convnextv2-tiny-1k-224-finetuned-four-five

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

• Loss: 0.6029
• Accuracy: 0.6544
• F1: 0.6540
• Precision: 0.6564
• Recall: 0.6544

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: 5e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
0.7064	0.9455	13	0.7007	0.5023	0.3813	0.4662	0.5023
0.6997	1.9636	27	0.6937	0.5276	0.4406	0.6010	0.5276
0.6895	2.9818	41	0.6864	0.5276	0.4815	0.5549	0.5276
0.6862	4.0	55	0.6769	0.5922	0.5887	0.5983	0.5922
0.6745	4.9455	68	0.6455	0.6336	0.6336	0.6344	0.6336
0.6443	5.9636	82	0.6340	0.6406	0.6399	0.6406	0.6406
0.6243	6.9818	96	0.6220	0.6590	0.6534	0.6661	0.6590
0.6243	8.0	110	0.6151	0.6705	0.6668	0.6754	0.6705
0.6248	8.9455	123	0.6104	0.6567	0.6565	0.6566	0.6567
0.6149	9.9636	137	0.6100	0.6751	0.6734	0.6816	0.6751
0.5968	10.9818	151	0.6026	0.6705	0.6705	0.6713	0.6705
0.5813	12.0	165	0.6028	0.6590	0.6589	0.6602	0.6590
0.5892	12.9455	178	0.6059	0.6452	0.6420	0.6537	0.6452
0.5738	13.9636	192	0.6029	0.6544	0.6541	0.6561	0.6544
0.5738	14.1818	195	0.6029	0.6544	0.6540	0.6564	0.6544

Framework versions

• Transformers 4.42.4
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1