MAE-CT-CPC-Dicotomized-v6-tricot

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: 2.5595
• Accuracy: 0.2564

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 3950

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.1005	0.0203	80	1.1025	0.3191
1.1073	1.0203	160	1.1015	0.4043
1.117	2.0203	240	1.1486	0.2340
1.0481	3.0203	320	1.1603	0.1702
1.0614	4.0203	400	1.2154	0.2979
0.9155	5.0203	480	1.1238	0.3830
0.9109	6.0203	560	1.1734	0.4468
0.6969	7.0203	640	1.2588	0.4468
0.6381	8.0203	720	1.2711	0.4255
0.5455	9.0203	800	1.3974	0.3830
0.4878	10.0203	880	1.2367	0.4468
0.3125	11.0203	960	1.5500	0.4468
0.5886	12.0203	1040	1.7877	0.3191
0.1826	13.0203	1120	1.8124	0.3404
0.3447	14.0203	1200	1.9852	0.4681
0.2065	15.0203	1280	2.3935	0.4043
0.3104	16.0203	1360	2.9981	0.3191
0.3517	17.0203	1440	2.5522	0.3830
0.0988	18.0203	1520	3.1463	0.4468
0.0532	19.0203	1600	2.8538	0.4468
0.1791	20.0203	1680	3.0306	0.4468
0.1584	21.0203	1760	3.4847	0.3830
0.016	22.0203	1840	3.4121	0.3191
0.0012	23.0203	1920	3.8550	0.3617
0.0005	24.0203	2000	3.9055	0.4043
0.0023	25.0203	2080	4.0501	0.4255
0.0009	26.0203	2160	4.2001	0.3404
0.0004	27.0203	2240	4.0130	0.3830
0.0162	28.0203	2320	4.0468	0.4043
0.0073	29.0203	2400	4.1919	0.4255
0.0012	30.0203	2480	4.0004	0.4255
0.0125	31.0203	2560	4.1151	0.3830
0.0005	32.0203	2640	4.3282	0.3830
0.0098	33.0203	2720	4.4689	0.3830
0.0036	34.0203	2800	4.3354	0.4043
0.0002	35.0203	2880	4.4605	0.3617
0.0002	36.0203	2960	4.1586	0.4255
0.0002	37.0203	3040	4.2574	0.4043
0.0002	38.0203	3120	4.6391	0.3830
0.0008	39.0203	3200	4.5526	0.3617
0.0001	40.0203	3280	4.5658	0.3830
0.0001	41.0203	3360	4.5712	0.3830
0.0001	42.0203	3440	4.6019	0.3830
0.0002	43.0203	3520	4.5915	0.4043
0.0001	44.0203	3600	4.6868	0.3830
0.0001	45.0203	3680	4.6619	0.3830
0.0002	46.0203	3760	4.7142	0.3617
0.0002	47.0203	3840	4.6525	0.3617
0.0001	48.0203	3920	4.6684	0.3617
0.0001	49.0076	3950	4.6668	0.3617

Framework versions

• Transformers 4.45.2
• Pytorch 2.0.1+cu117
• Datasets 3.0.1
• Tokenizers 0.20.0