Видео

I have a milvus-standalone v2.5.4 running through its docker-compose(from its documentation) in my EC2 instance. Which ports do i need to map or expose to avail this UI. I find it difficult running a python script everytime to check insertions into milvus.

Bro I NEED a video on importing json files to Zilliz, please share a link or something. And how I can prepare my json file with content and metadata

Thank You. I am a old time ROR programmer, jumping into AI. Although I know Python, I am a Ruby guy at heart. I will start digging into your stuff. This looks very promising.

Can you make one tutorial to add to droplet and integrate with n8n using openai

Hey, is there any other way for getting the relevant text , table and image information associated with that text ? Where the user will give query?

great tech from whyhow, but unfortunately their business execution is subpar

Oi

Glad to hear this presentation, the complex chains part is getting exciting!

Great Video, thanks!

Prompt : do llm not have limitations on tokens. Dont see how the code prevents an overflow of input.

Why splitting receipts? Does not make sense to me. A receipt is a single unit. If you split you lose information. You may loose overview of ingredients or the instructions.

That benchmarking tool is cool!

This is an interesting model of RAG, where instead of a vector database being used to augment a query to an LLM, you pass the query through the LLM first top retrieve something from the vector database. Doesn't seem to fit with the definition of RAG, which is Retrieval Augmented Generation, but here you defined LLM Augmented Retrieval.

why do I need langchain to use milvus?

where are the tech talks?

*Topics, Speakers & Timestamps:* RGB-X Model Development | _Daniel Gural_ --> 16:15 How Inkeep & Zilliz built an AI Assistant | _Robert Tran_ --> 47:57 Data Prep for LLM | _Santosh Borse_ --> 1:29:24

Thanks! Can you share your nootebook again ?

Thank you for organizing this amazing event!

Very good presentation, awesome we're pumping ruby as key player. I'll be using this on my next project, thanks

So annoying. Your intro outro welcoming is half of the video. Why you split that much?

jupyter mention that No module named 'milvus_lite', but I have from langchain_milvus import Milvus

Glad to see GraphRAG taking off! It's definitely the future of RAG 🙌

Would it be possible to share the notebook?

Great content, thanks!

I'm working on a Retrieval-Augmented Generation system using SEC 10-Q filings for multiple company like the one that you use in this demo and I’d like to know What are the most common and effective chunking strategies for those complex files?

Thank you, I will try.

really appreciate for your work! It's very helpful for me who just enter the field of graphrag!

I tried to reach you on your website, but you wouldn't accept my email. GPT4ALL is pretty good.

Jarrell Street

can you please share link to slides, the link above doesn't work

Very helpful :) 👍

it was very helpful

First view from india ❤

Milvus has similiarity search, hybrid search, and rich schema support. having different index options is not a bad thing. and milvus has c++ java python and go sdks and they all work fine even though they are community supported. I do not get your comparison guys

Failed to create new connection using: 8bd43c0453be43039caca825dc0105fd Traceback (most recent call last): client = MilvusClient("milvus_demo.db") File "C:\Users\z004zcyp\AppData\Local\Programs\Python\Python310\lib\site-packages\pymilvus\milvus_client\milvus_client.py", line 59, in __init__ self._using = self._create_connection( File "C:\Users\z004zcyp\AppData\Local\Programs\Python\Python310\lib\site-packages\pymilvus\milvus_client\milvus_client.py", line 656, in _create_connection raise ex from ex File "C:\Users\z004zcyp\AppData\Local\Programs\Python\Python310\lib\site-packages\pymilvus\milvus_client\milvus_client.py", line 653, in _create_connection connections.connect(using, user, password, db_name, token, uri=uri, **kwargs) File "C:\Users\z004zcyp\AppData\Local\Programs\Python\Python310\lib\site-packages\pymilvus\orm\connections.py", line 379, in connect from milvus_lite.server_manager import server_manager_instance ModuleNotFoundError: No module named 'milvus_lite' Can anyone help for this?

getting no module named milvus-lite, pip install milvus-lite giving error ERROR: Could not find a version that satisfies the requirement milvus-lite (from versions: none) ERROR: No matching distribution found for milvus-lite

Getting error, pk field should not be of type VARCHAR for auto Id : True

So, another copy pasta straight from langchains git.

29:45

Great video. #foAIs

this video is so good ! even after hackathon im coming back to refer this! thank you

thanks for this interesting demo

what about persistence? how it works?

Thank You for your informative seminar. I will also try to run your 'Demo Notebook'.

Great seminar, will run your notebook to learn more

thank you zilliz, this talk is very educational, will try out the notebook

Dude watch ur vide before uploading. Ur face is coming on top of the code all the time

Worst zoom ever... wow alex is the worst choice of any company..

Your chats don't appear on replay, so it would be helpful to share using a url shortener during the live and include links in your descriptions. I don't know why your chats disappear, but I learned the hard way that modifying the video in any way (even doing a trim) will result in RUclips removing the chat.

This is great! Thank you for the tutorial. Liked and followed! Could you please share the code? I would really like to try it out myself