MAE-CT-M1N0-M12_v8_split4_v3

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: 0.5156
• Accuracy: 0.8667

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: 4
• eval_batch_size: 4
• 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: 10500

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6865	0.0067	70	0.6839	0.6667
0.6859	1.0067	140	0.6229	0.6933
0.7131	2.0067	210	0.6232	0.6933
0.6056	3.0067	280	0.5851	0.6933
0.6318	4.0067	350	0.6402	0.68
0.5505	5.0067	420	0.4957	0.68
0.4649	6.0067	490	0.4274	0.7867
0.4421	7.0067	560	0.4528	0.7467
0.6176	8.0067	630	0.4277	0.7867
0.3803	9.0067	700	0.3763	0.8133
0.5473	10.0067	770	0.4343	0.8133
0.5326	11.0067	840	0.5099	0.8
0.7147	12.0067	910	0.4049	0.7867
0.5606	13.0067	980	0.5661	0.8133
0.4271	14.0067	1050	0.6158	0.7733
0.3684	15.0067	1120	0.5156	0.8667
0.4766	16.0067	1190	0.5960	0.8133
0.402	17.0067	1260	0.9327	0.8
0.2721	18.0067	1330	0.5997	0.8667
0.352	19.0067	1400	0.9081	0.8
0.6505	20.0067	1470	0.9743	0.7867
0.0024	21.0067	1540	0.9212	0.8
0.1791	22.0067	1610	1.0021	0.7867
0.3377	23.0067	1680	1.0045	0.8267
0.0004	24.0067	1750	0.9731	0.8267
0.0127	25.0067	1820	1.1212	0.8267
0.0325	26.0067	1890	1.0253	0.84
0.0002	27.0067	1960	1.0795	0.7867
0.0001	28.0067	2030	1.1357	0.7867
0.212	29.0067	2100	1.1049	0.8
0.0001	30.0067	2170	0.9523	0.8
0.2036	31.0067	2240	0.8127	0.8667
0.3654	32.0067	2310	1.1963	0.84
0.0009	33.0067	2380	1.3746	0.8133
0.0001	34.0067	2450	1.3530	0.7867
0.0001	35.0067	2520	1.4819	0.8
0.0003	36.0067	2590	1.3682	0.7867
0.0001	37.0067	2660	1.3876	0.8
0.0001	38.0067	2730	1.4598	0.8
0.0074	39.0067	2800	1.4145	0.7867
0.4399	40.0067	2870	1.2042	0.8
0.0001	41.0067	2940	1.2232	0.7733
0.0003	42.0067	3010	1.3577	0.7733
0.2268	43.0067	3080	1.3768	0.8
0.0001	44.0067	3150	1.4095	0.76
0.003	45.0067	3220	1.2064	0.8133
0.2623	46.0067	3290	1.5009	0.7867
0.0001	47.0067	3360	1.4357	0.8
0.0002	48.0067	3430	1.3622	0.8
0.0005	49.0067	3500	1.2478	0.8267
0.2139	50.0067	3570	1.0072	0.84
0.1948	51.0067	3640	1.4672	0.7867
0.4513	52.0067	3710	1.5611	0.7867
0.0003	53.0067	3780	1.6393	0.7867
0.0497	54.0067	3850	1.6415	0.7733
0.0001	55.0067	3920	1.5294	0.8133
0.0009	56.0067	3990	1.6254	0.7867
0.0	57.0067	4060	1.5758	0.7867
0.0001	58.0067	4130	1.3458	0.8133
0.0	59.0067	4200	1.4999	0.7867
0.0	60.0067	4270	1.5483	0.7867
0.0	61.0067	4340	1.4989	0.8133
0.1728	62.0067	4410	1.6545	0.7867
0.0003	63.0067	4480	1.5882	0.8
0.0017	64.0067	4550	1.8578	0.7333
0.0003	65.0067	4620	1.7840	0.7733
0.0	66.0067	4690	1.9174	0.76
0.0	67.0067	4760	2.0017	0.76
0.0	68.0067	4830	2.0249	0.76
0.1594	69.0067	4900	1.8066	0.7733
0.0	70.0067	4970	1.8688	0.7733
