The version number in the pyproject.toml file has been updated from 0.0.77 to 0.0.78. This is a chore commit as it does not introduce any new features or fix any bugs, but rather updates the version number to reflect the changes made in the code.
The version number in the pyproject.toml file has been updated from 0.0.76 to 0.0.77. This is a chore commit as it does not introduce any new features or fix any bugs, but only updates the version number.
🐛 fix(nodes.py): remove deepcopying of certain objects to improve performance
🐛 fix(loading.py): add error handling for empty documents in instantiate_textsplitter
The is_first_message logic in chat_manager.py was not correctly identifying the first message. The fix in this commit changes the comparison to <= 1 instead of == 0. In nodes.py, deepcopying of certain objects was removed to improve performance. This was done because deepcopying was not necessary for these objects. In loading.py, error handling was added to handle empty documents in instantiate_textsplitter. This was done to prevent errors when the source provided did not load correctly or was empty.
🔥 chore(test.yml): remove Python 3.11 from the matrix
Python 3.8 and 3.11 were removed from the matrix in lint.yml to reduce the number of Python versions being tested. Python 3.11 was removed from the matrix in test.yml as it is not yet released and therefore not supported.
Chroma uses onnxruntime which does not support 3.11 as of yet. This will be added back in once onnxruntime supports 3.11.
The display_name field was added to the LLM object in the test_agents_template.py and test_chains_template.py files. This field is used to display the name of the LLM in the UI.
🚀 Release 0.0.75
We are excited to announce the latest updates and features that have been added to Langflow! Here’s a summary of the recent changes:
New Features:
- Deploy Langflow using Langchain-Serve: Thanks to the hard work of @deepankarm, Langflow can now be deployed using Langchain-Serve to Jina AI Cloud. This integration brings enhanced functionality and improves the overall performance of Langflow. Just run `langflow --jcloud`.
- Copy and Paste Nodes and Flows: @anovazzi1 and @lucaseduoli have implemented the copy and paste functionality in Langflow. Now you can easily duplicate nodes and flows, improving productivity and workflow efficiency.
- Undo/Redo Functionality: @anovazzi1 and @lucaseduoli have also added an undo/redo feature to LangFlow, allowing you to easily revert and redo changes made to nodes and flows. This enhances the editing experience and provides greater control over your work.
Improvements and Fixes:
- Build: The latest update includes making Langchain-Serve optional, allowing for more flexibility in deployment.
- Documentation: The README has been updated to provide information about JCloud deployment.
- Validation Fixes: @ogabrielluiz has made several updates to improve validation in Langflow, including returning exception messages in responses and adding validation status icons and tooltips to node properties.
- Bug Fixes: Several bug fixes have been implemented, including resizing issues in the chat input, copy and paste bugs in the code area, and padding adjustments in the UI. Special thanks to @anovazzi1 and @lucaseduoli for their contributions in fixing these issues.
We appreciate the contributions from the following members:
- @deepankarm
- @anovazzi1
- @lucaseduoli
Thank you to everyone who has contributed to this release. Your efforts have made Langflow even better! If you have any feedback or suggestions, please don’t hesitate to reach out. Happy coding with Langflow! 🎉
- Modified UI styles in the app, making text more legible and buttons more visually appealing
- Changed chat trigger button to a gradient blue background to improve visibility and user experience
The position of the link to the Logspace website has been adjusted to be more centered on the page. The left margin has been increased by one unit to make it more visually appealing.
✨ feat(chatModal): add focus to chat input when modal is opened
The chat input was not resizing properly when the user typed more than one line of text. This was fixed by setting the height of the input to the scrollHeight of the input element. Additionally, the chat input is now focused when the chat modal is opened, which improves the user experience.
Fixes Text area is broken #328
✨ feat(chatModal): add focus to chat input when modal is opened
The chat input was not resizing properly when the user typed more than one line of text. This was fixed by setting the height of the input to the scrollHeight of the input element. Additionally, the chat input is now focused when the chat modal is opened, which improves the user experience.
The version number is now displayed in the footer of the application. The version number is fetched from the server using the /version endpoint and displayed in the footer. This allows users to easily identify which version of the application they are using.
🔖 chore(pyproject.toml): bump up package version to 0.0.75
The langchain package version has been updated to 0.0.176 to include the latest changes and improvements. The package version has also been bumped up to 0.0.75 to reflect the changes made to the dependencies.
The import statements for pathlib.Path and fastapi.staticfiles.StaticFiles were not used in the code and were therefore removed to improve code readability and maintainability.
This commit adds retry logic to the getAll API call in the TypesProvider
component. The retry logic will attempt to make the API call up to 5
times with an increasing delay between each attempt. If the API call is
successful, the state of the component will be updated with the
retrieved data. If the component is unmounted before the API call
resolves, the state will not be updated.
This commit adds retry logic to the getAll API call in the TypesProvider component. The retry logic will attempt to make the API call up to 5 times with an increasing delay between each attempt. If the API call is successful, the state of the component will be updated with the retrieved data. If the component is unmounted before the API call resolves, the state will not be updated.
