flan-t5-base-merged

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: 1.9282
• Rouge1: 0.4675
• Rouge2: 0.1579
• Rougel: 0.4313
• Bertscore: 0.8652
• Readability: 13.1666

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.06
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Bertscore	Readability
2.1647	1.0	3640	2.0145	0.4602	0.1545	0.4236	0.8626	13.5465
2.0892	2.0	7280	1.9591	0.4620	0.1549	0.4259	0.8636	13.3182
2.0151	3.0	10920	1.9376	0.4663	0.1571	0.4301	0.8648	13.2234
1.9793	4.0	14560	1.9282	0.4699	0.1599	0.4337	0.8656	13.1966
1.9679	5.0	18200	1.9269	0.4683	0.1583	0.4313	0.8653	13.2824

Framework versions

• Transformers 4.35.2
• Pytorch 2.2.1
• Datasets 2.19.0
• Tokenizers 0.15.2