Hugging Face's logo

samuellimabraz
/
t5-base-qlora-finetune-tweetsumm

PEFT
Safetensors
Generated from Trainer
Eval Results
Model card Files Community
t5-base-qlora-finetune-tweetsumm / README.md
samuellimabraz's picture
samuellimabraz
End of training
f16d96e verified
preview code
|
raw
history blame contribute delete
No virus
2.57 kB
metadata
base_model: google-t5/t5-base
datasets:
  - Andyrasika/TweetSumm-tuned
library_name: peft
license: apache-2.0
metrics:
  - rouge
  - f1
  - precision
  - recall
tags:
  - generated_from_trainer
model-index:
  - name: t5-base-qlora-finetune-tweetsumm-1724817707
    results:
      - task:
          type: summarization
          name: Summarization
        dataset:
          name: Andyrasika/TweetSumm-tuned
          type: Andyrasika/TweetSumm-tuned
        metrics:
          - type: rouge
            value: 0.4708
            name: Rouge1
          - type: f1
            value: 0.8942
            name: F1
          - type: precision
            value: 0.8941
            name: Precision
          - type: recall
            value: 0.8945
            name: Recall

t5-base-qlora-finetune-tweetsumm-1724817707

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

  • Loss: 1.7934
  • Rouge1: 0.4708
  • Rouge2: 0.2246
  • Rougel: 0.3984
  • Rougelsum: 0.4357
  • Gen Len: 49.4091
  • F1: 0.8942
  • Precision: 0.8941
  • Recall: 0.8945

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: 0.0005
  • 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
  • lr_scheduler_warmup_steps: 2
  • num_epochs: 3
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Rouge1 Rouge2 Rougel Rougelsum Gen Len F1 Precision Recall
2.062 1.0 110 1.8472 0.4633 0.2177 0.3919 0.428 49.7273 0.8911 0.8897 0.8927
1.7853 2.0 220 1.8120 0.4633 0.2203 0.3941 0.4285 49.4273 0.8953 0.8945 0.8963
1.5952 3.0 330 1.7934 0.4708 0.2246 0.3984 0.4357 49.4091 0.8942 0.8941 0.8945

Framework versions

  • PEFT 0.12.1.dev0
  • Transformers 4.44.0
  • Pytorch 2.4.0
  • Datasets 2.21.0
  • Tokenizers 0.19.1