["evidences"]),
333
- "judgments": str(data["judgments"]),
334
- "sub_w": str(data["sub_w"]),
335
  "sub_label": str(data["sub_label"]),
336
- "sub_aliases": str(data["sub_aliases"]),
337
- "obj_w": str(data["obj_w"]),
338
  "obj_label": str(data["obj_label"]),
339
- "obj_aliases": str(data["obj_aliases"]),
340
- "uuid": str(data["uuid"]),
341
  "masked_sentence": str(masked_sentence),
342
- "template": str(pred["template"]),
343
- "template_negated": str(pred["template_negated"]),
344
  }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
15
  """The LAMA Dataset"""
16
 
17
 
 
18
  import json
19
+ from fnmatch import fnmatch
20
 
21
  import datasets
22
 
44
 
45
  _LICENSE = "The Creative Commons Attribution-Noncommercial 4.0 International License. see https://github.com/facebookresearch/LAMA/blob/master/LICENSE"
46
 
47
+ _RELATIONS_URL = "https://s3.amazonaws.com/datasets.huggingface.co/lama/relations.jsonl"
48
+
49
+ _DATA_URL = "https://dl.fbaipublicfiles.com/LAMA/negated_data.tar.gz"
 
 
 
50
 
51
 
52
  class Lama(datasets.GeneratorBasedBuilder):
164
 
165
  def _split_generators(self, dl_manager):
166
  """Returns SplitGenerators."""
167
+ archive = dl_manager.download(_DATA_URL)
 
168
  if self.config.name == "trex":
169
+ relations_path = dl_manager.download(_RELATIONS_URL)
170
  return [
171
  datasets.SplitGenerator(
172
  name=datasets.Split.TRAIN,
173
  gen_kwargs={
174
+ "filepaths": ["TREx/*"],
175
+ "files": dl_manager.iter_archive(archive),
176
+ "relations_path": relations_path,
177
  },
178
  ),
179
  ]
182
  datasets.SplitGenerator(
183
  name=datasets.Split.TRAIN,
184
  gen_kwargs={
185
+ "filepaths": [
186
+ "Google_RE/date_of_birth_test.jsonl",
187
+ "Google_RE/place_of_birth_test.jsonl",
188
+ "Google_RE/place_of_death_test.jsonl",
 
 
 
189
  ],
190
+ "files": dl_manager.iter_archive(archive),
191
  },
192
  ),
193
  ]
196
  datasets.SplitGenerator(
197
  name=datasets.Split.TRAIN,
198
  gen_kwargs={
199
+ "filepaths": ["ConceptNet/test.jsonl"],
200
+ "files": dl_manager.iter_archive(archive),
201
  },
202
  ),
203
  ]
206
  datasets.SplitGenerator(
207
  name=datasets.Split.TRAIN,
208
  gen_kwargs={
209
+ "filepaths": ["Squad/test.jsonl"],
210
+ "files": dl_manager.iter_archive(archive),
211
  },
212
  ),
213
  ]
214
 
215
+ def _generate_examples(self, filepaths, files, relations_path=None):
216
  """Yields examples from the LAMA dataset."""
217
+ filepaths = list(filepaths)
218
  if self.config.name == "trex":
 
 
 
