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

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: 0.5662
• Accuracy: 0.7352
• F1: 0.7327
• Precision: 0.7370
• Recall: 0.7352

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
0.6983	0.9655	14	0.6720	0.5974	0.5694	0.6054	0.5974
0.6804	2.0	29	0.6609	0.6324	0.6190	0.6374	0.6324
0.6796	2.9655	43	0.6634	0.6083	0.6084	0.6084	0.6083
0.6886	4.0	58	0.6547	0.6171	0.6104	0.6161	0.6171
0.6577	4.9655	72	0.6577	0.6127	0.5724	0.6407	0.6127
0.6439	6.0	87	0.6196	0.6477	0.6339	0.6559	0.6477
0.602	6.9655	101	0.6125	0.6652	0.6585	0.6986	0.6652
0.5974	8.0	116	0.6224	0.6696	0.6601	0.7141	0.6696
0.5841	8.9655	130	0.5800	0.7002	0.7005	0.7011	0.7002
0.581	10.0	145	0.5822	0.7265	0.7262	0.7262	0.7265
0.5716	10.9655	159	0.5812	0.7068	0.7035	0.7083	0.7068
0.5611	12.0	174	0.5778	0.7221	0.7150	0.7319	0.7221
0.5411	12.9655	188	0.5652	0.7352	0.7341	0.7351	0.7352
0.5361	14.0	203	0.5670	0.7374	0.7347	0.7395	0.7374
0.5416	14.4828	210	0.5662	0.7352	0.7327	0.7370	0.7352

Framework versions

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