search

OpenPecha / Requests

RFWs and RFCs for all OpenPecha repositories
0 stars 0 forks source link

[RFW0104] Output format of Docx need to be same as input incase of mt document upload. #349

Open gangagyatso4364 opened 9 months ago

gangagyatso4364 commented 9 months ago

RFW0104: Output format of Docx need to be same as input in case of mt document upload.

Summary

Currently the model does not have a standard output for large document that are uploaded to the machine for translation. we need a standardised output format for documents that are big for model to process. the standard output will be used for all models except for the text to speech model.

Key Concepts

Context

models:

  1. Translation: a model that translate english to tibetan and vice versa
  2. text to speech: a model that translate tibetan text into audio.
  3. speech to text: a model that translates tibetan speech to text.
  4. ocr model: a model that converts a image into a text.

Outputs

the translation format should be like the picture below. Screenshot from 2023-12-25 11-15-15

Inputs

A large document (docx, txt) The translation machine models in Monlam AI website. Monlam AI

Timeline

Specify the expected delivery date for the project.

References

Include any relevant links or resources for additional context or information.

TenzinGayche commented 9 months ago

How can we make one format that works for both audio and text files? 🤔 It looks like we might need two different formats for each

ngawangtrinley commented 9 months ago

I'm confused... Is this RFW to define a standard or is it to add a document processing capability? If you're talking about adding document translation support, pdf OCR etc then this should be several RFWs