dit-base_tobacco

jordyvl/dit-base_tobacco

dit-base_tobacco is a image classification model from jordyvl. Use it when you need a label for an image. It is set up for transformers.

should probably proofread and complete it, then remove this comment. --

Downloads · 30 days

2% of all-time downloads

All-time downloads

644

Public

Repo size

687 MB

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.bin343 MB · 100%

At a glance

Task: Image Classification
Library: transformers
Model type: beit
Access: Public
Created: Jul 6, 2023
Updated: Jul 7, 2023
SHA: f38b4b71

Model card

From the Hugging Face model README

dit-base_tobacco

This model is a fine-tuned version of microsoft/dit-base on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.3120
Accuracy: 0.95
Brier Loss: 0.0965
Nll: 0.6372
F1 Micro: 0.9500
F1 Macro: 0.9545
Ece: 0.0560
Aurc: 0.0092

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: 2e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 16
total_train_batch_size: 128
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Brier Loss	Nll	F1 Micro	F1 Macro	Ece	Aurc
No log	0.96	6	2.4454	0.175	0.9193	8.6626	0.175	0.0676	0.2489	0.8592
No log	1.96	12	2.3287	0.175	0.9034	7.2049	0.175	0.0674	0.2590	0.8557
No log	2.96	18	2.0836	0.23	0.8528	3.3114	0.23	0.1544	0.2652	0.7357
No log	3.96	24	2.0456	0.315	0.8435	3.8932	0.315	0.1785	0.3010	0.6372
No log	4.96	30	1.8778	0.3	0.7820	3.0975	0.3	0.1659	0.2985	0.5174
No log	5.96	36	1.7247	0.365	0.7305	2.7808	0.3650	0.2235	0.2507	0.4036
No log	6.96	42	1.6610	0.38	0.7183	2.6958	0.38	0.2449	0.2538	0.4119
No log	7.96	48	1.4667	0.505	0.6417	2.4078	0.505	0.3653	0.2881	0.2656
No log	8.96	54	1.3427	0.58	0.6031	2.0381	0.58	0.5304	0.2885	0.2470
No log	9.96	60	1.1586	0.635	0.5217	1.8792	0.635	0.5496	0.2831	0.1697
No log	10.96	66	1.0108	0.71	0.4578	1.6886	0.7100	0.6273	0.2851	0.1340
No log	11.96	72	0.8648	0.75	0.3849	1.5408	0.75	0.6788	0.2530	0.0801
No log	12.96	78	0.7342	0.79	0.3327	1.3588	0.79	0.7264	0.2152	0.0575
No log	13.96	84	0.6024	0.835	0.2734	1.2694	0.835	0.7937	0.1876	0.0429
No log	14.96	90	0.5143	0.85	0.2386	1.1756	0.85	0.8175	0.1714	0.0363
No log	15.96	96	0.4429	0.865	0.2044	1.1080	0.865	0.8435	0.1380	0.0277
No log	16.96	102	0.3999	0.885	0.1854	1.0748	0.885	0.8673	0.1407	0.0274
No log	17.96	108	0.3635	0.88	0.1732	1.0361	0.88	0.8594	0.1117	0.0247
No log	18.96	114	0.3166	0.89	0.1454	1.0855	0.89	0.8682	0.0971	0.0196
No log	19.96	120	0.3137	0.905	0.1418	1.1614	0.905	0.8934	0.1041	0.0195
No log	20.96	126	0.3207	0.91	0.1408	1.1941	0.91	0.9002	0.0856	0.0198
No log	21.96	132	0.2753	0.925	0.1224	1.0928	0.925	0.9209	0.0858	0.0145
No log	22.96	138	0.2538	0.925	0.1169	1.0895	0.925	0.9187	0.0863	0.0111
No log	23.96	144	0.2691	0.935	0.1138	1.0767	0.935	0.9279	0.0730	0.0149
