flan-t5-base-finetuned-QLoRA-v2

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

  • Loss: 1.0254
  • Rouge1: 0.244
  • Rouge2: 0.111
  • Rougel: 0.2032
  • Rougelsum: 0.2292

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: 3e-05
  • train_batch_size: 4
  • eval_batch_size: 4
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 10

Training results

Training Loss Epoch Step Validation Loss Rouge1 Rouge2 Rougel Rougelsum
3.0551 1.0 500 2.2941 0.2336 0.1092 0.1969 0.217
1.6422 2.0 1000 1.1665 0.2459 0.1088 0.1991 0.227
1.4067 3.0 1500 1.0762 0.2462 0.1089 0.1982 0.2296
1.2856 4.0 2000 1.0518 0.2448 0.1112 0.2036 0.2298
1.3478 5.0 2500 1.0393 0.2458 0.1125 0.2056 0.2303
1.2114 6.0 3000 1.0340 0.2497 0.1145 0.2084 0.2333
1.3311 7.0 3500 1.0298 0.2479 0.1143 0.207 0.233
1.3081 8.0 4000 1.0270 0.2448 0.1112 0.2035 0.2301
1.1794 9.0 4500 1.0258 0.2449 0.1112 0.2036 0.2301
1.2407 10.0 5000 1.0254 0.244 0.111 0.2032 0.2292

Framework versions

  • PEFT 0.8.2
  • Transformers 4.37.0
  • Pytorch 2.1.2
  • Datasets 2.1.0
  • Tokenizers 0.15.1
Downloads last month
41
Inference API
Unable to determine this model’s pipeline type. Check the docs .

Model tree for RMWeerasinghe/flan-t5-base-finetuned-QLoRA-v2

Adapter
(130)
this model

Dataset used to train RMWeerasinghe/flan-t5-base-finetuned-QLoRA-v2