PG2-28b-pt-448-COND_GEN

This model is a fine-tuned version of google/paligemma2-28b-pt-448 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.4487
• Accuracy: 0.7

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: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Use 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_steps: 2
• num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.3753	0.2222	400	0.3052	0.7
0.2667	0.4444	800	0.2497	0.73
0.3405	0.6667	1200	0.2686	0.685
0.2995	0.8889	1600	0.2595	0.735
0.0841	1.1111	2000	0.4457	0.725
0.1271	1.3333	2400	0.3524	0.72
0.0906	1.5556	2800	0.4202	0.69
0.0532	1.7778	3200	0.4716	0.68
0.1139	2.0	3600	0.4487	0.7

Framework versions

• PEFT 0.14.0
• Transformers 4.48.0
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0