aws-cdk-neuronx-patterns 0.2.0 → 0.3.0

package/API.md

@@ -1434,43 +1434,342 @@ Batch terminates your jobs if they aren't finished.
 ---
 
 
-### NeuronxCompiler <a name="NeuronxCompiler" id="aws-cdk-neuronx-patterns.NeuronxCompiler"></a>
+### NeuronxCompilerBase <a name="NeuronxCompilerBase" id="aws-cdk-neuronx-patterns.NeuronxCompilerBase"></a>
+
+- *Implements:* <a href="#aws-cdk-neuronx-patterns.INeuronxCompiler">INeuronxCompiler</a>
+
+Abstract base class for Neuronx compilers.
+
+Provides the common orchestration logic (Lambda, CustomResource, WaitCondition)
+while subclasses define how to create the Batch compute environment and job definition.
+
+#### Initializers <a name="Initializers" id="aws-cdk-neuronx-patterns.NeuronxCompilerBase.Initializer"></a>
+
+```typescript
+import { NeuronxCompilerBase } from 'aws-cdk-neuronx-patterns'
+
+new NeuronxCompilerBase(scope: Construct, id: string, props: NeuronxCompilerBaseProps)
+```
+
+| **Name** | **Type** | **Description** |
+| --- | --- | --- |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBase.Initializer.parameter.scope">scope</a></code> | <code>constructs.Construct</code> | *No description.* |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBase.Initializer.parameter.id">id</a></code> | <code>string</code> | *No description.* |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBase.Initializer.parameter.props">props</a></code> | <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps">NeuronxCompilerBaseProps</a></code> | *No description.* |
+
+---
+
+##### `scope`<sup>Required</sup> <a name="scope" id="aws-cdk-neuronx-patterns.NeuronxCompilerBase.Initializer.parameter.scope"></a>
+
+- *Type:* constructs.Construct
+
+---
+
+##### `id`<sup>Required</sup> <a name="id" id="aws-cdk-neuronx-patterns.NeuronxCompilerBase.Initializer.parameter.id"></a>
+
+- *Type:* string
+
+---
+
+##### `props`<sup>Required</sup> <a name="props" id="aws-cdk-neuronx-patterns.NeuronxCompilerBase.Initializer.parameter.props"></a>
+
+- *Type:* <a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps">NeuronxCompilerBaseProps</a>
+
+---
+
+#### Methods <a name="Methods" id="Methods"></a>
+
+| **Name** | **Description** |
+| --- | --- |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBase.toString">toString</a></code> | Returns a string representation of this construct. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBase.with">with</a></code> | Applies one or more mixins to this construct. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBase.compile">compile</a></code> | *No description.* |
+
+---
+
+##### `toString` <a name="toString" id="aws-cdk-neuronx-patterns.NeuronxCompilerBase.toString"></a>
+
+```typescript
+public toString(): string
+```
+
+Returns a string representation of this construct.
+
+##### `with` <a name="with" id="aws-cdk-neuronx-patterns.NeuronxCompilerBase.with"></a>
+
+```typescript
+public with(mixins: ...IMixin[]): IConstruct
+```
+
+Applies one or more mixins to this construct.
+
+Mixins are applied in order. The list of constructs is captured at the
+start of the call, so constructs added by a mixin will not be visited.
+Use multiple `with()` calls if subsequent mixins should apply to added
+constructs.
+
+###### `mixins`<sup>Required</sup> <a name="mixins" id="aws-cdk-neuronx-patterns.NeuronxCompilerBase.with.parameter.mixins"></a>
+
+- *Type:* ...constructs.IMixin[]
+
+The mixins to apply.
+
+---
+
+##### `compile` <a name="compile" id="aws-cdk-neuronx-patterns.NeuronxCompilerBase.compile"></a>
+
+```typescript
+public compile(): NeuronxCompiledModel
+```
+
+#### Static Functions <a name="Static Functions" id="Static Functions"></a>
+
+| **Name** | **Description** |
+| --- | --- |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBase.isConstruct">isConstruct</a></code> | Checks if `x` is a construct. |
+
+---
+
+##### `isConstruct` <a name="isConstruct" id="aws-cdk-neuronx-patterns.NeuronxCompilerBase.isConstruct"></a>
+
+```typescript
+import { NeuronxCompilerBase } from 'aws-cdk-neuronx-patterns'
+
+NeuronxCompilerBase.isConstruct(x: any)
+```
+
+Checks if `x` is a construct.
+
+Use this method instead of `instanceof` to properly detect `Construct`
+instances, even when the construct library is symlinked.
+
+Explanation: in JavaScript, multiple copies of the `constructs` library on
+disk are seen as independent, completely different libraries. As a
+consequence, the class `Construct` in each copy of the `constructs` library
+is seen as a different class, and an instance of one class will not test as
+`instanceof` the other class. `npm install` will not create installations
+like this, but users may manually symlink construct libraries together or
+use a monorepo tool: in those cases, multiple copies of the `constructs`
+library can be accidentally installed, and `instanceof` will behave
+unpredictably. It is safest to avoid using `instanceof`, and using
+this type-testing method instead.
+
+###### `x`<sup>Required</sup> <a name="x" id="aws-cdk-neuronx-patterns.NeuronxCompilerBase.isConstruct.parameter.x"></a>
+
+- *Type:* any
+
+Any object.
+
+---
+
+#### Properties <a name="Properties" id="Properties"></a>
+
+| **Name** | **Type** | **Description** |
+| --- | --- | --- |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBase.property.node">node</a></code> | <code>constructs.Node</code> | The tree node. |
+
+---
+
+##### `node`<sup>Required</sup> <a name="node" id="aws-cdk-neuronx-patterns.NeuronxCompilerBase.property.node"></a>
+
+```typescript
+public readonly node: Node;
+```
+
+- *Type:* constructs.Node
+
+The tree node.
+
+---
+
+
+### NeuronxCrossCompiler <a name="NeuronxCrossCompiler" id="aws-cdk-neuronx-patterns.NeuronxCrossCompiler"></a>
+
+Neuronx cross-compiler construct.
+
+Compile the model on a non-Neuron instance and upload the artifacts to an S3 bucket.
+This avoids the need for expensive Neuron instances during the compilation phase.
+
+The compilation uses `vllm serve` which performs model tracing and neuronx-cc compilation
+entirely on CPU. The resulting artifacts are compatible with Neuron instances for inference.
+
+#### Initializers <a name="Initializers" id="aws-cdk-neuronx-patterns.NeuronxCrossCompiler.Initializer"></a>
+
+```typescript
+import { NeuronxCrossCompiler } from 'aws-cdk-neuronx-patterns'
+
+new NeuronxCrossCompiler(scope: Construct, id: string, props: NeuronxCrossCompilerProps)
+```
+
+| **Name** | **Type** | **Description** |
+| --- | --- | --- |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompiler.Initializer.parameter.scope">scope</a></code> | <code>constructs.Construct</code> | *No description.* |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompiler.Initializer.parameter.id">id</a></code> | <code>string</code> | *No description.* |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompiler.Initializer.parameter.props">props</a></code> | <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps">NeuronxCrossCompilerProps</a></code> | *No description.* |
+
+---
+
+##### `scope`<sup>Required</sup> <a name="scope" id="aws-cdk-neuronx-patterns.NeuronxCrossCompiler.Initializer.parameter.scope"></a>
+
+- *Type:* constructs.Construct
+
+---
+
+##### `id`<sup>Required</sup> <a name="id" id="aws-cdk-neuronx-patterns.NeuronxCrossCompiler.Initializer.parameter.id"></a>
+
+- *Type:* string
+
+---
+
+##### `props`<sup>Required</sup> <a name="props" id="aws-cdk-neuronx-patterns.NeuronxCrossCompiler.Initializer.parameter.props"></a>
+
+- *Type:* <a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps">NeuronxCrossCompilerProps</a>
+
+---
+
+#### Methods <a name="Methods" id="Methods"></a>
+
+| **Name** | **Description** |
+| --- | --- |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompiler.toString">toString</a></code> | Returns a string representation of this construct. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompiler.with">with</a></code> | Applies one or more mixins to this construct. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompiler.compile">compile</a></code> | *No description.* |
+
+---
+
+##### `toString` <a name="toString" id="aws-cdk-neuronx-patterns.NeuronxCrossCompiler.toString"></a>
+
+```typescript
+public toString(): string
+```
+
+Returns a string representation of this construct.
+
+##### `with` <a name="with" id="aws-cdk-neuronx-patterns.NeuronxCrossCompiler.with"></a>
+
+```typescript
+public with(mixins: ...IMixin[]): IConstruct
+```
+
+Applies one or more mixins to this construct.
+
+Mixins are applied in order. The list of constructs is captured at the
+start of the call, so constructs added by a mixin will not be visited.
+Use multiple `with()` calls if subsequent mixins should apply to added
+constructs.
+
+###### `mixins`<sup>Required</sup> <a name="mixins" id="aws-cdk-neuronx-patterns.NeuronxCrossCompiler.with.parameter.mixins"></a>
+
+- *Type:* ...constructs.IMixin[]
+
+The mixins to apply.
+
+---
+
+##### `compile` <a name="compile" id="aws-cdk-neuronx-patterns.NeuronxCrossCompiler.compile"></a>
+
+```typescript
+public compile(): NeuronxCompiledModel
+```
+
+#### Static Functions <a name="Static Functions" id="Static Functions"></a>
+
+| **Name** | **Description** |
+| --- | --- |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompiler.isConstruct">isConstruct</a></code> | Checks if `x` is a construct. |
+
+---
+
+##### `isConstruct` <a name="isConstruct" id="aws-cdk-neuronx-patterns.NeuronxCrossCompiler.isConstruct"></a>
+
+```typescript
+import { NeuronxCrossCompiler } from 'aws-cdk-neuronx-patterns'
+
+NeuronxCrossCompiler.isConstruct(x: any)
+```
+
+Checks if `x` is a construct.
+
+Use this method instead of `instanceof` to properly detect `Construct`
+instances, even when the construct library is symlinked.
+
+Explanation: in JavaScript, multiple copies of the `constructs` library on
+disk are seen as independent, completely different libraries. As a
+consequence, the class `Construct` in each copy of the `constructs` library
+is seen as a different class, and an instance of one class will not test as
+`instanceof` the other class. `npm install` will not create installations
+like this, but users may manually symlink construct libraries together or
+use a monorepo tool: in those cases, multiple copies of the `constructs`
+library can be accidentally installed, and `instanceof` will behave
+unpredictably. It is safest to avoid using `instanceof`, and using
+this type-testing method instead.
+
+###### `x`<sup>Required</sup> <a name="x" id="aws-cdk-neuronx-patterns.NeuronxCrossCompiler.isConstruct.parameter.x"></a>
+
+- *Type:* any
+
+Any object.
+
+---
+
+#### Properties <a name="Properties" id="Properties"></a>
+
+| **Name** | **Type** | **Description** |
+| --- | --- | --- |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompiler.property.node">node</a></code> | <code>constructs.Node</code> | The tree node. |
+
+---
+
+##### `node`<sup>Required</sup> <a name="node" id="aws-cdk-neuronx-patterns.NeuronxCrossCompiler.property.node"></a>
+
+```typescript
+public readonly node: Node;
+```
+
+- *Type:* constructs.Node
+
+The tree node.
+
+---
+
+
+### NeuronxNativeCompiler <a name="NeuronxNativeCompiler" id="aws-cdk-neuronx-patterns.NeuronxNativeCompiler"></a>
 
