rifailabs/asr-shikomori-swahili

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

asr-shikomori-swahili

This model is a fine-tuned version of openai/whisper-small on an unknown dataset. It achieves the following results on the evaluation set:

• —eval_loss: 0.8698
• —eval_wer: 39.5001
• —eval_cer: 13.7566
• —eval_runtime: 511.342
• —evalsamplesper_second: 1.862
• —evalstepsper_second: 0.233
• —epoch: 14.7059
• —step: 3500

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• —learning_rate: 1e-05
• —trainbatchsize: 16
• —evalbatchsize: 8
• —seed: 42
• —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• —lrschedulertype: linear
• —lrschedulerwarmup_steps: 500
• —training_steps: 8000
• —mixedprecisiontraining: Native AMP

Framework versions

• —Transformers 4.42.0.dev0
• —Pytorch 2.1.2
• —Datasets 2.18.0
• —Tokenizers 0.19.1