prompt_strategies.dpo.chatml
prompt_strategies.dpo.chatml
DPO strategies for chatml
Functions
| Name | Description |
|---|---|
| argilla_chat | for argilla/dpo-mix-7k conversations |
| icr | chatml transforms for datasets with system, input, chosen, rejected |
| intel | For Intel Orca DPO Pairs |
| ultra | for ultrafeedback binarized conversations |
argilla_chat
prompt_strategies.dpo.chatml.argilla_chat(cfg, **kwargs)for argilla/dpo-mix-7k conversations
icr
prompt_strategies.dpo.chatml.icr(cfg, **kwargs)chatml transforms for datasets with system, input, chosen, rejected ex. https://huggingface.co/datasets/argilla/distilabel-intel-orca-dpo-pairs
intel
prompt_strategies.dpo.chatml.intel(cfg, **kwargs)For Intel Orca DPO Pairs
ultra
prompt_strategies.dpo.chatml.ultra(cfg, **kwargs)for ultrafeedback binarized conversations