# {py:mod}`evorl.networks.linear`

```{py:module} evorl.networks.linear
```

```{autodoc2-docstring} evorl.networks.linear
:parser: autodoc2_docstrings_parser
:allowtitles:
```

## Module Contents

### Classes

````{list-table}
:class: autosummary longtable
:align: left

* - {py:obj}`MLP <evorl.networks.linear.MLP>`
  - ```{autodoc2-docstring} evorl.networks.linear.MLP
    :parser: autodoc2_docstrings_parser
    :summary:
    ```
* - {py:obj}`SNMLP <evorl.networks.linear.SNMLP>`
  - ```{autodoc2-docstring} evorl.networks.linear.SNMLP
    :parser: autodoc2_docstrings_parser
    :summary:
    ```
````

### Functions

````{list-table}
:class: autosummary longtable
:align: left

* - {py:obj}`make_discrete_q_network <evorl.networks.linear.make_discrete_q_network>`
  - ```{autodoc2-docstring} evorl.networks.linear.make_discrete_q_network
    :parser: autodoc2_docstrings_parser
    :summary:
    ```
* - {py:obj}`make_mlp <evorl.networks.linear.make_mlp>`
  - ```{autodoc2-docstring} evorl.networks.linear.make_mlp
    :parser: autodoc2_docstrings_parser
    :summary:
    ```
* - {py:obj}`make_policy_network <evorl.networks.linear.make_policy_network>`
  - ```{autodoc2-docstring} evorl.networks.linear.make_policy_network
    :parser: autodoc2_docstrings_parser
    :summary:
    ```
* - {py:obj}`make_q_network <evorl.networks.linear.make_q_network>`
  - ```{autodoc2-docstring} evorl.networks.linear.make_q_network
    :parser: autodoc2_docstrings_parser
    :summary:
    ```
* - {py:obj}`make_v_network <evorl.networks.linear.make_v_network>`
  - ```{autodoc2-docstring} evorl.networks.linear.make_v_network
    :parser: autodoc2_docstrings_parser
    :summary:
    ```
* - {py:obj}`make_vmap_mlp <evorl.networks.linear.make_vmap_mlp>`
  - ```{autodoc2-docstring} evorl.networks.linear.make_vmap_mlp
    :parser: autodoc2_docstrings_parser
    :summary:
    ```
````

### Data

````{list-table}
:class: autosummary longtable
:align: left

* - {py:obj}`ActivationFn <evorl.networks.linear.ActivationFn>`
  - ```{autodoc2-docstring} evorl.networks.linear.ActivationFn
    :parser: autodoc2_docstrings_parser
    :summary:
    ```
* - {py:obj}`Initializer <evorl.networks.linear.Initializer>`
  - ```{autodoc2-docstring} evorl.networks.linear.Initializer
    :parser: autodoc2_docstrings_parser
    :summary:
    ```
````

### API

````{py:data} ActivationFn
:canonical: evorl.networks.linear.ActivationFn
:value: >
   None

```{autodoc2-docstring} evorl.networks.linear.ActivationFn
:parser: autodoc2_docstrings_parser
```

````

````{py:data} Initializer
:canonical: evorl.networks.linear.Initializer
:value: >
   None

```{autodoc2-docstring} evorl.networks.linear.Initializer
:parser: autodoc2_docstrings_parser
```

````

`````{py:class} MLP
:canonical: evorl.networks.linear.MLP

Bases: {py:obj}`flax.linen.Module`

```{autodoc2-docstring} evorl.networks.linear.MLP
:parser: autodoc2_docstrings_parser
```

````{py:attribute} activation
:canonical: evorl.networks.linear.MLP.activation
:type: evorl.networks.linear.ActivationFn
:value: >
   None

```{autodoc2-docstring} evorl.networks.linear.MLP.activation
:parser: autodoc2_docstrings_parser
```

````

````{py:attribute} activation_final
:canonical: evorl.networks.linear.MLP.activation_final
:type: evorl.networks.linear.ActivationFn | None
:value: >
   None

```{autodoc2-docstring} evorl.networks.linear.MLP.activation_final
:parser: autodoc2_docstrings_parser
```

````

````{py:attribute} kernel_init
:canonical: evorl.networks.linear.MLP.kernel_init
:type: evorl.networks.linear.Initializer
:value: >
   'lecun_uniform(...)'

```{autodoc2-docstring} evorl.networks.linear.MLP.kernel_init
:parser: autodoc2_docstrings_parser
```

````

````{py:attribute} layer_sizes
:canonical: evorl.networks.linear.MLP.layer_sizes
:type: collections.abc.Sequence[int]
:value: >
   None

```{autodoc2-docstring} evorl.networks.linear.MLP.layer_sizes
:parser: autodoc2_docstrings_parser
```

````

````{py:attribute} norm_layer
:canonical: evorl.networks.linear.MLP.norm_layer
:type: flax.linen.Module | None
:value: >
   None

```{autodoc2-docstring} evorl.networks.linear.MLP.norm_layer
:parser: autodoc2_docstrings_parser
```

````

````{py:attribute} use_bias
:canonical: evorl.networks.linear.MLP.use_bias
:type: bool
:value: >
   True

```{autodoc2-docstring} evorl.networks.linear.MLP.use_bias
:parser: autodoc2_docstrings_parser
```

````

