emotion-classificationV3

This model is a fine-tuned version of /content/model/emotion-classificationV3/checkpoint-60 on FastJobs/Visual_Emotional_Analysis Dataset. It achieves the following results on the evaluation set:

• Loss: 0.5765
• Accuracy: 0.8438

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

FastJobs/Visual_Emotional_Analysis

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 1e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 143
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	5	0.6923	0.7937
0.5541	2.0	10	0.7871	0.8063
0.5541	3.0	15	0.7193	0.8313
0.5168	4.0	20	0.6446	0.825
0.5168	5.0	25	0.5653	0.8438
0.4627	6.0	30	0.7244	0.8063
0.4627	7.0	35	0.7213	0.7937
0.4516	8.0	40	0.6082	0.8313
0.4516	9.0	45	0.7545	0.8063
0.4339	10.0	50	0.5320	0.8562
0.4339	11.0	55	0.6222	0.8187
0.4233	12.0	60	0.6104	0.8438
0.4233	13.0	65	0.5913	0.825
0.3976	14.0	70	0.6852	0.8125
0.3976	15.0	75	0.6227	0.8125
0.3933	16.0	80	0.5550	0.825
0.3933	17.0	85	0.5438	0.8438
0.4359	18.0	90	0.5916	0.825
0.4359	19.0	95	0.6037	0.8063
0.3589	20.0	100	0.7102	0.8125

Framework versions

• Transformers 4.33.2
• Pytorch 2.0.1+cu118
• Datasets 2.14.5
• Tokenizers 0.13.3