sagittal-b4-v11-finetuned-segments

This model is a fine-tuned version of nvidia/mit-b4 on the jenniferlumeng/MiceSagittal dataset. It achieves the following results on the evaluation set:

• Loss: 0.2102
• Mean Iou: 0.7660
• Mean Accuracy: 0.8966
• Overall Accuracy: 0.9025
• Accuracy Background: nan
• Accuracy Olfactory bulb: 0.9311
• Accuracy Anterior olfactory nucleus: 0.8576
• Accuracy Basal ganglia: 0.8881
• Accuracy Cortex: 0.9547
• Accuracy Hypothalamus: 0.7991
• Accuracy Thalamus: 0.8460
• Accuracy Hippocampus: 0.9531
• Accuracy Midbrain: 0.8908
• Accuracy Cerebellum: 0.9482
• Accuracy Pons and medulla: 0.8971
• Iou Background: 0.0
• Iou Olfactory bulb: 0.9039
• Iou Anterior olfactory nucleus: 0.7535
• Iou Basal ganglia: 0.8169
• Iou Cortex: 0.9464
• Iou Hypothalamus: 0.6482
• Iou Thalamus: 0.8106
• Iou Hippocampus: 0.9292
• Iou Midbrain: 0.8164
• Iou Cerebellum: 0.9371
• Iou Pons and medulla: 0.8635

