metadata
license: llama3.1
library_name: transformers
base_model:
- nbeerbower/Llama-3.1-Saoirse-70B
datasets:
- nbeerbower/Schule-DPO
- nbeerbower/Arkhaios-DPO
- nbeerbower/Purpura-DPO
- antiven0m/physical-reasoning-dpo
- jondurbin/truthy-dpo-v0.1
llama3.1-kartoffeldes-70B
Llama-3.1-Saoirse-70B finetuned on various datasets.
Method
ORPO tuned with 8x A100 for 2 epochs.