learning_rate: 0.0003
train_batch_size: 8
eval_batch_size: 8
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-10 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_ratio: 0.05
num_epochs: 100.0

Training results

Training Loss	Epoch	Step	Validation Loss	Dice	Iou	Iou Fg	Iou Bg	Acc	Recall	F1
1.0434	1.0	18	1.0199	0.1678	0.0465	0.0930	0.0000	0.0930	1.0	0.1702
1.009	2.0	36	0.9674	0.1741	0.0698	0.0966	0.0429	0.1318	0.9988	0.1763
0.9442	3.0	54	0.9009	0.2972	0.3523	0.1707	0.5339	0.5747	0.9413	0.2916
0.8816	4.0	72	0.8400	0.4450	0.5286	0.2853	0.7719	0.7910	0.8973	0.4440
0.8463	5.0	90	0.8135	0.5264	0.6133	0.3776	0.8490	0.8617	0.9021	0.5482
0.826	6.0	108	0.7786	0.5945	0.6710	0.4499	0.8921	0.9008	0.8723	0.6206
0.8086	7.0	126	0.7453	0.6243	0.6935	0.4818	0.9052	0.9129	0.8706	0.6503
0.798	8.0	144	0.7347	0.6235	0.6929	0.4816	0.9042	0.9120	0.8791	0.6501
0.7845	9.0	162	0.7285	0.6329	0.7013	0.4929	0.9097	0.9170	0.8675	0.6604
0.7627	10.0	180	0.7249	0.6203	0.6920	0.4789	0.9050	0.9127	0.8629	0.6476
0.763	11.0	198	0.7102	0.6273	0.6964	0.4861	0.9068	0.9143	0.8715	0.6542
0.7468	12.0	216	0.6947	0.6486	0.7130	0.5099	0.9161	0.9228	0.8637	0.6754
0.7303	13.0	234	0.6942	0.6131	0.6833	0.4696	0.8969	0.9056	0.8990	0.6390
0.7231	14.0	252	0.6732	0.6565	0.7199	0.5200	0.9197	0.9262	0.8601	0.6842
0.7038	15.0	270	0.6594	0.6451	0.7093	0.5055	0.9130	0.9201	0.8780	0.6716
0.7059	16.0	288	0.6459	0.6562	0.7185	0.5186	0.9185	0.9251	0.8680	0.6830
0.69	17.0	306	0.6350	0.6670	0.7277	0.5317	0.9237	0.9298	0.8574	0.6943
0.6767	18.0	324	0.6186	0.6746	0.7324	0.5389	0.9260	0.9319	0.8563	0.7004
0.6743	19.0	342	0.6143	0.6713	0.7315	0.5378	0.9252	0.9312	0.8608	0.6994
0.6619	20.0	360	0.6166	0.6508	0.7112	0.5089	0.9136	0.9207	0.8836	0.6745
0.658	21.0	378	0.6008	0.6620	0.7221	0.5243	0.9200	0.9264	0.8718	0.6879
0.6306	22.0	396	0.5856	0.6765	0.7347	0.5426	0.9267	0.9326	0.8601	0.7035
0.6311	23.0	414	0.5789	0.6742	0.7320	0.5386	0.9254	0.9314	0.8616	0.7001
0.6238	24.0	432	0.5696	0.6647	0.7226	0.5256	0.9196	0.9262	0.8792	0.6890
0.6165	25.0	450	0.5520	0.6859	0.7415	0.5524	0.9306	0.9361	0.8479	0.7117
0.6044	26.0	468	0.5348	0.6974	0.7508	0.5662	0.9353	0.9403	0.8374	0.7230
0.603	27.0	486	0.5404	0.6867	0.7419	0.5534	0.9304	0.9359	0.8540	0.7125
0.5797	28.0	504	0.5338	0.6856	0.7417	0.5518	0.9316	0.9369	0.8350	0.7112
0.5753	29.0	522	0.5380	0.6770	0.7325	0.5388	0.9262	0.9321	0.8535	0.7003
0.5661	30.0	540	0.5295	0.6816	0.7359	0.5438	0.9280	0.9337	0.8500	0.7045
0.5645	31.0	558	0.5415	0.6738	0.7312	0.5355	0.9269	0.9326	0.8356	0.6975

Metric Plots

Best Metrics with TTA

Best threshold: 0.600 Metrics @ best: {'dice': 0.7461113333702087, 'iou': 0.7712731957435608, 'iou_fg': 0.5950379967689514, 'iou_bg': 0.9475083947181702}

Inference Results

Framework versions

Transformers 4.57.6
Pytorch 2.8.0+cu128
Datasets 4.5.0
Tokenizers 0.22.2

jmanuelc87/unet-blood-vessels

unet-blood-vessels

Model description

Training procedure

Training hyperparameters

Training results

Metric Plots

Best Metrics with TTA

Inference Results

Framework versions

jmanuelc87/
unet-blood-vessels