story2dialogue-SODA-T5-LoRA

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

• Loss: 1.7120
• Rouge1: 0.1270
• Rouge2: 0.0513
• Rougel: 0.1114
• Rougelsum: 0.1114
• Bleu: 0.0000
• Bleu1: 0.7504
• Bleu2: 0.3722
• Bleu3: 0.1851
• Bleu4: 0.0979
• Avg Jaccard: 0.0647
• Gen Length: 10.2755

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: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Bleu	Bleu1	Bleu2	Bleu3	Bleu4	Avg Jaccard	Gen Length
2.1429	0.2938	1000	1.9285	0.1306	0.0518	0.1135	0.1135	0.0000	0.7539	0.3521	0.1620	0.0764	0.0657	10.3782
2.0224	0.5876	2000	1.8245	0.1298	0.0518	0.1129	0.1129	0.0000	0.7606	0.3651	0.1718	0.0825	0.0661	10.3641
1.9642	0.8814	3000	1.7781	0.1291	0.0524	0.1128	0.1128	0.0000	0.7564	0.3765	0.1820	0.0903	0.0657	10.2462
1.918	1.1751	4000	1.7482	0.1282	0.0510	0.1117	0.1117	0.0000	0.7504	0.3745	0.1845	0.0937	0.0658	10.3853
1.9062	1.4689	5000	1.7299	0.1257	0.0490	0.1101	0.1101	0.0000	0.7484	0.3746	0.1868	0.0958	0.0649	10.3065
1.8854	1.7627	6000	1.7120	0.1270	0.0513	0.1114	0.1114	0.0000	0.7504	0.3722	0.1851	0.0979	0.0647	10.2755

Framework versions

• PEFT 0.15.2
• Transformers 4.52.4
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.2