Keras Tuner | Hyperparameter Tuning a Neural Network
Вставка
- Опубліковано 19 чер 2024
- KerasTuner is an easy-to-use, scalable hyperparameter optimization framework that solves the pain points of hyperparameter search. Easily configure your search space with a define-by-run syntax, then leverage one of the available search algorithms to find the best hyperparameter values for your models. KerasTuner comes with Bayesian Optimization, Hyperband, and Random Search algorithms built-in, and is also designed to be easy for researchers to extend in order to experiment with new search algorithms.
============================
Do you want to learn from me?
Check my affordable mentorship program at : learnwith.campusx.in
============================
📱 Grow with us:
CampusX' LinkedIn: / campusx-official
CampusX on Instagram for daily tips: / campusx.official
My LinkedIn: / nitish-singh-03412789
Discord: / discord
👍If you find this video helpful, consider giving it a thumbs up and subscribing for more educational videos on data science!
💭Share your thoughts, experiences, or questions in the comments below. I love hearing from you!
⌚Time Stamps⌚
00:00 - Intro
01:16 - Usage
02:25 - Code Example
#Keras #KerasTuner
You are the real data scientist. you are providing the best content regarding data science as compared to another channels.
You are just awesom. I've learned many things from your videos. Thank you so much. Lot of Love from London (UK). 👌👌👌👌👌
london mai wathman hai kya?
stop bragging abt yrself
i was afraid of learning Deep learning. because of you it become possible. Thank you!
The Scientist term fits you, thoroughly knowledgeable and well-researched information. Thank you 🙏
Hey, I am learning so much from all these videos and implementing them as much as I can. Will try this technique too.
Thank you so much for this insightful tutorial on hyperparameter tuning using Keras Tuner! Your clear explanations and step-by-step guidance made the entire process easy to understand and implement. I've struggled with optimizing my models before, but your video has provided me with the tools and confidence to tackle hyperparameter tuning effectively. Keep up the fantastic work! 🙌
There is one alternative of you, the real gems.
Amazing video, nice learning keras tuner
You're doing gods work Nitish Sir.
you are the best in teaching sir, you are gift to us sir long live
Thanks for the wonderful explanation.
you did a great job
Thank you Nitish !! 😃
simple and better explanation , keep up the good work please 👍
Thank you! Best teacher ever..
nice explanation, Thanks
It is really helping me lot.
Bhayya you teach very well ❤❤❤
I've learned many things from your videos
Hello dear, it is requested that kindly make video regarding Deep learning or pretrained models based ensemble techniques such as stacking , bagging, and boosting. Thanks dear
Hi , By using this I got an accuracy of 79 percent and val_Accuracy of 80.5 percent , And really i learned something different.
yeah i am also confused here ,at 26:00
Thank You Sir.
@CampusX Sir, where can I get the code written in this video? Have you posted it anywhere?
Hello Nitish! Great video. Only 1 doubt.
If we add BatchNormalization() to the layers, it is giving a consolidated set of hyperparameters (not for every layer but one set for all the layers) which is somehow performing better without BatchNormalization. Wanted to know how can we introduce BatchNormalization and still get hyperparameters for every layer.
Thank you!
Hi sir,
I recently come accross your 100days ML playlist. It looks lots of informative, I just wanted to ask is this your complete playlist? As I am seeing 66days and some maybe 5-6 other videos.
Nice video pls provide DL one notes
Love it sir
thank you
Thanks sir
Hello.
What if I want to optimize the learning rate, the number of epochs or callbacks such as ReduceLROnPlateau? Can I use keras tuner for that? Or is it already included in the search procedure?
you are the best
Majedar
Sir plz make a video series on UMAP
feature Extraction technique
Sir mujhe iloc samjh nahi aa raha aapne konse video mein padaya hai ek baar pata dijiye🙏
Awesome
Please create a complete video on Time Series Forcasting using ARIMA Model🙏
Hey when will you start RNN and CNN. Also which one will you start first, really hoping to study these topics from your channel!!!
If the optimum number of layers is getting set as 3 so why there are values till units8..it should be till units3 right?
When i first did it gave me rms prop now its giving adam😮
Nice
I guess we need to run it for more trails
Sir, Good video & well explained, Thanks a lot.
kindly share the code of this, i am not getting expected output. Thank you
brother if i used the same model for regression mean output layer is linear I am getting ccuracy as 0.0 what to do?
finished watching
we can use gridsearchcv also
Thanks for a very insightful video. Very helpful. One question I would like to understand - the initial model with 2 layers and small number of nodes performed close to or better than the recommendation by the Keras Hypeparameter Tuner. The network recommended by Keras Hyperparameter is much bigger and more complex. Why is that? After all that effort, our initial model performed better. Any thoughts?
This was for small dataset but when working with large dataset keras turner help
God NITISH
ISI dataset par kaam Kiya may nay aik project par
Hey, how can I contact you. I am struggling with some guidence issue in starting my career
Waha bhaiya harry potter ka book rakhee ho
Sir , does campusX provide any paid course ?
Sir, please update ML, DL and NLP playlists
Sir pls make a video on k fold cross validation
Sir maine abhut saara video dekha par usme acche se explain nahi Kiya gaya hai
Aap ek video banao aur statistics par bhi video banao
Apne playlist me dusre ka video daala hai
Sir it's very important for me I am watching only your video , getting understand with your video only
Humble request from honest student to best teacher
Bro next video kb upload hogi ?
sir please share the link of this notebook
Sir next video please jaldi upload karo
can you plz share your colab notebook
most of viewer not getting expected output (i also try more then 5 time ). sir please share this notebook code
best
tune search is not working . can anyone help.get hyperparameter shows list indx out of range
can u provide colab file link so that we can practice
Bro ANN kab tak complete hoga ? This month will it complete ?
Yes
59:52 {'units_0': 56, 'activation_0': 'relu', 'optimizer': 'adam'}
only this comes ! any help is much appreciated
At the very end we get to 75% accuracy what else we can try to improve it further
Bro, where can i find the code?
sir may you share this notebook
🙏
In the no of layers fine tuning i got 99% accuracy while in the last fine tuning i got only 65 %
Can someone explain this
Where is the code link 🙄
Can you please add link of source code in description of the video
Hit like if you 👍👍👍👍👍
where is code for this ?
sie plz share the code of this video. its request to you.
notebook link
brief info :
1. BEFORE TUNING our model(like with random neurons, activation function, optimizer), results were as follow:
loss: 0.4201 - accuracy: 0.7964 - val_loss: 0.4568 - val_accuracy: 0.8052
2. AFTER TUNING,
loss: 0.4576 - accuracy: 0.7883 - val_loss: 0.4793 - val_accuracy: 0.7922
😂😂
May be we can apply some early stopping technique as well, as we are running for so many epochs, causing overfitting
sir please provide the source code. I got some error(output not correct) althrough the code is same
Without 'hp' is giving best result 82%
Sir can you provide the code?
Bro share file link
source code
before learnint AI and DEEP learning...Learn Data Analysis
Admist mediocrity filled in the name of Data Science on Indian Internet, you stand out as an exception. Khush raho, bs ye playlist poori krdo!!!
Put hindi on the video title jeez
best