uncased-clf-digimag_v1

This model is a fine-tuned version of HooshvareLab/bert-fa-base-uncased-clf-digimag on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.2679
• Accuracy: 0.6413
• F1: 0.6418
• Precision: 0.6458

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: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision
No log	1.0	221	1.1415	0.4972	0.4716	0.5696
No log	2.0	442	0.9770	0.6016	0.6029	0.6063
1.0648	3.0	663	0.9529	0.6266	0.6271	0.6304
1.0648	4.0	884	0.9879	0.6470	0.6468	0.6636
0.6031	5.0	1105	1.0113	0.6311	0.6321	0.6345
0.6031	6.0	1326	1.0840	0.6356	0.6322	0.6363
0.329	7.0	1547	1.1273	0.6436	0.6438	0.6487
0.329	8.0	1768	1.2089	0.6300	0.6285	0.6291
0.329	9.0	1989	1.2486	0.6345	0.6352	0.6399
0.1809	10.0	2210	1.2679	0.6413	0.6418	0.6458

Framework versions

• Transformers 4.38.2
• Pytorch 2.1.0+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2