- Видео 134
- Просмотров 97 827
Rajib Deb
США
Добавлен 15 фев 2022
How to integrate crew ai with Phoenix
In this recording I show how to integrate crew ai with phoenix observability solution
code:github.com/rajib76/crewai_examples/blob/main/examples/03_simple_crewai_phoenix_integration.py
code:github.com/rajib76/crewai_examples/blob/main/examples/03_simple_crewai_phoenix_integration.py
Просмотров: 8
Видео
MultiOn | Exploration day#1
Просмотров 272 часа назад
In this recording I did an initial exploration of multion. I tried to do Q&A over langchain docs. I see a lot of hallucinations happening. I will explore more to see if there is a way to control the hallucination code:github.com/rajib76/multi_on_examples/blob/main/01_browsing_langchain_docs.py
Graph Memory in mem0 | How does search work
Просмотров 154 часа назад
In this recording, I show how does mem0 do a search in the Graph memory code:github.com/rajib76/memory_examples/blob/main/examples/03_search_graph_memory.py
MemoryGraph | mem0
Просмотров 559 часов назад
In this recording I demonstrate the Memory Graph capability of mem0 code:github.com/rajib76/memory_examples/blob/main/examples/02_graph_memory.py
LlamaIndex Workflow | Global context
Просмотров 9321 час назад
In this recording I show a more complex workflow implementation through llamaindex workflow code:github.com/rajib76/llamaindex/blob/main/examples/07_llama_index_workflow_global_context.py
Integrating mem0 with LlamaIndex workflow
Просмотров 8821 час назад
In this recording, I show how to integrate llamaindex workflow with mem0 code:github.com/rajib76/mem0/blob/main/cookbooks/mem0_example_with_llamaindex_workflow.py mem0: www.mem0.ai/
LLAMAINDEX | Workflow
Просмотров 305День назад
In this recording, I show the workflow feature of LlamaIndex code:github.com/rajib76/llamaindex/blob/main/examples/06_llama_index_workflow.py docs:docs.llamaindex.ai/en/stable/module_guides/workflow/
MEM0 Memory Exploration
Просмотров 9414 дней назад
In this recording, I explored the some of the features of mem0's hosted memory solution code: github.com/rajib76/memory_examples/tree/main/examples mem0: www.mem0.ai/
LangGraph Cloud and Studio | Part 2
Просмотров 4214 дней назад
This is the 2nd part of the langgraph studio exploration. Here I show how we can use langgraph studio after the deployment. And then how to use the shareable workflow like to integrate it externally through custom program Part1 vide is here : ruclips.net/video/qTPh_8xlu0M/видео.html Code used to test this out is here : github.com/rajib76/langgraph_examples/blob/main/09_how_to_use_langgraph_stud...
LangGraph Cloud and Studio | Part 1
Просмотров 7414 дней назад
This recording is part 1 of my exploration of LangGraph cloud and Studio. In this recording, I show how we can automate deployment of langgraph workflows using langgraph cloud. The code that was used to test out this capability is as below code:github.com/rajib76/langgraph_studio
How a reflective Agency can lead to better answer.
Просмотров 6528 дней назад
Here I show how a reflective agency can provide a better response code:github.com/rajib76/mistral_examples/blob/main/01_mistral_large_context_example.py
Model Graded Evaluation | Why it is not enough
Просмотров 56Месяц назад
Model graded evaluation cannot be the only means to evaluate the response of LLMs, you will need to combine other types of evaluations to create a composite score. In this recording I show why model graded evaluation is not enough. code:github.com/rajib76/ragas_examples/blob/main/05_answer_correctness.py
The rise of memory | How mem0 implemented it
Просмотров 830Месяц назад
In this recording, I show how mem0 has implemented the memory module github.com/mem0ai/mem0
LlamaIndex KG | Deduplication of nodes.
Просмотров 90Месяц назад
In this recording, I explain in details how LlamaIndex is doing the deduplication of the nodes after creating the knowledge graph code:github.com/rajib76/llamaindex/blob/main/examples/04_dedepulicating_the_graph.py ref: www.llamaindex.ai/blog/customizing-property-graph-index-in-llamaindex
Agent Interactions | Hierarchical pattern#2
Просмотров 57Месяц назад
In this recording I show another pattern for the hierarchical approach code: github.com/rajib76/langgraph_examples/blob/main/04_hierarchy_supervisor_pattern_02.py
Agent interactions | Hierarchical Structure
Просмотров 67Месяц назад
Agent interactions | Hierarchical Structure
Milvius | how to use cosine similarity
Просмотров 46Месяц назад
Milvius | how to use cosine similarity
How to create custom metrics in RAGAs
Просмотров 1062 месяца назад
How to create custom metrics in RAGAs
How to use GCP Vertex AI multimodal embedding
Просмотров 2312 месяца назад
How to use GCP Vertex AI multimodal embedding
Langgraph integrated with semantic memory
Просмотров 1852 месяца назад
Langgraph integrated with semantic memory
Create a reflection agent using Langgraph
Просмотров 1772 месяца назад
Create a reflection agent using Langgraph
How to create a property graph from unstructured data
Просмотров 1592 месяца назад
How to create a property graph from unstructured data
Associative Memory | How to implement
Просмотров 1412 месяца назад
Associative Memory | How to implement
UMAP | Way to identify semantic overlap
Просмотров 1013 месяца назад
UMAP | Way to identify semantic overlap
Langmem - Episode 2 | Managing context switching
Просмотров 753 месяца назад
Langmem - Episode 2 | Managing context switching
How is this different from AgentOps?
