flan-t5-plato

This model is a fine-tuned version of google/flan-t5-small on my philosophy-plato-qa dataset. It achieves the following results on the evaluation set:

• Loss: 1.6656
• Rouge1: 0.4143
• Rouge2: 0.3006
• Rougel: 0.3754
• Rougelsum: 0.3754
• Gen Len: 19.8982

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: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• 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
• num_epochs: 4
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
1.8847	1.0	3345	1.7122	0.4128	0.2977	0.3726	0.3725	19.9055
1.8401	2.0	6690	1.6803	0.4136	0.2995	0.3742	0.3741	19.8951
1.8116	3.0	10035	1.6679	0.4146	0.3007	0.3755	0.3754	19.8948
1.7995	4.0	13380	1.6656	0.4143	0.3006	0.3754	0.3754	19.8982

Framework versions

• Transformers 4.51.1
• Pytorch 2.5.1+cu124
• Datasets 3.5.0
• Tokenizers 0.21.0