MobileViT_Food_80epoch

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

• Loss: 0.7769
• Accuracy: 0.8053

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: 5e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 80

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
4.5902	0.9994	1183	4.5818	0.0286
4.2708	1.9996	2367	4.2247	0.1690
3.7077	2.9998	3551	3.5174	0.2602
3.271	4.0	4735	2.9216	0.3432
2.8193	4.9994	5918	2.4241	0.4276
2.4733	5.9996	7102	2.0284	0.5017
2.1674	6.9998	8286	1.7180	0.5674
1.9884	8.0	9470	1.5144	0.6122
1.7582	8.9994	10653	1.3711	0.6450
1.4781	9.9996	11837	1.2530	0.6689
1.6275	10.9998	13021	1.1598	0.6924
1.5292	12.0	14205	1.1260	0.7046
1.3675	12.9994	15388	1.0912	0.7122
1.3782	13.9996	16572	1.0276	0.7255
1.3084	14.9998	17756	1.0042	0.7345
1.1715	16.0	18940	0.9771	0.7427
1.2386	16.9994	20123	0.9601	0.7461
1.1787	17.9996	21307	0.9489	0.7472
1.1716	18.9998	22491	0.9360	0.7516
1.1363	20.0	23675	0.9129	0.7595
1.2677	20.9994	24858	0.9007	0.7633
1.2019	21.9996	26042	0.8869	0.7657
1.0633	22.9998	27226	0.8835	0.7656
1.0393	24.0	28410	0.8742	0.7693
0.9558	24.9994	29593	0.8704	0.7705
1.0596	25.9996	30777	0.8455	0.7764
1.0749	26.9998	31961	0.8431	0.7793
0.9913	28.0	33145	0.8332	0.7795
0.9477	28.9994	34328	0.8434	0.7777
0.9681	29.9996	35512	0.8215	0.7840
0.9356	30.9998	36696	0.8050	0.7888
0.806	32.0	37880	0.8152	0.7870
1.0011	32.9994	39063	0.8089	0.7843
0.9268	33.9996	40247	0.8018	0.7884
0.8209	34.9998	41431	0.8147	0.7876
0.8193	36.0	42615	0.8043	0.7893
0.8523	36.9994	43798	0.8014	0.7893
0.9134	37.9996	44982	0.7995	0.7895
0.9263	38.9998	46166	0.7928	0.7896
0.9393	40.0	47350	0.7951	0.7952
0.8028	40.9994	48533	0.7840	0.7967
0.8299	41.9996	49717	0.7994	0.7929
0.791	42.9998	50901	0.7873	0.7921
0.8739	44.0	52085	0.7869	0.7956
0.8777	44.9994	53268	0.7835	0.7952
0.8077	45.9996	54452	0.7815	0.7957
0.9119	46.9998	55636	0.7753	0.7984
0.9867	48.0	56820	0.7824	0.7969
0.8115	48.9994	58003	0.7852	0.7975
0.779	49.9996	59187	0.7815	0.7992
0.755	50.9998	60371	0.7796	0.8011
0.7529	52.0	61555	0.7739	0.8014
0.6878	52.9994	62738	0.7914	0.7989
0.744	53.9996	63922	0.7774	0.8002
0.7346	54.9998	65106	0.7679	0.8012
0.7672	56.0	66290	0.7696	0.7998
0.8018	56.9994	67473	0.7877	0.7987
0.7507	57.9996	68657	0.7903	0.7979
0.7632	58.9998	69841	0.7831	0.8010
0.7013	60.0	71025	0.7799	0.7985
0.7364	60.9994	72208	0.7527	0.8079
0.8036	61.9996	73392	0.7664	0.8010
0.74	62.9998	74576	0.7683	0.8022
0.6531	64.0	75760	0.7548	0.8021
0.7375	64.9994	76943	0.7623	0.8022
0.7228	65.9996	78127	0.7820	0.8028
0.7318	66.9998	79311	0.7625	0.8008
0.6529	68.0	80495	0.7693	0.8036
0.68	68.9994	81678	0.7371	0.8093
0.7396	69.9996	82862	0.7699	0.8040
0.7388	70.9998	84046	0.7596	0.8038
0.7135	72.0	85230	0.7607	0.8043
0.6667	72.9994	86413	0.7666	0.8034
0.6866	73.9996	87597	0.7640	0.8046
0.6601	74.9998	88781	0.7573	0.8037
0.7305	76.0	89965	0.7443	0.8094
0.7507	76.9994	91148	0.7636	0.8053
0.7073	77.9996	92332	0.7692	0.8033
0.688	78.9998	93516	0.7609	0.8044
0.6694	79.9493	94640	0.7769	0.8053

Framework versions

• Transformers 4.45.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.1
• Tokenizers 0.20.1