InternalError when trying to predict with an endpoint on AWS - amazon-web-services

Question

I use AWS SageMaker for a ML project and my TrainingJob for image classification has successfully finished. I use a separate inference.py for the inference, that looks like this:
import json
import logging
import sys
import os
import torch
import torch.nn as nn
import torch.nn.functional as F
import torchvision.models as models
import torchvision.transforms as transforms
import torchvision.transforms as T
from PIL import Image
import io
import requests
logger = logging.getLogger(__name__)
logger.setLevel(logging.DEBUG)
logger.addHandler(logging.StreamHandler(sys.stdout))
# define model
def Net():
model = models.__dict__['resnet50'](pretrained=True)
for param in model.parameters():
param.requires_grad = False
num_features = model.fc.in_features
model.fc = nn.Sequential(
nn.Linear(num_features, int(num_features/2)),
nn.Linear(int(num_features/2), 5))
return model
# load model parameters
def model_fn(model_dir):
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model = Net().to(device)
with open(os.path.join(model_dir, "model.pth"), "rb") as f:
model.load_state_dict(torch.load(f))
model.eval()
return model
# deserialize input
def input_fn(request_body, content_type):
if content_type == 'image/jpeg':
img = Image.open(io.BytesIO(request_body))
return img
else:
raise ValueError("This model only supports jpeg input")
# inference
def predict_fn(input_object, model):
transform = T.Compose([
transforms.Resize((224, 224)),
transforms.ToTensor(),
])
input_object=transform(input_object)
input_object=input_object.unsqueeze(0)
with torch.no_grad():
prediction = model(input_object)
return prediction
I created a predictor and made a prediction like this:
from sagemaker.pytorch import PyTorchModel
from sagemaker.predictor import Predictor
model_location=estimator.model_data
pytorch_model = PyTorchModel(model_data=model_location,
role=sagemaker.get_execution_role(),
entry_point='inference.py',
py_version='py3',
framework_version='1.4')
predictor = pytorch_model.deploy(initial_instance_count=1, instance_type='ml.m5.large')
import io
with open("test1.jpg", "rb") as image:
img = image.read()
response=predictor.predict(img, initial_args={"ContentType": "image/jpeg"})
I got the following error message:
InternalFailure Traceback (most recent call last)
/tmp/ipykernel_7063/3976847146.py in <cell line: 1>()
----> 1 response=predictor.predict(img, initial_args={"ContentType": "image/jpeg"})
~/anaconda3/envs/python3/lib/python3.8/site-packages/sagemaker/predictor.py in predict(self, data, initial_args, target_model, target_variant, inference_id)
159 data, initial_args, target_model, target_variant, inference_id
160 )
--> 161 response = self.sagemaker_session.sagemaker_runtime_client.invoke_endpoint(**request_args)
162 return self._handle_response(response)
163
~/anaconda3/envs/python3/lib/python3.8/site-packages/botocore/client.py in _api_call(self, *args, **kwargs)
393 "%s() only accepts keyword arguments." % py_operation_name)
394 # The "self" in this scope is referring to the BaseClient.
--> 395 return self._make_api_call(operation_name, kwargs)
396
397 _api_call.__name__ = str(py_operation_name)
~/anaconda3/envs/python3/lib/python3.8/site-packages/botocore/client.py in _make_api_call(self, operation_name, api_params)
723 error_code = parsed_response.get("Error", {}).get("Code")
724 error_class = self.exceptions.from_code(error_code)
--> 725 raise error_class(parsed_response, operation_name)
726 else:
727 return parsed_response
InternalFailure: An error occurred (InternalFailure) when calling the InvokeEndpoint operation (reached max retries: 4): An exception occurred while sending request to model. Please contact customer support regarding request 34c901aa-76... .
I also tried this code in different accounts, but the issue remained.