Metadata-Version: 2.4
Name: sorilm
Version: 0.0.0
Summary: Sori — an audio language model agent. Name reservation; the runtime is released with the model.
Author-email: Seonuk Kim <sukim@hi.snu.ac.kr>
Project-URL: Homepage, https://huggingface.co/snkii/Sori-1B
Keywords: audio,speech,music,audio-language-model,agent
Requires-Python: >=3.10
Description-Content-Type: text/markdown

# sorilm

Name reservation for **Sori**, an audio language model agent from the Human Interface Laboratory, Seoul National
University. Sori answers a question about a sound in natural language by planning and running a program in the
interpreter of its core model, **Sori-1B**.

This release contains no code. The runtime is published together with the model:
<https://huggingface.co/snkii/Sori-1B>
