vectorzhou

/

Qwen2.5-1.5B-Instruct-SFT-OpenHerm-llection-v0.1-NashMDPG-lora-0605154608-epoch-6

like 0

• "accelerate.state.PartialState",
• "transformers.trainer_utils.SaveStrategy",
• "transformers.trainer_utils.IntervalStrategy",
• "transformers.integrations.deepspeed.HfTrainerDeepSpeedConfig",
• "transformers.trainer_pt_utils.AcceleratorConfig",
• "accelerate.utils.dataclasses.DeepSpeedPlugin",
• "torch.device",
• "accelerate.utils.dataclasses.DistributedType",
• "torch.float16",
• "transformers.training_args.OptimizerNames",
• "trl.trainer.nash_md_config.NashMDConfig",
• "transformers.trainer_utils.HubStrategy",
• "transformers.integrations.deepspeed.HfDeepSpeedConfig",
• "transformers.trainer_utils.SchedulerType"

How to fix it?