vit-Mistral-UCM-without-captioning

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

• Loss: 0.7752
• Accuracy: 74.26
• Bleu-1: 0.8243
• Bleu-2: 0.7645
• Bleu-3: 0.7164
• Bleu-4: 0.6709
• Meteor: 0.8175
• Rouge-l: 0.7933
• Cider: 3.3030

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: 0.1
• 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	0.8768	72.09	0.4152	0.2825	0.2067	0.1530	0.3201	0.3517	0.3399
No log	2.0	296	0.7330	72.01	0.4497	0.3349	0.2616	0.2125	0.3385	0.3896	0.9183
No log	3.0	444	0.5821	72.81	0.8068	0.7453	0.6955	0.6512	0.7747	0.7695	3.1554
No log	4.0	592	0.5829	73.5	0.8522	0.7882	0.7329	0.6823	0.8173	0.8069	3.3459
No log	5.0	740	0.6105	72.64	0.8241	0.7669	0.7169	0.6723	0.8029	0.7968	3.2587
No log	6.0	888	0.6279	74.23	0.8280	0.7742	0.7272	0.6845	0.8246	0.8096	3.3476
0.6501	7.0	1036	0.6323	74.45	0.8648	0.8175	0.7766	0.7390	0.8395	0.8265	3.5956
0.6501	8.0	1184	0.6660	74.41	0.8551	0.7956	0.7462	0.7025	0.8141	0.8074	3.4659
0.6501	9.0	1332	0.6622	74.14	0.8649	0.8160	0.7718	0.7337	0.8467	0.8354	3.5959
0.6501	10.0	1480	0.6729	74.22	0.8345	0.7724	0.7250	0.6831	0.7891	0.7843	3.4044
0.6501	11.0	1628	0.7075	73.45	0.8164	0.7500	0.6938	0.6433	0.7997	0.7847	3.2171
0.6501	12.0	1776	0.7030	73.81	0.8256	0.7744	0.7298	0.6853	0.8056	0.7936	3.3060
0.6501	13.0	1924	0.7365	73.25	0.8213	0.7630	0.7178	0.6766	0.7975	0.7907	3.3430
0.2977	14.0	2072	0.7188	74.21	0.8429	0.7895	0.7433	0.7005	0.8267	0.8081	3.3791
0.2977	15.0	2220	0.7222	74.57	0.8542	0.8002	0.7551	0.7118	0.8351	0.8270	3.5997
0.2977	16.0	2368	0.7646	74.77	0.8527	0.7977	0.7511	0.7078	0.8247	0.8165	3.4923
0.2977	17.0	2516	0.7979	73.87	0.8278	0.7664	0.7173	0.6720	0.8007	0.7841	3.3060
0.2977	18.0	2664	0.7695	74.56	0.8475	0.7911	0.7435	0.6987	0.8262	0.8140	3.4166
0.2977	19.0	2812	0.7752	74.26	0.8243	0.7645	0.7164	0.6709	0.8175	0.7933	3.3030

Framework versions

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