0.1722	71.0067	5040	1.9031	0.7733
0.2082	72.0067	5110	1.2061	0.8133
0.0	73.0067	5180	1.5182	0.8133
0.0	74.0067	5250	1.2031	0.8267
0.0027	75.0067	5320	1.2114	0.8133
0.0001	76.0067	5390	1.3714	0.8267
0.0	77.0067	5460	1.3626	0.8267
0.0	78.0067	5530	1.5210	0.84
0.0	79.0067	5600	1.7948	0.8
0.0005	80.0067	5670	1.5987	0.7867
0.0	81.0067	5740	1.6562	0.8267
0.0	82.0067	5810	1.6416	0.8133
0.0	83.0067	5880	1.6684	0.8267
0.0467	84.0067	5950	1.9072	0.8
0.0002	85.0067	6020	1.9762	0.7733
0.0001	86.0067	6090	1.8163	0.8
0.0	87.0067	6160	1.7790	0.7867
0.0001	88.0067	6230	1.4023	0.8133
0.0	89.0067	6300	1.3033	0.8267
0.0	90.0067	6370	1.4240	0.8
0.0	91.0067	6440	1.7616	0.76
0.0	92.0067	6510	1.3589	0.8
0.0001	93.0067	6580	1.8171	0.7867
0.0	94.0067	6650	1.4888	0.8267
0.0	95.0067	6720	1.7894	0.8133
0.0	96.0067	6790	1.7989	0.8133
0.0	97.0067	6860	1.7690	0.8133
0.0	98.0067	6930	1.6816	0.8133
0.0	99.0067	7000	1.7260	0.8133
0.0	100.0067	7070	1.7433	0.8133
0.0	101.0067	7140	1.7458	0.8133
0.0	102.0067	7210	1.7581	0.8133
0.0	103.0067	7280	1.5385	0.84
0.0	104.0067	7350	1.5528	0.8267
0.0	105.0067	7420	1.5646	0.8267
0.0	106.0067	7490	1.5761	0.8267
0.0	107.0067	7560	1.5740	0.8267
0.0	108.0067	7630	1.5858	0.8267
0.0	109.0067	7700	1.5992	0.8267
0.0035	110.0067	7770	1.8796	0.8133
0.0	111.0067	7840	1.5757	0.8133
0.0	112.0067	7910	1.5459	0.8133
0.0	113.0067	7980	1.5457	0.8133
0.0	114.0067	8050	1.5464	0.8267
0.0	115.0067	8120	1.5455	0.8267
0.0	116.0067	8190	1.5476	0.8267
0.0	117.0067	8260	1.5904	0.8267
0.0	118.0067	8330	1.6196	0.84
0.0018	119.0067	8400	1.4688	0.84
0.0	120.0067	8470	1.6467	0.8267
0.0	121.0067	8540	1.8343	0.7867
0.2547	122.0067	8610	1.5052	0.8533
0.0	123.0067	8680	1.5886	0.84
0.0	124.0067	8750	1.4159	0.8533
0.0	125.0067	8820	1.4188	0.8533
0.0	126.0067	8890	1.4199	0.8533
0.0	127.0067	8960	1.4224	0.8533
0.0	128.0067	9030	1.4154	0.8533
0.0	129.0067	9100	1.4262	0.8533
0.0	130.0067	9170	1.4201	0.8667
0.0	131.0067	9240	1.4197	0.8667
0.2341	132.0067	9310	1.7014	0.8267
0.0	133.0067	9380	1.4320	0.8533
0.0	134.0067	9450	1.4451	0.84
0.0	135.0067	9520	1.4577	0.84
0.0	136.0067	9590	1.4622	0.8267
0.0	137.0067	9660	1.4703	0.8267
0.0	138.0067	9730	1.4797	0.8267
0.0	139.0067	9800	1.4841	0.8267
0.0	140.0067	9870	1.4888	0.8267
0.0	141.0067	9940	1.4930	0.8267
0.0	142.0067	10010	1.4959	0.8267
0.0	143.0067	10080	1.5002	0.8267
0.0	144.0067	10150	1.5562	0.8267
0.0	145.0067	10220	1.5572	0.8267
0.0	146.0067	10290	1.5577	0.8267
0.0	147.0067	10360	1.5579	0.8267
0.0	148.0067	10430	1.5576	0.8267
0.0	149.0067	10500	1.5577	0.8267

Framework versions

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