xls-r-1b-bem-natbed-combined-model-with-adapet

This model is a fine-tuned version of facebook/wav2vec2-xls-r-1b on the NATBED - BEM dataset. It achieves the following results on the evaluation set:

• Loss: 0.2050
• Wer: 1.0001

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: 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: 100
• num_epochs: 30.0
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.2077	0.2503	100	0.5981	1.0
0.2576	0.5006	200	0.2140	1.0
0.2238	0.7509	300	0.2080	1.0
0.1957	1.0013	400	0.2105	1.0
0.2047	1.2516	500	0.2102	1.0
0.1966	1.5019	600	0.2091	1.0
0.2367	1.7522	700	0.2085	1.0
0.183	2.0025	800	0.2090	1.0186
0.2109	2.2528	900	0.2044	1.0
0.2088	2.5031	1000	0.2050	1.0001
0.1981	2.7534	1100	0.2032	1.0002
0.1897	3.0038	1200	0.2064	1.0
0.1997	3.2541	1300	0.2072	1.0
0.2235	3.5044	1400	0.2062	1.3064

Framework versions

• Transformers 4.46.0.dev0
• Pytorch 2.4.1+cu121
• Datasets 3.0.1
• Tokenizers 0.20.0