Udap10/wav2vec2-large-mms-1b-dzo-colab-cerMetrices2

repo_name

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: 0.7776
• Cer: 0.2322

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: 0.001
• train_batch_size: 8
• 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_steps: 100
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Cer
No log	1.0	10	13.5720	1.6486
No log	2.0	20	10.3153	1.0010
No log	3.0	30	6.3205	0.8398
No log	4.0	40	3.8020	0.9725
No log	5.0	50	3.1634	0.9439
No log	6.0	60	2.7690	0.6942
No log	7.0	70	1.6732	0.3799
No log	8.0	80	1.1050	0.3083
No log	9.0	90	0.9242	0.2633
5.3111	10.0	100	0.7776	0.2322

Framework versions

• Transformers 4.49.0.dev0
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0