Commit 82880154 authored by Jason Phang's avatar Jason Phang
Browse files

add pile tasks

parent fa80f7bd
......@@ -174,8 +174,29 @@ TASK_REGISTRY = {
"reversed_words": unscramble.ReversedWords,
# Pile
"pile_enron": pile.PileEnronPerplexityTask,
"pile_ubuntu": pile.PileUbuntuPerplexityTask,
"arxiv": pile.PileArxivPerplexityTask,
"books3": pile.PileBooks3PerplexityTask,
"bookcorpus2": pile.PileBookCorpus2PerplexityTask,
"commoncrawl": pile.PileCommonCrawlPerplexityTask,
"dm-mathematics": pile.PileDmMathematicsPerplexityTask,
"enron": pile.PileEnronPerplexityTask,
"europarl": pile.PileEuroparlPerplexityTask,
"freelaw": pile.PileFreeLawPerplexityTask,
"github": pile.PileGithubPerplexityTask,
"gutenberg": pile.PileGutenbergPerplexityTask,
"hackernews": pile.PileHackernewsPerplexityTask,
"nih-exporter": pile.PileNIHExporterPerplexityTask,
"opensubtitles": pile.PileOpenSubtitlesPerplexityTask,
"openwebtext2": pile.PileOpenWebText2PerplexityTask,
"philpapers": pile.PilePhilPapersPerplexityTask,
"pile-cc": pile.PilePileCcPerplexityTask,
"pubmed-abstracts": pile.PilePubmedAbstractsPerplexityTask,
"pubmed-central": pile.PilePubmedCentralPerplexityTask,
"stackexchange": pile.PileStackExchangePerplexityTask,
"uspto": pile.PileUsptoPerplexityTask,
"ubuntu-irc": pile.PileUbuntuIrcPerplexityTask,
"wikipedia": pile.PileWikipediaPerplexityTask,
"youtubesubtitles": pile.PileYoutubeSubtitlesPerplexityTask,
}
......
......@@ -41,9 +41,93 @@ class PilePerplexityTask(PerplexityTask, abc.ABC):
return True
class PileArxivPerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "ArXiv"
class PileBooks3PerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "Books3"
class PileBookCorpus2PerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "BookCorpus2"
class PileCommonCrawlPerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "CommonCrawl"
class PileDmMathematicsPerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "DM Mathematics"
class PileEnronPerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "Enron Emails"
class PileUbuntuPerplexityTask(PilePerplexityTask):
class PileEuroparlPerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "EuroParl"
class PileFreeLawPerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "FreeLaw"
class PileGithubPerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "Github"
class PileGutenbergPerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "Gutenberg (PG-19)"
class PileHackernewsPerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "HackerNews"
class PileNIHExporterPerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "NIH ExPorter"
class PileOpenSubtitlesPerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "OpenSubtitles"
class PileOpenWebText2PerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "OpenWebText2"
class PilePhilPapersPerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "PhilPapers"
class PilePileCcPerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "Pile-CC"
class PilePubmedAbstractsPerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "PubMed Abstracts"
class PilePubmedCentralPerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "PubMed Central"
class PileStackExchangePerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "StackExchange"
class PileUsptoPerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "USPTO Backgrounds"
class PileUbuntuIrcPerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "Ubuntu IRC"
class PileWikipediaPerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "Wikipedia (en)"
class PileYoutubeSubtitlesPerplexityTask(PilePerplexityTask):
PILE_SET_NAME = "YoutubeSubtitles"
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment