CTMAE-P2-V5-3g-S4

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

• Loss: 0.4441
• Accuracy: 0.8444

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: 1e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 13050

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6729	0.0100	131	0.7202	0.4667
0.4727	1.0100	262	0.9200	0.4667
0.6907	2.0100	393	0.9521	0.4667
0.5985	3.0100	524	0.9436	0.4667
0.932	4.0100	655	1.0027	0.4667
0.5625	5.0100	786	0.8652	0.5556
0.6848	6.0100	917	0.8245	0.4667
1.9843	7.0100	1048	1.0328	0.4667
1.0259	8.0100	1179	1.1008	0.4667
0.6501	9.0100	1310	0.6334	0.7333
0.7425	10.0100	1441	1.9972	0.4667
0.3816	11.0100	1572	0.8999	0.6222
0.5828	12.0100	1703	0.6653	0.6444
0.6965	13.0100	1834	1.1995	0.5778
1.1441	14.0100	1965	1.1112	0.6444
1.0802	15.0100	2096	1.1944	0.6889
0.6764	16.0100	2227	1.2079	0.6667
0.9995	17.0100	2358	0.4441	0.8444
0.1524	18.0100	2489	0.6179	0.7111
0.6087	19.0100	2620	0.5499	0.7111
0.643	20.0100	2751	1.3079	0.5556
0.7549	21.0100	2882	0.5691	0.7778
0.0275	22.0100	3013	0.8075	0.7778
0.7877	23.0100	3144	0.6420	0.7556
0.1937	24.0100	3275	0.9011	0.6889
0.5489	25.0100	3406	0.8769	0.7111
0.636	26.0100	3537	0.5532	0.8222
0.65	27.0100	3668	1.0277	0.7556
1.752	28.0100	3799	0.8335	0.7556
0.8998	29.0100	3930	0.6492	0.8
0.5018	30.0100	4061	0.8418	0.8
0.8232	31.0100	4192	0.8180	0.7778
1.0924	32.0100	4323	0.7991	0.8
0.6881	33.0100	4454	1.5892	0.7111
0.6649	34.0100	4585	1.4467	0.6889
0.5765	35.0100	4716	1.2852	0.6889
0.535	36.0100	4847	0.8898	0.7778
0.4813	37.0100	4978	0.7436	0.8
0.5694	38.0100	5109	0.8707	0.7778
0.1397	39.0100	5240	1.1453	0.7556
0.2083	40.0100	5371	1.1627	0.7778
1.0281	41.0100	5502	1.7993	0.7111
0.7729	42.0100	5633	0.9857	0.7778
0.1594	43.0100	5764	2.0440	0.6444
0.0016	44.0100	5895	0.7688	0.8222
0.0029	45.0100	6026	1.5397	0.7111
0.0032	46.0100	6157	1.4149	0.7556
0.9143	47.0100	6288	1.1824	0.7778
1.1312	48.0100	6419	1.0552	0.8222
0.5491	49.0100	6550	1.4101	0.7333
0.0021	50.0100	6681	0.9631	0.8222
0.769	51.0100	6812	1.6253	0.7111
0.5188	52.0100	6943	0.8261	0.8
0.4272	53.0100	7074	1.2140	0.8222
0.5066	54.0100	7205	1.3431	0.7556
0.3655	55.0100	7336	1.4484	0.7778
0.3319	56.0100	7467	1.6030	0.7333
0.0106	57.0100	7598	1.2037	0.8444
0.0008	58.0100	7729	1.9449	0.7333
0.0693	59.0100	7860	1.9971	0.7333
0.2319	60.0100	7991	0.9876	0.8222
0.6032	61.0100	8122	1.5705	0.7333
0.0361	62.0100	8253	1.1494	0.8
0.4365	63.0100	8384	1.5111	0.7556
0.0002	64.0100	8515	1.5259	0.7778
0.0074	65.0100	8646	2.0437	0.7333
0.7648	66.0100	8777	1.5944	0.8
0.1898	67.0100	8908	1.7697	0.7556
0.3616	68.0100	9039	1.4570	0.7778
0.4445	69.0100	9170	1.8239	0.7111
0.0008	70.0100	9301	1.2796	0.8222
0.3047	71.0100	9432	1.3140	0.8
0.4618	72.0100	9563	1.1418	0.8
0.0014	73.0100	9694	1.4893	0.7778
0.1784	74.0100	9825	1.3635	0.8222
0.0585	75.0100	9956	1.3869	0.8222
0.0001	76.0100	10087	1.5997	0.7556
0.0001	77.0100	10218	1.2910	0.8222
0.2202	78.0100	10349	1.0942	0.8444
0.0001	79.0100	10480	1.6201	0.7778
0.0021	80.0100	10611	1.7737	0.7556
0.2677	81.0100	10742	1.7019	0.7556
0.4637	82.0100	10873	1.4919	0.7778
0.0001	83.0100	11004	1.4477	0.8222
0.0	84.0100	11135	1.5797	0.8
0.0	85.0100	11266	1.6962	0.7778
0.0001	86.0100	11397	1.3234	0.8222
0.0	87.0100	11528	1.6734	0.7778
0.0001	88.0100	11659	1.5921	0.8
0.0018	89.0100	11790	1.6718	0.7778
0.0	90.0100	11921	1.5656	0.8
0.3858	91.0100	12052	1.6168	0.7778
0.0	92.0100	12183	1.6664	0.8
0.0001	93.0100	12314	1.6439	0.8
0.0	94.0100	12445	1.4756	0.8222
0.0001	95.0100	12576	1.5124	0.8
0.4576	96.0100	12707	1.4554	0.8444
0.0	97.0100	12838	1.4442	0.8444
0.0	98.0100	12969	1.4453	0.8444
0.0	99.0062	13050	1.4455	0.8444

Framework versions

• Transformers 4.46.2
• Pytorch 2.0.1+cu117
• Datasets 3.0.1
• Tokenizers 0.20.0