mT5_multilingual_XLSum-finetuned-ar-wikilingua

This model is a fine-tuned version of csebuetnlp/mT5_multilingual_XLSum on the wiki_lingua dataset. It achieves the following results on the evaluation set:

• Loss: 3.6903
• Rouge-1: 24.47
• Rouge-2: 7.69
• Rouge-l: 20.04
• Gen Len: 39.64
• Bertscore: 72.63

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: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 250
• num_epochs: 8
• label_smoothing_factor: 0.1

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge-1	Rouge-2	Rouge-l	Gen Len	Bertscore
4.4406	1.0	5111	3.9582	22.35	6.84	18.39	34.78	71.94
4.0158	2.0	10222	3.8316	22.87	7.24	18.92	34.7	71.99
3.8626	3.0	15333	3.7695	23.65	7.5	19.6	35.53	72.31
3.7626	4.0	20444	3.7313	24.01	7.59	19.68	38.16	72.41
3.6934	5.0	25555	3.7118	24.37	7.77	19.93	39.36	72.47
3.6421	6.0	30666	3.7016	24.48	7.8	20.07	38.58	72.58
3.6073	7.0	35777	3.6907	24.31	7.83	20.13	38.07	72.5
3.5843	8.0	40888	3.6903	24.55	7.88	20.2	38.33	72.6

Framework versions

• Transformers 4.18.0
• Pytorch 1.10.0+cu111
• Datasets 2.1.0
• Tokenizers 0.12.1