t-m-6's picture
Create README.md
258d662 verified
metadata
language:
  - ja
base_model:
  - Rakuten/RakutenAI-2.0-mini-instruct

Base Model

Dataset

Parameter

  • learning_rate: 1e-6
  • num_train_epochs: 10
  • per_device_train_batch_size: 64
  • per_device_eval_batch_size: 64
  • gradient_accumulation_steps: 1