T5-medi

This model is a fine-tuned version of google/flan-t5-small on Medical QA dataset (wesley7137/qa_dataset). It is just for educational purpose and does not provide accurate results. It achieves the following results on the evaluation set:

• Loss: 0.7130
• Rouge1: 0.3189
• Rouge2: 0.2700
• Rougel: 0.3068
• Rougelsum: 0.3149

Model description

              This is a fine-tuned T5 model for Question-Answering tasks in Medical Field

Intended uses & limitations

May struggle with creative or subjective content. Requires fine-tuning for different tasks

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.001
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
No log	1.0	223	0.8782	0.3148	0.2630	0.3010	0.3098
No log	2.0	446	0.7820	0.3148	0.2650	0.3026	0.3111
1.1386	3.0	669	0.7456	0.3170	0.2681	0.3049	0.3131
1.1386	4.0	892	0.7259	0.3198	0.2699	0.3072	0.3156
0.7884	5.0	1115	0.7130	0.3189	0.2700	0.3068	0.3149

Framework versions

• Transformers 4.39.3
• Pytorch 2.1.2
• Datasets 2.18.0
• Tokenizers 0.15.2