`````

`````{py:class} SNMLP
:canonical: evorl.networks.linear.SNMLP

Bases: {py:obj}`flax.linen.Module`

```{autodoc2-docstring} evorl.networks.linear.SNMLP
:parser: autodoc2_docstrings_parser
```

````{py:attribute} activation
:canonical: evorl.networks.linear.SNMLP.activation
:type: evorl.networks.linear.ActivationFn
:value: >
   None

```{autodoc2-docstring} evorl.networks.linear.SNMLP.activation
:parser: autodoc2_docstrings_parser
```

````

````{py:attribute} activation_final
:canonical: evorl.networks.linear.SNMLP.activation_final
:type: evorl.networks.linear.ActivationFn | None
:value: >
   None

```{autodoc2-docstring} evorl.networks.linear.SNMLP.activation_final
:parser: autodoc2_docstrings_parser
```

````

````{py:attribute} kernel_init
:canonical: evorl.networks.linear.SNMLP.kernel_init
:type: evorl.networks.linear.Initializer
:value: >
   'lecun_uniform(...)'

```{autodoc2-docstring} evorl.networks.linear.SNMLP.kernel_init
:parser: autodoc2_docstrings_parser
```

````

````{py:attribute} layer_sizes
:canonical: evorl.networks.linear.SNMLP.layer_sizes
:type: collections.abc.Sequence[int]
:value: >
   None

```{autodoc2-docstring} evorl.networks.linear.SNMLP.layer_sizes
:parser: autodoc2_docstrings_parser
```

````

````{py:attribute} use_bias
:canonical: evorl.networks.linear.SNMLP.use_bias
:type: bool
:value: >
   True

```{autodoc2-docstring} evorl.networks.linear.SNMLP.use_bias
:parser: autodoc2_docstrings_parser
```

````

`````

````{py:function} make_discrete_q_network(action_size: int, n_stack: int = 1, hidden_layer_sizes: collections.abc.Sequence[int] = (256, 256), activation: evorl.networks.linear.ActivationFn = nn.relu, kernel_init: evorl.networks.linear.Initializer = jax.nn.initializers.lecun_uniform(), norm_layer_type: str = 'none', obs_key: str = '') -> flax.linen.Module
:canonical: evorl.networks.linear.make_discrete_q_network

```{autodoc2-docstring} evorl.networks.linear.make_discrete_q_network
:parser: autodoc2_docstrings_parser
```
````

````{py:function} make_mlp(layer_sizes: collections.abc.Sequence[int], activation: evorl.networks.linear.ActivationFn = nn.relu, kernel_init: evorl.networks.linear.Initializer = jax.nn.initializers.lecun_uniform(), activation_final: evorl.networks.linear.ActivationFn | None = None, use_bias: bool = True, norm_layer_type: str = 'none') -> flax.linen.Module
:canonical: evorl.networks.linear.make_mlp

```{autodoc2-docstring} evorl.networks.linear.make_mlp
:parser: autodoc2_docstrings_parser
```
````

````{py:function} make_policy_network(action_size: int, hidden_layer_sizes: collections.abc.Sequence[int] = (256, 256), use_bias: bool = True, activation: evorl.networks.linear.ActivationFn = nn.relu, activation_final: evorl.networks.linear.ActivationFn | None = None, norm_layer_type: str = 'none', obs_key: str = '') -> flax.linen.Module
:canonical: evorl.networks.linear.make_policy_network

```{autodoc2-docstring} evorl.networks.linear.make_policy_network
:parser: autodoc2_docstrings_parser
```
````

````{py:function} make_q_network(n_stack: int = 1, hidden_layer_sizes: collections.abc.Sequence[int] = (256, 256), activation: evorl.networks.linear.ActivationFn = nn.relu, kernel_init: evorl.networks.linear.Initializer = jax.nn.initializers.lecun_uniform(), norm_layer_type: str = 'none', obs_key: str = '') -> flax.linen.Module
:canonical: evorl.networks.linear.make_q_network

```{autodoc2-docstring} evorl.networks.linear.make_q_network
:parser: autodoc2_docstrings_parser
```
````

````{py:function} make_v_network(hidden_layer_sizes: collections.abc.Sequence[int] = (256, 256), activation: evorl.networks.linear.ActivationFn = nn.relu, kernel_init: evorl.networks.linear.Initializer = jax.nn.initializers.lecun_uniform(), norm_layer_type: str = 'none', obs_key: str = '') -> flax.linen.Module
:canonical: evorl.networks.linear.make_v_network

```{autodoc2-docstring} evorl.networks.linear.make_v_network
:parser: autodoc2_docstrings_parser
```
````

````{py:function} make_vmap_mlp(layer_sizes: collections.abc.Sequence[int], activation: evorl.networks.linear.ActivationFn = nn.relu, kernel_init: evorl.networks.linear.Initializer = jax.nn.initializers.lecun_uniform(), activation_final: evorl.networks.linear.ActivationFn | None = None, use_bias: bool = True, norm_layer_type: str = 'none', out_axes: int = -2)
:canonical: evorl.networks.linear.make_vmap_mlp

```{autodoc2-docstring} evorl.networks.linear.make_vmap_mlp
:parser: autodoc2_docstrings_parser
```
````