> This page is for version 0.5.0 · 26.06.
> For other versions, use one of these documentation indexes:
> - Nightly (default): https://docs.nvidia.com/nemo/automodel/nightly/llms.txt
> - Latest: https://docs.nvidia.com/nemo/automodel/latest/llms.txt
> - 0.7.0 · 26.10: https://docs.nvidia.com/nemo/automodel/v0.7/llms.txt
> - 0.5.0 · 26.06: https://docs.nvidia.com/nemo/automodel/v0.5/llms.txt
> - 0.4.0 · 26.04: https://docs.nvidia.com/nemo/automodel/v0.4/llms.txt

> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.nvidia.com/nemo/automodel/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.nvidia.com/nemo/automodel/_mcp/server.

# nemo_automodel.shared.peft_config

PEFT configuration without adapter or distributed runtime dependencies.

## Module Contents

### Classes

| Name                                                          | Description                                                                 |
| ------------------------------------------------------------- | --------------------------------------------------------------------------- |
| [`PeftConfig`](#nemo_automodel-shared-peft_config-PeftConfig) | Declarative adapter settings shared by execution and checkpoint boundaries. |

### API

```python
class nemo_automodel.shared.peft_config.PeftConfig(
    target_modules: list[str] | str | None = list(),
    exclude_modules: list[str] | str | None = list(),
    match_all_linear: bool = False,
    dim: int = 8,
    alpha: int = 32,
    use_dora: bool = False,
    dropout: float = 0.0,
    dropout_position: typing.Literal['pre', 'post'] = 'post',
    lora_A_init: str = 'xavier',
    lora_dtype: torch.dtype | str | None = None,
    use_memory_efficient_lora: bool = True,
    use_triton: bool = False,
    moe_rank_scaling: bool = False,
    expert_weight_format: typing.Literal['unquantized', 'mxfp4'] = 'unquantized'
)
```

Dataclass

Declarative adapter settings shared by execution and checkpoint boundaries.

**`alpha`** `int = 32`

---

**`dim`** `int = 8`

---

**`dropout`** `float = 0.0`

---

**`dropout_position`** `Literal['pre', 'post'] = 'post'`

---

**`exclude_modules`** `list[str] | str | None = field(default_factory=list)`

---

**`expert_weight_format`** `Literal['unquantized', 'mxfp4'] = 'unquantized'`

---

**`lora_A_init`** `str = 'xavier'`

---

**`lora_dtype`** `dtype | str | None = None`

---

**`match_all_linear`** `bool = False`

---

**`moe_rank_scaling`** `bool = False`

---

**`target_modules`** `list[str] | str | None = field(default_factory=list)`

---

**`use_dora`** `bool = False`

---

**`use_memory_efficient_lora`** `bool = True`

---

**`use_triton`** `bool = False`

---