Deep learning for (almost) any text classification problem (binary, multi-class, multi-label)

In this video, I will show you how to build a model for (almost) any text classification problem, i.e., binary classification, multi-class classification or multi-label classification. You can expand this to any language and any type of deep learning model (LSTM, GRU, Transformers, etc). In this video, I will be using BERT model from Transformers and Tez.

#NLP #DeepLearning #PyTorch

Please subscribe and like the video to help me keep motivated to make awesome videos like this one. :)

Follow me on:

[AIRobotica]

Very important video on this topic... Thanks a lot Sirji...🙏

[divyanshukumar]

I was waiting for this. Thank you so much, Sir. If possible, then please make a video on approaching Reinforcement Learning problem.

[sayedathar]

Amazing Video Thank You so much for sharing, I have been thinking of switching to Tez for Kaggle Competitions, Hopefully this will take me a long way for recent NLP Competition

[anuraagbeniwal]

Loved the video, can you please explain in a little bit more detail as to how do we approach multi-label classification in this.

[saistudent_sachinprabhu]

Thank you so much for everything :) Also congratulations on completing one year on youtube :D I've been following your work for more than a year and it has been helpful. Much Love :)

[HipHop-czos]

This video is insane🤩. A video on how to create our own python package.

[MrDanituga]

Hey Abhishek! Thank you so much for your video. That was a really fast hands-on. Just a question, how are you able to run the code in your Titan which is in another place, remotely. Basically, how do you run vs code via server and then access it on another machine?

[pedrocouto]

Amazing video! Can't wait to try it by my self. Many thanks, Sir!

[dv]

Amazing Guide for all types of text classification 🔥. Wating for 25days NLP😣. I hope it come soon.🙏

[vikasdubey]

Awesome. Loved the video. Really learnt a lot. Thanks a lot.

[mehparasaghir]

Hi, Brother Iam doing MSCS and doing my thesis in Mutlilabel text classification by busing Binary relevance, MLKNN, Label powerset and classifier chains. Kindly can you help me to guide a little for implementing thoes classifiers using python.

[gokulgupta]

Great informative video. One question why you had put 768 number in nn.Linear(768, num_classes). Means want to know any specific reason?

[williama]

Hi Abhishek, thanks for this youtube video. I found it good for me because I am just learning how to classify text using nlp. But then, while running your code, I have this error coming from line 87 of your code. It says TRAIN_BS not defined. I could not find anywhere in your code where TRAIN_BS or EPOCHS was defined. OR, does TRAIN_BS and train_bs mean the same thing ? Likewise, does EPOCHS and epochs mean the same thing? Thanks

[farahalaa]

hello, thanks for your video, I have a question and I hope you answer it, if we did a text classification where can we add word to classify it . this step to test our model accuracy . what I mean is after text classification step i need to add some words and the model will classify them

[debarchanbasu]

Hi, I'm getting an error here

AttributeError: 'TextModel' object has no attribute 'config'

Could someone tell me why? Apparently there seems to be something wrong with tez.callbacks .

[muhammadal-qurishi]

What is better/more reliable to evaluate your classifier sklearn or seqeval?
I used both of them in an NER task and they gave me completely different results.

[sumantthakur]

Thanks for this great video. For imbalanced dataset, where will you apply over-sampling of minority classes.

[sanketsharma]

Hi @Abhishek,

I want to learn NLP with practical exposure. Can you suggest me some resources to learn NLP?

[sperovita]

hi, could I use this to classify positive and negative sentiments using the IMDb dataset available on Kaggle?

[yveslacroix]

Hi very nice video and thanks for sharing. I have a question though : I am trying to run this on cpu (device="cpu"), with a Camembert model from transformers. I specify n_jobs=-1 but get stuck with an error "BrokenPipeError: [Errno 32] Broken pipe". Can you help me with this?