From 39795794ab1bf625d23c5ce7d56565223a8c239a Mon Sep 17 00:00:00 2001
From: "promptless[bot]" <179508745+promptless[bot]@users.noreply.github.com>
Date: Thu, 16 Oct 2025 18:04:26 +0000
Subject: [PATCH 1/2] Documentation updates from Promptless
---
docs.json | 10 +
integrations/dstack.mdx | 222 +++++++-----------
integrations/mods.mdx | 56 ++---
integrations/overview.mdx | 103 ++++++++
integrations/skypilot.mdx | 61 +++--
tutorials/integrations/crewai-integration.mdx | 139 +++++++++++
tutorials/integrations/n8n-integration.mdx | 143 +++++++++++
7 files changed, 551 insertions(+), 183 deletions(-)
create mode 100644 integrations/overview.mdx
create mode 100644 tutorials/integrations/crewai-integration.mdx
create mode 100644 tutorials/integrations/n8n-integration.mdx
diff --git a/docs.json b/docs.json
index c24f8737..4b08a87a 100644
--- a/docs.json
+++ b/docs.json
@@ -194,9 +194,15 @@
{
"group": "Integrations",
"pages": [
+ "integrations/overview",
+ "group": "Examples",
+ "pages":[
"integrations/dstack",
"integrations/mods",
"integrations/skypilot"
+ "tutorials/integrations/n8n-integration",
+ "tutorials/integrations/crewai-integration"
+ ]
]
},
{
@@ -472,6 +478,10 @@
},
"redirects": [
+ {
+ "source": "/serverless/development/integrations",
+ "destination": "/integrations/overview"
+ },
{
"source": "/references/faq",
"destination": "/references/troubleshooting/zero-gpus"
diff --git a/integrations/dstack.mdx b/integrations/dstack.mdx
index fc22078b..8e1c5558 100644
--- a/integrations/dstack.mdx
+++ b/integrations/dstack.mdx
@@ -3,46 +3,44 @@ title: "Manage Pods with dstack on Runpod"
sidebarTitle: "dstack"
---
-[dstack](https://dstack.ai/) is an open-source tool that simplifies the orchestration of Pods for AI and ML workloads. By defining your application and resource requirements in YAML configuration files, it automates the provisioning and management of cloud resources on Runpod, allowing you to focus on your application logic rather than the infrastructure.
+[dstack](https://dstack.ai/) is an open-source tool that automates Pod orchestration for AI and ML workloads. It lets you define your application and resource requirements in YAML files, then handles provisioning and managing cloud resources on Runpod so you can focus on your application instead of infrastructure.
-In this guide, we'll walk through setting up [dstack](https://dstack.ai/) with Runpod to deploy [vLLM](https://github.com/vllm-project/vllm). We'll serve the `meta-llama/Llama-3.1-8B-Instruct` model from Hugging Face using a Python environment.
+This guide shows you how to set up dstack with Runpod and deploy [vLLM](https://github.com/vllm-project/vllm) to serve the `meta-llama/Llama-3.1-8B-Instruct` model from Hugging Face.
-## Prerequisites
+## Requirements
-* [A Runpod account with an API key](/get-started/api-keys)
+You'll need:
-* On your local machine:
+* [A Runpod account with an API key](/get-started/api-keys).
+* Python 3.8 or higher installed on your local machine.
+* `pip` (or `pip3` on macOS).
+* Basic utilities like `curl`.
- * Python 3.8 or higher
- * `pip` (or `pip3` on macOS)
- * Basic utilities: `curl`
-
-* These instructions are applicable for macOS, Linux, and Windows systems.
+These instructions work on macOS, Linux, and Windows.
-**Windows Users**
+**Windows users**
-* It's recommended to use [WSL (Windows Subsystem for Linux)](https://docs.microsoft.com/en-us/windows/wsl/install) or tools like [Git Bash](https://gitforwindows.org/) to follow along with the Unix-like commands used in this tutorial
-* Alternatively, Windows users can use PowerShell or Command Prompt and adjust commands accordingly
+Use [WSL (Windows Subsystem for Linux)](https://docs.microsoft.com/en-us/windows/wsl/install) or [Git Bash](https://gitforwindows.org/) to follow along with the Unix-like commands in this guide. Alternatively, use PowerShell or Command Prompt and adjust commands as needed.
-## Installation
-
-### Setting Up the dstack Server
+## Set up dstack
-1. **Prepare Your Workspace**
+### Install and configure the server
- Open a terminal or command prompt and create a new directory for this tutorial:
+
+
+ Open a terminal and create a new directory:
```bash
mkdir runpod-dstack-tutorial
cd runpod-dstack-tutorial
```
+
-2. **Set Up a Python Virtual Environment**
-
+
```bash
@@ -78,10 +76,10 @@ In this guide, we'll walk through setting up [dstack](https://dstack.ai/) with R
+
-3. **Install dstack**
-
- Use `pip` to install dstack:
+
+ Install dstack using `pip`:
@@ -89,8 +87,6 @@ In this guide, we'll walk through setting up [dstack](https://dstack.ai/) with R
pip3 install -U "dstack[all]"
```
- **Note:** If `pip3` is not available, you may need to install it or use `pip`.
-
@@ -108,12 +104,14 @@ In this guide, we'll walk through setting up [dstack](https://dstack.ai/) with R
+
+
-### Configuring dstack for Runpod
+### Configure dstack for Runpod
-1. **Create the Global Configuration File**
-
- The following `config.yml` file is a **global configuration** used by [dstack](https://dstack.ai/) for all deployments on your computer. It's essential to place it in the correct configuration directory.
+
+
+ Create a `config.yml` file in the dstack configuration directory. This file stores your Runpod credentials for all dstack deployments.
* **Create the configuration directory:**
@@ -133,8 +131,6 @@ In this guide, we'll walk through setting up [dstack](https://dstack.ai/) with R
- **Command Prompt or PowerShell:**
-
```bash
mkdir %USERPROFILE%\.dstack\server
```
@@ -169,9 +165,7 @@ In this guide, we'll walk through setting up [dstack](https://dstack.ai/) with R
- * **Create the `config.yml` File**
-
- In the configuration directory, create a file named `config.yml` with the following content:
+ Create a file named `config.yml` with the following content:
```yml
projects:
@@ -183,17 +177,17 @@ In this guide, we'll walk through setting up [dstack](https://dstack.ai/) with R
api_key: YOUR_RUNPOD_API_KEY
```
- Replace `YOUR_RUNPOD_API_KEY` with the API key you obtained from Runpod.
-
-2. **Start the dstack Server**
+ Replace `YOUR_RUNPOD_API_KEY` with your actual Runpod API key.
+
- From the configuration directory, start the dstack server:
+
+ Start the dstack server:
```bash
dstack server
```
- You should see output indicating that the server is running:
+ You'll see output like this:
```bash
[INFO] Applying ~/.dstack/server/config.yml...
@@ -203,35 +197,33 @@ In this guide, we'll walk through setting up [dstack](https://dstack.ai/) with R
-The `ADMIN-TOKEN` displayed is important for accessing the dstack web UI.
+Save the `ADMIN-TOKEN` to access the dstack web UI.
+
-3. **Access the dstack Web UI**
-
-* Open your web browser and navigate to `http://127.0.0.1:3000`.
-* When prompted for an admin token, enter the `ADMIN-TOKEN` from the server output.
-* The web UI allows you to monitor and manage your deployments.
+
+ Open your browser and go to `http://127.0.0.1:3000`. Enter the `ADMIN-TOKEN` from the server output to access the web UI where you can monitor and manage deployments.
+
+
-## Deploying vLLM as a Task
-
-### Step 1: Configure the Deployment Task
-
-1. **Prepare for Deployment**
+## Deploy vLLM
-* Open a new terminal or command prompt window.
+### Configure the deployment
-* Navigate to your tutorial directory:
+
+
+ Open a new terminal and navigate to your tutorial directory:
```bash
cd runpod-dstack-tutorial
```
-* **Activate the Python Virtual Environment**
+ Activate the Python virtual environment:
@@ -264,21 +256,19 @@ The `ADMIN-TOKEN` displayed is important for accessing the dstack web UI.
+
-2. **Create a Directory for the Task**
-
-Create and navigate to a new directory for the deployment task:
+
+Create a new directory for the deployment:
```bash
mkdir task-vllm-llama
cd task-vllm-llama
```
+
-3. **Create the dstack Configuration File**
-
-* **Create the `.dstack.yml` File**
-
- Create a file named `.dstack.yml` (or `dstack.yml` if your system doesn't allow filenames starting with a dot) with the following content:
+
+ Create a file named `.dstack.yml` with the following content:
```yml
type: task
@@ -303,59 +293,51 @@ cd task-vllm-llama
-Replace `YOUR_HUGGING_FACE_HUB_TOKEN` with your actual [Hugging Face access token](https://huggingface.co/settings/tokens) (read-access is enough) or define the token in your environment variables. Without this token, the model cannot be downloaded as it is gated.
+Replace `YOUR_HUGGING_FACE_HUB_TOKEN` with your [Hugging Face access token](https://huggingface.co/settings/tokens). The model is gated and requires authentication to download.
+
+
-### Step 2: Initialize and Deploy the Task
-
-1. **Initialize dstack**
+### Initialize and deploy
-Run the following command **in the directory where your `.dstack.yml` file is located**:
+
+
+In the directory with your `.dstack.yml` file, run:
```bash
dstack init
```
+
-2. **Apply the Configuration**
-
-Deploy the task by applying the configuration:
+
+Deploy the task:
```bash
dstack apply
```
-* You will see an output summarizing the deployment configuration and available instances.
-
-* When prompted:
+ You'll see the deployment configuration and available instances. When prompted:
```bash
Submit the run vllm-llama-3.1-8b-instruct? [y/n]:
```
- Type `y` and press `Enter` to confirm.
+ Type `y` and press Enter.
-* The `ports` configuration provides port forwarding from the deployed pod to `localhost`, allowing you to access the deployed vLLM via `localhost:8000`.
+The `ports` configuration forwards the deployed Pod's port to `localhost:8000` on your machine.
+
-3. **Monitor the Deployment**
+
+dstack will provision the Pod, download the Docker image, install packages, download the model, and start the vLLM server. You'll see progress logs in the terminal.
-* After executing `dstack apply`, you'll see all the steps that dstack performs:
-
- * Provisioning the pod on Runpod.
- * Downloading the Docker image.
- * Installing required packages.
- * Downloading the model from Hugging Face.
- * Starting the vLLM server.
-
-* The logs of vLLM will be displayed in the terminal.
-
-* To monitor the logs at any time, run:
+To view logs at any time, run:
```bash
dstack logs vllm-llama-3.1-8b-instruct
```
-* Wait until you see logs indicating that vLLM is serving the model, such as:
+Wait until you see logs indicating the server is ready:
```
INFO: Started server process [1]
@@ -363,16 +345,14 @@ dstack apply
INFO: Application startup complete.
INFO: Uvicorn running on http://0.0.0.0:8000 (Press CTRL+C to quit)
```
+
+
-### Step 3: Test the Model Server
-
-1. **Access the Service**
-
-Since the `ports` configuration forwards port `8000` from the deployed pod to `localhost`, you can access the vLLM server via `http://localhost:8000`.
+### Test the deployment
-2. **Test the Service Using `curl`**
+The vLLM server is now accessible at `http://localhost:8000`.
-Use the following `curl` command to test the deployed model:
+Test it with `curl`:
@@ -428,9 +408,7 @@ curl.exe -Method Post http://localhost:8000/v1/chat/completions `
-3. **Verify the Response**
-
-You should receive a JSON response similar to the following:
+You'll receive a JSON response:
```json
{
@@ -460,45 +438,33 @@ You should receive a JSON response similar to the following:
}
```
-This confirms that the model is running and responding as expected.
-
-### Step 4: Clean Up
-
-To avoid incurring additional costs, it's important to stop the task when you're finished.
-
-1. **Stop the Task**
+### Clean up
-In the terminal where you ran `dstack apply`, you can stop the task by pressing `Ctrl + C`.
+Stop the task when you're done to avoid charges.
-You'll be prompted:
+Press `Ctrl + C` in the terminal where you ran `dstack apply`. When prompted:
```
Stop the run vllm-llama-3.1-8b-instruct before detaching? [y/n]:
```
-Type `y` and press `Enter` to confirm stopping the task.
+Type `y` and press Enter.
-2. **Terminate the Instance**
-
-The instance will terminate automatically after stopping the task.
-
-If you wish to ensure the instance is terminated immediately, you can run:
+The instance will terminate automatically. To ensure immediate termination, run:
```bash
dstack stop vllm-llama-3.1-8b-instruct
```
-3. **Verify Termination**
+Verify termination in your Runpod dashboard or the dstack web UI.
-Check your Runpod dashboard or the [dstack](https://dstack.ai/) web UI to ensure that the instance has been terminated.
+## Use volumes for persistent storage
-## Additional Tips: Using Volumes for Persistent Storage
+Volumes let you store data between runs and cache models to reduce startup times.
-If you need to retain data between runs or cache models to reduce startup times, you can use volumes.
+### Create a volume
-### Creating a Volume
-
-Create a separate [dstack](https://dstack.ai/) file named `volume.dstack.yml` with the following content:
+Create a file named `volume.dstack.yml`:
```yml
type: volume
@@ -513,7 +479,7 @@ size: 100GB
-The `region` ties your volume to a specific region, which then also ties your Pod to that same region.
+The `region` ties your volume to a specific region, which also ties your Pod to that region.
@@ -523,9 +489,7 @@ Apply the volume configuration:
dstack apply -f volume.dstack.yml
```
-This will create the volume named `llama31-volume`.
-
-### Using the Volume in Your Task
+### Use the volume in your task
Modify your `.dstack.yml` file to include the volume:
@@ -535,14 +499,6 @@ volumes:
path: /data
```
-This configuration will mount the volume to the `/data` directory inside your container.
-
-By doing this, you can store models and data persistently, which can be especially useful for large models that take time to download.
-
-For more information on using volumes with Runpod, refer to the [dstack blog on volumes](https://dstack.ai/blog/volumes-on-runpod/).
-
-***
-
-## Conclusion
+This mounts the volume to the `/data` directory inside your container, letting you store models and data persistently. This is useful for large models that take time to download.
-By leveraging [dstack](https://dstack.ai/) on Runpod, you can efficiently deploy and manage Pods, accelerating your development workflow and reducing operational overhead.
+For more information, see the [dstack blog on volumes](https://dstack.ai/blog/volumes-on-runpod/).
diff --git a/integrations/mods.mdx b/integrations/mods.mdx
index f5970c49..1000b1c5 100644
--- a/integrations/mods.mdx
+++ b/integrations/mods.mdx
@@ -3,28 +3,25 @@ title: "Running Runpod on Mods"
sidebarTitle: "Mods"
---
-[Mods](https://github.com/charmbracelet/mods) is an AI-powered tool designed for the command line and built to seamlessly integrate with pipelines. It provides a convenient way to interact with language models directly from your terminal.
+[Mods](https://github.com/charmbracelet/mods) is a command-line tool for interacting with language models. It integrates with Unix pipelines, letting you send command output directly to LLMs from your terminal.
-## How Mods Works
+## How Mods works
-Mods operates by reading standard input and prefacing it with a prompt supplied in the Mods arguments. It sends the input text to a language model (LLM) and prints out the generated result. Optionally, you can ask the LLM to format the response as Markdown. This allows you to "question" the output of a command, making it a powerful tool for interactive exploration and analysis. Additionally, Mods can work with standard input or an individually supplied argument prompt.
+Mods reads standard input (or a prompt you provide as an argument), sends it to a language model, and prints the result. You can prefix the input with a prompt from the Mods arguments, and optionally format the output as Markdown. This lets you pipe command output to an LLM for analysis or transformation.
-## Getting Started
+## Get started
-To start using Mods, follow these step-by-step instructions:
+
+
+ Get your API key from the [Runpod Settings](https://www.console.runpod.io/user/settings) page.
+
-1. **Obtain Your API Key**:
+
+ Follow the installation instructions for [Mods](https://github.com/charmbracelet/mods) based on your system.
+
- * Visit the [Runpod Settings](https://www.console.runpod.io/user/settings) page to retrieve your API key.
- * If you haven't created an account yet, you'll need to sign up before obtaining the key.
-
-2. **Install Mods**:
-
- * Refer to the different installation methods for [Mods](https://github.com/charmbracelet/mods) based on your preferred approach.
-
-3. **Configure Runpod**:
-
- * Update the `config_template.yml` file to use your Runpod configuration. Here's an example:
+
+ Update the `config_template.yml` file with your Runpod configuration:
```yml
runpod:
@@ -39,24 +36,21 @@ To start using Mods, follow these step-by-step instructions:
max-input-chars: 8192
```
- * `base-url`: Update your base-url with your specific endpoint.
-
- * `api-key-env`: Add your Runpod API key.
-
- * `openchat/openchat-3.5-1210`: Replace with the name of the model you want to use.
-
- * `aliases: ["openchat"]`: Replace with your preferred model alias.
-
- * `max-input-chars`: Update the maximum input characters allowed for your model.
-
-4. **Verify Your Setup**:
-
- * To ensure everything is set up correctly, pipe any command line output and pass it to `mods`.
+ Replace the following values:
+ * `base-url`: Your specific endpoint URL.
+ * `api-key-env`: Your Runpod API key.
+ * `openchat/openchat-3.5-1210`: The model name you want to use.
+ * `aliases: ["openchat"]`: Your preferred model alias.
+ * `max-input-chars`: The maximum input characters for your model.
+
- * Specify the Runpod API and model you want to use.
+
+ Test your setup by piping command output to Mods:
```sh
ls ~/Downloads | mods --api runpod --model openchat -f "tell my fortune based on these files" | glow
```
- * This command will list the files in your `~/Downloads` directory, pass them to Mods using the Runpod API and the specified model, and format the response as a fortune based on the files. The output will then be piped to `glow` for a visually appealing display.
