hubert-base-ls960-nmsqa-asr

Model Card for Model ID

<!-- Provide a quick summary of what the model is/does. -->

This model was pretrained using facebook/hubert-base-ls960 model on NMSQA dataset. The task is Automatic Speech Recognition (ASR) in which the questions and context sentences are used.
This is a checkpoint with WER 14.36 on dev set.

Model Details

Model Description

<!-- Provide a longer summary of what this model is. -->

The input of the models are from NMSQA dataset. The task of the dataset is Spoken QA, but in this model I used the sentences for ASR. The input audios are both from context and questions. This ASR model was trained on using training and dev set of NMSQA.

• Developed by: Merve Menevse
• Model type: Supervised ML
• Language(s) (NLP): English
• Finetuned from model [optional]: facebook/wav2vec2-base-960h

Uses

<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->

The model should be used as fine-tuned model for wav2vec2.

How to Get Started with the Model

 from transformers import AutoModel

 model = AutoModel.from_pretrained("menevsem/hubert-base-ls960-nmsqa-asr")

Training Details

Training Data

<!-- This should link to a Data Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->

The model was trained using voidful/NMSQA train and dev set.

Evaluation

For evalaution WER metric is used on dev set.