AIR-hl's picture
Update README.md
16fe4fe verified
metadata
license: apache-2.0
base_model:
  - wzhouad/zephyr-7B-WPO-FP
  - HuggingFaceH4/mistral-7b-sft-beta
tags:
  - wpo
  - mistral
  - alignment
datasets:
  - HuggingFaceH4/ultrafeedback_binarized
pipeline_tag: text-generation
library_name: transformers

following wzhouad/zephyr-7B-WPO-FP

Transfer original weights from float32 to bfloat16 type