This is the best training on this topic so far
Could you do integrating it with langchain please
please take a look here github.com/rajib76/memory_examples/blob/main/examples/01_remember_user_preference.py
does relying on events to trigger sufficient to cover all use cases?
@@user-wr4yl7tx3w some use case may need choreography some orchestration. So both patterns need to be considered. This is no new concept, microservices 101
How much effort was it to set something up like this? Is this scalable for more complex data?
Hi Sir - Cannot access ur git link, please can you keep the project open.
the number of open tabs made me understand memory at a different level
Does thie studio only works in Machine?
@@kitranet I am using it through langsmith. You need to have at least plus membership of langsmith. There is a desktop version of studio which works on certain Mac machines only
Is there git code for this?
Thanks for the mem0 video
@ajibdeb4059 Thanks for the video. I didn't find any content related to connecting agents to the knowledge base. Could you help with that?
Hello ... nice tutorial video ... What is the datatype of embedding in the code in python ?
Hi Rajib, I am in a dilemma, New to Langchain agent framwork. What i like is the abstractions and length of features it provides to build a true agent. What i fear is that it is too abstract and there are icebergs underneath which i dont have any clue. What if something goes wrong? Difficult to troubleshoot and may be i do not have the flexibilty enough to quickly do a work around. So i am thinking of devloping a very basic functionality which i will have full control. Any thoughts or your experience of taking Langchain to production? Your feedback is appreciated. THanks
Thanks for the video, loved how you went through every single step!
Its easy to see that the physics teacher agent was built after the math agent since the description in the physics teacher agent still references to mathematics and the instructions in the moderator related to the physics teacher agent mention maths not physics. The typical error of a copy & paste.
Nice video. Would like to see the next step for expanded Supervisor Agent capability. I appreciate that you are doing this without abstraction as i want to understand this before looking at a langgraph version, but would like to see that too :)
There is an orchestrator, but it could be exactly an event-driven orchestrator :) Great video!
Do we need to reindex all documents, everytime we add new document. Is there any way to run it programitically
Amazing explanation!! But if we don't change the existing prompts then, there's no point in creating 'Parser Lambda functions', right?
what about the prompt templates? Does this video only cover handling the response in the lambda function? What exactly are the prompt templates? How are those passed to the LLM?
Hi @rajibdeb4059, @EccleezyAvicii, do you know how to implement partial response? like if the Tool takes time to provide response, I will show temporary message to the user? Your prompt reply will be greatly appreciated.
I don't think I could've found a better video. As an intern I am so grateful thank you so much !
Great Explanation! However if we have to build a KB for prod, would it not be better to create a separate lambda function where we can have more control over how we create KB using Langchain and upload it to Vector DB? And then create a seprate RAG Agent with Lambda Function and Open API?
Can't we create a KB using LangChain because that way we have more control over how the chunking for the documents are done?
Good explanation Rajib, May I know how is it different from when we create a KG using Neo4j?
Which Authentication type you set for your tool ?
You can decompose the compound questions into simple questions. The Langchain page has a good example for this. Thanks for this video!
Can u provide the code that goes in the tool schema?
thank you! very clear approach and code!😶
Can you share the colab / github link? Awesome work, been following you for some months now.
Hi Rajib, great work. What should we do if we want to create index and use Vertex AI Matching Engine to index the multimodal embeddings ( including for example text and image embedding for each item)? how should this be done? Does matchingengine index support that?
Have you gotten the answer?
thanks for the video. I learnt a lot
You cant do Integration 🤣🤣🤣🤣🤣🤣
Thank you sir 💐
Hi Raj, Could you please load some videos for pdf document data extraction including bounding boxes
I tried the one in Langchain, but didn't give good results. Are you seeing good results with gpt3. 5?
Please cover Langgraph sir
great work
Sir, great tutorial as always, thanks. Can you please guide me on getting the API keys.
Search name and api key on youtube
Thanks Sir. Actually I am working on some project based on it and find difficult for me to find materials to understand the concept practically. After watching this video, I understand and implement it successfully and step ahead. At last Thanks Sir for this video.
what is the benefit of using this vs a straight forward vector search on a query-response pair? i can only think of a sql type search on historical conversational data that uses the neo4j graph nodes.
Semantic search is fuzzy. When you combine semantic search with associative memory, you will get a better match. But it depends on your use case. Some use cases are very complex where the question can map to multiple similar concepts, there the associative memory is useful
Thank you for articulating your thoughts on this interesting topic!! I’ve been experimenting lately w/ Neo4J. It will be neat to use PromptFoo or another evaluation tool to show the improvement between vector search and associative memory search.
beautifully explained!!!
You are mixing agents with tools. The idea of orchestration is good, however, the example you present is confusing because it treats agents as functions/tools Thanks for the effort, though.
Sir can u please tell me what you are teaching and what your all tutorials are about to ?
Is there any way that we can access agent apps created outside as an api call
How did you get the value of document_library_id?
Thanks for the video. Short and straightforward. Help me with the following questions? 1. Why did you load the separate csv which has the vector embeddings? I think we can create the vector embedding on the plot property itself. 2. Can you build a chatbot based on this? The example which you have used can be easily retrieved by using the simple cypher query. So can you do some more of text based search and let system perform a Symantec search on it? 3. How are these functions are different from LLM functions?
I couldn't see the data_dir when I ran this example in my linux machine, because of that I couldn't see the saved memory, also I feel there is no context from the prompt response
Hi Rajib! I always like your videos. Thank you for producing them. I'd like to suggest moving to vanilla python, at least for some of the videos. Some orgs are not able to use Langchain/graph/Semantic Kernel/etc. in prod (due to various reasons). Thank you for considering ;)
Is it different from knowledge graph,
Knowledge modeling is the process and KG is the technology