Whisper-WOLOF-5-hours-Google-Fleurs-Alffa-dataset

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

• Loss: 1.7146
• Wer: 49.5987
• Cer: 18.5026

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
• train_batch_size: 32
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
1.1824	5.8140	500	1.1937	50.6952	19.3233
0.0707	11.6279	1000	1.4496	50.4465	18.8349
0.006	17.4419	1500	1.5426	49.6100	19.6467
0.0014	23.2558	2000	1.6095	49.2596	18.9018
0.0008	29.0698	2500	1.6513	49.5309	19.6824
0.0005	34.8837	3000	1.6787	49.9265	19.4482
0.0004	40.6977	3500	1.7026	49.3840	19.0646
0.0003	46.5116	4000	1.7146	49.5987	18.5026

Framework versions

• Transformers 4.45.2
• Pytorch 2.1.0+cu118
• Datasets 3.0.1
• Tokenizers 0.20.1