Pytorch Tutorial 4- Solving Kaggle Pima Diabetes Prediction Using ANN With PyTorch Library

[@taherkayamkhani8214]

You are recording videos at midnight during the weekend. I am speechless, I have no words for your appreciation sir. SALUTE

[@SALESENGLISH2020]

Krish makes difficult things simple. That is the best way to teach.

[@midhileshmomidi2434]

Thanks Krish Its a complete Data Science package in this channel

[@rishiar4]

In order to not get those warnings, you can always use the following lines of code. from warnings import filterwarnings filterwarnings('ignore') This will ignore the unnecessary warnings.

[@manojtaleka954]

Thank you Krish for a wonderful video on using PyTorch for solving classification problem using ANN

[@ratulghosh3849]

Excellent tutorial with respect to PyTorch Library. Keep up the good work sir!

[@swarnalaxmipanda6964]

Thank you Krish Sir.. These tutorials helped me a lot..

[@thomsonblaze]

Great job, this came in very handy on dealing with pytroch

[@AmitSingh-wi8iv]

Thank you so much krish sir🙏🙏 it's very easy way to understand each line of code and concept 🙏once again Thank you so much sir

[@raunaksingh6]

Thank you Krish for sharing your knowledge

[@ajinkyapahinkar6463]

Would just like to to point out about loss.item() - After the loss is calculated it is stored as a single value inside a 1-D tensor. Loss.item() converts this single value in a 1-D tensor to a float value so that when you print it it doesn’t show as a tensor but a float value. Also appending loss as a tensor to the final_losses list would make it unable to plot via matplotlib hence there also we use loss.item before appending it to the list

[@pponvannan]

so simple and nicely explained.

[@thusharadunumalage709]

Thank you very much 👍

[@upendravijay1428]

in the forward pass, on last layer 'x=self.out(x)' you haven't applied sigmoid why?

[@sourabhagarwal4852]

because sir has used cross entropy as loss function where probability is default .... If sir had used BCEloss(i.e binary_cross entropy ) then sigmoid function will be applied.

[@sarbanidasgupta9595]

The tensor in the predict function is giving some numbers how do we interpret them?Doesnt look like probabilities as they are greater than one or negative.If we want to predict probabilities instead of 0,1 how do we do that?

[@prathameshsamant30]

For this statement( y_train=torch.LongTensor(y_train)) , I am getting error as 'can't convert np.ndarray of type numpy.object_. The only supported types are: float64, float32, float16, complex64, complex128, int64, int32, int16, int8, uint8, and bool'. Please guide me.

[@karthiavenger4577]

do you find it how you resolved i m getting same error !

[@narayans.adhikari262]

use: y_train=torch.LongTensor(y_train.values)

[@saipranavreddy4056]

import numpy as np you wont get it

[@saipranavreddy4056]

​@@karthiavenger4577 import numpy as np you wont get it

[@raina1696]

Sir i did not get that seaborn error , it worked even when Outcome feature has numerical values,i did in google colab

[@Audalingo_J]

why did you use model.forward(X_train) instead of just passing data to model(X_train)? I noticed that most people avoid using the .forward approach.

[@rishikambhampati2862]

Hello, thanks for the series Krish. However, when I try to predict the output for a data point as you did for each of the entries in X_test, 'Not Implemented error' is being raised during the run time. can you please help me with this. Thanks

[@adhirajmajumder]

Thank you so much sir... Actually I want to learn pytorch instead of tensor flow because it doesn't support on my machine...

[@manishsharma2211]

You can use google colab or kaggle

[@dhirajdhakal6350]

Hello Krish, don't we need to scale down the data ?

[@thepresistence5935]

Binary Cross entropy use panlame, 2 categories thana iruku

[@saurrav3801]

Is sklearn and pytorch are having same functionality or both are different?

[@apantazopoul]

I know that the purpose of this tutorial is pytorch but this task needs better feature engineering. Zeros represent missing values because there is no possibility for a person to have 0 glucose measurement for example.

[@krishnaik06]

True ...in the next video we will do more feature engineering stuff

[@sushilbastola8940]

How can i predict remaining useful life with training set, test set and actual RUL values given , Please help me

[@rohitmohite7997]

Hi @Krish sir, I am getting this error using for loop of epochs. ..............forward() takes 1 positional argument but 2 were given

[@shaikrasool1316]

Super

[@aniketanvekar5462]

Krish sir make more videos on pytorch

[@sippee93]

ValueError: Expected input batch_size (614) to match target batch_size (154). getting an error using the same dataset

[@dmlane_sougata]

Sir, Please upload CNN videos using PyTorch.

[@saurrav3801]

Bro how to use sklearn ml algorithms before making a pytorch model ....

[@vincecarter7500]

cant you get a better model using a ml model instead of an ann?

[@swapnadeepsarkar5850]

Sir can you make a tutorial on Face recognition using Siamese Neural networks...

[@suryagangadhar1735]

Hiii sir, can u start Julia after pytorch

[@Krishnaverma-xw7oe]

Hi Sir, I'm getting this error: ValueError Traceback (most recent call last) in 1 ##Creating tensors ----> 2 X_train = torch.Tensor(X_train) 3 X_test = torch.FloatTensor(X_test) 4 y_train = torch.LongTensor(y_train) 5 y_test = torch.LongTensor(y_test) ValueError: could not determine the shape of object type 'DataFrame' I've different dataset

[@cometolearn8598]

It needs to be FloatTensor(X_train) not Tensor(X_train)

[@suryakirannvs]

Not sure how but followed same steps and it took 1's and 0's for me

[@lakshmitejaswi7832]

Implement word2vec in pytorch

[@deepankarsharma7840]

why did we do i+=1 at : kzbin.info/www/bejne/pYSrqI19ab1mkNE

[@mdtarique.ansari3802]

Tu dekh kr q padha h bhaye 🥲

[@gullyboy9226]

import matplotlib.pyplot as plt %matplotlib inline plt.plot(range(epochs),final_losses) plt.ylabel('Loss') plt.xlabel('Epoch') i got error in that portion can you help me please

[@harshvardhansurana1968]

I am also getting error in this part, could you tell what is it?