 Neuronx compiler construct.
 
 Compile the model to work with Inferentia2 and Trainium1 and upload it to an S3 bucket.
 
-#### Initializers <a name="Initializers" id="aws-cdk-neuronx-patterns.NeuronxCompiler.Initializer"></a>
+#### Initializers <a name="Initializers" id="aws-cdk-neuronx-patterns.NeuronxNativeCompiler.Initializer"></a>
 
 ```typescript
-import { NeuronxCompiler } from 'aws-cdk-neuronx-patterns'
+import { NeuronxNativeCompiler } from 'aws-cdk-neuronx-patterns'
 
-new NeuronxCompiler(scope: Construct, id: string, props: NeuronxCompilerProps)
+new NeuronxNativeCompiler(scope: Construct, id: string, props: NeuronxNativeCompilerProps)
 ```
 
 | **Name** | **Type** | **Description** |
 | --- | --- | --- |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompiler.Initializer.parameter.scope">scope</a></code> | <code>constructs.Construct</code> | *No description.* |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompiler.Initializer.parameter.id">id</a></code> | <code>string</code> | *No description.* |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompiler.Initializer.parameter.props">props</a></code> | <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerProps">NeuronxCompilerProps</a></code> | *No description.* |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompiler.Initializer.parameter.scope">scope</a></code> | <code>constructs.Construct</code> | *No description.* |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompiler.Initializer.parameter.id">id</a></code> | <code>string</code> | *No description.* |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompiler.Initializer.parameter.props">props</a></code> | <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps">NeuronxNativeCompilerProps</a></code> | *No description.* |
 
 ---
 
-##### `scope`<sup>Required</sup> <a name="scope" id="aws-cdk-neuronx-patterns.NeuronxCompiler.Initializer.parameter.scope"></a>
+##### `scope`<sup>Required</sup> <a name="scope" id="aws-cdk-neuronx-patterns.NeuronxNativeCompiler.Initializer.parameter.scope"></a>
 
 - *Type:* constructs.Construct
 
 ---
 
-##### `id`<sup>Required</sup> <a name="id" id="aws-cdk-neuronx-patterns.NeuronxCompiler.Initializer.parameter.id"></a>
+##### `id`<sup>Required</sup> <a name="id" id="aws-cdk-neuronx-patterns.NeuronxNativeCompiler.Initializer.parameter.id"></a>
 
 - *Type:* string
 
 ---
 
-##### `props`<sup>Required</sup> <a name="props" id="aws-cdk-neuronx-patterns.NeuronxCompiler.Initializer.parameter.props"></a>
+##### `props`<sup>Required</sup> <a name="props" id="aws-cdk-neuronx-patterns.NeuronxNativeCompiler.Initializer.parameter.props"></a>
 
-- *Type:* <a href="#aws-cdk-neuronx-patterns.NeuronxCompilerProps">NeuronxCompilerProps</a>
+- *Type:* <a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps">NeuronxNativeCompilerProps</a>
 
 ---
 
@@ -1478,13 +1777,13 @@ new NeuronxCompiler(scope: Construct, id: string, props: NeuronxCompilerProps)
 
 | **Name** | **Description** |
 | --- | --- |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompiler.toString">toString</a></code> | Returns a string representation of this construct. |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompiler.with">with</a></code> | Applies one or more mixins to this construct. |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompiler.compile">compile</a></code> | *No description.* |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompiler.toString">toString</a></code> | Returns a string representation of this construct. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompiler.with">with</a></code> | Applies one or more mixins to this construct. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompiler.compile">compile</a></code> | *No description.* |
 
 ---
 
-##### `toString` <a name="toString" id="aws-cdk-neuronx-patterns.NeuronxCompiler.toString"></a>
+##### `toString` <a name="toString" id="aws-cdk-neuronx-patterns.NeuronxNativeCompiler.toString"></a>
 
 ```typescript
 public toString(): string
@@ -1492,7 +1791,7 @@ public toString(): string
 
 Returns a string representation of this construct.
 
-##### `with` <a name="with" id="aws-cdk-neuronx-patterns.NeuronxCompiler.with"></a>
+##### `with` <a name="with" id="aws-cdk-neuronx-patterns.NeuronxNativeCompiler.with"></a>
 
 ```typescript
 public with(mixins: ...IMixin[]): IConstruct
@@ -1505,7 +1804,7 @@ start of the call, so constructs added by a mixin will not be visited.
 Use multiple `with()` calls if subsequent mixins should apply to added
 constructs.
 
-###### `mixins`<sup>Required</sup> <a name="mixins" id="aws-cdk-neuronx-patterns.NeuronxCompiler.with.parameter.mixins"></a>
+###### `mixins`<sup>Required</sup> <a name="mixins" id="aws-cdk-neuronx-patterns.NeuronxNativeCompiler.with.parameter.mixins"></a>
 
 - *Type:* ...constructs.IMixin[]
 
@@ -1513,7 +1812,7 @@ The mixins to apply.
 
 ---
 
-##### `compile` <a name="compile" id="aws-cdk-neuronx-patterns.NeuronxCompiler.compile"></a>
+##### `compile` <a name="compile" id="aws-cdk-neuronx-patterns.NeuronxNativeCompiler.compile"></a>
 
 ```typescript
 public compile(): NeuronxCompiledModel
@@ -1523,16 +1822,16 @@ public compile(): NeuronxCompiledModel
 
 | **Name** | **Description** |
 | --- | --- |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompiler.isConstruct">isConstruct</a></code> | Checks if `x` is a construct. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompiler.isConstruct">isConstruct</a></code> | Checks if `x` is a construct. |
 
 ---
 
-##### `isConstruct` <a name="isConstruct" id="aws-cdk-neuronx-patterns.NeuronxCompiler.isConstruct"></a>
+##### `isConstruct` <a name="isConstruct" id="aws-cdk-neuronx-patterns.NeuronxNativeCompiler.isConstruct"></a>
 
 ```typescript
-import { NeuronxCompiler } from 'aws-cdk-neuronx-patterns'
+import { NeuronxNativeCompiler } from 'aws-cdk-neuronx-patterns'
 
-NeuronxCompiler.isConstruct(x: any)
+NeuronxNativeCompiler.isConstruct(x: any)
 ```
 
 Checks if `x` is a construct.
@@ -1551,7 +1850,7 @@ library can be accidentally installed, and `instanceof` will behave
 unpredictably. It is safest to avoid using `instanceof`, and using
 this type-testing method instead.
 
-###### `x`<sup>Required</sup> <a name="x" id="aws-cdk-neuronx-patterns.NeuronxCompiler.isConstruct.parameter.x"></a>
+###### `x`<sup>Required</sup> <a name="x" id="aws-cdk-neuronx-patterns.NeuronxNativeCompiler.isConstruct.parameter.x"></a>
 
 - *Type:* any
 
@@ -1563,11 +1862,11 @@ Any object.
 
 | **Name** | **Type** | **Description** |
 | --- | --- | --- |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompiler.property.node">node</a></code> | <code>constructs.Node</code> | The tree node. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompiler.property.node">node</a></code> | <code>constructs.Node</code> | The tree node. |
 
 ---
 
-##### `node`<sup>Required</sup> <a name="node" id="aws-cdk-neuronx-patterns.NeuronxCompiler.property.node"></a>
+##### `node`<sup>Required</sup> <a name="node" id="aws-cdk-neuronx-patterns.NeuronxNativeCompiler.property.node"></a>
 
