pt_model_test1 / README.md
4season's picture
Update README.md
9237182 verified
metadata
license: apache-2.0
language:
  - en

4season/pt_model_test1

Introduction

This model is test version, pt model.

We utilize state-of-the-art instruction fine-tuning methods.

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 1e-05
  • train_batch_size: 2
  • eval_batch_size: 8
  • seed: 42
  • distributed_type: multi-GPU
  • num_devices: 8
  • gradient_accumulation_steps: 8
  • total_train_batch_size: 128
  • total_eval_batch_size: 64
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • num_epochs: 1.0