bert-small-amharic-32k-l12-384

This model is a fine-tuned version of prajjwal1/bert-small on an unknown dataset. It achieves the following results on the evaluation set:

• eval_loss: 2.6525
• eval_model_preparation_time: 0.0031
• eval_runtime: 6.9115
• eval_samples_per_second: 2997.489
• eval_steps_per_second: 11.72
• epoch: 35.6989
• step: 297943

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: 0.0001
• train_batch_size: 256
• eval_batch_size: 256
• seed: 42
• optimizer: Use OptimizerNames.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_ratio: 0.1
• lr_scheduler_warmup_steps: 10000
• num_epochs: 40
• mixed_precision_training: Native AMP

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.1