Modelpublic

swadhindas324/vit-Mistral-RSICD-without-captioning

sourceHugging Faceupdated 2mo agoView on Hugging Face

0likes5downloads

Model Card

vit-Mistral-RSICD-without-captioning

This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset. It achieves the following results on the evaluation set:

—Loss: 1.6638
—Accuracy: 81.52
—Bleu-1: 0.5931
—Bleu-2: 0.4579
—Bleu-3: 0.3664
—Bleu-4: 0.3025
—Meteor: 0.5501
—Rouge-l: 0.5102
—Cider: 1.6319

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
—trainbatchsize: 64
—evalbatchsize: 64
—seed: 50
—optimizer: Use OptimizerNames.ADAMWTORCHFUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
—lrschedulertype: linear
—lrschedulerwarmup_steps: 0.1
—num_epochs: 128
—mixedprecisiontraining: 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	1.2860	79.79	0.4668	0.3027	0.2115	0.1537	0.4086	0.3975	0.4801
1.5997	2.0	1536	1.2245	79.68	0.6371	0.5021	0.4025	0.3265	0.6309	0.5761	1.6413
0.8304	3.0	2304	1.2246	81.33	0.6822	0.5578	0.4652	0.3916	0.6689	0.6262	2.1339
0.5919	4.0	3072	1.3626	81.74	0.6722	0.5462	0.4543	0.3841	0.6484	0.6051	2.0383
0.5919	5.0	3840	1.4524	80.57	0.6152	0.4803	0.3873	0.3184	0.5905	0.5435	1.6186
0.4163	6.0	4608	1.5060	82.13	0.6521	0.5248	0.4354	0.3683	0.6202	0.5846	1.9912
0.3441	7.0	5376	1.6478	80.98	0.6014	0.4637	0.3697	0.3019	0.5811	0.5338	1.6237
0.3101	8.0	6144	1.6638	81.52	0.5931	0.4579	0.3664	0.3025	0.5501	0.5102	1.6319

Framework versions

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