lewtun

/

Llama-3.1-8B-SFT-QLoRA-packing

like 0

• "transformers.trainer_utils.SchedulerType",
• "torch.float32",
• "torch.uint8",
• "transformers.utils.quantization_config.QuantizationMethod",
• "transformers.trainer_utils.IntervalStrategy",
• "transformers.training_args.OptimizerNames",
• "transformers.trainer_utils.HubStrategy",
• "accelerate.state.PartialState",
• "torch.device",
• "transformers.trainer_pt_utils.AcceleratorConfig",
• "accelerate.utils.dataclasses.DistributedType",
• "trl.trainer.sft_config.SFTConfig",
• "transformers.utils.quantization_config.BitsAndBytesConfig"

How to fix it?