sd-civitai-browser-plus
Extension to access CivitAI via WebUI: download, delete, scan for updates, list installed models, assign tags, and boost downloads with multi-threading.
Stars: 264
sd-civitai-browser-plus is an extension designed for Automatic1111's Stable Difussion Web UI, providing features to browse models from CivitAI, check for updates, download specific model versions hassle-free, assign tags to models, access model info quickly, and download models with high-speed using Aria2. The extension offers a sleek and intuitive user interface, actively maintained with feature requests welcome. It also addresses known issues like frozen downloads with possible solutions. The tool is actively developed with regular updates and bug fixes, ensuring a smooth user experience.
README:
Extension for Automatic1111's Stable Diffusion Web UI
- Explore a wide range of models at your fingertips.
- Easily spot new updates and identify already installed models while browsing.
- Ability to scan all installed models for available updates.
- Get the specific model version and file you need hassle-free.
- Download queue to avoid waiting for finished downloads.
- Assign tags by scanning all installed models for automatic use in image generation.
- A button for each model card in txt2img and img2img to load it into the extension.
- A button under each image in model info to send its generation info to txt2img.
- Maximize your bandwidth for lightning-fast downloads.
- Enjoy a clutter-free, user-friendly interface, designed to enhance your experience.
- Feel free to send me your feature requests, and I'll do my best to implement them!
If you're experiencing issues with broken or frozen downloads, there are two possible solutions you can try:
-
Revert to the old download method: A solution could be to disable the "Download models using Aria2" feature.
This will switch back to the old download method, which may resolve the issue. -
Disable Async DNS for Aria2: If you're using a DNS manager program like PortMaster, try turning on the "Disable Async DNS for Aria2" option.
These settings can be found under the "Settings" tab in Web-UI and then under the "CivitAI Browser+" tile.
-
Download the latest version from this site and unpack the .zip
-
Navigate to your extensions folder (Your SD folder/webui/extensions)
-
Place the unpacked folder inside the extensions folder
-
Restart SD-WebUI
- Feature: Added support for DoRA (Requires SD-WebUI v1.9)
- Bug fix: No longer rescans models that were previously not found on CivitAI
- Bug fix: Fixed placement of HTML & api_info files when custom images location was used.
- Bug fix: Fixed incorrect json naming.
- New Setting: Send model from the cards CivitAI button to the browser, instead of showing a popup.
- Bug fix: No longer fetch HTML info when the file already exists and overwrite is off.
- Bug fix: Fix NSFW models update functions.
- Bug fix: Fixed CivitAI button on NSFW model cards.
- Bug fix: NSFW filtering fixed.
- Temp fix: Quick fix for the new NSFW system.
- NSFW models may be returned when NSFW is disabled, this is an issue with the public CivitAI API.
- Search results currently do not get influenced by the NSFW setting, also an issue with the API.
- Bug fix: Better automatic coloring for white theme.
- Bug fix: Extension now works again with the latest CivitAI API version!
- Cleanup: Re-wrote backend API handling code.
- Bug fix: Hotfix for a change in the public API which broke searching.
- Bug fix: Fixed incorrect permission display on model page.
- Feature: Ability to set-up a custom proxy for API requests and downloads.
- Feature: Use image API for prompt info, should speed up loading.
- Feature: Optimized javascript code, improved webpage speed for some users.
- New setting: Proxy settings to set-up custom proxy.
- New setting: Toggle for saving description to model json file. (this displays description on the cards)
- Bug fix: Broken default sub folder option fixed #217
- Feature: Local images now work in HTML files as preview. (credit: mx)
- Feature: Updated available base models.
- Bug fix: Fixed prompt info and model selection after CivitAI API update.
- Bug fix: Fixed "/" missing from default path/sub-folder.
- Feature: (BETA) Download queue! rearrange download order and remove models from queue
- Will likely contain bugs, still not completely finished.
- Feature: Customizable sub folder insertion options, choose what sub folder options you want!
- New setting: Toggle per prompt example image buttons
- New setting: Insert sub folder options
- Bug fix: Add to queue fixed, now properly gets enabled.
- Bug fix: Symlinks now get correctly recognized and used.
- Bug fix: No longer creates accidental sub folder when bulk downloading.
- Feature: Ability to send individual parts of image generation data to txt2img.
