You can deploy this model and start earning money today!

NICOPOI-9/
segformer-b5-finetuned-ade20k-morphpadver1-hgo-coord-v9_mix_resample_40epochs

Name: segformer-b5-finetuned-ade20k-morphpadver1-hgo-coord-v9_mix_resample_40epochs
Availability: InStock

generated_from_trainer

nvidia/segformer-b5-finetuned-ade-640-640

finetune

other

endpoints_compatible

Views

No views yet

segformer-b5-finetuned-ade20k-morphpadver1-hgo-coord-v9_mix_resample_40epochs

This model is a fine-tuned version of nvidia/segformer-b5-finetuned-ade-640-640 on the NICOPOI-9/morphpad_coord_hgo_512_4class_v2 dataset. It achieves the following results on the evaluation set:

Loss: 0.5955
Mean Iou: 0.7753
Mean Accuracy: 0.8720
Overall Accuracy: 0.8743
Accuracy 0-0: 0.8115
Accuracy 0-90: 0.9059
Accuracy 90-0: 0.8813
Accuracy 90-90: 0.8893
Iou 0-0: 0.7564
Iou 0-90: 0.7853
Iou 90-0: 0.7894
Iou 90-90: 0.7700

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: 1
eval_batch_size: 1
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
num_epochs: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy 0-0	Accuracy 0-90	Accuracy 90-0	Accuracy 90-90	Iou 0-0	Iou 0-90	Iou 90-0	Iou 90-90
1.162	1.3638	4000	1.1730	0.3007	0.4629	0.4758	0.3342	0.5531	0.6459	0.3186	0.2606	0.3421	0.3576	0.2426
0.395	2.7276	8000	0.9973	0.4133	0.5816	0.5930	0.4473	0.6859	0.7131	0.4801	0.3740	0.4579	0.4501	0.3713
0.3669	4.0914	12000	0.7348	0.5875	0.7355	0.7422	0.6643	0.7867	0.8261	0.6650	0.5753	0.6040	0.6029	0.5679
0.1682	5.4552	16000	0.5960	0.6669	0.7992	0.8012	0.7239	0.7890	0.8507	0.8333	0.6476	0.6794	0.6793	0.6613
0.375	6.8190	20000	0.6153	0.6717	0.8059	0.8041	0.8256	0.7924	0.7818	0.8238	0.6541	0.6957	0.6718	0.6652
0.0555	8.1827	24000	0.5139	0.7146	0.8366	0.8337	0.8633	0.7891	0.8253	0.8688	0.7066	0.7163	0.7193	0.7164
0.0922	9.5465	28000	0.5495	0.7061	0.8232	0.8292	0.7713	0.8948	0.8771	0.7495	0.6945	0.7226	0.7186	0.6887
0.0843	10.9103	32000	0.5586	0.7233	0.8358	0.8409	0.7749	0.8844	0.8935	0.7903	0.7022	0.7486	0.7297	0.7127
0.0548	12.2741	36000	0.5068	0.7430	0.8519	0.8536	0.7967	0.8802	0.8552	0.8754	0.7230	0.7646	0.7530	0.7314
0.0703	13.6379	40000	0.4079	0.7834	0.8771	0.8791	0.8567	0.9133	0.8792	0.8592	0.7719	0.7917	0.7903	0.7796
0.0532	15.0017	44000	0.4843	0.7503	0.8549	0.8588	0.7706	0.8987	0.8868	0.8634	0.7228	0.7687	0.7628	0.7471
0.0702	16.3655	48000	0.4905	0.7572	0.8605	0.8626	0.8501	0.8941	0.8697	0.8280	0.7580	0.7694	0.7678	0.7335
0.0389	17.7293	52000	0.4224	0.7777	0.8726	0.8760	0.8338	0.9119	0.9028	0.8418	0.7611	0.7940	0.7879	0.7680
0.0441	19.0931	56000	0.5241	0.7439	0.8517	0.8537	0.8437	0.8851	0.8598	0.8180	0.7361	0.7455	0.7595	0.7346
0.0278	20.4569	60000	0.4326	0.7632	0.8623	0.8669	0.8353	0.9140	0.9066	0.7936	0.7663	0.7758	0.7795	0.7314
11.8254	21.8207	64000	0.4380	0.7765	0.8743	0.8747	0.8708	0.8755	0.8820	0.8687	0.7650	0.7907	0.7788	0.7716
0.0239	23.1845	68000	0.4872	0.7606	0.8630	0.8650	0.8762	0.8800	0.8908	0.8051	0.7477	0.7866	0.7649	0.7434
0.128	24.5482	72000	0.4604	0.7731	0.8700	0.8730	0.8505	0.9075	0.8951	0.8268	0.7637	0.7875	0.7842	0.7568
0.0482	25.9120	76000	0.5039	0.7559	0.8602	0.8620	0.8134	0.8757	0.8786	0.8732	0.7406	0.7746	0.7678	0.7407
0.0566	27.2758	80000	0.5290	0.7681	0.8671	0.8701	0.8100	0.9024	0.8928	0.8631	0.7410	0.7847	0.7865	0.7603
0.0148	28.6396	84000	0.7038	0.7441	0.8530	0.8541	0.8293	0.8652	0.8607	0.8570	0.7317	0.7596	0.7541	0.7311
0.0164	30.0034	88000	0.5058	0.7846	0.8776	0.8803	0.8808	0.9156	0.8982	0.8158	0.7733	0.8034	0.7946	0.7671
0.0019	31.3672	92000	0.5421	0.7682	0.8669	0.8697	0.8498	0.9145	0.8761	0.8272	0.7508	0.7768	0.7832	0.7619
0.0028	32.7310	96000	0.6268	0.7829	0.8773	0.8791	0.8674	0.9000	0.8917	0.8501	0.7662	0.7947	0.7993	0.7712
0.0103	34.0948	100000	0.5294	0.7730	0.8697	0.8731	0.8520	0.9142	0.8966	0.8160	0.7595	0.7844	0.7930	0.7552
0.0508	35.4586	104000	0.5242	0.7829	0.8763	0.8790	0.8294	0.9140	0.8913	0.8704	0.7691	0.7895	0.7964	0.7766
0.0102	36.8224	108000	0.6011	0.7772	0.8716	0.8756	0.8096	0.9239	0.8983	0.8546	0.7583	0.7816	0.7976	0.7714
0.0311	38.1862	112000	0.6471	0.7771	0.8727	0.8754	0.8152	0.9135	0.8833	0.8789	0.7553	0.7848	0.7924	0.7760
0.0089	39.5499	116000	0.5955	0.7753	0.8720	0.8743	0.8115	0.9059	0.8813	0.8893	0.7564	0.7853	0.7894	0.7700

Framework versions

Transformers 4.48.3
Pytorch 2.1.0
Datasets 3.2.0
Tokenizers 0.21.0

NICOPOI-9/segformer-b5-finetuned-ade20k-morphpadver1-hgo-coord-v9_mix_resample_40epochs

segformer-b5-finetuned-ade20k-morphpadver1-hgo-coord-v9_mix_resample_40epochs

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

NICOPOI-9/
segformer-b5-finetuned-ade20k-morphpadver1-hgo-coord-v9_mix_resample_40epochs