Modelpublic

pavitemple/finetuned-Accident-SingleLabel-Final-v4

sourceHugging Facecc-by-nc-4.0updated 3y agoView on Hugging Face

0likes5downloads

Model Card

finetuned-Accident-SingleLabel-Final-v4

This model is a fine-tuned version of MCG-NJU/videomae-base on an unknown dataset. It achieves the following results on the evaluation set:

—Loss: 1.2089
—Accuracy: 0.5588

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: 5e-05
—trainbatchsize: 8
—evalbatchsize: 8
—seed: 42
—distributed_type: multi-GPU
—num_devices: 2
—totaltrainbatch_size: 16
—totalevalbatch_size: 16
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—lrschedulerwarmup_ratio: 0.1
—training_steps: 65

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.06	4	1.8526	0.1304
No log	1.06	8	1.6872	0.3913
1.6209	2.06	12	1.4251	0.5652
1.6209	3.06	16	1.2785	0.5652
1.1184	4.06	20	1.1813	0.6522
1.1184	5.06	24	1.0827	0.4783
1.1184	6.06	28	0.9423	0.6522
1.0086	7.06	32	0.9538	0.6522
1.0086	8.06	36	0.8784	0.6087
0.7591	9.06	40	0.9870	0.6087
0.7591	10.06	44	0.9913	0.6522
0.7591	11.06	48	0.8661	0.6087
0.6925	12.06	52	0.8789	0.5652
0.6925	13.06	56	0.8263	0.6522
0.7497	14.06	60	1.0772	0.6522
0.7497	15.06	64	0.9689	0.6522
0.7497	16.02	65	0.8805	0.6522

Framework versions

—Transformers 4.35.0
—Pytorch 2.1.0
—Datasets 2.14.6
—Tokenizers 0.14.1