+ This lists files in your `~/Downloads` directory, sends them to Mods using the Runpod API and specified model, and pipes the output to `glow` for formatted display.
+
+
diff --git a/integrations/overview.mdx b/integrations/overview.mdx
new file mode 100644
index 00000000..ec88d2ab
--- /dev/null
+++ b/integrations/overview.mdx
@@ -0,0 +1,103 @@
+---
+title: "Integrate Runpod with external tools"
+sidebarTitle: "Overview"
+description: "Learn how to integrate Runpod endpoints with external tools and frameworks."
+---
+
+Runpod can be integrated with any system that supports custom endpoint configuration. If a library or framework lets you set a custom base URL, you can connect it to your Runpod endpoints without specialized adapters or connectors.
+
+## How it works
+
+Integration is straightforward: any library or framework that accepts a custom base URL for API calls will work with Runpod. This means you can use Runpod with tools like n8n, CrewAI, LangChain, and many others by simply pointing them to your Runpod endpoint URL.
+
+## Deployment options
+
+Runpod offers four deployment options for endpoint integrations:
+
+### Public Endpoints
+
+Public Endpoints are pre-deployed AI models that you can use without setting up your own Serverless endpoint. They're vLLM-compatible and return OpenAI-compatible responses, so you can get started quickly or test things out without deploying infrastructure.
+
+The following Public Endpoint URLs are available for OpenAI-compatible models:
+
+```
+# Public Endpoint for Qwen3 32B AWQ
+https://api.runpod.ai/v2/qwen3-32b-awq/openai/v1
+
+# Public Endpoint for ibm/IBM Granite-4.0-H-Small
+https://api.runpod.ai/v2/granite-4-0-h-small/openai/v1
+```
+
+For pricing details and usage information, see [Public Endpoints](/hub/public-endpoints).
+
+### vLLM
+
+[vLLM workers](/serverless/vllm/overview) provide an inference engine that returns [OpenAI-compatible responses](/serverless/vllm/openai-compatibility), making it ideal for tools that expect OpenAI's API format.
+
+When you deploy a vLLM endpoint, access it using the OpenAI-compatible API at:
+
+```
+https://api.runpod.ai/v2/ENDPOINT_ID/openai/v1
+```
+
+Where `ENDPOINT_ID` is your Serverless endpoint ID.
+
+### SGLang
+
+[SGLang workers](/serverless/sglang/overview) are an inference engine that returns OpenAI-compatible responses. It offers optimized performance for certain model types and use cases.
+
+To learn more, see the [runpod-workers/worker-sglang](https://github.com/runpod-workers/worker-sglang) repository on GitHub.
+
+### Load balancer
+
+[Load balancing endpoints](/serverless/load-balancing/overview) let you create custom endpoints where you define your own inputs and outputs. This gives you complete control over the API contract and is ideal when you need custom behavior beyond standard inference patterns.
+
+## Model configuration for compatibility
+
+Some models require specific vLLM environment variables to work with external tools and frameworks. You may need to set a custom chat template or [tool call parser](https://docs.vllm.ai/en/latest/features/tool_calling.html) to ensure your model returns responses in the format your integration expects.
+
+For example, you can configure the `Qwen/qwen3-32b-awq` model for OpenAI compatibility by adding these environment variables in your vLLM endpoint settings:
+
+```txt
+ENABLE_AUTO_TOOL_CHOICE=true
+REASONING_PARSER=qwen3
+TOOL_CALL_PARSER=hermes
+```
+
+These settings enable automatic tool choice selection and set the right parsers for the Qwen3 model to work with tools that expect OpenAI-formatted responses.
+
+For more information about tool calling configuration and available parsers, see the [vLLM tool calling documentation](https://docs.vllm.ai/en/latest/features/tool_calling.html).
+
+## Integration tutorials
+
+Follow these step-by-step tutorials to integrate Runpod with popular tools:
+
+
+
+ Connect Runpod to n8n for AI-powered workflow automation.
+
+
+
+ Use Runpod to power autonomous AI agents in CrewAI.
+
+
+
+## Compatible frameworks
+
+The same integration pattern works with any framework that supports custom OpenAI-compatible endpoints, including:
+
+* **CrewAI**: A framework for orchestrating role-playing autonomous AI agents.
+* **LangChain**: A framework for developing applications powered by language models.
+* **AutoGen**: Microsoft's framework for building multi-agent conversational systems.
+* **Haystack**: An end-to-end framework for building search systems and question answering.
+* **n8n**: A workflow automation tool with AI integration capabilities.
+
+Configure these frameworks to use your Runpod endpoint URL as the base URL, and provide your Runpod API key for authentication.
+
+## Third-party integrations
+
+For infrastructure management and orchestration, Runpod integrates with:
+
+* [**dstack**](/integrations/dstack): Simplified Pod orchestration for AI/ML workloads.
+* [**SkyPilot**](/integrations/skypilot): Multi-cloud execution framework.
+* [**Mods**](/integrations/mods): AI-powered command-line tool.
diff --git a/integrations/skypilot.mdx b/integrations/skypilot.mdx
index c48d63be..853f3304 100644
--- a/integrations/skypilot.mdx
+++ b/integrations/skypilot.mdx
@@ -3,41 +3,56 @@ title: "Running Runpod on SkyPilot"
sidebarTitle: "SkyPilot"
---
-[SkyPilot](https://skypilot.readthedocs.io/en/latest/) is a framework for executing LLMs, AI, and batch jobs on any cloud, offering maximum cost savings, highest GPU availability, and managed execution.
+[SkyPilot](https://skypilot.readthedocs.io/en/latest/) is a framework for running LLMs, AI, and batch jobs on any cloud.
-This integration leverages the Runpod CLI infrastructure, streamlining the process of spinning up on-demand pods and deploying serverless endpoints with SkyPilot.
+This integration uses the Runpod CLI infrastructure to spin up on-demand Pods and deploy Serverless endpoints with SkyPilot.
-## Getting started
+## Get started
-To begin using Runpod with SkyPilot, follow these steps:
+
+
+ Get your API key from the [Runpod Settings](https://www.console.runpod.io/user/settings) page.
+
-1. **Obtain Your API Key**: Visit the [Runpod Settings](https://www.console.runpod.io/user/settings) page to get your API key. If you haven't created an account yet, you'll need to do so before obtaining the key.
-
-2. **Install Runpod**: Use the following command to install the latest version of Runpod:
+
+ Install the latest version of Runpod:
```sh
pip install "runpod>=1.6"
```
+
-3. **Configure Runpod**: Enter `runpod config` in your CLI and paste your API key when prompted.
+
+ Run `runpod config` and paste your API key when prompted.
+
-4. **Install SkyPilot Runpod Cloud**: Execute the following command to install the [SkyPilot Runpod cloud](https://skypilot.readthedocs.io/en/latest/getting-started/installation.html#runpod):
+
+ Install the [SkyPilot Runpod cloud](https://skypilot.readthedocs.io/en/latest/getting-started/installation.html#runpod):
```sh
pip install "skypilot-nightly[runpod]"
```
+
-5. **Verify Your Setup**: Run `sky check` to ensure your credentials are correctly set up and you're ready to proceed.
-
-## Running a Project
+
+ Run `sky check` to verify your credentials are set up correctly.
+
+
-After setting up your environment, you can seamlessly spin up a cluster in minutes:
+## Run a project
-1. **Create a New Project Directory**: Run `mkdir hello-sky` to create a new directory for your project.
+
+
+ Create a new directory for your project:
-2. **Navigate to Your Project Directory**: Change into your project directory with `cd hello-sky`.
+ ```sh
+ mkdir hello-sky
+ cd hello-sky
+ ```
+
-3. **Create a Configuration File**: Enter `cat > hello_sky.yaml` and input the following configuration details:
+
+ Create a file named `hello_sky.yaml` with the following content:
```yml
resources:
@@ -60,9 +75,17 @@ After setting up your environment, you can seamlessly spin up a cluster in minut
echo "Hello, SkyPilot!"
conda env list
```
+
-4. **Launch Your Project**: With your configuration file created, launch your project on the cluster by running `sky launch -c mycluster hello_sky.yaml`.
+
+ Launch your project on the cluster:
-5. **Confirm Your GPU Type**: You should see the available GPU options on Secure Cloud appear in your command line. Once you confirm your GPU type, your cluster will start spinning up.
+ ```sh
+ sky launch -c mycluster hello_sky.yaml
+ ```
+
-With this integration, you can leverage the power of Runpod and SkyPilot to efficiently run your LLMs, AI, and batch jobs on any cloud.
+
+ You'll see the available GPU options. Confirm your GPU type and the cluster will start spinning up.
+
+
diff --git a/tutorials/integrations/crewai-integration.mdx b/tutorials/integrations/crewai-integration.mdx
new file mode 100644
index 00000000..8a98d6e6
--- /dev/null
+++ b/tutorials/integrations/crewai-integration.mdx
@@ -0,0 +1,139 @@
+---
+title: "Integrate Runpod with CrewAI"
+description: "Learn how to deploy a vLLM worker on Runpod and connect it to CrewAI for orchestrating autonomous AI agents."
+---
+
+Learn how to integrate Runpod Serverless with CrewAI, a framework for orchestrating role-playing autonomous AI agents. By the end of this tutorial, you'll have a vLLM endpoint running on Runpod that you can use to power your CrewAI agents.
+
+## What you'll learn
+
+In this tutorial, you'll learn how to:
+
+* Deploy a vLLM worker on Runpod Serverless.
+* Configure your vLLM endpoint for OpenAI compatibility.
+* Connect CrewAI to your Runpod endpoint.
+* Test your integration with a simple agent.
+
+## Requirements
+
+* You've [created a Runpod account](/get-started/manage-accounts).
+* You've created a [Runpod API key](/get-started/api-keys).
+* You have CrewAI installed in your development environment.
+* (Optional) For gated models, you've created a [Hugging Face access token](https://huggingface.co/docs/hub/en/security-tokens).
+
+## Step 1: Deploy a vLLM worker on Runpod
+
+First, you'll deploy a vLLM worker that will serve your language model.
+
+
+
+ Open the [Runpod console](https://www.console.runpod.io/serverless) and navigate to the Serverless page.
+
+
+
+ Under **Quick Deploy**, find **Serverless vLLM** and click **Configure**.
+
+
+
+ In the deployment modal:
+
+ * Select a vLLM version (latest stable recommended).
+ * Under **Hugging Face Models**, enter your model name (e.g., `openchat/openchat-3.5-0106`).
+ * If using a gated model like Llama, enter your **Hugging Face Token**.
+ * Click **Next**.
+
+
+
+ In the vLLM settings modal, under **LLM Settings**:
+
+ * Set **Max Model Length** to `8192` (or an appropriate context length for your model).
+ * If your CrewAI agents require tool calling, add environment variables like `ENABLE_AUTO_TOOL_CHOICE=true` and appropriate parsers. See the [vLLM tool calling documentation](https://docs.vllm.ai/en/latest/features/tool_calling.html) for details.
+ * Click **Next**.
+
+
+
+ Review the endpoint settings:
+
+ * Set your desired GPU type and worker count.
+ * Configure scaling settings based on your expected workload.
+ * Click **Deploy**.
+
+
+
+ Your endpoint will now begin initializing. This may take several minutes while Runpod provisions resources and downloads your model. Wait until the status shows as **Running**.
+
+
+
+ Once deployed, navigate to your endpoint in the Runpod console and copy the **Endpoint ID**. You'll need this to connect CrewAI.
+
+
+
+
+
+For more details on vLLM deployment options, see [Deploy a vLLM worker](/serverless/vllm/get-started).
+
+
+
+## Step 2: Connect CrewAI to your Runpod endpoint
+
+Now you'll configure CrewAI to use your Runpod endpoint as an OpenAI-compatible API.
+
+
+
+ Open the CrewAI dashboard and look for the **LLM connections** section.
+
+
+
+ Under **Provider**, select **custom-openai-compatible** from the dropdown menu.
+
+
+
+ Configure the connection with your Runpod credentials:
+
+ * For `OPENAI_API_KEY`, use your Runpod API Key. You can find or create API keys in the [Runpod console](https://www.runpod.io/console/user/settings).
+
+
+
+ For `OPENAI_API_BASE`, add the base URL for your vLLM's OpenAI-compatible endpoint:
+
+ ```
+ https://api.runpod.ai/v2/ENDPOINT_ID/openai/v1
+ ```
+
+ Replace `ENDPOINT_ID` with your actual endpoint ID from Step 1.
+
+
+
+ Click **Fetch Available Models** to test the connection. If successful, CrewAI will retrieve the list of models available on your endpoint.
+
+
+
+## Step 3: Test your integration
+
+Verify that your CrewAI agents can use your Runpod endpoint.
+
+
+
+ Create a simple CrewAI agent that uses your Runpod endpoint for its language model.
+
+
+
+ Assign a simple task to your agent and run it to verify that it can communicate with your Runpod endpoint.
+
+
+
+ Monitor requests from your CrewAI agents in the endpoint details page of the Runpod console.
+
+
+
+ Confirm that your agent is receiving appropriate responses from your model running on Runpod.
+
+
+
+## Next steps
+
+Now that you've integrated Runpod with CrewAI, you can:
+
+* Build complex multi-agent systems using your Runpod endpoint to serve the necessary models.
+* Explore other [integration options](/integrations/overview).
+* Learn more about [OpenAI compatibility](/serverless/vllm/openai-compatibility) features in vLLM.
diff --git a/tutorials/integrations/n8n-integration.mdx b/tutorials/integrations/n8n-integration.mdx
new file mode 100644
index 00000000..0f609168
--- /dev/null
+++ b/tutorials/integrations/n8n-integration.mdx
@@ -0,0 +1,143 @@
+---
+title: "Integrate Runpod with n8n"
+description: "Learn how to deploy a vLLM worker on Runpod and connect it to n8n for AI-powered workflow automation."
+---
+
+Learn how to integrate Runpod Serverless with n8n, a workflow automation tool. By the end of this tutorial, you'll have a vLLM endpoint running on Runpod that you can use within your n8n workflows.
+
+## What you'll learn
+
+In this tutorial, you'll learn how to:
+
+* Deploy a vLLM worker on Runpod Serverless.
+* Configure your vLLM endpoint for OpenAI compatibility.
+* Connect n8n to your Runpod endpoint.
+* Test your integration with a simple workflow.
+
+## Requirements
+
+* You've [created a Runpod account](/get-started/manage-accounts).
+* You've created a [Runpod API key](/get-started/api-keys).
+* You have an n8n instance (cloud or self-hosted).
+* (Optional) For gated models, you've created a [Hugging Face access token](https://huggingface.co/docs/hub/en/security-tokens).
+
+## Step 1: Deploy a vLLM worker on Runpod
+
+First, you'll deploy a vLLM worker that will serve your language model.
+
+
+
+ Open the [Runpod console](https://www.console.runpod.io/serverless) and navigate to the Serverless page.
+
+
+
+ Under **Quick Deploy**, find **Serverless vLLM** and click **Configure**.
+
+
+
+ In the deployment modal:
+
+ * Select a vLLM version (latest stable recommended).
+ * Under **Hugging Face Models**, enter your model name (e.g., `openchat/openchat-3.5-0106`).
+ * If using a gated model like Llama, enter your **Hugging Face Token**.
+ * Click **Next**.
+
+
+
+ In the vLLM settings modal, under **LLM Settings**:
+
+ * Set **Max Model Length** to `8192` (or an appropriate context length for your model).
+ * If your n8n workflows require tool calling, add environment variables like `ENABLE_AUTO_TOOL_CHOICE=true` and appropriate parsers. See the [vLLM tool calling documentation](https://docs.vllm.ai/en/latest/features/tool_calling.html) for details.
+ * Click **Next**.
+
+
+
+ Review the endpoint settings:
+
+ * Set your desired GPU type and worker count.
+ * Configure scaling settings based on your expected workload.
+ * Click **Deploy**.
+
+
+
+ Your endpoint will now begin initializing. This may take several minutes while Runpod provisions resources and downloads your model. Wait until the status shows as **Running**.
+
+
+
+ Once deployed, navigate to your endpoint in the Runpod console and copy the **Endpoint ID**. You'll need this to connect n8n.
+
+
+
+
+
+For more details on vLLM deployment options, see [Deploy a vLLM worker](/serverless/vllm/get-started).
+
+
+
+## Step 2: Connect n8n to your Runpod endpoint
+
+Now you'll configure n8n to use your Runpod endpoint as an OpenAI-compatible API.
+
+
+
+ In your n8n workflow, add a new **OpenAI Chat Model** node to your canvas.
+
+
+
+ Click the dropdown under **Credential to connect with** and select **Create new credential**.
+
+
+
+ Under **API Key**, add your Runpod API Key. You can find or create API keys in the [Runpod console](https://www.runpod.io/console/user/settings).
+
+
+
+ Under **Base URL**, replace the default OpenAI URL with your Runpod endpoint URL:
+
+ ```
+ https://api.runpod.ai/v2/ENDPOINT_ID/openai/v1
+ ```
+
+ Replace `ENDPOINT_ID` with your actual endpoint ID from Step 1.
+
+
+
+ Click **Save**. n8n will automatically test your endpoint connection. If successful, you can start using the node in your workflow.
+
+
+
+## Step 3: Test your integration
+
+Create a simple workflow to test your integration.
+
+
+
+ Add a **Manual Trigger** node and connect it to your **OpenAI Chat Model** node.
+
+
+
+ In the **OpenAI Chat Model** node, add a test message like "Hello, what can you help me with?"
+
+
+
+ Click **Execute Workflow** in n8n. You should see a response from your model running on Runpod.
+
+
+
+ Monitor requests from your n8n workflow in the endpoint details page of the Runpod console.
+
+
+
+
+
+The n8n chat feature may have trouble parsing output from vLLM depending on your model. If you experience issues, try adjusting your model's output format or testing with a different model.
