This model is a fine-tuned version of nvidia/segformer-b5-finetuned-ade-640-640 on the AllanK24/apple-dms-materials dataset. It achieves the following results on the evaluation set:

Loss: 1.2034
Mean Iou: 0.4595
Mean Accuracy: 0.5491
Overall Accuracy: 0.8239
Accuracy Animal Skin: 0.5919
Iou Animal Skin: 0.4132
Accuracy Bone Teeth Horn: 0.0
Iou Bone Teeth Horn: 0.0
Accuracy Brickwork: 0.7186
Iou Brickwork: 0.5878
Accuracy Cardboard: 0.7042
Iou Cardboard: 0.5463
Accuracy Carpet Rug: 0.8807
Iou Carpet Rug: 0.7729
Accuracy Ceiling Tile: 0.8715
Iou Ceiling Tile: 0.7782
Accuracy Ceramic: 0.7940
Iou Ceramic: 0.6572
Accuracy Chalkboard Blackboard: 0.7596
Iou Chalkboard Blackboard: 0.6748
Accuracy Clutter: 0.0002
Iou Clutter: 0.0002
Accuracy Concrete: 0.5929
Iou Concrete: 0.4385
Accuracy Cork Corkboard: 0.0
Iou Cork Corkboard: 0.0
Accuracy Engineered Stone: 0.0222
Iou Engineered Stone: 0.0222
Accuracy Fabric Cloth: 0.9068
Iou Fabric Cloth: 0.7972
Accuracy Fiberglass Wool: 0.0
Iou Fiberglass Wool: 0.0
Accuracy Fire: 0.4037
Iou Fire: 0.3584
Accuracy Foliage: 0.9342
Iou Foliage: 0.8377
Accuracy Food: 0.8947
Iou Food: 0.7873
Accuracy Fur: 0.9217
Iou Fur: 0.8129
Accuracy Gemstone Quartz: 0.0520
Iou Gemstone Quartz: 0.0381
Accuracy Glass: 0.7609
Iou Glass: 0.6317
Accuracy Hair: 0.8536
Iou Hair: 0.7276
Accuracy Ice: 0.4306
Iou Ice: 0.3222
Accuracy Leather: 0.6377
Iou Leather: 0.5276
Accuracy Liquid Non-water: 0.0682
Iou Liquid Non-water: 0.0655
Accuracy Metal: 0.5094
Iou Metal: 0.3766
Accuracy Mirror: 0.6544
Iou Mirror: 0.5306
Accuracy Paint Plaster Enamel: 0.8744
Iou Paint Plaster Enamel: 0.7547
Accuracy Paper: 0.7554
Iou Paper: 0.6071
Accuracy Pearl: 0.0
Iou Pearl: 0.0
Accuracy Photograph Painting: 0.4862
Iou Photograph Painting: 0.3616
Accuracy Plastic Clear: 0.3503
Iou Plastic Clear: 0.2723
Accuracy Plastic Non-clear: 0.5506
Iou Plastic Non-clear: 0.4171
Accuracy Rubber Latex: 0.3110
Iou Rubber Latex: 0.2692
Accuracy Sand: 0.7507
Iou Sand: 0.5800
Accuracy Skin Lips: 0.8420
Iou Skin Lips: 0.7321
Accuracy Sky: 0.9571
Iou Sky: 0.9293
Accuracy Snow: 0.7200
Iou Snow: 0.6666
Accuracy Soap: 0.0
Iou Soap: 0.0
Accuracy Soil Mud: 0.5783
Iou Soil Mud: 0.4410
Accuracy Sponge: 0.0
Iou Sponge: 0.0
Accuracy Stone Natural: 0.6875
Iou Stone Natural: 0.5539
Accuracy Stone Polished: 0.3631
Iou Stone Polished: 0.3075
Accuracy Styrofoam: 0.0
Iou Styrofoam: 0.0
Accuracy Tile: 0.8126
Iou Tile: 0.6766
Accuracy Wallpaper: 0.6609
Iou Wallpaper: 0.5012
Accuracy Water: 0.9142
Iou Water: 0.8189
Accuracy Wax: 0.5010
Iou Wax: 0.4732
Accuracy Whiteboard: 0.8216
Iou Whiteboard: 0.6930
Accuracy Wicker: 0.6048
Iou Wicker: 0.5193
Accuracy Wood: 0.8714
Iou Wood: 0.7396
Accuracy Wood Tree: 0.5341
Iou Wood Tree: 0.3701
Accuracy Asphalt: 0.6402
Iou Asphalt: 0.5071

