ft5

This model is a fine-tuned version of reza-alipour/ft5 on the reza-alipour/Text-Edit-Instruct-Preprocessed-4m dataset. It achieves the following results on the evaluation set:

  • Loss: 0.3032
  • Rouge1: 86.5313
  • Rouge2: 75.3351
  • Rougel: 85.9565
  • Rougelsum: 85.9785
  • Gen Len: 24.5688

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: 7e-05
  • train_batch_size: 4
  • eval_batch_size: 4
  • seed: 1337
  • gradient_accumulation_steps: 8
  • total_train_batch_size: 32
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: constant
  • num_epochs: 1.0

Training results

Training Loss Epoch Step Validation Loss Rouge1 Rouge2 Rougel Rougelsum Gen Len
0.3601 1.0 31042 0.3032 86.5313 75.3351 85.9565 85.9785 24.5688

Framework versions

  • Transformers 4.36.0.dev0
  • Pytorch 2.1.1
  • Datasets 2.15.0
  • Tokenizers 0.15.0
Downloads last month
2
Safetensors
Model size
0.8B params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for reza-alipour/ft5

Unable to build the model tree, the base model loops to the model itself. Learn more.

Dataset used to train reza-alipour/ft5

Evaluation results

  • Rouge1 on reza-alipour/Text-Edit-Instruct-Preprocessed-4m
    self-reported
    86.531