blip-finetuned-colab100

This model is a fine-tuned version of Salesforce/blip-image-captioning-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.9787

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: 9e-06
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 64
• 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_steps: 50
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
3.3479	1.4444	150	2.9834
1.9243	2.8889	300	1.7804
1.377	4.3285	450	1.3534
1.176	5.7729	600	1.1680
1.0378	7.2126	750	1.0528
0.9697	8.6570	900	0.9787

Framework versions

• Transformers 4.51.3
• Pytorch 2.5.1+cu121
• Datasets 3.6.0
• Tokenizers 0.21.1