Morph is a python-centric full-stack framework for building and deploying AI apps.
- Fast to start π - Allows you to get up and running with just three commands.
- Deploy and operate π - Easily deploy your AI apps and manage them in production. Managed cloud is available for user authentication and secure data connection.
- No HTML/CSS knowledge requiredπ° - With Markdown-based syntax and pre-made components, you can create flexible, visually appealing designs without writing a single line of HTML or CSS.
- Customizable π οΈ - Chain Python and SQL for advanced AI workflows. Custom CSS and custom React components are available for building tailored UI.
- Install morph
pip install morph-data
- Create a new project
morph new
- Start dev server
morph serve
- Visit
http://localhsot:8080
on browser.
Understanding the concept of developing an AI app in Morph will let you do a flying start.
- Develop the AI workflow in Python and give it an alias.
- Create an .mdx file. Each .mdx file becomes a page of your app.
- Place the component in the MDX file and specify the alias to connect to.
.
ββ pages
β ββ index.mdx
ββ python
ββ chat.py
- Create each files in
python
andpages
directories.
Python: Using Langchain to create a AI workflow.
import morph
from morph import MorphGlobalContext
from morph_lib.stream import stream_chat
from langchain_openai import ChatOpenAI
from langchain_core.messages import HumanMessage
@morph.func
def langchain_chat(context: MorphGlobalContext):
llm = ChatOpenAI(model="gpt-4o")
messages = [HumanMessage(context.vars["prompt"])]
for token in llm.stream(messages):
yield stream_chat(token.content)
MDX: Define the page and connect the data.
# π¦π Langchain Chat
<Chat postData="langchain_chat" height={300} />
- Run
morph serve
to open the app!
Visit https://docs.morph-data.io for more documentation.
Thanks for your interest in helping improve Morph β€οΈ
- Before contributing, please read the CONTRIBUTING.md.
- If you find any issues, please let us know and open an issue.
Morph is Apache 2.0 licensed.