swadhindas324/swin-Mistral-UCM-without-captioning

swin-Mistral-UCM-captioning

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

• Loss: 0.7974
• Accuracy: 74.93
• Bleu-1: 0.8485
• Bleu-2: 0.7909
• Bleu-3: 0.7437
• Bleu-4: 0.7027
• Meteor: 0.8063
• Rouge-l: 0.8060
• Cider: 3.4543

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: 8
• 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	148	0.7681	69.91	0.3560	0.2384	0.1671	0.1215	0.2719	0.2982	0.4223
No log	2.0	296	0.6346	69.59	0.7218	0.6492	0.5907	0.5395	0.7064	0.6989	2.7599
No log	3.0	444	0.6058	75.06	0.8570	0.7971	0.7496	0.7049	0.8177	0.8083	3.3427
No log	4.0	592	0.6234	73.7	0.8099	0.7363	0.6792	0.6276	0.7820	0.7740	3.2220
No log	5.0	740	0.6354	73.61	0.8249	0.7579	0.7078	0.6650	0.7911	0.7806	3.2551
No log	6.0	888	0.6600	74.44	0.8429	0.7862	0.7373	0.6933	0.8238	0.8193	3.4049
0.5906	7.0	1036	0.6738	74.57	0.8390	0.7797	0.7328	0.6912	0.7946	0.7973	3.3445
0.5906	8.0	1184	0.7451	73.86	0.8469	0.7902	0.7409	0.6979	0.8163	0.8066	3.4162
0.5906	9.0	1332	0.7000	75.19	0.8503	0.7928	0.7448	0.7017	0.8172	0.8065	3.5555
0.5906	10.0	1480	0.7449	74.66	0.8552	0.7931	0.7448	0.7011	0.8157	0.8053	3.4820
0.5906	11.0	1628	0.7309	74.96	0.8498	0.8003	0.7594	0.7227	0.7923	0.7952	3.4882
0.5906	12.0	1776	0.7576	74.54	0.8356	0.7747	0.7271	0.6839	0.8087	0.8012	3.3285
0.5906	13.0	1924	0.7656	74.75	0.8474	0.7953	0.7535	0.7159	0.8160	0.8147	3.5224
0.2906	14.0	2072	0.7898	74.23	0.8329	0.7736	0.7267	0.6855	0.7933	0.7874	3.4272
0.2906	15.0	2220	0.8058	74.46	0.8438	0.7815	0.7309	0.6851	0.8047	0.7996	3.5010
0.2906	16.0	2368	0.7974	74.93	0.8485	0.7909	0.7437	0.7027	0.8063	0.8060	3.4543

Framework versions

• Transformers 5.12.1
• Pytorch 2.12.0+cu130
• Datasets 5.0.0
• Tokenizers 0.22.2