No log	24.96	150	0.2775	0.935	0.1131	1.0538	0.935	0.9292	0.0676	0.0157
No log	25.96	156	0.2544	0.94	0.1011	1.0266	0.94	0.9292	0.0643	0.0131
No log	26.96	162	0.2637	0.945	0.1013	1.0337	0.945	0.9384	0.0648	0.0150
No log	27.96	168	0.2787	0.94	0.1089	1.0202	0.94	0.9348	0.0685	0.0161
No log	28.96	174	0.2794	0.935	0.1091	1.0099	0.935	0.9306	0.0671	0.0143
No log	29.96	180	0.2631	0.935	0.1025	0.9815	0.935	0.9306	0.0575	0.0129
No log	30.96	186	0.2616	0.945	0.1009	0.9683	0.945	0.9401	0.0674	0.0120
No log	31.96	192	0.2726	0.935	0.1074	0.9598	0.935	0.9346	0.0641	0.0100
No log	32.96	198	0.2765	0.935	0.1058	0.9067	0.935	0.9321	0.0696	0.0101
No log	33.96	204	0.2662	0.95	0.0965	0.8891	0.9500	0.9522	0.0672	0.0120
No log	34.96	210	0.2761	0.935	0.1019	0.8893	0.935	0.9338	0.0597	0.0134
No log	35.96	216	0.2729	0.945	0.0961	0.8807	0.945	0.9419	0.0552	0.0119
No log	36.96	222	0.2741	0.94	0.1037	0.8782	0.94	0.9356	0.0645	0.0086
No log	37.96	228	0.2686	0.94	0.0994	0.8423	0.94	0.9356	0.0592	0.0085
No log	38.96	234	0.2712	0.95	0.0906	0.8179	0.9500	0.9545	0.0610	0.0105
No log	39.96	240	0.2644	0.95	0.0870	0.8240	0.9500	0.9443	0.0510	0.0110
No log	40.96	246	0.2653	0.95	0.0932	0.8386	0.9500	0.9525	0.0572	0.0118
No log	41.96	252	0.2724	0.955	0.0939	0.8369	0.955	0.9573	0.0602	0.0104
No log	42.96	258	0.2552	0.95	0.0868	0.8079	0.9500	0.9522	0.0539	0.0079
No log	43.96	264	0.2629	0.95	0.0879	0.7800	0.9500	0.9545	0.0526	0.0080
No log	44.96	270	0.2664	0.955	0.0864	0.7660	0.955	0.9575	0.0515	0.0086
No log	45.96	276	0.2777	0.945	0.0948	0.7670	0.945	0.9513	0.0524	0.0096
No log	46.96	282	0.2824	0.94	0.1014	0.7799	0.94	0.9436	0.0570	0.0093
No log	47.96	288	0.2699	0.95	0.0896	0.7706	0.9500	0.9546	0.0528	0.0087
No log	48.96	294	0.2809	0.945	0.0950	0.7691	0.945	0.9480	0.0475	0.0087
No log	49.96	300	0.2827	0.945	0.0940	0.7635	0.945	0.9447	0.0571	0.0091
No log	50.96	306	0.2781	0.945	0.0921	0.7591	0.945	0.9478	0.0552	0.0090
No log	51.96	312	0.2834	0.95	0.0946	0.7572	0.9500	0.9484	0.0549	0.0089
No log	52.96	318	0.2986	0.94	0.0994	0.7541	0.94	0.9363	0.0605	0.0091
No log	53.96	324	0.2957	0.94	0.1016	0.7447	0.94	0.9385	0.0562	0.0086
No log	54.96	330	0.2991	0.94	0.1047	0.7392	0.94	0.9377	0.0592	0.0102
No log	55.96	336	0.3027	0.94	0.1031	0.7235	0.94	0.9377	0.0572	0.0113
No log	56.96	342	0.2945	0.945	0.0968	0.7143	0.945	0.9470	0.0581	0.0104
No log	57.96	348	0.2935	0.94	0.0955	0.7046	0.94	0.9459	0.0569	0.0097
No log	58.96	354	0.2909	0.94	0.0934	0.6969	0.94	0.9459	0.0544	0.0092
No log	59.96	360	0.2973	0.95	0.0939	0.6964	0.9500	0.9545	0.0524	0.0082
No log	60.96	366	0.3222	0.93	0.1108	0.7078	0.93	0.9266	0.0586	0.0088
No log	61.96	372	0.3247	0.935	0.1093	0.7743	0.935	0.9353	0.0622	0.0091
No log	62.96	378	0.3125	0.945	0.1003	0.7651	0.945	0.9453	0.0559	0.0089
No log	63.96	384	0.3035	0.945	0.0993	0.7515	0.945	0.9476	0.0545	0.0088
