Files
pythia410m-sft-tldr/code/configs/vllm_hh.yml
ModelHub XC 4bb3a5553f 初始化项目,由ModelHub XC社区提供模型
Model: mnoukhov/pythia410m-sft-tldr
Source: Original Platform
2026-05-30 14:21:21 +08:00

12 lines
233 B
YAML

output_dir: hh-rlhf-gen_pythia2.8b-dpo
dataset_name: "sophiex/hh-rlhf"
model_name: null
lora_model: True
train_split: train[:20000]
batch_size: 4
max_prompt_length: 256
max_new_tokens: 256
temperature: 0.7
top_p: 1.0
dtype: bfloat16