hardy-bee-220

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

• Loss: 0.0965
• Accuracy: 0.9648
• Precision: 0.9660
• Recall: 0.9648
• F1: 0.9650
• Roc Auc: 0.9978

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
1.3971	1.0	17	1.3628	0.3997	0.4611	0.3997	0.3922	0.6197
1.343	2.0	34	1.2959	0.4206	0.5256	0.4206	0.4184	0.7270
1.218	3.0	51	1.1328	0.5104	0.5209	0.5104	0.5114	0.7725
1.0866	4.0	68	1.0127	0.5482	0.5251	0.5482	0.5135	0.7935
0.979	5.0	85	0.9274	0.5182	0.5688	0.5182	0.5264	0.7989
0.7709	6.0	102	0.8019	0.5508	0.5959	0.5508	0.5541	0.8207
0.6793	7.0	119	0.7300	0.5729	0.6291	0.5729	0.5779	0.8376
0.7004	8.0	136	0.7979	0.5820	0.6178	0.5820	0.5900	0.8433
0.5933	9.0	153	0.7219	0.5456	0.6553	0.5456	0.5559	0.8498
0.5292	10.0	170	0.6114	0.6888	0.7024	0.6888	0.6911	0.8923
0.4391	11.0	187	0.5453	0.6758	0.7536	0.6758	0.6758	0.9062
0.3856	12.0	204	0.5515	0.6810	0.7299	0.6810	0.6793	0.9156
0.3171	13.0	221	0.4486	0.7188	0.7829	0.7188	0.7172	0.9330
0.3035	14.0	238	0.4295	0.7227	0.7859	0.7227	0.7248	0.9364
0.2196	15.0	255	0.3438	0.8229	0.8298	0.8229	0.8243	0.9563
0.1842	16.0	272	0.2979	0.8190	0.8539	0.8190	0.8180	0.9690
0.1423	17.0	289	0.2747	0.8646	0.8638	0.8646	0.8638	0.9717
0.1387	18.0	306	0.3680	0.7904	0.8535	0.7904	0.7934	0.9793
0.1137	19.0	323	0.2358	0.8490	0.8744	0.8490	0.8486	0.9824
0.1013	20.0	340	0.1700	0.9193	0.9191	0.9193	0.9191	0.9894
0.0763	21.0	357	0.1573	0.9193	0.9241	0.9193	0.9199	0.9906
0.0648	22.0	374	0.1423	0.9323	0.9336	0.9323	0.9324	0.9915
0.0433	23.0	391	0.1344	0.9414	0.9413	0.9414	0.9412	0.9933
0.0392	24.0	408	0.1444	0.9427	0.9444	0.9427	0.9423	0.9938
0.0282	25.0	425	0.1134	0.9622	0.9627	0.9622	0.9623	0.9952
0.0249	26.0	442	0.1243	0.9466	0.9500	0.9466	0.9470	0.9953
0.015	27.0	459	0.1377	0.9336	0.9379	0.9336	0.9339	0.9959
0.0175	28.0	476	0.1320	0.9492	0.9497	0.9492	0.9493	0.9954
0.029	29.0	493	0.1202	0.9583	0.9592	0.9583	0.9582	0.9961
0.0138	30.0	510	0.0889	0.9714	0.9714	0.9714	0.9714	0.9976
0.0135	31.0	527	0.1064	0.9622	0.9635	0.9622	0.9624	0.9969
0.0076	32.0	544	0.1238	0.9427	0.9466	0.9427	0.9428	0.9969
0.0098	33.0	561	0.0871	0.9635	0.9643	0.9635	0.9636	0.9974
0.0066	34.0	578	0.1342	0.9518	0.9547	0.9518	0.9522	0.9968
0.0055	35.0	595	0.0965	0.9648	0.9660	0.9648	0.9650	0.9978

Framework versions

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