Create a Lambda layer for 3rd party library used in Lambda function

Next, we're going to setup a lambda layer to execute lambda function. A Lambda layer is an archive containing additional code, such as libraries, dependencies, or even custom runtimes.

1. Click "Lambda" in "Services" to enter the console of Lambda function.

2. Click "Layers" in the sidebar of console.

3. Click "Create layer" to create a layer to hold all the necessary libraries (e.g. TensorFlow-Lite).

4. Fill "Name": "tf-lite" and click "Upload" to upload the following zip file. Then select "Python 3.9" in the dropdown list under "Compatible runtimes". Finally, click "Create".

21MB

lambda_layer_tflite.zip

archive

5. Then you can find it in the console.