Rfank2021

Cache parsed expression

6

We tested that cache ast parsing result reduced more than 50% run time for our user case, it would be great if it can be merged. The cache is an...

Current HuggingFaceTokenizer can only read tokenizer.json and can only configure a little, there are additional configs in tokenizer_config.json that is not used by HuggingFaceTokenizer. For example, https://huggingface.co/sentence-transformers/stsb-roberta-base-v2/tree/main has a tokenizer_config.json,...