vit-Mistral-RSICD-without-captioning

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

• Loss: 2.1520
• Accuracy: 78.79
• Bleu-1: 0.6471
• Bleu-2: 0.4703
• Bleu-3: 0.3606
• Bleu-4: 0.2880
• Meteor: 0.4808
• Rouge-l: 0.4839
• Cider: 0.8155

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: 0.0001
• train_batch_size: 64
• eval_batch_size: 64
• seed: 50
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 1024
• num_epochs: 128
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Bleu-1	Bleu-2	Bleu-3	Bleu-4	Meteor	Rouge-l	Cider
No log	1.0	768	1.2533	78.33	0.6350	0.4600	0.3526	0.2798	0.4738	0.4775	0.7666
1.1016	2.0	1536	1.3913	78.35	0.6360	0.4668	0.3615	0.2915	0.4900	0.4875	0.8142
0.5840	3.0	2304	1.6044	78.97	0.6547	0.4822	0.3765	0.3058	0.4847	0.4903	0.8581
0.3908	4.0	3072	1.7479	78.75	0.6460	0.4721	0.3638	0.2907	0.4822	0.4861	0.8140
0.3908	5.0	3840	1.8525	78.93	0.6473	0.4681	0.3577	0.2852	0.4730	0.4805	0.8030
0.2978	6.0	4608	1.9700	78.72	0.6459	0.4750	0.3694	0.2986	0.4844	0.4873	0.8337
0.2641	7.0	5376	2.0651	78.94	0.6516	0.4756	0.3644	0.2904	0.4753	0.4798	0.8092
0.2456	8.0	6144	2.1520	78.79	0.6471	0.4703	0.3606	0.2880	0.4808	0.4839	0.8155

Framework versions

• Transformers 5.12.1
• Pytorch 2.12.1+cu130
• Datasets 5.0.0
• Tokenizers 0.22.2