 ```typescript
 public readonly node: Node;
@@ -4652,43 +4951,88 @@ The task definition to use for tasks in the service. TaskDefinition or TaskImage
 
 ---
 
-### ModelConfig <a name="ModelConfig" id="aws-cdk-neuronx-patterns.ModelConfig"></a>
+### ComputeEnvironmentResult <a name="ComputeEnvironmentResult" id="aws-cdk-neuronx-patterns.ComputeEnvironmentResult"></a>
 
-#### Initializer <a name="Initializer" id="aws-cdk-neuronx-patterns.ModelConfig.Initializer"></a>
+Result of creating a compute environment.
+
+#### Initializer <a name="Initializer" id="aws-cdk-neuronx-patterns.ComputeEnvironmentResult.Initializer"></a>
 
 ```typescript
-import { ModelConfig } from 'aws-cdk-neuronx-patterns'
+import { ComputeEnvironmentResult } from 'aws-cdk-neuronx-patterns'
 
-const modelConfig: ModelConfig = { ... }
+const computeEnvironmentResult: ComputeEnvironmentResult = { ... }
 ```
 
 #### Properties <a name="Properties" id="Properties"></a>
 
 | **Name** | **Type** | **Description** |
 | --- | --- | --- |
-| <code><a href="#aws-cdk-neuronx-patterns.ModelConfig.property.attentionHeads">attentionHeads</a></code> | <code>number</code> | *No description.* |
-| <code><a href="#aws-cdk-neuronx-patterns.ModelConfig.property.embeddingDimension">embeddingDimension</a></code> | <code>number</code> | *No description.* |
-| <code><a href="#aws-cdk-neuronx-patterns.ModelConfig.property.layers">layers</a></code> | <code>number</code> | *No description.* |
+| <code><a href="#aws-cdk-neuronx-patterns.ComputeEnvironmentResult.property.computeEnvironment">computeEnvironment</a></code> | <code>aws-cdk-lib.aws_batch.IComputeEnvironment</code> | The compute environment. |
+| <code><a href="#aws-cdk-neuronx-patterns.ComputeEnvironmentResult.property.instanceRole">instanceRole</a></code> | <code>aws-cdk-lib.aws_iam.IRole</code> | The instance role associated with the compute environment. |
 
 ---
 
-##### `attentionHeads`<sup>Required</sup> <a name="attentionHeads" id="aws-cdk-neuronx-patterns.ModelConfig.property.attentionHeads"></a>
+##### `computeEnvironment`<sup>Required</sup> <a name="computeEnvironment" id="aws-cdk-neuronx-patterns.ComputeEnvironmentResult.property.computeEnvironment"></a>
 
 ```typescript
-public readonly attentionHeads: number;
+public readonly computeEnvironment: IComputeEnvironment;
 ```
 
-- *Type:* number
+- *Type:* aws-cdk-lib.aws_batch.IComputeEnvironment
+
+The compute environment.
 
 ---
 
-##### `embeddingDimension`<sup>Required</sup> <a name="embeddingDimension" id="aws-cdk-neuronx-patterns.ModelConfig.property.embeddingDimension"></a>
+##### `instanceRole`<sup>Required</sup> <a name="instanceRole" id="aws-cdk-neuronx-patterns.ComputeEnvironmentResult.property.instanceRole"></a>
 
 ```typescript
-public readonly embeddingDimension: number;
+public readonly instanceRole: IRole;
 ```
 
-- *Type:* number
+- *Type:* aws-cdk-lib.aws_iam.IRole
+
+The instance role associated with the compute environment.
+
+---
+
+### ModelConfig <a name="ModelConfig" id="aws-cdk-neuronx-patterns.ModelConfig"></a>
+
+#### Initializer <a name="Initializer" id="aws-cdk-neuronx-patterns.ModelConfig.Initializer"></a>
+
+```typescript
+import { ModelConfig } from 'aws-cdk-neuronx-patterns'
+
+const modelConfig: ModelConfig = { ... }
+```
+
+#### Properties <a name="Properties" id="Properties"></a>
+
+| **Name** | **Type** | **Description** |
+| --- | --- | --- |
+| <code><a href="#aws-cdk-neuronx-patterns.ModelConfig.property.attentionHeads">attentionHeads</a></code> | <code>number</code> | *No description.* |
+| <code><a href="#aws-cdk-neuronx-patterns.ModelConfig.property.embeddingDimension">embeddingDimension</a></code> | <code>number</code> | *No description.* |
+| <code><a href="#aws-cdk-neuronx-patterns.ModelConfig.property.layers">layers</a></code> | <code>number</code> | *No description.* |
+
+---
+
+##### `attentionHeads`<sup>Required</sup> <a name="attentionHeads" id="aws-cdk-neuronx-patterns.ModelConfig.property.attentionHeads"></a>
+
+```typescript
+public readonly attentionHeads: number;
+```
+
+- *Type:* number
+
+---
+
+##### `embeddingDimension`<sup>Required</sup> <a name="embeddingDimension" id="aws-cdk-neuronx-patterns.ModelConfig.property.embeddingDimension"></a>
+
+```typescript
+public readonly embeddingDimension: number;
+```
+
+- *Type:* number
 
