Bound network-request endpoint chunks

Stabilize long-form semantic and speaker boundaries

Stabilize network speech generation context

Harden clean atomic inference recovery

Naturalize reversible short-path URLs

Harden network endpoint and local verification

Harden clean speaker inference release

Harden hybrid network inference recovery

Harden network generation and row-local recovery

Harden production inference normalization and boundaries

Harden fail-closed inference selection

Add coverage-adaptive chunk refill

Bound active pace correction to gate margin

Remove artificial onset chunk splits

Add dual-ASR whole-output verification

Stabilize verified sequence fallback

Harden hosted inference verification

Add fail-closed stable speaker inference

Fix Space dependency resolution

Pin verified model and speaker encoder revisions

Catch weak stop signals near sentence end

Stabilize very short utterances

Prevent speech after the requested text

Stop cleanly at natural text boundaries

Stop earlier without clipping final speech

Stop forcing generated tail speech

Tighten onset pacing and endpoint cap

Tighten endpoint stop window

Stabilize onset pace and endpoint stopping

Use stable production inference profile

Add speech-rate slider (pitch-preserving time-stretch post-processing, shared across tabs)

Add adjustable cfg_value / inference_timesteps sliders (shared across tabs, defaults = model-card recommendation)

Voice clone: single reference-clip mode only (drop centroid/ECAPA stable mode + speechbrain)

Update for checkpoint step_0006000

Recreate Space (zero-a10g) after region corruption; streaming included

initial commit