metadata
base_model: openbmb/VoxCPM1.5
library_name: voxcpm
pipeline_tag: text-to-speech
tags:
- lora
- voxcpm
- text-to-speech
A.M.N. VoxCPM1.5 LoRA
Single-speaker LoRA adapter selected from checkpoint step 90 after human listening on 110 curated A.M.N. voice clips.
- Base model:
openbmb/VoxCPM1.5 - LoRA: LM + DiT, rank 16, alpha 32
- Native output: 44.1 kHz
- Intended use: the AMNESIAC fictional interrogator voice
Synthetic speech should be disclosed where appropriate. Do not use this adapter for impersonation, fraud, or disinformation.