faiq-wav2vec2-large-xlsr-indo-demo-rtx4090-batch32

This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the common_voice_11_0 dataset. It achieves the following results on the evaluation set:

• Loss: 0.3705
• Wer: 0.4296

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: 32
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
4.559	5.84	400	2.8099	1.0
0.9265	11.68	800	0.3847	0.5260
0.2592	17.52	1200	0.3658	0.4685
0.176	23.36	1600	0.3862	0.4422
0.1451	29.2	2000	0.3705	0.4296

Framework versions

• Transformers 4.28.0
• Pytorch 2.0.1
• Datasets 2.6.1
• Tokenizers 0.13.3