DanSumT5-smallV_59491

This model is a fine-tuned version of Danish-summarisation/DanSumT5-small on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.4715
• Rouge1: 34.3324
• Rouge2: 11.2932
• Rougel: 20.8529
• Rougelsum: 31.803
• Gen Len: 126.0591

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: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 11

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	0.97	29	2.6350	32.3407	9.3733	18.8907	29.6731	126.4726
No log	1.98	59	2.5732	33.2864	10.0998	19.4234	30.5031	126.5021
No log	2.99	89	2.5433	34.0216	10.7327	20.0836	31.3728	125.9873
No log	4.0	119	2.5209	34.1794	10.8803	20.414	31.4901	126.654
No log	4.97	148	2.5065	34.1598	10.9122	20.5029	31.5882	126.443
No log	5.98	178	2.4942	34.3057	10.8906	20.6468	31.628	126.8354
No log	6.99	208	2.4843	34.2147	10.9465	20.4998	31.5286	126.3418
No log	8.0	238	2.4768	34.3636	11.2347	20.6723	31.7775	126.3165
No log	8.97	267	2.4736	34.1256	11.3376	20.6823	31.6962	126.2447
No log	9.98	297	2.4718	34.3458	11.2773	20.8056	31.8051	125.9789
No log	10.72	319	2.4715	34.3324	11.2932	20.8529	31.803	126.0591

Framework versions

• Transformers 4.30.2
• Pytorch 1.12.1+git7548e2f
• Datasets 2.13.2
• Tokenizers 0.13.3