long_text_unbalanced_smaller_original_text

This model is a fine-tuned version of weny22/sum_model_t5_saved on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.4320
• Rouge1: 0.2054
• Rouge2: 0.0753
• Rougel: 0.1641
• Rougelsum: 0.1639
• Gen Len: 18.966

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: 0.002
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	72	2.7101	0.1773	0.0483	0.1397	0.1396	18.9987
No log	2.0	144	2.3778	0.1907	0.0614	0.1506	0.1506	18.9913
No log	3.0	216	2.2716	0.1898	0.0615	0.1506	0.1507	18.9973
No log	4.0	288	2.2295	0.1922	0.0619	0.1527	0.1528	18.992
No log	5.0	360	2.2176	0.1881	0.0611	0.149	0.1493	18.9773
No log	6.0	432	2.2287	0.192	0.0646	0.1534	0.1535	18.978
3.1377	7.0	504	2.1991	0.1949	0.0658	0.1547	0.1547	18.9867
3.1377	8.0	576	2.2397	0.198	0.0692	0.1588	0.1588	18.9767
3.1377	9.0	648	2.2651	0.1981	0.0693	0.1592	0.1592	18.9593
3.1377	10.0	720	2.2766	0.2032	0.0716	0.162	0.162	18.954
3.1377	11.0	792	2.2676	0.197	0.0663	0.1568	0.1568	18.9733
3.1377	12.0	864	2.3104	0.2024	0.0717	0.1603	0.1604	18.9593
3.1377	13.0	936	2.3127	0.2031	0.0716	0.162	0.1622	18.98
1.8667	14.0	1008	2.3402	0.2025	0.0717	0.1615	0.1613	18.972
1.8667	15.0	1080	2.3727	0.2037	0.0738	0.163	0.1628	18.9787
1.8667	16.0	1152	2.3787	0.2058	0.0747	0.1632	0.1631	18.9747
1.8667	17.0	1224	2.3898	0.2032	0.0735	0.1618	0.1618	18.9753
1.8667	18.0	1296	2.4053	0.2034	0.0738	0.1629	0.1626	18.982
1.8667	19.0	1368	2.4229	0.2021	0.073	0.1612	0.161	18.9693
1.8667	20.0	1440	2.4320	0.2054	0.0753	0.1641	0.1639	18.966

Framework versions

• Transformers 4.38.2
• Pytorch 2.1.2+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2