MOSS-TTSD tested by Argusic Runner

What the test found

MOSS-TTSD v1.0 model loads from HuggingFace, processor encodes conversations, forward pass returns logits, and generate() produces audio codes and text tokens on CPU with bfloat16 precision

End-to-end verification with actual HuggingFace model weights (OpenMOSS-Team/MOSS-TTSD-v1.0). Model loads, forward pass returns logits/hidden_states, generate() produces audio codes and text tokens. All Python modules import cleanly. No test suite exists in the repo. Gradio demo and SGLang request scripts import but require runtime servers.

Argusic installed MOSS-TTSD in 6 minutes and launched it; verification reached Verified running for real. 2 of 2 recorded errors were worked through (see the timeline below).

Runsall runs of MOSS-TTSD

This summary is drawn from the agent's recorded report for this run. Every figure traces to the log and recording above; nothing here is authored.