assasinatee
/

SWEAR

Audio-Text-to-Text

qwen2_5_omni_thinker

image-text-to-text

audio-reasoning

Model card Files Files and versions

SWEAR

SWEAR model (merged full weights on Qwen2.5-Omni-7B) for SWEAR: Section-Wise Efficient Audio Reasoning via Self-Rated Difficulty.

This is the merged full-model release. No separate LoRA adapter is required.

The model was trained on the choice-only, no-S2TT subset of DiCoTA.

Paper: arXiv:2610.13081
Code: https://github.com/assasinatee/SWEAR
Dataset: https://huggingface.co/datasets/assasinatee/DiCoTA

Downloads last month: -

Safetensors

Model size

9B params

Tensor type

BF16

·

Inference Providers NEW

Audio-Text-to-Text

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for assasinatee/SWEAR

Base model

Qwen/Qwen2.5-Omni-7B

Finetuned

(73)

this model

Paper for assasinatee/SWEAR

SWEAR: Section-Wise Efficient Audio Reasoning via Self-Rated Difficulty

Paper • 2610.13081 • Published 4 days ago