w2v2-bert-Wolof-10-hours-Google-Fleurs-dataset

This model is a fine-tuned version of facebook/w2v-bert-2.0 on the fleurs dataset. It achieves the following results on the evaluation set:

• Loss: 1.1192
• Wer: 0.3997
• Cer: 0.1251

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: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 31

Training results

Training Loss	Epoch	Step	Cer	Validation Loss	Wer
1.403	5.23	400	0.1672	0.6614	0.4857
0.4459	10.46	800	0.1432	0.6289	0.4476
0.2611	15.69	1200	0.1402	0.6713	0.4298
0.1019	21.01	1600	0.8813	0.4052	0.1288
0.0291	26.24	2000	1.1192	0.3997	0.1251

Framework versions

• Transformers 4.38.1
• Pytorch 2.1.0+cu118
• Datasets 2.17.0
• Tokenizers 0.15.2