segformer-b0-finetuned-agriculture-freeze-encoder

This model is a fine-tuned version of nvidia/mit-b0 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4279
• Mean Iou: 0.2863
• Mean Accuracy: 0.3449
• Overall Accuracy: 0.3991
• Accuracy Unlabeled: nan
• Accuracy Nutrient Deficiency: 0.3911
• Accuracy Planter Skip: 0.2441
• Accuracy Water: 0.7100
• Accuracy Waterway: 0.1217
• Accuracy Weed Cluster: 0.2574
• Iou Unlabeled: 0.0
• Iou Nutrient Deficiency: 0.3885
• Iou Planter Skip: 0.2436
• Iou Water: 0.7074
• Iou Waterway: 0.1213
• Iou Weed Cluster: 0.2569

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: 1e-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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Unlabeled	Accuracy Nutrient Deficiency	Accuracy Planter Skip	Accuracy Water	Accuracy Waterway	Accuracy Weed Cluster	Iou Unlabeled	Iou Nutrient Deficiency	Iou Planter Skip	Iou Water	Iou Waterway	Iou Weed Cluster
0.2023	1.0	8145	0.5192	0.1103	0.1327	0.1902	nan	0.2070	0.0003	0.3457	0.0000	0.1105	0.0	0.2057	0.0003	0.3455	0.0000	0.1103
0.7172	2.0	16290	0.4974	0.1282	0.1543	0.2138	nan	0.2617	0.0332	0.3582	0.0098	0.1083	0.0	0.2601	0.0332	0.3582	0.0098	0.1082
0.6844	3.0	24435	0.4657	0.2032	0.2445	0.3092	nan	0.3512	0.1223	0.5564	0.0384	0.1544	0.0	0.3492	0.1220	0.5554	0.0382	0.1543
0.2052	4.0	32580	0.4671	0.1912	0.2299	0.2961	nan	0.3261	0.1340	0.4389	0.0347	0.2160	0.0	0.3245	0.1338	0.4389	0.0346	0.2154
0.6564	5.0	40725	0.4468	0.2317	0.2788	0.3460	nan	0.3663	0.1487	0.5721	0.0793	0.2278	0.0	0.3642	0.1485	0.5715	0.0790	0.2272
0.1997	6.0	48870	0.4483	0.2392	0.2879	0.3446	nan	0.3524	0.1821	0.6219	0.0772	0.2059	0.0	0.3501	0.1817	0.6209	0.0769	0.2055
0.3586	7.0	57015	0.4413	0.2379	0.2860	0.3492	nan	0.3676	0.2073	0.5656	0.0505	0.2392	0.0	0.3661	0.2069	0.5653	0.0504	0.2387
0.7879	8.0	65160	0.4369	0.2501	0.3008	0.3597	nan	0.3632	0.2320	0.6003	0.0585	0.2497	0.0	0.3618	0.2315	0.6000	0.0584	0.2491
1.137	9.0	73305	0.4393	0.2649	0.3189	0.3853	nan	0.4236	0.2202	0.6417	0.0772	0.2316	0.0	0.4212	0.2197	0.6405	0.0770	0.2312
0.2625	10.0	81450	0.4388	0.2540	0.3057	0.3850	nan	0.4493	0.2058	0.5276	0.0705	0.2755	0.0	0.4460	0.2054	0.5275	0.0704	0.2748
0.2108	11.0	89595	0.4308	0.2845	0.3427	0.4149	nan	0.4475	0.2446	0.6482	0.0891	0.2838	0.0	0.4438	0.2440	0.6472	0.0889	0.2830
0.3237	12.0	97740	0.4251	0.2858	0.3440	0.4225	nan	0.4322	0.2372	0.6314	0.0854	0.3336	0.0	0.4296	0.2367	0.6309	0.0853	0.3322
1.0289	13.0	105885	0.4488	0.2604	0.3138	0.3823	nan	0.4623	0.2111	0.6343	0.0744	0.1869	0.0	0.4575	0.2107	0.6332	0.0742	0.1867
0.6843	14.0	114030	0.4253	0.2922	0.3519	0.4252	nan	0.4515	0.2267	0.6901	0.1089	0.2824	0.0	0.4481	0.2263	0.6884	0.1086	0.2815
0.2695	15.0	122175	0.4299	0.2856	0.3437	0.3878	nan	0.3812	0.2818	0.6873	0.1207	0.2472	0.0	0.3801	0.2811	0.6858	0.1203	0.2466
0.3991	16.0	130320	0.4225	0.2938	0.3534	0.4137	nan	0.4213	0.2714	0.6712	0.1131	0.2898	0.0	0.4198	0.2708	0.6702	0.1129	0.2888
0.7352	17.0	138465	0.4303	0.2732	0.3288	0.3894	nan	0.4176	0.2558	0.5941	0.1024	0.2740	0.0	0.4150	0.2553	0.5939	0.1022	0.2731
0.6884	18.0	146610	0.4243	0.2956	0.3556	0.4135	nan	0.4154	0.2735	0.6575	0.1294	0.3024	0.0	0.4137	0.2728	0.6568	0.1290	0.3013
0.3863	19.0	154755	0.4249	0.2861	0.3445	0.4184	nan	0.4597	0.2254	0.6370	0.1138	0.2864	0.0	0.4561	0.2251	0.6365	0.1135	0.2858
0.3208	20.0	162900	0.4279	0.2863	0.3449	0.3991	nan	0.3911	0.2441	0.7100	0.1217	0.2574	0.0	0.3885	0.2436	0.7074	0.1213	0.2569

Framework versions

• Transformers 4.39.1
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2