- Feature: Added compatibility for stable-diffusion-webui-forge fork.
- New setting: Use local images in the HTML
- Does not work in combination with the "Use local HTML file for model info" option!
- New setting: Store the HTML and api_info in the custom images location
- Bug fix: New HTML model info now scales with width so it should always fit.
- Bug fix: Various bug fixes to the "Update model info & tags" function.
- Bug fix: Auto save all images now uses correctly uses custom image path if set.
- Bug fix: "Save model info" button should no longer return errors.
- Bug fix: Old download method (non Aria2) should now work again.
- Feature: New txt2img and img2img model info overlay on CivitAI button press.
- Feature: Base Model as sub folder option.
- Feature: Ability to multi-download to selected folder.
- Feature: Use the same folder as older versions when updating using multi-download.
- Feature: txt2img and img2img CivitAI buttons can use local HTML file, toggle in settings.
- Note: Save images no longer saves .html and API info, save model info does this instead now.
- New setting: Save API info of model when saving model info.
- New setting: Automatically save all images after download.
- New setting: Use local HTML file for model info.
- Bug fix: better JSON decode, now forces UTF-8
- Bug fix: Now uses the proper default file when using multi-download
- Bug fix: Hide early access models fix, now works when published_at does not exist in API.
- Bug fix: Fix attempt for queue clearing upon download fail.
- Bug fix: Removed default API Key since it gets blocked after many downloads.
- Because of this it's now required for some downloads to use a personal CivitAI key, this can be set in the the settings tab of SD-WebUI under the CivitAI Browser+ tab.
- Bug fix: Fixed bug when selecting a model from txt2img/img2img that doesn't exist on CivitAI.
- Bug fix: Changed model selection to Model ID instead of model name
- This previously caused issues when 2 models were named the same.
- Bug fix: Fixed an issue where the default file was not properly used by default.
- Bug fix: Fixed some tiles not being selectable due to having "'" in its title
- Bug fix: Now automatically removes residual Aria2 files.
- Bug fix: Fix version detection for non standard SD-WebUI versions.
- Bug fix: Retry to fetch ModelID if previously not found in update functions.
- Bug fix: Style fix for when the Lobe theme is used in SD-WebUI
- Bug fix: Better required packages import error catching.
- Bug fix: Fixed CivitAI button scaling in txt2img and img2img tabs.
- Bug fix: Added ability to handle models that have no hashes saved.
- Bug fix: Generate hash toggle in update models was inverted (silly mistake, sry bout that)
- Bug fix: Better error detection if no model IDs were retrieved during update functions.
- Bug fix: Better error handling if a local model does not exist on CivitAI
- Bug fix: Fixed an
api_response
issue in the update model functions - Bug fix: Reverted automatically retrieving base models to fix startup issues
- Bug fix: Better error description if a model no longer exists on CivitAI
- Bug fix: Primary file is now used as default file.
- Bug fix: Search after updating models no longer returns errors.
- Feature: Extension now automatically retrieves latest base models from CivitAI.
- Bug fix: Hotfix for functionality with SD.Next
- Feature: A toggle for One-Time hash generation for externally downloaded models.
- Feature: Updated extension settings layout for SD-WebUI 1.7.0 and higher.
- Bug fix: Set default value of Lora & LoCON combination based on SD-WebUI version.
- Bug fix: LORA models with embedding files now get placed inside embeddings folder.
- Bug fix: Better tile count handling to avoid issues with incorrect tile count.
- Bug fix: Better settings saving/loading to prevent writing issues.
- Bug fix: Early Access models now get correctly hidden/detected.
- Bug fix: Better timeout/offline server detection for options in "Update Models" tab.
- Bug fix: Better error detection if required packages were not installed/imported.
- Bug fix: Download button now displays as "Add to queue" during active download.
- Feature: Send to txt2img, Send any image in the model info to txt2img.
- Feature: Added new Base model filters:
- SD 1.5 LCM, SDXL 1.0 LCM, SDXL Distilled, SDXL Turbo, SVD, SVD XT
- Feature: Hide installed models filter toggle.
- Feature: Better display of permissions and tags in model info.
- New setting: Append sub folders to custom image path.
- New setting: Toggle gif/video playback, Disable if videos are causing high CPU usage.
- Bug fix: Better handling if hash is not found.
- Feature: Download queue! Ability to add downloads to a queue. (Finally!)
