mT5-TextSimp-LT-BatchSize8-lr5e-5

This model is a fine-tuned version of google/mt5-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0983
• Rouge1: 0.6245
• Rouge2: 0.4439
• Rougel: 0.6142
• Sacrebleu: 35.7192
• Gen Len: 38.0501

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: 8

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Sacrebleu	Gen Len
35.3898	0.96	200	27.6372	0.0019	0.0	0.0018	0.0003	512.0
3.5712	1.91	400	1.9615	0.0171	0.0	0.0167	0.0225	39.0501
0.6489	2.87	600	0.5638	0.0052	0.0	0.0051	0.0256	39.0501
0.6017	3.83	800	3.2823	0.2419	0.1287	0.2318	0.6457	130.3556
0.3784	4.78	1000	0.1340	0.5092	0.3277	0.4978	26.7005	38.0549
0.1521	5.74	1200	0.1092	0.5782	0.3973	0.5672	33.2443	38.0501
0.2096	6.7	1400	0.1001	0.6149	0.4342	0.6046	34.6518	38.0501
0.1719	7.66	1600	0.0983	0.6245	0.4439	0.6142	35.7192	38.0501

Framework versions

• Transformers 4.33.0
• Pytorch 2.1.2+cu121
• Datasets 2.14.4
• Tokenizers 0.13.3