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: 0.001
train_batch_size: 32
eval_batch_size: 16
seed: 42
distributed_type: multi-GPU
num_devices: 8
total_train_batch_size: 256
total_eval_batch_size: 128
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 0.1
num_epochs: 20
label_smoothing_factor: 0.1

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Animal Skin	Iou Animal Skin	Accuracy Brickwork	Iou Brickwork	Accuracy Cardboard	Iou Cardboard	Accuracy Carpet Rug	Iou Carpet Rug	Accuracy Ceiling Tile	Iou Ceiling Tile	Accuracy Ceramic	Iou Ceramic	Accuracy Chalkboard Blackboard	Iou Chalkboard Blackboard	Accuracy Clutter	Iou Clutter	Accuracy Concrete	Iou Concrete	Accuracy Engineered Stone	Iou Engineered Stone	Accuracy Fabric Cloth	Iou Fabric Cloth	Accuracy Fire	Iou Fire	Accuracy Foliage	Iou Foliage	Accuracy Food	Iou Food	Accuracy Fur	Iou Fur	Accuracy Gemstone Quartz	Iou Gemstone Quartz	Accuracy Glass	Iou Glass	Accuracy Hair	Iou Hair	Accuracy Ice	Iou Ice	Accuracy Leather	Iou Leather	Accuracy Liquid Non-water	Iou Liquid Non-water	Accuracy Metal	Iou Metal	Accuracy Mirror	Iou Mirror	Accuracy Paint Plaster Enamel	Iou Paint Plaster Enamel	Accuracy Paper	Iou Paper	Accuracy Photograph Painting	Iou Photograph Painting	Accuracy Plastic Clear	Iou Plastic Clear	Accuracy Plastic Non-clear	Iou Plastic Non-clear	Accuracy Rubber Latex	Iou Rubber Latex	Accuracy Sand	Iou Sand	Accuracy Skin Lips	Iou Skin Lips	Accuracy Sky	Iou Sky	Accuracy Snow	Iou Snow	Accuracy Soil Mud	Iou Soil Mud	Accuracy Stone Natural	Iou Stone Natural	Accuracy Stone Polished	Iou Stone Polished	Accuracy Tile	Iou Tile	Accuracy Wallpaper	Iou Wallpaper	Accuracy Water	Iou Water	Accuracy Wax	Iou Wax	Accuracy Whiteboard	Iou Whiteboard	Accuracy Wicker	Iou Wicker	Accuracy Wood	Iou Wood	Accuracy Wood Tree	Iou Wood Tree	Accuracy Asphalt	Iou Asphalt
2.4072	1.7045	150	1.4449	0.2578	0.3205	0.7517	0.0	0.0	0.4948	0.3764	0.2556	0.2399	0.7925	0.6477	0.8360	0.7015	0.6684	0.5413	0.0	0.0	0.0	0.0	0.4411	0.2861	0.0	0.0	0.8774	0.7198	0.0	0.0	0.9342	0.7885	0.8921	0.7134	0.9111	0.7167	0.0	0.0	0.6624	0.5147	0.7914	0.6443	0.0	0.0	0.0095	0.0094	0.0	0.0	0.2845	0.2221	0.3094	0.2818	0.8804	0.6824	0.7382	0.5148	0.3239	0.2400	0.0046	0.0046	0.4120	0.3004	0.0	0.0	0.0	0.0	0.7532	0.6192	0.9614	0.9131	0.0	0.0	0.4185	0.2607	0.5552	0.4013	0.0011	0.0011	0.6613	0.5585	0.0000	0.0000	0.9010	0.7757	0.0	0.0	0.0805	0.0802	0.0002	0.0002	0.8155	0.6505	0.0000	0.0000	0.0009	0.0009
