Files
pythia410m-sft-tldr/code/configs/dpo1b_relabel_comparisons.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

13 lines
354 B
YAML

output_dir: /home/toolkit/huggingface/openai_summarize_comparisons_relabelled_margin
mode: relabel
model_name: mnoukhov/pythia410m-tldrprompt-dpo1b-adapter
dataset_name: mnoukhov/openai_summarize_comparisons_tldrprompt
eval_split: train
use_peft: False
beta: 0.5
load_in_8bit: False
bf16: False
fp16: True
per_device_eval_batch_size: 8
warmup_steps: 150