# bring your own LLM based functions
# These can be any function or callable
# for now lets use example ones
from stringdale.chat import Chat
from stringdale.db import ChromaClientStringDALE
String Diagram Agent Language Express (StringDale) is a library for building complex LLM agentic workflows easily using string diagrams.
Motivation
We found current agentic frameworks unnecessarily complex:
- Too many levels of abstractions and boilerplate code.
- Too many interfaces to learn
- To many wrappers to build for our code to fit these frameworks
As a response, we create stringdale which aims to:
- Have a small interface - making it easier to learn.
- Non-intrusive - You can use your own functions in stringdale, without fitting them to a complex interface.
- Have few levels of abstraction - we don’t create tons of wrappers you have to navigate, you bring your own functions and understand exactly what is happening at all times.
- Make tracing and debugging simple.
Installation
# in an environment with python>=3.10
pip install stringdalestringdale uses graphviz to draw your diagrams in jupyter notebooks.
Make sure you download and install graphviz on your system if you want diagrams to be rendered.
Usage
Lets look at how to build a RAG workflow
Lets make an example vector db:
chroma_client = ChromaClient()
dog_docs = [
{
'id': 'dog1',
'text': 'The Golden Retriever is a friendly, intelligent breed known for its golden coat. They make excellent family pets and are great with children.',
'metadata': {'breed': 'Golden Retriever'}
},
{
'id': 'dog2',
'text': 'German Shepherds are loyal, protective dogs often used in police work. They are highly trainable and good at various tasks.',
'metadata': {'breed': 'German Shepherd'}
},
{
'id': 'dog3',
'text': 'The Golden Retriever is a friendly, intelligent breed with a beautiful golden coat. They are wonderful family pets that get along well with kids.',
'metadata': {'breed': 'Golden Retriever'}
},
{
'id': 'dog4',
'text': 'Huskies are energetic working dogs bred for cold climates. They have thick fur and often blue eyes.',
'metadata': {'breed': 'Husky'}
},
{
'id': 'dog5',
'text': 'Siberian Huskies are active working dogs that thrive in cold weather. They are known for their thick coats and striking blue eyes.',
'metadata': {'breed': 'Husky'}
}
]
chroma_client.add_collection("dog_docs",exists_ok=True)
chroma_client.upsert(collection_name="dog_docs",docs=dog_docs)
chroma_client.list(collection_name='dog_docs')[{'id': 'dog1',
'text': 'The Golden Retriever is a friendly, intelligent breed known for its golden coat. They make excellent family pets and are great with children.',
'metadata': {'breed': 'Golden Retriever'},
'embedding': array([0.05813659, 0.02785078, 0.00372152, ..., 0.0241031 , 0.04067278,
0.00027733])},
{'id': 'dog2',
'text': 'German Shepherds are loyal, protective dogs often used in police work. They are highly trainable and good at various tasks.',
'metadata': {'breed': 'German Shepherd'},
'embedding': array([ 0.03066194, 0.00074497, 0.01366658, ..., -0.0081617 ,
-0.02541033, -0.01031713])},
{'id': 'dog3',
'text': 'The Golden Retriever is a friendly, intelligent breed with a beautiful golden coat. They are wonderful family pets that get along well with kids.',
'metadata': {'breed': 'Golden Retriever'},
'embedding': array([0.04065013, 0.02587523, 0.01385626, ..., 0.02761683, 0.03554016,
0.00553581])},
{'id': 'dog4',
'text': 'Huskies are energetic working dogs bred for cold climates. They have thick fur and often blue eyes.',
'metadata': {'breed': 'Husky'},
'embedding': array([-0.02400597, -0.00190817, -0.02052087, ..., -0.02349378,
-0.04244469, -0.02794758])},
{'id': 'dog5',
'text': 'Siberian Huskies are active working dogs that thrive in cold weather. They are known for their thick coats and striking blue eyes.',
'metadata': {'breed': 'Husky'},
'embedding': array([-0.01651848, 0.00581224, -0.02498457, ..., -0.00659932,
-0.03063537, -0.02282516])}]rag_chat = Chat(
model='gpt-4o-mini',
messages=[
{'role':'system','content':'''
You are a helpful assistant that answers questions about dogs.
I found the following documents that may be relevant to the user's question:
{% for doc in docs %}
{{doc['text']}}
{% endfor %}
'''},
{'role':'user','content':'{{question}}'},
]
)# import the basic building blocks
from stringdale import V,E,Define# define your diagram schema
with Define('hello world') as RAG:
V('get_docs',chroma_client.query)
E('Start->get_docs(**)')
V('chat',rag_chat)
E('get_docs->chat(docs=.)')
