Check_Model_2

This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on the common_voice dataset. It achieves the following results on the evaluation set:

• Loss: 0.3499
• Wer: 0.2729
• Cer: 0.0673

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: 0.0003
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
3.8708	3.23	400	0.7345	0.7259	0.2034
0.4247	6.45	800	0.4128	0.4268	0.1102
0.2047	9.68	1200	0.3726	0.3795	0.0930
0.1422	12.9	1600	0.3690	0.3514	0.0884
0.1139	16.13	2000	0.3811	0.3160	0.0794
0.089	19.35	2400	0.3650	0.2895	0.0731
0.0709	22.58	2800	0.3629	0.2944	0.0727
0.0594	25.81	3200	0.3538	0.2779	0.0692
0.0478	29.03	3600	0.3499	0.2729	0.0673

Framework versions

• Transformers 4.31.0
• Pytorch 2.0.1+cu117
• Datasets 1.18.3
• Tokenizers 0.13.3