system HF staff commited on
Commit
b59f946
0 Parent(s):

Update files from the datasets library (from 1.0.0)

Browse files

Release notes: https://github.com/huggingface/datasets/releases/tag/1.0.0

.gitattributes ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ *.7z filter=lfs diff=lfs merge=lfs -text
2
+ *.arrow filter=lfs diff=lfs merge=lfs -text
3
+ *.bin filter=lfs diff=lfs merge=lfs -text
4
+ *.bin.* filter=lfs diff=lfs merge=lfs -text
5
+ *.bz2 filter=lfs diff=lfs merge=lfs -text
6
+ *.ftz filter=lfs diff=lfs merge=lfs -text
7
+ *.gz filter=lfs diff=lfs merge=lfs -text
8
+ *.h5 filter=lfs diff=lfs merge=lfs -text
9
+ *.joblib filter=lfs diff=lfs merge=lfs -text
10
+ *.lfs.* filter=lfs diff=lfs merge=lfs -text
11
+ *.model filter=lfs diff=lfs merge=lfs -text
12
+ *.msgpack filter=lfs diff=lfs merge=lfs -text
13
+ *.onnx filter=lfs diff=lfs merge=lfs -text
14
+ *.ot filter=lfs diff=lfs merge=lfs -text
15
+ *.parquet filter=lfs diff=lfs merge=lfs -text
16
+ *.pb filter=lfs diff=lfs merge=lfs -text
17
+ *.pt filter=lfs diff=lfs merge=lfs -text
18
+ *.pth filter=lfs diff=lfs merge=lfs -text
19
+ *.rar filter=lfs diff=lfs merge=lfs -text
20
+ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
21
+ *.tar.* filter=lfs diff=lfs merge=lfs -text
22
+ *.tflite filter=lfs diff=lfs merge=lfs -text
23
+ *.tgz filter=lfs diff=lfs merge=lfs -text
24
+ *.xz filter=lfs diff=lfs merge=lfs -text
25
+ *.zip filter=lfs diff=lfs merge=lfs -text
26
+ *.zstandard filter=lfs diff=lfs merge=lfs -text
27
+ *tfevents* filter=lfs diff=lfs merge=lfs -text
dataset_infos.json ADDED
@@ -0,0 +1 @@
 
1
+ {"18828_alt.atheism": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\ndoes not include cross-posts and includes only the \"From\" and \"Subject\" headers.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "18828_alt.atheism", "version": {"version_str": "3.0.0", "description": null, "datasets_version_to_prepare": null, "major": 3, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1669711, "num_examples": 799, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz": {"num_bytes": 14666916, "checksum": "380962ab5174e64eb8fee393d539d0793e03eb2b59a69df2318398935f4a48e2"}}, "download_size": 14666916, "dataset_size": 1669711, "size_in_bytes": 16336627}, "18828_comp.graphics": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\ndoes not include cross-posts and includes only the \"From\" and \"Subject\" headers.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "18828_comp.graphics", "version": {"version_str": "3.0.0", "description": null, "datasets_version_to_prepare": null, "major": 3, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1661443, "num_examples": 973, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz": {"num_bytes": 14666916, "checksum": "380962ab5174e64eb8fee393d539d0793e03eb2b59a69df2318398935f4a48e2"}}, "download_size": 14666916, "dataset_size": 1661443, "size_in_bytes": 16328359}, "18828_comp.os.ms-windows.misc": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\ndoes not include cross-posts and includes only the \"From\" and \"Subject\" headers.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "18828_comp.os.ms-windows.misc", "version": {"version_str": "3.0.0", "description": null, "datasets_version_to_prepare": null, "major": 3, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 2378987, "num_examples": 985, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz": {"num_bytes": 14666916, "checksum": "380962ab5174e64eb8fee393d539d0793e03eb2b59a69df2318398935f4a48e2"}}, "download_size": 14666916, "dataset_size": 2378987, "size_in_bytes": 17045903}, "18828_comp.sys.ibm.pc.hardware": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\ndoes not include cross-posts and includes only the \"From\" and \"Subject\" headers.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "18828_comp.sys.ibm.pc.hardware", "version": {"version_str": "3.0.0", "description": null, "datasets_version_to_prepare": null, "major": 3, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1185433, "num_examples": 982, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz": {"num_bytes": 14666916, "checksum": "380962ab5174e64eb8fee393d539d0793e03eb2b59a69df2318398935f4a48e2"}}, "download_size": 14666916, "dataset_size": 1185433, "size_in_bytes": 15852349}, "18828_comp.sys.mac.hardware": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\ndoes not include cross-posts and includes only the \"From\" and \"Subject\" headers.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "18828_comp.sys.mac.hardware", "version": {"version_str": "3.0.0", "description": null, "datasets_version_to_prepare": null, "major": 3, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1056506, "num_examples": 961, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz": {"num_bytes": 14666916, "checksum": "380962ab5174e64eb8fee393d539d0793e03eb2b59a69df2318398935f4a48e2"}}, "download_size": 14666916, "dataset_size": 1056506, "size_in_bytes": 15723422}, "18828_comp.windows.x": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\ndoes not include cross-posts and includes only the \"From\" and \"Subject\" headers.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "18828_comp.windows.x", "version": {"version_str": "3.0.0", "description": null, "datasets_version_to_prepare": null, "major": 3, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1876543, "num_examples": 980, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz": {"num_bytes": 14666916, "checksum": "380962ab5174e64eb8fee393d539d0793e03eb2b59a69df2318398935f4a48e2"}}, "download_size": 14666916, "dataset_size": 1876543, "size_in_bytes": 16543459}, "18828_misc.forsale": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\ndoes not include cross-posts and includes only the \"From\" and \"Subject\" headers.