+
+
+
+## Next steps
+
+Now that you've integrated Runpod with n8n, you can:
+
+* Build complex AI-powered workflows using your Runpod endpoints.
+* Explore other [integration options](/integrations/overview) with Runpod.
+* Learn about [OpenAI compatibility](/serverless/vllm/openai-compatibility) features in vLLM.
From fb2713c64ad4842fbcd267d58d044115f01b1f2c Mon Sep 17 00:00:00 2001
From: Mo King
Date: Thu, 6 Nov 2025 15:40:02 -0500
Subject: [PATCH 2/2] Add n8n integration guide
---
.github/ISSUE_TEMPLATE/bug-report.md | 2 +-
.../ISSUE_TEMPLATE/documentation-request.md | 2 +-
docs.json | 29 +-
hub/public-endpoint-reference.mdx | 110 ++++++-
images/serverless-endpoint-id.png | Bin 0 -> 268289 bytes
.../crewai-integration.mdx | 85 +++--
integrations/n8n-integration.mdx | 164 ++++++++++
integrations/overview.mdx | 51 ++-
overview.mdx | 11 +
pods/choose-a-pod.mdx | 2 +-
pods/configuration/use-ssh.mdx | 116 ++++---
pods/connect-to-a-pod.mdx | 4 +-
pods/manage-pods.mdx | 6 +-
pods/overview.mdx | 6 +-
pods/pricing.mdx | 4 +-
pods/storage/create-network-volumes.mdx | 2 +-
pods/storage/types.mdx | 22 +-
references/troubleshooting/storage-full.mdx | 6 +-
runpodctl/reference/runpodctl-create-pod.mdx | 2 +-
runpodctl/reference/runpodctl-create-pods.mdx | 2 +-
.../endpoints/endpoint-configurations.mdx | 4 +-
serverless/endpoints/job-states.mdx | 1 +
serverless/endpoints/manage-endpoints.mdx | 3 +-
serverless/endpoints/overview.mdx | 42 +--
serverless/endpoints/send-requests.mdx | 2 +-
serverless/load-balancing/build-a-worker.mdx | 16 +-
serverless/load-balancing/overview.mdx | 75 ++++-
serverless/load-balancing/vllm-worker.mdx | 16 +-
serverless/overview.mdx | 4 +-
serverless/pricing.mdx | 3 +-
serverless/storage/overview.mdx | 6 +-
serverless/vllm/environment-variables.mdx | 3 +-
serverless/vllm/get-started.mdx | 90 +++--
serverless/vllm/openai-compatibility.mdx | 193 +++++++----
serverless/vllm/overview.mdx | 133 +++++---
serverless/vllm/vllm-requests.mdx | 307 ++++++++++++------
serverless/workers/concurrent-handler.mdx | 3 +-
serverless/workers/custom-worker.mdx | 3 +-
serverless/workers/deploy.mdx | 2 +-
serverless/workers/github-integration.mdx | 2 +-
serverless/workers/handler-functions.mdx | 20 +-
serverless/workers/overview.mdx | 3 +-
storage/network-volumes.mdx | 30 +-
storage/s3-api.mdx | 4 +
tutorials/integrations/n8n-integration.mdx | 143 --------
tutorials/pods/comfyui.mdx | 2 +-
46 files changed, 1077 insertions(+), 659 deletions(-)
create mode 100644 images/serverless-endpoint-id.png
rename {tutorials/integrations => integrations}/crewai-integration.mdx (55%)
create mode 100644 integrations/n8n-integration.mdx
delete mode 100644 tutorials/integrations/n8n-integration.mdx
diff --git a/.github/ISSUE_TEMPLATE/bug-report.md b/.github/ISSUE_TEMPLATE/bug-report.md
index 9f674ae4..e6a982a7 100644
--- a/.github/ISSUE_TEMPLATE/bug-report.md
+++ b/.github/ISSUE_TEMPLATE/bug-report.md
@@ -3,7 +3,7 @@ name: Bug report
about: Create a report to help us improve
title: "[Bug]"
labels: ''
-assignees: AllysonRosenthal, rachfop
+assignees: muhsinking
---
diff --git a/.github/ISSUE_TEMPLATE/documentation-request.md b/.github/ISSUE_TEMPLATE/documentation-request.md
index 9406a085..d656e95f 100644
--- a/.github/ISSUE_TEMPLATE/documentation-request.md
+++ b/.github/ISSUE_TEMPLATE/documentation-request.md
@@ -3,7 +3,7 @@ name: Documentation request
about: Suggest an idea for this project
title: "[Request]"
labels: ''
-assignees: AllysonRosenthal, rachfop
+assignees: muhsinking
---
diff --git a/docs.json b/docs.json
index 4b08a87a..ccf6e549 100644
--- a/docs.json
+++ b/docs.json
@@ -142,6 +142,21 @@
"group": "Storage",
"pages": ["storage/network-volumes", "storage/s3-api"]
},
+ {
+ "group": "Integrations",
+ "pages": [
+ "integrations/overview",
+ {
+ "group": "Guides",
+ "pages":[
+ "integrations/n8n-integration",
+ "integrations/dstack",
+ "integrations/mods",
+ "integrations/skypilot"
+ ]
+ }
+ ]
+ },
{
"group": "Hub",
"pages": [
@@ -191,20 +206,6 @@
"references/troubleshooting/manage-payment-cards"
]
},
- {
- "group": "Integrations",
- "pages": [
- "integrations/overview",
- "group": "Examples",
- "pages":[
- "integrations/dstack",
- "integrations/mods",
- "integrations/skypilot"
- "tutorials/integrations/n8n-integration",
- "tutorials/integrations/crewai-integration"
- ]
- ]
- },
{
"group": "Migrations",
"pages": [
diff --git a/hub/public-endpoint-reference.mdx b/hub/public-endpoint-reference.mdx
index 5878a6eb..d998811b 100644
--- a/hub/public-endpoint-reference.mdx
+++ b/hub/public-endpoint-reference.mdx
@@ -13,6 +13,7 @@ The following models are currently available:
| Model | Description | Endpoint URL | Type | Price |
|-------|-------------|----------|------|-------|
+| **IBM Granite-4.0-H-Small** | A 32B parameter long-context instruct model. | `https://api.runpod.ai/v2/granite-4-0-h-small/` | Text | \$0.01 per 1000 tokens |
| **Qwen3 32B AWQ** | The latest LLM in the Qwen series, offering advancements in reasoning, instruction-following, agent capabilities, and multilingual support. | `https://api.runpod.ai/v2/qwen3-32b-awq/` | Text | \$0.01 per 1000 tokens |
| **Flux Dev** | Offers exceptional prompt adherence, high visual fidelity, and rich image detail. | `https://api.runpod.ai/v2/black-forest-labs-flux-1-dev/` | Image | \$.02 per megapixel |
| **Flux Schnell** | Fastest and most lightweight FLUX model, ideal for local development, prototyping, and personal use. | `https://api.runpod.ai/v2/black-forest-labs-flux-1-schnell/` | Image | \$.0024 per megapixel |
@@ -24,7 +25,11 @@ The following models are currently available:
| **Seedream 4.0 Edit** | New-generation image editing with unified generation and editing architecture. | `https://api.runpod.ai/v2/seedream-v4-edit/` | Image | \$0.027 per megapixel
| **Seedream 3.0** | Native high-resolution bilingual image generation (Chinese-English). | `https://api.runpod.ai/v2/seedream-3-0-t2i/` | Image | \$0.03 per megapixel
| **Nano Banana Edit** | Google's state-of-the-art image editing model. | `https://api.runpod.ai/v2/nano-banana-edit/` | Image | \$0.027 per megapixel
+| **InfiniteTalk** | Audio-driven video generation model that creates talking or singing videos from a single image and audio input. | `https://api.runpod.ai/v2/infinitetalk/` | Video | \$0.25 per video generation |
+| **Kling v2.1 I2V Pro** | Professional-grade image-to-video with enhanced visual fidelity. | `https://api.runpod.ai/v2/kling-v2-1-i2v-pro/` | Video | \$0.36 per 5 seconds of video |
| **Seedance 1.0 Pro** | High-performance video generation with multi-shot storytelling. | `https://api.runpod.ai/v2/seedance-1-0-pro/` | Video | \$0.62 per 5 seconds of video |
+| **SORA 2 I2V** | OpenAI's Sora 2 is a video and audio generation model. | `https://api.runpod.ai/v2/sora-2-i2v/` | Video | \$0.40 per video generation |
+| **SORA 2 Pro I2V** | OpenAI's Sora 2 Pro is a professional-grade video and audio generation model. | `https://api.runpod.ai/v2/sora-2-pro-i2v/` | Video | \$1.20 per video generation |
| **WAN 2.5** | Image-to-video generation model. | `https://api.runpod.ai/v2/wan-2-5/` | Video | \$0.50 per 5 seconds of video
| **WAN 2.2 I2V 720p LoRA** | Open-source video generation with LoRA support. | `https://api.runpod.ai/v2/wan-2-2-t2v-720-lora/` | Video | \$0.35 per 5 seconds of video
| **WAN 2.2 I2V 720p** | Open-source AI video generation model that uses a diffusion transformer architecture for image-to-video generation. | `https://api.runpod.ai/v2/wan-2-2-i2v-720/` | Video | \$0.30 per 5 seconds of video
@@ -101,6 +106,43 @@ Flux Schnell is optimized for speed and real-time applications:
Flux Schnell is optimized for speed and works best with lower step counts. Using higher values may not improve quality significantly.
+### IBM Granite-4.0-H-Small
+
+IBM Granite-4.0-H-Small is a 32B parameter long-context instruct model.
+
+```json
+{
+ "input": {
+ "messages": [
+ {
+ "role": "system",
+ "content": "You are a helpful assistant. Please ensure responses are professional, accurate, and safe."
+ },
+ {
+ "role": "user",
+ "content": "What is Runpod?"
+ }
+ ],
+ "sampling_params": {
+ "max_tokens": 512,
+ "temperature": 0.7,
+ "seed": -1,
+ "top_k": -1,
+ "top_p": 1
+ }
+ }
+}
+```
+
+| Parameter | Type | Required | Default | Range | Description |
+|-----------|------|----------|---------|-------|-------------|
+| `messages` | array | Yes | - | - | Array of message objects with role and content. |
+| `sampling_params.max_tokens` | integer | No | 512 | - | Maximum number of tokens to generate. |
+| `sampling_params.temperature` | float | No | 0.7 | 0.0-1.0 | Controls randomness in generation. Lower values make output more deterministic. |
+| `sampling_params.seed` | integer | No | -1 | - | Seed for reproducible results. The default value (-1) will generate a random seed. |
+| `sampling_params.top_k` | integer | No | -1 | - | Restricts sampling to the top K most probable tokens. |
+| `sampling_params.top_p` | float | No | 1 | 0.0-1.0 | Nucleus sampling threshold. |
+
### Qwen3 32B AWQ
Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models.
@@ -394,6 +436,30 @@ Seedream 4.0 Edit provides advanced image editing capabilities with the same uni
| `enable_safety_checker` | boolean | No | true | Enable content safety checking. |
+### InfiniteTalk
+
+InfiniteTalk is an audio-driven video generation model that creates talking or singing videos from a single image and audio input.
+
+```json
+{
+ "input": {
+ "prompt": "a cartoon computer talking",
+ "image": "https://image.runpod.ai/assets/meigen-ai/poddy.jpg",
+ "audio": "https://image.runpod.ai/assets/meigen-ai/audio.wav",
+ "size": "480p",
+ "enable_safety_checker": true
+ }
+}
+```
+
+| Parameter | Type | Required | Default | Description |
+|-----------|------|----------|---------|-------------|
+| `prompt` | string | Yes | - | Text description of the desired video. |
+| `image` | string | Yes | - | URL of the source image to animate. |
+| `audio` | string | Yes | - | URL of the audio file to drive the animation. |
+| `size` | enum | Yes | "480p" | Output video resolution. Valid options are `480p` and `720p`. |
+| `enable_safety_checker` | boolean | No | true | Enable content safety checking. |
+
### Kling v2.1 I2V Pro
Kling 2.1 Pro generates videos from static images with additional control parameters.
@@ -445,6 +511,48 @@ Seedance 1.0 Pro is a high-performance video generation model with multi-shot st
| `size` | string | No | "1920x1080" | Video dimensions. |
| `image` | string | No | "" | Optional source image URL for image-to-video generation. |
+### SORA 2 I2V
+
+OpenAI's Sora 2 is a video and audio generation model.
+
+```json
+{
+ "input": {
+ "prompt": "Action: The mechs headlamps flicker a few times. It then slowly and laboriously pushes itself up with a damaged mechanical arm, sparks flying from gaps in its armor. Ambient Sound: Distant, continuous explosions (low rumbles); the sizzle and crackle of short-circuiting electricity from within the mech; heavy, grinding metallic sounds as the mech rises; faint, garbled static from a damaged comms unit.Character Dialogue: (Processed mechanical voice, weary but firm) `No retreat.`",
+ "image": "https://image.runpod.ai/assets/sora-2-i2v/example.jpeg",
+ "duration": 4
+ }
+}
+```
+
+| Parameter | Type | Required | Default | Description |
+|-----------|------|----------|---------|-------------|
+| `prompt` | string | Yes | - | Text description of the desired video, including action, ambient sound, and character dialogue. |
+| `image` | string | Yes | - | URL of the source image to animate. |
+| `duration` | integer | Yes | 4 | Video duration in seconds. Valid options: 4, 8, or 12. |
+
+### SORA 2 Pro I2V
+
+OpenAI's Sora 2 Pro is a professional-grade video and audio generation model.
+
+```json
+{
+ "input": {
+ "prompt": "Action: She opened her hands\n\nAmbient Sound: The soft crackling of the dying fire in the oven; a high-pitched, happy little ding sound from the timer; the warm, persistent sizzle of butter melting on a nearby stovetop.\n\nCharacter Dialogue: (Voice is high-pitched, bubbly, and enthusiastic) \"Welcome to my bakery!\"\n\n\n",
+ "image": "https://image.runpod.ai/assets/sora-2-pro-i2v/example.jpeg",
+ "size": "720p",
+ "duration": 4
+ }
+}
+```
+
+| Parameter | Type | Required | Default | Description |
+|-----------|------|----------|---------|-------------|
+| `prompt` | string | Yes | - | Text description of the desired video, including action, ambient sound, and character dialogue. |
+| `image` | string | Yes | - | URL of the source image to animate. |
+| `size` | string | No | "720p" | Output video resolution. |
+| `duration` | integer | Yes | 4 | Video duration in seconds. Valid options: 4, 8, or 12. |
+
### Whisper V3 Large
Whisper V3 Large is a state-of-the-art automatic speech recognition model that transcribes audio to text.
