vit-msn-small-wbc-classifier-cells-separated-dataset-agregates-25

This model is a fine-tuned version of facebook/vit-msn-small on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.1790
• Accuracy: 0.9401

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: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 256
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.351	0.9937	119	0.2523	0.9151
0.3364	1.9958	239	0.2355	0.9195
0.2999	2.9979	359	0.2384	0.9169
0.2861	4.0	479	0.1902	0.9341
0.3014	4.9937	598	0.2154	0.9290
0.292	5.9958	718	0.1764	0.9383
0.2441	6.9979	838	0.1894	0.9348
0.2416	8.0	958	0.1913	0.9349
0.2642	8.9937	1077	0.1738	0.9385
0.2482	9.9958	1197	0.1911	0.9371
0.2279	10.9979	1317	0.1867	0.9381
0.2331	12.0	1437	0.1814	0.9389
0.2208	12.9937	1556	0.1790	0.9401
0.2326	13.9958	1676	0.1926	0.9366
0.1899	14.9979	1796	0.1975	0.9372
0.1822	16.0	1916	0.2052	0.9352
0.1837	16.9937	2035	0.2078	0.9364
0.1712	17.9958	2155	0.2345	0.9288
0.1715	18.9979	2275	0.2156	0.9368
0.1516	20.0	2395	0.2279	0.9368
0.1504	20.9937	2514	0.2213	0.9382
0.139	21.9958	2634	0.2247	0.9370
0.1264	22.9979	2754	0.2357	0.9384
0.1266	24.0	2874	0.2360	0.9381
0.1144	24.8434	2975	0.2370	0.9375

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.2.0
• Tokenizers 0.19.1