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "18828_misc.forsale", "version": {"version_str": "3.0.0", "description": null, "datasets_version_to_prepare": null, "major": 3, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 925368, "num_examples": 972, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz": {"num_bytes": 14666916, "checksum": "380962ab5174e64eb8fee393d539d0793e03eb2b59a69df2318398935f4a48e2"}}, "download_size": 14666916, "dataset_size": 925368, "size_in_bytes": 15592284}, "18828_rec.autos": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\ndoes not include cross-posts and includes only the \"From\" and \"Subject\" headers.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "18828_rec.autos", "version": {"version_str": "3.0.0", "description": null, "datasets_version_to_prepare": null, "major": 3, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1295555, "num_examples": 990, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz": {"num_bytes": 14666916, "checksum": "380962ab5174e64eb8fee393d539d0793e03eb2b59a69df2318398935f4a48e2"}}, "download_size": 14666916, "dataset_size": 1295555, "size_in_bytes": 15962471}, "18828_rec.motorcycles": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\ndoes not include cross-posts and includes only the \"From\" and \"Subject\" headers.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "18828_rec.motorcycles", "version": {"version_str": "3.0.0", "description": null, "datasets_version_to_prepare": null, "major": 3, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1206741, "num_examples": 994, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz": {"num_bytes": 14666916, "checksum": "380962ab5174e64eb8fee393d539d0793e03eb2b59a69df2318398935f4a48e2"}}, "download_size": 14666916, "dataset_size": 1206741, "size_in_bytes": 15873657}, "18828_rec.sport.baseball": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\ndoes not include cross-posts and includes only the \"From\" and \"Subject\" headers.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "18828_rec.sport.baseball", "version": {"version_str": "3.0.0", "description": null, "datasets_version_to_prepare": null, "major": 3, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1369801, "num_examples": 994, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz": {"num_bytes": 14666916, "checksum": "380962ab5174e64eb8fee393d539d0793e03eb2b59a69df2318398935f4a48e2"}}, "download_size": 14666916, "dataset_size": 1369801, "size_in_bytes": 16036717}, "18828_rec.sport.hockey": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\ndoes not include cross-posts and includes only the \"From\" and \"Subject\" headers.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "18828_rec.sport.hockey", "version": {"version_str": "3.0.0", "description": null, "datasets_version_to_prepare": null, "major": 3, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1758344, "num_examples": 999, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz": {"num_bytes": 14666916, "checksum": "380962ab5174e64eb8fee393d539d0793e03eb2b59a69df2318398935f4a48e2"}}, "download_size": 14666916, "dataset_size": 1758344, "size_in_bytes": 16425260}, "18828_sci.crypt": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\ndoes not include cross-posts and includes only the \"From\" and \"Subject\" headers.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "18828_sci.crypt", "version": {"version_str": "3.0.0", "description": null, "datasets_version_to_prepare": null, "major": 3, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 2050975, "num_examples": 991, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz": {"num_bytes": 14666916, "checksum": "380962ab5174e64eb8fee393d539d0793e03eb2b59a69df2318398935f4a48e2"}}, "download_size": 14666916, "dataset_size": 2050975, "size_in_bytes": 16717891}, "18828_sci.electronics": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\ndoes not include cross-posts and includes only the \"From\" and \"Subject\" headers.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "18828_sci.electronics", "version": {"version_str": "3.0.0", "description": null, "datasets_version_to_prepare": null, "major": 3, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1237421, "num_examples": 981, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz": {"num_bytes": 14666916, "checksum": "380962ab5174e64eb8fee393d539d0793e03eb2b59a69df2318398935f4a48e2"}}, "download_size": 14666916, "dataset_size": 1237421, "size_in_bytes": 15904337}, "18828_sci.med": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\ndoes not include cross-posts and includes only the \"From\" and \"Subject\" headers.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "18828_sci.med", "version": {"version_str": "3.0.0", "description": null, "datasets_version_to_prepare": null, "major": 3, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1886611, "num_examples": 990, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz": {"num_bytes": 14666916, "checksum": "380962ab5174e64eb8fee393d539d0793e03eb2b59a69df2318398935f4a48e2"}}, "download_size": 14666916, "dataset_size": 1886611, "size_in_bytes": 16553527}, "18828_sci.space": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\ndoes not include cross-posts and includes only the \"From\" and \"Subject\" headers.