The Flowise SDK for Python provides an easy way to interact with the Flowise API for creating predictions, supporting both streaming and non-streaming responses. This SDK allows users to create predictions with customizable options, including history, file uploads, and more.
- Support for streaming and non-streaming API responses
- Ability to include message history and file uploads
You can install the SDK via pip:
pip install flowise
Upgrade version:
pip install --upgrade flowise
from flowise import Flowise, PredictionData, IMessage, IFileUpload
def example_non_streaming():
# Initialize Flowise client
client = Flowise()
# Create a prediction without streaming
completion = client.create_prediction(
PredictionData(
chatflowId="abc",
question="What is the capital of France?",
streaming=False # Non-streaming mode
)
)
# Process and print the full response
for response in completion:
print("Non-streaming response:", response)
def example_streaming():
# Initialize Flowise client
client = Flowise()
# Create a prediction with streaming enabled
completion = client.create_prediction(
PredictionData(
chatflowId="abc",
question="Tell me a joke!",
streaming=True # Enable streaming
)
)
# Process and print each streamed chunk
print("Streaming response:")
for chunk in completion:
print(chunk)
if __name__ == "__main__":
# Run the non-streaming example
example_non_streaming()
# Run the streaming example
example_streaming()
- Increment version on
setup.py
pip install wheel
python setup.py sdist bdist_wheel
twine upload --skip-existing dist/*