hf-space: drop binary assets — HF requires Xet for non-text files

Push to huggingface.co/spaces/Tushar9802/sakhi was rejected because
workstation_demo.gif + workstation_demo.mp4 (added in 42295ef two days
ago but never pushed to hf) trip HF Hub's binary-files-must-go-through-Xet
rule. Removing them so the Space can pick up today's docs / deps / frontend
commits. The GIF embed is removed from this branch's README to avoid a
broken image on the Space landing page.

GitHub (origin/master) keeps the GIF — that repo allows binaries
inline up to 100 MB. The 3-minute YouTube demo link in the top callouts
covers the same visual on both surfaces.

Re-adding the GIF to hf-main via Xet/LFS is a post-submission cleanup
if the Space landing page wants it back.

README.md

@@ -23,8 +23,6 @@ Offline-first tool that converts Hindi home visit conversations into structured
 
 **▶ Pull the Unsloth fine-tune:** [`tusharbrisingr9802/sakhi`](https://ollama.com/tusharbrisingr9802/sakhi) on the Ollama registry — `ollama pull tusharbrisingr9802/sakhi` fetches the LoRA-fine-tuned Gemma 4 E4B behind the A/B numbers below. The base model (`gemma4:e4b-it-q4_K_M`) is what ships in the live pipeline; this is the side-by-side comparison artifact for the Unsloth track.
 
-![Workstation demo: Hindi audio → form + danger signs (30 s)](workstation_demo.gif)
-
 ## Problem
 
 India's ASHA workers conduct 50M+ maternal/child health home visits per year across rural areas. Every visit ends with paper forms filled from memory, then physically carried to the Primary Health Center. Danger signs observed in the field — preeclampsia, postpartum hemorrhage, neonatal distress — often never reach the system in time for intervention.