tsf-gs-rots-wtoken-DRPT0.3-r224-f150-6.6-h768-i3072-p32-b8-e50

This model is a fine-tuned version of facebook/timesformer-base-finetuned-k400 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.2019
• Accuracy: 0.6043

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: 8
• eval_batch_size: 8
• 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: 5400
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.1714	0.0202	109	1.1130	0.3369
1.1431	1.0202	218	1.1018	0.3369
1.1363	2.0202	327	1.1018	0.3369
1.1117	3.0202	436	1.1062	0.3369
1.0901	4.0202	545	1.1024	0.3369
1.061	5.0202	654	1.1049	0.3262
1.1047	6.0202	763	1.0996	0.3262
1.0911	7.0202	872	1.1056	0.3369
1.1305	8.0202	981	1.0994	0.3262
1.1145	9.0202	1090	1.0978	0.3262
1.1075	10.0202	1199	1.0921	0.3369
1.0886	11.0202	1308	1.0732	0.4118
1.2003	12.0202	1417	1.1430	0.3369
1.1006	13.0202	1526	1.0699	0.4599
1.0594	14.0202	1635	1.0786	0.3743
1.0283	15.0202	1744	1.0056	0.4813
1.0804	16.0202	1853	1.0239	0.4332
1.1055	17.0202	1962	1.0384	0.5027
1.0577	18.0202	2071	0.9683	0.4813
0.9961	19.0202	2180	1.1497	0.3422
0.9365	20.0202	2289	1.1324	0.4064
1.0121	21.0202	2398	1.0838	0.5080
0.8774	22.0202	2507	1.1312	0.5722
0.8922	23.0202	2616	1.1480	0.5241
1.0667	24.0202	2725	0.9482	0.6096
0.825	25.0202	2834	0.8572	0.6524
0.985	26.0202	2943	0.9455	0.6043
0.8331	27.0202	3052	0.7787	0.6310
0.8799	28.0202	3161	0.8728	0.5989
0.7616	29.0202	3270	0.7851	0.6952
0.6198	30.0202	3379	0.8857	0.6524
0.9501	31.0202	3488	0.7847	0.6791
0.708	32.0202	3597	0.7956	0.6684
0.697	33.0202	3706	0.8540	0.6471
0.7463	34.0202	3815	0.7873	0.7005
0.6234	35.0202	3924	1.0743	0.5615
0.7174	36.0202	4033	0.7159	0.7219
0.6577	37.0202	4142	0.8100	0.6417
0.7482	38.0202	4251	0.8508	0.6631
0.6993	39.0202	4360	0.8796	0.6684
0.6782	40.0202	4469	0.9814	0.6578
0.614	41.0202	4578	0.8482	0.6631
0.7516	42.0202	4687	0.9654	0.6471
0.7425	43.0202	4796	1.0050	0.6257
0.6818	44.0202	4905	1.1266	0.6096
0.4599	45.0202	5014	1.1824	0.6364
0.7561	46.0202	5123	1.1567	0.6417
0.7543	47.0202	5232	1.2946	0.5882
0.4198	48.0202	5341	1.2598	0.5882
0.6238	49.0109	5400	1.2019	0.6043

Framework versions

• Transformers 4.41.2
• Pytorch 1.13.0+cu117
• Datasets 2.20.0
• Tokenizers 0.19.1