dialouge_summarization_model

This model is a fine-tuned version of t5-small on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.5091
• Rouge1: 0.3626
• Rouge2: 0.1277
• Rougel: 0.3026
• Rougelsum: 0.3025
• Gen Len: 18.818

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: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
1.7341	1.0	779	1.6330	0.3367	0.102	0.2786	0.2784	18.802
1.4962	2.0	1558	1.5773	0.3461	0.1095	0.2848	0.2845	18.832
1.4727	3.0	2337	1.5615	0.3508	0.1169	0.2923	0.2921	18.786
1.4291	4.0	3116	1.5377	0.3544	0.1184	0.2945	0.2941	18.756
1.4146	5.0	3895	1.5317	0.355	0.1205	0.2955	0.2953	18.774
1.3913	6.0	4674	1.5183	0.3592	0.1247	0.3009	0.3007	18.794
1.3877	7.0	5453	1.5153	0.3611	0.1252	0.3009	0.3008	18.806
1.3744	8.0	6232	1.5105	0.3635	0.1284	0.303	0.3029	18.812
1.3627	9.0	7011	1.5106	0.3644	0.1291	0.3038	0.3037	18.824
1.3624	10.0	7790	1.5091	0.3626	0.1277	0.3026	0.3025	18.818

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.0
• Tokenizers 0.19.1