Running pytorch program from shell

I pulled a pytorch based program from github, and they execute using a shell script. The model trains, and then terminates. I then inserted at the end:, '/saved_model/test.bin')

It saves to the file, but then when i try to load the file in a python console, it does not retain the class name:

'model = torch.load(‘saved_model/test.bin’)`



and it states there is not eval.

What am i doing wrong? The code is below, and said line is at the bottom prior to the last if statement. My goal is to save the model, load it in a console, and start passing new data through it to make predictions. The author of this model is not answering which is why I am here.

import sys
import os
import torch
import random
import numpy as np
from tqdm import tqdm
from torch.autograd import Variable
from torch.nn.parameter import Parameter
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
import math
import pdb
from DGCNN_embedding import DGCNN
from mlp_dropout import MLPClassifier, MLPRegression
from sklearn import metrics
from util import cmd_args, load_data
import matplotlib.pyplot as plt

class Classifier(nn.Module):
    def __init__(self, regression=False):
        super(Classifier, self).__init__()
        self.regression = regression
        if == 'DGCNN':
            model = DGCNN
            print('unknown gm %s' %

        if == 'DGCNN':
            self.gnn = model(latent_dim=cmd_args.latent_dim,
        out_dim = cmd_args.out_dim
        if out_dim == 0:
            if == 'DGCNN':
                out_dim = self.gnn.dense_dim
                out_dim = cmd_args.latent_dim
        self.mlp = MLPClassifier(input_size=out_dim, hidden_size=cmd_args.hidden, num_class=cmd_args.num_class, with_dropout=cmd_args.dropout)
        if regression:
            self.mlp = MLPRegression(input_size=out_dim, hidden_size=cmd_args.hidden, with_dropout=cmd_args.dropout)

    def PrepareFeatureLabel(self, batch_graph):
        if self.regression:
            labels = torch.FloatTensor(len(batch_graph))
            labels = torch.LongTensor(len(batch_graph))
        n_nodes = 0

        if batch_graph[0].node_tags is not None:
            node_tag_flag = True
            concat_tag = []
            node_tag_flag = False

        if batch_graph[0].node_features is not None:
            node_feat_flag = True
            concat_feat = []
            node_feat_flag = False

        if cmd_args.edge_feat_dim > 0:
            edge_feat_flag = True
            concat_edge_feat = []
            edge_feat_flag = False

        for i in range(len(batch_graph)):
            labels[i] = batch_graph[i].label
            n_nodes += batch_graph[i].num_nodes
            if node_tag_flag == True:
                concat_tag += batch_graph[i].node_tags
            if node_feat_flag == True:
                tmp = torch.from_numpy(batch_graph[i].node_features).type('torch.FloatTensor')
            if edge_feat_flag == True:
                if batch_graph[i].edge_features is not None:  # in case no edge in graph[i]
                    tmp = torch.from_numpy(batch_graph[i].edge_features).type('torch.FloatTensor')

        if node_tag_flag == True:
            concat_tag = torch.LongTensor(concat_tag).view(-1, 1)
            node_tag = torch.zeros(n_nodes, cmd_args.feat_dim)
            node_tag.scatter_(1, concat_tag, 1)

        if node_feat_flag == True:
            node_feat =, 0)

        if node_feat_flag and node_tag_flag:
            # concatenate one-hot embedding of node tags (node labels) with continuous node features
            node_feat =[node_tag.type_as(node_feat), node_feat], 1)
        elif node_feat_flag == False and node_tag_flag == True:
            node_feat = node_tag
        elif node_feat_flag == True and node_tag_flag == False:
            node_feat = torch.ones(n_nodes, 1)  # use all-one vector as node features
        if edge_feat_flag == True:
            edge_feat =, 0)

        if cmd_args.mode == 'gpu':
            node_feat = node_feat.cuda()
            labels = labels.cuda()
            if edge_feat_flag == True:
                edge_feat = edge_feat.cuda()

        if edge_feat_flag == True:
            return node_feat, edge_feat, labels
        return node_feat, labels

    def forward(self, batch_graph):
        feature_label = self.PrepareFeatureLabel(batch_graph)
        if len(feature_label) == 2:
            node_feat, labels = feature_label
            edge_feat = None
        elif len(feature_label) == 3:
            node_feat, edge_feat, labels = feature_label
        embed = self.gnn(batch_graph, node_feat, edge_feat)
        return self.mlp(embed, labels)

    def output_features(self, batch_graph):
        feature_label = self.PrepareFeatureLabel(batch_graph)
        if len(feature_label) == 2:
            node_feat, labels = feature_label
            edge_feat = None
        elif len(feature_label) == 3:
            node_feat, edge_feat, labels = feature_label
        embed = self.gnn(batch_graph, node_feat, edge_feat)
        return embed, labels

def loop_dataset(g_list, classifier, sample_idxes, optimizer=None, bsize=cmd_args.batch_size):
    total_loss = []
    total_iters = (len(sample_idxes) + (bsize - 1) * (optimizer is None)) // bsize
    pbar = tqdm(range(total_iters), unit='batch')
    all_targets = []
    all_scores = []