@@ -737,4 +845,4 @@ An open-source video generation model for creating 720p videos from text prompts
| `flow_shift` | integer | No | 5 | - | Controls the motion flow in the generated video. |
| `seed` | integer | No | -1 | - | Provide a seed for reproducible results. The default value (-1) will generate a random seed. |
| `enable_prompt_optimization` | boolean | No | false | - | Whether to automatically optimize the prompt. |
-| `enable_safety_checker` | boolean | No | true | - | Whether to run safety checks on the output. |
\ No newline at end of file
+| `enable_safety_checker` | boolean | No | true | - | Whether to run safety checks on the output. |
diff --git a/images/serverless-endpoint-id.png b/images/serverless-endpoint-id.png
new file mode 100644
index 0000000000000000000000000000000000000000..c3bbc94dd09d5555667cb4848fa47cc1e76f2a01
GIT binary patch
literal 268289
zcmeFZXIK=?)&+_vsDP*-pkzfP2}qWVk^~8oVMt0&GsI!Yhzb$}l$>*h8R8Izq#!xx
zIDq6ZLzJ9ubIyCdbDr`l1P2F?^#(q0W|=O?6bFY$!BSH4jhv(;{Tl~cGfNxb)T@9Q-
zkdlyJR=o3#L_+=s`}^OI{N-Qaej+h``KCBMhybB4Rg{)bA4qej_?;Gc(>jQpkd|1j$U(4mJ8J`N@NtN8*U
zcM36$3}M3Km&MnN#Qi-T9GM6oCJ`3jD)Vvdk37PZ(K&ISWySL6v@Mwv8EHKA^rfbc
zdhK{HaNgr|vJn=aem|y>Qgc^9=hYW(vzQ8|vGB$SZJOEJVP*cW{E=5A?2C(f3<6?f
z1HZaNidr#zb$r*2LpsZ~ouF+XPRB5R{nHDZ8}_KjOpj`&v!3ZwXO#IIN_PBot$FUh
zOOGIFyJ~z4R%(6HLe%Nbqu;#nd#C5HM2s%tX!_pR7ImapRwB#n?wWB3UaWn<9kqwV
ztrSdB5(QW6KZywQ7fe6Nqz|X!ddr~pli9fVz4nu^YsdM&i~F5)-YC?ccIzEVU`7KJ
zxH!}347)nM5UMN?xIdcNKfG=H10_wObn-srl^BNyJ$WiK{Tni2C33l0V+|1mQ`b(Y
zN^lc4Lr~>(ud<$*NxLon1vOJoD9!!&Mb7#GZH87iwMG
zes~VOkw5?QW5VhxPXF<71nbxg%g-u$qWSOc(#c^ay&tb_UgMticKP{P1IM`4M`P@K
zc~<$2A~kJT@w_Mtw{>n2E(NXBo3
z-VnHf{UjXz)3@ZIK?ol9jd+f@xTW~w^@$v3!KAnYC`w-_Gk~geQB@1O&0cGq(18&EBci!ee3Rr3S1Az-e2nj
z*PFjnkA8Pa@S)no^`gJ?>BFl#*TcR`EqgxETjG93)_HA5;x6-*jvth7i1~fsDFhh&
z;}@b9xJ*BE>~88^5%{iO=_d4@u=m!KPtT1HU++*hNc_G_`jezsHsLP+?8j@5e!U_J
z`;_tMf&VoXdY*5GuWpu*u*jCv54DQTKXV|ddSCVs_Vx2Crftf@PewUHc^@fLHJ4vS
z1UvLT<0VdfZBHwZua*7_JkivDGw@W|}0`JL!H
z0(bWBDBqdz4-F7~Od^+TqP+b?I_w}rMD?yEjU~m(^A*!EWZN@St+ew2HT6dS8Ks
zSD6Wj7R&;Va}L~lzxU!}>aUz119z-C`*M171}gKm!jw{V`vrdo{Kj=-ImJ2Uu6-Sz
zW|`SGG;iHKT+l00;QG2$HYk-(o=ufa%Q|3K2~G&=LCTG5j{A(i#Nv(XB9$%0&E66@
z;zvJx+i7lNobb+R`3|%I>LR2g#3)n^1wm0zBvjvt)G2#KbD4GsY;L5$_9Qg6QLlNj
zr>5^o#5g4n%KK(9wxIYXTjZc^S};%mQIj0FqvpFpKA<_!V>WEIW#((<
z-nkyH)-|@cvY6i$-&M`7@U$7!S(a0L&=qFXR^C~rXYJa(T^otiNE>>Y@0Pzb>^3a2
zhPy_&=FZPcWf;%GAIU!g4k)*ou$j13u7!CctW~X0O;Fvjms1;GJL>h(bK*wBjnKPc
zcOw}ff;qx=)Y~*x)U-m}G>tTbG;1`5A!YYtMI4%19U1BctArMwgB?-A^6u)(ISneo
z=?20Rf`kcd35B6q!NioJ=1z^vS{!>EkD`M)D)b*9jP)r{{MBAiM_72(5DJRYt;|Lh
z!s?rb-^mt77xlQ`5jhcEbm4XN`0cdQxih(O7#;mlt9(SMaLKr!#lP8xYG-t!Gk&AC
zsT%!xA$+xTEE_#C7`>spa@^W8sk~f)-gwXw72Y(~-Vmg;agpAz8-a5>;^Ys`;EMttHjQ?lJ0su`gDhcyyVe5d5=HSjK&U15-n-&n?$U3
zJLneMZTEW+3JyYB>Y$KYG}Ts^+8Z4=v28k-E8Z67}m+Us%
zp{Qj|*{&rT7x5N8eZ8+_7<(qtC~T}DW1akI;nCU^Tcjvh&gT`7p|LJ@%d-F6
zHyXt(MKW1vD2K3&X!(rx#@2^9QKk$gl#*G3$D~N^;d-upHX=Dh4DI#=(|K6>!dy9V
zEeyoduZQq6X9?%9p+mL$$n3bovet;*H@a(OE44ffcf3}vHeQdcpE#?_
zM#E2+8{dCs4>h7gdp>nJnGHB;iw?n4wp4zd94nIVwSH2Q%agOyPB(YH?DEiat15q4
z87@9jU*_t%$yr|yLbX6)HDR@@p0)GAXYn+)`cwS-QoAB9(Z}mM$FsMO=<>u$#FLIA
z4pzpNat&V@w%3h&fltWyum-je+rkLWx&2duw-T;pa@S{eiE%9CaGuRneG@W(iecFN
zR~05cfA?<4<1sO`znVgRHXsZ8Q6I01gF)lJCmYPp#`PskWS%q9P6}aC`#?
z?+OXdRp96f@QrhY5(n=e$2d6hSE&BySoI3?U-#hR;C!~kx%SsRAmH=j>nrdBl>Y1U
zYM39+b>J@&;OA2c?!WHFhooHn*D>Bj-8e7QB<19QPc>r)Q&Wheg{{+Q$Y23*;-=jz
z9Y-7-a>k3FD{`ujH*s)q$1K&gowOAdgp6&$9PdnQjZ8UQz;+kU!4Yv00uI5ZPVeYl
zz%~#^As5j{|F}a4IKDW|`H21>w>VjgKGIfvLoaFTU`o%=@r>iyBQXMcdU_ED6Eh)I
zDVe{D1AmD=vT$;;6XN8A!C)LP9u8Xvb51TnK|#)E+??Fp?7$uDj;;`=cP{J@M}~hs
zL@pjC^v2S~)J9v%5)5=T&@?eFeqMf&f7J7TF8#Zu
z|17HEXzCzo3kFI$iT!)9{wn;R7yoy`f7Ge-A9eC@bN^?Z|8eO*ie9{gkgBPptqt^|
z6Ez@~PGUf@%TE9AQab-GCdS3f`&Xg=Jo~>(XkS|5KhOT}63PyiKu5g0Xqp(;Up4&a
zxxelg;k+39|1gGsTG~I(0;VQLAj0`SW-CTOrj;mygCl_>C-p+zSfpg`Lzm(s9zjYn=PY=0x`qziJ*C5>U!72Cut~wkas*pd|0#r{g
zdF{reNh{S0uD@*$t^oDbzpF>W@5Y_Vwyy#ojQ-}`aex*-$N#%}e2DJS+l4>0z1H-1
zoddMkj_7aedGJVL=85FgjXel>URUH!X0q_=D2`tyJT
z^_ckM_{0;rGBD%+-GEBW_!<9=MO?endgltRz=Lg>H$;Cl8{h(9I4yto(*HG_zZ`Y88#Izz7$9q_0t^59H
zu39$1BHM;?o>t*H*0|B{7W_q{jPp81XQJLac|z+H%f3-9dwBO!FPU-!@aKW=qe7&`
zc4Du?jax!eHAHtRvcYZcL6V1=-?*#>gZUwomAzTyPqp82)@DQMY-_Q^RQyib{ovD$^Rw<@xNfu)?UO}9rvGY)!{C=R)?AI
zUh#sCD8*QjK7R0cj9r~VVsEPC_3D+y2?De-Uhb4mSuZ(9g6+;yJ(z++uXRwL)wxRS
zcxDOiS?xTW4>PpYuXe_kp{t!;dT!H**@T`hA2&_mvt9mnH9&k|Nr&Zmeh$HljQH_L
zA&!^pphP`c06flHq9E*yc{=29O&}^q;fE_k*ePf5sz4;0R{8OI^)w~F4fRv`WmoJt
z^rfrGPl7L$p6TfG?=5gsNdD-}n+ALqy}j#o$C6Xi?el$i-IDp*YmVkh<$F6J
z!s{j0h?OwaR7wGe(VQ)+>3E{tGyinAjmc&Y6YIP++Y}MS$;oqzns;;^wKf$q%WpkA
z`8!*xBmF+5!DX{o4-Xh8#0P<@w2j%X4;u@V5>d{-m%ph_4ApraQ&5(%AR8hEZVe&y
zC+dDL6prtXgtBhaI$=h~yiVo}M|XO}t(CKp^`Qu}ND^l#uGX!S2Ys)3@c
zK`k!GiDGU~XTn1tXYBaI`N^4b6M@;if-+Uc>V3)mvANLm3ISTB7hd=xZCzbR!}CSh
z%9wJ6c%l%rc)D^IiWz0Ox`#NBKA2|`UufpGQavFn#=kU5vVH|(F~M&$gRW|B$YfV$
z`7w}z@Y~INRh=-9p7?rFw_i(_ss7IU>_kx53!^Q6jYjX;+(f9?be?9{7r#$cRqR-S
z^IdoAfh@560?acdlh^R0Uogg8{5%r6F~yL1$OFw%N)(#FBG#V&0SW#jFp;@sZWHU`
z`<;w%9tk#26owUR9FZUU9KY4DKY09~Fn6I8|6zXy8KAd4$+M1Ri>fSu%
z<&IEHI?sRqMRi1F8qm75gy^<7(!nd+>?aPLNxN0zmDg3b>3mG(5Zm*
z{fMRHvvSkV_FVs=Jn@a;Gbw~YjcNJ0%QD+--AwsO;jg(HcNbn>0$7-Z0hnTH-V-(O
zG(}S&^KNBod|6t({r>d%)~6TIeAb%dpF+V`(4j95%Pj(Bc^j`wI7Du4-q=f1d7KQ4Z@m5}({Uc@ZP
z7y<1=L9J`+R|c{{(FioA!|*5%;A=aYn6TB&-G}S2Ix=yVIm2n_e(g<9?z+u;cP_(o
zJ|yHb&PyctX1k0GMG$mcZ=mcKYE#M^b&m)5Mw~)0M^;UNri@i%1jUeGX{R+XOlrNO
z#?^LAtA0DPHAQg{H;A0Q$oq6M8{D!7Rtd)@xvvLzat3#bc^-{nD`a6;vLu3W*}V@t
zNK;c3)yFiwiW=&+BROL{()v?gNZNG9*sFwd^B21AknerJntUU2zSw-9
z<9JN@ysD=b8#=TYsLjkBDd9vHysxfvP=HJ5+IRcx4D$hG}=smX$6+2Rl
z7xy-p|1A~Eq?}Z?;k_{dsal@ts--*fUvnr>O>5%eI1e)PXvNdE_G{mz3qAe5M#;(R
zb2&mqa08a1HAZEi;tcC+z;otEOY=Ta{PkTPwI|5N)*h;+l_pn<>+&pf
zdlTzlCj05m4(hWw8Un-0ZKolmQ!eZdeZ{7|B+}^U-3@X25|LF}@YwNdhHSE^TzfCC
zv(n^IpVQ-)rR9)BbE^e4+Ws6B0qtq8cgYpgZanYu)3%EYBko>0{!RpUD&?}Dtg-U>
zsAVgS32$%^feT+pL#Dp3r{1(|z0Z$v%hzBNF3S>hG}9f*BcdStaML6wzKobg7|&rl
znO4|&aIscYAi|sRL1E3d(L9D-7a7>e+16QX+)_Qv`D-8#<>Y>}Qh(2SKsJQB{QRiU
zP#Q6Zs7}{Cl^+~&=(C7qxd-Rl
zV?8>xZo87wN8#4?NbmErH)0Pi%?kD4xMU@<-*0;zBoAlUhxfW4?Ju2~@nk9ATA<)#
z5Eh>BIx1~}%P+8|YMcxPj}=4EABno@+}Gh
zI&zc1{BnSs_e|MWnb}qkKU(TZu3*(@Xpyj3Bodb
z{6VP5w!ZaH9@K)YuDwOYe9FmH~
zjFnFC+i26ZGckDLTG^++See!-YsJ6)fU?v`W9Bxk$e>wI(t~ir1d3H$HrB3K(T)y;
zW-~h^irssf*6v6A!C;Lvq5EurxSQtIK;Sm$O2NBESt$;bNn7=|!}h@giM5f>S7WIJ
zY+hA%Mk-X=of%o|$w#t{H3j`T>HT8ao}*Q0!0ex+!S)=w-1~zU++;J97x`^%wI8V~
zgAW;0oQkG(d#=9dxYRv1SAU|Di$DC*m;Tb^*oWv|&R3s!7PTj1p8KM2Aw>cAw;}q~
zx}|1a;1NV3*P*m0S345pFF%l_l*P?W@Fdt(BX6<7E|$+-hS@H?E2p~52@f)8WvuX1
zV*j(gX$?wFO_eaEMB#OGjky)Hzszzc#>m8UsOVjzU%VLU%Bysl={r?%3msDx*og{R
zCABQjla;~@)sOy=24Xrfy*!`zX2rDXk81P+iI)2a{crhei?{-@G3)i`#>3nKB2&E=
zKqsCb-1+m7+LKG6L%Vv)%Mklg;}&P#qqDGQ1%&foD9QAWMhFq_lS@0Hdl>PS|N@?v9K
z`4YcDreGQhwBmSy&H$mNYNq^{cf?@-YoEr0j4kPYE<4YqZvCeMX$wDd)tEZUf)EcS
z{i=#>8jOk9T{-Ld)N5qrYQ>x}7V~+ItBX>L6nv!sCjAzq2}SQBZN|7AKsqLx4nEtQ
z)g1kU@OK_ZLv!LV<2Q)`J&SyoGbqn}N(#XosTIGv1k5J_NLwVWZCCI$Hb#rpn9t>T
zNTA{TuBu~5k;;Bg=AIp`T!*SOvHKD?*9Yzka0H_S*_8S#8AtR5o|h3oy^%&u+aEmY
z*UCS9;dUwMfTki%R@_tB+5Gt&Oty667=;#x7}2@@n#!pkA|(7iTopijyq=Gnb~S%z
z;w5MLWR*k_zSAEY1}}6@6k3^7FCK>ObX@={(U)s6t0=54iZ@N##6sy0j$H^ZP%A(2
zhnZCXm-}#MD@h2gehngh9BFOq>WY22rkP@~J}C%h26{%f)_-D0k)PHW|@
zu>%RNwQOoFA&HXK=ta~lsLn%&x&ATF9(H1>xm<{DTq-1T-(`>`&u5fLvllb#
zihDW
z(Rvu6TDpGwNtLJDXjDobB!ymS8aBsh(%a=ZA0_
zQMbnt%6}m1^vH91ec~h8#Eam%{DxB#uX|QUY%M&Ge&1Ti4oi)q4FGXIrqp7J+~w|x
zym?J`(xQz;jY%5u_B-@rfq~TbhYpLCogDfo(Auyo5Xa~|iY5VALmN&<^tefZd4a;b
zJE1t1C;g*#(}QW2UtETMa4_vr`*~>&q>*>?^vi<*c^xrt#~_M&(kit5Tp)$SZ>^)$
z+h}uk(EJgnWGK=)+EdH&>UHAsca4bKky!l^DEi9-HOWJh%Q8dy_YwsXzQmIjT1+j-
zFPkv&J{Ov^-%_EX?t2bbYXyiEr%Fu2S>8yo()b)mwE69>*0Fc>gpRcWV5>%#)we*WH0u^6?_=J46dvGJ9WGLTNHw;*b>@3h
z5Cw)p{5&9c_q9&OP1_pssT7(O-VH<_ju#nrC%cg8bn3IW_U0wnaZ6c1nu`a
zT{HCCZo4#7$MGqW@Vz~$5HHlEF`Gn)J?wqh2$u<_Og(Af27-d97%?175YBeE0$@i`
z2L@wOCm6ZlPNo746@yF&L6^h!{LOeCVK+=hZ>MQQu~`@EM46?e#hPLqFOrtyy(_VdrK&t-qrwoI)YIer&)ctsB+=!b>xT7rs$wY&Ph@jL!K6?9KUBo?`(gQdqU%qJ
zRz#WQRS8QAy>_iNksxG7ke2eCwve1a#M@aW@x$C!%{R)TyBXMVB+n036NXDENr^Wp
z4c8$6Tk{sk@@}k`?u&G2bOIRQX+-kVbP#ft-)VWX_bCC9BJ|25)iHw31i?lpb8io;
z*ooD^tK_Szxn^CUt^94&Lt*HuISL=~LEVg;x6(jC#1|+=<*-Y&=!v(RYn5FuXZp2z$Lw_47
z6G43D&n(#L;25NS4$*rwQfO$+S-0u&$AIHcp0nvgg>_CFewJ=q`4GhTVtwX^WPv7p
zb7q==h(^fi8PFNgQSU*Goe6hE>pDj5`_M^@th#Aja!-_6F_8X~(^xUjz(ipRh$X+x
z=#6}XT9mkMYA9f6WmC0-+Ep6Fd%(35qHBdefu~?2O#sh<3w2uW{ihX
z=TpS+soxq0xM_QzVF^4`5_!5Pa6b2;S&|{U#XQ6TJ#o73eP)41?fNI*J3RVyOcWuV
zr2t=b9CPAYnHn
zD}IuM9W55+g+Jxy(5rkKD!O-7&5n(AMf9K>=529)?5#YiD1bI%*MHV>JIj(N&24ve
z;uhj{bIatJ*XcCfO}wDnG(vC1&wA*j;p6RjO05Rj3{ztNXwuEX|4JBkF)V7
zt_G=c-gC9V?YBo|EYAC)bRlE8tQvW7tXYO$hfOgjITX;fMeXymUj()qX{^tr+KH}c
zkmOru9mvlY6j1V84O+Nkky?c8%cI48aK|T#PYti2+7N7|KOi1)2Y^tBYj1c`V6K$G
zkZBd9xW)m`NNL`$+mBgTKn6|Dp93u7=xo!5loDHhr|0qBxU%;)-CiTzS$35f^S368
zfh@S^VXAjWvh!9N#d2FWBGHLDo|GwMLZE(&Z{U!bHBscjI_op5I;TZG+o=Xxml@Jj
zjdcK*pPej;|Jeo2<7!C@p1TWTH--f1A=0XKiiO+8WQ)-JP1r4gwVSO0nCDn1AxxLzJ^
zJ)l(1l5A1h8Y_3kvQAXmO9J-IJ%~5L5b7aNf)`P~znafE7Q|_`c))VOL2P)RtEf?U
z`97hTY_n$v=V|6g*=a(r?kCFDFW`o+k<@uJwZHum&*`=GFkdCzENAlk)G)N#lk-$U
zHHZH4K=G#g8Y1p{*J4+^&F*SsYDKnMHV4Zsa4la%wt7Z
zglyaA@qro{7wq+Z
z-7(nx{FHIZ&5G$p4dZGOT$@U{q^NmjWx`g%g1A;-XX$hGDq+{FKg9A+7cUjOC!*q8
z+8q8WYLM*B2m&oxT41p2!$L3B{?uo#C3J#X%bq%tz#7WqjOkJd8vAs!PhJ
z^X$0Zo2H%V(md3uM_p-O!)%_
zF3bLFS}nBtZL}acd@}{-h1^4QiL@*mHO
zdkQ(M09odEMcpA(?RCUy?)!+B
zp%onZ)xOT&b6LTuHDT?`L-`fK4|v4NjxLhcfQ2$B$ilq!QiPN8g>p@HUEe8MqzH3iWJ_vdZpwoIlmJy^N_(U%V
zuM@Mnqwx$;=>4nap@hs&^^Xh!XxYctDe8Z}ndj&2?x^=ZoyPR0e*8Ajp_>Kjvw^{@
zrd-XxjgTcbe;Pwto2R~r2O?@^ah#B!Gt7|%$jyzguWyS|kJoW0d2B0d=4-Ndba!)X
z8=eG)!}G0%>{9r=Tn#QoHgBe`AyA;g%-7K_k?FviL8xot;P`9>B*pR8GOdD2?2-7QO3bojjvI7eZYQtZ9BY6zNopptgWa`WrMEuC`5C1Sv7Z3`~S
zS(vlhHTR{lPrn*0>Gx^heYyA7FMVio
ze3*^vV3i|<0Ho_SJkJ_Sn#?{TR`J(y!iw;Ovh^2~KTuY^o)-
zh*ZT9??}yoiKv`Ci5iP4_YFw*H>RxtaOxjhSX0+K${_4)bDKv
zsy5B)Dp}TU)K$0gdUb0=v+6_LJBcr|Bw%H0C0FsO`I~CkM)`i!Q{qUs-jvZ{zO6#!
