MAE-CT-M1N0-M12_v8_split5_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: 1.1517
• Accuracy: 0.8701

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: 10350

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.685	0.0068	70	0.6757	0.7792
0.5601	1.0068	140	0.6218	0.6234
0.6632	2.0068	210	0.6157	0.6234
0.5153	3.0068	280	0.5660	0.6364
0.5008	4.0068	350	0.5238	0.7662
0.4879	5.0068	420	0.5012	0.7792
0.3636	6.0068	490	0.5640	0.7013
0.7238	7.0068	560	0.5756	0.7013
0.3339	8.0068	630	0.9895	0.6883
0.4152	9.0068	700	0.5031	0.8182
0.3126	10.0068	770	0.5350	0.7273
0.4479	11.0068	840	0.4278	0.8312
0.5548	12.0068	910	0.6865	0.7013
0.1509	13.0068	980	0.8144	0.7143
0.4038	14.0068	1050	0.6039	0.7922
0.2748	15.0068	1120	1.1834	0.7662
0.4552	16.0068	1190	0.7594	0.7532
0.5584	17.0068	1260	0.9481	0.7922
0.0919	18.0068	1330	1.0080	0.7662
0.2309	19.0068	1400	0.8453	0.8182
0.191	20.0068	1470	1.0695	0.7662
0.2013	21.0068	1540	1.4657	0.7403
0.6645	22.0068	1610	1.0602	0.8052
0.1083	23.0068	1680	1.2148	0.7532
0.0885	24.0068	1750	1.2008	0.7792
0.0015	25.0068	1820	1.2987	0.7532
0.2372	26.0068	1890	1.6225	0.7532
0.001	27.0068	1960	1.1689	0.7662
0.0006	28.0068	2030	1.3817	0.7532
0.0002	29.0068	2100	1.7143	0.7273
0.0012	30.0068	2170	1.8865	0.7273
0.153	31.0068	2240	2.4574	0.6623
0.1308	32.0068	2310	1.1800	0.8052
0.0002	33.0068	2380	1.2817	0.7792
0.0001	34.0068	2450	1.2770	0.7792
0.0001	35.0068	2520	1.2779	0.7922
0.0001	36.0068	2590	1.3971	0.7792
0.0001	37.0068	2660	1.1263	0.8182
0.0001	38.0068	2730	1.1233	0.8182
0.0675	39.0068	2800	1.4885	0.7662
0.0002	40.0068	2870	1.8406	0.7013
0.0001	41.0068	2940	1.9085	0.7532
0.0005	42.0068	3010	1.9380	0.7143
0.1589	43.0068	3080	0.9674	0.8312
0.0001	44.0068	3150	1.5574	0.7403
0.0353	45.0068	3220	1.1688	0.8312
0.0001	46.0068	3290	1.7684	0.7143
0.0002	47.0068	3360	1.3363	0.7792
0.1237	48.0068	3430	1.2230	0.7922
0.0001	49.0068	3500	1.4665	0.7792
0.0	50.0068	3570	1.5472	0.7662
0.1479	51.0068	3640	2.3369	0.7273
0.0001	52.0068	3710	2.2529	0.6753
0.1081	53.0068	3780	1.4745	0.7273
0.0002	54.0068	3850	1.5813	0.7403
0.0119	55.0068	3920	1.6007	0.7662
0.1478	56.0068	3990	2.3310	0.7143
0.0001	57.0068	4060	1.4788	0.8052
0.0001	58.0068	4130	1.1851	0.8442
0.0001	59.0068	4200	1.1920	0.8571
0.0904	60.0068	4270	1.1858	0.8312
0.0001	61.0068	4340	1.4534	0.7662
0.0017	62.0068	4410	1.6716	0.7792
0.0001	63.0068	4480	2.2017	0.6883
0.3407	64.0068	4550	1.2424	0.8052
0.0001	65.0068	4620	1.5786	0.7792
0.0002	66.0068	4690	1.3379	0.8182
0.0005	67.0068	4760	1.1517	0.8701
0.0	68.0068	4830	1.5294	0.7792
0.0	69.0068	4900	2.4381	0.6883
