hiera-base-224-in1k-hf-finetuned-stroke-binary

This model is a fine-tuned version of facebook/hiera-base-224-in1k-hf on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.1032
Accuracy: 0.9828
F1: 0.9828
Precision: 0.9829
Recall: 0.9828

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: 2e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 32
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: cosine
lr_scheduler_warmup_ratio: 0.1
num_epochs: 48
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
0.6239	0.6202	100	0.5808	0.6920	0.6271	0.7644	0.6920
0.4449	1.2357	200	0.4067	0.8304	0.8208	0.8482	0.8304
0.3462	1.8558	300	0.3286	0.8675	0.8620	0.8807	0.8675
0.271	2.4713	400	0.2234	0.9204	0.9201	0.9202	0.9204
0.2472	3.0868	500	0.1890	0.9272	0.9264	0.9285	0.9272
0.2765	3.7070	600	0.1951	0.9313	0.9313	0.9314	0.9313
0.2277	4.3225	700	0.1878	0.9299	0.9297	0.9297	0.9299
0.2463	4.9426	800	0.1769	0.9313	0.9307	0.9318	0.9313
0.2181	5.5581	900	0.1948	0.9258	0.9244	0.9308	0.9258
0.223	6.1736	1000	0.1736	0.9362	0.9358	0.9366	0.9362
0.231	6.7938	1100	0.3210	0.8765	0.8705	0.8963	0.8765
0.2022	7.4093	1200	0.1485	0.9475	0.9471	0.9483	0.9475
0.2114	8.0248	1300	0.2115	0.9118	0.9092	0.9215	0.9118
0.2093	8.6450	1400	0.1446	0.9480	0.9477	0.9482	0.9480
0.1736	9.2605	1500	0.1410	0.9484	0.9481	0.9490	0.9484
0.1587	9.8806	1600	0.1912	0.9317	0.9321	0.9340	0.9317
0.1561	10.4961	1700	0.1361	0.9512	0.9507	0.9524	0.9512
0.1536	11.1116	1800	0.1555	0.9389	0.9388	0.9388	0.9389
0.155	11.7318	1900	0.1216	0.9602	0.9600	0.9604	0.9602
0.1533	12.3473	2000	0.1359	0.9502	0.9499	0.9509	0.9502
0.1382	12.9674	2100	0.1394	0.9539	0.9535	0.9549	0.9539
0.1581	13.5829	2200	0.1107	0.9602	0.9599	0.9611	0.9602
0.1256	14.1984	2300	0.1255	0.9588	0.9586	0.9592	0.9588
0.1284	14.8186	2400	0.1777	0.9448	0.9440	0.9485	0.9448
0.1763	15.4341	2500	0.1201	0.9543	0.9540	0.9552	0.9543
0.1532	16.0496	2600	0.1272	0.9539	0.9535	0.9547	0.9539
0.1286	16.6698	2700	0.1221	0.9638	0.9638	0.9638	0.9638
0.1291	17.2853	2800	0.1137	0.9548	0.9544	0.9560	0.9548
0.1127	17.9054	2900	0.1271	0.9607	0.9608	0.9612	0.9607
0.1346	18.5209	3000	0.1024	0.9647	0.9645	0.9654	0.9647
0.1032	19.1364	3100	0.1438	0.9593	0.9590	0.9602	0.9593
0.1125	19.7566	3200	0.1158	0.9575	0.9574	0.9574	0.9575
0.1147	20.3721	3300	0.1350	0.9552	0.9548	0.9571	0.9552
0.1013	20.9922	3400	0.1616	0.9534	0.9529	0.9555	0.9534
0.1071	21.6078	3500	0.1299	0.9625	0.9622	0.9632	0.9625
0.0795	22.2233	3600	0.0954	0.9729	0.9729	0.9731	0.9729
0.1168	22.8434	3700	0.1308	0.9661	0.9660	0.9662	0.9661
0.1007	23.4589	3800	0.1528	0.9566	0.9561	0.9587	0.9566
0.08	24.0744	3900	0.1415	0.9607	0.9604	0.9616	0.9607
0.086	24.6946	4000	0.0843	0.9729	0.9728	0.9729	0.9729
0.0945	25.3101	4100	0.1114	0.9711	0.9710	0.9711	0.9711
0.0904	25.9302	4200	0.1066	0.9747	0.9746	0.9748	0.9747
0.0836	26.5457	4300	0.1061	0.9715	0.9714	0.9717	0.9715
0.0851	27.1612	4400	0.1305	0.9683	0.9681	0.9691	0.9683
0.0561	27.7814	4500	0.1142	0.9720	0.9718	0.9725	0.9720
0.0681	28.3969	4600	0.1083	0.9733	0.9732	0.9734	0.9733
0.0672	29.0124	4700	0.1157	0.9724	0.9724	0.9724	0.9724
0.0728	29.6326	4800	0.1083	0.9751	0.9750	0.9754	0.9751
0.0915	30.2481	4900	0.1164	0.9729	0.9728	0.9730	0.9729
0.0637	30.8682	5000	0.1171	0.9738	0.9738	0.9739	0.9738
0.0764	31.4837	5100	0.1100	0.9701	0.9701	0.9703	0.9701
0.063	32.0992	5200	0.0833	0.9778	0.9778	0.9778	0.9778
0.0494	32.7194	5300	0.0947	0.9778	0.9778	0.9779	0.9778
0.0481	33.3349	5400	0.0962	0.9792	0.9792	0.9792	0.9792
0.0648	33.9550	5500	0.0997	0.9783	0.9782	0.9784	0.9783
0.0516	34.5705	5600	0.1097	0.9796	0.9796	0.9798	0.9796
0.0533	35.1860	5700	0.1054	0.9769	0.9769	0.9769	0.9769
0.0413	35.8062	5800	0.1080	0.9769	0.9769	0.9771	0.9769
0.0454	36.4217	5900	0.1113	0.9774	0.9773	0.9777	0.9774
0.0436	37.0372	6000	0.1058	0.9787	0.9787	0.9788	0.9787
0.0425	37.6574	6100	0.0945	0.9810	0.9810	0.9810	0.9810
0.0529	38.2729	6200	0.0931	0.9815	0.9814	0.9815	0.9815
0.0613	38.8930	6300	0.1019	0.9801	0.9801	0.9801	0.9801
0.0605	39.5085	6400	0.0960	0.9810	0.9810	0.9811	0.9810
0.0402	40.1240	6500	0.0935	0.9824	0.9823	0.9824	0.9824
0.0467	40.7442	6600	0.0992	0.9810	0.9810	0.9811	0.9810
0.0471	41.3597	6700	0.0957	0.9801	0.9801	0.9801	0.9801
0.0524	41.9798	6800	0.0981	0.9810	0.9810	0.9810	0.9810
0.0308	42.5953	6900	0.1033	0.9824	0.9823	0.9825	0.9824
0.0472	43.2109	7000	0.1014	0.9815	0.9814	0.9815	0.9815
0.0453	43.8310	7100	0.0970	0.9801	0.9801	0.9801	0.9801
0.0303	44.4465	7200	0.1041	0.9828	0.9828	0.9830	0.9828
0.0311	45.0620	7300	0.1032	0.9828	0.9828	0.9829	0.9828
0.0443	45.6822	7400	0.1004	0.9824	0.9823	0.9825	0.9824
0.045	46.2977	7500	0.1004	0.9824	0.9823	0.9825	0.9824
0.036	46.9178	7600	0.1007	0.9824	0.9823	0.9825	0.9824
0.0379	47.5333	7700	0.1007	0.9824	0.9823	0.9825	0.9824