Introducing the fastest way to create and deploy a RAG application built using @embedchain on @streamlit We saw that some @embedchain users deploy a RAG application to streamlit once they have created it.

So we have added native support for it in the framework itself.

We spent $100 to learn how to build & run a RAG application in production.

Read further so that you don’t end up doing the same mistakes.

👇🏻👇🏻👇🏻 We launched @SadhguruJV AI few days back & users asked ~3.6K queries.

This gave us a ton of learnings around RAG systems in production & we are incorporating this to @embedchain

FYI: We used LLM & embedding model from @OpenAI

https://twitter.com/taranjeetio/status/1737328611030651160

📇 @embedchain now supports a wide range of LLMs

• OpenAI
• Anthropic’s Claude
• Cohere
• GPT4All
• JinaChat
• Llama2
• Vertex AI

via @LangChainAI The best part is all these LLM models are supported via @LangChainAI

So you can set the variables for Langsmith and can quickly debug your RAG application using Langsmith.

We @embedchain now support LangSmith,

a platform to debug & monitor

LLM applications by @LangChainAI Some context:

@embedchain allows you to create
ChatGPT like apps on top of
custom dataset.

Here is an example of a personal assistant
which has my notes in pdf format

Create ChatGPT like chatbot over any open source project.

Introducing a new data type called "code_docs_page" in @embedchain

Powered by @LangChainAI

Watch the video below to see how we create a chat bot over Langchain & Guardrails docs In this video, we show how to create and use a @LangChainAI docs bot.

📢 Major announcement

Embedchain now supports asking questions to pdf files without an internet connection, for free and no data leaves your machine.

Powered by
@nomic_ai 's gpt4all
@huggingface 's sentence transformers
& @LangChainAI

Watch the video 👇🏻 to learn how In this video, we have taken Naval's Almanack book and stored it locally on disk.

We then
• create a chatbot
• add dataset using .add
• query using .query

Introducing fastest way to create & get a chatbot running:

Embedchain Replit template,

Has 2 examples:

• Youtube Channel to Bot
• Chat with Multiple PDF files

Built using Embedchain, hosted on @replit
powered by @langchain and @trychroma First let's understand how to use the template.

• Open the template
• Click on the "Use Template" button on the top right
• Your own repl created with a readymade bot
• Now just click on the "Run" button

https://t.co/YAsWMu9ASnreplit.com/@taranjeetio/E…

👶What if Embeddings and LangChain had a baby?

🎉Introducing EmbedChain: a framework to easily create LLM powered bots over any dataset.

✨Powered by our favorite @LangChainAI and @trychroma Some context on why this is needed👇🏻

Everyone is building ChatGPT like bot over their datasets. Some examples:

• BibleGPT, QuranGPT, KrishnaGPT
• Chatbot over youtube channel
• Chatbot over Notion, PDF, Quarterly reports

Introducing Youtube to ChatGPT like Bot

• Just enter the channel or video URL
• wait for some time and
• a ChatGPT-like bot trained on your entire content is ready.

The link is in the next tweet Try it out:

embedstore.ai/add/youtube-ch…

🧠 AI newsletters have been really helpful in my journey as an AI learner & builder.

💡 Have picked up a lot of things & I stay constantly updated thanks to these.

📅 Curated a schedule of all the popular AI newsletters.

Check it out: ainewsletterschedule.com

(all time in UTC)

Ever wondered how @LangChainAI calculates OpenAI pricing?

This thread is a dive into LangChain's source to learn how it maintains and calculates OpenAI pricing. LangChain has two functions:

• One function maintains the model to cost mapping

AI is moving very fast. Bit difficult to catch up
& learn what’s happening in AI.

To solve this for myself & others,
I am starting a newsletter

“Keeping up with AI”. First of all, it’s not a curation newsletter.

There are already many popular & great newsletters like BenBites from @bentossell,
The Prompt from @anitakirkovska
and many more.

Special thanks for the effort that they put in 🙌

Releasing @LangChainAI Python Docs Embedding

• Now create chat apps over Langchain docs using this free and updated embedding. To build a chat app using this embedding, simply load the embeddings

I was a Senior Product Manager at a Series C startup backed by YC, Sequoia.

Here is how I would have automated my job as a product manager using ChatGPT.

cookup.ai/category/produ… Create Product Requirement Doc

Eg - PRD for group feature in WhatsApp

A thread of all the use cases where ChatGPT lacks and its output is not correct: Promotion Criteria

ChatGPT has some bias around race and gender and hence lacks in giving correct promotion criteria

https://twitter.com/spiantado/status/1599462375887114240

All the amazing use cases that can be solved using ChatGPT

👇🏼👇🏼👇🏼 Programming
• Clone repo
• Convert JSONL into JSON array
• Convert text to AWS IAM policy
• Create a summarizer app
• Create react component
• Explain programming problems in a particular style
• Explain the regular expression
• Find & explain bugs in code

🚨🚨🚨 Big Announcement

Launching CookupAI - No-code AI App builder.

cookup.ai

Write Prompt. Hit Publish.

Opened 25 spots. RT this & we might give you one. 2/ Context 👇🏼👇🏼👇🏼

When ChatGPT was released, it gave new superpowers to the world. We are daily discovering something new about the model but are limited to only sharing a screenshot of it.

1/ @danielgross has written about a new search engine called Boogle (Better Google Search).

I tried creating Boogle using ChatGPT.

The results are astonishing. 🤯🤯🤯

👇🏼👇🏼👇🏼 2/ First, context about Boogle - a query expander, understands query & its topic and adds relevant site operators for it.

Eg:

“reverse list in python” site:stackoverflow.com

“pixel 7 reviews” site: reddit.com

Read Daniel’s post to find more about Boogle.

1/ OpenAI released ChatGPT few hours ago.

I played around with it. Sharing some observations about its output. 🔍 2/ ChatGPT is good at conversing in a human and natural way. Asked for gift suggestions for my mom. It first added a filler paragraph, suggestions, and then a conclusion. This is the same as a human would answer.

1/ OpenAI recently released a new GPT3 model called text-davinci-003. It has several enhancements over the previous versions.

I compared text-davinci-002 and text-davinci-003 with the same prompt and input settings.

Here are the observations 🔍 2/ davinci003 doesn't make assumptions even at some temperature value, whereas davinci002 does.

Here davinci002 assumed that a CSS library would be available and hence its output contained class names. davinci003 did not assume & its output is inline CSS styles.