Update README.md

Update README.md

Add hackathon submission tags

Pre-submission cleanup: remove scratch files, fix internal comments

Fix resubmit/state-reset bug: fresh output on first click

Keep the hero spectrum reactive across pause/play cycles (+ seek-safe)

Give the processing overlay room to breathe (fix cramped 'composing' stage)

Audio-reactive hero: real FFT via shadow <audio>, + in-body JS bootstrap

Make the hero spectrum dance with the actual audio (Web Audio AnalyserNode)

Fix result-panel UI: accurate seam ribbon, full waveform, instant readout, intro replays

Add three flagship experiences: onboarding, seam reveal, live hero

Swap demo to Track0000 — my own unfinished song from 2016

Ship the bundled PUSHBACK demo clip (via Git LFS)

Fix demo load: copy bundled clip to a temp file for Gradio cache

Fix 'Try the demo' on the Space: allow serving the bundled clip

Make CODA a premium audio instrument with a 4-beat journey

Redesign UI: ethereal dark-studio theme for CODA

Fix resubmit UI bugs: persist length across runs, drop autoplay

Harden best-of-N selection: reject dropouts and squashed takes

Autoplay the finished song; remove temporary diagnostics

Add temporary ZeroGPU diagnostic instrumentation

Fix H200 fp16 instability: pin MATH SDPA, fp32 autoencoder, no TF32

Fix ZeroGPU audio quality and stuck-processing UI

Restore the verified-good whole-buffer normalization

Fix ZeroGPU continuation: bound the lead-in, normalize the tail, shrink the GPU window

Authenticate to the HF Hub with HF_TOKEN before loading gated SA3 weights

Revert to gradio SDK with runtime --no-deps SA3 install for ZeroGPU

Switch Space to Docker SDK to bypass stable-audio-tools torch pin

Fix torch pin for ZeroGPU: 2.7.1+cu128 → 2.8.0+cu124

Pin the continuation seed so the app reproduces the verified take

Rebuild CODA on Stable Audio 3 Small Music

Rebuild CODA around a working continuation engine; strip the scope creep

Mount the handoff player reliably: build it visible inside the hidden result group (gradio 6 drops visible=False audio components that a later update tries to show)

Trim editor-local entries out of the tracked gitignore

Tests track the window-safe planner; add energy-guard + re-roll coverage; pin transformers to the GPU-verified 4.57.6

Real music out: window-safe passes, stem-sum prompt, level anchor + session console

Fix: instant visual feedback when "Finish this song" is clicked

Remove mp3 from git, will upload via CLI

Switch from LFS to Xet for mp3

Track mp3 with LFS

PUSHBACK demo preset + lo-fi input enhancement (hears the song, not the noise)

The handoff, liner notes, share caption, and the craft story

UI elevation: live waveform painting, streaming lyrics, vinyl poster v2

Full pipeline overhaul: lyrics, poster, meter detection, staged UI

Pure audio continuation: drop text pull, add vocal gate + mastering pass

Max out audio context (20s prompt), equal-gain seam, generate() logging

Fix generic "8-bit" output: fp32 conditioning, 15s audio prompt, song-arc structure

Fix music continuation: chained 60s+ generation, native-rate stitching, seamless crossfade

Fix footer year: Build Small 2026

Fix torchcodec crash, gradio 6 launch args, and polish UI