metadata

license: apache-2.0
base_model: riotu-lab/ArabianGPT-01B
tags:
  - generated_from_trainer
metrics:
  - bleu
  - rouge
model-index:
  - name: res_nw_yem
    results: []

res_nw_yem

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: 0.4720
Bleu: 0.2297
Rouge1: 0.5777
Rouge2: 0.3341
Rougel: 0.5758

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: 20.0

Training results

Training Loss	Epoch	Step	Bleu	Validation Loss	Rouge1	Rouge2	Rougel
7.8314	1.0	153	0.0104	2.3597	0.0983	0.0018	0.0964
0.9995	2.0	306	0.1161	0.5538	0.4736	0.1854	0.4728
0.4848	3.0	459	0.1472	0.5034	0.5008	0.2174	0.5002
0.3823	4.0	612	0.1912	0.4827	0.5331	0.2744	0.5314
0.293	5.0	765	0.2096	0.4732	0.5619	0.3099	0.5587
0.2239	6.0	918	0.2297	0.4720	0.5777	0.3341	0.5758
0.1766	7.0	1071	0.2302	0.4737	0.5885	0.3468	0.5872
0.1434	8.0	1224	0.2451	0.4758	0.5938	0.3648	0.5903
0.1202	9.0	1377	0.2509	0.4828	0.6047	0.3683	0.6015
0.1042	10.0	1530	0.2439	0.4852	0.5980	0.3704	0.5951
0.0955	11.0	1683	0.2560	0.4885	0.6116	0.3795	0.6086
0.0875	12.0	1836	0.4914	0.2551	0.6098	0.3802	0.6056
0.0825	13.0	1989	0.4981	0.2627	0.6145	0.3872	0.6108
0.0787	14.0	2142	0.5005	0.2590	0.6138	0.3797	0.6101
0.0757	15.0	2295	0.5020	0.2632	0.6171	0.3850	0.6141
0.074	16.0	2448	0.5049	0.2594	0.6161	0.3829	0.6125

Framework versions

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