babylama-attentionchange

This model is a fine-tuned version of babylm/babyllama-100m-2024 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 3.9233

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: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 64
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
4.3065	0.9999	5559	4.3348
4.0853	1.9999	11119	4.0786
3.9166	2.9998	16678	3.9717
3.7768	3.9999	22238	3.9284
3.7544	4.9994	27795	3.9233

Framework versions

• Transformers 4.46.2
• Pytorch 2.5.1+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3