Skip to main content

Script Nodes

Script Nodes allow users to integrate custom Python scripts into aiXplain pipelines. They are particularly useful when:

Modifying or transforming data between pipeline nodes. Implementing custom logic not covered by built-in nodes. Formatting outputs to meet specific requirements. This guide demonstrates how to create a pipeline with a Script Node, modify the output of a speech recognition model, and execute a custom Python script.

Creating a Pipeline with a Script Node​

A Script Node allows users to insert a Python script within an aiXplain pipeline. Below, we create a pipeline with the following flow:

input_node -> asset_node (speech recognition) -> script_node (custom processing) -> output_node

Initialize the Pipeline​

from aixplain.factories import PipelineFactory

pipeline = PipelineFactory.init("Example Scripted Pipeline")

input_node = pipeline.input()
asset_node = pipeline.speech_recognition("60ddefab8d38c51c5885ee38")

Writing a Custom Script for the Script Node​

In this example, we write a simple script that modifies the output of the speech recognition node by formatting it into JSON.

with open("example_script.py", "w") as f:
  f.write("""
def main(speakers):
    # build the response
    response = []
    for i, speaker in enumerate(speakers):
        print(f"Processing speaker at index={i}")
        data = speaker["data"]
        # convert data into a json formatted data
        json_data = f'{{"url": "{data}"}}'
        response.append(
            {
                "index": i,
                "success": True,
                "input_type": "text",
                "is_url": False,
                "details": {},
                "data": json_data,
                "input": json_data,
            }
        )
    return response
""")

Adding the Script Node to the Pipeline​

Once the script is ready, we integrate it into the pipeline using a Script Node.

script_node = pipeline.script(script_path="example_script.py")

The script above accepts an input parameter called speakers and an output parameter called data, we need to create them in our node specs accordingly.

script_node.inputs.create_param("speakers", data_type="text")
script_node.outputs.create_param("data", data_type="text")

Linking Nodes Together​

input_node --(input-source_audio)--> asset_node --(data-speakers)--> script_node --(data-output) --> output_node
input_node.outputs.input.link(asset_node.inputs.source_audio)
asset_node.outputs.data.link(script_node.inputs.speakers)
output_node = script_node.use_output("data")

Running the Pipeline​

Now that the pipeline is fully configured, save and run it with an audio input.

Save the Pipeline​

pipeline.save(save_as_asset=True)

Run the Pipeline​

pipeline.run("https://homepage.ntu.edu.tw/~karchung/miniconversations/mc1.mp3", **{ "version": "3.0" })
Show output

Delete the Pipeline (Optional)​

If the pipeline is no longer needed, it can be deleted.

pipeline.delete()

Script Nodes provide a flexible way to incorporate custom processing within aiXplain pipelines. They allow developers to:

  • Modify intermediate outputs.
  • Format and structure data before passing it to the next node.
  • Implement custom logic within AI-driven workflows.

By leveraging Script Nodes, you can enhance the capabilities of aiXplain pipelines for a variety of use cases.