🦜 | A mock worker that can be used to emulate the behavior of different types of endpoints. Docker Link
{"input":
{
"mock_return": "Anything here will be returned as the output of the worker.",
"mock_delay": 0, // The number of seconds to wait before returning output, raising error or crashing. If generator is enabled, this is the delay between each yeild.
"mock_progress": {
"updates": [], // A list of progress updates to send back to the RunPod API.
"wait_time": 0 // The number of seconds to wait before sending a progress update.
},
"mock_error": false, // If true, the worker will raise an error.
"mock_crash": false, // If true, the worker will crash (kills the processes)
"mock_refresh": false, // If true, the refresh_worker flag is enabled.
"mock_external": {
"error_job_return": false, // If true, the job will fail to return the job results.
},
"mock_logs": [ // A list of log messages to generate, these are sent at the start of the job.
{
"level": "info", // The log level to use. Options: debug, info, warn, error
"message": "This is a log message." // The log message to send.
}
]
}
}
To test the generator handler override the docker command with the following:
python3 -u /handler.py --generator
python3 -u /handler.py --async_generator
To return the stream aggregation include --return_aggregate_stream
argument.
# To test multiple yeild outputs, set the `mock_return` to a list of values.
{
"input": {
"mock_return": ["value1", "value2", "value3"],
"mock_delay": 0,
"mock_error": false,
"mock_crash": false
}
}