Text Generation AI - Next Word Prediction in Python

[@swapbriarASMR]

26:08 Best part. Thanks for the video, very informative!

[@ruthvikas]

Chat gpt has some competition here 😂

[@artistpw]

This seems very helpful for what I'm going to try to write. Thanks!

[@alessandroswissborg2536]

I haven't understood a jack, but I stuck till the end of the video. I guess it's a good start! 😅😂

[@Ayush_.]

please provide source code and dataset.

[@yashgajjar4838]

Thank you! It was a nice explanation.

[@girishkrishnan8228]

Hi, by any chance do you know how to measure the accuracy of prediction (of future words) by your model. If so, how accurate are the autocomplete predictions?

[@m.s.mohamedwaseem1935]

it was a really nice video i really liked it. but not very understandable but i got the grasp of what i have to do to make a text generation AI. thanks

[@ElCalvoYT]

Which tensorflow version are you using?? Because the RMSprop is not working

[@doggolegend]

for me, i used "pip install keras" then i imported them like this: from keras._tf_keras.keras.models import Sequential, load_model from keras._tf_keras.keras.layers import LSTM, Dense, Activation from keras._tf_keras.keras.optimizers import RMSprop

[@gangs0846]

Great Video and explanation! Did you demonstrate how to create a model by yourself same as you used here?

[@shinrafahell]

Thanks Rob. I love pandas!

[@Pythonist_01]

Hello so please maybe we want to make this work in real time with an app or a Website how would we do that Please can you do a video for that 🙏🥺

[@opheliaschwuchow6977]

Hi! Thanks for the video, how would you do batch learning if your data was too large to fit into memory?

[@batnoy779]

I am using google colab, they provide 12GB of RAM. with big dataset it could take a lot of time but still work

[@kareemmamdouh6402]

the link of the dataset please

[@chiranjeevinaidu3660]

Quick question can I train this to run a cli application just based on keywords and appropriate commands and arguments for that keyword, this can make automation a breeze

[@batnoy779]

I trained my LSTM model on a dataset with about 2000 articles, and the model's accuracy was 30%. Can it get higher accuracy with larger dataset like mine? Thank you for your video, it's so helpful

[@spxix2961]

hie bro can u tell me did u perform embedding of all the words in that article or u just did tokenization and trained the model on it dividing it into training and testing data ??

[@batnoy779]

@@spxix2961 I used some pre-trained tokenizer (the language is Vietnamese) like phoBERT,... Then I split the dataset into train and test sets.

[@ASOSIYXABAR]

how can i see your "pre_tarained....." ai model

[@guilhermefigueiredo-234]

Does any one know a larger dataset?

[@GeraldBosley-cf8ng]

You can create your own database of non fiction books, articles on political websites, Reddit posts, and ebooks and convert to a csv form using python. Machine Learning is more than writing the algorithm, that's the easiest part of Machine Learning lol.

[@BrunoJPP78]

Amazing!

[@Saidov2024]

Good job

[@mk007__]

i cant import RMSprop but i can import rmsprop_v2 but it is not callable, whats the solution?

[@SudhanvaMS64]

It sounds like you're encountering an issue with importing the RMSprop optimizer in TensorFlow. The rmsprop_v2 module might be available but not callable directly. One potential solution is to import the RMSprop optimizer from tensorflow.keras.optimizers module. Here's how you can do it: from tensorflow.keras.optimizers import RMSprop Then, you can use RMSprop as your optimizer in your code. For example: optimizer = RMSprop(learning_rate=0.001) This should resolve the issue

[@srinivasasatyavathi6056]

how pre_tained_models folder came

[@peaceful3767]

Code?

[@sridhardornadula3913]

Hello...bro... please make a python code on giving input text and search in the given data and excute perticular text output

[@sridhardornadula3913]

*perticular text as output

[@emreay9577]

I wrote the same message of yours to chatgpt, here is the result: # Define the data and associated code to execute data = { "hello": "print('Hello, world!')", "goodbye": "print('Goodbye, cruel world!')" } # Prompt the user to enter a search term search_term = input("Enter a search term: ") # Search for the term in the data if search_term in data: # If a match is found, execute the corresponding code exec(data[search_term]) else: # If no match is found, print an error message print("No match found for", search_term) I don't really understand what you need but I hope chatgpt does. :)

[@Alexander-ms2ct]

It’s called Natural Language Processing, also the Transformer Architecture. There are books you can read.

[@rictrix1503]

11:55

[@philtoa334]

Thanks_Nice.

[@Mr.Aswin_Das]

you sounds like Chris Pratt😁😁

[@RedCloudServices]

Would this be easier with Langchain?

[@jamesknight0001]

?

[@nauvalzulfikar4884]

hehe... n_words.. hehe.. this is the only time it's acceptable to say n_words.. hehe..

[@codewithtaus]

♥️♥️♥️♥️♥️

[@mugudev]

1st View

[@thechoosen4240]

Good job bro, JESUS IS COMING BACK VERY SOON;WATCH AND PREPARE

[@simssim262]

In the era of Transformers this guy is using LSTMs 😂