git-base-appliances

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: 3.4281
• Wer Score: 3.2039

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: 16
• total_train_batch_size: 32
• 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: 50
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer Score
84.9149	3.6893	100	2.9324	3.8749
39.5513	7.3991	200	2.6633	2.6427
28.5436	11.1088	300	2.7566	3.0925
20.9713	14.7982	400	2.8737	3.2221
15.2719	18.5079	500	2.9953	2.9620
11.4351	22.2177	600	3.1084	3.0657
8.6091	25.9070	700	3.1823	3.1642
6.4907	29.6168	800	3.2530	3.0930
5.0524	33.3265	900	3.3025	3.1223
4.0674	37.0363	1000	3.3509	3.1089
3.3508	40.7256	1100	3.3843	3.1288
2.8752	44.4354	1200	3.4132	3.1570
2.5338	48.1451	1300	3.4281	3.2039

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0