Derek Hsu
Actually you can use llama3 by vllm backend now.
No, if you want to use a customized key phrase extractor, you can create your own `DocumentStore` and pass by `docstore` parameter.
> You have not given the whole data, but `context.data_details` requires there is another outer nested for-loop isn't it ? Original data is too large to give, so I ignore...
Thanks for the fix! Everything looks good now. I did notice that the text is still untranslated—will that also be addressed?
So there is no way to specified EasyOCR lang by CLI? I found the default langs supported are en, fr, de and es.
Nobody addresses this problem for mouths. Is this project died?