vqamed_exp

This model is a fine-tuned version of google/paligemma-3b-pt-224 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.7403

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: 10
• eval_batch_size: 10
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 40
• 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
• lr_scheduler_warmup_ratio: 0.03
• training_steps: 1000

Training results

Training Loss	Epoch	Step	Validation Loss
1.5638	0.3125	100	1.1130
1.2452	0.625	200	0.9256
1.0757	0.9375	300	0.8467
0.8429	1.25	400	0.8236
0.9455	1.5625	500	0.7959
0.8045	1.875	600	0.7693
0.6543	2.1875	700	0.7652
0.7764	2.5	800	0.7567
0.7508	2.8125	900	0.7401
0.593	3.125	1000	0.7403

Framework versions

• PEFT 0.11.1
• Transformers 4.46.2
• Pytorch 2.4.1+cu121
• Datasets 2.19.1
• Tokenizers 0.20.0