mms-1b-all-bemgen-combined-m50f50-dnn-42-0.06

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

• Loss: 14.5630
• Cer: 2.9735

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: 30.0
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Cer
14.3833	0.8439	100	14.5629	2.9738
13.9155	1.6835	200	14.5631	2.9731
13.3928	2.5232	300	14.5629	2.9734
14.5477	3.3629	400	14.5628	2.9737
13.9813	4.2025	500	14.5629	2.9736
15.4189	5.0422	600	14.5628	2.9738
12.5819	5.8861	700	14.5631	2.9734

Framework versions

• Transformers 4.53.0
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.4