No log	64.96	390	0.3002	0.945	0.0973	0.7408	0.945	0.9476	0.0537	0.0091
No log	65.96	396	0.3023	0.95	0.0965	0.7321	0.9500	0.9545	0.0523	0.0095
No log	66.96	402	0.3075	0.945	0.1007	0.7323	0.945	0.9477	0.0540	0.0096
No log	67.96	408	0.3062	0.945	0.0999	0.6682	0.945	0.9514	0.0525	0.0098
No log	68.96	414	0.3182	0.945	0.0968	0.6809	0.945	0.9432	0.0485	0.0115
No log	69.96	420	0.3272	0.945	0.0972	0.6879	0.945	0.9432	0.0513	0.0132
No log	70.96	426	0.3210	0.945	0.0973	0.7545	0.945	0.9488	0.0522	0.0124
No log	71.96	432	0.3194	0.945	0.1027	0.7464	0.945	0.9514	0.0546	0.0108
No log	72.96	438	0.3236	0.94	0.1067	0.7486	0.94	0.9427	0.0587	0.0097
No log	73.96	444	0.3166	0.94	0.1049	0.6751	0.94	0.9427	0.0597	0.0096
No log	74.96	450	0.3062	0.945	0.0982	0.6702	0.945	0.9514	0.0526	0.0100
No log	75.96	456	0.3018	0.95	0.0948	0.6823	0.9500	0.9545	0.0523	0.0102
No log	76.96	462	0.3062	0.95	0.0951	0.7444	0.9500	0.9545	0.0522	0.0109
No log	77.96	468	0.3072	0.95	0.0933	0.7437	0.9500	0.9545	0.0501	0.0118
No log	78.96	474	0.3095	0.95	0.0943	0.6749	0.9500	0.9545	0.0512	0.0121
No log	79.96	480	0.3097	0.945	0.0968	0.6654	0.945	0.9514	0.0576	0.0116
No log	80.96	486	0.3094	0.95	0.0967	0.6581	0.9500	0.9545	0.0526	0.0112
No log	81.96	492	0.3109	0.95	0.0954	0.6549	0.9500	0.9545	0.0507	0.0115
No log	82.96	498	0.3104	0.95	0.0949	0.7168	0.9500	0.9545	0.0521	0.0113
0.3747	83.96	504	0.3122	0.95	0.0949	0.7130	0.9500	0.9545	0.0513	0.0111
0.3747	84.96	510	0.3140	0.95	0.0944	0.7116	0.9500	0.9545	0.0534	0.0113
0.3747	85.96	516	0.3175	0.95	0.0949	0.7100	0.9500	0.9545	0.0544	0.0113
0.3747	86.96	522	0.3187	0.95	0.0958	0.7072	0.9500	0.9545	0.0537	0.0111
0.3747	87.96	528	0.3191	0.95	0.0967	0.6428	0.9500	0.9545	0.0536	0.0103
0.3747	88.96	534	0.3168	0.95	0.0963	0.6438	0.9500	0.9545	0.0542	0.0102
0.3747	89.96	540	0.3136	0.95	0.0963	0.6418	0.9500	0.9545	0.0554	0.0099
0.3747	90.96	546	0.3117	0.95	0.0963	0.6407	0.9500	0.9545	0.0533	0.0097
0.3747	91.96	552	0.3113	0.95	0.0964	0.6403	0.9500	0.9545	0.0528	0.0091
0.3747	92.96	558	0.3112	0.95	0.0968	0.6401	0.9500	0.9545	0.0517	0.0091
0.3747	93.96	564	0.3109	0.95	0.0967	0.6393	0.9500	0.9545	0.0563	0.0091
0.3747	94.96	570	0.3112	0.95	0.0969	0.6370	0.9500	0.9545	0.0567	0.0092
0.3747	95.96	576	0.3118	0.95	0.0971	0.6364	0.9500	0.9545	0.0568	0.0091
0.3747	96.96	582	0.3120	0.95	0.0969	0.6377	0.9500	0.9545	0.0564	0.0092
0.3747	97.96	588	0.3121	0.95	0.0966	0.6379	0.9500	0.9545	0.0560	0.0092
0.3747	98.96	594	0.3121	0.95	0.0965	0.6374	0.9500	0.9545	0.0560	0.0092
0.3747	99.96	600	0.3120	0.95	0.0965	0.6372	0.9500	0.9545	0.0560	0.0092

Framework versions

Transformers 4.26.1
Pytorch 1.13.1.post200
Datasets 2.9.0
Tokenizers 0.13.2

Task: Image Classification
Library: transformers
Type: beit
Created: Jul 6, 2023
Updated: Jul 7, 2023