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "18828_sci.space", "version": {"version_str": "3.0.0", "description": null, "datasets_version_to_prepare": null, "major": 3, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1813051, "num_examples": 987, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz": {"num_bytes": 14666916, "checksum": "380962ab5174e64eb8fee393d539d0793e03eb2b59a69df2318398935f4a48e2"}}, "download_size": 14666916, "dataset_size": 1813051, "size_in_bytes": 16479967}, "18828_soc.religion.christian": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\ndoes not include cross-posts and includes only the \"From\" and \"Subject\" headers.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "18828_soc.religion.christian", "version": {"version_str": "3.0.0", "description": null, "datasets_version_to_prepare": null, "major": 3, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 2307736, "num_examples": 997, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz": {"num_bytes": 14666916, "checksum": "380962ab5174e64eb8fee393d539d0793e03eb2b59a69df2318398935f4a48e2"}}, "download_size": 14666916, "dataset_size": 2307736, "size_in_bytes": 16974652}, "18828_talk.politics.guns": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\ndoes not include cross-posts and includes only the \"From\" and \"Subject\" headers.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "18828_talk.politics.guns", "version": {"version_str": "3.0.0", "description": null, "datasets_version_to_prepare": null, "major": 3, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1923220, "num_examples": 910, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz": {"num_bytes": 14666916, "checksum": "380962ab5174e64eb8fee393d539d0793e03eb2b59a69df2318398935f4a48e2"}}, "download_size": 14666916, "dataset_size": 1923220, "size_in_bytes": 16590136}, "18828_talk.politics.mideast": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\ndoes not include cross-posts and includes only the \"From\" and \"Subject\" headers.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "18828_talk.politics.mideast", "version": {"version_str": "3.0.0", "description": null, "datasets_version_to_prepare": null, "major": 3, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 2910560, "num_examples": 940, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz": {"num_bytes": 14666916, "checksum": "380962ab5174e64eb8fee393d539d0793e03eb2b59a69df2318398935f4a48e2"}}, "download_size": 14666916, "dataset_size": 2910560, "size_in_bytes": 17577476}, "18828_talk.politics.misc": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\ndoes not include cross-posts and includes only the \"From\" and \"Subject\" headers.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "18828_talk.politics.misc", "version": {"version_str": "3.0.0", "description": null, "datasets_version_to_prepare": null, "major": 3, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 2103003, "num_examples": 775, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz": {"num_bytes": 14666916, "checksum": "380962ab5174e64eb8fee393d539d0793e03eb2b59a69df2318398935f4a48e2"}}, "download_size": 14666916, "dataset_size": 2103003, "size_in_bytes": 16769919}, "18828_talk.religion.misc": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\ndoes not include cross-posts and includes only the \"From\" and \"Subject\" headers.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "18828_talk.religion.misc", "version": {"version_str": "3.0.0", "description": null, "datasets_version_to_prepare": null, "major": 3, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1374419, "num_examples": 628, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz": {"num_bytes": 14666916, "checksum": "380962ab5174e64eb8fee393d539d0793e03eb2b59a69df2318398935f4a48e2"}}, "download_size": 14666916, "dataset_size": 1374419, "size_in_bytes": 16041335}, "19997_alt.atheism": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nthe original, unmodified version.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "19997_alt.atheism", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 2562527, "num_examples": 1000, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz": {"num_bytes": 17332201, "checksum": "ab7bbf82b7831f7dbb1a09d9312f66fa78565c8de25526999b0d66f69d37e414"}}, "download_size": 17332201, "dataset_size": 2562527, "size_in_bytes": 19894728}, "19997_comp.graphics": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nthe original, unmodified version.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "19997_comp.graphics", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 2181923, "num_examples": 1000, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz": {"num_bytes": 17332201, "checksum": "ab7bbf82b7831f7dbb1a09d9312f66fa78565c8de25526999b0d66f69d37e414"}}, "download_size": 17332201, "dataset_size": 2181923, "size_in_bytes": 19514124}, "19997_comp.os.ms-windows.misc": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nthe original, unmodified version.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "19997_comp.os.ms-windows.misc", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 2899010, "num_examples": 1000, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz": {"num_bytes": 17332201, "checksum": "ab7bbf82b7831f7dbb1a09d9312f66fa78565c8de25526999b0d66f69d37e414"}}, "download_size": 17332201, "dataset_size": 2899010, "size_in_bytes": 20231211}, "19997_comp.sys.ibm.pc.hardware": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nthe original, unmodified version.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "19997_comp.sys.ibm.pc.hardware", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1671416, "num_examples": 1000, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz": {"num_bytes": 17332201, "checksum": "ab7bbf82b7831f7dbb1a09d9312f66fa78565c8de25526999b0d66f69d37e414"}}, "download_size": 17332201, "dataset_size": 1671416, "size_in_bytes": 19003617}, "19997_comp.sys.mac.hardware": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nthe original, unmodified version.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "19997_comp.sys.mac.hardware", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1581131, "num_examples": 1000, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz": {"num_bytes": 17332201, "checksum": "ab7bbf82b7831f7dbb1a09d9312f66fa78565c8de25526999b0d66f69d37e414"}}, "download_size": 17332201, "dataset_size": 1581131, "size_in_bytes": 18913332}, "19997_comp.windows.x": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nthe original, unmodified version.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "19997_comp.windows.x", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 2418523, "num_examples": 1000, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz": {"num_bytes": 17332201, "checksum": "ab7bbf82b7831f7dbb1a09d9312f66fa78565c8de25526999b0d66f69d37e414"}}, "download_size": 17332201, "dataset_size": 2418523, "size_in_bytes": 19750724}, "19997_misc.forsale": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nthe original, unmodified version.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "19997_misc.forsale", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1412262, "num_examples": 1000, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz": {"num_bytes": 17332201, "checksum": "ab7bbf82b7831f7dbb1a09d9312f66fa78565c8de25526999b0d66f69d37e414"}}, "download_size": 17332201, "dataset_size": 1412262, "size_in_bytes": 18744463}, "19997_rec.autos": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nthe original, unmodified version.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "19997_rec.autos", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1780752, "num_examples": 1000, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz": {"num_bytes": 17332201, "checksum": "ab7bbf82b7831f7dbb1a09d9312f66fa78565c8de25526999b0d66f69d37e414"}}, "download_size": 17332201, "dataset_size": 1780752, "size_in_bytes": 19112953}, "19997_rec.motorcycles": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nthe original, unmodified version.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "19997_rec.motorcycles", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1678214, "num_examples": 1000, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz": {"num_bytes": 17332201, "checksum": "ab7bbf82b7831f7dbb1a09d9312f66fa78565c8de25526999b0d66f69d37e414"}}, "download_size": 17332201, "dataset_size": 1678214, "size_in_bytes": 19010415}, "19997_rec.sport.baseball": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nthe original, unmodified version.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "19997_rec.sport.baseball", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1835682, "num_examples": 1000, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz": {"num_bytes": 17332201, "checksum": "ab7bbf82b7831f7dbb1a09d9312f66fa78565c8de25526999b0d66f69d37e414"}}, "download_size": 17332201, "dataset_size": 1835682, "size_in_bytes": 19167883}, "19997_rec.sport.hockey": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nthe original, unmodified version.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "19997_rec.sport.hockey", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 2207532, "num_examples": 1000, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz": {"num_bytes": 17332201, "checksum": "ab7bbf82b7831f7dbb1a09d9312f66fa78565c8de25526999b0d66f69d37e414"}}, "download_size": 17332201, "dataset_size": 2207532, "size_in_bytes": 19539733}, "19997_sci.crypt": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nthe original, unmodified version.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "19997_sci.crypt", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 2608085, "num_examples": 1000, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz": {"num_bytes": 17332201, "checksum": "ab7bbf82b7831f7dbb1a09d9312f66fa78565c8de25526999b0d66f69d37e414"}}, "download_size": 17332201, "dataset_size": 2608085, "size_in_bytes": 19940286}, "19997_sci.electronics": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nthe original, unmodified version.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "19997_sci.electronics", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1732449, "num_examples": 1000, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz": {"num_bytes": 17332201, "checksum": "ab7bbf82b7831f7dbb1a09d9312f66fa78565c8de25526999b0d66f69d37e414"}}, "download_size": 17332201, "dataset_size": 1732449, "size_in_bytes": 19064650}, "19997_sci.med": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nthe original, unmodified version.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "19997_sci.med", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 2389039, "num_examples": 1000, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz": {"num_bytes": 17332201, "checksum": "ab7bbf82b7831f7dbb1a09d9312f66fa78565c8de25526999b0d66f69d37e414"}}, "download_size": 17332201, "dataset_size": 2389039, "size_in_bytes": 19721240}, "19997_sci.space": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nthe original, unmodified version.