whisper-medium-atcosim

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

• Loss: 0.0493
• Wer: 79.0563

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: 128
• eval_batch_size: 128
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 10
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.129	8.33	500	0.0419	2.3014
0.0004	16.67	1000	0.0428	17.1282
0.0002	25.0	1500	0.0426	10.4510
0.0012	33.33	2000	0.0427	39.6277
0.0003	41.67	2500	0.0423	58.2515
0.0	50.0	3000	0.0456	54.7324
0.0	58.33	3500	0.0468	61.8402
0.0	66.67	4000	0.0477	68.6794
0.0	75.0	4500	0.0483	72.7172
0.0	83.33	5000	0.0488	76.2502
0.0	91.67	5500	0.0491	78.2228
0.0	100.0	6000	0.0493	79.0563

Framework versions

• Transformers 4.36.2
• Pytorch 2.1.2
• Datasets 2.15.0
• Tokenizers 0.15.0