AI & machine learning2021 · shared on Hacker News
The Pile: An 800GB Dataset of Diverse Text for Language Modeling
Explore The Pile: An 800GB Dataset of Diverse Text for Language Modeling, its original source and Hacker News discussions. Discover community perspectives and save it to your reading list.
GO BEYOND THE ABSTRACT
The conversation around the idea.
Original Hacker News threads. Different perspectives, questions, and things worth thinking about.
Showing up to 100 related threads. A paper may appear in a comment rather than the original submission.
Links found in the discussion
Up to 50 recent source links, in HN item order.
Comment #47851548 Comment #42973261 Comment #39842390 Comment #38621946 Comment #38296516 Comment #37415976 Comment #37354015 Comment #36823596 Comment #36659041 Comment #36318068 Comment #36134481 Comment #35610781 Comment #34743665 Comment #34549511 Comment #34451559 Comment #34357280 Comment #34298310 Comment #34047062 Comment #33927432 Comment #33861206 Comment #31180740 Comment #28941658 Comment #27727687 Comment #26624096
Ygameshot911 ·
Meta torrented & seeded 81.7 TB dataset containing copyrighted data
1,270 points938 comments
Yjasondavies ·
DBRX: A new open LLM
866 points343 comments
Ycanada_dry ·
GPT-J-6B – A 6 billion parameter, autoregressive text generation model
636 points146 comments
Ycrazypython ·
Textsynth: Bellard's free GPT-NeoX-20B, GPT-J playground and paid API
263 points134 comments
Ywebmaven ·
The makers of Eleuther hope it will be an open source alternative to GPT-3
145 points95 comments
Ysohkamyung ·
Greater than 99% consensus on human caused climate change in the literature
137 points282 comments
Yskilled ·
X updates Terms to allow it to use data for AI training
92 points56 comments
YFanaHOVA ·
LLaMA2 isn't “Open Source” – and why it doesn't matter
28 points6 comments
Yoccamschainsaw ·
Just know stuff. (Or, how to achieve success in a machine learning PhD)
27 points3 comments
Yaresende ·
OpenSubtitles is not open anymore
17 points8 comments
Ycolesantiago ·
Google Announces Layoffs: A difficult decision to set us up for the future
9 points0 comments
Yeqmvii ·
Tell HN: ChatGPT can reply like a specific Reddit or HN user, including you
8 points1 comments
YJuliusSu ·
How good could GTP-4 get if copyright was not an issue?
6 points3 comments
Yijk ·
Scaling Data-Constrained Language Models
4 points0 comments
YTomte ·
Sarah Silverman is suing OpenAI and Meta for copyright infringement
4 points0 comments
Ytosh ·
The Pile: An 800GB Dataset of Diverse Text for Language Modeling
3 points1 comments
Yandyxor ·
The Pile: An 800GB Dataset of Diverse Text for Language Modeling
2 points1 comments
Ynigamanth ·
The Limits of GPT-3 and Similar Large Language Models
2 points0 comments
Ytomduncalf ·
David Guetta uses ChatGPT and uberduck.ai to create fake Eminem rap for DJ set
2 points0 comments
Ycharlysl ·
The Pile: a dataset for language modeling [pdf]
2 points0 comments
Yllmmadness ·
Even 'uncensored' models can't say what they want
2 points1 comments
Ysharemywin ·
Could an open source/data GPT like chatbot be created?
1 points0 comments
Ymfabbri77 ·
ChatGPT playing a not so simple game
1 points0 comments
Yasim ·
Ask HN: Can you crowdfund the compute for GPT?
1 points0 comments
Ytim_sw ·
RedPajama: Reproduction of Llama with Friendly License (Stanford, MILA Quebec)
1 points0 comments
Yskilled ·
What OpenAI Really Wants
1 points1 comments
Ybirriel ·
Phi-2: The surprising power of small language models
1 points0 comments