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "19997_sci.space", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 2351661, "num_examples": 1000, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz": {"num_bytes": 17332201, "checksum": "ab7bbf82b7831f7dbb1a09d9312f66fa78565c8de25526999b0d66f69d37e414"}}, "download_size": 17332201, "dataset_size": 2351661, "size_in_bytes": 19683862}, "19997_soc.religion.christian": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nthe original, unmodified version.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "19997_soc.religion.christian", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 2743268, "num_examples": 997, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz": {"num_bytes": 17332201, "checksum": "ab7bbf82b7831f7dbb1a09d9312f66fa78565c8de25526999b0d66f69d37e414"}}, "download_size": 17332201, "dataset_size": 2743268, "size_in_bytes": 20075469}, "19997_talk.politics.guns": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nthe original, unmodified version.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "19997_talk.politics.guns", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 2639593, "num_examples": 1000, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz": {"num_bytes": 17332201, "checksum": "ab7bbf82b7831f7dbb1a09d9312f66fa78565c8de25526999b0d66f69d37e414"}}, "download_size": 17332201, "dataset_size": 2639593, "size_in_bytes": 19971794}, "19997_talk.politics.mideast": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nthe original, unmodified version.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "19997_talk.politics.mideast", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 3696181, "num_examples": 1000, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz": {"num_bytes": 17332201, "checksum": "ab7bbf82b7831f7dbb1a09d9312f66fa78565c8de25526999b0d66f69d37e414"}}, "download_size": 17332201, "dataset_size": 3696181, "size_in_bytes": 21028382}, "19997_talk.politics.misc": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nthe original, unmodified version.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "19997_talk.politics.misc", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 3169433, "num_examples": 1000, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz": {"num_bytes": 17332201, "checksum": "ab7bbf82b7831f7dbb1a09d9312f66fa78565c8de25526999b0d66f69d37e414"}}, "download_size": 17332201, "dataset_size": 3169433, "size_in_bytes": 20501634}, "19997_talk.religion.misc": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nthe original, unmodified version.", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "19997_talk.religion.misc", "version": {"version_str": "1.0.0", "description": null, "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 2658950, "num_examples": 1000, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz": {"num_bytes": 17332201, "checksum": "ab7bbf82b7831f7dbb1a09d9312f66fa78565c8de25526999b0d66f69d37e414"}}, "download_size": 17332201, "dataset_size": 2658950, "size_in_bytes": 19991151}, "bydate_alt.atheism": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nsorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "bydate_alt.atheism", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1042344, "num_examples": 480, "dataset_name": "newsgroups"}, "test": {"name": "test", "num_bytes": 1042344, "num_examples": 480, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz": {"num_bytes": 14464277, "checksum": "8f1b2514ca22a5ade8fbb9cfa5727df95fa587f4c87b786e15c759fa66d95610"}}, "download_size": 14464277, "dataset_size": 2084688, "size_in_bytes": 16548965}, "bydate_comp.graphics": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nsorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "bydate_comp.graphics", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 911811, "num_examples": 584, "dataset_name": "newsgroups"}, "test": {"name": "test", "num_bytes": 911811, "num_examples": 584, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz": {"num_bytes": 14464277, "checksum": "8f1b2514ca22a5ade8fbb9cfa5727df95fa587f4c87b786e15c759fa66d95610"}}, "download_size": 14464277, "dataset_size": 1823622, "size_in_bytes": 16287899}, "bydate_comp.os.ms-windows.misc": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nsorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "bydate_comp.os.ms-windows.misc", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1771136, "num_examples": 591, "dataset_name": "newsgroups"}, "test": {"name": "test", "num_bytes": 1771136, "num_examples": 591, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz": {"num_bytes": 14464277, "checksum": "8f1b2514ca22a5ade8fbb9cfa5727df95fa587f4c87b786e15c759fa66d95610"}}, "download_size": 14464277, "dataset_size": 3542272, "size_in_bytes": 18006549}, "bydate_comp.sys.ibm.pc.hardware": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nsorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "bydate_comp.sys.ibm.pc.hardware", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 800594, "num_examples": 590, "dataset_name": "newsgroups"}, "test": {"name": "test", "num_bytes": 800594, "num_examples": 590, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz": {"num_bytes": 14464277, "checksum": "8f1b2514ca22a5ade8fbb9cfa5727df95fa587f4c87b786e15c759fa66d95610"}}, "download_size": 14464277, "dataset_size": 1601188, "size_in_bytes": 16065465}, "bydate_comp.sys.mac.hardware": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nsorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "bydate_comp.sys.mac.hardware", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 696457, "num_examples": 578, "dataset_name": "newsgroups"}, "test": {"name": "test", "num_bytes": 696457, "num_examples": 578, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz": {"num_bytes": 