NAVER LABS Europe Submission to the Instruction-following Track
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915318334488576 |
|---|---|
| author | Lee, Beomseok Boito, Marcely Zanon Besacier, Laurent Calapodescu, Ioan |
| author_facet | Lee, Beomseok Boito, Marcely Zanon Besacier, Laurent Calapodescu, Ioan |
| contents | In this paper we describe NAVER LABS Europe submission to the instruction-following speech processing short track at IWSLT 2025. We participate in the constrained settings, developing systems that can simultaneously perform ASR, ST, and SQA tasks from English speech input into the following target languages: Chinese, Italian, and German. Our solution leverages two pretrained modules: (1) a speech-to-LLM embedding projector trained using representations from the SeamlessM4T-v2-large speech encoder; and (2) LoRA adapters trained on text data on top of a Llama-3.1-8B-Instruct. These modules are jointly loaded and further instruction-tuned for 1K steps on multilingual and multimodal data to form our final system submitted for evaluation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_01808 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | NAVER LABS Europe Submission to the Instruction-following Track Lee, Beomseok Boito, Marcely Zanon Besacier, Laurent Calapodescu, Ioan Computation and Language In this paper we describe NAVER LABS Europe submission to the instruction-following speech processing short track at IWSLT 2025. We participate in the constrained settings, developing systems that can simultaneously perform ASR, ST, and SQA tasks from English speech input into the following target languages: Chinese, Italian, and German. Our solution leverages two pretrained modules: (1) a speech-to-LLM embedding projector trained using representations from the SeamlessM4T-v2-large speech encoder; and (2) LoRA adapters trained on text data on top of a Llama-3.1-8B-Instruct. These modules are jointly loaded and further instruction-tuned for 1K steps on multilingual and multimodal data to form our final system submitted for evaluation. |
| title | NAVER LABS Europe Submission to the Instruction-following Track |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2506.01808 |