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: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Background	Accuracy Olfactory bulb	Accuracy Anterior olfactory nucleus	Accuracy Basal ganglia	Accuracy Cortex	Accuracy Hypothalamus	Accuracy Thalamus	Accuracy Hippocampus	Accuracy Midbrain	Accuracy Cerebellum	Accuracy Pons and medulla	Iou Background	Iou Olfactory bulb	Iou Anterior olfactory nucleus	Iou Basal ganglia	Iou Cortex	Iou Hypothalamus	Iou Thalamus	Iou Hippocampus	Iou Midbrain	Iou Cerebellum	Iou Pons and medulla
0.679	2.8571	20	0.9745	0.6058	0.7741	0.8090	nan	0.8511	0.3173	0.7203	0.9142	0.7325	0.7090	0.9034	0.7282	0.9504	0.9147	0.0	0.7558	0.2904	0.5982	0.7974	0.6685	0.5337	0.7721	0.5572	0.9323	0.7581
0.3016	5.7143	40	0.3587	0.7273	0.8768	0.8823	nan	0.9822	0.8005	0.9478	0.9440	0.6822	0.8097	0.9763	0.7227	0.9769	0.9253	0.0	0.9301	0.7134	0.7860	0.9003	0.6332	0.7492	0.8546	0.6615	0.9086	0.8635
0.1624	8.5714	60	0.2448	0.7492	0.8943	0.8995	nan	0.9893	0.8487	0.8965	0.9607	0.7695	0.7685	0.9800	0.7855	0.9873	0.9567	0.0	0.9430	0.6882	0.8037	0.9349	0.6812	0.7179	0.9229	0.7106	0.9520	0.8865
0.1345	11.4286	80	0.2230	0.7558	0.8896	0.9019	nan	0.9778	0.8016	0.9463	0.9728	0.6755	0.8240	0.9634	0.8346	0.9680	0.9321	0.0	0.9270	0.7009	0.8164	0.9578	0.6296	0.7628	0.9486	0.7469	0.9470	0.8766
0.0976	14.2857	100	0.1891	0.7781	0.9114	0.9200	nan	0.9805	0.8293	0.9469	0.9676	0.7870	0.8236	0.9832	0.8752	0.9811	0.9393	0.0	0.9437	0.7270	0.8437	0.9653	0.7230	0.7625	0.9627	0.7735	0.9634	0.8941
0.0881	17.1429	120	0.1893	0.7803	0.9135	0.9214	nan	0.9807	0.8364	0.9315	0.9758	0.7977	0.8714	0.9677	0.8514	0.9779	0.9439	0.0	0.9431	0.7231	0.8374	0.9699	0.7346	0.7965	0.9561	0.7745	0.9599	0.8885
0.0873	20.0	140	0.1849	0.7739	0.9054	0.9158	nan	0.9813	0.8181	0.9328	0.9729	0.7932	0.8266	0.9580	0.8558	0.9728	0.9421	0.0	0.9382	0.7194	0.8417	0.9664	0.7251	0.7666	0.9478	0.7653	0.9597	0.8831
0.076	22.8571	160	0.1871	0.7765	0.9086	0.9192	nan	0.9823	0.8126	0.9226	0.9777	0.8067	0.8463	0.9570	0.8531	0.9776	0.9498	0.0	0.9369	0.7113	0.8346	0.9719	0.7353	0.7829	0.9478	0.7706	0.9621	0.8881
0.0948	25.7143	180	0.1850	0.7727	0.9047	0.9139	nan	0.9730	0.8330	0.9127	0.9667	0.8194	0.8083	0.9538	0.8653	0.9759	0.9384	0.0	0.9374	0.7165	0.8343	0.9599	0.7339	0.7588	0.9442	0.7697	0.9559	0.8895
0.0646	28.5714	200	0.1813	0.7756	0.9067	0.9137	nan	0.9731	0.8354	0.9258	0.9588	0.7750	0.8675	0.9635	0.8512	0.9749	0.9422	0.0	0.9414	0.7175	0.8305	0.9569	0.7235	0.7973	0.9530	0.7725	0.9562	0.8827
0.0636	31.4286	220	0.1894	0.7729	0.9034	0.9130	nan	0.9754	0.8133	0.9279	0.9673	0.7529	0.8737	0.9581	0.8483	0.9725	0.9448	0.0	0.9383	0.7088	0.8227	0.9625	0.7090	0.7979	0.9488	0.7750	0.9556	0.8831
0.0688	34.2857	240	0.1906	0.7761	0.9079	0.9174	nan	0.9803	0.8062	0.9240	0.9747	0.7894	0.8718	0.9650	0.8394	0.9792	0.9491	0.0	0.9350	0.7046	0.8302	0.9689	0.7313	0.7982	0.9527	0.7717	0.9579	0.8864
0.0635	37.1429	260	0.1874	0.7783	0.9093	0.9189	nan	0.9814	0.8162	0.9153	0.9752	0.8063	0.8667	0.9536	0.8567	0.9783	0.9436	0.0	0.9397	0.7071	0.8299	0.9683	0.7441	0.7990	0.9457	0.7800	0.9584	0.8887
0.0693	40.0	280	0.1869	0.7785	0.9094	0.9180	nan	0.9779	0.8151	0.9170	0.9727	0.8047	0.8740	0.9628	0.8455	0.9772	0.9471	0.0	0.9398	0.7103	0.8302	0.9678	0.7427	0.8030	0.9514	0.7747	0.9586	0.8853
0.0585	42.8571	300	0.1889	0.7770	0.9081	0.9168	nan	0.9778	0.8248	0.9108	0.9720	0.8021	0.8638	0.9564	0.8545	0.9768	0.9426	0.0	0.9389	0.7094	0.8293	0.9663	0.7435	0.7943	0.9470	0.7763	0.9559	0.8859
0.0783	45.7143	320	0.1880	0.7772	0.9078	0.9166	nan	0.9768	0.8193	0.9141	0.9708	0.7949	0.8700	0.9590	0.8567	0.9741	0.9426	0.0	0.9385	0.7093	0.8312	0.9651	0.7414	0.7979	0.9484	0.7769	0.9560	0.8842
0.0588	48.5714	340	0.1859	0.7774	0.9076	0.9167	nan	0.9767	0.8169	0.9159	0.9709	0.8001	0.8695	0.9546	0.8556	0.9744	0.9415	0.0	0.9380	0.7104	0.8323	0.9656	0.7425	0.7984	0.9455	0.7779	0.9563	0.8848

Framework versions

• Transformers 4.52.2
• Pytorch 2.6.0+cu124
• Datasets 2.16.1
• Tokenizers 0.21.1