219
  all_rels = {}
220
  with open(relations_path, encoding="utf-8") as f:
221
  for row in f:
222
  data = json.loads(row)
223
  all_rels[data["relation"]] = data
224
  id_ = -1
225
+ inside_trec_directory = False
226
+ for path, f in files:
227
+ if any(fnmatch(path, pattern) for pattern in filepaths):
228
+ inside_trec_directory = True
229
  for row in f:
230
  data = json.loads(row)
231
  pred = all_rels.get(data["predicate_id"], {})
247
  "description": str(pred.get("description", "")),
248
  "type": str(pred.get("type", "")),
249
  }
250
+ elif inside_trec_directory:
251
+ break
252
  elif self.config.name == "conceptnet":
253
  id_ = -1
254
+ for path, f in files:
255
+ if not filepaths:
256
+ break
257
+ if path in list(filepaths):
258
+ for row in f:
259
+ data = json.loads(row)
260
+ if data.get("negated") is not None:
261
+ for masked_sentence, negated in zip(data["masked_sentences"], data["negated"]):
262
+ id_ += 1
263
+ yield id_, {
264
+ "uuid": str(data["uuid"]),
265
+ "sub": str(data.get("sub", "")),
266
+ "obj": str(data.get("obj", "")),
267
+ "pred": str(data["pred"]),
268
+ "obj_label": str(data["obj_label"]),
269
+ "masked_sentence": str(masked_sentence),
270
+ "negated": str(negated),
271
+ }
272
+ else:
273
+ for masked_sentence in data["masked_sentences"]:
274
+ id_ += 1
275
+ yield id_, {
276
+ "uuid": str(data["uuid"]),
277
+ "sub": str(data.get("sub", "")),
278
+ "obj": str(data.get("obj", "")),
279
+ "pred": str(data["pred"]),
280
+ "obj_label": str(data["obj_label"]),
281
+ "masked_sentence": str(masked_sentence),
282
+ "negated": str(""),
283
+ }
284
+ filepaths.remove(path)
285
  elif self.config.name == "squad":
286
  id_ = -1
287
+ for path, f in files:
288
+ if not filepaths:
289
+ break
290
+ if path in filepaths:
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
291
  for row in f:
292
  data = json.loads(row)
293
  for masked_sentence in data["masked_sentences"]:
294
  id_ += 1
295
  yield id_, {
296
+ "id": str(data["id"]),
 
 
 
 
 
297
  "sub_label": str(data["sub_label"]),
 
 
298
  "obj_label": str(data["obj_label"]),
299
+ "negated": str(data.get("negated", "")),
 
300
  "masked_sentence": str(masked_sentence),
 
 
301
  }
302
+ filepaths.remove(path)
303
+ elif self.config.name == "google_re":
304
+ id_ = -1
305
+ for path, f in files:
306
+ if path in filepaths:
307
+ if not filepaths:
308
+ break
309
+ if path in filepaths:
310
+ # from https://github.com/facebookresearch/LAMA/blob/master/scripts/run_experiments.py
311
+ if "place_of_birth" in path:
312
+ pred = {
313
+ "relation": "place_of_birth",
314
+ "template": "[X] was born in [Y] .",
315
+ "template_negated": "[X] was not born in [Y] .",
316
+ }
317
+ elif "date_of_birth" in path:
318
+ pred = {
319
+ "relation": "date_of_birth",
320
+ "template": "[X] (born [Y]).",
321
+ "template_negated": "[X] (not born [Y]).",
322
+ }
323
+ else:
324
+ pred = {
325
+ "relation": "place_of_death",
326
+ "template": "[X] died in [Y] .",
327
+ "template_negated": "[X] did not die in [Y] .",
328
+ }
329
+ for row in f:
330
+ data = json.loads(row)
331
+ for masked_sentence in data["masked_sentences"]:
332
+ id_ += 1
333
+ yield id_, {
334
+ "pred": str(data["pred"]),
335
+ "sub": str(data["sub"]),
336
+ "obj": str(data["obj"]),
337
+ "evidences": str(data["evidences"]),
338
+ "judgments": str(data["judgments"]),
339
+ "sub_w": str(data["sub_w"]),
340
+ "sub_label": str(data["sub_label"]),
341
+ "sub_aliases": str(data["sub_aliases"]),
342
+ "obj_w": str(data["obj_w"]),
343
+ "obj_label": str(data["obj_label"]),
344
+ "obj_aliases": str(data["obj_aliases"]),
345
+ "uuid": str(data["uuid"]),
346
+ "masked_sentence": str(masked_sentence),
347
+ "template": str(pred["template"]),
348
+ "template_negated": str(pred["template_negated"]),
349
+ }
350
+ filepaths.remove(path)