segformer-b0-finetuned-segments-construction-1

yiming19/segformer-b0-finetuned-segments-construction-1

segformer-b0-finetuned-segments-construction-1 is a image segmentation model from yiming19. Use it for the image segmentation task on the model card, and read the license before you ship it in a product. It is set up for transformers. The card lists the license as other.

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

Downloads · 30 days

1% of all-time downloads

All-time downloads

1.5K

Public

Repo size

134 MB

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.bin14.9 MB · 100%

At a glance

Task: Image Segmentation
Library: transformers
License: other
Model type: segformer
Access: Public
Created: Mar 12, 2023
Updated: Mar 19, 2023
SHA: c78e1694

Model card

From the Hugging Face model README

segformer-b0-finetuned-segments-construction-1

This model is a fine-tuned version of nvidia/mit-b0 on the yiming19/construction_place dataset. It achieves the following results on the evaluation set:

Loss: 0.2796
Mean Iou: 0.3218
Mean Accuracy: 0.5305
Overall Accuracy: 0.9276
Accuracy Unlabeled: nan
Accuracy Ruler: 0.8954
Accuracy Socket: 0.0
Accuracy Wall: 0.9644
Accuracy Window: nan
Accuracy Heater: nan
Accuracy Floor: 0.6710
Accuracy Ceiling: 0.0
Accuracy Skirting: nan
Accuracy Door: 0.6525
Accuracy Light: nan
Iou Unlabeled: nan
Iou Ruler: 0.7222
Iou Socket: 0.0
Iou Wall: 0.9553
Iou Window: 0.0
Iou Heater: nan
Iou Floor: 0.2630
Iou Ceiling: 0.0
Iou Skirting: 0.0
Iou Door: 0.6342
Iou Light: nan

