babylama-scratch

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: 3.9140

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
4.3038	0.9999	5559	4.3162
4.0495	1.9999	11119	4.0657
3.8894	3.0000	16679	3.9579
3.769	4.0	22239	3.9166
3.73	4.9993	27795	3.9140

Framework versions

• Transformers 4.44.2
• Pytorch 2.5.0+cu121
• Datasets 3.1.0
• Tokenizers 0.19.1