Kushtrim/whisper-base-sq

OpenAI Whisper Tiny for Albanian

Model Description

OpenAI Whisper Base for Albanian is a specialized automatic speech recognition (ASR) model. It is a fine-tuned version of the base OpenAI Whisper Base model, trained specifically on the Mozilla Common Voice 14 dataset. The primary objective of this model is to transcribe spoken Albanian language into text.

Training Data

The OpenAI Whisper Base for Albanian model is fine-tuned on a small-scale dataset from Mozilla Common Voice 14. While the dataset offers a diverse collection of Albanian language audio recordings and corresponding transcriptions, it's important to note that the model's overall quality is impacted by the limited size of the training data (~1 hour).

Authors

The base OpenAI Whisper Tiny model is developed by the team at OpenAI, and the fine-tuning on the Albanian dataset for this specialized version is performed by Kushtrim Visoka.

Citation

If you use this model, please consider citing this repository.