zMV3*+k^Yh}{M$F&j%dzOAY1#~K3rkwJ>m)a1*pi+b^2Ka=!n
z&e=-b9xvy?qdFPv5}cnO+%Y5nIEFNUa-Ii<4CgRlIU79iG*~3Y6EXutr$8mWIaR};(r${@9dlgn
zvL&2q8Jn;*HM_jt-W)<@h%w3^>IOKvJD{GiQga(16;D0MfAkU2j|i-YO_ch-D?A8(
zpFFJz;^Sjs*m`_cF;9v0$X;oUER+TbL4yjptBkX)jj~px)xR_Gl0JCA_zXOL?|4bs
zvzfPC@!2g}?!m<__YORWkuJ!R7TF4<>*jWQ4(A@zb+rc3pwlCLic(I3$(qduM+1Y1
zrUD%vOlK?)SI515U;KpgzECd`(bihRxYOAs@nQfGlvc*G>@GkjJsv-*?5{tEt|#=j
z0_26(JU7~9?z#0C!2G)6%_}85r)}rV{URMiK3bu_4`uu|XlkG?Usk)jHX_rdM0+Wz
zv;dmPyEVDYvHO50YYbk2zPp2;$9>+$bUu%4@Q1njRYh;9H4W8SWnU^DF7qc6HyTkIe3X?{pK|zyIX9akn
zE5w?bK-j1|Kcxeyzg4F?#oBtSWufnC=Bc*@s|+-BC)e9G{ir7Eh@Jp^#l`Y1o|hZI
zxoI39jnuE4opo{O-KcO2j1`=f?n$b}yf!?Nw`d2J!Z5k@XMz#Q@_r$wW?k_UA=n8_
z1l4Qip)&KHbrNf64o*ZnnF7IcV4+7mrpfVA%9FM1yT*h@+|*E;+7rL2$(kDHEPDY&
z(EV_u#V;#a?{DXx+*EL3nA8U;)gCzm?Jr#&+G#;;?aTWlx!P)V;-F6sZQXtFybX7?
zb5%{%g1)XHsP1*{CD~5Z^a1Zo*3N0=hdMrL7NA8Y)i(U#hVQhApZ6KKE;OYf>P9<)
z9|%vKZ8hZYWius}Go5b>98Z*T#(;ih#yNQ%9G5D~b|XB`mb}jt=F3cC!Q=B)#a><}
z7i%6}d)pnH<7cPGl80^FgWIy~x(045&|~z**;~$f?jOAT${s~5V?r^0m!k|uCOeJm
zkYR)lIG!JJVU
zD_Znqa|G^;6~dda?oCM*JxZh-EqXUXF;BKPzYuNcpr3{fPQXq9B#+@igCvBiSf5v>
zOD|KNta^8N0+TK7E;5*zzNqzntEyg{6lu{Q0#(@$GCXU>uN*exFk0%${*Ly)n~rS}
zKdjP#yAAO-C|ERKyL7nab!YMlGvnt!aNQO=@V&q9?Eq^U$x-2XVms%z3*;VgF*RW>
zeZcy1ulCcG(e60Dd>dn_udI>K3Y#Iom(;RE>J(Ze0mwXJd}fCNUC9!zw{M<&_WPr6
zl>O{ayS0Ru6s)p53id5EAMX9gu!r1ZzL9mYmQZFH{!ypN?QBKaoBx-+6OhE)+T?y?
zkfpGu+9FJTZr?!uG=?IX#tkkJ*!)61qVAySa}5_vb!AZj4mmscJUuaTe)_F-DA&FD
zU1Q-s%0zG5;7nWo%xy?99(1BVP?hhA@ln>!9|k`fN!?%kEPfEOR=1^nv8qXT*eZ)Y
zikDS_OxLBO9Xh~Snj()c?VCHy!an(8c09&M>V@3mF4?Xgy)w4SmdNr4-UIAe^-fG0
zk16=<2GV6YM{sksP(db?^B!BXu>QI!DYgV5(=g9%(&RvlL%X5bQ2TA0okhj74uZo<
z2wI5Zv5ql7eXx$!AFR*vzWepz(0REx^@D5Rj>*YEydqqYBr_qw?dDaQfG9ae?}=e{
z0DB2JE#D^?#vdt!?1GS1iQC1gnu>Af>HFRym2n@Ib|;VOP8L4bjx2R2A+f^cu^+hD
zb<0bD+!}9((QP~4@Vp_4LqE*Iq2Xj5-N=6PBagL^PkhRao=bH;%$`2^sC-^~Ix&w?
z4NP64x&c{!oP+{^@1_TmOZcq13$R>A7%h11VX%b#vISfc49uqq;
zg5DU9i0pJ40y3I|0ytaqeVvZzTH9pzO*KH}^F=CNp5@%ON5qr>M>-SWHc*g$1OGUw
z%j4ILh%`X``x)M@5$sJ8NsO5`->4~lW<}z<@MdPVDd=W(Aqs?&+nFvm(N9&B+jQNG
zhWm$lJF^=bc&*Sa#!cPBOdaklVG6xH8w_5N46?)swvh&N?ly6XSxlcapO*rVx(er4
zz+=ZHW!3i!!2$+IPVW`Rz0F3$5_C0;0;TYdhxz#q+}0EAj2WfnS@XNKhS8lqiK7uF
zrx^sc4O3h`(r?mmBPn?;?5d(lqy(DU0h9FUq_6;}M<~W67*NUpJNk4?kN24%!CHg0
zxc~tO#SQ~0mzB*eD*iIR^9UX;llLsv*yfRAx2J(}bh+9lQ8|!j2Q=
z;C#4l#qh+mdXy>cA11pFAXQ*o>lnXuO8k|i`9$C?v$C7Anq2~t{;k@O*91hwZAT@}
zm)G>qXVIGi^|7dNF
ztVgE&h-e^ylB%nwLFLh(+>H+#;6)spOrE}%SiDgNK{p@m{(~y&PI#$8^2b7t2-VrQ
z!=C=b9lqhf8&u7ASda(6+CxBTVOTYJo#0Z^QbPK|*Wz#3I74XihaL<6it0%gv&|+>
z_)5SBtQU4
zO#S>Vplz<4-7d5EW(~x=5vg;O-@E>*1x!#zZmikuf};}3LB#T!GL8mllL{@lCagh
zg|PRpnyxs(#tQ{WzeA@wKW;E3-*)$}@gsn^4{|!=hz~0rV)k+JHpi<-U5o4!n6$u-
zmsbF&YYbSP`xrj3r+1vPQ`UbM2Daav`B4V1eyllIYRpJU!hWn)?o=m}%(2q#l=^w)d
z`f~`u>=p&(U1@2N&ssk{<#$g_#DF2E5643Uepl2)Qw|%sSX+C5c{(h0drBw=x2M7f
zt}BiH3jS=f;a+XilEChK)S9d6TDaW|Eh?`tP+KnNND#GkpF*2bO>2XXqp>VbXWP7z
zkp-;IXqb$P^c25iI{V?!hi&x}ZZ>N`tqcCWOa+HqbE&|skRKd=FHa5phOJNPOq98X
zE5-|r0c5iNGts{AMRw5W=DQI<=lAI&Pyd$bMIZD0?P!NoNe<4`_`RO9!
zzPDapW!t^=K|1w=lu1VEB*1wV1IsTu+15h3jQgB1-HGDn8=U$^gb9tYJmw$5NyK^v
zPk$2@tnZ#=5NN3yUD3F!{#JdY@!q+-b)D8m!<`do1E&seyAQ6BLwUK
zxP2j1WJD=yhI}7-^mTphh`9o21gH@|xJC{T&Xs8ydJ0;V8!c_=>)JH*65N~FSoh*#
z+g%--(1f3S$D4BYTYMd*a<}IfnSF9t?Pmo^08zAuOZd{wEyf+Yj|Y2KHKIIGp?oV7
zY4jH90&-WV7cAoz`)Xn%Shezvy6_jrQ|P{wA+9guFPVx7?9n%&8(r#JJ
zNBOTb^cf{S@6llELcKZxT0EgF0;P*tuU-$nIi4Vx{#MzmBN_u5<9C?9#gW4Wka4Q9
zqxI)FR6J%L3cXLfbq`+)SWXVE*X%%G`)VyOwW7Nn@^{++GGP=^mG$~X@8V>Yqw|@E
z*XxQrjXY4Y*r}druYPy==5tuqy7?U>bWLPq;l)T=4l_L9!&+Wz51SUkMgZ-UPxy$4
z^`J(-(&a}tY9=SysbcE0m`CygKxCG>JWZ1*>JUTwD80n-38Q$m1k+6Gr+r>D?X<4m
zzn>3!YJio4$A(gi>3uJP88)E=^WV0(e2`!v%s57gsch|It347)Jd!NDws~YF;9RHE
zuBfa~+r-%L@VQ-ul&PCS(7?LB@<&OHY_npf-72zj0A%x-y?M`oRHtS)dJIqy4k4}S
zC@loBR2h1HA^XnAC!TR*?x8&u#=4;yHG1}DdRr|`&K6x~F&cvw+8wLqaNFHNON(H)
zJmF&b;q|?4XNQfO69-3eF|Mxm&>K0rWuYGP>6*YU0^835;HhQgwLbm0JQ1&jNQzob
zE?~L6WxUMNS`g}3L$FXUeym^uQ_8cb!%kzlU5Rt+mHI2lw<;9s7t8rhvN{X$
zkzgQ8$5bsPAw}JtWg%>d3z|(e>)nx&{Za7vT4ARlXnb5BAg*Y@L&&(papDAekboR`
zf(p>R=x~^tt$N`T7Mxn4QOIbdX-C$~h5C?X=YG92O+;{{fb2A(hyvF^4K5<&7
zZBJI|!3~*&S_G~OAXq0etr)$1tzCg0J+j;8
zY!YJKS7F>jOo8*Yn~S=0ER?E#KyhX6vZ-sL9rV>%21tXa3NKRB*LIfqx?fCg+H=Ev(xP15Z}|K*l^@
z#fGD#Corg2YWpu&A4$pS+Y!-%JLFgt*%q8f5X!fsx94W$EL%#jveEEdypPn9#Ym;2L7Irh^SeR|BoL3PJA?NgphA>`zgF4ee
z{TBVee}y#Z(HrWE2uPtB^1+&1Kqhe7ExpJzCP@&Nk`&k|0Sg?$E&E?nGL2BZ^?NXA
z++?Uw$ApYMA8PUvbUrITA!Z(~712JF4;on8zHZRWXN3Gm!NQM^31No%9w#LB`Br1PdvHVh@V^r5f7hg_o-M
zunLRXK2Kc%Tz4VV)^Q3is~kK>JncevzKa91QFmbf9%0rUlVsPs{9quP++%pq>S^Tq
z?-TIm1u%<}%wCj7$-e%LZ$Sf!2GvDC
zg^G*{7l?ay0#zy$L%_;!=#kBqqFJ?k>*!7^{#-08E$4V2yDYXf>tp#!Fw%L^;@8I^
z>mflJO6@hO2^|30bwEH97cd}uFR?6fr=p6!-%QxOVDrUz{!pG77(fa#Bm77Z;wfmc
zcl>Jk5<0QmT_sJlb-H*xdBu3FYt4l->cq__QOt?e`{jpoJ|&fjcwslSiu$Ut@~+8<
zx82*Z@RWdNgvxl8TG#eYi=kYZfED}pk#(Uq9CXOemZzROyUVuGGhMD<{XyOmjm;&$
z77&KZN}y>$vl@m)!=~kUtxddyD-^uiH$vDQvR=o&)_>b_`)N+V(K@=?$V}P1imED#
z$&11g)p-6hfRt;3pl-C-xEecFJ;6&H+Wc8%p=P~|FiQd5^Y+1R4Rr0~dly9DM)ve7
zix+JzWDZr<2L5IZXy*8W;TZ@BFMb=Kc(8$2R~J2EIz8B*?rD@Ic4wa;!4?vUYZp(h
ztrupkX}&?Pfzc?&KFqhnu`wjJnX1KQAd1t_Y;710Wv`BNPp$yyM9q<1q^Rk()JW4C
zul;ub>&0_x5B~dTt-5}ZsIg$ycnn2bbgXW$D!%@~_
z6&;>YLu6l3a;g+$DlC;r02$_sA4oHYVl7xeX#yZCx-0p-d|wbV4zGqfvHw~fhOZ5~
zPg*lpfG0Eyci#>2&+f!kSiBQ;`|WU*NWcmGa@Y(MM>^9Iq$XO(ILy-aLDc
z{k<}PS?(LGeuEv#hrQ5pcU-Dl5fhI-ZoC~O$xalH>WS@tNg<^?Xxc9^BCp!w3y9y<
z9<^lWsbz;$hzk*$rfqy+SicWwHG@`012>3&>W*ce6NFsfuUk8o)c17)c(Vm0*!p7Y
z+GgrU1Hzv(+`ghCT+q{921R{BR=7cJyC&+z`?>JNwcgZpf#H~=sKTU3)6aVO{Izr)
z(e)e1VSmGBZ~Z|BYof|>$aEa9rCBg~3y>8Z?{Dlt%)V-*&7r3uuHL_U>?sG7^fRBl
z_VpBC(j86)c7JxQ^DItj2_QNOVh2t!z3Cy<0CYvoArkj#^qV}VR~;+N_aXaLdJau6
z*?eJ5&t@o(f8D-v6^A`BN*F%0t+iCHt}5qAchrJtzvInuJLz_iqk*F7`H4e9_3yC}
zZtNs94Rtq6H%eH57P|Be&|SnMZtm*HMa=dL9Es#39SoAhtQBi|cH(WjZ?YRCIBMCM
z3VHD^S9a@K-OP8Kg1|5~_GrRF-RN@~w+{9S*J-*5^zp`Y!h9ja)T5!D6j0Ira$jS$
z%%E6&wc+Y};pF9NSX&sQ^Xj0SH|b3}>neLjU@z#hK5S+Db<3F-3=QmyF<2_EN%8ee
zBU={IpR->U9oQZl8$&=huWXDfPXE#pI&e$Fnpk%dIlQ-
zAp86;0D!k|X=EALT@Zqa);6Ew_rZ1}Gt#`oPQmx~wRc8HJ#2my02SpSuMNw2+TD$fPtvzE##4M^qi?&wn+A}sGL5iZNJ%Wf>5ro8w*q*!3
z?>y)CKS!LC`*VNZ*ZaC&*W;bZ<6!FH9T->w={7Os(%>2j_F45MWkvJI-fdLdM&@;D
zB={mmx_swVoAzm1Vvf)d+%jr)xWTOR8-`6@S2t}}&2u9@BWXt|PhU9w15Gi56p}hN
z2*(G97`~`*9%jUj;u$?Sj;Xv&^bIJ79u+#`2fYw|?ZHrLuB#Ea`mRR19Ldofer4^>
z9tAWha<
zH}n;cb@QKX&rHN;mr3*daD|IkRCYtrb@nL-;hL{>t*gDn$j1`F+Th79
z^r}5VG7;s6<5JdaA*(f|m=nqy%9^Uq+I!LTncs{2s6Nh1sN31|Bt~P-Ft}_&C<{$)
z4@zs@eW@x{BRh#PW4n`w%!q{0rm0%XD*Nu?hetC8{M{>RrI96C*lQ10`tF(O6YaA{
zs_aX`)yY`W$yc#I9|N=R4l{E)UpMkpWe}U=6!9-&(F(OF-o`$<)%9Jx`{-ypiKUPK
zuPCI!$M(V)lkC;Kpd2#BJdEe)K4iMF;Xu!GLN<;rRWB7l_4e6P#(hGm4?hNc&fdz!
