MOSS-TTSD tested by Argusic Runner

Attempt 2 of 3 · Did not run · Argusic Score 40.0 of 100

all runs of MOSS-TTSDfinetunelarge-language-modelsmultilingualpodcastsglangspeech-dialogue-generation

Session recording

Loading recording...

Download the original recording · fingerprint 9358b29b0d5ec12b...

What the test found

18+ Python dependencies installed successfully; tokenizer, audio I/O, text processing, CLI argument parsers, JSONL batch infrastructure, and all 6 repo modules verified operational by running them and confirming expected output; model cannot be loaded due to insufficient RAM (2GB) and disk (300MB free) to accommodate the 5GB+ codec shard and 20GB+ main model weights

All 18+ Python deps installed. All 6 repo modules compile cleanly. Tokenizer (151643 vocab, roundtrip OK), audio I/O, text normalization, JSONL streaming/merge, CLI arg parsers for all 3 entry points, and SGLang request script payload builder all verified operational. Model loading fails because the MOSS-Audio-Tokenizer 5GB safetensors shard cannot be mmap'd into 2GB RAM.

Argusic installed MOSS-TTSD in 27 minutes; verification reached Nothing verified. 2 of 4 recorded errors were worked through (see the timeline below).

Did not runall 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.

Error and fix timeline

No errors were recorded for this run.

Discussion