feat: improve model speed

chore: revert changes

feat: improve model speed

fix: remove dumb rules

feat: improve model by using two models and adding rules for corrections

feat: increase the beam size for stricter translations

fix: restore batch translation

feat: change model to a domain focused model

fix: improve model quality

fix: remove/disable queue

fix: remove fastApi completely

fix: remove versions of packages

feat: integrate fast api

fix: revert changes

fix: revert changes

fix: revert changes

fix: remove docker file

feat: initial Hugging Face Spaces translation app

feat: switch to Gradio SDK with NLLB-200 translation model

fix: pin gradio 5.x to fix audioop/Python 3.13 incompatibility

feat: switch to Gradio SDK with NLLB-200 translation model

fix(space): add HF token auth, reduce compute, stabilize model load

fix(space): add HF token auth, reduce compute, stabilize model load

fix(space): add HF token auth, reduce compute, stabilize model load

fix(space): add HF token auth, reduce compute, stabilize model load

fix(space): add HF token auth, reduce compute, stabilize model load

fix(space): add HF token auth, reduce compute, stabilize model load

fix(gradio): eliminate duplicate model loads and stabilize startup

fix(gradio): disable SSR and harden translation handlers

feat: enable quering on space

feat: enable api call in space

feat: implement NLLB translation app with Gradio

refactor: offload translation to Hugging Face Space and add robust API handling

feat: implement NLLB translation app with Gradio

feat: enable batch translation

Merge branch 'main' of https://huggingface.co/spaces/vinjex/translation-ml

feat: initial Hugging Face Spaces translation app

initial commit