zo9OdfNhS;*V-#yq=Ex9)wh!X#ccbauo`lX9M-_d=8AJ{+HL&VezJobQ=^1x{x;n>{
z33q*wd2N)ectRr9D{p2JnB9ynSsH~9yN$MI1Q}Ex#w2_vfc^wt-mvCDiZcX5DIax@
z10LQliJA49hirCO{0Gnq-`t^g>KEwnOvMhF+!p0BJU^TCf}`4fGXVGJa!y3agW5!N
zzJVio7?>mH_Z;s;Tx_8|k1iPP4t9`^4f#T%0!W!}bN3{)0KUqX4j4rl7u$n7(Pr<*Vizw
zLPXKV6qx%6DJO=;vmCK4wBo7;XMgR((-YihW&{}
zos0cM*-)`V_y)P=BE!p$SWMo`A-UY7^C{$WqE%!PXdMXe87;xhdSpCLPvAT-!zI5Y
zOp}QI1R~aQxcxvp^|ABQ)|+7U(EmL!Isfm#q|p^7a;z@~?HW+9w(3^!MnGAs(9exPhk!$KL2{MrlVGSG*#Wp2F$*fAXj1S}8m
zZNd3Ts9yv(teB@OtU3&JMF+e0JYmVUb9x~hzlUWFfv*rpV5ci$rW+0r$mtpm%H7!K
z*sG{!)bgT*ZD~!$KIIVc(@%n~_`rRI9k$hG9!h%@0B6EorkZ45=YAdA-%l;Qm$E(V~
zC^zVQ4LqWFkMyf$)3=pWo3I;OOSN=x;4IuxNe=pfMLzra~(6P48wbsfdn{>o$
zj@bB<<4M|3HH6PRx$7)bQCUgh)~Jgy`t@LSXB)>(LmSSN`oR*bR92m@~k=!qbhua`X`U=(O}o~w&{9#
znx$ykKA77gf!oOa7xDuRNAQI4d$1CMOMmu4W38xlan^~pkPG|>QfmW?m1ka^Yy0qV
zpEzeN|FI4vy|Sxpudw8ipGl$Eiv>Z}DdDeq?jT;P&|fibT6{Owk$M??MtVgo0Y^87
z83<2c-^wWiK#Xpr)d&JwhLjtE2v8DJ?JRhv6Qz0JzGL+LbBz>dy97
z2~=irLa8c+I1dVGHnfuU6a$Y$W`>;VwpK9+X4G%6kWHYy=*{tF<9;~KV&kL$80LIn
z?fe_&Sv8R14x9vo6#MufKCq?V?r?v5-k7Y<@YB4bP*kn<@YQ{d1oPrUG`TEr4G7(xMh`+W9+wfqPX3TKG^6sY$;ZM_^9(PESz{N`|8QpvV1G2C*GGp6
zqglKQC1>k!)Cwa}lMbp)ce$)bAOBVL%mf)YyllwL%E=7hUc{*$q918vIJEalg(pwD
ztaQWq_n*PB5=W}$1XmO13sjuJX=SkzpF|L-6U8i0d(XVe&g&4wY}NAuVdvY2^+G5#
zbG(WwuMjr9IrnMo!Tx(5mT(zBM2LRu=i5l|p}x>UjK;G+pIQu0LTOs!u=MCNdVnz#
zmcR*Zdadz$v9Z*!2hq)k&{fCVQ9L_i`##8*LAl&;b!$yg-_X>(uZnvl<-6|5ONiVX
z$4L}wsqdvF@zo8=jd_!h<*CZy+6#(gbLkbIBA>4@3=Hznd%|DS?k!)$oCw_ZyYT4wL(wSy6IKuWJ=Zk&{xgw_`s6yn~pYb_%*YJ
zB&;PEk+HA{kmjhjcz=2*^e_1a{BWrLYm4M_MO|8ucQo10lw@>LKdRKNg)>skj!{lW
zzli0VXZW21+Qp&g*Y9%qcCn$XeF>?7Q*uE$yb2*x-01FLBVRJNmRKYZyDMm-j=fGE
zZ6yH#2d9?ZNK_qRt~u1VEglqzhLS1T!|A)AB#^wQjRpxCEFZ+aJ4*i;GA!H`()J!Y
zU1`=v%l+AH5uES)1&HuJvJP10kE}a$k3y=h?=+X>!k1=uLynATWpcM6lJTFKaYwnS
z5hKj1VrdcTvavNT_{p9j+LOcRP;P}DnwDBCNzwhqux2l{kUCk1#P{XX6|Ul1mUiK`zbSq`jIseP)pt@@~x5?+{H66uc?Q-
zh{Li!XAmyjey%oWo290T>mPWdLb40UWp3#7DV{)~_h$q{1{?@g;JSqk20xe3RWJ)L
zWyK~P{5jR7wj@x@G4aIOji?JO3C6ap54g^`k6{|SU5P#$t;E)mK7{L8jBKmC^1v`5
zZ(zRYNXU))9c=aU1;V%HZ{=F$!J0(w&jkL8N%`QMOzsxtV)(h+7EPIz7o;nyGDnV{-?!fkVeBbFQxbez_@P4I&PW}zB}luP;BxCV_5nMk
zKP<3F@+s`&awk@XZd;Tscrrc^!u!~NFIcxGQ~o^=`fx${^iebJ4>m>;VQZ6kJVQ{2
zRS1#6WaGPavgNpBdwLoFP!nFxp4iPX?XT1FON#}Y@Po2$V4sfa4LDe9P)CD?z!Amw
z0%jvgwr8Y@AM01-dF9--Vn{(nnANQw-R1yydL?uxxc$q|KnwM4f9H)Up>UF*e_pKq
z@SoA3i42(7+wg=>o3BhbBbeB>%&F`}Um>@kIY>N%BF9aaMU;A1cQ-Rb_fh|XI@f)J
z^`|hC9*s_e`+KlJkVnKpC7%|YQoY3F-dAr+iVS^BV)z@rRXB8D3==|uxIUoJ%d|Gx
z!98ozbVgaY51`~y{C^nc{%U{8^>Wi<&D|GlruAc8ZAtBH`tayV7u)2Lj%T{9VFOVJ
zUl|&EFm3Qzj5l-*SFYpV2?r64pXhX=zJnsRrw<4Zx`&^%GpB*pO?r$fgAv(AI6RPd&=e4m+5h#lK!+$6(KS9oX>O
zK5|j6VV@A7=D3d(@-)sH&okr#0{-$PE)
zmf4hTuS8&WCo84n%2CW8ww*fAQb%{Dwd~jByH{vso-O!$StOKNl*jqfC+|WHKA!<>
zid*jqish*A^oQ_puK&*W>-7q5th|>OGY+%WVw(Lp<@=;L|S$?+>
z9X3PlSGqdPTHj3~EZqJ)4QW=7oBq#T18WEz3TSl0iZu52%Q{~;`!KiI<_MIx%}hlink5)ipVq7>qBNQ3`#V2^Y8
z58IohzbMUMgq%A$FE~}OvQJdRuFTbIC_qcW<1&MTUvra~!-ft&#cDutvWn1Ti{?Tp
zp(Q%i8{zcwaa+;juLQ$-jb4{ATHJAZLy}3Y`_5g913IzpBv{t;D72A-dD5v5DpBTp
zI`a^1^YEfhkc(U~V$|DqPPj-^8rsWwVN%%o^bdN-gMOm*K+ijoYww)pBqe5p8tzQ`
z4?n!ve^Its+gSiFsA%(Bm0doa^=QVs-Yvu8Q7}3xZ(z$#
zz2y=Y|WEk{F*E!gC`%$DE{i)+A@A#bKt_7R?KQdpOxu%(-cxJ)y|s!m3+CBivS
zOlM9jYD*LuPG%x!9+WB2^wm+Kmg=x}PT7C!f;#wc{?|}1Y^oc)R(>Fm=Om*93a*$l
zKdKSZE~7wBH&9zL8wg=Niya+=vRXB$^6EPZi9+C0Fg
z$X`aQBsc&Jwsi+1vU=?~^~8r5bJ%kR_Zmx7ANZfmfntu8AJve)R->0O^hUPs`x((s
z-QSf>R!<_+b<2D@qeG|q(jN5kd=TH}T?wcASTf{%r&O4H#=O?=xzDhg0Qdl|!40B2
zY=-0w!ybD~<7RDZ;m7o<=4XiwTjoJSa7Y!8kjL^+yUb`XsR$bAoU5-~p0|N&u2*7K
z6J0OQrAioeFh$68V=J?_hU+v5%gj2WAJgt}khYRzY{kqb9=b>HbPnhuB%pFc%vern
za1Q50NKa_(YK*Q+-l`DP?1t7#m`x&i(-XIYuoAUAQi{=)1d}083|z)5vJO_kW+uoD
zvM9wBjBK6@5cx=*RUv4vwC?QYRq{iluVjvZik^MYv8pRWKtp0q)`}||eHD2Kk|H7C
z`}VTeW!Vea%$0G>IF?wT4Ib(v-OVeV+V8iP^E4ukQBVaVps^c
z86tXMyF$}M(Nz5_2{yN`Q91e{H<9M`6INtyUg(zl!$p5gEHA
zWu0ye6)OOToN`V|J^4F$xM%AC3gZGSo;1*`Z>OVkK8I-HgcRU*1zz;pZ4v4)+0Ye(
zlR3OBi4dqM475?kL@+GEqx87^Oi{U5BrSaYFF$kZ6`1CCYH
zGsT2lh26tkCC>KFHF+`B!!d5rZdkM-*v
zXEVHwfChJ`I=+xYNaPl`xogRY9CWfPV*Vr3ChRNCM=r*
z@{on7t)cXpmUwl-hdij&=&^)#^Sw7P?JC1lzjTie7BMlwjY|x(He2Et4H9sn*f@MR
zfe+4HQpCc(hI19=+mWt@C{tQynGaredv!u|`)q5bRZ48$5uBwNQ^d!Ayutfe=d(?2qJ;f7
zKZ;f(n8D)kl-gMs$)<`OO0x9~fx2epR%%D3Y(KMt=?ZC}Qm1uk@-L>y#I$i&7}W`IW3q%nJ83gjo6|>
zTX>X#4H2^_vLk1cT5kfg0A>&xDt`+LHjd2;jp;c#9zNbDm9_3D?($wcgowK}wC0Om
zTZJVdeF}Jn?|Q~PAF1L#Xw!AyoZQ!a7Q4UN9-DeeIj*B~(?AnzZ~lL``J>?26}8dK
z{1@rgvDVzN*{Yut&&To`845AaxhgKp!D{PEOO3`=nbAQ#85m?+OR(gZ?Ff&TdjlAhN{Ub|jE6x4P
zgnhVJH1J4`{^{uvHC3(RdHyiYT%aV?Nf3(QY;^AJxo~Kb*6%1EQ~N~kiV}VSa$8TI
z<@V8@I{3V71N)OfT1-ct-jG+hv0%thGy^w5N8L}gp{&dr)9gI;P?-bCV;icd8vKP#
zslA*8^iAar(7YfsDCj}mxRuEq)?j?&vT~@G6_;ZsGp?-9x))FX{QV3xvQ_{TeWZhx
z!&C(8LId&>ghsuk_yIEt_~+ATubj6kD2qQ%>GtV~hc3;jX#X!K_5G!rhf!az7>lk}vp>1u-NF8^yYk=*-?|w3-%(x&x
zg(WQq7qdTcoB7-dcfv#?j!z|u=Nr~s1`)ZE1>0Ox5
z)*PgkDECkl->AL{2fFyx*zQFjp*TY;XdY!;esVD9)7iEC!Jrn_bLI^$#BfOd6@&rULQs=56%Y?H
z(wT18tAhTHJ-MPxjbz!wy$Z31wW-_((0dK!n&2c*Px$bBW?)2eCqw8QNo2J)Tx?}h
z-}a`+(hTgu5|VQu)|%l_UBlQpjp7Rb*U&xQDS9=6Mh8
zSwxYjV!fW6f13f7gJZ~zT)_B>uKZXRm{4wI{~euk6wczE-~d|lu>{;BQDj5(4)MW!
zc&B4Fje;?3DO#=tM_On5-n_4*nGN5eVs2bAB)<9oPaazgi3DiCE_~p?6?3UaoNqjv
zTf-|`wsgJv!1C7nvZlMIzrL26%Vdhp+;!e4k4x7LCt|O0M{|SA{9~+Rt(ZDR3`(wH
zq~GTE`idP=8V;;Hlt-dogfj0QknVt!1YM8NuUmr3znfD(l^uCB5dAz>JH)1Iaeh>k
zc!(vhJ2Kb6ITJes2U}b6M7QVVa^@9HoyzvfBxIX)8n4UhR@`ep3gV=jvLp%)Zqdrn
zHC9dg{&Azcs|QpPK`Sn_*PshbO`}1K2esKJbs35u_9~woY&kQua1!)z22a#vvxL`k
zcs`HR8skw=7S#PO=K1$FV>tcXgNjDex{c*1`{O4NdGgd@YUY)-h>(b-#k_)
z3;+9kHiyIZL-Z#3ui45BdLKif*~`m+bV+iU+lfNF2fyWkokhcwNam><)Bb-_avRVO
z5!C|2b!Mq<)VUrQ%GDzJ;9+&*oz2GHVq$kpgGD20YUGhBbUE06wH+SdO$
z;4gpA-B67RZxK9k2Mcb4n&c`3GNW@uP0L=xARuR8_298;a^(Ovb+ES)yE6F_mPKEk
z!0{+QiHKOCXdj92$_EWZcUF$@xYggU-i0P5VOK;ab@w;o3O};|ckKPn&g)o>i1>yx
zE#leC=m_0FFJbkb3`4p;_=ocawmf6RB>OqYAz&ZFMiv(ir>_S_ssZ?-m*s5FbH#`?wS
z*2U!<0FMPuQ>012WTGsU(U
zf}BRat8s)HfoCL+A2BBqJ(&^2aM=eV#T?CLP23nZgS}>j1@qw=)W}nH!nX}l-ps+*
zB{(n|9>dD(e0Du&F^bG3yIO|RyN5GgCC(<^7fZ-x!x3RT(9rzUQb?nhn~I4e^EA2q
z->$brtM;)!8PSkc`Ju)=y*lFvXCULU9X%#fo?d17x;X~AHrab%fDCi#&97v+%Fxsl
z*}WIK+*l?d9!Y4O|HzXnn!s~9oTGo=ZPf>RRu&d@2;?%uYU
z?pUqM;bHRrq;jXoZb~;*9BO334P;)nna6Z*T&&5~t?Au;Fma{VB<;{QW9zfQ#n1^Y
z(O7G7zROR3E}R1y&Nu*8^ip-g_;JK%%>rnm-Q)wu>NT@QDTk{ow@d3>m-WjHo|O>K
z-23M|i!MieN`4{H$J8M#7rdTJHh1%J%6NW#BP>qF7H1L5bWlXN;PI`xQwP^FdeBtq
zp|dIn>8#_cdnP(xnxOiFi4^f^txO+K79kYMD8oZzYTr>V?EXdjkcP17aIKYbo#b&L3ev@$rCN)XF~}&+xtnv}sUiJYaKM
zIeH~tI%J^y^JqbWjOklAGQ4gyxP0QTzJ2Epf$Qi;ij-rb?0;qZOJjI%0E!o&@}$>5
zqM(`QmH#`yR?`RMz$4>BzV4I7C1P9XCa^)yG1jOvxyP}&GgUU=%v$4ApY_S)6H!N6N_A()5?)5@*m!5Ymt@
zV!B-pwm0A7oGK>=Wx>~kiWx3_iTr!%@rNSO%u!ckuBzX}EX>5&k2`sNywv!K(`Xh@a%^k6x7$@URn9_A|4_Rf*wZK
z4V%iHWH!;?R_8q0s@0TpjP}0PuE%J9K3@IIxsPA&jL+PFxra8p5{~eAk2Zk+U0xl1
zmGNq{_LcwMy%P|~N(0(#ARfA-S$VB*!5!6vr$~)k!+lmCrXTf*21dLA
z6BF_R+!3?y$_v~DHXpGl2_1$&W2dAxLLB|GB#BJ&zUy!0oKEcAd=AIG6fgu(^k(!^
zCabmE5%(jn8X2;>F>u
zL9*qwy1F|GkJ_VSkhJxetHXCn*isCfHQ-mC&~HwJ!K>!~m1Gq-aHj80|82=)JFjmIwcto23~N20v=<$~?$<
z&S~~3QFY=mgtdddGKyQ%cmnqx9F)5ymp`MkJ?}-xVdV#xXKV%tM5E
zp5kYa>&S-BW0VD0+rclZ2Gq(opTf=kzv+vEE!`ps!X-;`b-nL3$-@PiyuY2O-tYIq
z&a>_UB=nn8$@g_ZOH>N!YJi6Ese94ROQ!iSzAVb
zMb_~s2Bg8~qfL5_Oud^T+yJMs_hAdU0?tTJy@wbr|8}lC)a25xgucY9OOm+KnYpzApv4dZ0%xBN7U=*3RED>JqFz8kNxc7eySnJ*wtI(k5ipn6P6!E^#2;Vfh64Z
z$FDTVI`-Ye6l)m}OE%U@j4u)@f
z0G`?2UttlYmD(ogx~L=ZYYqTclpFsd`C?tTds;8>l;~>T@{@jBb{YmJ%noWX+up^
z+WW7<{63%}y&U7A(TSi%lbU(;fT7a2r_O=c!-_dpoVxw*lXTLgQ?0BlI*e^nOT=p7
zg3p;hMQGAdV65A#9^VFxV`bZ;27~h{cVbUE@F|=Yna*KPbAYq}-iRP=i=H?h%g@zF
z4RD{Ee|r!t{BEM?=-av@%^X|3mdAJBZQxdu_Y3#WCt>ZEFN!hNJ}`1Mld~iH)7!?$n>ix9#`><$Ji9tyh??^_w*H3jezW_u9J+pnv4K=|rT$RrzG~(*!^U
z`Vco0%EZ+!+|R!J$FlnBv3#hGJL$zUGYy2hshv%MFs(2Eli|Xs(LL4l^A~RiJoVAu
zPi@b1VGi13Xj+wgobUtK!2B-`huusY-+lnxsRbvkEphGv=^~Qo@E4^hW<;yrPt~0*^J;-K^U#_ByO}}tUDqqMP{t$cEZbVQ-v?Tr4?Jq)F7&VvGlssw9giEr`++5Lb}%zn_x3$K2c`-Jeby;0;n=XJW=|!fJ6aT
zM(Ei4yHN}3`hAZSPegOH~q;8YN4h2{pX$Q2uM}mzFo1(h1ftkh%Mp0IHs5@o%@ZykiT`6BC<|a$3h$_Sdo^0gHa5
z;lZDFd2nn1=H+(thE-kH4kE-3B=Y$BUVhvyXVAB7g=D*}Wb45gp$R#Ii>LoHymuyn
z^4Yh|N3{2YMNDFH)#NJ+6aBH1b_&MnRl)E(O1@t;%IkyA&eC-sUw6B8)u_|}nMQNa
z^6O`5T?Tn-O&^Eau!vCs&brmz{DswGEaFWHyFjQMpXsIZw#My#rObTG^QjUt3AZ(7
z36cq5)$ZgKsLM5CowJ<4+>z(T7j4f+64}@O)F4VKGv;JuT_>lGzyv~oTHfX0bG~b1
zV&rN8WgR$>7F5+jvE=
zL-7JDuC6L^nQy(jC&SnoAu#YJ`vCAVQC@N@D7vK@+Hd-9W$~MvONgI%)#EQh&Z*uK
zJGz$qJN4
z(e0eToZv(a_OZJY)EUtD3HwfgZf3K%;YSJ4Ucf4slnsINEwxs!cPpIr@)m)>!2K-y2MNeAwp0mv=QDPEk>0
zybDck8?xmshG9F{UQ3lcnH_i|`<(Smo!j)nw*>!`UPo*PYvu9W8f@zNe&gWjy2`A!
