5c_2

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: 3.8264
• Accuracy: 0.44

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: 5
• eval_batch_size: 5
• 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: 4600

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.3192	0.0102	47	1.3455	0.4
0.8974	1.0102	94	1.4305	0.4
0.9209	2.0102	141	1.5682	0.4
0.7915	3.0102	188	1.6475	0.4
0.9571	4.0102	235	1.4177	0.4
0.9085	5.0102	282	1.4976	0.4
0.8815	6.0102	329	1.6249	0.4
0.9245	7.0102	376	1.7363	0.4
0.9179	8.0102	423	1.4075	0.4
1.0738	9.0102	470	1.9536	0.4
0.5865	10.0102	517	1.6996	0.4
0.4752	11.0102	564	2.1711	0.4
0.7409	12.0102	611	1.7252	0.36
0.7534	13.0102	658	1.9988	0.4
0.6109	14.0102	705	1.7449	0.36
0.4217	15.0102	752	2.9984	0.4
0.8409	16.0102	799	1.6709	0.36
0.6114	17.0102	846	1.9014	0.36
0.6806	18.0102	893	1.8763	0.36
0.5359	19.0102	940	2.1816	0.36
0.5989	20.0102	987	1.7239	0.32
0.4145	21.0102	1034	2.2460	0.32
0.3841	22.0102	1081	2.4509	0.36
0.4356	23.0102	1128	2.1125	0.36
0.2509	24.0102	1175	2.6513	0.32
0.4963	25.0102	1222	2.8019	0.4
0.1915	26.0102	1269	2.4637	0.32
0.1269	27.0102	1316	2.8957	0.36
0.3599	28.0102	1363	2.5853	0.36
0.399	29.0102	1410	3.3633	0.4
0.205	30.0102	1457	3.0276	0.32
0.0945	31.0102	1504	3.3960	0.4
0.3376	32.0102	1551	3.0445	0.32
0.2407	33.0102	1598	2.8461	0.32
0.1653	34.0102	1645	3.1737	0.36
0.187	35.0102	1692	3.5642	0.32
0.2339	36.0102	1739	3.6020	0.4
0.1097	37.0102	1786	3.5631	0.4
0.2859	38.0102	1833	3.6048	0.36
0.0123	39.0102	1880	4.2022	0.4
0.0062	40.0102	1927	4.2564	0.36
0.031	41.0102	1974	4.0465	0.4
0.1045	42.0102	2021	3.5379	0.36
0.0025	43.0102	2068	4.1880	0.4
0.2103	44.0102	2115	4.4486	0.32
0.0035	45.0102	2162	3.7883	0.32
0.0117	46.0102	2209	3.8264	0.44
0.0027	47.0102	2256	4.2371	0.32
0.0174	48.0102	2303	4.0451	0.4
0.0199	49.0102	2350	4.0996	0.4
0.0082	50.0102	2397	4.5682	0.36
0.0186	51.0102	2444	4.0036	0.36
0.1483	52.0102	2491	3.8019	0.36
0.1276	53.0102	2538	3.9253	0.4
0.0601	54.0102	2585	4.5047	0.4
0.0027	55.0102	2632	4.5747	0.36
0.0055	56.0102	2679	4.2363	0.32
0.0338	57.0102	2726	4.3328	0.36
0.0005	58.0102	2773	4.5897	0.36
0.0489	59.0102	2820	4.7412	0.32
0.11	60.0102	2867	4.7991	0.36
0.0006	61.0102	2914	4.8250	0.32
0.0008	62.0102	2961	4.7567	0.32
0.0004	63.0102	3008	4.4867	0.36
0.0877	64.0102	3055	4.8180	0.36
0.0009	65.0102	3102	4.3209	0.4
0.0004	66.0102	3149	4.3730	0.36
0.0005	67.0102	3196	4.0573	0.44
0.0288	68.0102	3243	3.7278	0.44
0.0014	69.0102	3290	4.9681	0.36
0.0002	70.0102	3337	4.8522	0.4
0.0009	71.0102	3384	4.9470	0.4
0.0004	72.0102	3431	4.8706	0.36
0.0016	73.0102	3478	4.8785	0.32
0.0003	74.0102	3525	4.9980	0.36
0.0003	75.0102	3572	4.7280	0.36
0.0003	76.0102	3619	5.0809	0.36
0.0005	77.0102	3666	4.8118	0.4
0.0003	78.0102	3713	4.7439	0.4
0.0003	79.0102	3760	4.9703	0.4
0.0004	80.0102	3807	4.5657	0.4
0.0004	81.0102	3854	4.5084	0.44
0.1261	82.0102	3901	4.8884	0.44
0.0002	83.0102	3948	4.8646	0.4
0.0002	84.0102	3995	4.8225	0.4
0.0003	85.0102	4042	4.7205	0.4
0.0008	86.0102	4089	4.7888	0.44
0.0004	87.0102	4136	4.8506	0.44
0.0004	88.0102	4183	4.8165	0.44
0.0006	89.0102	4230	4.6865	0.44
0.0002	90.0102	4277	4.6192	0.4
0.0002	91.0102	4324	4.6489	0.4
0.0005	92.0102	4371	4.7074	0.4
0.0002	93.0102	4418	4.6926	0.4
0.0012	94.0102	4465	4.7289	0.36
0.0002	95.0102	4512	4.7505	0.36
0.0002	96.0102	4559	4.7498	0.36
0.0002	97.0089	4600	4.7523	0.36

Framework versions

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