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: 0.4171
Accuracy: 0.9079

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
2.3212	0.02	75	2.2295	0.2581
1.7775	1.02	150	1.7800	0.3871
1.0633	2.02	225	0.9655	0.5484
0.3783	3.02	300	0.5901	0.7419
0.756	4.02	375	0.9142	0.6774
0.5186	5.02	450	0.7384	0.7742
0.3714	6.02	525	1.1662	0.7742
0.0263	7.02	600	0.9102	0.8065
0.0848	8.02	675	0.1210	0.9355
0.0028	9.02	750	0.2374	0.9355
0.0863	10.02	825	0.1728	0.9677
0.0018	11.02	900	0.3043	0.9355
0.1662	12.02	975	0.4244	0.9032
0.0132	13.02	1050	0.6966	0.8710
0.0019	14.02	1125	0.5602	0.8710
0.0012	15.02	1200	0.2649	0.8710
0.0061	16.02	1275	0.7361	0.8710
0.0012	17.02	1350	0.2821	0.9355
0.001	18.02	1425	0.3117	0.9355
0.001	19.02	1500	0.1193	0.9677
0.0021	20.02	1575	0.2413	0.9355
0.0009	21.02	1650	0.1641	0.9677
0.0016	22.02	1725	0.1333	0.9677
0.0007	23.02	1800	0.1060	0.9677
0.0008	24.02	1875	0.1112	0.9677
0.0006	25.02	1950	0.0160	1.0
0.0006	26.02	2025	0.0291	0.9677
0.0006	27.02	2100	0.0860	0.9677
0.0005	28.02	2175	0.1080	0.9677
0.0006	29.02	2250	0.1120	0.9677
0.0006	30.02	2325	0.0593	0.9677
0.0005	31.02	2400	0.1660	0.9677
0.0005	32.02	2475	0.0454	0.9677
0.0007	33.02	2550	0.1015	0.9677
0.0005	34.02	2625	0.1712	0.9677
0.0005	35.02	2700	0.1600	0.9677
0.0004	36.02	2775	0.1618	0.9677
0.0004	37.02	2850	0.1468	0.9677
0.0004	38.02	2925	0.1167	0.9677
0.0004	39.02	3000	0.1278	0.9677
0.0004	40.02	3075	0.1200	0.9677
0.0004	41.02	3150	0.1200	0.9677
0.0004	42.02	3225	0.1230	0.9677
0.0004	43.02	3300	0.1323	0.9677
0.0004	44.02	3375	0.1283	0.9677
0.0004	45.02	3450	0.1330	0.9677
0.0004	46.02	3525	0.1341	0.9677
0.0004	47.02	3600	0.1305	0.9677
0.0004	48.02	3675	0.1309	0.9677
0.0004	49.02	3750	0.1304	0.9677

Framework versions

Transformers 4.24.0
Pytorch 1.8.0+cu111
Datasets 2.7.1
Tokenizers 0.13.2