Skip to content

duplexjev 0.2.1 + released weights

Choose a tag to compare

View all tags
@BKtensor BKtensor released this 24 Sep 12:54
· 76 commits to main since this release
v0.2.1
97f07a3

Weights (Hugging Face, adventists-ai): 7 connectors for Qwen3-ASR-0.6B → Qwen3-32B (A: cross-attention fusion, B: last layer; content, gender, emotion and the mixed-objective A-Para) and 2 encoder repositories. Emotion connectors are CC BY-NC 4.0 (ESD is research-only); the rest Apache-2.0.

duplexjev 0.2.1 (PyPI):

  • Audio is padded to whole audio tokens at the end instead of the start. Start padding shifted every stack group against the frames the connector was trained on and cost up to 6 points on emotion.
  • text_model / audio_model overrides are applied before the config is built, so local copies work offline.

Through the package with default prompts, every released connector is within 0–5 points of the paper's numbers (see the model cards).

Assets 2
Loading