1.3990	3.4091	300	1.3133	0.3801	0.4630	0.7905	0.4450	0.3464	0.6816	0.5528	0.4601	0.4046	0.8308	0.7031	0.7919	0.7265	0.6966	0.5831	0.7040	0.5738	0.0	0.0	0.5429	0.3618	0.0	0.0	0.9026	0.7472	0.0	0.0	0.9325	0.8212	0.9245	0.7434	0.9159	0.7748	0.0	0.0	0.7021	0.5645	0.8353	0.6965	0.0	0.0	0.3668	0.3479	0.0	0.0	0.4559	0.3104	0.3370	0.3199	0.8592	0.7260	0.7652	0.5610	0.4873	0.3273	0.2128	0.1725	0.4911	0.3617	0.1564	0.1468	0.4271	0.3653	0.8131	0.6772	0.9616	0.9266	0.6957	0.6421	0.6067	0.4118	0.7065	0.4841	0.1680	0.1595	0.6999	0.6089	0.4695	0.4052	0.9016	0.7899	0.0	0.0	0.6906	0.6108	0.4598	0.4157	0.8465	0.7007	0.5139	0.2878	0.6187	0.4081
1.2984	5.1136	450	1.2672	0.4053	0.4935	0.8027	0.5377	0.4599	0.7569	0.5266	0.5888	0.4684	0.8900	0.7097	0.9150	0.7043	0.7718	0.6052	0.7690	0.6036	0.0	0.0	0.4767	0.3858	0.0	0.0	0.9001	0.7640	0.1058	0.1058	0.9395	0.8231	0.8951	0.7786	0.9156	0.7769	0.0030	0.0030	0.6977	0.5836	0.8343	0.7132	0.0079	0.0079	0.5743	0.4689	0.0	0.0	0.4227	0.3237	0.6460	0.4528	0.8694	0.7319	0.7818	0.5750	0.4853	0.3335	0.2137	0.1878	0.5107	0.3812	0.2340	0.2144	0.6079	0.5246	0.8271	0.7010	0.9481	0.9233	0.7303	0.6386	0.5652	0.4197	0.6280	0.5240	0.1726	0.1623	0.7902	0.6499	0.5501	0.4354	0.9058	0.8073	0.0	0.0	0.7339	0.6428	0.5207	0.4605	0.8336	0.7149	0.4709	0.3277	0.6325	0.4558
1.2528	6.8182	600	1.2495	0.4196	0.4977	0.8087	0.5753	0.3916	0.6549	0.5835	0.6285	0.5044	0.8440	0.7424	0.8285	0.7575	0.7541	0.6171	0.7248	0.6579	0.0	0.0	0.5763	0.3990	0.0	0.0	0.8926	0.7778	0.3272	0.3073	0.9401	0.8258	0.8886	0.7857	0.9280	0.7899	0.0260	0.0258	0.7000	0.5930	0.8390	0.7188	0.1350	0.1325	0.5763	0.4895	0.0	0.0	0.4931	0.3470	0.4065	0.3754	0.8799	0.7388	0.7473	0.5911	0.4561	0.3425	0.2520	0.2200	0.5689	0.3979	0.2475	0.2258	0.4387	0.3781	0.8365	0.7118	0.9476	0.9226	0.8025	0.6941	0.4164	0.3637	0.6760	0.5351	0.3475	0.2733	0.7974	0.6507	0.5979	0.4890	0.9093	0.8148	0.0	0.0	0.8104	0.6805	0.5475	0.4783	0.8541	0.7223	0.5126	0.3208	0.4953	0.4457
