BERT Java-8k
#1
by
akmalkadi
- opened
Greetings,
I have some questions, and I would really appreciate your answers to understand the paper clearly.
1 - Are you planning to add (BERT Java-8k, BERT Java-16k, and BERT Java-32k) to Hugging Face?
2 - Are JavaBERT-uncased and JavaBERT-cased fine-tuned versions of BERT-uncased and BERT-cased?
Hey,
I am not planning to upload the Java-specific tokenizers to this hub, as they do not perform as good as expected. Therefore, I don't see any use for them. Both models are built from the ground up and do not make use of other pretrained models.
neltds
changed discussion status to
closed