paligemma_khatt_paragraph_mix_448_ep1_r128_linearlr-2

This model is a fine-tuned version of google/paligemma-3b-mix-448 on the khatt-paragraphs dataset. It achieves the following results on the evaluation set:

• Loss: 0.8302

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: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 4
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 2
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss
1.1635	0.1014	100	1.1339
0.9899	0.2027	200	0.9917
0.7868	0.3041	300	0.9362
0.9343	0.4055	400	0.9000
0.9738	0.5068	500	0.8741
0.8056	0.6082	600	0.8583
0.9728	0.7096	700	0.8417
0.8998	0.8109	800	0.8319
0.7644	0.9123	900	0.8302

Framework versions

• PEFT 0.11.1
• Transformers 4.42.3
• Pytorch 2.1.2
• Datasets 2.20.0
• Tokenizers 0.19.1