build-small-hackathon/glossolalia

readme: link the public inputs dataset

revert to working model load (first-call verified)

zerogpu: load model on cpu in parent, move to gpu per call (fixes consecutive-call crash)

readme: demo + social links, data rationale; revert to working model load; ghost dial 3/4 distinct

zerogpu: eager model load fixes consecutive-call crash; ghost dial 3/4 distinct

README: validation evidence describes the reproducible graded slide, not run-specific transcription strings

README audit fixes: real scalar-conditioner architecture, kill em dashes, correct voice labels, honest validation evidence

submission README: tags (track:wood + badges) + accurate writeup

voice clone: transcode upload to 24k wav + auto-transcribe blank ref

fix: @gpu_task back on speak/morph (was mis-attached to DSP helper); dry cache as path not array

ZeroGPU: load F5-TTS on CPU, move to GPU inside @spaces.GPU (proper ZeroGPU pattern)

ZeroGPU: eager-load F5-TTS at module init (fix Low-level CUDA init error)

post-fx applies live: cache dry voice, re-render on dropdown change (no re-inference)

org: minimal packages (ffmpeg + libsndfile1)

org: drop torchcodec pins (bypassed via soundfile patch)

soundfile load patch (bypass torchcodec)

soundfile load patch (bypass torchcodec)

org: add FFmpeg dev shared libs so torchcodec core libs load

ZeroGPU: plain torch 2.9.1 + torchcodec 0.9.1+cu128 only (CUDA 12.8 match)

ZeroGPU: pin full torch triple to +cu128 builds (torchcodec 0.9.1+cu128 links libnvrtc.so.12)

ZeroGPU: pull torch stack from cu128 index so torchcodec CUDA libs match

ZeroGPU stack: torch 2.9.1 + spaces

guarded @spaces.GPU decorator (no-op off ZeroGPU)

mirror data/voices/v9.txt

mirror data/voices/v8.txt

mirror data/voices/v7.txt

mirror data/voices/v6.txt

mirror data/voices/v5.txt

mirror data/voices/v4.txt

mirror data/voices/v3.txt

mirror data/voices/v2.txt

mirror data/voices/v1.txt

mirror data/voices/v9.wav

mirror data/voices/v8.wav

mirror data/voices/v7.wav

mirror data/voices/v6.wav

mirror data/voices/v5.wav

mirror data/voices/v4.wav

mirror data/voices/v3.wav

mirror data/voices/v2.wav

mirror data/voices/v1.wav

mirror static/knob.js

mirror static/style.css

mirror static/index.html

mirror data/phoneme_lm.npz

mirror data/cmudict.dict

mirror scripts/post_fx.py

mirror scripts/corrupt_phonemes.py

mirror scripts/mondegreen.py

mirror scripts/__init__.py