results_t5small

This model is a fine-tuned version of t5-small on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3877
• Rouge1: 0.9032
• Rouge2: 0.8342
• Rougel: 0.8846
• Gen Len: 239.3236

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: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Gen Len
0.6906	0.4494	200	0.5593	0.8712	0.7784	0.8476	238.173
0.4308	0.8989	400	0.4529	0.8891	0.8139	0.8699	239.2697
0.4254	1.3483	600	0.4310	0.8934	0.8208	0.8749	239.3236
0.353	1.7978	800	0.4214	0.8967	0.826	0.8782	239.2989
0.4512	2.2472	1000	0.4137	0.898	0.8279	0.8797	239.3056
0.3438	2.6966	1200	0.4069	0.8993	0.829	0.8808	239.3191
0.4563	3.1461	1400	0.4038	0.8993	0.8297	0.881	239.3236
0.3577	3.5955	1600	0.3988	0.9003	0.8307	0.8818	239.3236
0.2402	4.0449	1800	0.3982	0.9011	0.832	0.8828	239.3213
0.3261	4.4944	2000	0.3941	0.9014	0.8319	0.883	239.3236
0.4044	4.9438	2200	0.3936	0.9014	0.8326	0.8833	239.3236
0.2152	5.3933	2400	0.3943	0.9022	0.8326	0.8835	239.3236
0.5285	5.8427	2600	0.3897	0.902	0.8333	0.8839	239.3236
0.2584	6.2921	2800	0.3921	0.9022	0.8332	0.8839	239.3213
0.2767	6.7416	3000	0.3904	0.9026	0.8341	0.8844	239.3236
0.3118	7.1910	3200	0.3903	0.903	0.834	0.8843	239.3236
0.2102	7.6404	3400	0.3894	0.9033	0.8342	0.8848	239.3236
0.2771	8.0899	3600	0.3886	0.903	0.8342	0.8844	239.3236
0.3008	8.5393	3800	0.3882	0.9028	0.8343	0.8845	239.3236
0.2221	8.9888	4000	0.3877	0.9032	0.8346	0.8847	239.3236
0.3443	9.4382	4200	0.3874	0.9034	0.8344	0.8847	239.3236
0.3818	9.8876	4400	0.3877	0.9032	0.8342	0.8846	239.3236

Framework versions

• Transformers 4.41.2
• Pytorch 2.2.1+cu121
• Datasets 2.19.2
• Tokenizers 0.19.1