dinov2-base_winter-wheat_fertiliser-identification_dnd-ww2020

This model is a fine-tuned version of facebook/dinov2-base on the winter-wheat_fertiliser-identification_dnd-ww2020 dataset. It achieves the following results on the evaluation set:

Loss: 0.4946
Accuracy: 0.7985

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: 32
eval_batch_size: 1
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.7679	1.0	38	1.7781	0.3060
1.7069	2.0	76	1.5920	0.3806
1.5476	3.0	114	1.5259	0.3955
1.3803	4.0	152	1.3005	0.4851
1.3015	5.0	190	1.1809	0.5299
1.2157	6.0	228	1.1946	0.4851
1.1173	7.0	266	1.0651	0.5821
1.1138	8.0	304	1.0565	0.5522
1.0272	9.0	342	1.0698	0.5299
1.0161	10.0	380	1.0412	0.6045
1.0413	11.0	418	0.9661	0.5746
0.9973	12.0	456	1.0308	0.5746
0.9868	13.0	494	1.0036	0.6045
0.921	14.0	532	0.8507	0.6418
1.0229	15.0	570	0.8769	0.6567
0.9111	16.0	608	0.9176	0.6343
0.7588	17.0	646	0.8129	0.6343
0.8978	18.0	684	0.8439	0.6493
0.8469	19.0	722	0.7913	0.6791
0.8492	20.0	760	0.8750	0.6343
0.6988	21.0	798	0.8203	0.6716
0.9145	22.0	836	0.7737	0.6791
0.739	23.0	874	0.7167	0.7313
0.8208	24.0	912	0.7586	0.6716
0.6627	25.0	950	0.7110	0.7090
0.7707	26.0	988	0.7468	0.7015
0.5777	27.0	1026	0.7886	0.6642
0.5762	28.0	1064	0.6722	0.7612
0.6887	29.0	1102	0.7041	0.7015
0.7005	30.0	1140	0.6601	0.7239
0.6654	31.0	1178	0.7200	0.7090
0.5987	32.0	1216	0.7395	0.6791
0.571	33.0	1254	0.6474	0.7463
0.6837	34.0	1292	0.7545	0.6940
0.6899	35.0	1330	0.7235	0.7313
0.693	36.0	1368	0.6012	0.7612
0.5643	37.0	1406	0.6165	0.7164
0.5575	38.0	1444	0.5896	0.7537
0.4765	39.0	1482	0.6218	0.7836
0.4907	40.0	1520	0.6871	0.7239
0.5793	41.0	1558	0.6265	0.7687
0.5973	42.0	1596	0.6109	0.7612
0.5992	43.0	1634	0.6445	0.7388
0.5493	44.0	1672	0.5963	0.7687
0.5219	45.0	1710	0.5561	0.7910
0.5304	46.0	1748	0.6020	0.7537
0.5922	47.0	1786	0.5408	0.7836
0.5281	48.0	1824	0.6592	0.7090
0.5409	49.0	1862	0.5544	0.8060
0.4781	50.0	1900	0.5589	0.7985
0.4613	51.0	1938	0.5880	0.7612
0.5482	52.0	1976	0.6220	0.7836
0.6053	53.0	2014	0.5958	0.7910
0.5964	54.0	2052	0.6246	0.7687
0.4663	55.0	2090	0.5476	0.7687
0.5479	56.0	2128	0.5846	0.7537
0.5953	57.0	2166	0.5644	0.7761
0.3778	58.0	2204	0.6519	0.7463
0.4446	59.0	2242	0.5997	0.7537
0.5251	60.0	2280	0.5263	0.7761
0.5537	61.0	2318	0.5772	0.7836
0.4679	62.0	2356	0.5421	0.8060
0.4351	63.0	2394	0.5701	0.7836
0.4324	64.0	2432	0.5260	0.7836
0.5806	65.0	2470	0.6130	0.7612
0.4776	66.0	2508	0.5693	0.7687
0.4108	67.0	2546	0.5180	0.7985
0.3895	68.0	2584	0.5509	0.7687
0.4178	69.0	2622	0.5365	0.7761
0.4593	70.0	2660	0.5168	0.7761
0.4258	71.0	2698	0.5182	0.7910
0.4273	72.0	2736	0.5040	0.8060
0.4166	73.0	2774	0.5033	0.7910
0.4199	74.0	2812	0.5319	0.7985
0.5292	75.0	2850	0.5206	0.8060
0.4566	76.0	2888	0.5389	0.8134
0.4747	77.0	2926	0.5050	0.8060
0.3542	78.0	2964	0.5782	0.7687
0.3562	79.0	3002	0.5190	0.7985
0.3806	80.0	3040	0.5066	0.7910
0.4889	81.0	3078	0.5292	0.7836
0.4626	82.0	3116	0.5193	0.7985
0.3558	83.0	3154	0.5166	0.7687
0.3881	84.0	3192	0.5112	0.7910
0.3715	85.0	3230	0.5349	0.7910
0.428	86.0	3268	0.5158	0.7985
0.3854	87.0	3306	0.5081	0.7985
0.3323	88.0	3344	0.5212	0.7985
0.3842	89.0	3382	0.4974	0.7985
0.386	90.0	3420	0.4942	0.8060
0.4029	91.0	3458	0.5034	0.8060
0.4314	92.0	3496	0.5172	0.8060
0.3171	93.0	3534	0.5088	0.8060
0.4585	94.0	3572	0.4855	0.7910
0.3147	95.0	3610	0.4954	0.7910
0.386	96.0	3648	0.4879	0.8060
0.2971	97.0	3686	0.5056	0.8134
0.2969	98.0	3724	0.5015	0.8134
0.3909	99.0	3762	0.4971	0.7985
0.373	100.0	3800	0.4946	0.7985

