git-base-pokemon

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: 0.4035
• Wer Score: 1.5183

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: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 8
• 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
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer Score
7.2724	2.0	50	4.6456	10.9317
2.6854	4.0	100	0.8998	10.6017
0.5411	6.0	150	0.4014	0.9533
0.2944	8.0	200	0.3795	0.945
0.1998	10.0	250	0.3885	1.17
0.1466	12.0	300	0.3911	0.9633
0.1161	14.0	350	0.3987	0.9633
0.0994	16.0	400	0.4011	1.24
0.0891	18.0	450	0.4027	1.2417
0.0838	20.0	500	0.4035	1.5183

Framework versions

• Transformers 4.53.0
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.2