Expand demo 1 description (real-time streaming, barge-in)

Fix demo video playback across browsers/devices (faststart remux, MP4 brand fix, stable CDN embed with fallback)

Update Speech Generation (TTS) metrics to final report: add CER row, revise UTMOS/STOI, rename TTS->Speech Generation

Update technical report

Update both demo videos (health_demo, vad_demo)

Encoder key feature: BEST-RQ -> Random Quantization

Reword decoder key feature (Efficient Streaming Speech Decoder)

Update technical report (v2026-07-28, 31p); remove internal assets/README.md

Replace demo 2 video (vad_demo.mp4) with updated version

Replace demo 1 video (health_demo.mp4) with updated version

Upload vad_demo.mp4

Rename figure caption to 'A.X K2 ALM architecture'

Simplify 'dual-head VAD' to 'VAD'

Remove unused demo poster JPGs from repo (used only as wiki attachments)

Shorten Forgetting-Free and VAD key-feature descriptions

Update architecture figure (detailed, white bg); add encoder & decoder key features

Loudness-match demo videos to -16 LUFS; shorten architecture caption; back-channel example 네→아

Refine model card: tighten summary, merge table cells (HTML), 2 demos + VAD video, drop Limitations/footnote, plan-to-release note

Sync model card to latest: fix eval/spec tables to report, drop Architecture Highlights, Korean-only, relabel demos, add first demo video

Update README.md

initial commit