From aa6c3c14b4ff8fbd5d40b100f4aae71fb359d6ae Mon Sep 17 00:00:00 2001 From: Ahmed Elnaggar Date: Tue, 6 Oct 2020 15:32:52 +0200 Subject: [PATCH] typo fix (#7611) It should be T5-3B not T5-3M. --- model_cards/Rostlab/prot_t5_xl_bfd/README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/model_cards/Rostlab/prot_t5_xl_bfd/README.md b/model_cards/Rostlab/prot_t5_xl_bfd/README.md index b6e4100ec..418d4c32d 100644 --- a/model_cards/Rostlab/prot_t5_xl_bfd/README.md +++ b/model_cards/Rostlab/prot_t5_xl_bfd/README.md @@ -20,7 +20,7 @@ This means it was pretrained on the raw protein sequences only, with no humans l publicly available data) with an automatic process to generate inputs and labels from those protein sequences. One important difference between this T5 model and the original T5 version is the denosing objective. -The original T5-3M model was pretrained using a span denosing objective, while this model was pre-trained with a Bart-like MLM denosing objective. +The original T5-3B model was pretrained using a span denosing objective, while this model was pre-trained with a Bart-like MLM denosing objective. The masking probability is consistent with the original T5 training by randomly masking 15% of the amino acids in the input. It has been shown that the features extracted from this self-supervised model (LM-embeddings) captured important biophysical properties governing protein shape.