LLM_Teached_Pegasus_FS

This model is a fine-tuned version of google/pegasus-large on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.6167
• Rouge1: 0.4649
• Rouge2: 0.2096
• Rougel: 0.3686
• Rougelsum: 0.3688
• Gen Len: 30.6191
• Precision: 0.9102
• Recall: 0.9083
• F1: 0.9091

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: 24
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 96
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 16
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len	Precision	Recall	F1
No log	1.0	208	1.8075	0.411	0.1689	0.3152	0.3155	29.9091	0.901	0.897	0.8988
No log	2.0	417	1.7312	0.4379	0.1893	0.3442	0.3446	29.9073	0.9059	0.9024	0.904
2.0112	3.0	625	1.6987	0.4475	0.1978	0.352	0.3525	30.0173	0.9075	0.9039	0.9055
2.0112	4.0	834	1.6768	0.4514	0.1981	0.357	0.3573	30.0618	0.9082	0.9047	0.9063
1.7647	5.0	1042	1.6617	0.4537	0.2003	0.3592	0.3595	30.3264	0.9084	0.9055	0.9068
1.7647	6.0	1251	1.6502	0.4554	0.2021	0.3607	0.361	30.0827	0.9089	0.9057	0.9072
1.7647	7.0	1459	1.6416	0.4592	0.2052	0.3639	0.3641	30.0218	0.9099	0.9064	0.908
1.6948	8.0	1668	1.6360	0.4612	0.2054	0.3649	0.365	30.7827	0.909	0.9074	0.9081
1.6948	9.0	1876	1.6302	0.4621	0.2062	0.3645	0.3647	30.6291	0.9095	0.9074	0.9083
1.6501	10.0	2085	1.6265	0.4606	0.2051	0.3651	0.3655	30.4818	0.9095	0.9073	0.9083
1.6501	11.0	2293	1.6230	0.4625	0.2073	0.3658	0.366	30.8064	0.9097	0.908	0.9087
1.6222	12.0	2502	1.6205	0.4644	0.2082	0.3674	0.3679	30.5527	0.9103	0.9081	0.909
1.6222	13.0	2710	1.6188	0.4648	0.2087	0.3681	0.3683	30.8055	0.9101	0.9083	0.909
1.6222	14.0	2919	1.6172	0.4654	0.2097	0.3685	0.3689	30.6709	0.9104	0.9084	0.9093
1.6048	15.0	3127	1.6169	0.465	0.21	0.3693	0.3697	30.6309	0.9104	0.9084	0.9093
1.6048	15.96	3328	1.6167	0.4649	0.2096	0.3686	0.3688	30.6191	0.9102	0.9083	0.9091

Framework versions

• Transformers 4.36.0
• Pytorch 2.0.1+cu117
• Datasets 2.14.5
• Tokenizers 0.15.0