In all the videos by Bappy, he is speaking superfast, his intent is good but all the great information is lost because at the end his way of presentation is difficult to understand.
The lecture is informative. But few suggestions 1. please do mention that no one uses such ( 01_ data_ingestion...) file names in live projects. 2. The structure/template is too complex... it can be simplified. 3. Use callable & typing to simplify flow and avoid creating objects for everything.
Hello Krish sir can you implement end to end project based on lang chain for any used case like question answering or text classification or ner or any which to use growing fast in field of LLM
Thanks for the video Krish! Do you have tested different software like Aim, ClearML besides MLFlow? Tbf, I just want to log experiments and save the models :D
You are machine to make videos on data science sir ji.. Lots of love and respect on you. Thank you is very small word to describe your dedication to teach us.
Greetings sir, I have a request, please mention the prerequisite of the end to end in description or comment box, I am a beginner and enrolled in Ds masters 2.0 course, get tempted by this end to end projects of yours but hesitate, do i have knowledge to understand this will i be able to do it? With al this question i avoid this projects
if you're a beginner i would suggest first focus on your data science skills (machine learning , numpy, pandas, matplotlib, sklearn etc..) and python, then when you feel confortable with your jupyter notebook and can do the basics steps of a machine learning project (EDA, cleaning the data, scaling/encoding the features, train a model and then make predictions) then you should be able to understand the full video (NB: watch krish video on docker before doing this project)
Hey sir> thanks for this video.we love your videos. I will keep it small. i had i issue while depolying error was "STOP AND REMOVE CONTAINER IF RUNNING" if anybody can know the issue plz comment down.Thanking again.
Hi Krish, there is a bug in the data validation schema it's working and checking last column only in the schema.yaml if we comment all the columns except last one then also validation status is true.
Excellent! My comment is: Your updated requirements.txt file is well-suited for an MLOps project. It includes essential libraries for data manipulation, machine learning, visualization, web service creation, and more. BUT, by pinning versions and using a virtual environment, you ensure a reproducible and isolated environment for your project.
I think you should add data transformation steps as well in this and I have one doubt in general we do save preprocessing.pkl file and we call that in prediction pipeline and in while flask app as well please explain that thing as well
Hi , I was going through the data validation code, I think you should add a ''break'' in the if statement otherwise it always gives the validation status of the last column. for col in all_cols: if col not in all_schema: validation_status = False with open(self.config.STATUS_FILE, 'w') as f: f.write(f"Validation status: {validation_status}") break else: validation_status = True with open(self.config.STATUS_FILE, 'w') as f: f.write(f"Validation status: {validation_status}") return validation_status
To create the Python environment write the following commands on your terminal- 1) (To create a new virtual environment) 2) (To activate your environment) 3) once you are in the environment run "" to install the packages
@krihnaik06 Hello, I am following the video series since long I did the projects also. but now I am facing following issue in data ingestion:- artifacts/data_ingestion folder created successfully but facing above issue.