- Feature: Checkboxes to download multiple models at once.
- This will automatically use the first version and first file of the selected model(s).
- Will use the default sub folders per content type defined in sub folder settings.
- Feature: "Select all" button to select all downloadable models at once.
- Feature: "Open on CivitAI" button when viewing a models metadata in txt2img or img2img.
- Will only display if the model's info has been saved to the .json after v3.0
- Feature: Ability to rename model filename
- Note that it's not recommended to change the filename since some checks rely on it.
- Bug fix: Fixed display of saved .html files.
- Bug fix: Removed potential illegal characters from file name/path name.
- Bug fix: Fixed case sensitive sorting of sub folders.
- Feature: "Overwrite any existing previews, tags or descriptions" Toggle in Update tab.
- Feature: Added content type "All" to model scanning to select all content types.
- Bug fix: Folders starting with "." now no longer show sub folders.
- Bug fix: Added headers to simulate browser request. (May fix issues for users from Russia)
- Feature: New button on each model card in txt2img and img2img to view it in the extension.
- Feature: NSFW toggle now properly impacts search results.
- Feature: Ability to set [\Model Name] & [\Model Name\Version Name] as default sub folders.
- New setting: Hide sub folders that start with a '.'
- Bug fix: Preview HTML is now emptied when loading a new page.
- Bug fix: Buttons now correctly display when loading new page.
- Bug fix: Fixed compatibility with SD.Next. (again)
- Bug fix: Emptied tags, base model, and filename upon loading new page.
- Bug fix: Filter change detection fixed
- Feature: Ability to download/update model preview images in Update Models tab.
- Feature: "Update model tags" changed to "Update model info & tags".
- The option now saves tags, description and base model version.
- This also applies to the browser, saved tags is changed to save model info.
- Bug fix: Archived models are now hidden since they cannot be used.
- New setting: Custom save images location
- New setting: Default sub folders
- Any sub folders you have will be able to be selected as default, per content type.
- If a content type doesn't appear, then it means there are no subfolders in that type.
- Bug fix: Unreleased models caused a crash, now hidden by default since they can't be used.
- New setting: Show console logs during update scanning.
- Bug fix: Scan for update no longer prints incorrect info about outdated models.
- Bug fix: Removed bad logic which triggered the same function multiple times.
- Cleanup: Optimized functions and improved the speed of selecting models.
- Feature: Filter option to show favorited models. (requires personal API key)
- Feature: Back to top button when viewing model details.
- New setting: Page navigation as header. (keeps page navigation always visible at the top)
- Bug fix: Aria2 now restarts when UI is reloaded.
- Bug fix: SHA256 error fixed if .json files don't contain it.
- Bug fix: Cleaned up javascript code.
- New setting: Hide early access models (EA models are only downloadable by supporters)
- New setting: Personal CivitAI API key (Text field to insert personal API key)
- Useful for CivitAI supporters, you can use your own API Key to allow downloading Early Access models
- Bug fix: Extension now works with
no gradio queue
flag. - Bug fix: Auto disable Aria2 on MacOS due to incompatibility.
- Bug fix: Now properly works on SD.Next again.
- Bug fix: Download progression and cancelling is no longer broken on old download method.
- Bug fix: Extension now correctly downloads models where it is required to be logged in.
- Bug fix: Extension no longer attempts to install already installed requirements.
- Bug fix: Removed pre-load of default page, caused issues for some users.
- Bug fix: Fixed internal model naming, caused issues when model names included '
- Bug fix: Different host for .svg icons, caused issues with MalwareBytes.
- Bug fix: Preview saving was broken due to passing the wrong file path.
- Feature: Base Model filter now impacts search results.
- Feature: Ability to input model URL into search bar to find corresponding model.
- Bug fix: Adetailer models now get placed in the correct folder
- Bug fix: Page slider broke the Next Page button when loaded from "Update Models".
- Bug fix: "Save settings as default" button inserted broken .json data.
- Bug fix: Triggering "Scan for available updates" twice resulted in an error.
- Bug fix: LORA content type was broken when "Treat LoCon as LORA" was turned on.
- Feature: Custom page handling when scanning models.
- Bug fix: Model scan feature now works for large model count (+900)
- Bug fix: Better broken .json error handling
- Bug fix: Gifs did not display properly.