 ---
 
@@ -5846,234 +6190,627 @@ Automatically added to the job definition.
 ##### `gpu`<sup>Optional</sup> <a name="gpu" id="aws-cdk-neuronx-patterns.NeuronxBatchProps.property.gpu"></a>
 
 ```typescript
-public readonly gpu: number;
+public readonly gpu: number;
+```
+
+- *Type:* number
+- *Default:* no gpus
+
+The number of physical GPUs to reserve for the container.
+
+Make sure that the number of GPUs reserved for all containers in a job doesn't exceed
+the number of available GPUs on the compute resource that the job is launched on.
+
+---
+
+##### `privileged`<sup>Optional</sup> <a name="privileged" id="aws-cdk-neuronx-patterns.NeuronxBatchProps.property.privileged"></a>
+
+```typescript
+public readonly privileged: boolean;
+```
+
+- *Type:* boolean
+- *Default:* false
+
+When this parameter is true, the container is given elevated permissions on the host container instance (similar to the root user).
+
+---
+
+##### `ulimits`<sup>Optional</sup> <a name="ulimits" id="aws-cdk-neuronx-patterns.NeuronxBatchProps.property.ulimits"></a>
+
+```typescript
+public readonly ulimits: Ulimit[];
+```
+
+- *Type:* aws-cdk-lib.aws_batch.Ulimit[]
+- *Default:* no ulimits
+
+Limits to set for the user this docker container will run as.
+
+---
+
+##### `neuronxInstanceType`<sup>Required</sup> <a name="neuronxInstanceType" id="aws-cdk-neuronx-patterns.NeuronxBatchProps.property.neuronxInstanceType"></a>
+
+```typescript
+public readonly neuronxInstanceType: INeuronxInstanceType;
+```
+
+- *Type:* <a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a>
+
+The instance type of worker instance.
+
+---
+
+##### `volumeSize`<sup>Required</sup> <a name="volumeSize" id="aws-cdk-neuronx-patterns.NeuronxBatchProps.property.volumeSize"></a>
+
+```typescript
+public readonly volumeSize: Size;
+```
+
+- *Type:* aws-cdk-lib.Size
+- *Default:* N bilion parameters * 5GiB EBS
+
+The root volume of worker instance.
+
+---
+
+##### `vpc`<sup>Required</sup> <a name="vpc" id="aws-cdk-neuronx-patterns.NeuronxBatchProps.property.vpc"></a>
+
+```typescript
+public readonly vpc: IVpc;
+```
+
+- *Type:* aws-cdk-lib.aws_ec2.IVpc
+
+VPC in which this will launch worker instance.
+
+---
+
+##### `spot`<sup>Optional</sup> <a name="spot" id="aws-cdk-neuronx-patterns.NeuronxBatchProps.property.spot"></a>
+
+```typescript
+public readonly spot: boolean;
+```
+
+- *Type:* boolean
+- *Default:* false
+
+Whether or not to use spot instances.
+
+Spot instances are less expensive EC2 instances that can be reclaimed by EC2 at any time;
+your job will be given two minutes of notice before reclamation.
+
+---
+
+##### `vpcSubnets`<sup>Optional</sup> <a name="vpcSubnets" id="aws-cdk-neuronx-patterns.NeuronxBatchProps.property.vpcSubnets"></a>
+
+```typescript
+public readonly vpcSubnets: SubnetSelection;
+```
+
+- *Type:* aws-cdk-lib.aws_ec2.SubnetSelection
+- *Default:* new subnets will be created
+
+The VPC Subnets this Compute Environment will launch instances in.
+
+---
+
+### NeuronxCompiledModel <a name="NeuronxCompiledModel" id="aws-cdk-neuronx-patterns.NeuronxCompiledModel"></a>
+
+The model compiled by Neuronx compiler.
+
+#### Initializer <a name="Initializer" id="aws-cdk-neuronx-patterns.NeuronxCompiledModel.Initializer"></a>
+
+```typescript
+import { NeuronxCompiledModel } from 'aws-cdk-neuronx-patterns'
+
+const neuronxCompiledModel: NeuronxCompiledModel = { ... }
+```
+
+#### Properties <a name="Properties" id="Properties"></a>
+
+| **Name** | **Type** | **Description** |
+| --- | --- | --- |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.bucket">bucket</a></code> | <code>aws-cdk-lib.aws_s3.IBucket</code> | The bucket to upload compiled artifacts. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.modelName">modelName</a></code> | <code>string</code> | The model name. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.recommendedInstanceType">recommendedInstanceType</a></code> | <code><a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a></code> | The recommended Neuron instance type for running inference with this compiled model. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.s3Prefix">s3Prefix</a></code> | <code>string</code> | S3 prefix that compiled artifact uploaded. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.s3Uri">s3Uri</a></code> | <code>string</code> | S3 URL that compiled artifact uploaded. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.weightSize">weightSize</a></code> | <code>aws-cdk-lib.Size</code> | The weight size of the model. |
+
+---
+
+##### `bucket`<sup>Required</sup> <a name="bucket" id="aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.bucket"></a>
+
+```typescript
+public readonly bucket: IBucket;
+```
+
+- *Type:* aws-cdk-lib.aws_s3.IBucket
+
+The bucket to upload compiled artifacts.
+
+---
+
+##### `modelName`<sup>Required</sup> <a name="modelName" id="aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.modelName"></a>
+
+```typescript
+public readonly modelName: string;
+```
+
+- *Type:* string
+
+The model name.
+
+---
+
+##### `recommendedInstanceType`<sup>Required</sup> <a name="recommendedInstanceType" id="aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.recommendedInstanceType"></a>
+
+```typescript
+public readonly recommendedInstanceType: INeuronxInstanceType;
+```
+
+- *Type:* <a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a>
+
+The recommended Neuron instance type for running inference with this compiled model.
+
+---
+
+##### `s3Prefix`<sup>Required</sup> <a name="s3Prefix" id="aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.s3Prefix"></a>
+
+```typescript
+public readonly s3Prefix: string;
+```
+
+- *Type:* string
+
+S3 prefix that compiled artifact uploaded.
+
+---
+
+##### `s3Uri`<sup>Required</sup> <a name="s3Uri" id="aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.s3Uri"></a>
+
+```typescript
+public readonly s3Uri: string;
+```
+
+- *Type:* string
+
+S3 URL that compiled artifact uploaded.
+
+---
+
+##### `weightSize`<sup>Required</sup> <a name="weightSize" id="aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.weightSize"></a>
+
+```typescript
+public readonly weightSize: Size;
+```
+
+- *Type:* aws-cdk-lib.Size
+
+The weight size of the model.
+
+---
+
+### NeuronxCompilerBaseProps <a name="NeuronxCompilerBaseProps" id="aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps"></a>
+
+Common props for NeuronxCompilerBase.
+
+#### Initializer <a name="Initializer" id="aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.Initializer"></a>
+
+```typescript
+import { NeuronxCompilerBaseProps } from 'aws-cdk-neuronx-patterns'
+
+const neuronxCompilerBaseProps: NeuronxCompilerBaseProps = { ... }
+```
+
+#### Properties <a name="Properties" id="Properties"></a>
+
+| **Name** | **Type** | **Description** |
+| --- | --- | --- |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.artifactS3Prefix">artifactS3Prefix</a></code> | <code>string</code> | S3 Prefix that compiled artifact uploaded. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.bucket">bucket</a></code> | <code>aws-cdk-lib.aws_s3.IBucket</code> | The bucket to upload compiled artifacts. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.image">image</a></code> | <code><a href="#aws-cdk-neuronx-patterns.INeuronxContainerImage">INeuronxContainerImage</a></code> | An image of the container where the compile job is executed. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.model">model</a></code> | <code><a href="#aws-cdk-neuronx-patterns.Model">Model</a></code> | The model to be compiled. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.neuronxInstanceType">neuronxInstanceType</a></code> | <code><a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a></code> | The instance type of compile worker instance. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.vpc">vpc</a></code> | <code>aws-cdk-lib.aws_ec2.IVpc</code> | VPC in which this will launch compile worker instance. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.command">command</a></code> | <code>string[]</code> | The command to run in the container. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.environment">environment</a></code> | <code>{[ key: string ]: string}</code> | The environment variables to pass to the container. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.secrets">secrets</a></code> | <code>{[ key: string ]: aws-cdk-lib.aws_batch.Secret}</code> | Secrets to pass to the container. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.spot">spot</a></code> | <code>boolean</code> | Whether or not to use spot instances. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.volumeSize">volumeSize</a></code> | <code>aws-cdk-lib.Size</code> | The root volume of worker instance. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.vpcSubnets">vpcSubnets</a></code> | <code>aws-cdk-lib.aws_ec2.SubnetSelection</code> | The VPC Subnets this Compute Environment will launch instances in. |
+
+---
+
+##### `artifactS3Prefix`<sup>Required</sup> <a name="artifactS3Prefix" id="aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.artifactS3Prefix"></a>
+
+```typescript
+public readonly artifactS3Prefix: string;
+```
+
+- *Type:* string
+
+S3 Prefix that compiled artifact uploaded.
+
+This property is not depends on compile job finish.
+
+---
+
+##### `bucket`<sup>Required</sup> <a name="bucket" id="aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.bucket"></a>
+
+```typescript
+public readonly bucket: IBucket;
+```
+
+- *Type:* aws-cdk-lib.aws_s3.IBucket
+
+The bucket to upload compiled artifacts.
+
+---
+
+##### `image`<sup>Required</sup> <a name="image" id="aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.image"></a>
+
+```typescript
+public readonly image: INeuronxContainerImage;
+```
+
+- *Type:* <a href="#aws-cdk-neuronx-patterns.INeuronxContainerImage">INeuronxContainerImage</a>
+
+An image of the container where the compile job is executed.
+
+---
+
+##### `model`<sup>Required</sup> <a name="model" id="aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.model"></a>
+
+```typescript
+public readonly model: Model;
+```
+
+- *Type:* <a href="#aws-cdk-neuronx-patterns.Model">Model</a>
+
+The model to be compiled.
+
+---
+
+##### `neuronxInstanceType`<sup>Required</sup> <a name="neuronxInstanceType" id="aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.neuronxInstanceType"></a>
+
+```typescript
+public readonly neuronxInstanceType: INeuronxInstanceType;
+```
+
+- *Type:* <a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a>
+
+The instance type of compile worker instance.
+
+---
+
+##### `vpc`<sup>Required</sup> <a name="vpc" id="aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.vpc"></a>
+
+```typescript
+public readonly vpc: IVpc;
+```
+
+- *Type:* aws-cdk-lib.aws_ec2.IVpc
+
+VPC in which this will launch compile worker instance.
+
+---
+
+##### `command`<sup>Optional</sup> <a name="command" id="aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.command"></a>
+
+```typescript
+public readonly command: string[];
+```
+
+- *Type:* string[]
+
+The command to run in the container.
+
+---
+
+##### `environment`<sup>Optional</sup> <a name="environment" id="aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.environment"></a>
+
+```typescript
+public readonly environment: {[ key: string ]: string};
+```
+
+- *Type:* {[ key: string ]: string}
+- *Default:* No environment variables.
+
+The environment variables to pass to the container.
+
+This is only applicable when using container runtime.
+
+---
+
+##### `secrets`<sup>Optional</sup> <a name="secrets" id="aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.secrets"></a>
+
+```typescript
+public readonly secrets: {[ key: string ]: Secret};
+```
+
+- *Type:* {[ key: string ]: aws-cdk-lib.aws_batch.Secret}
+
+Secrets to pass to the container.
+
+---
+
+##### `spot`<sup>Optional</sup> <a name="spot" id="aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.spot"></a>
+
+```typescript
+public readonly spot: boolean;
+```
+
+- *Type:* boolean
+- *Default:* false
+
+Whether or not to use spot instances.
+
+Spot instances are less expensive EC2 instances that can be reclaimed by EC2 at any time; your job will be given two minutes of notice before reclamation.
+
+---
+
+##### `volumeSize`<sup>Optional</sup> <a name="volumeSize" id="aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.volumeSize"></a>
+
+```typescript
+public readonly volumeSize: Size;
+```
+
+- *Type:* aws-cdk-lib.Size
+- *Default:* N billion parameters * 5GiB EBS
+
+The root volume of worker instance.
+
+---
+
+##### `vpcSubnets`<sup>Optional</sup> <a name="vpcSubnets" id="aws-cdk-neuronx-patterns.NeuronxCompilerBaseProps.property.vpcSubnets"></a>
+
+```typescript
+public readonly vpcSubnets: SubnetSelection;
 ```
 
-- *Type:* number
-- *Default:* no gpus
-
-The number of physical GPUs to reserve for the container.
+- *Type:* aws-cdk-lib.aws_ec2.SubnetSelection
+- *Default:* new subnets will be created
 
-Make sure that the number of GPUs reserved for all containers in a job doesn't exceed
-the number of available GPUs on the compute resource that the job is launched on.
+The VPC Subnets this Compute Environment will launch instances in.
 
