dataset_name: CarperAI/openai_summarize_tldr
model_name: /home/toolkit/huggingface/tldr_sft_pythia14m_fp32
reward_model_name: /home/toolkit/huggingface/tldr_gptrm_sft_pythia14m_fp32_2epochs
gold_reward_model_name: /home/toolkit/huggingface/tldr_gptrm_sft_pythia410m_amp16_lr5e-6_1epoch
steps: 2000
load_in_8bit: False
bf16: False
fp16: False
batch_size: 128
mini_batch_size: 32
gradient_accumulation_steps: 1
learning_rate: 1e-6
eval_freq: 1