results_fixed

This model is a fine-tuned version of riotu-lab/ArabianGPT-01B on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.1411
• Bleu: 0.2987
• Rouge1: 0.5831
• Rouge2: 0.3405
• Rougel: 0.5413

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 3.0

Training results

Training Loss	Epoch	Step	Bleu	Validation Loss	Rouge1	Rouge2	Rougel
3.9816	1.0846	500	0.2270	2.8213	0.3851	0.1625	0.3198
3.9816	2.0	922	0.2481	2.4793	0.4405	0.2030	0.3844
2.6514	3.0	1383	0.2658	2.2931	0.4828	0.2385	0.4327
2.3308	4.0	1844	0.2797	2.1801	0.5114	0.2679	0.4659
2.1322	5.0	2305	0.2886	2.1133	0.5264	0.2867	0.4852
1.9942	6.0	2766	0.2926	2.0649	0.5417	0.2993	0.5003
1.8884	7.0	3227	0.2967	2.0354	0.5529	0.3108	0.5117
1.8003	8.0	3688	0.2973	2.0164	0.5597	0.3199	0.5204
1.7305	9.0	4149	0.3053	2.0036	0.5660	0.3262	0.5266
1.6672	10.0	4610	0.3072	1.9933	0.5704	0.3319	0.5325
1.6132	11.0	5071	0.3093	1.9886	0.5737	0.3366	0.5363
1.5659	12.0	5532	0.3099	1.9834	0.5777	0.3397	0.5396
1.5659	13.0	5993	0.3117	1.9819	0.5796	0.3423	0.5418
1.5244	14.0	6454	0.3126	1.9798	0.5833	0.3452	0.5451
1.4884	15.0	6915	0.3130	1.9793	0.5832	0.3461	0.5454
1.4594	16.0	7376	0.3133	1.9800	0.5846	0.3469	0.5466
1.4361	17.0	7837	0.3151	1.9799	0.5865	0.3493	0.5485
1.4159	18.0	8298	0.3149	1.9809	0.5865	0.3495	0.5486
1.4159	19.0	8398	0.3099	2.0767	0.5858	0.3476	0.5471
1.6189	20.0	8840	0.3119	2.0654	0.5862	0.3489	0.5479

Framework versions

• Transformers 4.45.0.dev0
• Pytorch 2.3.1+cu121
• Datasets 2.19.2
• Tokenizers 0.19.1