14464277, "checksum": "8f1b2514ca22a5ade8fbb9cfa5727df95fa587f4c87b786e15c759fa66d95610"}}, "download_size": 14464277, "dataset_size": 1392914, "size_in_bytes": 15857191}, "bydate_comp.windows.x": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nsorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "bydate_comp.windows.x", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1243613, "num_examples": 593, "dataset_name": "newsgroups"}, "test": {"name": "test", "num_bytes": 1243613, "num_examples": 593, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz": {"num_bytes": 14464277, "checksum": "8f1b2514ca22a5ade8fbb9cfa5727df95fa587f4c87b786e15c759fa66d95610"}}, "download_size": 14464277, "dataset_size": 2487226, "size_in_bytes": 16951503}, "bydate_misc.forsale": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nsorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "bydate_misc.forsale", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 611358, "num_examples": 585, "dataset_name": "newsgroups"}, "test": {"name": "test", "num_bytes": 611358, "num_examples": 585, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz": {"num_bytes": 14464277, "checksum": "8f1b2514ca22a5ade8fbb9cfa5727df95fa587f4c87b786e15c759fa66d95610"}}, "download_size": 14464277, "dataset_size": 1222716, "size_in_bytes": 15686993}, "bydate_rec.autos": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nsorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "bydate_rec.autos", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 860796, "num_examples": 594, "dataset_name": "newsgroups"}, "test": {"name": "test", "num_bytes": 860796, "num_examples": 594, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz": {"num_bytes": 14464277, "checksum": "8f1b2514ca22a5ade8fbb9cfa5727df95fa587f4c87b786e15c759fa66d95610"}}, "download_size": 14464277, "dataset_size": 1721592, "size_in_bytes": 16185869}, "bydate_rec.motorcycles": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nsorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "bydate_rec.motorcycles", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 811301, "num_examples": 598, "dataset_name": "newsgroups"}, "test": {"name": "test", "num_bytes": 811301, "num_examples": 598, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz": {"num_bytes": 14464277, "checksum": "8f1b2514ca22a5ade8fbb9cfa5727df95fa587f4c87b786e15c759fa66d95610"}}, "download_size": 14464277, "dataset_size": 1622602, "size_in_bytes": 16086879}, "bydate_rec.sport.baseball": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nsorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "bydate_rec.sport.baseball", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 850890, "num_examples": 597, "dataset_name": "newsgroups"}, "test": {"name": "test", "num_bytes": 850890, "num_examples": 597, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz": {"num_bytes": 14464277, "checksum": "8f1b2514ca22a5ade8fbb9cfa5727df95fa587f4c87b786e15c759fa66d95610"}}, "download_size": 14464277, "dataset_size": 1701780, "size_in_bytes": 16166057}, "bydate_rec.sport.hockey": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nsorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "bydate_rec.sport.hockey", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1189802, "num_examples": 600, "dataset_name": "newsgroups"}, "test": {"name": "test", "num_bytes": 1189802, "num_examples": 600, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz": {"num_bytes": 14464277, "checksum": "8f1b2514ca22a5ade8fbb9cfa5727df95fa587f4c87b786e15c759fa66d95610"}}, "download_size": 14464277, "dataset_size": 2379604, "size_in_bytes": 16843881}, "bydate_sci.crypt": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nsorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "bydate_sci.crypt", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1502598, "num_examples": 595, "dataset_name": "newsgroups"}, "test": {"name": "test", "num_bytes": 1502598, "num_examples": 595, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz": {"num_bytes": 14464277, "checksum": "8f1b2514ca22a5ade8fbb9cfa5727df95fa587f4c87b786e15c759fa66d95610"}}, "download_size": 14464277, "dataset_size": 3005196, "size_in_bytes": 17469473}, "bydate_sci.electronics": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nsorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "bydate_sci.electronics", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 815004, "num_examples": 591, "dataset_name": "newsgroups"}, "test": {"name": "test", "num_bytes": 815004, "num_examples": 591, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz": {"num_bytes": 14464277, "checksum": "8f1b2514ca22a5ade8fbb9cfa5727df95fa587f4c87b786e15c759fa66d95610"}}, "download_size": 14464277, "dataset_size": 1630008, "size_in_bytes": 16094285}, "bydate_sci.med": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nsorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "bydate_sci.med", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1195351, "num_examples": 594, "dataset_name": "newsgroups"}, "test": {"name": "test", "num_bytes": 1195351, "num_examples": 594, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz": {"num_bytes": 14464277, "checksum": "8f1b2514ca22a5ade8fbb9cfa5727df95fa587f4c87b786e15c759fa66d95610"}}, "download_size": 14464277, "dataset_size": 2390702, "size_in_bytes": 16854979}, "bydate_sci.space": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nsorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "bydate_sci.space", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1198115, "num_examples": 