    n_samples = 0
    for pos in pbar:
        selected_idx = sample_idxes[pos * bsize : (pos + 1) * bsize]

        batch_graph = [g_list[idx] for idx in selected_idx]
        targets = [g_list[idx].label for idx in selected_idx]
        all_targets += targets
        if classifier.regression:
            pred, mae, loss = classifier(batch_graph)
            all_scores.append(pred.cpu().detach())  # for binary classification
            logits, loss, acc = classifier(batch_graph)
            all_scores.append(logits[:, 1].cpu().detach())  # for binary classification

        if optimizer is not None:

        loss =
        if classifier.regression:
            pbar.set_description('MSE_loss: %0.5f MAE_loss: %0.5f' % (loss, mae) )
            total_loss.append( np.array([loss, mae]) * len(selected_idx))
            pbar.set_description('loss: %0.5f acc: %0.5f' % (loss, acc) )
            total_loss.append( np.array([loss, acc]) * len(selected_idx))

        n_samples += len(selected_idx)
    if optimizer is None:
        assert n_samples == len(sample_idxes)
    total_loss = np.array(total_loss)
    avg_loss = np.sum(total_loss, 0) / n_samples
    all_scores =
    # np.savetxt('test_scores.txt', all_scores)  # output test predictions
    if not classifier.regression and cmd_args.printAUC:
        all_targets = np.array(all_targets)
        fpr, tpr, _ = metrics.roc_curve(all_targets, all_scores, pos_label=1)
        auc = metrics.auc(fpr, tpr)
        avg_loss = np.concatenate((avg_loss, [auc]))
        avg_loss = np.concatenate((avg_loss, [0.0]))
    return avg_loss

if __name__ == '__main__':

    train_graphs, test_graphs = load_data()
    print('# train: %d, # test: %d' % (len(train_graphs), len(test_graphs)))

    if cmd_args.sortpooling_k <= 1:
        num_nodes_list = sorted([g.num_nodes for g in train_graphs + test_graphs])
        cmd_args.sortpooling_k = num_nodes_list[int(math.ceil(cmd_args.sortpooling_k * len(num_nodes_list))) - 1]
        cmd_args.sortpooling_k = max(10, cmd_args.sortpooling_k)
        print('k used in SortPooling is: ' + str(cmd_args.sortpooling_k))

    classifier = Classifier()
    if cmd_args.mode == 'gpu':
        classifier = classifier.cuda()

    optimizer = optim.Adam(classifier.parameters(), lr=cmd_args.learning_rate)

    train_idxes = list(range(len(train_graphs)))
    best_loss = None
    for epoch in range(cmd_args.num_epochs):
        avg_loss = loop_dataset(train_graphs, classifier, train_idxes, optimizer=optimizer)
        if not cmd_args.printAUC:
            avg_loss[2] = 0.0
        print('\033[92maverage training of epoch %d: loss %.5f acc %.5f auc %.5f\033[0m' % (epoch, avg_loss[0], avg_loss[1], avg_loss[2]))

        test_loss = loop_dataset(test_graphs, classifier, list(range(len(test_graphs))))
        if not cmd_args.printAUC:
            test_loss[2] = 0.0
        print('\033[93maverage test of epoch %d: loss %.5f acc %.5f auc %.5f\033[0m' % (epoch, test_loss[0], test_loss[1], test_loss[2]))

    with open( + '_acc_results.txt', 'a+') as f:
        f.write(str(test_loss[1]) + '\n')

    if cmd_args.printAUC:
        with open( + '_auc_results.txt', 'a+') as f:
            f.write(str(test_loss[2]) + '\n')

    model = Classifier(), '/home/anthony/PycharmProjects/DGCNN/pytorch_DGCNN-master/saved_model/test.bin')

    if cmd_args.extract_features:
        features, labels = classifier.output_features(train_graphs)
        labels = labels.type('torch.FloatTensor')
        np.savetxt('extracted_features_train.txt',[labels.unsqueeze(1), features.cpu()], dim=1).detach().numpy(), '%.4f')
        features, labels = classifier.output_features(test_graphs)
        labels = labels.type('torch.FloatTensor')
        np.savetxt('extracted_features_test.txt',[labels.unsqueeze(1), features.cpu()], dim=1).detach().numpy(), '%.4f')

Saves the model.state_dict() (which is best practice).

It saves to the file, but then when i try to load the file in a python console, it does not retain the class name:

model = torch.load(‘saved_model/test.bin’)

Loads the saved state dict and assigns it to model.

This is because you only ave the state dict.

You likely want

model = Classifier()
sd = torch.load(...)

I make this kind of error (that an object isn’t quite what the user thinks it is), too. When things go wrong, I tend to print things a lot to see if they look like what I expect.

Best regards


Thank you for the info, I will give it a shot. The program though is started by a shell script, is there a better way to run it then in the terminal? Because by doing that, with the code you gave me, will not give me the desired output model. I tried to use subprocess, but I am still not seeing a way to put model = ... at the end of my script, and have it saved as a variable in the python console. Did that make sense at all?

Also, it’s stating that when I use model.load_state_dict(sd), the return is TypeError: load_state_dict() missing 1 required positional argument: 'state_dict'

I double checked, and the code as presented above works for me.