def build_new_network():
# Gets a valid model before proceeding
partial_model = helper.get_model(None)
# Gets correct learn rate and hidden layers before proceeding
learn_rate, hidden_layers = helper.get_network_inputs()
# Buils the network
model, criterion, optimizer = helper.build_network(learn_rate,
hidden_layers,
partial_model)
model.criterion = criterion
model.optimizer = optimizer
return model
def get_model():
model_file = '../data/all-files-for-ml/' + 'all_mnb' + '.pkl'
mnb = joblib.load(open(model_file, 'rb'))
data = pd.read_csv("../data/all-files-for-ml/all_x.csv")
df = pd.DataFrame(data)
cols = df.columns
features = cols # = symptoms
features_raw = [str(features[x]) for x in range(len(features))]
# convert feature array into dict of symptom: index
feature_dict = {}
for i,f in enumerate(features):
feature_dict[f] = i
return mnb, features, feature_dict
MODEL, LABELS, LABELS_DICT = get_model()
@app.route('/', defaults={'path': ''})
@app.route('/<path:path>')
def serve(path):
path_dir = os.path.abspath("../app/backend") #path react build
if path != "" and os.path.exists(os.path.join(path_dir, path)):
return send_from_directory(os.path.join(path_dir), path)
else:
return render_template("react.html")
@limiter.limit("500 per hour")
@app.route('/api/labels', methods = ['GET'])
def labels():
from helper import get_tensor, get_model
import json
model = get_model()
with open('mappings.json', 'r') as f:
mappings = json.load(f)
with open('remidies.json', 'r') as f:
remidies = json.load(f)
def get_disease_name(image_bytes):
tensor = get_tensor(image_bytes)
outputs = model.forward(tensor)
_, prediction = outputs.max(1)
category = prediction.item()
disease = mappings[str(category)]
return disease, remidies[disease]
type=int)
parser.add_argument('--epochs',
help="Amount of epochs the network is trained (default=5)",
default=5,
type=int)
parser.add_argument('--gpu',
help="Use the gpu to accelerate training",
action='store_true')
args = parser.parse_args()
train_dir = args.data_directory[0] + '/train'
valid_dir = args.data_directory[0] + '/valid'
(trainloader, validationloader,
class_to_idx) = helper.get_dataloaders(train_dir, valid_dir)
if not torch.cuda.is_available() and args.gpu:
raise (
"No gpu available to train the network. Please remove the --gpu argument to train using the cpu"
)
device = ('cuda' if args.gpu else 'cpu')
(model, optimizer) = helper.get_model(args.arch, args.learning_rate,
args.hidden_units)
helper.train_model(model, trainloader, validationloader, optimizer, device,
args.epochs)
if args.save_dir != None:
model.class_to_idx = class_to_idx
torch.save(model, args.save_dir)
import os, time
import helper
from dataset import classes_names
from PIL import Image
import numpy as np
import pandas as pd
test_data_dir = 'test'
if __name__ == '__main__':
model_path = helper.train_and_evaluate()
model = helper.get_model(model_path)
img_names = []
img_pils = []
for img_name in os.listdir(test_data_dir):
img = Image.open(os.path.join(test_data_dir, img_name))
img = img.convert('RGB')
img_names.append(img_name)
img_pils.append(img)
predicts = helper.predict(model, img_pils) # outputs are classes indexes
predict_classes = [classes_names[cls] for cls in predicts] # names here
# img_names = ['1.png', '2.png']
# predict_classes = ['hello', 'world']
pd_data = {'file': img_names, 'species': predict_classes}
df = pd.DataFrame(pd_data)