Framework versions

PEFT 0.13.2
Transformers 4.45.2
Pytorch 2.4.1
Datasets 3.0.1
Tokenizers 0.20.1

加载中…

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 0.0001

train_batch_size: 32

eval_batch_size: 1

seed: 42

optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08

lr_scheduler_type: linear

num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.7679	1.0	38	1.7781	0.3060
1.7069	2.0	76	1.5920	0.3806
1.5476	3.0	114	1.5259	0.3955
1.3803	4.0	152	1.3005	0.4851
1.3015	5.0	190	1.1809	0.5299
1.2157	6.0	228	1.1946	0.4851
1.1173	7.0	266	1.0651	0.5821
1.1138	8.0	304	1.0565	0.5522
1.0272	9.0	342	1.0698	0.5299
1.0161	10.0	380	1.0412	0.6045
1.0413	11.0	418	0.9661	0.5746
0.9973	12.0	456	1.0308	0.5746
0.9868	13.0	494	1.0036	0.6045
0.921	14.0	532	0.8507	0.6418
1.0229	15.0	570	0.8769	0.6567
0.9111	16.0	608	0.9176	0.6343
0.7588	17.0	646	0.8129	0.6343
0.8978	18.0	684	0.8439	0.6493
0.8469	19.0	722	0.7913	0.6791
0.8492	20.0	760	0.8750	0.6343
0.6988	21.0	798	0.8203	0.6716
0.9145	22.0	836	0.7737	0.6791
0.739	23.0	874	0.7167	0.7313
0.8208	24.0	912	0.7586	0.6716
0.6627	25.0	950	0.7110	0.7090
0.7707	26.0	988	0.7468	0.7015
0.5777	27.0	1026	0.7886	0.6642
0.5762	28.0	1064	0.6722	0.7612
0.6887	29.0	1102	0.7041	0.7015
0.7005	30.0	1140	0.6601	0.7239
0.6654	31.0	1178	0.7200	0.7090
0.5987	32.0	1216	0.7395	0.6791
0.571	33.0	1254	0.6474	0.7463
0.6837	34.0	1292	0.7545	0.6940
0.6899	35.0	1330	0.7235	0.7313
0.693	36.0	1368	0.6012	0.7612
0.5643	37.0	1406	0.6165	0.7164
0.5575	38.0	1444	0.5896	0.7537
0.4765	39.0	1482	0.6218	0.7836
0.4907	40.0	1520	0.6871	0.7239
0.5793	41.0	1558	0.6265	0.7687
0.5973	42.0	1596	0.6109	0.7612
0.5992	43.0	1634	0.6445	0.7388
0.5493	44.0	1672	0.5963	0.7687
0.5219	45.0	1710	0.5561	0.7910
0.5304	46.0	1748	0.6020	0.7537
0.5922	47.0	1786	0.5408	0.7836
0.5281	48.0	1824	0.6592	0.7090
0.5409	49.0	1862	0.5544	0.8060
0.4781	50.0	1900	0.5589	0.7985
0.4613	51.0	1938	0.5880	0.7612
0.5482	52.0	1976	0.6220	0.7836
