CTMAE-P2-V2-S3

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: 1.3307
• Accuracy: 0.7826

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: 3250

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6138	0.02	65	0.7342	0.5435
0.5607	1.02	130	0.7364	0.5435
0.5525	2.02	195	0.9435	0.5435
0.6208	3.02	260	0.7286	0.5870
0.6099	4.02	325	0.7308	0.5435
0.4065	5.02	390	1.0573	0.5435
0.4338	6.02	455	0.6411	0.5870
0.4706	7.02	520	0.5360	0.6957
0.4147	8.02	585	1.0753	0.5870
0.6016	9.02	650	1.1059	0.5652
0.8252	10.02	715	0.5255	0.6739
0.3944	11.02	780	0.8291	0.5870
0.8652	12.02	845	0.7266	0.5652
0.4453	13.02	910	0.8914	0.6304
0.3375	14.02	975	1.4725	0.5870
0.5423	15.02	1040	0.6293	0.6739
0.4718	16.02	1105	0.9326	0.5870
0.4305	17.02	1170	0.6471	0.6957
0.1434	18.02	1235	1.1059	0.6739
0.3129	19.02	1300	0.7987	0.7391
0.2712	20.02	1365	0.9583	0.6957
0.5669	21.02	1430	1.9777	0.5652
0.3252	22.02	1495	0.9683	0.6739
0.1086	23.02	1560	1.0589	0.7391
0.2289	24.02	1625	1.0725	0.7174
0.2527	25.02	1690	1.0045	0.7609
0.433	26.02	1755	1.1574	0.6304
0.3203	27.02	1820	1.0995	0.7174
0.65	28.02	1885	1.4326	0.6957
0.1041	29.02	1950	1.2175	0.7174
0.0569	30.02	2015	1.4499	0.7174
0.2142	31.02	2080	1.3656	0.7609
0.343	32.02	2145	1.3127	0.7609
0.0331	33.02	2210	1.5137	0.6957
0.1634	34.02	2275	1.4774	0.7174
0.5358	35.02	2340	1.5174	0.6739
0.0396	36.02	2405	1.4475	0.6957
0.7272	37.02	2470	1.6262	0.7174
0.3445	38.02	2535	1.7874	0.7174
0.2743	39.02	2600	1.5739	0.7174
0.2592	40.02	2665	1.6093	0.7174
0.1367	41.02	2730	1.5615	0.7174
0.0459	42.02	2795	1.5238	0.7174
0.0102	43.02	2860	1.4447	0.7174
0.0084	44.02	2925	1.3307	0.7826
0.106	45.02	2990	1.3326	0.7609
0.0796	46.02	3055	1.3509	0.7391
0.0004	47.02	3120	1.5341	0.6522
0.0004	48.02	3185	1.3930	0.7391
0.0033	49.02	3250	1.4388	0.7174

Framework versions

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