wav2vec2-large-xls-r-vi-colab

This model is a fine-tuned version of facebook/wav2vec2-base on the common_voice_13_0 dataset. It achieves the following results on the evaluation set:

• Loss: 3.4884
• Wer: 1.0
• Cer: 1.0

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: 1e-05
• train_batch_size: 32
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 0.1
• num_epochs: 80
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
9.4752	7.1111	160	4.4992	1.0	1.0
4.2035	14.2222	320	3.9228	1.0	1.0
3.7611	21.3333	480	3.6584	1.0	1.0
3.5825	28.4444	640	3.5584	1.0	1.0
3.5044	35.5556	800	3.5285	1.0	1.0
3.4669	42.6667	960	3.5226	1.0	1.0
3.4382	49.7778	1120	3.5093	1.0	1.0
3.4183	56.8889	1280	3.4942	1.0	1.0
3.4002	64.0	1440	3.4957	1.0	1.0
3.3871	71.1111	1600	3.4896	1.0	1.0
3.382	78.2222	1760	3.4884	1.0	1.0

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1