0.6053	53.0	2014	0.5958	0.7910
0.5964	54.0	2052	0.6246	0.7687
0.4663	55.0	2090	0.5476	0.7687
0.5479	56.0	2128	0.5846	0.7537
0.5953	57.0	2166	0.5644	0.7761
0.3778	58.0	2204	0.6519	0.7463
0.4446	59.0	2242	0.5997	0.7537
0.5251	60.0	2280	0.5263	0.7761
0.5537	61.0	2318	0.5772	0.7836
0.4679	62.0	2356	0.5421	0.8060
0.4351	63.0	2394	0.5701	0.7836
0.4324	64.0	2432	0.5260	0.7836
0.5806	65.0	2470	0.6130	0.7612
0.4776	66.0	2508	0.5693	0.7687
0.4108	67.0	2546	0.5180	0.7985
0.3895	68.0	2584	0.5509	0.7687
0.4178	69.0	2622	0.5365	0.7761
0.4593	70.0	2660	0.5168	0.7761
0.4258	71.0	2698	0.5182	0.7910
0.4273	72.0	2736	0.5040	0.8060
0.4166	73.0	2774	0.5033	0.7910
0.4199	74.0	2812	0.5319	0.7985
0.5292	75.0	2850	0.5206	0.8060
0.4566	76.0	2888	0.5389	0.8134
0.4747	77.0	2926	0.5050	0.8060
0.3542	78.0	2964	0.5782	0.7687
0.3562	79.0	3002	0.5190	0.7985
0.3806	80.0	3040	0.5066	0.7910
0.4889	81.0	3078	0.5292	0.7836
0.4626	82.0	3116	0.5193	0.7985
0.3558	83.0	3154	0.5166	0.7687
0.3881	84.0	3192	0.5112	0.7910
0.3715	85.0	3230	0.5349	0.7910
0.428	86.0	3268	0.5158	0.7985
0.3854	87.0	3306	0.5081	0.7985
0.3323	88.0	3344	0.5212	0.7985
0.3842	89.0	3382	0.4974	0.7985
0.386	90.0	3420	0.4942	0.8060
0.4029	91.0	3458	0.5034	0.8060
0.4314	92.0	3496	0.5172	0.8060
0.3171	93.0	3534	0.5088	0.8060
0.4585	94.0	3572	0.4855	0.7910
0.3147	95.0	3610	0.4954	0.7910
0.386	96.0	3648	0.4879	0.8060
0.2971	97.0	3686	0.5056	0.8134
0.2969	98.0	3724	0.5015	0.8134
0.3909	99.0	3762	0.4971	0.7985
0.373	100.0	3800	0.4946	0.7985

Framework versions

PEFT 0.13.2

Transformers 4.45.2

Pytorch 2.4.1

Datasets 3.0.1

Tokenizers 0.20.1

fengchen025/dinov2-base_winter-wheat_fertiliser-identification_dnd-ww2020

dinov2-base_winter-wheat_fertiliser-identification_dnd-ww2020

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

dinov2-base_winter-wheat_fertiliser-identification_dnd-ww2020

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions