I've been trying to deploy my model to the AI platform for Prediction through the console on my vm instance, but I've gotten the error "(gcloud.beta.ai-platform.versions.create) Create Version failed. Bad model detected with error: "Failed to load model: Unexpected error when loading the model: problem in predictor - ModuleNotFoundError: No module named 'torchvision' (Error code: 0)"
I need to include both torch and torchvision. I followed the steps in this question Cannot deploy trained model to Google Cloud Ai-Platform with custom prediction routine: Model requires more memory than allowed, but I couldn't fetch the files pointed to by user gogasca. I tried downloading this .whl file from Pytorch website and uploading it to my cloud storage but got the same error that there is no module torchvision, even though this version is supposed to include both torch and torchvision. Also tried using Cloud AI compatible packages here, but they don't include torchvision.
I tried pointing to two separate .whl files for torch and torchvision in the --package-uris arguments, those point to files in my cloud storage, but then I got the error that the memory capacity was exceeded. This is strange, because collectively their size is around 130Mb. An example of my command that resulted in absence of torchvision looked like this:
gcloud beta ai-platform versions create version_1 \
--model online_pred_1 \
--runtime-version 1.15 \
--python-version 3.7 \
--origin gs://BUCKET/model-dir \
--package-uris gs://BUCKET/staging-dir/my_package-0.1.tar.gz,gs://BUCKET/torchvision-dir/torch-1.4.0+cpu-cp37-cp37m-linux_x86_64.whl \
--prediction-class predictor.MyPredictor
I've tried pointing to different combinations of .whl files that I obtained from different sources, but got either the no module error or not enough memory. I don't understand how the modules interact in this case and why the compiler thinks there is no such module. How can I resolve this? Or alternatively, how can I compile a package myself that include both torch and torchvision. Can you please give detailed answers because I'm not very familiar with package management and bash scripting.
Here's the code I used, torch_model.py:
from torch import nn
class EthnicityClassifier44(nn.Module):
def __init__(self, num_classes=2):
super().__init__()
self.conv1 = nn.Conv2d(3, 32, kernel_size=7, stride=1, padding=3)
self.maxpool1 = nn.MaxPool2d(kernel_size=2, stride=2)
self.conv22 = nn.Conv2d(32, 32, kernel_size=3, stride=1, padding=1)
self.maxpool2 = nn.MaxPool2d(kernel_size=2, stride=2)
self.conv3 = nn.Conv2d(32, 64, kernel_size=3, stride=1, padding=1)
self.maxpool3 = nn.MaxPool2d(kernel_size=2, stride=2)
self.conv4 = nn.Conv2d(64, 128, kernel_size=3, stride=1, padding=1)
self.maxpool4 = nn.MaxPool2d(kernel_size=2, stride=2)
self.relu = nn.ReLU(inplace=False)
self.fc1 = nn.Linear(8*8*128, 128)
self.fc2 = nn.Linear(128, 128)
self.fc4 = nn.Linear(128, num_classes)
def forward(self, x):
x = self.relu(self.conv1(x))
x = self.maxpool1(x)
x = self.relu(self.conv22(x))
x = self.maxpool2(x)
x = self.maxpool3(self.relu(self.conv3(x)))
x = self.maxpool4(self.relu(self.conv4(x)))
x = self.relu(self.fc1(x.view(x.shape[0], -1)))
x = self.relu(self.fc2(x))
x = self.fc4(x)
return x
This is predictor_py:
from facenet_pytorch import MTCNN, InceptionResnetV1, extract_face
import torch
import torchvision
from torchvision import transforms
from torch.nn import functional as F
from PIL import Image
from sklearn.externals import joblib
import numpy as np
import os
import torch_model
class MyPredictor(object):
import torch
import torchvision
def __init__(self, model, preprocessor, device):
"""Stores artifacts for prediction. Only initialized via `from_path`.
"""
self._resnet = model
self._mtcnn_mult = preprocessor
self._device = device
self.get_std_tensor = transforms.Compose([
np.float32,
np.uint8,
transforms.ToTensor(),
])
self.tensor2pil = transforms.ToPILImage(mode='RGB')
self.trans_resnet = transforms.Compose([
transforms.Resize((100, 100)),
np.float32,
transforms.ToTensor()
])
def predict(self, instances, **kwargs):
pil_transform = transforms.Resize((512, 512))
imarr = np.asarray(instances)
pil_im = Image.fromarray(imarr)
image = pil_im.convert('RGB')
pil_im_512 = pil_transform(image)
boxes, _ = self._mtcnn_mult(pil_im_512)
box = boxes[0]
face_tensor = extract_face(pil_im_512, box, margin=40)
std_tensor = self.get_std_tensor(face_tensor.permute(1, 2, 0))
cropped_pil_im = self.tensor2pil(std_tensor)
face_tensor = self.trans_resnet(cropped_pil_im)
face_tensor4d = face_tensor.unsqueeze(0)
face_tensor4d = face_tensor4d.to(self._device)
prediction = self._resnet(face_tensor4d)
preds = F.softmax(prediction, dim=1).detach().numpy().reshape(-1)
print('probability of (class1, class2) = ({:.4f}, {:.4f})'.format(preds[0], preds[1]))
return preds.tolist()
@classmethod
def from_path(cls, model_dir):
import torch
import torchvision
import torch_model
model_path = os.path.join(model_dir, 'class44_M40RefinedExtra_bin_no_norm_7860.joblib')
classifier = joblib.load(model_path)
mtcnn_path = os.path.join(model_dir, 'mtcnn_mult.joblib')
mtcnn_mult = joblib.load(mtcnn_path)
device_path = os.path.join(model_dir, 'device_cpu.joblib')
device = joblib.load(device_path)
return cls(classifier, mtcnn_mult, device)
And setup.py:
from setuptools import setup
REQUIRED_PACKAGES = ['opencv-python-headless', 'facenet-pytorch']
setup(
name="my_package",
version="0.1",
include_package_data=True,
scripts=["predictor.py", "torch_model.py"],
install_requires=REQUIRED_PACKAGES
)
