mt5-base-finetuned-test_30483_prefix_summarize

This model is a fine-tuned version of google/mt5-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.3527
• Rouge1: 23.6141
• Rouge2: 7.1791
• Rougel: 16.0152
• Rougelsum: 21.8213
• Gen Len: 69.64

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: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
7.4668	1.25	500	2.6478	10.4597	4.2457	8.7184	9.8473	18.49
3.3231	2.5	1000	2.5146	17.7315	6.0795	13.5384	16.6839	40.55
3.0956	3.75	1500	2.4512	20.2871	6.4051	14.8508	18.8768	51.35
2.9928	5.0	2000	2.4180	21.4196	6.629	15.2607	20.1471	57.92
2.8802	6.25	2500	2.4030	21.7949	6.7926	15.2506	20.338	61.05
2.8243	7.5	3000	2.3856	21.7075	6.7397	15.0044	20.1744	61.19
2.7646	8.75	3500	2.3847	22.4137	6.7644	14.9987	20.7797	63.81
2.7096	10.0	4000	2.3691	22.3403	6.9812	15.5411	20.6166	62.79
2.6758	11.25	4500	2.3612	23.6542	7.2355	15.9979	21.9807	69.83
2.6579	12.5	5000	2.3556	23.7473	7.5446	16.0314	21.917	69.75
2.651	13.75	5500	2.3557	23.9711	7.5018	16.2033	22.2811	69.29
2.639	15.0	6000	2.3527	23.6141	7.1791	16.0152	21.8213	69.64

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.0
• Datasets 2.12.0
• Tokenizers 0.13.3