Tamasheq-XLSR-53-mixed-Kabyle

This model is a fine-tuned version of ad019el/tamasheq-99-1 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.7107
• Wer: 0.6480

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: 16
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 200
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
7.2189	5.1	500	1.2682	0.8246
0.8718	10.2	1000	0.6082	0.6467
0.5876	15.3	1500	0.7394	0.6896
0.5031	20.41	2000	0.7304	0.6735
0.447	25.51	2500	0.7107	0.6480

Framework versions

• Transformers 4.20.1
• Pytorch 1.11.0
• Datasets 2.1.0
• Tokenizers 0.12.1