mms-1b-all-bemgen-combined-gdro

This model is a fine-tuned version of facebook/mms-1b-all on the BEMGEN - BEM dataset. It achieves the following results on the evaluation set:

• Loss: 17.8045
• Wer: 0.5344

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: 4
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• 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_steps: 100
• num_epochs: 5.0
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
651.3737	0.5076	100	199.4144	1.0
200.6862	1.0152	200	66.1207	0.9503
134.4785	1.5228	300	23.4666	0.5956
111.6127	2.0305	400	19.1133	0.5569
110.8015	2.5381	500	18.9017	0.5372
102.4243	3.0457	600	18.3297	0.5449
103.9713	3.5533	700	18.8923	0.5445
102.1721	4.0609	800	17.8276	0.5343
102.6848	4.5685	900	17.8989	0.5297

Framework versions

• Transformers 4.52.4
• Pytorch 2.9.0+cu128
• Datasets 4.4.1
• Tokenizers 0.21.4