README.md

metadata

library_name: transformers
license: apache-2.0
base_model: facebook/vit-msn-base
tags:
  - generated_from_trainer
datasets:
  - imagefolder
metrics:
  - accuracy
model-index:
  - name: vit-msn-base-finetuned-lf-invalidation
    results:
      - task:
          name: Image Classification
          type: image-classification
        dataset:
          name: imagefolder
          type: imagefolder
          config: default
          split: test
          args: default
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.9234042553191489

vit-msn-base-finetuned-lf-invalidation

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

• Loss: 0.2414
• Accuracy: 0.9234

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: 80

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.96	6	0.6512	0.6957
0.7053	1.92	12	0.6311	0.6809
0.7053	2.88	18	0.5361	0.7277
0.5163	4.0	25	0.3341	0.8681
0.3242	4.96	31	0.3167	0.8809
0.3242	5.92	37	0.3960	0.8191
0.2779	6.88	43	0.3818	0.8255
0.2348	8.0	50	0.5019	0.7362
0.2348	8.96	56	0.2944	0.8851
0.26	9.92	62	0.2414	0.9234
0.26	10.88	68	0.3664	0.8298
0.2778	12.0	75	0.2505	0.9043
0.2271	12.96	81	0.6277	0.6298
0.2271	13.92	87	0.2753	0.8745
0.2488	14.88	93	0.6249	0.6957
0.2729	16.0	100	0.5195	0.7149
0.2729	16.96	106	0.7984	0.5745
0.3261	17.92	112	0.4631	0.7723
0.3261	18.88	118	1.1010	0.5149
0.2212	20.0	125	0.2337	0.9170
0.2802	20.96	131	0.4638	0.7574
0.2802	21.92	137	0.3859	0.8362
0.2112	22.88	143	0.6708	0.6894
0.2231	24.0	150	0.3387	0.8681
0.2231	24.96	156	0.7045	0.6553
0.2037	25.92	162	0.3958	0.8277
0.2037	26.88	168	0.5082	0.7702
0.1845	28.0	175	0.5991	0.7234
0.1898	28.96	181	0.5108	0.7617
0.1898	29.92	187	0.2720	0.9085
0.2118	30.88	193	0.4936	0.7851
0.2097	32.0	200	0.3748	0.8404
0.2097	32.96	206	0.5048	0.7766
0.1704	33.92	212	0.4368	0.7957
0.1704	34.88	218	0.6959	0.6830
0.1962	36.0	225	1.0097	0.5957
0.1686	36.96	231	0.4992	0.7915
0.1686	37.92	237	0.5374	0.7574
0.1855	38.88	243	0.3710	0.8340
0.1528	40.0	250	0.3631	0.8447
0.1528	40.96	256	0.5589	0.7681
0.1523	41.92	262	0.5147	0.7809
0.1523	42.88	268	0.5299	0.7638
0.1709	44.0	275	0.5937	0.7447
0.1527	44.96	281	0.5969	0.7383
0.1527	45.92	287	0.6439	0.7255
0.1397	46.88	293	0.7721	0.6723
0.1538	48.0	300	0.5768	0.7702
0.1538	48.96	306	0.5801	0.7596
0.1466	49.92	312	0.5673	0.7574
0.1466	50.88	318	0.6469	0.7085
0.1302	52.0	325	0.7276	0.6957
0.1565	52.96	331	0.8247	0.6723
0.1565	53.92	337	0.4811	0.7979
0.1267	54.88	343	0.6373	0.7021
0.1424	56.0	350	0.7252	0.6723
0.1424	56.96	356	0.5697	0.7489
0.1053	57.92	362	0.7067	0.6957
0.1053	58.88	368	0.6577	0.7064
0.1301	60.0	375	0.5326	0.7745
0.0906	60.96	381	0.5468	0.7851
0.0906	61.92	387	0.4413	0.8277
0.0974	62.88	393	0.5479	0.7660
0.1133	64.0	400	0.7109	0.7043
0.1133	64.96	406	0.5735	0.7617
0.1189	65.92	412	0.4084	0.8298
0.1189	66.88	418	0.5716	0.7489
0.1064	68.0	425	0.5537	0.7553
0.1084	68.96	431	0.4569	0.8021
0.1084	69.92	437	0.5227	0.7617
0.1054	70.88	443	0.5995	0.7277
0.1005	72.0	450	0.5560	0.7638
0.1005	72.96	456	0.4550	0.8064
0.1028	73.92	462	0.4404	0.8234
0.1028	74.88	468	0.4761	0.7957
0.0917	76.0	475	0.5278	0.7681
0.1009	76.8	480	0.5346	0.7617

Framework versions

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