P2Peek 0.8.7 — Model credits and license information

P2Peek · published

  • Added model credits, project links and bundled license files to About.
  • Target speaker extraction uses a model developed by Ke Zhang and released through the REAL-TSE WeSep project. REAL-TSE WeSep
  • P2Peek converted the model to ONNX and adapted it for local processing. The original authors and projects do not endorse P2Peek.
  • The REAL-TSE author's portions are provided under Apache-2.0. The license for the pretrained WeSpeaker ECAPA weights is being clarified with upstream.
P2Peek 0.8.7 — Model credits and license information · P2Peek patch notes · SteamPeaks