convnext-small-224-leicester_binary

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

• Loss: 0.1283
• F1: 0.9620

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: 64
• eval_batch_size: 128
• seed: 1337
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30.0
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	F1
No log	1.0	7	0.5143	0.8608
0.5872	2.0	14	0.4215	0.8608
0.3903	3.0	21	0.4127	0.8608
0.3903	4.0	28	0.3605	0.8608
0.3163	5.0	35	0.3152	0.8608
0.2942	6.0	42	0.2942	0.8608
0.2942	7.0	49	0.2669	0.8608
0.2755	8.0	56	0.2316	0.8608
0.2281	9.0	63	0.2104	0.8608
0.2076	10.0	70	0.1938	0.8608
0.2076	11.0	77	0.1803	0.8608
0.1832	12.0	84	0.1704	0.8608
0.1758	13.0	91	0.1650	0.8608
0.1758	14.0	98	0.1714	0.8608
0.167	15.0	105	0.1575	0.8608
0.1519	16.0	112	0.1549	0.8608
0.1519	17.0	119	0.1705	0.8608
0.1422	18.0	126	0.1478	0.8608
0.1444	19.0	133	0.1437	0.8608
0.1396	20.0	140	0.1398	0.8608
0.1396	21.0	147	0.1351	0.8608
0.1293	22.0	154	0.1370	0.8987
0.1361	23.0	161	0.1335	0.8987
0.1361	24.0	168	0.1311	0.9367
0.1246	25.0	175	0.1289	0.9620
0.1211	26.0	182	0.1283	0.9620
0.1211	27.0	189	0.1294	0.9620
0.1182	28.0	196	0.1306	0.9620
0.1172	29.0	203	0.1312	0.9620
0.1102	30.0	210	0.1318	0.9620

Framework versions

• Transformers 4.26.0.dev0
• Pytorch 1.12.1+cu113
• Datasets 2.7.1
• Tokenizers 0.13.2