base_sami_22k_ftallpseudo_ftlabelled_sami_parliament_fd0

This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 206.2647
• Wer: 0.3949
• Cer: 0.1327

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: 0.0005
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.25
• num_epochs: 60.0
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
889.4263	1.0	446	207.5438	0.3984	0.1279
736.1366	2.0	892	206.1312	0.3966	0.1333
712.5319	3.0	1338	247.7773	0.4225	0.1328
744.1979	4.0	1784	226.4229	0.4165	0.1384
714.2005	5.0	2230	262.0536	0.4904	0.1599
763.4344	6.0	2676	304.6938	0.4645	0.1596
785.2655	7.0	3122	297.9980	0.4811	0.1673

Framework versions

• Transformers 4.48.3
• Pytorch 2.5.1
• Datasets 3.2.0
• Tokenizers 0.21.0