---

license: mit language:

• en pipeline_tag: automatic-speech-recognition

---

About

This model was created to support experiments for evaluating phonetic transcription with the Buckeye corpus as part of https://github.com/ginic/multipa. This is a version of facebook/wav2vec2-large-xlsr-53 fine tuned on a specific subset of the Buckeye corpus. For details about specific model parameters, please view the config.json here or training scripts in the scripts/buckeye_experiments folder of the GitHub repository.

Experiment Details

The entire train split of the Buckeye corpus was used to train this model. The only data excluded are samples in the train split that are too short (< 0.1 seconds) or too long (>12 seconds) to be used to train the model

Goals:

• Include the largest amount of training data possible.
• Can be used with a different corpus (e.g. TIMIT, Speech Accent Archive) for evaluation to test generalization to other dialects and language varieties.