593, "dataset_name": "newsgroups"}, "test": {"name": "test", "num_bytes": 1198115, "num_examples": 593, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz": {"num_bytes": 14464277, "checksum": "8f1b2514ca22a5ade8fbb9cfa5727df95fa587f4c87b786e15c759fa66d95610"}}, "download_size": 14464277, "dataset_size": 2396230, "size_in_bytes": 16860507}, "bydate_soc.religion.christian": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nsorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "bydate_soc.religion.christian", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1358197, "num_examples": 599, "dataset_name": "newsgroups"}, "test": {"name": "test", "num_bytes": 1358197, "num_examples": 599, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz": {"num_bytes": 14464277, "checksum": "8f1b2514ca22a5ade8fbb9cfa5727df95fa587f4c87b786e15c759fa66d95610"}}, "download_size": 14464277, "dataset_size": 2716394, "size_in_bytes": 17180671}, "bydate_talk.politics.guns": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nsorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "bydate_talk.politics.guns", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1313157, "num_examples": 546, "dataset_name": "newsgroups"}, "test": {"name": "test", "num_bytes": 1313157, "num_examples": 546, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz": {"num_bytes": 14464277, "checksum": "8f1b2514ca22a5ade8fbb9cfa5727df95fa587f4c87b786e15c759fa66d95610"}}, "download_size": 14464277, "dataset_size": 2626314, "size_in_bytes": 17090591}, "bydate_talk.politics.mideast": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nsorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "bydate_talk.politics.mideast", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1765975, "num_examples": 564, "dataset_name": "newsgroups"}, "test": {"name": "test", "num_bytes": 1765975, "num_examples": 564, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz": {"num_bytes": 14464277, "checksum": "8f1b2514ca22a5ade8fbb9cfa5727df95fa587f4c87b786e15c759fa66d95610"}}, "download_size": 14464277, "dataset_size": 3531950, "size_in_bytes": 17996227}, "bydate_talk.politics.misc": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nsorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "bydate_talk.politics.misc", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1328175, "num_examples": 465, "dataset_name": "newsgroups"}, "test": {"name": "test", "num_bytes": 1328175, "num_examples": 465, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz": {"num_bytes": 14464277, "checksum": "8f1b2514ca22a5ade8fbb9cfa5727df95fa587f4c87b786e15c759fa66d95610"}}, "download_size": 14464277, "dataset_size": 2656350, "size_in_bytes": 17120627}, "bydate_talk.religion.misc": {"description": "\nThe 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across \n20 different newsgroups. To the best of my knowledge, it was originally collected by Ken Lang, probably for his Newsweeder: \nLearning to filter netnews paper, though he does not explicitly mention this collection. The 20 newsgroups collection has become \na popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.\n\nsorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)", "citation": "\n\n", "homepage": "http://qwone.com/~jason/20Newsgroups/", "license": "", "features": {"text": {"dtype": "string", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "newsgroups", "config_name": "bydate_talk.religion.misc", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 835857, "num_examples": 377, "dataset_name": "newsgroups"}, "test": {"name": "test", "num_bytes": 835857, "num_examples": 377, "dataset_name": "newsgroups"}}, "download_checksums": {"http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz": {"num_bytes": 14464277, "checksum": "8f1b2514ca22a5ade8fbb9cfa5727df95fa587f4c87b786e15c759fa66d95610"}}, "download_size": 14464277, "dataset_size": 1671714, "size_in_bytes": 16135991}}
dummy/18828_alt.atheism/3.0.0/dummy_data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:01d27e67178e23decedb4ece5d1af0d995d29b1de50048911dc5e8d2c65ba456
3
+ size 19084
newsgroup.py ADDED
@@ -0,0 +1,162 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # coding=utf-8
2
+ # Copyright 2020 The TensorFlow Datasets Authors and the HuggingFace Datasets Authors.
3
+ #
4
+ # Licensed under the Apache License, Version 2.0 (the "License");
5
+ # you may not use this file except in compliance with the License.
6
+ # You may obtain a copy of the License at
7
+ #
8
+ # http://www.apache.org/licenses/LICENSE-2.0
9
+ #
10
+ # Unless required by applicable law or agreed to in writing, software
11
+ # distributed under the License is distributed on an "AS IS" BASIS,
12
+ # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13
+ # See the License for the specific language governing permissions and
14
+ # limitations under the License.
15
+
16
+ # Lint as: python3
17
+ """20Newsgroup dataset"""
18
+
19
+ from __future__ import absolute_import, division, print_function
20
+
21
+ import os
22
+
23
+ import datasets
24
+
25
+
26
+ _CITATION = """
27
+ @inproceedings{Lang95,
28
+ author = {Ken Lang},
29
+ title = {Newsweeder: Learning to filter netnews}
30
+ year = {1995}
31
+ booktitle = {Proceedings of the Twelfth International Conference on Machine Learning}
32
+ pages = {331-339}
33
+ }
34
+ """
35
+
36
+ _DESCRIPTION = """
37
+ The 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across
38
+ 20 different newsgroups. The 20 newsgroups collection has become a popular data set for experiments in text applications of
39
+ machine learning techniques, such as text classification and text clustering.
