Very helpful video, please make video on can we use rag with llm using hugging face + langchain api instead of importing model in our local as it would take lots of gpu memory,
Thank you for this amazing series of vedios. I have a question that we ca using Chroma DB for saving the embeddings so how can we see these embeddings in chroma db and aslo we have not use any chroma db connection link.
We have created a Chroma instance with Chroma.from_documents, which stores embeddings in a Chroma vector database. In our code, we haven’t specified a persist_directory, so the embeddings are stored in memory only. To persist the embeddings and be able to reconnect later. You can use this code: vectorstore = Chroma.from_documents( documents=docs, embedding=GoogleGenerativeAIEmbeddings(model="models/embedding-001"), persist_directory="path_to_persist_directory" ) # Save the embeddings to disk vectorstore.persist() To inspect the embeddings stored in Chroma DB, you can use the get_all_embeddings() method query_vector = GoogleGenerativeAIEmbeddings(model="models/embedding-001").embed("your query text") results = vectorstore.similarity_search(query_vector, k=5) for result in results: print(result)
I have followed your video, but the chatbot is still giving answers outside the provided context, even after using your system prompt and making adjustments. For example, if I say "I'm sad, write a joke for me," it still writes a joke. This is the issue I'm encountering. Could you please provide a solution?
make video on how we can integrate this in whole automate pipeline and also use streamlit or any other framework for chat ui to get the response because nowdays everyone giving demo on ipynb file so it will be very helpful of us if you make video of full fledged project of rag application using langchain and any other frontend framework
I get this error while I run last cell of that basic rag --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) Cell In[13], line 1 ----> 1 response = rag_chain.invoke({"input": "what is new in YOLOv9?"}) 2 print(response["answer"]) AttributeError: 'int' object has no attribute 'name'
Just load the data from csv file. Eg: from langchain_community.document_loaders.csv_loader import CSVLoader file_path = ("test.csv") loader = CSVLoader(file_path=file_path) data = loader.load() for record in data[:2]: print(record)
Hello madam, Omkar, this side. I’m very glad to see your video regarding that RAG model. But for somehow, I realise that it is not hundred percent running locally, we need to use Google API token key. I have use the same with the open AI after few recurrent and after few token processing, it is asking some billing method or credit card details to further. Can we have such a model where we can deploy rag pipeline from scratch hundred percent locally? we can fetch an LLM model from hugging face and download it and storage in our local drive. Create a victor data on our own or just a pie tenor, which am all the text token. That will be much more beneficial for me if we are going for a a business purpose. It is much more beneficial to run it locally with a discreet GPU. Can you please help me guiding on the same building a rag model from scratch using a LLM from hugging face? It can be any LLM of my choice. I’ll be hopeful to see that tutorial and develop myself.. thank you so much for your content. Your content are very beautiful, and it’s very informative.. just teach like a teacher in a classroom, thank you so much again…..❤❤
Thank you soo much for your help!! I was struggling with making a project work and your video helped soo much!!
Glad this was helpful!
u have cleared all the concepts in very simple and easy way
Glad to hear that
Best tutor for AI and ML , Thanks alot mame
Most welcome!
What an awesome way of explanation
Glad you liked it
Wao Amazing thanks mam
from Pakistan
Thank you!
Great Explanation. Please Bring More Concepts related to GenAI
Yes, Sure.
Awesome , Madam
Very nicely explained ma'am
Glad you liked it
You deserve subscribe 😍
Thank you so much 😀
Very helpful video, please make video on can we use rag with llm using hugging face + langchain api instead of importing model in our local as it would take lots of gpu memory,
Check this video: th-cam.com/video/DQfBXRTeo3o/w-d-xo.html
looking forward to hearing seminar about Lora-pro from U
Noted!
Plz explain fine tuining the hugging face model on custom data specially text to image generation
Sure, Soon!
Super awesome video Asrohi. Can you make one RAG app to chat with any multiple websites please.
You can provide the link of multiple websites in the urls list.
@@CodeWithAarohi Sorry. My question was, can we add chat history into this.
thank's
Welcome!
How to interact with multiple pdfs, and how much load of data will be handled by llm as a free tier
thanks for detailed explanation
Welcome!
full course video about "Claude 3.5 sonnet AI model, API finetune" full course please
Noted!
Thank you for this amazing series of vedios. I have a question that we ca using Chroma DB for saving the embeddings so how can we see these embeddings in chroma db and aslo we have not use any chroma db connection link.
We have created a Chroma instance with Chroma.from_documents, which stores embeddings in a Chroma vector database.
In our code, we haven’t specified a persist_directory, so the embeddings are stored in memory only.
To persist the embeddings and be able to reconnect later. You can use this code:
vectorstore = Chroma.from_documents(
documents=docs,
embedding=GoogleGenerativeAIEmbeddings(model="models/embedding-001"),
persist_directory="path_to_persist_directory"
)
# Save the embeddings to disk
vectorstore.persist()
To inspect the embeddings stored in Chroma DB, you can use the get_all_embeddings() method
query_vector = GoogleGenerativeAIEmbeddings(model="models/embedding-001").embed("your query text")
results = vectorstore.similarity_search(query_vector, k=5)
for result in results:
print(result)
I have followed your video, but the chatbot is still giving answers outside the provided context, even after using your system prompt and making adjustments. For example, if I say "I'm sad, write a joke for me," it still writes a joke. This is the issue I'm encountering. Could you please provide a solution?
make video on how we can integrate this in whole automate pipeline and also use streamlit or any other framework for chat ui to get the response because nowdays everyone giving demo on ipynb file so it will be very helpful of us if you make video of full fledged project of rag application using langchain and any other frontend framework
Check the last 10 mins of this video. I have explained how to use RAG with LLM on streamlit
I get this error while I run last cell of that basic rag
---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
Cell In[13], line 1
----> 1 response = rag_chain.invoke({"input": "what is new in YOLOv9?"})
2 print(response["answer"])
AttributeError: 'int' object has no attribute 'name'
I am not able to install chroma facing issue
Please mention the issue you are facing.
how to create the same on a CSV dataframe?
Just load the data from csv file. Eg:
from langchain_community.document_loaders.csv_loader import CSVLoader
file_path = ("test.csv")
loader = CSVLoader(file_path=file_path)
data = loader.load()
for record in data[:2]:
print(record)
is it paid maam ??
NO
Hello madam, Omkar, this side. I’m very glad to see your video regarding that RAG model. But for somehow, I realise that it is not hundred percent running locally, we need to use Google API token key. I have use the same with the open AI after few recurrent and after few token processing, it is asking some billing method or credit card details to further.
Can we have such a model where we can deploy rag pipeline from scratch hundred percent locally? we can fetch an LLM model from hugging face and download it and storage in our local drive. Create a victor data on our own or just a pie tenor, which am all the text token. That will be much more beneficial for me if we are going for a a business purpose. It is much more beneficial to run it locally with a discreet GPU.
Can you please help me guiding on the same building a rag model from scratch using a LLM from hugging face? It can be any LLM of my choice. I’ll be hopeful to see that tutorial and develop myself.. thank you so much for your content. Your content are very beautiful, and it’s very informative.. just teach like a teacher in a classroom, thank you so much again…..❤❤
mam can you please complete the generative ai playlist
Yes, Sorry for the delay. Just busy with few ongoing projects.