Vlad Precup comments

Repositories
Issues
Comments

Results 3 comments of


                                            Vlad Precup

[New Model] DocFormer: End-to-End Transformer for Document Understanding

Hello there, @uakarsh. Has this initiative of integrating DocFormer into Transformers been discontinued in the meantime?

need to be able to set return data size for each request in batch transform

Depending on the serving tool you use, i.e. TorchServe (TS) / Multi-model Server, you can change the maximum request/response size by setting the TS_MAX_RESPONSE_SIZE / MMS_MAX_RESPONSE_SIZE / ... environment variables...

[feat] Distribute NLTK tokenizers used in the core package

NLTK also poses problems for the AWS Lambda runtime. Also, agree on the fact that the size of the tokenizer itself is quite large. This is roughly 15% of the...