finetune_colpali_v1_2-vdsid_french-4bit

This model is a fine-tuned version of vidore/colpaligemma-3b-pt-448-base on the vidore/vdsid_french dataset. It achieves the following results on the evaluation set:

• Loss: 0.0193

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
• gradient_accumulation_steps: 4
• 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: 100
• num_epochs: 1.5

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0034	1	0.0579
0.0093	0.3404	100	0.0258
0.0083	0.6809	200	0.0256
0.005	1.0213	300	0.0205
0.0083	1.3617	400	0.0193

Framework versions

• Transformers 4.45.1
• Pytorch 2.4.1+cu121
• Datasets 3.0.1
• Tokenizers 0.20.0