_nougat_JawiCharJawi_AHR

This model is a fine-tuned version of bustamiyusoef/_nougat_JawiChar_Jawi on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.9923

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: 6
• total_train_batch_size: 48
• 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: 30

Training results

Training Loss	Epoch	Step	Validation Loss
No log	1.0	9	2.3985
15.1378	2.0	18	2.2738
13.1108	3.0	27	2.2290
12.6153	4.0	36	2.1573
12.2237	5.0	45	2.1169
11.8336	6.0	54	2.0919
11.5763	7.0	63	2.0773
11.3269	8.0	72	2.0492
11.1202	9.0	81	2.0450
10.2585	10.0	90	2.0421
10.2585	11.0	99	2.0222
10.864	12.0	108	2.0165
10.7603	13.0	117	2.0047
10.6177	14.0	126	2.0067
10.5024	15.0	135	2.0107
10.3865	16.0	144	2.0011
10.3026	17.0	153	1.9990
10.2636	18.0	162	1.9944
10.2232	19.0	171	1.9910
9.356	20.0	180	1.9946
9.356	21.0	189	1.9937
10.1019	22.0	198	1.9923

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0