zvT8Zos#}pV50m3>e+T$~3gruiYby;oq_hiX!f)$hfYdAMQVlaYMXS#3Q)QWj#1Ov`
zn3hfpy5SJYMZmJS+8WFty9KsineuQn-NPO2SEZFVb2MJS%yz_UERa`yUTVDPayZb6
zlax%3Cm!D;*Cww1XzdF6ZE6iyhk~x~)PlGlDf=%yDLkX8CN^5~
z;=;3t_m9fL-kr!Vl7!;pArI1%)RxVAgG31?L!3bKR1B!^s>J|RA>gm}r*@OIs!+_I
zQ`*#YyX;q~n@+ldN(Dbw)LQ4>W5=o=Cfkvd%XdU4Hv|+MyKkKei&FhLBm#K{o6sfU
z?%PxjfCW{{+($bS60hQ7Qt?!(=LbE6G1OpvW3KzF`v8@!%qjHImVD#5_4x}WQ^oZBOeC|16Jw#PGhV87v)7|$!JBmP=nHLzY
zUXP!?cBRdN|Lx{~D011SK@~Nm%3$b9E>%?j>(aPwgm|U7d{Qc;xt*x1w+z%pxr?V?
ztz=g90d8W_avsxDOG*$-a5$^urGslbqMN1mIjzK(+dTK`y1;F@*%<{&iSaiLziW8q
z(2y2ne`5I{by>N}kS{*y;MPKU0#Mm?&cvvA4QSeWPjCtGku-=ZL%j8KiF~v
z)B{C=XjQBou-#G0z?(MkwV2<*OIAFO{he$!q%ozZFCXbxL7JeAJX*uA-*UkpMDr<*
zDY`bb7h0Q47TJu6o-Y@`*b32S3+vh_>g9iz?Y3U-nG(C6zAwZ}yOxH1;J}jrhFjng
zS_Q5Gc%6quZX3cTs`66eQyWi?zZBXus#)^UqM0U0h-hF!P
z1iA5`8}S|Kjv8pRZ&2mc2)^1$j^b!su9~(x&El{-b({2IrQ}BM~&yNQoIPb<0+a_9J2U}9lL`L
zR?q&@_0xMw_0tOOYX?GFBaQ<;_Y;J%xm>}+2UJt*{AE_-mXy965uW~_wUTV!*qHNS
z!2M1s%*F2~-5PZKCUNz@{>(>?ZC}=^r*HLYt>q_+{#U~VpS-t^@zI(sn>*Vns1UVK
zYjMGemu5L(%OO>;(=PwtCyVs)fWN+tO(jM(hSJgU(lQ5&R2h=;V=hGgyxVx?w<1FG
zx2;@f>6-MSiC-6=PQ=D(I!d56gV4bk3Dy`8IPu6XLyq?ki_h|*V8MmUqLKGYa8=E6
zKD+S^{`=WQSX49cyNDj%ky2evpW_9&5`n09M!3zD{YUMeVl=Jfv;!X^56W%KSu1%f
zdiaxBB1zh;7Sn#dFOtgh+uzD84s0IfL#)6BPvk3)>L;uh5LXo|XTBJLMqLbX57&|%
zxnZuHv}@)^Xu-wRZ$qL>;poCYTPua2_>cEg*hAFLhE@p_8&&6vR2a`!xs1KdmA~=N
zM?ZI-93(#RE}GgXCFkq;Y_{3%h^tb`dF8RMIBxxZRF+sj3p9Y~;unA1%Qsnmc6sT-
zcUGA`+-oAj2U-^OH_}sHHv!ym-2AFD6Klyt1(+8MfC)T`=f~uVhDQf8;-23so=)HN
z)5p@?c#%7T{RWpN?aIQ5~7ScqY&tDMM=;EQt<^F0wR|6*f
z+7ke!WnjJuuJ;@?Qy|~@34B@0Yy
z%n9d&^rGjLjazF+<)goGpn1KyAH=J?_9pWz^BaNk+DMl9=sQET
zQm=nKp$`^68ZI$;5zo<#pD~PZoZg2dv3{zD>HfYtf*oXv+Yu~L2yTQ@tzI);xUtIE
z!N|&I0~uh#82NYKQ6&wdpmsOk5W9)u$0(12Tm!{&g^3`B%dQwuUrAeF{yE%GQ}H~P
zP?ucEaBo2Ja`V8}+FZNl&NO6sUp?fe<*crdb^yXO^9~oa%@7Mgk>0E~
z8Yr4X(7yLF#oEU9&SEQ{3DKv%KDFQ@09=e(y0?VX80htpVzt|Z4Jj;b+#$c;**yv2
zm=t2c3Ry8LqsKG05!-J{H1!#uRtot0Xt)4L6GxJy?7Qs(gQ2hwi1oNnRu(Fp%hR7H
ztIbxCD#(lWSA1(!eZ#f`ToW%k~hYT^4*3%nyGjEMaGK5ZH~=43>#p7@?(G$26xoA@n4tp1Oz`d
zOKeXYWvWUB?r71={=>z#-qN^U?3edNKIE_-Jxv}dTHE(pDPKVwc7*KbCkRJ9%Wjn%
zt=1r%Qo5}g9Jsn7x>7$Xp`gd!s=2mLP}};{|9R#>7)-Dtu{fMI>#w^&AHo=ZE#{j&~
z{WaP&&jasxo~z%o4e9X<6-KM>6ITejBeer1CLSe&$l)__^D6);SKYX+nMVo&Fw=q4Q_19%&(&
z0_{iMdrw*}uRDEW3fP|!sh&?4Q_C94y|BEuO#a&%n#y6~KB3F?<9IBgX0)O%A5yZP
zk82BFwYKCy-t_pvDp4x)!KC9JTM!}vV`le0C^WZdw5H5iR)%o22Ms}aP;`L#YvNDk
zZtJ_EUG-bT`yCha)s9l~Nw!!t=5?*XfeyDwEM&h8Tqk8HT@PA5ETh*Yi_>?$lu1kl
zx8A_Lf{PvJ2$?^ucA7oo^Y688BMNM@w2=c8hD?vk$n-#7(43sYT?%USgOaP@*B_)
zTGqU>UpQ?;IR2-|P*OJE=d`Y~{mf(MQt6
zH0ZSe>f-jT?;0a-)}|Wbj>wBjr*h+2u9|h})?&wi;i1*oBt_Cpse2QEO}})y7X06e
zgdV%ME*4f5M*F=#VH(nV(OhYFrVaA?WtYd*(2`aIA+S<*AVN_5itZE4Fkt0`BuD#2
z>>9114Q^EinM#f(Ce2uQ=y*qjZdfFRI0F*Z*w%Bo%f>y#S+%zTk%t$CA`_*mN
zk>6o4Gk)ARgGQBPg>)MGSwX_vw)b^{Dk#msH}@uuEZ1Png18d0xggv7uTKmuZqzI6
zN)%RASx>LAY4?`B^dM-LmD1z~Flz2?eY_&&Kb1moYr4%TxZ$r=n(AiOpuPy;J1$%;
z#$9sX)L@}&9sFT!{hO&QYi5v;{TMj+s?O+zsz%_rUk(0BYbdN>O2p=8<~tpG2VV|nET(}{Aq+2QHx?r!#CXz4wN_lbZSCF
z9p=qh6n589n7<@mh@dCRAtZuaolDbtH^gRCKX%X3i#D*
zDCiTFf?rA5fm8Jsb1T9x#=;r45VxCPmhic!W|?~~<=MQtb8=M|07XUXOoFl3yqD&;
zir^u&VFyE|ILEfPU9PI#=ertm3VJ`t+?dF+KlmrY_cmO0JaXW0LA}uFivB+qz+Snl
z_J>sYWEYuE&Cg-(TGh>Bf4>=3>h$LMaP`jd_sQ@=mTlzm6TJ_Y#vf63B`ieSY{{`d
zq%qbmI;+AI>8x&kh@68%sAJ0V30qsKqvlp2EM~0eC|UyXz`~mFIw1J&m8+J-H{N@#
z_SFlWKHAN_O~dv!Yz9ISoFN)fcPg!ptM|rygs!|7E{(2`6mg01cx=NB^H?;drf@1P
zan!lniwrXrTv?j-Z2K?H@ma$|7h&@EKR16WzYtM2)YW+JdzPna2~L?PKsffkc3jJ4
z`}`wl;a)1KFWr?6e`K6?A8EI7`=I3q>E6M^%+(0b&X~1DV71fD`lAb^e->_K$B}!n
zfRQ*Nr!6$`2!e^#9J@11!&+A+J_>g03@S7*-G!!r`l^(oknTM|{pCxNh@iRX!J9kh
zS-KXC&m9lHNdG>*QxU-^JN2kl;I7xmO)NlNUYqIMwn=mia{!H&Qtp^pe)^zvB_jLu
zb8t04zV2?I72oU2U1*%`-~nd`Qq!|qvZBXq3}%1X!K2&wPD~_0a2Am-K1$0zl(w2}789`hXzm$t%`g4LUn`Mtw_Y~5ekI2$OZZ^?
zQWE#63*G7^&&}S%8T4>VrJs?l2b+6z68MnF0*swV4-K9#4M_+l?i^T
z=?;MCndmd}HiI$2o>+R{?>Ko87vJ&=0Jx+Zp%@hib?r
z-%7S7*pBvz7=0)5)A#Hif8o?C4GvtY1BSjOL%3jaN^4@LEV*~dRh}ZhF*1JO@});4>_t9e64$6
zrA%XM4@b%)*(t}ro)5`+?M$jDKkU(uYD4x6N
zNjz^@SvD?tS`~EpbfkQVZIji)hPO-qd4qc7{glAFal@tNZSU>LQzNyvBGt3ppD`zY
z*We&)8(_d^4UnpeKF7b|TeD+1c`lt)*gE|M>MKP5ly=}DYbrh81~hrNkw`a;N@U^>
z7&n`A{^_^HSu$S9+53jyZ#hOa>-zRLp1lFZdCPt&-Wda75#G;?aZ;BfmGp#6XJy%2
z)uVoB$%EgxClTx0CgVOtP8C8Q@Bh)w|HzD!sa(~_ENDyts=}>cru~X1U*z1Fay4(S
z{BpVUlyc1dw{eW)wEjDhzedtL0kiLTh5LQiD+K;(T9{xFl$JQ5wQBhYwjVh|c;c^rvP0M%YTmJKK;)@T!5x%gogSy<=W-
z#yc61k^f=uJENNF)~!_(6hTx#1f)ol4pM@26ai@>-B6`>BArkY5do3jdy(Ekq_;o>
zq!W6F(0dg^uiy6EbMCm`IPZ7Q&+pGWMt)?Btev&X+G{;)KJ%ILN6C_NH}Sk-*O3Z<
za0a6LNz=|U)Ki~`x&OQ;CXrMvVh`RqkR3_V9{v0ZoBYCI3hQZB^T@O!$)Z^&Mci`)
zOTvuN*8tG}st`1GpY^Ld^c>Kw|GMlpctF3DgAzG-5^>-Kou;dCcV!#uqd*^%yVYBL
z(E8q0%2sRdf&hHUyChUXYitlq-RBj^-!>MT^!2B9%=O#k+~t7>8C6vg>N;k|_hXdt
z$q@+K!wuh+oHqm#@X0KBB-o^C@q2VdO6t>f>GdO3);E4r148etvWt!^K*bdUw4Rlf
zgTku@{b^z@_vl2Pc*D$W=&a?8S~i4!pp!IHHL@DuW5Te#sIXdy5ar#AIT96r%^LPD
zfsxrJ`3Z}n)R2Iprp_?+)w7i@!^=cbqzfpy3}rkJcuJv06sQh9Z8P&;K3gcVY+DEm
zV!KKI&f)g2ZV&dJ=MGF}Odn6BzV{ni$tWb6P~V$G4^-YCe@hq}B`VzaRJMFGY#+kM|C%Jp-bp
zS>X)>^&*czF`>c0=?%iNOtfua^u#F#LN91XTA$?XaGQ&(^dUrh#3C-3V*na0PzEKg
zk6!ZlZO8OhaWd?1%XDF4v~JxRYIuU{%GwhT|HCIjI(x_}gBN>>YOd
zM;|2Ygnb7fC`s;|`=6v*p5Dj#AYZioijYcw>|;-Iz^7R#_v*=}mr#YKdPxY|r=Nmv
z(xux%w~NRb7&Yats9ra-88+YCSM@}PT$gZ|zqs=hH2=Nadbrc?*Sodn8zmVpAKtui
zpH=v}AM>t!#*%}t6n)aotgq;47NBw>kY`MfKP3u;@W8WuEWtzg-Crub{HBO=H_yaA
zBZYBA`PL7bn}D>|V=#1J>Z6G}trprF4I7nfeo0$8EqD7S{WT!0p}pZgrI9*Q>X;Jf
zvI`l?53G6TK3z`R-{_1&6pEt%WcYm*4>gOOxSGc~s~o^ptgm&xi3yllrYa-VZcwGa
z|9~sA?413jdCKTGw)9FWPQ0j8KU`noe4Gyzjo0vz_qyb1I0Zt9_N=cWJg_
z=hAyivKui8pPFv4L0;}icxUEJN7F9WBTb8cO~oq5+{+?@RD?#SKEZ!e)FyPl4N^$999&
zzL*Spm=yUb8pH^9)YVH-T;Wcvfa^Tcc17^8y7^q*zkTzL<$aHZR=!pdB2c^7P4+uJ
zfctP#0NqKom9^6~R*0ab6jDu$L@^Coo^Zj$E6J~vwgQH{IiRykk
zjn7l*02!xTOJwKu#t+c#V?z#4wd4E*ASrqxTQFvE9R9R9f8+d6J+dtUgzWW&pnW*bGGp7%?yP-zdu-j?_%H3it
zhOjS{U6Q$KT#QV(tjFUYP|FgB^o;Mh*+~L|Lz&|eWD~yEfWhFy@#b0!hWtx!aTKYg
zaJBMOXOs9d%mAJDhqt!9Zi2n>`Oi-;&HK_DWk-UC3pItf
z_H+6GG3kM&G}6N;4-)`nd-ft}9vzG<+9b?E!cVTH%szRTq2@6T@oz2C1kzF7d+a0@_V
z%Q#%$ZyxJBdspaSH(0I9ub!<#?&R1Ousn1b#Rr%D^ELnnfS5e}W3@})NLtm?hJRC(LaSqIBVKQK9&OjGr+EU(
zP;Pf>lLRVOG;@oET{rWDy>e>dni}f5_GL;;6eej8?sKgf*
zX{M`H8P8x`0s&!TutM)v@<)$iv)2k;b^Tr(5}pB>;W~>;O-~?d{PO^qZoZphx-R|v
zA8+uVi~PrP`N!%1&#&!*J_R+?PtX4GbpP$w|8dOiXP@R-lJ92!&x8NR&;R;LzXJ5T
z_(WLr%b#KJzx?4G(C&-VcK}=Y*P(xTj*H3Hz9l6yS_J&>&7|o6wEM^%{pv6O<-f1r
zCz({rAtsU0ay|F{Q@<~Oc5`~He*ae||Nh^?e+Jb``}ouU(C^=a@2{Tn--GWjhmF+V
zgYPdU!M|s|f1iH-%(#Ege1EkP{=Mw}<%0S5vinzI*xxbVU#-s9eEyF4{wngu5>p>QQ5vjv
zi0XtoCX@f=IRi6fc%;-pv<=5x=lZZs!iL9`y^atMFNUg0NSUhgd2Y`WemTi0vwJjH
z`&fi$SF)H(-own-+&fEE52X(P`XxM-
z_HWsOtn|>Bcg12H%$UKJqs8@WBM8qa77i^=AwbygteL#F8EC~Q=nUGK3t5?~J^E+D>NjlHY!pdARiBchpWtgOu
zga+?384iG7OXPIo*s>-kWHU0gzu}f5+CZY>Wf{v2ej=`APJdr^)*9+|Yv+hYv?>Xq
zTlppBp4oo~BI#4N_4vOf)w)n6J(s2x657A|V)d>Htd@Q>weMZ#jAi2Yo{ln~`A#gq
zQNrMK!z93~hs~~BCf5~D`DN|{^NSI1f{+xZFjfjMw8m_?DSZ_Ma>j>=Y3GY_f9_Z{
zh8;vm=<&<06wozQ_+AZE4f#_FTL+6*v2UOMVY@1&oE8haZQD+rlzw!1@CHQ!*uw_Q
zFR!28;n?h7UU6eT3|4me5y!31M?He%=eKw(1Q>KFYR0$#xq#a~rw1k%iGW{dA|Nkh
zmi2lK@Km0v_gI~63?>)4ARywcn&O+L!F=~6=nkgb@q02J!(lyMYPLV+x;+*;zub92
zrj1EO#=2<(Udy&bn%D9)ENC4}C#MhjQY8PI^
zg~I~Ur&EyAqxcW0aGb9B+Jt<5K^ZsRjk-ewA6Dd^n;?&Y2;}g|`?y9ejr^jUIu-UN
z2u;y=Ko~3bn2cJ`+B61~kLXgIvb+Pz1Dr3R#}Q$2RqksoD;S{AG_%VrC4!>05${Y{
zcSDn@|K~l0q*66A&)ZTzN%OQ2iEG1kVQa&Maq5IBDIzUMV-NEC;d16>%NT4#PrL+X
