summary_vacancy

This model is a fine-tuned version of d0rj/rut5-base-summ on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.9030
• Rouge1: 0.2908
• Rouge2: 0.0952
• Rougel: 0.2728
• Rougelsum: 0.2728
• Gen Len: 65.6667

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: 1
• eval_batch_size: 1
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 25
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	60	2.7805	0.4024	0.1111	0.3511	0.3511	70.3333
No log	2.0	120	2.7797	0.1746	0.0	0.127	0.127	34.3333
No log	3.0	180	2.7807	0.2222	0.0	0.1746	0.1746	38.0
No log	4.0	240	2.7872	0.3966	0.1524	0.349	0.349	46.6667
No log	5.0	300	2.7995	0.2758	0.1176	0.2758	0.2758	50.3333
No log	6.0	360	2.8120	0.2758	0.1176	0.2758	0.2758	49.3333
No log	7.0	420	2.8270	0.2758	0.1176	0.2758	0.2758	51.0
No log	8.0	480	2.8371	0.2758	0.1176	0.2758	0.2758	47.6667
1.9576	9.0	540	2.8480	0.2728	0.1143	0.2728	0.2728	73.3333
1.9576	10.0	600	2.8573	0.2804	0.1333	0.2804	0.2804	63.6667
1.9576	11.0	660	2.8661	0.2888	0.1333	0.2888	0.2888	48.6667
1.9576	12.0	720	2.8718	0.2787	0.1081	0.2787	0.2787	51.0
1.9576	13.0	780	2.8786	0.1075	0.0	0.1075	0.1075	49.6667
1.9576	14.0	840	2.8817	0.2827	0.1333	0.2827	0.2827	57.3333
1.9576	15.0	900	2.8865	0.293	0.1333	0.293	0.293	39.0
1.9576	16.0	960	2.8900	0.2997	0.1333	0.2997	0.2997	57.6667
1.7164	17.0	1020	2.8941	0.2997	0.1333	0.2997	0.2997	71.6667
1.7164	18.0	1080	2.8968	0.2997	0.1143	0.2997	0.2997	76.3333
1.7164	19.0	1140	2.8983	0.293	0.1143	0.275	0.275	88.6667
1.7164	20.0	1200	2.8998	0.3194	0.0926	0.3019	0.3019	77.3333
1.7164	21.0	1260	2.9005	0.2908	0.0952	0.2728	0.2728	65.6667
1.7164	22.0	1320	2.9020	0.2908	0.0952	0.2728	0.2728	65.6667
1.7164	23.0	1380	2.9027	0.2908	0.0952	0.2728	0.2728	65.6667
1.7164	24.0	1440	2.9026	0.2908	0.0952	0.2728	0.2728	65.6667
1.647	25.0	1500	2.9030	0.2908	0.0952	0.2728	0.2728	65.6667

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.17.1
• Tokenizers 0.15.2