 ---
 
-##### `privileged`<sup>Optional</sup> <a name="privileged" id="aws-cdk-neuronx-patterns.NeuronxBatchProps.property.privileged"></a>
+### NeuronxCrossCompilerProps <a name="NeuronxCrossCompilerProps" id="aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps"></a>
+
+Props of NeuronxCrossCompiler.
+
+#### Initializer <a name="Initializer" id="aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.Initializer"></a>
 
 ```typescript
-public readonly privileged: boolean;
+import { NeuronxCrossCompilerProps } from 'aws-cdk-neuronx-patterns'
+
+const neuronxCrossCompilerProps: NeuronxCrossCompilerProps = { ... }
 ```
 
-- *Type:* boolean
-- *Default:* false
+#### Properties <a name="Properties" id="Properties"></a>
 
-When this parameter is true, the container is given elevated permissions on the host container instance (similar to the root user).
+| **Name** | **Type** | **Description** |
+| --- | --- | --- |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.artifactS3Prefix">artifactS3Prefix</a></code> | <code>string</code> | S3 Prefix that compiled artifact uploaded. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.bucket">bucket</a></code> | <code>aws-cdk-lib.aws_s3.IBucket</code> | The bucket to upload compiled artifacts. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.image">image</a></code> | <code><a href="#aws-cdk-neuronx-patterns.INeuronxContainerImage">INeuronxContainerImage</a></code> | An image of the container where the compile job is executed. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.model">model</a></code> | <code><a href="#aws-cdk-neuronx-patterns.Model">Model</a></code> | The model to be compiled. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.neuronxInstanceType">neuronxInstanceType</a></code> | <code><a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a></code> | The instance type of compile worker instance. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.vpc">vpc</a></code> | <code>aws-cdk-lib.aws_ec2.IVpc</code> | VPC in which this will launch compile worker instance. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.command">command</a></code> | <code>string[]</code> | The command to run in the container. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.environment">environment</a></code> | <code>{[ key: string ]: string}</code> | The environment variables to pass to the container. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.secrets">secrets</a></code> | <code>{[ key: string ]: aws-cdk-lib.aws_batch.Secret}</code> | Secrets to pass to the container. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.spot">spot</a></code> | <code>boolean</code> | Whether or not to use spot instances. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.volumeSize">volumeSize</a></code> | <code>aws-cdk-lib.Size</code> | The root volume of worker instance. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.vpcSubnets">vpcSubnets</a></code> | <code>aws-cdk-lib.aws_ec2.SubnetSelection</code> | The VPC Subnets this Compute Environment will launch instances in. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.compileInstanceType">compileInstanceType</a></code> | <code>aws-cdk-lib.aws_ec2.InstanceType</code> | The EC2 instance type to use for cross-compilation. |
 
 ---
 
-##### `ulimits`<sup>Optional</sup> <a name="ulimits" id="aws-cdk-neuronx-patterns.NeuronxBatchProps.property.ulimits"></a>
+##### `artifactS3Prefix`<sup>Required</sup> <a name="artifactS3Prefix" id="aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.artifactS3Prefix"></a>
 
 ```typescript
-public readonly ulimits: Ulimit[];
+public readonly artifactS3Prefix: string;
 ```
 
-- *Type:* aws-cdk-lib.aws_batch.Ulimit[]
-- *Default:* no ulimits
+- *Type:* string
 
-Limits to set for the user this docker container will run as.
+S3 Prefix that compiled artifact uploaded.
+
+This property is not depends on compile job finish.
 
 ---
 
-##### `neuronxInstanceType`<sup>Required</sup> <a name="neuronxInstanceType" id="aws-cdk-neuronx-patterns.NeuronxBatchProps.property.neuronxInstanceType"></a>
+##### `bucket`<sup>Required</sup> <a name="bucket" id="aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.bucket"></a>
 
 ```typescript
-public readonly neuronxInstanceType: INeuronxInstanceType;
+public readonly bucket: IBucket;
 ```
 
-- *Type:* <a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a>
+- *Type:* aws-cdk-lib.aws_s3.IBucket
 
-The instance type of worker instance.
+The bucket to upload compiled artifacts.
 
 ---
 
-##### `volumeSize`<sup>Required</sup> <a name="volumeSize" id="aws-cdk-neuronx-patterns.NeuronxBatchProps.property.volumeSize"></a>
+##### `image`<sup>Required</sup> <a name="image" id="aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.image"></a>
 
 ```typescript
-public readonly volumeSize: Size;
+public readonly image: INeuronxContainerImage;
 ```
 
-- *Type:* aws-cdk-lib.Size
-- *Default:* N bilion parameters * 5GiB EBS
+- *Type:* <a href="#aws-cdk-neuronx-patterns.INeuronxContainerImage">INeuronxContainerImage</a>
 
-The root volume of worker instance.
+An image of the container where the compile job is executed.
 
 ---
 
-##### `vpc`<sup>Required</sup> <a name="vpc" id="aws-cdk-neuronx-patterns.NeuronxBatchProps.property.vpc"></a>
+##### `model`<sup>Required</sup> <a name="model" id="aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.model"></a>
 
 ```typescript
-public readonly vpc: IVpc;
+public readonly model: Model;
 ```
 
-- *Type:* aws-cdk-lib.aws_ec2.IVpc
+- *Type:* <a href="#aws-cdk-neuronx-patterns.Model">Model</a>
 
-VPC in which this will launch worker instance.
+The model to be compiled.
 
 ---
 
-##### `spot`<sup>Optional</sup> <a name="spot" id="aws-cdk-neuronx-patterns.NeuronxBatchProps.property.spot"></a>
+##### `neuronxInstanceType`<sup>Required</sup> <a name="neuronxInstanceType" id="aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.neuronxInstanceType"></a>
 
 ```typescript
-public readonly spot: boolean;
+public readonly neuronxInstanceType: INeuronxInstanceType;
 ```
 
-- *Type:* boolean
-- *Default:* false
-
-Whether or not to use spot instances.
+- *Type:* <a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a>
 
-Spot instances are less expensive EC2 instances that can be reclaimed by EC2 at any time;
-your job will be given two minutes of notice before reclamation.
+The instance type of compile worker instance.
 
 ---
 
-##### `vpcSubnets`<sup>Optional</sup> <a name="vpcSubnets" id="aws-cdk-neuronx-patterns.NeuronxBatchProps.property.vpcSubnets"></a>
+##### `vpc`<sup>Required</sup> <a name="vpc" id="aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.vpc"></a>
 
 ```typescript
-public readonly vpcSubnets: SubnetSelection;
+public readonly vpc: IVpc;
 ```
 
-- *Type:* aws-cdk-lib.aws_ec2.SubnetSelection
-- *Default:* new subnets will be created
+- *Type:* aws-cdk-lib.aws_ec2.IVpc
 
-The VPC Subnets this Compute Environment will launch instances in.
+VPC in which this will launch compile worker instance.
 
 ---
 
-### NeuronxCompiledModel <a name="NeuronxCompiledModel" id="aws-cdk-neuronx-patterns.NeuronxCompiledModel"></a>
-
-#### Initializer <a name="Initializer" id="aws-cdk-neuronx-patterns.NeuronxCompiledModel.Initializer"></a>
+##### `command`<sup>Optional</sup> <a name="command" id="aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.command"></a>
 
 ```typescript
-import { NeuronxCompiledModel } from 'aws-cdk-neuronx-patterns'
-
-const neuronxCompiledModel: NeuronxCompiledModel = { ... }
+public readonly command: string[];
 ```
 
-#### Properties <a name="Properties" id="Properties"></a>
+- *Type:* string[]
 
-| **Name** | **Type** | **Description** |
-| --- | --- | --- |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.bucket">bucket</a></code> | <code>aws-cdk-lib.aws_s3.IBucket</code> | The bucket to upload compiled artifacts. |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.compileTimeInstanceType">compileTimeInstanceType</a></code> | <code><a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a></code> | *No description.* |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.modelName">modelName</a></code> | <code>string</code> | The model name. |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.s3Prefix">s3Prefix</a></code> | <code>string</code> | S3 prefix that compiled artifact uploaded. |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.s3Uri">s3Uri</a></code> | <code>string</code> | S3 URL that compiled artifact uploaded. |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.weightSize">weightSize</a></code> | <code>aws-cdk-lib.Size</code> | *No description.* |
+The command to run in the container.
 
 ---
 
-##### `bucket`<sup>Required</sup> <a name="bucket" id="aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.bucket"></a>
+##### `environment`<sup>Optional</sup> <a name="environment" id="aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.environment"></a>
 
 ```typescript
-public readonly bucket: IBucket;
+public readonly environment: {[ key: string ]: string};
 ```
 
-- *Type:* aws-cdk-lib.aws_s3.IBucket
+- *Type:* {[ key: string ]: string}
+- *Default:* No environment variables.
 
-The bucket to upload compiled artifacts.
+The environment variables to pass to the container.
+
+This is only applicable when using container runtime.
 
