From 2d8340a91fe930b394902695c22f9dd892c309bd Mon Sep 17 00:00:00 2001 From: Patrick von Platen Date: Sat, 2 May 2020 10:25:22 +0200 Subject: [PATCH] [Reformer] Move model card to google model (#4113) * correct model card * remove model card from patrick von platen --- .../reformer-crime-and-punishment}/README.md | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) rename model_cards/{patrickvonplaten/reformer-crime-and-punish => google/reformer-crime-and-punishment}/README.md (80%) diff --git a/model_cards/patrickvonplaten/reformer-crime-and-punish/README.md b/model_cards/google/reformer-crime-and-punishment/README.md similarity index 80% rename from model_cards/patrickvonplaten/reformer-crime-and-punish/README.md rename to model_cards/google/reformer-crime-and-punishment/README.md index 2b4603b54c..0886727834 100644 --- a/model_cards/patrickvonplaten/reformer-crime-and-punish/README.md +++ b/model_cards/google/reformer-crime-and-punishment/README.md @@ -3,9 +3,8 @@ Crime and Punishment text was taken from `gs://trax-ml/reformer/crime-and-punish Model was trained in flax using colab notebook proposed by authors: https://colab.research.google.com/github/google/trax/blob/master/trax/models/reformer/text_generation.ipynb Weights were converted to Hugging Face PyTorch `ReformerModelWithLMHead`. -Model is used as a proof of concept that the forward pass works for a `ReformerModelWithLMHead`. -Given that the model was trained only for 30mins on a ~0.5M tokens dataset and has only 320 tokens, -the generation results are reasonable: +The model is a character level language model. Language from in `crime-and-punishment` style +can be generated as follows: ```python model = ReformerModelWithLMHead.from_pretrained("patrickvonplaten/reformer-crime-and-punish")