segformer-b0-scene-parse-150

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

• Loss: 4.1486
• Mean Iou: 0.0000
• Mean Accuracy: 0.0001
• Overall Accuracy: 0.0001

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: 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: 2

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy
4.9721	0.025	1	5.0059	0.0007	0.0148	0.0063
4.9475	0.05	2	5.0027	0.0007	0.0140	0.0060
4.9457	0.075	3	4.9996	0.0008	0.0144	0.0063
4.9923	0.1	4	4.9959	0.0008	0.0142	0.0063
5.0016	0.125	5	4.9912	0.0009	0.0153	0.0069
4.9753	0.15	6	4.9876	0.0008	0.0149	0.0070
4.8799	0.175	7	4.9824	0.0006	0.0108	0.0051
4.9689	0.2	8	4.9767	0.0006	0.0095	0.0045
4.9046	0.225	9	4.9697	0.0006	0.0093	0.0044
4.8772	0.25	10	4.9629	0.0005	0.0074	0.0035
4.7839	0.275	11	4.9574	0.0005	0.0084	0.0038
4.9577	0.3	12	4.9500	0.0005	0.0074	0.0031
4.8491	0.325	13	4.9411	0.0004	0.0067	0.0026
4.8449	0.35	14	4.9340	0.0004	0.0070	0.0026
4.8899	0.375	15	4.9229	0.0003	0.0051	0.0020
4.7924	0.4	16	4.9163	0.0003	0.0050	0.0019
4.7651	0.425	17	4.9072	0.0003	0.0043	0.0016
4.7951	0.45	18	4.8953	0.0002	0.0035	0.0013
4.7355	0.475	19	4.8865	0.0002	0.0028	0.0010
4.7461	0.5	20	4.8723	0.0002	0.0026	0.0008
4.704	0.525	21	4.8606	0.0002	0.0022	0.0007
4.7775	0.55	22	4.8484	0.0001	0.0017	0.0006
4.7081	0.575	23	4.8331	0.0001	0.0013	0.0004
4.7681	0.6	24	4.8187	0.0001	0.0009	0.0003
4.7297	0.625	25	4.8037	0.0001	0.0008	0.0003
4.8181	0.65	26	4.7882	0.0001	0.0007	0.0002
4.833	0.675	27	4.7748	0.0001	0.0006	0.0002
4.7222	0.7	28	4.7575	0.0000	0.0004	0.0002
4.6457	0.725	29	4.7389	0.0000	0.0004	0.0002
4.7089	0.75	30	4.7236	0.0000	0.0005	0.0002
4.543	0.775	31	4.7079	0.0001	0.0006	0.0002
4.5529	0.8	32	4.6963	0.0001	0.0006	0.0003
4.7005	0.825	33	4.6759	0.0001	0.0006	0.0003
4.4735	0.85	34	4.6630	0.0001	0.0008	0.0004
4.6562	0.875	35	4.6468	0.0001	0.0009	0.0004
4.5902	0.9	36	4.6274	0.0001	0.0008	0.0004
4.4974	0.925	37	4.6125	0.0001	0.0008	0.0004
4.524	0.95	38	4.5967	0.0001	0.0011	0.0005
4.5527	0.975	39	4.5826	0.0001	0.0011	0.0005
4.5165	1.0	40	4.5627	0.0001	0.0010	0.0005
4.6337	1.025	41	4.5502	0.0001	0.0012	0.0006
4.4551	1.05	42	4.5425	0.0001	0.0012	0.0005
4.4697	1.075	43	4.5294	0.0001	0.0006	0.0003
4.4967	1.1	44	4.5065	0.0001	0.0007	0.0003
4.4839	1.125	45	4.4896	0.0000	0.0004	0.0002
4.4394	1.15	46	4.4699	0.0000	0.0003	0.0001
4.4557	1.175	47	4.4511	0.0000	0.0003	0.0001
4.2669	1.2	48	4.4475	0.0000	0.0003	0.0001
4.3143	1.225	49	4.4325	0.0000	0.0002	0.0001
4.4519	1.25	50	4.4195	0.0000	0.0002	0.0001
4.5376	1.275	51	4.4092	0.0000	0.0001	0.0001
4.2617	1.3	52	4.4058	0.0000	0.0001	0.0000
4.2813	1.325	53	4.3936	0.0000	0.0001	0.0000
4.5218	1.35	54	4.3867	0.0000	0.0002	0.0001
4.4805	1.375	55	4.3691	0.0000	0.0002	0.0001
4.184	1.4	56	4.3574	0.0000	0.0002	0.0001
4.2208	1.425	57	4.3606	0.0000	0.0001	0.0001
4.5288	1.45	58	4.3579	0.0000	0.0001	0.0001
4.3959	1.475	59	4.3421	0.0000	0.0001	0.0000
4.2578	1.5	60	4.3403	0.0000	0.0001	0.0000
4.3504	1.525	61	4.3307	0.0000	0.0001	0.0000
4.2364	1.55	62	4.3177	0.0000	0.0001	0.0000
4.3248	1.575	63	4.2924	0.0000	0.0000	0.0000
4.2771	1.6	64	4.2698	0.0000	0.0000	0.0000
4.2447	1.625	65	4.2533	0.0000	0.0000	0.0000
4.4481	1.65	66	4.2418	0.0000	0.0000	0.0000
4.1369	1.675	67	4.2374	0.0000	0.0000	0.0000
4.2266	1.7	68	4.2305	0.0000	0.0000	0.0000
4.5113	1.725	69	4.2225	0.0000	0.0000	0.0000
4.4737	1.75	70	4.2077	0.0000	0.0000	0.0000
4.4571	1.775	71	4.1960	0.0000	0.0001	0.0000
4.2179	1.8	72	4.1824	0.0000	0.0001	0.0000
4.5426	1.825	73	4.1654	0.0000	0.0002	0.0001
4.3632	1.85	74	4.1572	0.0000	0.0002	0.0001
4.2132	1.875	75	4.1628	0.0000	0.0002	0.0001
4.3442	1.9	76	4.1621	0.0000	0.0001	0.0000
4.4454	1.925	77	4.1647	0.0000	0.0001	0.0000
4.1564	1.95	78	4.1691	0.0000	0.0001	0.0000
4.5028	1.975	79	4.1513	0.0000	0.0002	0.0001
4.3814	2.0	80	4.1486	0.0000	0.0001	0.0001

Framework versions

• Transformers 4.55.2
• Pytorch 2.6.0+cu124
• Datasets 4.0.0
• Tokenizers 0.21.4