Translingo: Instant sentence translator

background
workflow of the application
User interface of webapp 1
User interface of webapp 2

Inspiration

While travelling to different place, it become hard to communicate with others because not everyone knows English or the language we are speaking.

What it does

Translingo: instant sentence translator, is a web application that can translate sentence of any language to english.

How we built it

This application is build using clarifai and langchain. Firstly we used Llama 2 model available on langchain's integration of clarifai for getting the translations. We defined specific prompts for getting the output. After that, we used pyttxs3 for voice integration. and finally for building the ui, streamlit is used and then deployed the final project on stramlit cloud.

Challenges we ran into

Complete voice integration for every language. So far voice integration is only done for english language. Finding the langchain integration of llama2 model available on clarifai. Experimenting different prompts to get the desired outputs.

Accomplishments that we're proud of

Complete deployment on streamlit cloud. Successfully integrated the llama2 model. Designed a prompt to get the desired output.

What we learned

How to integrate clarifai models using langchain. How to give prompts using PromptTemplate available on langchain. How to integrate voice using Python text to speech.