wav2vec2-large-xls-r-300m-tr

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: 0.4074
• Wer: 0.4227

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
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
3.9399	4.21	400	0.7252	0.7387
0.4147	8.42	800	0.4693	0.5201
0.1855	12.63	1200	0.4584	0.4848
0.1256	16.84	1600	0.4464	0.4708
0.0948	21.05	2000	0.4261	0.4389
0.0714	25.26	2400	0.4331	0.4349
0.0532	29.47	2800	0.4074	0.4227

Framework versions

• Transformers 4.11.3
• Pytorch 1.10.0+cu111
• Datasets 1.18.3
• Tokenizers 0.10.3