‪@LangChain‬

Рет қаралды 40,329

10 ай бұрын

Langchain is a great tool for building applications powered by LLMs. It provides a series of agents that lets users interact with the LLMs on their specific use case with natural language.
In this video, we learn how to set up and use the Pandas Dataframe Agent with GPT-4 from OpenAI.
Find the notebooks and the data here: github.com/AssemblyAI/youtube...
▬▬▬▬▬▬▬▬▬▬▬▬ CONNECT ▬▬▬▬▬▬▬▬▬▬▬▬
🖥️ Website: www.assemblyai.com/?...
🐦 Twitter: / assemblyai
🦾 Discord: / discord
▶️ Subscribe: kzfaq.info?...
🔥 We're hiring! Check our open roles: www.assemblyai.com/careers
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
#MachineLearning #DeepLearning

Пікірлер: 30

@GeorgeZoto 9 ай бұрын

Excellent tutorial and to the point, thank you for this hands on session, looking forward to checking it out :)

@vidivy 2 ай бұрын

This was awesome - clear, concise and relatable. Thanks

@JoeSmith-kn5wo 9 ай бұрын

Great video! I'm doing some data analysis with pandas right now and this could come in handy.

@gvasvas 9 ай бұрын

Excellent tutorial, easy to understand.

@AssemblyAI 9 ай бұрын

Glad it was helpful!

@kunalsoni7681 9 ай бұрын

enjoyable session :)

@WilliamDye-willdye 9 ай бұрын

Very nice walkthrough. Now my problem is to find a data set for my own demo project that has an plausible chance of producing even a small profit. Perhaps an analysis of popular children's books, fed into a coloring book generator. The book might not actually sell, but it seems reasonable to start thinking early about how to use AI to make stuff that people will want to buy.

@VISHVANI9 6 ай бұрын

Good one, thank you!

@AssemblyAI 6 ай бұрын

You're welcome!

@alinakapshyk452 9 ай бұрын

Great video. My question is can we use it in Power Bi with questions embedded in the form of tubs, or as a list of questions with data sources on Azure?

@joaquinmenendez7234 8 ай бұрын

Do you have some recommendations about tuning up the prompt to get better outputs? Like passing dataframe schema, some explanation, etc

@rasalive4062 9 ай бұрын

Thanks a lot it was great leanring, I just wanted to mention that chatgpt already knows this dataset so it's able to understand where to find the specific things. If you try something on custom data set it does not provide such in depth answers

@RonaldDraxer-rb4qm 6 ай бұрын

Thats not how it works. It works by generating code and executing it behind the scene

@justinchung982 6 ай бұрын

any chance you could demonstrate using a model running locally (e.g. LLama2) please?

@DK-dp3kk 6 ай бұрын

Great video thank you. I use the Falcon model to save money by the way. Can you summarize text that is within the pandas data frame? How?

@nicky_rads 9 ай бұрын

Nice video ! AI agents have a lot of potential, very cool to see.

@ozgurdugmeci6071 7 ай бұрын

is it possible to add conversational memory to the pandas agent?

@walidmaly3 9 ай бұрын

Thnaks a lot.. I am getting this error OutputParserException: "Could not parse LLM output: `I need to use the `python_repl_ast` tool to execute the command.`" Can you please help? If I do this it works query="What is the average sepal_length?" query=query+ " using tool python_repl_ast" agent.run( query)

@goldenfishes3695 4 ай бұрын

for those wondering, the agent supports openai only

@gunnamvenkatganesh2386 3 ай бұрын

how to get the action input formula as the output

@MuhammadFaizanMumtaz3 9 ай бұрын

I don't know you're name but I love you a lot.

@user-lz8wv7rp1o 9 ай бұрын

Can it work for pyspark? Great

@devpriyashivani1855 7 ай бұрын

When the list of rows in response is very large, ... is returned in the response instead of the actual values. How can we get the actual values?

@texasfossilguy 6 ай бұрын

specifically what are you wanting? It is not good practice to read every row in a giant dataframe because it will use RAM. If you want specific rows ask for those.

@texasfossilguy 6 ай бұрын

import pandas as pd fruits_df = pd.read_csv('fruits.csv') with pd.option_context('display.max_rows', None,): print(fruits_df) option context is temporary change, so the next time code is executed in another cell, you will see ... again. For permanent option change use: import pandas as pd fruits_df = pd.read_csv('fruits.csv') pd.set_option('display.max_rows', None) print(fruits_df)