resnet-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: 0.9403
• Accuracy: 74.87
• Bleu-1: 0.8253
• Bleu-2: 0.7628
• Bleu-3: 0.7115
• Bleu-4: 0.6639
• Meteor: 0.7936
• Rouge-l: 0.7886
• Cider: 3.3492

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	3.7133	68.76	0.3122	0.1569	0.0696	0.0127	0.1541	0.2755	0.1638
No log	2.0	296	0.8074	68.51	0.5354	0.4448	0.3813	0.3399	0.4908	0.4964	1.6647
No log	3.0	444	0.6727	71.49	0.7345	0.6642	0.6116	0.5656	0.6883	0.6901	2.6803
No log	4.0	592	0.6691	73.58	0.7725	0.7067	0.6550	0.6119	0.7222	0.7216	2.9514
No log	5.0	740	0.6746	72.18	0.7416	0.6648	0.6084	0.5603	0.7249	0.7136	2.8120
No log	6.0	888	0.6908	72.03	0.7481	0.6802	0.6275	0.5812	0.7281	0.7203	2.8980
1.2018	7.0	1036	0.6925	73.67	0.8014	0.7398	0.6918	0.6529	0.7484	0.7444	3.0107
1.2018	8.0	1184	0.7252	72.52	0.7954	0.7345	0.6870	0.6441	0.7685	0.7581	3.0965
1.2018	9.0	1332	0.7115	74.62	0.8097	0.7536	0.7093	0.6715	0.7810	0.7752	3.2966
1.2018	10.0	1480	0.7696	73.12	0.7502	0.6852	0.6318	0.5853	0.7303	0.7236	2.8921
1.2018	11.0	1628	0.7353	74.16	0.8021	0.7392	0.6863	0.6362	0.7667	0.7649	3.0826
1.2018	12.0	1776	0.7867	74.34	0.7866	0.7239	0.6728	0.6257	0.7581	0.7559	3.0359
1.2018	13.0	1924	0.7661	73.71	0.8110	0.7453	0.6931	0.6468	0.7781	0.7642	3.1120
0.3380	14.0	2072	0.7747	74.86	0.8239	0.7649	0.7183	0.6754	0.7841	0.7748	3.3324
0.3380	15.0	2220	0.8137	73.92	0.8082	0.7473	0.6961	0.6487	0.7850	0.7750	3.2459
0.3380	16.0	2368	0.8149	73.73	0.8099	0.7518	0.7016	0.6543	0.7811	0.7746	3.2421
0.3380	17.0	2516	0.8380	74.14	0.8155	0.7477	0.6952	0.6497	0.7857	0.7724	3.2162
0.3380	18.0	2664	0.8416	74.23	0.8151	0.7513	0.6971	0.6473	0.8155	0.7874	3.2287
0.3380	19.0	2812	0.8520	74.04	0.8240	0.7589	0.7032	0.6517	0.7976	0.7818	3.2306
0.3380	20.0	2960	0.8757	73.97	0.8032	0.7418	0.6928	0.6484	0.7872	0.7742	3.2376
0.2863	21.0	3108	0.8525	74.29	0.8237	0.7628	0.7123	0.6673	0.8052	0.7863	3.3082
0.2863	22.0	3256	0.8676	74.7	0.8301	0.7650	0.7122	0.6642	0.8047	0.7891	3.3572
0.2863	23.0	3404	0.8635	73.93	0.8389	0.7770	0.7279	0.6820	0.8099	0.8027	3.2788
0.2863	24.0	3552	0.9010	74.64	0.8283	0.7625	0.7090	0.6593	0.8015	0.7858	3.3640
0.2863	25.0	3700	0.9004	74.62	0.8258	0.7628	0.7108	0.6643	0.7919	0.7830	3.2338
0.2863	26.0	3848	0.8713	74.69	0.8441	0.7861	0.7361	0.6904	0.8319	0.8172	3.4987
0.2863	27.0	3996	0.8921	74.74	0.8220	0.7694	0.7253	0.6840	0.7981	0.7866	3.2820
0.2676	28.0	4144	0.8975	74.71	0.8098	0.7448	0.6914	0.6432	0.7834	0.7742	3.2960
0.2676	29.0	4292	0.8785	75.44	0.8362	0.7824	0.7347	0.6899	0.8137	0.8045	3.5087
0.2676	30.0	4440	0.8890	74.9	0.8190	0.7622	0.7126	0.6672	0.7928	0.7832	3.2668
0.2676	31.0	4588	0.8992	74.38	0.8164	0.7554	0.7050	0.6608	0.7779	0.7686	3.2444
0.2676	32.0	4736	0.8994	73.88	0.8401	0.7823	0.7336	0.6910	0.8171	0.7997	3.4254
0.2676	33.0	4884	0.9135	74.72	0.8175	0.7548	0.7031	0.6574	0.7985	0.7814	3.2915
0.2676	34.0	5032	0.9085	74.79	0.8165	0.7537	0.7023	0.6552	0.7901	0.7828	3.3517
0.2576	35.0	5180	0.9409	74.79	0.8326	0.7740	0.7237	0.6764	0.8134	0.7989	3.3713
0.2576	36.0	5328	0.9403	74.87	0.8253	0.7628	0.7115	0.6639	0.7936	0.7886	3.3492

Framework versions

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