metadata

license: cc-by-nc-4.0
tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: videomae-base-finetuned-ucf101-subset
    results: []

videomae-base-finetuned-ucf101-subset

This model is a fine-tuned version of MCG-NJU/videomae-base on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.2483
Accuracy: 0.9432

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: 4
eval_batch_size: 4
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
training_steps: 3750

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.3832	0.02	75	2.1700	0.3784
1.8551	1.02	150	1.8236	0.3784
1.0117	2.02	225	1.1747	0.5135
0.6169	3.02	300	0.4409	0.8108
0.3897	4.02	375	0.6103	0.8108
0.3564	5.02	450	0.9210	0.7838
0.4998	6.02	525	0.6993	0.8378
0.0605	7.02	600	0.1617	0.9189
0.0814	8.02	675	0.6548	0.8378
0.0312	9.02	750	0.5517	0.8649
0.023	10.02	825	0.3978	0.9459
0.0021	11.02	900	0.3968	0.9189
0.1367	12.02	975	0.0432	0.9730
0.0021	13.02	1050	0.1839	0.9730
0.2373	14.02	1125	0.0755	0.9730
0.0015	15.02	1200	0.1486	0.9459
0.0013	16.02	1275	0.0174	1.0
0.1707	17.02	1350	0.5296	0.8919
0.0014	18.02	1425	0.0230	1.0
0.0011	19.02	1500	0.5438	0.8919
0.0011	20.02	1575	0.6957	0.8378
0.0008	21.02	1650	0.2705	0.9189
0.0028	22.02	1725	0.1965	0.9730
0.0007	23.02	1800	0.1783	0.9730
0.0008	24.02	1875	0.1809	0.9730
0.0006	25.02	1950	0.1793	0.9730
0.0009	26.02	2025	0.0970	0.9730
0.0006	27.02	2100	0.2483	0.9459
0.0006	28.02	2175	0.3035	0.9459
0.0006	29.02	2250	0.2314	0.9459
0.0005	30.02	2325	0.1906	0.9459
0.0005	31.02	2400	0.0814	0.9730
0.0005	32.02	2475	0.0881	0.9459
0.0005	33.02	2550	0.0798	0.9459
0.0005	34.02	2625	0.0706	0.9459
0.0005	35.02	2700	0.0949	0.9459
0.0004	36.02	2775	0.0868	0.9459
0.0004	37.02	2850	0.0595	0.9730
0.0005	38.02	2925	0.1342	0.9730
0.0004	39.02	3000	0.1594	0.9730
0.0004	40.02	3075	0.1488	0.9730
0.0004	41.02	3150	0.1434	0.9730
0.0004	42.02	3225	0.1149	0.9730
0.0004	43.02	3300	0.1119	0.9730
0.0004	44.02	3375	0.1119	0.9730
0.0004	45.02	3450	0.1096	0.9730
0.0004	46.02	3525	0.1096	0.9730
0.0004	47.02	3600	0.1085	0.9730
0.0004	48.02	3675	0.1032	0.9730
0.0004	49.02	3750	0.1064	0.9730

Framework versions

Transformers 4.24.0
Pytorch 1.8.0+cu111
Datasets 2.7.1
Tokenizers 0.13.2