Eimhin03/outout_model_MCV_data_only_20000_steps

outout_model_MCV_data_only

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

• Loss: 3.3098
• Wer: 70.8861

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: 0.0001
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• optimizer: Use OptimizerNames.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: 200
• training_steps: 20000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.6635	2.5458	1250	2.3923	418.9873
0.2117	5.0916	2500	2.6428	153.4177
0.1714	7.6375	3750	2.8463	101.5190
0.0456	10.1833	5000	2.9152	76.9620
0.0413	12.7291	6250	3.1064	76.2025
0.0353	15.2749	7500	3.2051	79.7468
0.0127	17.8208	8750	3.4358	77.9747
0.0267	20.3666	10000	3.3111	74.1772
0.0043	22.9124	11250	3.3488	78.4810
0.0143	25.4582	12500	3.3206	73.1646
0.0096	28.0041	13750	3.3887	74.9367
0.0001	30.5499	15000	3.3206	72.6582
0.0000	33.0957	16250	3.3218	73.4177
0.0000	35.6415	17500	3.3122	71.3924
0.0000	38.1874	18750	3.3104	71.3924
0.0000	40.7332	20000	3.3098	70.8861

Framework versions

• Transformers 5.0.1.dev0
• Pytorch 2.9.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.2