 ---
 
-##### `compileTimeInstanceType`<sup>Required</sup> <a name="compileTimeInstanceType" id="aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.compileTimeInstanceType"></a>
+##### `secrets`<sup>Optional</sup> <a name="secrets" id="aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.secrets"></a>
 
 ```typescript
-public readonly compileTimeInstanceType: INeuronxInstanceType;
+public readonly secrets: {[ key: string ]: Secret};
 ```
 
-- *Type:* <a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a>
+- *Type:* {[ key: string ]: aws-cdk-lib.aws_batch.Secret}
+
+Secrets to pass to the container.
 
 ---
 
-##### `modelName`<sup>Required</sup> <a name="modelName" id="aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.modelName"></a>
+##### `spot`<sup>Optional</sup> <a name="spot" id="aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.spot"></a>
 
 ```typescript
-public readonly modelName: string;
+public readonly spot: boolean;
 ```
 
-- *Type:* string
+- *Type:* boolean
+- *Default:* false
 
-The model name.
+Whether or not to use spot instances.
+
+Spot instances are less expensive EC2 instances that can be reclaimed by EC2 at any time; your job will be given two minutes of notice before reclamation.
 
 ---
 
-##### `s3Prefix`<sup>Required</sup> <a name="s3Prefix" id="aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.s3Prefix"></a>
+##### `volumeSize`<sup>Optional</sup> <a name="volumeSize" id="aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.volumeSize"></a>
 
 ```typescript
-public readonly s3Prefix: string;
+public readonly volumeSize: Size;
 ```
 
-- *Type:* string
+- *Type:* aws-cdk-lib.Size
+- *Default:* N billion parameters * 5GiB EBS
 
-S3 prefix that compiled artifact uploaded.
+The root volume of worker instance.
 
 ---
 
-##### `s3Uri`<sup>Required</sup> <a name="s3Uri" id="aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.s3Uri"></a>
+##### `vpcSubnets`<sup>Optional</sup> <a name="vpcSubnets" id="aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.vpcSubnets"></a>
 
 ```typescript
-public readonly s3Uri: string;
+public readonly vpcSubnets: SubnetSelection;
 ```
 
-- *Type:* string
+- *Type:* aws-cdk-lib.aws_ec2.SubnetSelection
+- *Default:* new subnets will be created
 
-S3 URL that compiled artifact uploaded.
+The VPC Subnets this Compute Environment will launch instances in.
 
 ---
 
-##### `weightSize`<sup>Required</sup> <a name="weightSize" id="aws-cdk-neuronx-patterns.NeuronxCompiledModel.property.weightSize"></a>
+##### `compileInstanceType`<sup>Optional</sup> <a name="compileInstanceType" id="aws-cdk-neuronx-patterns.NeuronxCrossCompilerProps.property.compileInstanceType"></a>
 
 ```typescript
-public readonly weightSize: Size;
+public readonly compileInstanceType: InstanceType;
 ```
 
-- *Type:* aws-cdk-lib.Size
+- *Type:* aws-cdk-lib.aws_ec2.InstanceType
+- *Default:* ec2.InstanceType.of(ec2.InstanceClass.C7I, ec2.InstanceSize.XLARGE4)
+
+The EC2 instance type to use for cross-compilation.
+
+This should be a non-Neuron instance type with sufficient memory and CPU
+for model compilation.
 
 ---
 
-### NeuronxCompilerProps <a name="NeuronxCompilerProps" id="aws-cdk-neuronx-patterns.NeuronxCompilerProps"></a>
+### NeuronxNativeCompilerProps <a name="NeuronxNativeCompilerProps" id="aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps"></a>
 
-Props of NeuronxCompiler.
+Props of NeuronxNativeCompiler.
 
-#### Initializer <a name="Initializer" id="aws-cdk-neuronx-patterns.NeuronxCompilerProps.Initializer"></a>
+#### Initializer <a name="Initializer" id="aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.Initializer"></a>
 
 ```typescript
-import { NeuronxCompilerProps } from 'aws-cdk-neuronx-patterns'
+import { NeuronxNativeCompilerProps } from 'aws-cdk-neuronx-patterns'
 
-const neuronxCompilerProps: NeuronxCompilerProps = { ... }
+const neuronxNativeCompilerProps: NeuronxNativeCompilerProps = { ... }
 ```
 
 #### Properties <a name="Properties" id="Properties"></a>
 
 | **Name** | **Type** | **Description** |
 | --- | --- | --- |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.artifactS3Prefix">artifactS3Prefix</a></code> | <code>string</code> | S3 Prefix that compiled artifact uploaded. |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.bucket">bucket</a></code> | <code>aws-cdk-lib.aws_s3.IBucket</code> | The bucket to upload compiled artifacts. |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.image">image</a></code> | <code><a href="#aws-cdk-neuronx-patterns.INeuronxContainerImage">INeuronxContainerImage</a></code> | An image of the container where the compile job is executed. |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.model">model</a></code> | <code><a href="#aws-cdk-neuronx-patterns.Model">Model</a></code> | The model to be compiled. |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.neuronxInstanceType">neuronxInstanceType</a></code> | <code><a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a></code> | The instance type of compile worker instance. |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.vpc">vpc</a></code> | <code>aws-cdk-lib.aws_ec2.IVpc</code> | VPC in which this will launch compile worker instance. |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.command">command</a></code> | <code>string[]</code> | *No description.* |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.environment">environment</a></code> | <code>{[ key: string ]: string}</code> | The environment variables to pass to the container. |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.secrets">secrets</a></code> | <code>{[ key: string ]: aws-cdk-lib.aws_batch.Secret}</code> | Secrets to pass to the container. |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.spot">spot</a></code> | <code>boolean</code> | Whether or not to use spot instances. |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.volumeSize">volumeSize</a></code> | <code>aws-cdk-lib.Size</code> | The root volume of worker instance. |
-| <code><a href="#aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.vpcSubnets">vpcSubnets</a></code> | <code>aws-cdk-lib.aws_ec2.SubnetSelection</code> | The VPC Subnets this Compute Environment will launch instances in. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.artifactS3Prefix">artifactS3Prefix</a></code> | <code>string</code> | S3 Prefix that compiled artifact uploaded. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.bucket">bucket</a></code> | <code>aws-cdk-lib.aws_s3.IBucket</code> | The bucket to upload compiled artifacts. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.image">image</a></code> | <code><a href="#aws-cdk-neuronx-patterns.INeuronxContainerImage">INeuronxContainerImage</a></code> | An image of the container where the compile job is executed. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.model">model</a></code> | <code><a href="#aws-cdk-neuronx-patterns.Model">Model</a></code> | The model to be compiled. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.neuronxInstanceType">neuronxInstanceType</a></code> | <code><a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a></code> | The instance type of compile worker instance. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.vpc">vpc</a></code> | <code>aws-cdk-lib.aws_ec2.IVpc</code> | VPC in which this will launch compile worker instance. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.command">command</a></code> | <code>string[]</code> | The command to run in the container. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.environment">environment</a></code> | <code>{[ key: string ]: string}</code> | The environment variables to pass to the container. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.secrets">secrets</a></code> | <code>{[ key: string ]: aws-cdk-lib.aws_batch.Secret}</code> | Secrets to pass to the container. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.spot">spot</a></code> | <code>boolean</code> | Whether or not to use spot instances. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.volumeSize">volumeSize</a></code> | <code>aws-cdk-lib.Size</code> | The root volume of worker instance. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.vpcSubnets">vpcSubnets</a></code> | <code>aws-cdk-lib.aws_ec2.SubnetSelection</code> | The VPC Subnets this Compute Environment will launch instances in. |
 
 ---
 
-##### `artifactS3Prefix`<sup>Required</sup> <a name="artifactS3Prefix" id="aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.artifactS3Prefix"></a>
+##### `artifactS3Prefix`<sup>Required</sup> <a name="artifactS3Prefix" id="aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.artifactS3Prefix"></a>
 
 ```typescript
 public readonly artifactS3Prefix: string;
@@ -6087,7 +6824,7 @@ This property is not depends on compile job finish.
 
 ---
 
-##### `bucket`<sup>Required</sup> <a name="bucket" id="aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.bucket"></a>
+##### `bucket`<sup>Required</sup> <a name="bucket" id="aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.bucket"></a>
 
 ```typescript
 public readonly bucket: IBucket;
@@ -6099,7 +6836,7 @@ The bucket to upload compiled artifacts.
 
 ---
 
-##### `image`<sup>Required</sup> <a name="image" id="aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.image"></a>
+##### `image`<sup>Required</sup> <a name="image" id="aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.image"></a>
 
 ```typescript
 public readonly image: INeuronxContainerImage;
@@ -6111,7 +6848,7 @@ An image of the container where the compile job is executed.
 
 ---
 
-##### `model`<sup>Required</sup> <a name="model" id="aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.model"></a>
+##### `model`<sup>Required</sup> <a name="model" id="aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.model"></a>
 
 ```typescript
 public readonly model: Model;
@@ -6123,7 +6860,7 @@ The model to be compiled.
 
 ---
 
-##### `neuronxInstanceType`<sup>Required</sup> <a name="neuronxInstanceType" id="aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.neuronxInstanceType"></a>
+##### `neuronxInstanceType`<sup>Required</sup> <a name="neuronxInstanceType" id="aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.neuronxInstanceType"></a>
 
 ```typescript
 public readonly neuronxInstanceType: INeuronxInstanceType;
@@ -6135,7 +6872,7 @@ The instance type of compile worker instance.
 
 ---
 
-##### `vpc`<sup>Required</sup> <a name="vpc" id="aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.vpc"></a>
+##### `vpc`<sup>Required</sup> <a name="vpc" id="aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.vpc"></a>
 
 ```typescript
 public readonly vpc: IVpc;
@@ -6147,7 +6884,7 @@ VPC in which this will launch compile worker instance.
 