1.2135	8.5227	750	1.2290	0.4342	0.5214	0.8143	0.6203	0.4142	0.7305	0.5797	0.6914	0.5401	0.8752	0.7536	0.8329	0.7580	0.7703	0.6371	0.7187	0.6591	0.0	0.0	0.5394	0.4299	0.0000	0.0000	0.8946	0.7815	0.3434	0.3172	0.9311	0.8338	0.8779	0.7703	0.9144	0.7920	0.0085	0.0085	0.7178	0.5973	0.8442	0.7254	0.3141	0.2712	0.6169	0.4893	0.0	0.0	0.5195	0.3628	0.6515	0.4918	0.8759	0.7443	0.7334	0.5965	0.4649	0.3484	0.3211	0.2609	0.5688	0.4031	0.2562	0.2371	0.6650	0.5444	0.8330	0.7190	0.9577	0.9272	0.8162	0.6634	0.6250	0.4663	0.6883	0.5439	0.3291	0.2834	0.7752	0.6697	0.6250	0.4744	0.8906	0.8042	0.0001	0.0001	0.7257	0.6293	0.5792	0.4998	0.8637	0.7307	0.4777	0.3581	0.6278	0.4623
1.1877	10.2273	900	1.2292	0.4442	0.5338	0.8165	0.5770	0.4171	0.7129	0.5991	0.6820	0.5432	0.8799	0.7584	0.8554	0.7655	0.8171	0.6304	0.7526	0.6774	0.0	0.0	0.4948	0.4033	0.0050	0.0050	0.8954	0.7887	0.4021	0.3626	0.9286	0.8346	0.9221	0.7890	0.9204	0.8061	0.1092	0.0795	0.7363	0.6191	0.8521	0.7181	0.3807	0.2694	0.6077	0.4966	0.0009	0.0009	0.4906	0.3649	0.6234	0.5035	0.8667	0.7466	0.7384	0.6016	0.4760	0.3539	0.2805	0.2332	0.5477	0.4037	0.2761	0.2402	0.7465	0.5730	0.8425	0.7246	0.9513	0.9237	0.5854	0.5711	0.5269	0.4192	0.7194	0.5437	0.3607	0.2903	0.8077	0.6618	0.7178	0.4896	0.9197	0.8004	0.3123	0.3085	0.8222	0.6938	0.5565	0.4956	0.8856	0.7242	0.5326	0.3672	0.6411	0.5012
1.1586	11.9318	1050	1.2173	0.4516	0.5397	0.8192	0.5662	0.4071	0.6991	0.5886	0.6916	0.5322	0.8844	0.7597	0.8293	0.7639	0.8030	0.6546	0.7217	0.6640	0.0	0.0	0.5317	0.4116	0.0127	0.0127	0.9065	0.7892	0.4642	0.4039	0.9430	0.8305	0.8858	0.7794	0.9360	0.8210	0.0316	0.0272	0.7733	0.6167	0.8573	0.7192	0.4196	0.3409	0.5973	0.4944	0.0010	0.0010	0.4949	0.3692	0.6570	0.5270	0.8741	0.7509	0.7398	0.6006	0.4913	0.3575	0.3290	0.2625	0.5336	0.4074	0.2930	0.2645	0.7138	0.5559	0.8076	0.7209	0.9512	0.9273	0.6992	0.6482	0.5835	0.4204	0.6793	0.5473	0.3977	0.3145	0.7997	0.6677	0.6787	0.4938	0.9127	0.8135	0.4497	0.4311	0.7829	0.6877	0.6242	0.5430	0.8548	0.7351	0.4529	0.3589	0.7084	0.4602
1.1399	13.6364	1200	1.2118	0.4561	0.5435	0.8210	0.6019	0.4311	0.7346	0.5916	0.7060	0.5364	0.8627	0.7713	0.8943	0.7671	0.7741	0.6454	0.7586	0.6725	0.0008	0.0008	0.6304	0.4400	0.0176	0.0176	0.9061	0.7930	0.4432	0.3895	0.9352	0.8392	0.9155	0.7876	0.9164	0.8174	0.0243	0.0208	0.7520	0.6269	0.8522	0.7243	0.4369	0.3233	0.5995	0.4999	0.0429	0.0426	0.4900	0.3669	0.5806	0.4994	0.8672	0.7518	0.7426	0.5987	0.4594	0.3458	0.3483	0.2707	0.5962	0.4193	0.2881	0.2535	0.7219	0.6194	0.8408	0.7279	0.9523	0.9275	0.7298	0.6660	0.5498	0.4262	0.7179	0.5519	0.3332	0.2860	0.8309	0.6682	0.6154	0.4970	0.9196	0.8154	0.4870	0.4638	0.7963	0.6841	0.5863	0.5092	0.8660	0.7391	0.5257	0.3792	0.6132	0.5124
