tuanio

/

MoE-LLaVA-Qwen1.5-1.8Bx4-Top2

like 1

• "torch.bfloat16",
• "transformers.trainer_utils.IntervalStrategy",
• "moellava.train.train.TrainingArguments",
• "accelerate.utils.dataclasses.DistributedType",
• "transformers.trainer_utils.HubStrategy",
• "transformers.trainer_utils.SchedulerType",
• "accelerate.utils.dataclasses.DeepSpeedPlugin",
• "transformers.training_args.OptimizerNames",
• "accelerate.state.PartialState",
• "transformers.integrations.deepspeed.HfTrainerDeepSpeedConfig",
• "torch.device"

How to fix it?