git-base-naruto

This model is a fine-tuned version of microsoft/git-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0400
• Wer Score: 0.3529

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: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 4
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 50
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer Score
7.3756	5.8824	50	4.6183	7.9118
2.5329	11.7647	100	0.6340	7.0
0.199	17.6471	150	0.0438	0.7941
0.0155	23.5294	200	0.0390	0.8529
0.0051	29.4118	250	0.0385	0.3529
0.0025	35.2941	300	0.0392	0.3235
0.0018	41.1765	350	0.0397	0.3529
0.0016	47.0588	400	0.0400	0.3529

Framework versions

• Transformers 4.40.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1