Translate one segment at a time for the local CPU backend

Shrink translation batch size so results appear incrementally during Translate All

Link local model name to its HF repo, unbold "local CPU model", clarify wording

Add usage explanation to top of app, covering local vs OpenRouter model choice

Switch translation backend from Gemini to OpenRouter, add test suite

Enable ZeroGPU on local backend, remove deprecated Docker files

Merge remote-tracking branch 'space/main'

Configure Space for gradio SDK (ZeroGPU) and add academic translation prompt

initial commit

Initial simplified TranslationUI: plain-text upload/download, single Buddhist translation prompt, Gemini or local CPU model