Pierre-Arthur/T5_small_eurlexsum_8Epochs

T5_small_eurlexsum

This model is a fine-tuned version of t5-small on the eur-lex-sum dataset. It achieves the following results on the evaluation set:

• Loss: 0.9360
• Rouge1: 0.2288
• Rouge2: 0.1816
• Rougel: 0.2157
• Rougelsum: 0.2158
• Gen Len: 19.0

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 8

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	71	1.4482	0.1743	0.0982	0.1509	0.1511	19.0
No log	2.0	142	1.1661	0.193	0.1257	0.1731	0.1734	19.0
No log	3.0	213	1.0651	0.2072	0.1483	0.1892	0.1896	19.0
No log	4.0	284	1.0053	0.2167	0.1638	0.2017	0.2019	19.0
No log	5.0	355	0.9706	0.222	0.1731	0.2082	0.2079	19.0
No log	6.0	426	0.9510	0.2253	0.1771	0.2114	0.2114	19.0
No log	7.0	497	0.9393	0.2263	0.1785	0.2134	0.2133	19.0
1.4549	8.0	568	0.9360	0.2288	0.1816	0.2157	0.2158	19.0

Framework versions

• Transformers 4.31.0
• Pytorch 2.0.1+cu118
• Datasets 2.13.1
• Tokenizers 0.13.3