hiyouga/LlamaFactory · error · ValueError

Please specify `export_dir` to save model.

Error message

Please specify `export_dir` to save model.

What it means

`export_model` (src/llamafactory/train/tuner.py:180) requires a destination; if `model_args.export_dir` is None it refuses to proceed because model.save_pretrained would have nowhere to write. export_dir is the `export_dir:` key of the export YAML.

Source

Thrown at src/llamafactory/train/tuner.py:180

def run_exp(args: Optional[dict[str, Any]] = None, callbacks: Optional[list["TrainerCallback"]] = None) -> None:
    args = read_args(args)
    if "-h" in args or "--help" in args:
        get_train_args(args)

    ray_args = get_ray_args(args)
    callbacks = callbacks or []
    if ray_args.use_ray:
        _ray_training_function(ray_args, config={"args": args, "callbacks": callbacks})
    else:
        _training_function(config={"args": args, "callbacks": callbacks})


def export_model(args: Optional[dict[str, Any]] = None) -> None:
    model_args, data_args, finetuning_args, _ = get_infer_args(args)

    if model_args.export_dir is None:
        raise ValueError("Please specify `export_dir` to save model.")

    if model_args.adapter_name_or_path is not None and model_args.export_quantization_bit is not None:
        raise ValueError("Please merge adapters before quantizing the model.")

    tokenizer_module = load_tokenizer(model_args)
    tokenizer = tokenizer_module["tokenizer"]
    processor = tokenizer_module["processor"]
    template = get_template_and_fix_tokenizer(tokenizer, data_args)
    model = load_model(tokenizer, model_args, finetuning_args)  # must after fixing tokenizer to resize vocab

    if getattr(model, "quantization_method", None) is not None and model_args.adapter_name_or_path is not None:
        raise ValueError("Cannot merge adapters to a quantized model.")

    if not isinstance(model, PreTrainedModel):
        raise ValueError("The model is not a `PreTrainedModel`, export aborted.")

    if getattr(model, "quantization_method", None) is not None:  # quantized model adopts float16 type
        setattr(model.config, "torch_dtype", torch.float16)

View on GitHub (pinned to f28afaf635)

Solutions

1. Add `export_dir: /path/to/output` to the export YAML.
2. If calling programmatically, pass {'export_dir': ...} in the args dict.
3. Ensure the directory is writable and has space for the sharded weights.

Example fix

# before (yaml)
model_name_or_path: meta-llama/Llama-3-8b
adapter_name_or_path: output/lora

# after
model_name_or_path: meta-llama/Llama-3-8b
adapter_name_or_path: output/lora
export_dir: output/merged-model

assert export_dir, "export config must set export_dir"

When it happens

Trigger: Running `llamafactory-cli export` with a YAML missing the `export_dir:` line, or calling export_model(args) with a dict that omits it.

Common situations: Quickly built export configs from a train config where export_dir was never added; automation scripts that template export configs.

Related errors

• Please merge adapters before quantizing the model.
• Cannot merge adapters to a quantized model.
• Please specify peft_config to merge and export model.
• Please upgrade `transformers` to 4.34.0
• Unable to process key {key}

AI-assisted analysis of hiyouga/LlamaFactory@f28afaf635 (2026-08-14). Data as JSON: /api/errors/fdee1f5949b38ebc. Report an issue: GitHub.