RT @AkiraTOSEI: arxiv.org/abs/2104.08691 Instead of fine-tuning the parameters of the learned model, the task is performed by tuning the input prompting method. It is possible to reach the same level as fine-tuning, with five orders of magnitude fewer tuning parameters. pic.twitter.com/7tKdAipmTo
posted at 01:44:55
torchtextのバージョンが上がったせいで、import torchtext.legacy as torchtextと回避しないといけないが
posted at 01:40:21
この辺をみんなに試してもらうとよいかな / 【実装解説】日本語版BERTでlivedoorニュース分類：Google Colaboratoryで（PyTorch） qiita.com/sugulu_Ogawa_I… #Qiita
posted at 01:25:29