 ---
 
-##### `command`<sup>Optional</sup> <a name="command" id="aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.command"></a>
+##### `command`<sup>Optional</sup> <a name="command" id="aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.command"></a>
 
 ```typescript
 public readonly command: string[];
@@ -6155,9 +6892,11 @@ public readonly command: string[];
 
 - *Type:* string[]
 
+The command to run in the container.
+
 ---
 
-##### `environment`<sup>Optional</sup> <a name="environment" id="aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.environment"></a>
+##### `environment`<sup>Optional</sup> <a name="environment" id="aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.environment"></a>
 
 ```typescript
 public readonly environment: {[ key: string ]: string};
@@ -6172,7 +6911,7 @@ This is only applicable when using container runtime.
 
 ---
 
-##### `secrets`<sup>Optional</sup> <a name="secrets" id="aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.secrets"></a>
+##### `secrets`<sup>Optional</sup> <a name="secrets" id="aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.secrets"></a>
 
 ```typescript
 public readonly secrets: {[ key: string ]: Secret};
@@ -6184,7 +6923,7 @@ Secrets to pass to the container.
 
 ---
 
-##### `spot`<sup>Optional</sup> <a name="spot" id="aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.spot"></a>
+##### `spot`<sup>Optional</sup> <a name="spot" id="aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.spot"></a>
 
 ```typescript
 public readonly spot: boolean;
@@ -6199,20 +6938,20 @@ Spot instances are less expensive EC2 instances that can be reclaimed by EC2 at
 
 ---
 
-##### `volumeSize`<sup>Optional</sup> <a name="volumeSize" id="aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.volumeSize"></a>
+##### `volumeSize`<sup>Optional</sup> <a name="volumeSize" id="aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.volumeSize"></a>
 
 ```typescript
 public readonly volumeSize: Size;
 ```
 
 - *Type:* aws-cdk-lib.Size
-- *Default:* N bilion parameters * 5GiB EBS
+- *Default:* N billion parameters * 5GiB EBS
 
 The root volume of worker instance.
 
 ---
 
-##### `vpcSubnets`<sup>Optional</sup> <a name="vpcSubnets" id="aws-cdk-neuronx-patterns.NeuronxCompilerProps.property.vpcSubnets"></a>
+##### `vpcSubnets`<sup>Optional</sup> <a name="vpcSubnets" id="aws-cdk-neuronx-patterns.NeuronxNativeCompilerProps.property.vpcSubnets"></a>
 
 ```typescript
 public readonly vpcSubnets: SubnetSelection;
@@ -10554,11 +11293,11 @@ const vllmNxdInferenceCompiledModel: VllmNxdInferenceCompiledModel = { ... }
 | **Name** | **Type** | **Description** |
 | --- | --- | --- |
 | <code><a href="#aws-cdk-neuronx-patterns.VllmNxdInferenceCompiledModel.property.bucket">bucket</a></code> | <code>aws-cdk-lib.aws_s3.IBucket</code> | The bucket to upload compiled artifacts. |
-| <code><a href="#aws-cdk-neuronx-patterns.VllmNxdInferenceCompiledModel.property.compileTimeInstanceType">compileTimeInstanceType</a></code> | <code><a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a></code> | *No description.* |
 | <code><a href="#aws-cdk-neuronx-patterns.VllmNxdInferenceCompiledModel.property.modelName">modelName</a></code> | <code>string</code> | The model name. |
+| <code><a href="#aws-cdk-neuronx-patterns.VllmNxdInferenceCompiledModel.property.recommendedInstanceType">recommendedInstanceType</a></code> | <code><a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a></code> | The recommended Neuron instance type for running inference with this compiled model. |
 | <code><a href="#aws-cdk-neuronx-patterns.VllmNxdInferenceCompiledModel.property.s3Prefix">s3Prefix</a></code> | <code>string</code> | S3 prefix that compiled artifact uploaded. |
 | <code><a href="#aws-cdk-neuronx-patterns.VllmNxdInferenceCompiledModel.property.s3Uri">s3Uri</a></code> | <code>string</code> | S3 URL that compiled artifact uploaded. |
-| <code><a href="#aws-cdk-neuronx-patterns.VllmNxdInferenceCompiledModel.property.weightSize">weightSize</a></code> | <code>aws-cdk-lib.Size</code> | *No description.* |
+| <code><a href="#aws-cdk-neuronx-patterns.VllmNxdInferenceCompiledModel.property.weightSize">weightSize</a></code> | <code>aws-cdk-lib.Size</code> | The weight size of the model. |
 | <code><a href="#aws-cdk-neuronx-patterns.VllmNxdInferenceCompiledModel.property.vllmArgs">vllmArgs</a></code> | <code><a href="#aws-cdk-neuronx-patterns.VllmEngineArguments">VllmEngineArguments</a></code> | Passed to the vllm engine at compile time. |
 
 ---
@@ -10575,25 +11314,27 @@ The bucket to upload compiled artifacts.
 
 ---
 
-##### `compileTimeInstanceType`<sup>Required</sup> <a name="compileTimeInstanceType" id="aws-cdk-neuronx-patterns.VllmNxdInferenceCompiledModel.property.compileTimeInstanceType"></a>
+##### `modelName`<sup>Required</sup> <a name="modelName" id="aws-cdk-neuronx-patterns.VllmNxdInferenceCompiledModel.property.modelName"></a>
 
 ```typescript
-public readonly compileTimeInstanceType: INeuronxInstanceType;
+public readonly modelName: string;
 ```
 
-- *Type:* <a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a>
+- *Type:* string
+
+The model name.
 
 ---
 
-##### `modelName`<sup>Required</sup> <a name="modelName" id="aws-cdk-neuronx-patterns.VllmNxdInferenceCompiledModel.property.modelName"></a>
+##### `recommendedInstanceType`<sup>Required</sup> <a name="recommendedInstanceType" id="aws-cdk-neuronx-patterns.VllmNxdInferenceCompiledModel.property.recommendedInstanceType"></a>
 
 ```typescript
-public readonly modelName: string;
+public readonly recommendedInstanceType: INeuronxInstanceType;
 ```
 
-- *Type:* string
+- *Type:* <a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a>
 
-The model name.
+The recommended Neuron instance type for running inference with this compiled model.
 
 ---
 
@@ -10629,6 +11370,8 @@ public readonly weightSize: Size;
 
 - *Type:* aws-cdk-lib.Size
 
+The weight size of the model.
+
 ---
 
 ##### `vllmArgs`<sup>Required</sup> <a name="vllmArgs" id="aws-cdk-neuronx-patterns.VllmNxdInferenceCompiledModel.property.vllmArgs"></a>
@@ -10662,6 +11405,7 @@ const vllmNxdInferenceCompileProps: VllmNxdInferenceCompileProps = { ... }
 | <code><a href="#aws-cdk-neuronx-patterns.VllmNxdInferenceCompileProps.property.bucket">bucket</a></code> | <code>aws-cdk-lib.aws_s3.IBucket</code> | The bucket to upload compiled artifacts. |
 | <code><a href="#aws-cdk-neuronx-patterns.VllmNxdInferenceCompileProps.property.model">model</a></code> | <code><a href="#aws-cdk-neuronx-patterns.Model">Model</a></code> | The model to be compiled. |
 | <code><a href="#aws-cdk-neuronx-patterns.VllmNxdInferenceCompileProps.property.vpc">vpc</a></code> | <code>aws-cdk-lib.aws_ec2.IVpc</code> | VPC in which this will launch compile worker instance. |
+| <code><a href="#aws-cdk-neuronx-patterns.VllmNxdInferenceCompileProps.property.compileInstanceType">compileInstanceType</a></code> | <code>aws-cdk-lib.aws_ec2.InstanceType</code> | The EC2 instance type to use for cross-compilation. |
 | <code><a href="#aws-cdk-neuronx-patterns.VllmNxdInferenceCompileProps.property.environment">environment</a></code> | <code>{[ key: string ]: string}</code> | The environment variables to pass to the container. |
 | <code><a href="#aws-cdk-neuronx-patterns.VllmNxdInferenceCompileProps.property.image">image</a></code> | <code><a href="#aws-cdk-neuronx-patterns.INeuronxContainerImage">INeuronxContainerImage</a></code> | An image of the container where the compile job is executed. |
 | <code><a href="#aws-cdk-neuronx-patterns.VllmNxdInferenceCompileProps.property.neuronxInstanceType">neuronxInstanceType</a></code> | <code><a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a></code> | The instance type of compile worker instance. |
@@ -10708,6 +11452,21 @@ VPC in which this will launch compile worker instance.
 
 ---
 
+##### `compileInstanceType`<sup>Optional</sup> <a name="compileInstanceType" id="aws-cdk-neuronx-patterns.VllmNxdInferenceCompileProps.property.compileInstanceType"></a>
+
+```typescript
+public readonly compileInstanceType: InstanceType;
+```
+
+- *Type:* aws-cdk-lib.aws_ec2.InstanceType
+- *Default:* Automatically selected based on model size
+
+The EC2 instance type to use for cross-compilation.
+
+This should be a non-Neuron instance type with sufficient memory for model compilation.
+
+---
+
 ##### `environment`<sup>Optional</sup> <a name="environment" id="aws-cdk-neuronx-patterns.VllmNxdInferenceCompileProps.property.environment"></a>
 
