Datasets:
Update files from the datasets library (from 1.16.0)
Browse filesRelease notes: https://github.com/huggingface/datasets/releases/tag/1.16.0
eli5.py
CHANGED
@@ -115,9 +115,9 @@ def _open_compressed_file(f_name, f_type):
|
|
115 |
# download a file, extract posts from desired subreddit, then remove from disk
|
116 |
def _download_and_select_lines(dl_manager, f_url, mode, st_time):
|
117 |
# download and pre-process original posts
|
118 |
-
logger.info("downloading {} {
|
119 |
f_downloaded_path = dl_manager.download(f_url)
|
120 |
-
logger.info("decompressing and filtering {} {
|
121 |
f, fh = _open_compressed_file(f_downloaded_path, f_url.split(".")[-1])
|
122 |
lines = dict([(name, []) for name in _SUB_REDDITS])
|
123 |
for line in f:
|
@@ -130,7 +130,7 @@ def _download_and_select_lines(dl_manager, f_url, mode, st_time):
|
|
130 |
os.remove(f_downloaded_path)
|
131 |
os.remove(f_downloaded_path + ".json")
|
132 |
os.remove(f_downloaded_path + ".lock")
|
133 |
-
logger.info("tokenizing and selecting {} {
|
134 |
processed_items = dict([(name, []) for name in _SUB_REDDITS])
|
135 |
if mode == "submissions":
|
136 |
key_list = ["id", "score", "url", "title", "selftext", "subreddit"]
|
@@ -146,9 +146,7 @@ def _download_and_select_lines(dl_manager, f_url, mode, st_time):
|
|
146 |
else:
|
147 |
reddit_res[k] = line[k]
|
148 |
processed_items[name] += [reddit_res]
|
149 |
-
logger.info(
|
150 |
-
"Total found {} {} {:.2f}".format(sum([len(ls) for ls in processed_items.values()]), mode, time() - st_time)
|
151 |
-
)
|
152 |
return processed_items
|
153 |
|
154 |
|
@@ -191,7 +189,7 @@ def _download_and_filter_reddit(dl_manager, start_year=2011, start_month=7, end_
|
|
191 |
for dct in processed_submissions[name]:
|
192 |
qa_dict[name][dct["id"]] = dct
|
193 |
else:
|
194 |
-
logger.info("Could not find submissions dump file for year {:4d} month {:2d}"
|
195 |
# then all answers
|
196 |
for year in range(start_year, end_year + 1):
|
197 |
start_mth = start_month if year == start_year else 1
|
@@ -210,7 +208,7 @@ def _download_and_filter_reddit(dl_manager, start_year=2011, start_month=7, end_
|
|
210 |
merged_comments += 1
|
211 |
qa_dict[name][did]["comments"] = qa_dict[name][did].get("comments", []) + [dct]
|
212 |
else:
|
213 |
-
logger.info("Could not find comments dump file for year {:4d} month {:2d}"
|
214 |
# then post-process
|
215 |
res = {}
|
216 |
for name in _SUB_REDDITS:
|
@@ -354,7 +352,7 @@ class Eli5(datasets.GeneratorBasedBuilder):
|
|
354 |
]
|
355 |
|
356 |
def _generate_examples(self, split, subreddit_name):
|
357 |
-
logger.info("generating examples from = {}, {} set"
|
358 |
if split in self.data_split.get(subreddit_name, []):
|
359 |
id_list = self.data_split[subreddit_name][split]
|
360 |
data = [
|
@@ -386,7 +384,7 @@ class Eli5(datasets.GeneratorBasedBuilder):
|
|
386 |
for i, ans in enumerate(example["comments"]):
|
387 |
txt = ans["body"][0]
|
388 |
for j, _ in enumerate(ans["body"][1]):
|
389 |
-
txt = txt.replace("_URL_{}_"
|
390 |
answer_texts += [txt.strip()]
|
391 |
yield id_, {
|
392 |
"q_id": id_,
|
|
|
115 |
# download a file, extract posts from desired subreddit, then remove from disk
|
116 |
def _download_and_select_lines(dl_manager, f_url, mode, st_time):
|
117 |
# download and pre-process original posts
|
118 |
+
logger.info(f"downloading {f_url} {time() - st_time:.2f}")
|
119 |
f_downloaded_path = dl_manager.download(f_url)
|
120 |
+
logger.info(f"decompressing and filtering {f_url} {time() - st_time:.2f}")
|
121 |
f, fh = _open_compressed_file(f_downloaded_path, f_url.split(".")[-1])
|
122 |
lines = dict([(name, []) for name in _SUB_REDDITS])
|
123 |
for line in f:
|
|
|
130 |
os.remove(f_downloaded_path)
|
131 |
os.remove(f_downloaded_path + ".json")
|
132 |
os.remove(f_downloaded_path + ".lock")
|
133 |
+
logger.info("tokenizing and selecting {f_url} {time() - st_time:.2f}")
|
134 |
processed_items = dict([(name, []) for name in _SUB_REDDITS])
|
135 |
if mode == "submissions":
|
136 |
key_list = ["id", "score", "url", "title", "selftext", "subreddit"]
|
|
|
146 |
else:
|
147 |
reddit_res[k] = line[k]
|
148 |
processed_items[name] += [reddit_res]
|
149 |
+
logger.info(f"Total found {sum([len(ls) for ls in processed_items.values()])} {mode} {time() - st_time:.2f}")
|
|
|
|
|
150 |
return processed_items
|
151 |
|
152 |
|
|
|
189 |
for dct in processed_submissions[name]:
|
190 |
qa_dict[name][dct["id"]] = dct
|
191 |
else:
|
192 |
+
logger.info(f"Could not find submissions dump file for year {year:4d} month {month:2d}")
|
193 |
# then all answers
|
194 |
for year in range(start_year, end_year + 1):
|
195 |
start_mth = start_month if year == start_year else 1
|
|
|
208 |
merged_comments += 1
|
209 |
qa_dict[name][did]["comments"] = qa_dict[name][did].get("comments", []) + [dct]
|
210 |
else:
|
211 |
+
logger.info(f"Could not find comments dump file for year {year:4d} month {month:2d}")
|
212 |
# then post-process
|
213 |
res = {}
|
214 |
for name in _SUB_REDDITS:
|
|
|
352 |
]
|
353 |
|
354 |
def _generate_examples(self, split, subreddit_name):
|
355 |
+
logger.info(f"generating examples from = {subreddit_name}, {split} set")
|
356 |
if split in self.data_split.get(subreddit_name, []):
|
357 |
id_list = self.data_split[subreddit_name][split]
|
358 |
data = [
|
|
|
384 |
for i, ans in enumerate(example["comments"]):
|
385 |
txt = ans["body"][0]
|
386 |
for j, _ in enumerate(ans["body"][1]):
|
387 |
+
txt = txt.replace(f"_URL_{j}_", f"_URL_{map_url_indices[(i, j)]}_")
|
388 |
answer_texts += [txt.strip()]
|
389 |
yield id_, {
|
390 |
"q_id": id_,
|