textQA

Training/Evaluation scripts for pipeline baseline in SLUE QA benchmark proposed in the paper SLUE Phase-2: A Benchmark Suite of Diverse Spoken Language Understanding Tasks

*** Note: these instructions are a work in progress and will be updated over the next few days

Fine-tuning

Fine-tune the text QA model with the groundtruth transcripts of the SLUE-SQA-5 fine-tune set

 sh train.sh

Evaluate the text QA model with the groundtruth transcripts of the SLUE-SQA-5 dev set

sh eval_groundtruth.sh

Name		Name	Last commit message	Last commit date
Latest commit History 6 Commits
.gitignore		.gitignore
README.md		README.md
dataset.py		dataset.py
eval.py		eval.py
eval_asr.sh		eval_asr.sh
eval_groundtruth.sh		eval_groundtruth.sh
scheduler.py		scheduler.py
train.py		train.py
train.sh		train.sh
utils.py		utils.py