whisper-a-nomimose-again

This model is a fine-tuned version of openai/whisper-small on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0252
• Wer: 192.9941

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: 0.0004
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 132
• num_epochs: 9
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.2606	0.9217	100	0.2230	39.7493
0.2758	1.8387	200	0.1259	41.2979
0.0716	2.7558	300	0.0491	24.1888
0.0433	3.6728	400	0.0618	25.7375
0.0339	4.5899	500	0.0383	184.0708
0.0211	5.5069	600	0.0418	139.3068
0.0154	6.4240	700	0.0424	195.4277
0.0095	7.3410	800	0.0251	186.2832
0.0059	8.2581	900	0.0252	192.9941

Framework versions

• Transformers 4.47.0.dev0
• Pytorch 2.4.0
• Datasets 3.0.1
• Tokenizers 0.20.0