def submit():
global list_images
global student_image
global row
global row2
student_admno = E4.get()
student_name = E1.get()
student_class = E2.get()
student_rollno = E3.get()
if ((student_admno == '') or (not student_admno.isdigit())
or (student_name == '') or (student_class not in classes)
or (len(student_rollno) < 1) or (len(student_rollno) > 2)
or (len(list_images) != 50) or (not student_rollno.isdigit())):
showerror("Error Submit", "Something is wrong in your inputs..!!!")
root.focus_set()
elif (student_admno in unique_ids_list):
showerror("Error Submit",
"A student in this admission number already exists")
root.focus_set()
elif (askyesno(
"Confirm Submit",
"Hope you have entered right datas.\nAre you sure to submit ?")
):
global root_label
root.lift()
root_label.destroy()
root_label = Label(F0, text="Photo")
root_label.pack(pady=5)
root.geometry("400x300+500+50")
E1.delete(0, END)
E2.delete(0, END)
E3.delete(0, END)
E4.delete(0, END)
f = open(unique_ids, "a")
f.write(student_admno + '\n')
f.close()
for sheet in book.worksheets:
sheet['A{0}'.format(row)] = student_admno
sheet['B{0}'.format(row)] = student_name
book.save(os.path.join("Database", "Students", "Attendence.xlsx"))
row += 1
sheet2['A{0}'.format(row2)].value = student_admno
sheet2['B{0}'.format(row2)].value = student_name
sheet2['C{0}'.format(row2)] = "Students"
book2.save(os.path.join("Database", "teachers_and_students.xlsx"))
row2 += 1
cv2.imwrite(
os.path.join("Database", "Students", "Images", student_class,
student_rollno + ".jpg"), student_image)
trainer.train(list_images, int(student_admno))
else:
root.focus_set()
"baseline_ptcone20",
"baseline_ptcone30",
"baseline_ptcone40",
"baseline_ptvarcone20",
"baseline_ptvarcone30",
"baseline_ptvarcone40",
"baseline_eflowcone20_over_pt",
"trk_vtx_type",
]
options["training_split"] = 0.7
options["batch_size"] = 32
options["n_epochs"] = 10
options["n_layers"] = 3
options["hidden_neurons"] = 256
options["intrinsic_dimensions"] = 1024 # only matters for deep sets
options["output_neurons"] = 2
options["device"] = args.device
options["save_model"] = True
options[
"model_save_path"] = options["output_folder"] + "test_gru_model.pth"
options["train_BDT"] = True
t0 = time.time()
print("number of epochs planned:", options["n_epochs"])
print("input data:", options["input_data"].split("/")[-1])
print("batch_size:", options["batch_size"])
print("device:", args.device)
print("architecture:", options["architecture_type"])
trainer.train(options)
print("total runtime :", time.time() - t0)
torch.cuda.empty_cache()
data_filepath = "../colores.csv"
img_filepath = "../images/tanzania.jpg"
x = np.genfromtxt(data_filepath,
dtype=np.float128,
delimiter=",",
usecols=(0, 1, 2))
y = np.genfromtxt(data_filepath, dtype=np.string0, delimiter=",", usecols=(3))
class_array = []
new_y = []
for i in range(y.size):
if y[i] not in class_array:
class_array.append(y[i])
new_y.append([class_array.index(y[i])])
color_length = len(class_array)
new_y = np.array(new_y, dtype=float128)
x = x / 255.
new_y = new_y / float(color_length - 1)
NN = Neural_Network(3, 11, 1)
trainer = trainer(NN)
trainer.train(x, new_y)
print(get_predominant_color(NN, img_filepath, class_array))
torch.cuda.manual_seed_all(random_seed)
torch.backends.cudnn.deterministic = True
torch.backends.cudnn.benchmark = False
np.random.seed(random_seed)
# Setting the config for each stage
train_config = get_config(mode='train')
dev_config = get_config(mode='dev')
test_config = get_config(mode='test')
print(train_config)
# Creating pytorch dataloaders
train_data_loader = get_loader(train_config, shuffle=True)
dev_data_loader = get_loader(dev_config, shuffle=False)
test_data_loader = get_loader(test_config, shuffle=False)
# Trainer is a wrapper for model training, validating and testing
trainer = trainer(train_config,
train_data_loader,
dev_data_loader,
test_data_loader,
is_train=True)
# Build the model
trainer.build()
# Train the model (test scores will be returned based on val performance)
trainer.train()