Fix download links the multi_news dataset

PiperOrigin-RevId: 796933010

Author: The TensorFlow Datasets Authors

tensorflow_datasets/datasets/multi_news/README.md

@@ -1,8 +1,13 @@
-Multi-News, consists of news articles and human-written summaries
-of these articles from the site newser.com.
-Each summary is professionally written by editors and
-includes links to the original articles cited.
-
-There are two features:
-  - document: text of news articles seperated by special token "|||||".
-  - summary: news summary.
+# Multi-News Dataset
+
+Multi-News consists of news articles and human-written summaries of these
+articles from the news site `newser.com`. Each summary is professionally written
+by editors and includes links to the original articles cited.
+
+This is the first large-scale dataset for multi-document summarization on news
+articles.
+
+Each record has two features:
+
+* `document`: Texts of news articles, separated by special token "|||||".
+* `summary`: Summary of the news.

tensorflow_datasets/datasets/multi_news/TAGS.txt

@@ -0,0 +1,6 @@
+content.data-type.text # Contains text data.
+content.subject.news # Relates to news.
+content.language.en # Contains text in language English / en.
+ml.task.abstractive-text-summarization # Relates to Abstractive Text Summarization, a machine learning task.
+ml.task.natural-language-understanding # Relates to Natural Language Understanding, a machine learning task.
+ml.task.text-summarization # Relates to Text Summarization, a machine learning task.

tensorflow_datasets/datasets/multi_news/checksums.tsv

@@ -1,6 +1,6 @@
-https://huggingface.co/datasets/alexfabbri/multi_news/raw/main/data/test.src.cleaned	133	d04c4581d52321a30c246d2caa72853ee7f28c6b7a3985ee436f54c4bc264315	test.src.cleaned
-https://huggingface.co/datasets/alexfabbri/multi_news/raw/main/data/test.tgt	132	afba4aa26d95bb557c0eaa0cb8f7495af2104f1e43f4b5f9ef429b8752477abd	test.tgt
-https://huggingface.co/datasets/alexfabbri/multi_news/raw/main/data/train.src.cleaned	134	75f87b786ff1982bf1bd5803c6a7377d1834b81956ac680a6955789ba047cc0b	train.src.cleaned
-https://huggingface.co/datasets/alexfabbri/multi_news/raw/main/data/train.tgt	133	9f1e9b290a6aae1aa67bd5b361c934ee9db32486e5cd97d83184c097ef8b27e5	train.tgt
-https://huggingface.co/datasets/alexfabbri/multi_news/raw/main/data/val.src.cleaned	133	8df3ef6bd1882094de8120fa635c3abf758e10427f81f306aaa4786df7b57861	val.src.cleaned
-https://huggingface.co/datasets/alexfabbri/multi_news/raw/main/data/val.tgt	132	9c0377a443ea92b17449f7df17f1cdfa7c7ebbfe3a45f2f8cd7b3e0ffb47b1df	val.tgt
+https://huggingface.co/datasets/alexfabbri/multi_news/resolve/main/data/test.src.cleaned	68999509	138d3ac2dc899cbcd2e3745aaa94d1c1db55fb7058d9df4ba3ef2dac05a3a186	test.src.cleaned
+https://huggingface.co/datasets/alexfabbri/multi_news/resolve/main/data/test.tgt	7309099	fa97cf91a62ae82a0af6da88f2ddf8e06eb4e3b90f7971d8e0c516436518fae3	test.tgt
+https://huggingface.co/datasets/alexfabbri/multi_news/resolve/main/data/train.src.cleaned	547512283	627781c8ce55d528fcdacd495db45583a915e2d24b7983b0a5a6693ede933bb1	train.src.cleaned
+https://huggingface.co/datasets/alexfabbri/multi_news/resolve/main/data/train.tgt	58793912	e9e82b8f413b0f1ed4eb7c883f93bb744f829c218c1608b6ba7615d687d07121	train.tgt
+https://huggingface.co/datasets/alexfabbri/multi_news/resolve/main/data/val.src.cleaned	66875522	f0a43902da366eea2b882e39ddd4c0975ad44aba6b61095a2ea90362e9e2bb65	val.src.cleaned
+https://huggingface.co/datasets/alexfabbri/multi_news/resolve/main/data/val.tgt	7295302	bb08a078e0cb2b8ca9cc0fe3bfbe9d4098dee706bd00eb97449155e41b880157	val.tgt

tensorflow_datasets/datasets/multi_news/multi_news_dataset_builder.py

@@ -19,7 +19,7 @@
 import tensorflow_datasets.public_api as tfds
 
 _URL_PATH = (
-    "https://huggingface.co/datasets/alexfabbri/multi_news/raw/main/data/"
+    "https://huggingface.co/datasets/alexfabbri/multi_news/resolve/main/data/"
 )
 _LICENSE = "For non-commercial research and educational purposes only"
 
@@ -31,10 +31,11 @@
 class Builder(tfds.core.GeneratorBasedBuilder):
   """DatasetBuilder for multi_news dataset."""
 
-  VERSION = tfds.core.Version("2.0.0")
+  VERSION = tfds.core.Version("2.1.0")
   RELEASE_NOTES = {
       "1.0.0": "Initial release.",
       "2.0.0": "Update the dataset with valid URLs.",
+      "2.1.0": "Update the dataset with cleaned URLs.",
   }
 
   def _info(self) -> tfds.core.DatasetInfo: