mt5-small-persian-dataset

This model is a fine-tuned version of google/mt5-small on the pn_summary dataset. It achieves the following results on the evaluation set:

• Loss: 1.9495
• Rouge1: 28.6676
• Rouge2: 12.4796
• Rougel: 26.0552
• Rougelsum: 26.0624

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: 5.6e-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
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
4.3518	1.0	1139	2.2437	24.8632	10.0197	22.5098	22.5578
2.8053	2.0	2278	2.1084	26.7746	11.2038	24.1918	24.1592
2.5852	3.0	3417	2.0525	27.3138	11.6092	24.795	24.765
2.4537	4.0	4556	2.0333	27.8395	11.92	25.1716	25.1786
2.3629	5.0	5695	1.9973	28.4229	12.2162	25.7546	25.7399
2.3007	6.0	6834	1.9752	28.259	12.3229	25.6448	25.6348
2.2527	7.0	7973	1.9605	28.7359	12.608	26.0384	26.0478
2.2227	8.0	9112	1.9571	28.5958	12.4125	25.9516	25.9815
2.1983	9.0	10251	1.9557	28.5015	12.4138	25.8887	25.8967
2.1769	10.0	11390	1.9495	28.6676	12.4796	26.0552	26.0624

Framework versions

• Transformers 4.26.1
• Pytorch 1.13.1+cu116
• Datasets 2.10.1
• Tokenizers 0.13.2