master-mahdi/intent_detection_fa

my_model

This model is a fine-tuned version of HooshvareLab/bert-fa-base-uncased-clf-persiannews on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.8573
• Accuracy: 0.8695

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: 2e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	360	1.0779	0.7772
1.8476	2.0	720	0.7038	0.8436
0.6038	3.0	1080	0.6356	0.8490
0.6038	4.0	1440	0.6477	0.8470
0.3037	5.0	1800	0.6521	0.8515
0.1723	6.0	2160	0.6938	0.8554
0.1066	7.0	2520	0.6942	0.8519
0.1066	8.0	2880	0.7426	0.8559
0.0636	9.0	3240	0.7769	0.8549
0.0464	10.0	3600	0.8179	0.8549
0.0464	11.0	3960	0.8457	0.8544
0.029	12.0	4320	0.8509	0.8554
0.0241	13.0	4680	0.8534	0.8529
0.0158	14.0	5040	0.8793	0.8549
0.0158	15.0	5400	0.9053	0.8544
0.0131	16.0	5760	0.9009	0.8539
0.012	17.0	6120	0.9124	0.8554
0.012	18.0	6480	0.9092	0.8549
0.0114	19.0	6840	0.9117	0.8529
0.0079	20.0	7200	0.9129	0.8519

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1