hiyouga/LlamaFactory · error · AttributeError

Please update `transformers`.

Error message

Please update `transformers`.

What it means

Error "Please update `transformers`." thrown in hiyouga/LlamaFactory.

Source

Thrown at src/llamafactory/train/dpo/trainer.py:90

        self._precomputed_eval_ref_log_probs = False
        self._peft_has_been_casted_to_bf16 = False

        self.ref_model = ref_model
        self._stored_metrics = defaultdict(lambda: defaultdict(list))

        # dpo hyperparams
        self.beta = finetuning_args.pref_beta
        self.loss_type = finetuning_args.pref_loss
        self.ftx_gamma = finetuning_args.pref_ftx
        self.bco_gemma = finetuning_args.pref_bco_weight
        self.label_smoothing = finetuning_args.dpo_label_smoothing
        self.simpo_gamma = finetuning_args.simpo_gamma
        self.ld_alpha = finetuning_args.ld_alpha

        Trainer.__init__(self, model=model, **kwargs)
        self.model_accepts_loss_kwargs = False  # overwrite trainer's default behavior
        if not hasattr(self, "accelerator"):
            raise AttributeError("Please update `transformers`.")

        warnings.simplefilter("ignore")  # remove gc warnings on ref model

        if ref_model is not None:
            if self.is_deepspeed_enabled:
                if not (
                    getattr(ref_model, "is_loaded_in_8bit", False) or getattr(ref_model, "is_loaded_in_4bit", False)
                ):  # quantized models are already set on the correct device
                    self.ref_model = prepare_deepspeed(self.ref_model, self.accelerator)
            elif self.is_fsdp_enabled:
                if self.accelerator.is_fsdp2:
                    from accelerate.utils.fsdp_utils import fsdp2_prepare_model

                    self.ref_model = fsdp2_prepare_model(self.accelerator, self.ref_model)
                else:
                    self.ref_model = prepare_fsdp(self.ref_model, self.accelerator)
            else:
                self.ref_model = self.accelerator.prepare_model(self.ref_model, evaluation_mode=True)

View on GitHub (pinned to f28afaf635)

Solutions

1. Upgrade transformers to a version whose Trainer sets self.accelerator (a recent 4.x release): pip install -U transformers.

Example fix

pip install -U transformers

When it happens

Trigger: Thrown at src/llamafactory/train/dpo/trainer.py:90 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of hiyouga/LlamaFactory@f28afaf635 (2026-08-14). Data as JSON: /api/errors/0c9f26a022868298. Report an issue: GitHub.