GIT_inf_w_bbox_caption_ep5

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

• Loss: 0.0748
• Rouge1: 7.4067
• Rouge2: 3.0184
• Rougel: 6.6214
• Rougelsum: 6.6313
• Gen Len: 215.256

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: 32
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
0.0794	1.0	777	0.0800	6.7935	2.5273	6.016	6.0274	215.258
0.0704	2.0	1554	0.0774	6.861	2.638	6.1681	6.1833	215.258
0.0637	3.0	2331	0.0761	7.1364	2.862	6.4308	6.4446	215.258
0.0595	4.0	3108	0.0750	7.2911	2.9116	6.5386	6.5461	215.258
0.0544	5.0	3885	0.0748	7.4067	3.0184	6.6214	6.6313	215.256

Framework versions

• Transformers 4.38.1
• Pytorch 2.2.1+cu121
• Datasets 2.17.1
• Tokenizers 0.15.2