bart-large-xsum-sentence-paraphrased-10-cnt-supervised-basic

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

• Loss: 2.1986
• Rouge1: 0.3258
• Rouge2: 0.1195
• Rougel: 0.2587
• Rougelsum: 0.2587

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: 1e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.06
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
2.7279	1.6667	5	2.2952	0.1646	0.0223	0.1266	0.1265
1.9318	3.3333	10	2.0213	0.2401	0.0725	0.191	0.1909
1.4017	5.0	15	1.9251	0.3029	0.1091	0.2401	0.2401
1.2787	6.6667	20	1.9361	0.3227	0.1212	0.2564	0.2563
0.6613	8.3333	25	1.9616	0.3272	0.1226	0.2594	0.2594
0.6822	10.0	30	2.0036	0.3287	0.1238	0.2607	0.2607
0.5634	11.6667	35	2.0476	0.3298	0.1251	0.263	0.2629
0.418	13.3333	40	2.0928	0.3317	0.1253	0.2642	0.2641
0.3908	15.0	45	2.1372	0.328	0.1219	0.2608	0.2606
0.2241	16.6667	50	2.1689	0.3261	0.1208	0.2593	0.2593
0.2268	18.3333	55	2.1893	0.3251	0.1192	0.2586	0.2585
0.2135	20.0	60	2.1986	0.3258	0.1195	0.2587	0.2587

Framework versions

• Transformers 4.57.3
• Pytorch 2.9.1+cu128
• Datasets 3.6.0
• Tokenizers 0.22.1