_Nougat_base_Edv_Ar_Jw_03

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

• Loss: 2849.9761

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: 0.0001
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 10
• total_train_batch_size: 80
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss
14.8948	1.0	61	3.7409
246.1179	2.0	122	181.8641
1736.7211	3.0	183	390.1302
3414.9794	4.0	244	663.5850
10090.4463	5.0	305	1313.3225
14502.3713	6.0	366	1777.3622
17956.3125	7.0	427	2115.5103
20723.0988	8.0	488	2345.2673
22772.8725	9.0	549	2503.4954
25527.135	10.0	610	2615.0542
25991.4225	11.0	671	2695.8262
26450.9	12.0	732	2757.9531
27287.9525	13.0	793	2801.9192
28178.5075	14.0	854	2831.1147
28008.415	15.0	915	2846.3933
28184.5925	16.0	976	2848.7292
28243.1675	17.0	1037	2849.6929
28126.525	18.0	1098	2849.9456
28380.8175	19.0	1159	2849.9773
28139.025	19.6788	1200	2849.9761

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cu121
• Tokenizers 0.21.0