polar-pond-221

This model is a fine-tuned version of facebook/convnextv2-base-1k-224 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1960
• Accuracy: 0.9688
• Precision: 0.9703
• Recall: 0.9688
• F1: 0.9683
• Roc Auc: 0.9987

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: 0.0001
• train_batch_size: 256
• eval_batch_size: 256
• 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: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1	Roc Auc
1.3843	1.0	17	1.3585	0.4896	0.2458	0.4896	0.3273	0.7094
1.3004	2.0	34	1.2273	0.4180	0.5441	0.4180	0.4640	0.7541
1.0841	3.0	51	1.1291	0.5430	0.5094	0.5430	0.4415	0.8077
0.8792	4.0	68	0.7692	0.4831	0.6620	0.4831	0.5065	0.7978
0.7057	5.0	85	0.7192	0.6510	0.6349	0.6510	0.6354	0.8585
0.6424	6.0	102	0.6292	0.5299	0.6840	0.5299	0.5244	0.8628
0.5829	7.0	119	0.5684	0.5898	0.7109	0.5898	0.6047	0.8724
0.4313	8.0	136	0.3756	0.7930	0.7945	0.7930	0.7936	0.9476
0.2881	9.0	153	0.2655	0.8516	0.8713	0.8516	0.8530	0.9716
0.1871	10.0	170	0.3171	0.8060	0.8729	0.8060	0.8089	0.9834
0.158	11.0	187	0.1419	0.9440	0.9441	0.9440	0.9439	0.9921
0.1137	12.0	204	0.1567	0.9245	0.9283	0.9245	0.9232	0.9932
0.0658	13.0	221	0.1298	0.9453	0.9462	0.9453	0.9455	0.9944
0.0696	14.0	238	0.1345	0.9466	0.9470	0.9466	0.9467	0.9948
0.043	15.0	255	0.1541	0.9674	0.9684	0.9674	0.9674	0.9972
0.0393	16.0	272	0.0805	0.9622	0.9633	0.9622	0.9624	0.9973
0.0339	17.0	289	0.1905	0.9466	0.9522	0.9466	0.9469	0.9966
0.0466	18.0	306	0.1001	0.9401	0.9468	0.9401	0.9413	0.9975
0.0322	19.0	323	0.0643	0.9792	0.9800	0.9792	0.9792	0.9990
0.0184	20.0	340	0.1204	0.9844	0.9846	0.9844	0.9842	0.9985
0.0201	21.0	357	0.0876	0.9779	0.9786	0.9779	0.9778	0.9992
0.0212	22.0	374	0.1340	0.9570	0.9611	0.9570	0.9574	0.9985
0.0168	23.0	391	0.0807	0.9661	0.9689	0.9661	0.9664	0.9992
0.0158	24.0	408	0.1210	0.9740	0.9745	0.9740	0.9738	0.9983
0.0135	25.0	425	0.1960	0.9688	0.9703	0.9688	0.9683	0.9987

Framework versions

• Transformers 4.52.3
• Pytorch 2.7.0+cpu
• Datasets 3.6.0
• Tokenizers 0.21.0