SOLAR_merge_DPOv3 / README.md
genne's picture
Update README.md
8881819 verified
|
raw
history blame contribute delete
No virus
395 Bytes
metadata
base_model: ENERGY-DRINK-LOVE/SOLAR_merge
tags:
  - trl
  - dpo
  - generated_from_trainer
model-index:
  - name: nhn_dpo_v3_SOLAR_merge_DPO
    results: []
license: apache-2.0

Model

  • trained on custom DPO dataset
    • dedup
    • ~20000??

Base Moel

  • ENERGY-DRINK-LOVE/SOLAR_merge

Framework versions

  • Transformers 4.38.1
  • Pytorch 2.2.1+cu118
  • Datasets 2.17.1
  • Tokenizers 0.15.2