PhysicalSciencePegasusLargeModel

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: 5.4857
• Rouge1: 45.4096
• Rouge2: 14.1811
• Rougel: 31.2204
• Rougelsum: 40.9948
• Bertscore Precision: 78.8464
• Bertscore Recall: 82.0345
• Bertscore F1: 80.4009
• Bleu: 0.0934
• Gen Len: 192.7149

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: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 16
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Bertscore Precision	Bertscore Recall	Bertscore F1	Bleu	Gen Len
6.8224	0.0620	100	6.4676	35.0969	8.6135	23.5193	31.2379	75.6321	79.8224	77.6611	0.0532	192.7149
6.3975	0.1239	200	6.1277	38.4317	11.0631	27.2821	34.6261	76.5444	80.6289	78.5221	0.0704	192.7149
6.1059	0.1859	300	6.0085	39.425	11.6567	27.8193	35.5195	76.8342	80.9269	78.8164	0.0749	192.7149
6.1215	0.2478	400	5.9209	40.7923	12.1231	28.4792	36.7207	77.1433	81.1253	79.0738	0.0786	192.7149
6.0244	0.3098	500	5.8054	41.3668	12.6049	28.8963	37.3873	77.1495	81.267	79.1445	0.0830	192.7149
5.9814	0.3717	600	5.7414	43.1237	13.2946	29.7215	38.645	77.6801	81.5336	79.5507	0.0875	192.7149
5.9421	0.4337	700	5.6719	43.8248	13.4876	30.0523	39.3496	77.8832	81.6572	79.7164	0.0887	192.7149
5.7744	0.4957	800	5.6277	44.5131	13.6831	30.4494	39.9332	78.1354	81.7356	79.8859	0.0896	192.7149
5.7227	0.5576	900	5.5916	44.8391	13.8057	30.6822	40.4155	78.4064	81.8258	80.0711	0.0902	192.7149
5.7532	0.6196	1000	5.5600	44.5759	13.8594	30.7325	40.1139	78.3708	81.8401	80.0593	0.0909	192.7149
5.7156	0.6815	1100	5.5421	45.2882	13.9686	31.0389	40.7447	78.6902	81.9492	80.2784	0.0916	192.7149
5.6819	0.7435	1200	5.5229	45.3689	14.1424	31.0474	40.781	78.6599	81.9628	80.269	0.0925	192.7149
5.7394	0.8055	1300	5.5060	45.4608	14.1555	31.1182	40.9486	78.7697	82.0031	80.3458	0.0930	192.7149
5.7374	0.8674	1400	5.4912	44.9845	14.0026	31.0462	40.6411	78.6797	81.9501	80.2733	0.0919	192.7149
5.6422	0.9294	1500	5.4874	45.5215	14.1925	31.2402	41.0589	78.8731	82.0393	80.4171	0.0934	192.7149
5.6744	0.9913	1600	5.4857	45.4096	14.1811	31.2204	40.9948	78.8464	82.0345	80.4009	0.0934	192.7149

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.1+cu121
• Datasets 2.19.2
• Tokenizers 0.19.1