This model is a fine-tuned version of facebook/w2v-bert-2.0 on the common_voice_16_0 dataset. It achieves the following results on the evaluation set:

Loss: 0.2080
Wer: 0.1643

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: 5e-05
train_batch_size: 16
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 32
optimizer: Use 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: 500
num_epochs: 12
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.758	1.5973	300	0.2420	0.2881
0.1988	3.192	600	0.2179	0.2280
0.124	4.7893	900	0.1846	0.2048
0.0755	6.384	1200	0.1841	0.1897
0.0446	7.9813	1500	0.1830	0.1729
0.0221	9.576	1800	0.1967	0.1706
0.0097	11.1707	2100	0.2080	0.1643

Framework versions

Transformers 4.53.0.dev0
Pytorch 2.7.1+cu126
Datasets 2.15.0
Tokenizers 0.21.1

Sargis001/w2v-bert-2.0-armenian002-CV16.0

w2v-bert-2.0-armenian002-CV16.0

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Sargis001/
w2v-bert-2.0-armenian002-CV16.0