The purpose is to build base code for user using Pytorch library
Since I am coming from C and C++ background , to see "main" files make me happy :)
Therefore , I seperate the related class and files into folders such as
required_classes auxiliry classses
dataset contains raw and processed images.
main_function main python file that will be runned.
- python3.x
- torch
- numpy
- time
- torch.utils.data
- torchvision.datasets
- shutil
#Dataset
TRAIN_PATH="C:/Users/talha/Desktop/Dataset/train"
TEST_PATH="C:/Users/talha/Desktop/Dataset/test"
VALIDATION_PATH="C:/Users/talha/Desktop/Dataset/validation"
model = Model()
my=ModelTrain()
batch_size = 64
my.dataset_load(TRAIN_PATH, TEST_PATH, VALIDATION_PATH, batch_size)
my.GPU_Usage(True)
print(my.use_gpu)
#Dataset info
my.dataset_info("all")
#logging
import sys
sys.path.insert(1, 'C:/Users/talha/Desktop/Github_sources/pycodes')
import tk_logging
tk = tk_logging.Tk_logging("ModelTrain","log.txt","DEBUG")
loggerr = tk.logger
my.logger = loggerr
#Debugger
from icecream import ic
icdebugger = ic
icdebugger.configureOutput(prefix='') # writes Debug before each printing
icdebugger.enable()
my.debugger = icdebugger
#Parameters
learning_rate = 1e-3
n_epochs = 21
optimizer = optim.Adam(params=model.parameters(), lr=learning_rate,weight_decay=1e-5) #,,weight_decay=1e-5
criteria = nn.MSELoss()
loss_values,accuracy_values=my.train(model, criteria, n_epochs, optimizer)
for any problem , don't hesitate to contact me from Linkedin 👍 👍