mms-1b-all-bigc-combined-15hrs-62

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

• Loss: 1.3579
• Wer: 0.7166

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: 62
• 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	Wer
6.6815	0.2160	100	1.9230	0.9608
0.9004	0.4320	200	1.4664	0.7384
0.759	0.6479	300	1.3800	0.7253
0.7673	0.8639	400	1.3579	0.7169
0.7218	1.0799	500	1.3380	0.7056
0.6999	1.2959	600	1.3271	0.6997
0.7227	1.5119	700	1.3676	0.7041
0.6687	1.7279	800	1.5271	0.7032
0.6869	1.9438	900	1.3698	0.7023

Framework versions

• Transformers 4.53.0.dev0
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.0