E('Start->chat(question=query)')
E('chat->End')
RAG.draw()
# run the diagram
diagram = RAG()
diagram_input = {
'query':'tell me some stuff about golden retrievers',
'collection_name':'dog_docs',
'k':2
}
for trace in diagram.run(diagram_input):
# you can write any logging,tracing logic here
trace.pprint(file='log_file.txt')
pass
# after running, access the output
diagram.output{'role': 'assistant',
'content': 'The Golden Retriever is a friendly and intelligent breed known for its beautiful golden coat. They make excellent family pets and are particularly good with children, making them wonderful companions for families.',
'meta': {'input_tokens': 198, 'output_tokens': 42}}! cat log_file.txt
! rm log_file.txtNode Start:
{ 'input': { 'collection_name': 'dog_docs',
'k': 2,
'q': 'tell me some stuff about golden retrievers'},
'output': { 'collection_name': 'dog_docs',
'k': 2,
'q': 'tell me some stuff about golden retrievers'}}
================================================================================
Node Start:
{ 'input': { 'collection_name': 'dog_docs',
'k': 2,
'query': 'tell me some stuff about golden retrievers'},
'output': { 'collection_name': 'dog_docs',
'k': 2,
'query': 'tell me some stuff about golden retrievers'}}
================================================================================
Node get_docs:
{ 'input': { 'collection_name': 'dog_docs',
'k': 2,
'query': 'tell me some stuff about golden retrievers'},
'output': [ { 'distance': 0.6993070840835571,
'id': 'dog3',
'metadata': {'breed': 'Golden Retriever'},
'text': 'The Golden Retriever is a friendly, intelligent breed '
'with a beautiful golden coat. They are wonderful '
'family pets that get along well with kids.'},
{ 'distance': 0.7122190594673157,
'id': 'dog1',
'metadata': {'breed': 'Golden Retriever'},
'text': 'The Golden Retriever is a friendly, intelligent breed '
'known for its golden coat. They make excellent family '
'pets and are great with children.'}]}
================================================================================
Node chat:
{ 'input': { 'docs': [ { 'distance': 0.6993070840835571,
'id': 'dog3',
'metadata': {'breed': 'Golden Retriever'},
'text': 'The Golden Retriever is a friendly, '
'intelligent breed with a beautiful golden '
'coat. They are wonderful family pets that '
'get along well with kids.'},
{ 'distance': 0.7122190594673157,
'id': 'dog1',
'metadata': {'breed': 'Golden Retriever'},
'text': 'The Golden Retriever is a friendly, '
'intelligent breed known for its golden coat. '
'They make excellent family pets and are '
'great with children.'}],
'question': 'tell me some stuff about golden retrievers'},
'output': { 'content': 'The Golden Retriever is a friendly and intelligent '
'breed known for its beautiful golden coat. They make '
'excellent family pets and are particularly good with '
'children, making them wonderful companions for '
'families.',
'meta': {'input_tokens': 198, 'output_tokens': 42},
'role': 'assistant'}}
================================================================================
Node End:
{ 'input': { 0: { 'content': 'The Golden Retriever is a friendly and '
'intelligent breed known for its beautiful golden '
'coat. They make excellent family pets and are '
'particularly good with children, making them '
'wonderful companions for families.',
'meta': {'input_tokens': 198, 'output_tokens': 42},
'role': 'assistant'}},
'output': { 'content': 'The Golden Retriever is a friendly and intelligent '
'breed known for its beautiful golden coat. They make '
'excellent family pets and are particularly good with '
'children, making them wonderful companions for '
'families.',
'meta': {'input_tokens': 198, 'output_tokens': 42},
'role': 'assistant'}}
================================================================================Reporting Bugs
Found bugs? Missing features? Feel free to open an issue
Contributing
Want to contribute to stringdale? Great! Feel free to submit a PR or discuss a feature you would like to add in the Issues.
To get a working development environment:
# clone the repo
git clone https://github.com/DeanLight/stringdale.git
cd stringdale
# install stringdale in editable dev mode
pip install -e ".[dev]"
# run pre-commit to install commit hooks
pre-commit installstringdale is developed using nbdev
# Make changes under nbs/ directory
# ...
# clean and export notebooks into a python library
nbdev_clean && nbdev_export
# make sure tests pass
nbdev_test
# Changing the docs? make sure that they are rendering correctly
nbdev_docs