0.0032	70.0068	4970	1.7952	0.7532
0.0	71.0068	5040	3.0253	0.6753
0.214	72.0068	5110	1.9327	0.7143
0.0	73.0068	5180	2.0236	0.7532
0.0	74.0068	5250	1.9076	0.7662
0.0	75.0068	5320	1.7070	0.8052
0.0003	76.0068	5390	1.8621	0.7532
0.0	77.0068	5460	1.8847	0.7662
0.0	78.0068	5530	1.8880	0.7662
0.0001	79.0068	5600	1.8182	0.7792
0.0	80.0068	5670	1.7965	0.8052
0.0001	81.0068	5740	3.0536	0.6753
0.0005	82.0068	5810	1.5427	0.7922
0.0006	83.0068	5880	1.8892	0.7403
0.0001	84.0068	5950	1.9648	0.7403
0.0	85.0068	6020	1.7625	0.7532
0.1655	86.0068	6090	1.6751	0.7662
0.0	87.0068	6160	1.8559	0.7403
0.0	88.0068	6230	1.8886	0.7532
0.0	89.0068	6300	1.8957	0.7532
0.0	90.0068	6370	1.8181	0.7662
0.0	91.0068	6440	1.8299	0.7532
0.0	92.0068	6510	1.5186	0.8182
0.0393	93.0068	6580	1.9234	0.7792
0.0	94.0068	6650	2.1199	0.7273
0.0	95.0068	6720	2.1309	0.7403
0.0009	96.0068	6790	1.9311	0.7532
0.0001	97.0068	6860	1.7858	0.7792
0.0894	98.0068	6930	1.5577	0.8052
0.0	99.0068	7000	1.8138	0.7792
0.0	100.0068	7070	2.0068	0.7532
0.0163	101.0068	7140	1.8340	0.7922
0.0	102.0068	7210	1.3226	0.8312
0.0	103.0068	7280	2.4607	0.7532
0.0683	104.0068	7350	1.7550	0.7922
0.0	105.0068	7420	1.4900	0.8312
0.0	106.0068	7490	1.5684	0.7662
0.0	107.0068	7560	1.7333	0.8052
0.0	108.0068	7630	1.4233	0.7922
0.0001	109.0068	7700	1.7542	0.7792
0.0	110.0068	7770	1.4554	0.8052
0.0	111.0068	7840	1.3538	0.8571
0.0	112.0068	7910	1.4165	0.8571
0.0	113.0068	7980	1.4229	0.8571
0.0	114.0068	8050	1.4191	0.8571
0.0	115.0068	8120	1.4364	0.8571
0.0	116.0068	8190	1.4575	0.8312
0.0	117.0068	8260	1.4640	0.8312
0.0	118.0068	8330	1.4807	0.8312
0.0	119.0068	8400	1.5030	0.8312
0.0	120.0068	8470	1.5188	0.8312
0.0	121.0068	8540	1.5642	0.8182
0.0	122.0068	8610	1.5663	0.8182
0.0	123.0068	8680	1.5686	0.8182
0.0	124.0068	8750	1.4284	0.8571
0.0	125.0068	8820	1.4352	0.8571
0.0	126.0068	8890	1.4392	0.8571
0.0	127.0068	8960	1.5200	0.8442
0.0	128.0068	9030	1.5244	0.8442
0.0	129.0068	9100	1.5282	0.8442
0.0	130.0068	9170	1.5338	0.8442
0.0	131.0068	9240	1.5489	0.8442
0.0	132.0068	9310	1.5530	0.8442
0.0	133.0068	9380	1.5586	0.8442
0.0	134.0068	9450	1.5642	0.8442
0.0	135.0068	9520	1.5596	0.8442
0.0	136.0068	9590	1.5681	0.8442
0.0	137.0068	9660	1.4498	0.8182
0.0	138.0068	9730	1.6159	0.8312
0.0	139.0068	9800	1.6950	0.8182
0.0	140.0068	9870	1.6978	0.8182
0.0	141.0068	9940	1.6985	0.8182
0.0	142.0068	10010	1.6995	0.8182
0.0	143.0068	10080	1.7037	0.8052
0.0	144.0068	10150	1.7056	0.8052
0.0	145.0068	10220	1.7054	0.8052
0.0	146.0068	10290	1.7054	0.8052
0.0	147.0058	10350	1.7041	0.8052

Framework versions

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