xls-r-asr_xh-run7

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

• Wer (Validation): 55.13%
• Wer (Test): 52.11%

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0003
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• 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 (Train)
6.1294	1.59	250	3.0258	1.0
1.8726	3.18	500	0.7953	0.8526
0.6065	4.78	750	0.5099	0.6676
0.3769	6.37	1000	0.4647	0.6016
0.2652	7.96	1250	0.4966	0.5967
0.1992	9.55	1500	0.4320	0.5507
0.1573	11.15	1750	0.4762	0.5475
0.1252	12.74	2000	0.5147	0.5157

Framework versions

• Transformers 4.28.0
• Pytorch 2.0.1+cu117
• Datasets 2.14.4
• Tokenizers 0.13.3