———– blog ———–
Open AIのGPT-2アップデート前に試しましたが、びっくりするほど凄かった。そして、GoogleもBERT(Bidirectional Encoder Representations from Transformers)が日本語を含む世界約70言語にも対応したとのことで。
アップデートBERTは試しておりませんが、GPT-2は本当にびっくりしました。ちなみに試したGPT-2もアップデート前のものです。
ご興味がございましたら、Open AIのGPT-2やGoogleのBERT関係の公式リンクなどをざっくりとですが置いておきます。但しも・・・。
Yesterday I launched Talk to Transformer, a site where you can try out @OpenAI's new text-generating language model on your own custom text. https://t.co/ncOL1QSRyS One example is below. @AlecRad @gdb pic.twitter.com/mcW6sipp3S
— Adam King (@AdamDanielKing) 2019年5月7日
We've trained an unsupervised language model that can generate coherent paragraphs and perform rudimentary reading comprehension, machine translation, question answering, and summarization — all without task-specific training: https://t.co/sY30aQM7hU pic.twitter.com/360bGgoea3
— OpenAI (@OpenAI) 2019年2月14日
We've fine-tuned GPT-2 using human feedback for tasks such as summarizing articles, matching the preferences of human labelers (if not always our own). We're hoping this brings safety methods closer to machines learning values by talking with humans. https://t.co/ok9jeMP5zj
— OpenAI (@OpenAI) 2019年9月19日
Hebrew, Hindi, Hungarian, Icelandic, Indonesian, Italian, Japanese, Javanese, Kannada, Kazakh, Khmer, Korean, Kurdish, Kyrgyz, Lao, Latvian, Lithuanian, Macedonian
Malay (Brunei Darussalam & Malaysia), Malayalam, Maltese, Marathi, Mongolian, Nepali, Norwegian, Polish…. (MORE)— Google SearchLiaison (@searchliaison) 2019年12月9日
And yes, Spanish!
— Google SearchLiaison (@searchliaison) 2019年12月9日
BERT is rolling out for: Afrikaans, Albanian, Amharic, Arabic, Armenian, Azeri, Basque, Belarusian, Bulgarian, Catalan, Chinese (Simplified & Taiwan), Croatian, Czech, Danish, Dutch, English, Estonian, Farsi, Finnish, French, Galician, Georgian, German, Greek, Gujarati…. (MORE)
— Google SearchLiaison (@searchliaison) 2019年12月9日
The non-profit/for-profit/investor partnership is held together by a set of legal documents that are entirely novel (=bad term in legal docs), are non-public + unclear, have no case precedence, yet promise to wed operation to a vague (and already re-interpreted) OpenAI Charter.
— Smerity (@Smerity) July 22, 2019
Many of their blog posts and spoken positions end up influencing govt policy and public opinion on the future of AI through amplified pseudo-credibility due to "Open", "Musk founded", repeatedly hyped statements, and a sheen from their now distant non-profit good will era.
— Smerity (@Smerity) July 22, 2019
"We'll clarify our position in an upcoming blog post."
— Smerity (@Smerity) July 22, 2019
github.com/openai/gpt-2
talktotransformer.com
blog.google/products/search/search-language-understanding-bert
github.com/google-research/bert
arxiv.org/abs/1810.04805