output_model

This model is a fine-tuned version of facebook/wav2vec2-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.2884
• Wer: 0.4752

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 1000
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
2.8783	2.4752	500	2.7618	0.9999
1.4069	4.9505	1000	1.0853	0.6936
0.6264	7.4257	1500	0.9955	0.6014
0.3864	9.9010	2000	1.0460	0.5675
0.2714	12.3762	2500	0.9830	0.5422
0.2099	14.8515	3000	1.0333	0.5296
0.1615	17.3267	3500	1.1575	0.5203
0.1248	19.8020	4000	1.1311	0.4956
0.1032	22.2772	4500	1.3206	0.4953
0.0834	24.7525	5000	1.2094	0.4855
0.0655	27.2277	5500	1.2966	0.4763
0.052	29.7030	6000	1.2884	0.4752

Framework versions

• Transformers 4.40.0
• Pytorch 2.2.1+cu121
• Datasets 2.19.0
• Tokenizers 0.19.1