- Bug fix: Video's no longer save as preview since they cannot be used.
- Bug fix: Filter window was not hidden by default.
- Feature: Redesign of UI.
- Feature: New dropdown with filter settings.
- Feature: Button to save current filter settings as default. (requires restart)
- Feature: Tag box can now be typed in to save custom tags.
- Feature: Delete function removes any unpacked files.
- Feature: Updated available content types:
- Upscaler
- MotionModule
- Wildcards
- Workflows
- Other
- Feature: Videos can now also be displayed on preview cards and in the model info.
- Feature: Automatically scans upscaler type by looking through model's description.
- Feature: Automatically identify correct folder for wildcards based on extension.
- Bug fix: Version ID got saved instead of correct Model ID after download.
- Bug fix: [Installed] tag was only assigned to latest installed version.
- Bug fix: Folder location didn't update when selecting different version/file.
- Bug fix: Version scanning didn't properly scan sha256 in uppercase.
- Feature: You can now refresh by pressing Ctrl+Enter and Alt+Enter.
- Bug fix: Auto unpack feature was unpacking unintended archives, now only unpacks .zip.
- New setting: Option to toggle automatically unpacking .zip models.
- Bug fix: Error wasn't catched when file path was incorrect.
- Feature: Able to download multiple files from each version.
- Bug fix: Models did not get deleted properly when in nested folders.
- Bug fix: Wrong sha256 was being saved after downloading.
- Bug fix: Wrong default folder was used when installed model got selected.
- Feature: File deletion now uses both SHA256 and file name to detect correct file.
- New setting: Option to toggle automatically inserting 2 default sub folders.
- New setting: Option to toggle installing LoCON's in LORA folder.
- Bug fix: Default file was incorrect when selecting a model.
- Bug fix: Next Page caused an error when changing content type.
- Feature: Ability to load all selected installed models into browser in Update Models tab.
- Feature: Installed/outdated models check is now done using SHA256 + file name.
- Feature: Ability to select multiple content Types when searching and scanning.
- Feature: Greatly improved speed of model scanning if model ID is saved in .json
-
Feature: Redesign of model page by ManOrMonster
-
Model page changes (https://github.com/BlafKing/sd-civitai-browser-plus/pull/33)
- Redesigned the look of the model page.
- Added link to model page on CivitAI. Click on model name to open.
- Added link to uploader/creator page on CivitAI. Click creator name to open.
- Added CivitAI avatar display.
- Separate description section.
- First sample image is marked with data attribute and downloaded as preview image instead of grabbing first in model HTML. This guarantees that the first sample image (not avatar or image in description) is used when downloading the model.
- Sample images are marked with data attribute so that only they are downloaded when using "Save Images" (no description images or avatar).
- Removed trained tags from info since they are displayed above.
- Each sample image has its own section.
- Sample images zoom in when clicked, zoom out when clicking anywhere.
- Forced width is removed from sample image URLs so that nice big images can be viewed.
- Metadata is arranged so that the most commonly used data is at the top, no more searching for prompts.
- Extra metadata is in accordion labeled "More details...". This is especially useful to hide insanely large ComfyUI JSON.
- Feature: Error detection during Aria2 downloads.
- Bug fix: Avoid starting Aria2 RPC multiple times with better port check.
- Bug fix: Fixed dynamic tile status updates after deleting/downloading.
- Feature: Ability to scan all installed models for available updates.
- Feature: Model ID and sha256 get saved to .json after scanning or downloading a model.
- Bug fix: Fixed crash when base model is not found.
- Bug fix: No longer overwrite sha256 and model ID in existing .json.
- Bug fix: Fixed pathing for Unix systems
- Bug fix: Extra checks to prevent deleting unintentional files.
- Feature: Models get moved to trash instead of fully deleted.
- Feature: Update tags for all installed models!
- Feature: Tabs for Browsing and updating Tags.
- Feature: Buttons to select which folders to update tags in.
- Feature: Added Civit AI settings tab
- New setting: Disable downloading with Aria2. (will use old download method instead)
- New setting: Disable using Async DNS. (can fix issues for some users who use DNS managing programs)
- New setting: Show Aria2 logs in the CMD.
- New setting: Set the amount of connections when downloading a model with Aria2.
(The optimal connection count is different per user, try to find the lowest option which still gives you full bandwidth speed)
- Feature: Included Motrix Aria2 version.
