sfulay

/

zephyr-7b-dpo-full-gpt_consistent-reward-scale-1-rpo

like 0

• "transformers.trainer_utils.HubStrategy",
• "transformers.integrations.deepspeed.HfTrainerDeepSpeedConfig",
• "transformers.training_args.OptimizerNames",
• "torch.device",
• "transformers.trainer_pt_utils.AcceleratorConfig",
• "accelerate.utils.dataclasses.DistributedType",
• "trl.trainer.dpo_config.DPOConfig",
• "transformers.integrations.deepspeed.HfDeepSpeedConfig",
• "accelerate.utils.dataclasses.DeepSpeedPlugin",
• "transformers.trainer_utils.IntervalStrategy",
• "torch.bfloat16",
• "transformers.trainer_utils.SchedulerType",
• "trl.trainer.dpo_config.FDivergenceType",
• "accelerate.state.PartialState"

How to fix it?