1.1277	15.3409	1350	1.2074	0.4583	0.5481	0.8226	0.6260	0.4119	0.7198	0.5857	0.6843	0.5416	0.8742	0.7736	0.8388	0.7753	0.8063	0.6508	0.7789	0.6673	0.0024	0.0024	0.5795	0.4343	0.0248	0.0247	0.9003	0.7958	0.4193	0.3657	0.9321	0.8382	0.9087	0.7912	0.9179	0.8142	0.0989	0.0656	0.7688	0.6272	0.8561	0.7252	0.4175	0.3389	0.6451	0.5270	0.0454	0.0438	0.5150	0.3780	0.6339	0.5214	0.8740	0.7529	0.7352	0.5974	0.4953	0.3711	0.3408	0.2627	0.5594	0.4187	0.3179	0.2724	0.6807	0.5720	0.8441	0.7311	0.9559	0.9290	0.7019	0.6476	0.5451	0.4274	0.7032	0.5553	0.3695	0.3016	0.8066	0.6733	0.6593	0.5000	0.9216	0.8170	0.5001	0.4732	0.8365	0.6763	0.6116	0.5203	0.8706	0.7401	0.5433	0.3744	0.6339	0.5159
1.1180	17.0455	1500	1.2048	0.4590	0.5483	0.8236	0.5971	0.4126	0.7227	0.5873	0.6905	0.5425	0.8868	0.7705	0.8609	0.7794	0.7970	0.6544	0.7682	0.6706	0.0002	0.0002	0.5648	0.4334	0.0198	0.0198	0.9012	0.7995	0.4367	0.3867	0.9358	0.8380	0.8963	0.7848	0.9176	0.8196	0.0530	0.0394	0.7480	0.6298	0.8560	0.7270	0.4547	0.3181	0.6335	0.5257	0.0571	0.0548	0.4838	0.3700	0.6502	0.5322	0.8832	0.7526	0.7496	0.6083	0.4652	0.3534	0.3334	0.2658	0.5738	0.4199	0.3058	0.2652	0.7437	0.5886	0.8400	0.7329	0.9563	0.9286	0.6840	0.6401	0.5765	0.4365	0.7028	0.5567	0.3814	0.3114	0.8084	0.6781	0.6943	0.5076	0.9141	0.8178	0.5000	0.4724	0.8304	0.6900	0.5997	0.5191	0.8579	0.7400	0.5401	0.3793	0.6390	0.5088
1.1121	18.75	1650	1.2034	0.4595	0.5491	0.8239	0.5919	0.4132	0.7186	0.5878	0.7042	0.5463	0.8807	0.7729	0.8715	0.7782	0.7940	0.6572	0.7596	0.6748	0.0002	0.0002	0.5929	0.4385	0.0222	0.0222	0.9068	0.7972	0.4037	0.3584	0.9342	0.8377	0.8947	0.7873	0.9217	0.8129	0.0520	0.0381	0.7609	0.6317	0.8536	0.7276	0.4306	0.3222	0.6377	0.5276	0.0682	0.0655	0.5094	0.3766	0.6544	0.5306	0.8744	0.7547	0.7554	0.6071	0.4862	0.3616	0.3503	0.2723	0.5506	0.4171	0.3110	0.2692	0.7507	0.5800	0.8420	0.7321	0.9571	0.9293	0.7200	0.6666	0.5783	0.4410	0.6875	0.5539	0.3631	0.3075	0.8126	0.6766	0.6609	0.5012	0.9142	0.8189	0.5010	0.4732	0.8216	0.6930	0.6048	0.5193	0.8714	0.7396	0.5341	0.3701	0.6402	0.5071

Framework versions

Transformers 5.0.0
Pytorch 2.9.1+cu128
Datasets 4.5.0
Tokenizers 0.22.2

AllanK24/segformer-b5-finetuned-apple-dms-run11

segformer-b5-finetuned-apple-dms-run11

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

AllanK24/
segformer-b5-finetuned-apple-dms-run11