WAXAL-sid_asr-w2v-bert-2.0

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

• Loss: 0.6255
• Cer: 0.1208

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: 4
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Cer
4.3471	1.7143	300	0.6681	0.1773
1.2452	3.4286	600	0.5896	0.1315
1.0122	5.1429	900	0.5721	0.1241
0.8201	6.8571	1200	0.5715	0.1183
0.6385	8.5714	1500	0.6075	0.1199
0.6385	10.0	1750	0.6255	0.1208

Framework versions

• Transformers 5.12.1
• Pytorch 2.11.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.2