40
+ """
41
+
42
+ _DOWNLOAD_URL = {
43
+ "bydate": "http://qwone.com/~jason/20Newsgroups/20news-bydate.tar.gz",
44
+ "19997": "http://qwone.com/~jason/20Newsgroups/20news-19997.tar.gz",
45
+ "18828": "http://qwone.com/~jason/20Newsgroups/20news-18828.tar.gz",
46
+ }
47
+ _NEWS_GROUPS = [
48
+ "comp.graphics",
49
+ "comp.os.ms-windows.misc",
50
+ "comp.sys.ibm.pc.hardware",
51
+ "comp.sys.mac.hardware",
52
+ "comp.windows.x",
53
+ "rec.autos",
54
+ "rec.motorcycles",
55
+ "rec.sport.baseball",
56
+ "rec.sport.hockey",
57
+ "sci.crypt",
58
+ "sci.electronics",
59
+ "sci.med",
60
+ "sci.space",
61
+ "misc.forsale",
62
+ "talk.politics.misc",
63
+ "talk.politics.guns",
64
+ "talk.politics.mideast",
65
+ "talk.religion.misc",
66
+ "alt.atheism",
67
+ "soc.religion.christian",
68
+ ]
69
+ _VERSIONS = {"19997": "1.0.0", "bydate": "2.0.0", "18828": "3.0.0"}
70
+
71
+ _DESC = {
72
+ "19997": "the original, unmodified version.",
73
+ "bydate": "sorted by date into training(60%) and test(40%) sets, does not include cross-posts (duplicates) and does not include newsgroup-identifying headers (Xref, Newsgroups, Path, Followup-To, Date)",
74
+ "18828": 'does not include cross-posts and includes only the "From" and "Subject" headers.',
75
+ }
76
+ _CONFIG_NAMES = []
77
+ for version in _VERSIONS:
78
+ for group in _NEWS_GROUPS:
79
+ _CONFIG_NAMES.append(version + "_" + group)
80
+
81
+ _CONFIG_NAMES = sorted(_CONFIG_NAMES)
82
+
83
+
84
+ class NewsgroupConfig(datasets.BuilderConfig):
85
+ """BuilderConfig for 20Newsgroup."""
86
+
87
+ def __init__(self, sub_dir, **kwargs):
88
+ """Constructs a 20Newsgroup.
89
+
90
+ Args:
91
+ sub_dirs: str
92
+ **kwargs: keyword arguments forwarded to super.
93
+ """
94
+
95
+ super(NewsgroupConfig, self).__init__(**kwargs)
96
+ self.sub_dir = sub_dir
97
+
98
+
99
+ class Newsgroups(datasets.GeneratorBasedBuilder):
100
+
101
+ BUILDER_CONFIGS = [
102
+ NewsgroupConfig(
103
+ name=name,
104
+ description=_DESC[name.split("_")[0]],
105
+ sub_dir=name.split("_")[1],
106
+ version=datasets.Version(_VERSIONS[name.split("_")[0]]),
107
+ )
108
+ for name in _CONFIG_NAMES
109
+ ]
110
+
111
+ def _info(self):
112
+ return datasets.DatasetInfo(
113
+ description=_DESCRIPTION + "\n" + self.config.description,
114
+ features=datasets.Features(
115
+ {
116
+ "text": datasets.Value("string"),
117
+ }
118
+ ),
119
+ homepage="http://qwone.com/~jason/20Newsgroups/",
120
+ citation=_CITATION,
121
+ )
122
+
123
+ def _split_generators(self, dl_manager):
124
+ url = _DOWNLOAD_URL[self.config.name.split("_")[0]]
125
+ path = dl_manager.download_and_extract(url)
126
+ if self.config.name.startswith("bydate"):
127
+
128
+ return [
129
+ datasets.SplitGenerator(
130
+ name=datasets.Split.TRAIN,
131
+ gen_kwargs={"files_path": os.path.join(path, "20news-bydate-train", self.config.sub_dir)},
132
+ ),
133
+ datasets.SplitGenerator(
134
+ name=datasets.Split.TEST,
135
+ gen_kwargs={"files_path": os.path.join(path, "20news-bydate-train", self.config.sub_dir)},
136
+ ),
137
+ ]
138
+ elif self.config.name.startswith("19997"):
139
+ return [
140
+ datasets.SplitGenerator(
141
+ name=datasets.Split.TRAIN,
142
+ gen_kwargs={"files_path": os.path.join(path, "20_newsgroups", self.config.sub_dir)},
143
+ )
144
+ ]
145
+ else:
146
+ return [
147
+ datasets.SplitGenerator(
148
+ name=datasets.Split.TRAIN,
149
+ gen_kwargs={"files_path": os.path.join(path, "20news-18828", self.config.sub_dir)},
150
+ )
151
+ ]
152
+
153
+ def _generate_examples(self, files_path):
154
+ """Yields examples."""
155
+ files = sorted(os.listdir(files_path))
156
+ for id_, file in enumerate(files):
157
+ filepath = os.path.join(files_path, file)
158
+ with open(
159
+ filepath, encoding="utf8", errors="ignore"
160
+ ) as f: # here we can ignore byte encoded tokens. we only have a very few and in most case it happens at the end of the file (kind of \FF)
161
+ text = f.read()
162
+ yield id_, {"text": text}