paligemma-cnmc-ft

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: 0.2339

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: 2e-05
• train_batch_size: 2
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 170
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.9645	17	1.3682
No log	1.9858	35	1.1198
1.1664	2.9504	52	0.6518
1.1664	3.9716	70	0.3661
1.1664	4.9929	88	0.3079
0.3897	5.9574	105	0.2835
0.3897	6.9787	123	0.2548
0.3897	8.0	141	0.2513
0.2665	8.9645	158	0.2098
0.2665	9.9858	176	0.2031
0.2665	10.9504	193	0.2482
0.1931	11.9716	211	0.2339

Framework versions

• PEFT 0.11.1
• Transformers 4.43.0.dev0
• Pytorch 2.3.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1