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: 6e-05
train_batch_size: 2
eval_batch_size: 2
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Unlabeled	Accuracy Ruler	Accuracy Wall	Accuracy Window	Accuracy Heater	Accuracy Floor	Accuracy Skirting	Accuracy Door	Accuracy Light	Iou Unlabeled	Iou Ruler	Iou Wall	Iou Heater	Iou Floor	Iou Door	Iou Light
1.8126	1.43	20	2.1233	0.1955	0.5160	0.8448	nan	0.8191	0.8868	nan	nan	0.9618	nan	0.4281	nan	0.0	0.5555	0.8845	0.0	0.2971	0.4135	0.0
1.905	2.86	40	1.3611	0.1827	0.4921	0.8505	nan	0.9275	0.9139	nan	nan	0.9627	nan	0.1484	nan	nan	0.5404	0.9095	0.0	0.2289	0.1484	0.0
1.1072	4.29	60	1.0502	0.2327	0.5517	0.8903	nan	0.9108	0.9266	nan	nan	0.9367	nan	0.5360	nan	nan	0.5301	0.9206	0.0	0.3475	0.5284	0.0
1.0076	5.71	80	0.8802	0.2744	0.5609	0.9089	nan	0.8208	0.9410	nan	nan	0.9532	nan	0.6505	nan	nan	0.5500	0.9277	0.0	0.3688	0.6227	nan
1.5533	7.14	100	0.8991	0.2846	0.5514	0.8878	nan	0.8918	0.9243	nan	nan	0.9591	nan	0.5332	nan	nan	0.5262	0.9169	nan	0.3209	0.5132	nan
0.9912	8.57	120	0.9340	0.2891	0.5652	0.8854	nan	0.9478	0.9151	nan	nan	0.9438	nan	0.5844	nan	nan	0.5059	0.9098	nan	0.3424	0.5544	nan
0.784	10.0	140	0.7017	0.3140	0.5984	0.9173	nan	0.9136	0.9305	nan	nan	0.8971	nan	0.8493	nan	nan	0.5324	0.9224	0.0	0.5805	0.7909	nan
0.5636	11.43	160	0.6925	0.3573	0.5978	0.9280	nan	0.8714	0.9412	nan	nan	0.8868	nan	0.8876	nan	nan	0.5701	0.9308	nan	0.5638	0.7935	nan
1.0692	12.86	180	0.7313	0.2931	0.5724	0.8981	nan	0.9587	0.9231	nan	nan	0.8880	nan	0.6647	nan	nan	0.4988	0.9182	nan	0.3342	0.5932	nan
0.7603	14.29	200	0.6907	0.2577	0.5744	0.9001	nan	0.9619	0.9251	nan	nan	0.8930	nan	0.6661	nan	nan	0.4939	0.9208	0.0	0.3219	0.5824	nan
0.9509	15.71	220	0.5110	0.3682	0.6069	0.9324	nan	0.9355	0.9417	nan	nan	0.8453	nan	0.9191	nan	nan	0.5671	0.9334	nan	0.6050	0.8403	nan
0.4254	17.14	240	0.5925	0.2961	0.5629	0.9023	nan	0.9646	0.9295	nan	nan	0.8261	nan	0.6569	nan	nan	0.5302	0.9243	nan	0.3138	0.6009	nan
0.3839	18.57	260	0.4226	0.3537	0.5479	0.9367	nan	0.9108	0.9540	nan	nan	0.5102	nan	0.9124	nan	nan	0.6643	0.9426	nan	0.3868	0.8361	nan
0.7441	20.0	280	0.5084	0.3533	0.5993	0.9277	nan	0.9691	0.9391	nan	nan	0.8075	nan	0.8801	nan	nan	0.5527	0.9333	nan	0.5197	0.8208	nan
0.4374	21.43	300	0.4683	0.3038	0.5549	0.9173	nan	0.9662	0.9480	nan	nan	0.7594	nan	0.6558	nan	nan	0.6024	0.9419	nan	0.2804	0.6056	nan
0.6224	22.86	320	0.4100	0.3810	0.5960	0.9374	nan	0.9704	0.9460	nan	nan	0.7131	nan	0.9467	nan	nan	0.5986	0.9401	nan	0.6197	0.8898	nan
0.4473	24.29	340	0.3933	0.3368	0.5431	0.9336	nan	0.9212	0.9620	nan	nan	0.6197	nan	0.7556	nan	nan	0.7221	0.9521	nan	0.2958	0.7245	nan
0.3364	25.71	360	0.4336	0.2976	0.5125	0.9134	nan	0.9408	0.9544	nan	nan	0.6075	nan	0.5721	nan	nan	0.6918	0.9481	nan	0.1998	0.5411	nan
0.281	27.14	380	0.3795	0.3689	0.5760	0.9420	nan	0.9250	0.9589	nan	nan	0.6859	nan	0.8863	nan	nan	0.7108	0.9518	nan	0.4576	0.8305	nan
0.3198	28.57	400	0.4023	0.3158	0.5143	0.9238	nan	0.9120	0.9610	nan	nan	0.5580	nan	0.6550	nan	nan	0.7210	0.9519	nan	0.2238	0.6293	nan
0.4624	30.0	420	0.3565	0.3770	0.5774	0.9475	nan	0.9408	0.9613	nan	nan	0.6287	nan	0.9337	nan	nan	0.6855	0.9539	nan	0.4943	0.8827	nan
0.2356	31.43	440	0.3940	0.3100	0.5349	0.9221	nan	0.9268	0.9602	nan	nan	0.7187	nan	0.6040	nan	nan	0.7005	0.9536	nan	0.2474	0.5781	nan
0.3931	32.86	460	0.3516	0.3162	0.5570	0.9258	nan	0.9338	0.9598	nan	nan	0.8124	nan	0.6362	nan	nan	0.6824	0.9542	nan	0.2888	0.6040	nan
0.2431	34.29	480	0.4011	0.2955	0.5291	0.9138	nan	0.9242	0.9583	nan	nan	0.7864	nan	0.5058	nan	nan	0.6954	0.9520	nan	0.2331	0.4832	nan
0.2131	35.71	500	0.2847	0.3764	0.5613	0.9487	nan	0.8877	0.9679	nan	nan	0.6103	nan	0.9020	nan	nan	0.7330	0.9571	nan	0.4539	0.8669	nan
0.4151	37.14	520	0.3176	0.3186	0.5239	0.9256	nan	0.8930	0.9640	nan	nan	0.6505	nan	0.6356	nan	nan	0.7251	0.9544	nan	0.2507	0.6187	nan
0.2408	38.57	540	0.3267	0.3071	0.5361	0.9208	nan	0.9264	0.9600	nan	nan	0.7441	nan	0.5859	nan	nan	0.6868	0.9538	nan	0.2526	0.5635	nan
0.2274	40.0	560	0.2875	0.3396	0.5471	0.9349	nan	0.9098	0.9626	nan	nan	0.6456	nan	0.7649	nan	nan	0.7018	0.9547	nan	0.3216	0.7387	nan
0.2452	41.43	580	0.2998	0.3181	0.5357	0.9279	nan	0.9089	0.9642	nan	nan	0.6932	nan	0.6480	nan	nan	0.7057	0.9562	nan	0.2578	0.6252	nan
0.2922	42.86	600	0.2957	0.3131	0.5246	0.9255	nan	0.9056	0.9643	nan	nan	0.6535	nan	0.6242	nan	nan	0.7103	0.9563	nan	0.2347	0.6037	nan
0.3704	44.29	620	0.3290	0.3172	0.5429	0.9247	nan	0.9246	0.9583	nan	nan	0.7123	nan	0.6621	nan	nan	0.6856	0.9527	nan	0.2707	0.6286	nan
0.2482	45.71	640	0.2995	0.3251	0.5368	0.9276	nan	0.9018	0.9617	nan	nan	0.6795	nan	0.6779	nan	nan	0.7154	0.9538	nan	0.2790	0.6528	nan
0.2798	47.14	660	0.2808	0.3323	0.5374	0.9319	nan	0.8938	0.9644	nan	nan	0.6554	nan	0.7110	nan	nan	0.7218	0.9554	nan	0.2919	0.6894	nan
0.2746	48.57	680	0.2695	0.3265	0.5341	0.9299	nan	0.8947	0.9642	nan	nan	0.6597	nan	0.6861	nan	nan	0.7198	0.9554	nan	0.2735	0.6633	nan
0.2169	50.0	700	0.2796	0.3218	0.5305	0.9276	nan	0.8954	0.9644	nan	nan	0.6710	nan	0.6525	nan	nan	0.7222	0.9553	nan	0.2630	0.6342	nan

Framework versions

Transformers 4.25.1
Pytorch 1.13.1
Datasets 2.10.1
Tokenizers 0.13.0.dev0

Task: Image Segmentation
Library: transformers
Type: segformer
License: other
Created: Mar 12, 2023
Updated: Mar 19, 2023