z6I*x0{$jn;ZbyjGb*J>m*J6`+G18|g@4Yge#U=op)aTbD0f)>Z5?HUYl@;ht6NlB#Bf|%SI%#idilrj|<1)rcuH$
zV`RQ=T}g+{L?5}aQ|I08wCuAbQ3I`*_#wJ9Px*z$bC8_6z#gA}#
z#6vj@7|8TH9%$8a`Hy{knn2{tlzN>0QrcGa7d5kkK6h-PM!s=zT`}BQ8}7a4x9U}C
zp6Y!_=>)Jmm17DtyQfIZEegRCe^|-FKBS=6{dMZoY_73;ivTO3VQs7=b8P~V1lZ5#
z2$jRDrbMRwRX7B2%iiy7sR`LxeC$D-sYG;4ijR0+IXK;dv0(WLg~xJ$+hC-iE)8i91^OgCzY=P?w<)VY76l2}aRpP+)p@efwNZ7M6(NeVn=;~1Vn#F09CM^{(bv|xxnIdq_N
zQg&e2V<1a?YH`I4q*(|yD-FQ_OS2;%YVxzO;xZG}tq1RBi!Jc#8K>gIcJyDTYlupeZlsA)6(N)39Xdv5gl)
z1c9E&+M~QO?@6t_%xt{YQw4e8IK&U-`K-d;<9JfWRb|$ReJV$4k
zs`#QwJs@*(7&$pqZq?tFDt*^!#N$`dw7~H2oe@b$jQ(D1$7^Q^5uRss-lh_qV)wB{
z9u+z{23Qt*uz`wSZs_aL?8JQ4`byCq^AOSavK18_vk6(JQzh%d*@=pEU1nYHQCpG9
z7MS#zsAZC=TCNtJdf6Kvx4HT8^0vsL6q^cE0Wqe`afxscAUI3&y0`O|S@t&tI}ow$
zRrcc-Y9V+T<$`ua<1p?n)45oGE*X4+toH3WE~RO*VE
z5NPvb95UaZ0jl2>E5k#{!cML2PAh<|jcl7HmDSjf2rFz(g|PQ}x$pEIY=jSJPi$e`_t$tG$E-9h>os~JXKfiv8
zzSzG|iy(6%4)dX6-JpCx_
z5|u0l9e|o1Fa2UWGGK@p;Mqt)*l3RaxD*-V{i0iCD?jm}YWsN$=Fn~ubNy_}X1p{8
z*AyYt1DFR!+D`)1>DAiwCRUwzUee9G$pDo@++tX$W6^_nTAg{Vis{L#iI&+q>6l?{
zc^9u*g8ibt1LWEHY$G0JJ!HuBhe~!s>Q2|H8&GXoUyn?xKKppEuG#N0!L3>ie;Bw3
ztI0ss@d@Kf^~#L4Hav!_g0_7DdCSD8XOf)iSModB;=cRdPQPbiqiBw;qgR!^6-+yA
zyjs!!Tx8g(eH03-DTmIcS0NQTj2q|lT*TCv3`y3m<9VtCt54-Q
z;bOuhAv&Oq`4Ke#ipBC4Sh6F_3Wt$1ESZ#%_%(eoOR(mV>-f0C
zTeSs4li6Tqzj<}XO)SZcn-}eV_}#eekrwZ6&Q8(;Jc%VOyYY(l^K%}O*M;QHv_=wF
zvuRU%mEwvRm9)?cP``o?<#QU*eVqBbG|$YoRc#K*{%#pKA3w!qxb{m)?u|)>;15<8SzB$&z1uT;2s58rOHTo1w`V0sfR-D+Bg(xLV@6T+@
zp5(Om4tm731v2UWblo?&s2d>iA%)9kgv!h@L^Q{zZSeEZd~JbGkEgapuV#)if(VWF
zRIWBA`^|{Uy?4&8+PG}q?HOM7Ub%jYIotN+EkF%)nkmNpVzaleKhWZL9C}}c4PM(x
zsG14Y&n)JhLj&3x&wKcR!_I`L`94+W(qc!6rx>)ntvNw*$O9aNveWLYvp
zZ(MHDqxVbefHZLMt82HGH0Ia}z4YyY2;4=;@-R@vtCMxbf6)znRxysUQvcyGg=%61
zxR-A-ZHmkhI{hwjJo=Ts@n@LDj;L<`gZvqb{*ZV!ejh-!QkGn(vGMAMf7jk7k&Q}c
zd<0`#{9H3cpdu)Gs36xz7p#ds?VQo;xlim5-wiT!QMR2Z??1nE+daYe?R~F#IxXq_
z48gD*WO!PjPE~LTfuD6n`);ZICVKp$J(xvd5RF#=|LQB|U-=grF`9ytUe>!&g)0Vr&~Prrh9OSDV`CbMRE%uhFGx98;-K
z%mCoA#U|C+hynZ#i!P^uuQ
zYnxcWlm4#4N%Ht-ciNQ?|0ZjtgX+uQsqFG-X0WHx5@2-27bbq_E^s~x5Uj@BI;o8`
zL-q)OqCik_FIoO^g!(i}9QwzZPjj`hfp;v|ln}DvnvNuiD9H9Ezp}8L^&jm;V
zI{ZgWM`6#v&vUf*urYUhZ1u_-FY?5fGB-@o37>a}hJ~%jQ-kd2RG{m8&-MDsR+p0?
zg7qFpC9Nh1Y*qYBT=1`7hZWjavHQ@)jZ-U7Z$pE=Z+1gY~?GiG6RHt|#P
z;ILFoGBccQSZaYlF->qm4mQ}Xo(|4*%`r*Zz=YO!Wo~i5KPj!5T>csHrX(tGK1aV$
zLNyyBS9!Pc!;)d<)(IuFSnTrWj$K2_8wnD_?cBW^?Co?$G4~(|;Gi@l9SMk)hL
z6gus-M9LHY){fvf!+7w7!{8gceXa9j<9Z<8&7|Ar*5Y|jKYii(_Uj`qBH`TCP62+r
zGbI5!G)^wZqayLU;L=m3W$Dvth(YqXewp1-WVOdeWjAk}WUVQ@%(_QiL4lrat|?1y
zMb)9@R*D>p=G0I*txU=!N$7}S{quSk$8tXY&`-wYT@wxq{s%b$6~AMNDEWRVC6Zk>
zZnKkJxLq9JWlnPwRS|8#usjtnz)1k+6VR8h7538cc4}p!^58oMIOcxK@6p)sP
zeV^UkRT_CVG1@>`aE)Dg;q+SFLTFt#Yhkbg3o*iFoyKi?3t#?^Q)2wZSrkWyz;F1>
z1<7k+g2|z_49<&kv^|!vJBUZj264fMYrR|9?B#lgb!wI|%qt_FVPHpwFp?Vb#A>I^
zXt{+Vt%9&O2Bx{H@u;PiB&+-mDX}9ldQR3Rdurr-Hw=BZP_0PGcZQT@%Fh1DgDcbl
z`GKA#wl)MhzsnPUgqU~U(~VNoF*w_bj}7^RN}s^COiS4l@B%1XuW{Zt
zkFqUvfcPB~hdvj?efDl8OJyf2?pZ@|eif`vp2>Gwg}LavReZxhE}N^mZ3`c#?DZ^0
z(_T~aho26ng?O|x-TNu8D&hY7U1oTCulP~``wz>$BjhEBPq
z8LjtbnR&gY`{l(BJOt-{FI9T}{L7uAtWBYl(<#Z+*DZ+#`EJ@GgyRk7j63PiHbwKI
zCSB%5rTvI)r~7~kSrT-*BrcF;-c0(r07cd^5ZkoQ;7NqE3+2u1$|*NiJ+P
zteg(LzN>q*`!ZW*qn&mL%XWUt4%R~5&cGTiM62Fa3814^DR*(J*
zyI6$c*3M>``R+_M?|W
zc0vDS>C@cuOGEp5;U2J%;t1!COwqP1_qzFt~oh*KYM^0Ie63hrZ#D?^%n)POIVGIl2^V
zQ2Bj8`#~VOH&^n2vJeI3UTQl|(Q0JaxxqtN^K1tMM_OJh1=#x_&UXzpGdgD*3*sqW
zz&rOGAex10MFn7GZ3K%q44;kxs8Qmy_h~JzuOW;^e4R5RlbTnZ`z9HFzak&AEhsew
zMwbXGc^cSoz6ALN9dEZW3eb#Jq(mOnpY1z!h6K=Ee!*=L+NTws=xEKI#2~$}Ea4DJ
z>{2WysVMGCXjz{?W!rd{!64z0cFVnjjoNzWT{X3x_Ry1)9;Y|GDuH5ipvjv!0CaNN
zS~H0GtZ9&<7A|?Xy09D|c^QLNY+TGyd`Faz6IE3?<{#$)WCPxR#ZR-k9T
zAG$yAU0Uv_=-}eIVR3#i4QkaR?l^xrJGsA1fK%gr--bh!CYhffpTaoB4~211B~M?O
zoXSzKopEzM?1W>67kxS0JHQSp#X?lutJ)-RuZQ|RT%Ok~yKEReVj&Sff4)MqHA++Z
zeQBgv*Cdoq)O>xo&_V?4S75;?O;7yh>QD$xWVSllr-4?#x{jw8;jZ}7W3}(PdYvoH
zBuV|_B0%{VRX6`hBixT;ce`_OsLUH9VH$z~=*ef=GjD~9v(6Ekbvafu#lE=L>bC{`
zs4u2g8Z2Q}_1wb{MKU(ChTRYdC8d+R^VUwe_YdNc@a>;Ic-tw364}9)i!MjM8R^^<
zjG;gM<$(peXs7MRLMG9FVe#+AZ)+&DgzPoCM`SD|V}{jqA_jhO*ni$C5N9JAu}`%+
zbq?}+!%?oy-VzdhqY)*F(>|{lGg5wx!{!ahD&#&h^ys#ocrH{tox~5ftbcpi5lnpu
z)t0ayqv+mqiNb8-6)vnZa}%vOp@kvj94hETm)vf(_IzIOk9o7qK8wgk?`_3DyN#m;
zOa89WdNXKRD7S^k;IHKewL2s5KtK!0KHfMLx)a99!ksm#-EivCRv$*AH+-!f
zD59HRFA1lYWIVv3e{N=eLl(cotnU^9;(%4%p
ztn}#-s-1s(+#(Is8F8PCm=C_zWk^DdB`eB>`hD3`p!Rzmm0g8T_?ZuWDMmpkr@+{u
zEHtzH3eMyoL}=gaHGao=Gn~$D(K14fUgv{F>A@yZB@{_~vLv0Ef3-W5cKKmMcS+{m
zbSbSi>9c7F;EuOU_6w#@zCj$zK?y%g0Rp}3Ac_>Xgzcnsl=uneMOfQ&leR?i5qf&t
z^o`2NqDJ|M0okLNX|BMXkQ%3z_d6*&)z<}xi`mY<_lh!^+=Bc_H$B`!FzEA*(@HZ)
z!h@$gQ(vDbAJE1<@F!w4eKU}+P7x={?htm0LD`N!tO+%>{C+8gTZYQ+nEP=N*16V#
zk=#5+Cw6>WciW?s@#^-@Rae2pgjahMlFIq3?dE8i9SX{pJQ8MA_h){uiAeogG~ej*
z?4T}U>nXk+0j0IuHne55LLOC1D{iIVDdgT~UvZ+rtL0ap-k`6rY>=cQyRVhM&iWv~
zE*9i9P-`*nZrPdPTQ-OO)N(Htu2mVH6({p8(YTQ?vAelo;fK4ub)Gu@_a)LGsy!#qksKTMRW8tw{Y0SJ`-lF_z~>mA6EI-;XVe@Ccu#v#TOjzXPEI=@H0KyY$PFQA(MW>d0
z%k7d;$lb7ggKEH<`1TZ(SufNjz5t?`_=D*DC5p_!VwwQaUK(=NxQMvVZy7Iar^@A4
zp7$>eCo%dS_unw*W0*w(DlvI59QN73`b`@u=K#ho6i7vJQA1j)=Twi>rt
zI|NB;r!^(tGE1q5eXB8X?9$0qVZTL{tC7g>nNx2OBw6OIpJVnc{i8hgE@1{S;GOf4Lyh1RVq%iVyeWX!B9X=$&dIENARj5j|3$2iG=MV?sy0y{Ya*Y
z4MTKfV3il+MeV1Ed#KTO%*Vzq;*8Yw4~I_kt?Js%EI#5#_S3
zg+WkFi!(js(Ao3dDX{;4@i;V`Z^RflsJowjOn~`DmgmY>Qmg;x~lp
z&@MyXXWyC-22_!}TO>w-520~>XmUrXdscRW2mZ=cQR)PqGdy+YTHIbWPs
za3keP$nV^$ux=~nN?UmP{hRL4r$%*^i$-=z_!j=!XCuPqb#*Pr8&p!U1k=a6Zew~{
zB;)uD1ww2`TE$=428JDfQ+17$sn4<2uTn17%X@)8vbVIB!J#5cPF(LLlW?PxUY4mC
zy^W6Q3|0Y)F^Mb`w8!J)=?W5m+!t?aLU!i|m7sCYR;O;b(qhQ>`&RI?NkwynrfIB9
z^-j=4Qpb?huN72u0{6Jr3$W_V{>F%4R8B=a|$sLQLJjNvP$y7!XN5A
zossz2S>D&7ARh3OjZ6Z>vhNoDO4kI$YOoSF196+>qCm
zIU2Kqo3KG1YknWJ8qA5W_dKk2@j$YPS+Wb8kZ1j=aMf%&40fJLCql9f)v`)-
zPc#3o7lQ4Y*1y&b)<4_Id@J(n20f`9-thtz4~Wh*+dSf!IQzd~7IW`OAb7{67yTTpsol`XF2)ygi^C)qisy-q
z65t%3m5K+TbJd(P)%xiXlQs@GkdC#8mS#z@ostmJk*TqIh|tr9e1j6R>gi-S^QiT0
zidXj}OE^uj&V)0ew+=vY^7Ym0oly&B&6VR?VAl~IJ}zb-8qvx+pReu9nvpuN!}Z}d
zzXz(~Z1kTU<_5QFb3;50F!vJW54Xpusc6|tLbzl`2H8i`-^{zjyhk&21&)+5kAvQ8
z-&w0~H`b@UzxH~>Lp}HHa{#K~Jbe0S%7B)V5l`4_prY$^inLFB4hblus?>z@uZ0@-
zXZV5QTCKE|ed%k3J`a>x?5tk3(gHo6pOP=_Zfd+|{=01F~w{fn~
z>q$JdtV8C}*@d4J#Fp}glFrMY<=(X79hWhWes@X>Qu!|H8kC~@i69)J
zdIx6qdxgMfj-N3HMp2}k%YO2(IIC)W1pr9bK(EqH=?y^koG`A~9G>5oXU{V^mwGMx
z=#6jAA4R?3Qbfy4+oXBUD{_k^P3QyeItSM-6gxf9hUxG!tHIM~m*E`zSl^4LVi18L
zyQ+Quug29M0OVn$(TOZ#K-UeVxIhicJm6U#s&vSIX#vz30C`|a@z+aLL{Yw=u#h-v
z|FB?~_DQKy?zqG##{sN+-z%TszgVhytT&C#KanqeE&ce!&+h@rW5MBc0~}w)_gxUaUXxy)ekzMgV)Z0YGB;|x++!${K#|b9Q@GtnornzLvc&@g!6aJ*u
zV=)XxLOby*o#d{?@ZLMzUOGb5vUU}isD*5~pLZ<|a`GGEttjuXR<($eGJflM33`(F
zN}CCmep4FCsY9k(Y*ss&zEn2f(N##Fk}OP3Ii)4;M>0qs{A$cp`m?W_nUxA%wZY*|
z5OMHXY2@3Oo!3x`T8h_VhxYYbu7&06*0Ef=*3!CQnF9SVjVCv#h;R?jA0L@@$0xLOX%Q1BXxXY)j_=ah(N2hcl&$QjiqV|BIvR$jy>!>`xU{ZV3!!JVsDJ;_
zH!iSlw1q7Fs&I^T5TIw5NMY#S0~o(mysxm86wT(`*V>-|=ky}*4ioJri}gzuIFL8M
zSIbKkUqSHmOfo(=&X9INZDB2A{PTuBT3cxC#EtNS|lcdH*?K_E63AOGkq=b$6&Yhf>?f
zD5c+Wqy?rahNn%F*khja{R5o&ZVc7?Tk`K3aL0?Y#>kBPS`VH-ukVrEUDNLAn1{(0D~Dln`{CXPR)zlMNZXwsDnlv~7*Ymm~K
zN}OnM7{_1Il(gnYYpJ
zL4O-mho!9Z85_lTjq%=6{4>}s~PhNxl4
z;~(KzG~BS%nwedD^Na_!TexlVj+&h!^pc5cc6fGsLeOSPDwA_DoI{S@U4MteNA5Ej
zpMe9Zt+a(28rHooGWbL(_iak4LV~*MWcyFO#}BZg0Qck;E?>pFl~{?Lq#a0_eS4?^
zUle0sT&lD_PypBimB!yDe}LFHM6WaB!?FJKdnZi~b%_k<#@$^(miJYsPA=XA#s@~f
zex4xjxwh_pD-(HyfggvaVcd19-Z})WB=F~onpH=G9