flan-t5-base-mimic-med-reports

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

• Loss: 0.2032
• Rouge1: 52.8742
• Rouge2: 42.4294
• Rougel: 51.1178
• Rougelsum: 51.8773
• Meteor: 47.6053
• Bleu4: 14.2811
• Bleu-p1: 61.1865
• Bleu-p2: 43.5135
• Bleu-p3: 33.9223
• Bleu-p4: 25.8304
• Gen Len: 13.4702

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Meteor	Bleu4	Bleu-p1	Bleu-p2	Bleu-p3	Bleu-p4	Gen Len
0.2627	1.0	10497	0.2235	51.0687	39.674	49.6154	50.1361	45.6438	13.1526	60.4125	40.8134	31.1521	23.6299	13.2415
0.2376	2.0	20994	0.2102	51.5603	40.8339	49.8247	50.5212	46.2244	13.1941	60.8733	42.1622	32.8697	24.9374	13.1225
0.23	3.0	31491	0.2051	52.5731	41.7381	50.8502	51.6767	47.2270	14.0337	60.9681	42.8231	33.1248	25.1256	13.4702
0.2288	4.0	41988	0.2032	52.8742	42.4294	51.1178	51.8773	47.6053	14.2811	61.1865	43.5135	33.9223	25.8304	13.4702

Framework versions

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