 ```typescript
@@ -11788,6 +12547,9 @@ new NeuronxInstanceType()
 | <code><a href="#aws-cdk-neuronx-patterns.NeuronxInstanceType.property.INF2_XLARGE">INF2_XLARGE</a></code> | <code><a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a></code> | inf2.xlarge. |
 | <code><a href="#aws-cdk-neuronx-patterns.NeuronxInstanceType.property.TRN1_2XLARGE">TRN1_2XLARGE</a></code> | <code><a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a></code> | trn1.2xlarge. |
 | <code><a href="#aws-cdk-neuronx-patterns.NeuronxInstanceType.property.TRN1_32XLARGE">TRN1_32XLARGE</a></code> | <code><a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a></code> | trn1.32xlarge. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxInstanceType.property.TRN2_3XLARGE">TRN2_3XLARGE</a></code> | <code><a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a></code> | trn2.3xlarge. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxInstanceType.property.TRN2_48XLARGE">TRN2_48XLARGE</a></code> | <code><a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a></code> | trn2.48xlarge. |
+| <code><a href="#aws-cdk-neuronx-patterns.NeuronxInstanceType.property.TRN2U_48XLARGE">TRN2U_48XLARGE</a></code> | <code><a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a></code> | trn2u.48xlarge. |
 
 ---
 
@@ -11863,6 +12625,42 @@ trn1.32xlarge.
 
 ---
 
+##### `TRN2_3XLARGE`<sup>Required</sup> <a name="TRN2_3XLARGE" id="aws-cdk-neuronx-patterns.NeuronxInstanceType.property.TRN2_3XLARGE"></a>
+
+```typescript
+public readonly TRN2_3XLARGE: INeuronxInstanceType;
+```
+
+- *Type:* <a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a>
+
+trn2.3xlarge.
+
+---
+
+##### `TRN2_48XLARGE`<sup>Required</sup> <a name="TRN2_48XLARGE" id="aws-cdk-neuronx-patterns.NeuronxInstanceType.property.TRN2_48XLARGE"></a>
+
+```typescript
+public readonly TRN2_48XLARGE: INeuronxInstanceType;
+```
+
+- *Type:* <a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a>
+
+trn2.48xlarge.
+
+---
+
+##### `TRN2U_48XLARGE`<sup>Required</sup> <a name="TRN2U_48XLARGE" id="aws-cdk-neuronx-patterns.NeuronxInstanceType.property.TRN2U_48XLARGE"></a>
+
+```typescript
+public readonly TRN2U_48XLARGE: INeuronxInstanceType;
+```
+
+- *Type:* <a href="#aws-cdk-neuronx-patterns.INeuronxInstanceType">INeuronxInstanceType</a>
+
+trn2u.48xlarge.
+
+---
+
 ### Parameters <a name="Parameters" id="aws-cdk-neuronx-patterns.Parameters"></a>
 
 Represents the amount of parameters.
@@ -12679,6 +13477,73 @@ public readonly neuronxCores: number;
 ---
 
 
+### Trainium2Chips <a name="Trainium2Chips" id="aws-cdk-neuronx-patterns.Trainium2Chips"></a>
+
+- *Implements:* <a href="#aws-cdk-neuronx-patterns.IAcceleratorChips">IAcceleratorChips</a>
+
+#### Initializers <a name="Initializers" id="aws-cdk-neuronx-patterns.Trainium2Chips.Initializer"></a>
+
+```typescript
+import { Trainium2Chips } from 'aws-cdk-neuronx-patterns'
+
+new Trainium2Chips(chips: number)
+```
+
+| **Name** | **Type** | **Description** |
+| --- | --- | --- |
+| <code><a href="#aws-cdk-neuronx-patterns.Trainium2Chips.Initializer.parameter.chips">chips</a></code> | <code>number</code> | *No description.* |
+
+---
+
+##### `chips`<sup>Required</sup> <a name="chips" id="aws-cdk-neuronx-patterns.Trainium2Chips.Initializer.parameter.chips"></a>
+
+- *Type:* number
+
+---
+
+
+
+#### Properties <a name="Properties" id="Properties"></a>
+
+| **Name** | **Type** | **Description** |
+| --- | --- | --- |
+| <code><a href="#aws-cdk-neuronx-patterns.Trainium2Chips.property.acceleratorMemory">acceleratorMemory</a></code> | <code>aws-cdk-lib.Size</code> | *No description.* |
+| <code><a href="#aws-cdk-neuronx-patterns.Trainium2Chips.property.chips">chips</a></code> | <code>number</code> | *No description.* |
+| <code><a href="#aws-cdk-neuronx-patterns.Trainium2Chips.property.neuronxCores">neuronxCores</a></code> | <code>number</code> | *No description.* |
+
+---
+
+##### `acceleratorMemory`<sup>Required</sup> <a name="acceleratorMemory" id="aws-cdk-neuronx-patterns.Trainium2Chips.property.acceleratorMemory"></a>
+
+```typescript
+public readonly acceleratorMemory: Size;
+```
+
+- *Type:* aws-cdk-lib.Size
+
+---
+
+##### `chips`<sup>Required</sup> <a name="chips" id="aws-cdk-neuronx-patterns.Trainium2Chips.property.chips"></a>
+
+```typescript
+public readonly chips: number;
+```
+
+- *Type:* number
+
+---
+
+##### `neuronxCores`<sup>Required</sup> <a name="neuronxCores" id="aws-cdk-neuronx-patterns.Trainium2Chips.property.neuronxCores"></a>
+
+```typescript
+public readonly neuronxCores: number;
+```
+
+- *Type:* number
+
+---
+
+
 ### VllmEngineArgumentsParser <a name="VllmEngineArgumentsParser" id="aws-cdk-neuronx-patterns.VllmEngineArgumentsParser"></a>
 
 #### Initializers <a name="Initializers" id="aws-cdk-neuronx-patterns.VllmEngineArgumentsParser.Initializer"></a>
@@ -13117,7 +13982,7 @@ The neuronx SDK version.
 
 ### IAcceleratorChips <a name="IAcceleratorChips" id="aws-cdk-neuronx-patterns.IAcceleratorChips"></a>
 
-- *Implemented By:* <a href="#aws-cdk-neuronx-patterns.Inferentia2Chips">Inferentia2Chips</a>, <a href="#aws-cdk-neuronx-patterns.Trainium1Chips">Trainium1Chips</a>, <a href="#aws-cdk-neuronx-patterns.IAcceleratorChips">IAcceleratorChips</a>
+- *Implemented By:* <a href="#aws-cdk-neuronx-patterns.Inferentia2Chips">Inferentia2Chips</a>, <a href="#aws-cdk-neuronx-patterns.Trainium1Chips">Trainium1Chips</a>, <a href="#aws-cdk-neuronx-patterns.Trainium2Chips">Trainium2Chips</a>, <a href="#aws-cdk-neuronx-patterns.IAcceleratorChips">IAcceleratorChips</a>
 
 
 #### Properties <a name="Properties" id="Properties"></a>
@@ -13160,6 +14025,27 @@ public readonly neuronxCores: number;
 
 ---
 
+### INeuronxCompiler <a name="INeuronxCompiler" id="aws-cdk-neuronx-patterns.INeuronxCompiler"></a>
+
+- *Implemented By:* <a href="#aws-cdk-neuronx-patterns.NeuronxCompilerBase">NeuronxCompilerBase</a>, <a href="#aws-cdk-neuronx-patterns.NeuronxCrossCompiler">NeuronxCrossCompiler</a>, <a href="#aws-cdk-neuronx-patterns.NeuronxNativeCompiler">NeuronxNativeCompiler</a>, <a href="#aws-cdk-neuronx-patterns.INeuronxCompiler">INeuronxCompiler</a>
+
+Interface for Neuronx compilers.
+
+#### Methods <a name="Methods" id="Methods"></a>
+
+| **Name** | **Description** |
+| --- | --- |
+| <code><a href="#aws-cdk-neuronx-patterns.INeuronxCompiler.compile">compile</a></code> | *No description.* |
+
+---
+
+##### `compile` <a name="compile" id="aws-cdk-neuronx-patterns.INeuronxCompiler.compile"></a>
+
+```typescript
+public compile(): NeuronxCompiledModel
+```
+
+
 ### INeuronxContainerImage <a name="INeuronxContainerImage" id="aws-cdk-neuronx-patterns.INeuronxContainerImage"></a>
 
 - *Implemented By:* <a href="#aws-cdk-neuronx-patterns.VllmNxdInferenceCompileImage">VllmNxdInferenceCompileImage</a>, <a href="#aws-cdk-neuronx-patterns.VllmNxdInferenceEcsImage">VllmNxdInferenceEcsImage</a>, <a href="#aws-cdk-neuronx-patterns.VllmNxdInferenceEcsImageBase">VllmNxdInferenceEcsImageBase</a>, <a href="#aws-cdk-neuronx-patterns.INeuronxContainerImage">INeuronxContainerImage</a>