fine-tuned-bart-15-epochs-1500-input-256-output

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

• Loss: 0.8826
• Rouge1: 0.1734
• Rouge2: 0.0367
• Rougel: 0.1299
• Rougelsum: 0.1304
• Gen Len: 37.9

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: 0.0001
• 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_ratio: 0.1
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	151	4.7528	0.0	0.0	0.0	0.0	13.65
No log	2.0	302	1.0942	0.1251	0.0251	0.1187	0.1205	10.18
No log	3.0	453	0.9616	0.0757	0.0193	0.0606	0.0612	18.24
3.6176	4.0	604	0.9080	0.119	0.0278	0.0913	0.0926	31.52
3.6176	5.0	755	0.8831	0.1595	0.0325	0.1307	0.1316	30.88
3.6176	6.0	906	0.8731	0.1295	0.0276	0.0917	0.0922	39.26
0.7542	7.0	1057	0.8610	0.1595	0.0379	0.1159	0.1169	48.91
0.7542	8.0	1208	0.8556	0.172	0.0407	0.1286	0.1286	51.53
0.7542	9.0	1359	0.8619	0.1576	0.0272	0.1262	0.1263	27.97
0.5404	10.0	1510	0.8606	0.1763	0.0473	0.138	0.139	36.5
0.5404	11.0	1661	0.8644	0.1924	0.0364	0.1538	0.1537	32.84
0.5404	12.0	1812	0.8701	0.1832	0.036	0.1386	0.139	38.88
0.5404	13.0	1963	0.8776	0.1817	0.0399	0.1383	0.1398	34.36
0.4139	14.0	2114	0.8821	0.1888	0.0399	0.1407	0.1418	37.35
0.4139	15.0	2265	0.8826	0.1734	0.0367	0.1299	0.1304	37.9

Framework versions

• Transformers 4.36.2
• Pytorch 1.12.1+cu113
• Datasets 2.16.1
• Tokenizers 0.15.0