facebook-mms-1b-all-km-indomain5-withoutac

This model is a fine-tuned version of facebook/mms-1b-all on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2768
• Wer: 0.1952

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: 3e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 32
• 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: cosine
• lr_scheduler_warmup_steps: 100
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
15.3586	4.7853	100	0.4774	0.4006
2.4590	9.5399	200	0.2889	0.2386
1.7887	14.2945	300	0.2907	0.2188
1.4262	19.0491	400	0.2744	0.2079
1.2769	23.8344	500	0.2792	0.1968
1.2507	28.5890	600	0.2770	0.1942
1.2507	30.0	630	0.2768	0.1952

Framework versions

• Transformers 5.7.0
• Pytorch 2.10.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.2