When I try to run a processing job using the Serverless Inference Benchmarking tool, I run into this error: FileNotFoundError: [Errno 2] No such file or directory: '/opt/ml/processing/input/data\invoke_args_examples.pkl'.
When I follow my S3 bucket where the input/ folder is, it has the invoke_args_examples.pkl in an input_1/ subfolder rather than data/.
I followed the example outlined in the github repo.
example_invoke_file = convert_invoke_args_to_pkl(example_invoke_args)
processor = run_as_sagemaker_job(
role=role,
model_name=MODEL_NAME,
invoke_args_examples_file= example_invoke_file,
concurrency_benchmark_max_conc=[1, 10, 100]
)
print(
f"Once the job is finished, the outputs will be uploaded to {processor.latest_job.outputs[0].destination}"
)
I believe I need to change the folder to run the tool properly but I'm not sure how to do so.