finetuned-Accident-MultipleLabels-Video-subset-v2-new3

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: 1.7434
• Accuracy: 0.3333

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
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 35

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.06	2	1.9011	0.1719
No log	1.06	4	1.8469	0.3281
No log	2.06	6	1.8533	0.3281
No log	3.06	8	1.9122	0.3281
1.6429	4.06	10	1.9870	0.3125
1.6429	5.06	12	1.9921	0.2969
1.6429	6.06	14	1.9683	0.3125
1.6429	7.06	16	1.9313	0.3125
1.6429	8.06	18	1.9203	0.3125
1.4144	9.06	20	1.9251	0.3125
1.4144	10.06	22	1.9316	0.2969
1.4144	11.06	24	1.9455	0.2969
1.4144	12.06	26	1.9723	0.2969
1.4144	13.06	28	1.9897	0.2969
1.3307	14.06	30	2.0000	0.2969
1.3307	15.06	32	2.0063	0.2969
1.3307	16.06	34	2.0072	0.2969
1.3307	17.03	35	2.0082	0.2969

Framework versions

• Transformers 4.35.2
• Pytorch 2.2.0.dev20231202+cu121
• Datasets 2.15.0
• Tokenizers 0.15.0