Huge Difference in losses

S18MLOL0007 · July 12, 2020, 7:52am

losses in video lecture is 0.034 and mine is about 0.34…

Aayushkh_333 · July 12, 2020, 7:58am

Hey @S18MLOL0007, I request you to upload your code on google drive and share the link here. I need to look at the code to find out the details !

Thanks !

S18MLOL0007 · July 12, 2020, 10:29am

https://drive.google.com/file/d/12ki8jbYk1UdG__UAbSQ0GR0e3GYQgFhM/view?usp=sharing

Aayushkh_333 · July 12, 2020, 11:36am

Hey @S18MLOL0007, if you see this code :

github.com

prateek27/deep-neural-network/blob/master/Neural_Network_MLP_Final.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## MultiLayer Perceptron/ Feed Forward Architecture\n",
    "### Implementing a 3 Layer Architecture  from Scratch!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from matplotlib import pyplot as plt\n",
    "from sklearn.datasets import make_moons,make_circles,make_classification\n"

This file has been truncated. show original

There is not much difference at 500 epochs. Try to play with hyperparameters and you may achieve even less than what sir has achieved in the video. Also run your code for more epochs and see the final loss value that you obtain !

Since all the weights are randomly initialised , the exact values would never match. Just see that there is not a hugh difference.

I hope this resolves your doubt !
Please mark the doubt as resolved in your doubts section
Happy Learning

Aayushkh_333 · July 21, 2020, 5:17am

I hope I’ve cleared your doubt. I ask you to please rate your experience here
Your feedback is very important. It helps us improve our platform and hence provide you
the learning experience you deserve.

On the off chance, you still have some questions or not find the answers satisfactory, you may reopen
the doubt.