- Feature: Max connections per server set to 64 and split file set 64.
- Feature: Aria2 is now shipped with this extension for Linux as well.
- Cleanup: Split up script into multiple files for improved oversight/readability.
- Cleanup: Centered model icons
- Bug fix: Added back old download function if aria2 fails.
- Feature: Faster downloads by using Aria2.
- Feature: More info about current download: Speed, ETA, File Size and % completion.
- Feature: Sub Folder list now contains 2 default options:
/{Model name}
&/{Model name}/{Version name}
- Feature: Ability to download different file types per version.
- Feature: NSFW Toggle is now dynamic.
- Feature: Version list now dynamically updates after download.
- Cleanup: Rearranged/Resized UI elements.
- Bug fix: Downloading models now uses file ID instead of names.
- Bug fix: NSFW Toggle no longer hides images tagged as "Soft".
- Bug fix: Fixed each model load running twice.
- Bug fix: Download button did not get re-enabled properly.
- Bug fix: Tile status did not get updated properly when download failed.
- Feature: Base Model filtering dims tiles instead of hiding.
- Bug fix: NSFW Blur increases with tile size.
- Bug fix: Dynamic tile status after installation & deletion now correctly detects other versions.
- Feature: Introduced seperate download progress bar, browse while downloading.
- Feature: no more force refresh after installing, cancelling and deleting.
- Feature: Added toggle to sort Tiles by date.
- Feature: Dynamic changing of tile borders after installation & deletion.
- Removal: 'Auto delete old version' removed since it relied on a reload.
- Bug fix: Page count is now always correclty read when refreshing.
(You can fill in the page number you'd like to visit and press refresh to go to that page) - Feature: 'Filter Base Model' to dynamically hide any unselected Base models.
(Please note: This does not impact search results, since the CivitAI API does not yet support this)
- Feature: Slider to change tile size.
- Feature: Download Folder textbox which can be used to define a custom download path.
- Feature: Sub Folder Dropdown to select any available subfolder(s) as download location.
- Feature: Display a timed out message instead of an error icon.
- Bug fix: Nested files can now be detected as installed or outdated.
- Bug fix: Auto selects corresponding folder of any installed models.
- Bug fix: Better cancellation logic to prevent downloads from continuing.
- Feature: Download progress bar is now on web page instead of CMD.
- Feature: Added Cancel and Delete buttons.
- Feature: Download button will now change according to circumstances:
- Cancel button if there's a current download.
- Delete button if the selected version is installed.
- Cleanup: Better margin fixes with theme detection.
- Bug fix: Delete option now also removes .json files.
- Bug fix: Buttons are now disabled during download. (except cancel button)
- Bug fix: Fixed tag saving bugs/oversights.
- Bug fix: Trained tags display now do not include the model itself.
- Feature: 'Save Tags' button saves tags to a .json file which gets used in image creaton.
(If a LORA with saved tags is used it will automatically input all tags into the txt box in image creation) - Feature: 'Save tags after download' toggle to automatically save .json tags.
- Cleanup: Removed "Get model info" button.
- Cleanup: Removed download link box.
- Cleanup: Removed "No" from search options.
- Cleanup: Added border radius to cards.
- Cleanup: Improved padding based on if Lobe theme is being used.
- Feature: Automatically saves preview image when downloading a model.
- Feature: Added [installed] text suffix for any versions that are installed in the 'Version' tab.
- Cleanup: Changed 'Model Filename' from a dropbox to a textbox.
- Cleanup: Made bottom textboxes non typeable.
- Cleanup: Disabled bottom buttons when no model is selected.
- Bug fix: Margin error on the latest tile.
- Bug fix: Version checking is now case sensitive.
- Bug fix: Default verison in version tab shows installed version.
- Feature: Dropdown box which can filter by time period.
- Cleanup: 'Content type' changed from buttons to a dropdown box.
- Bug fix: Fixed tiles not reloading when already pressed.
- Feature: 'Refresh' now reloads the current page unless any options have been changed.
- Feature: Made the page refresh after a download and made it load during one.
- Feature: Orange glow for any outdated installed packages.
- Feature: 'Delete old version after download' option.
- Feature: Ability to manually fill in a page number to load the corresponding page.
- Cleanup: Removed new folder option.
- Cleanup: Made the glow around frames always visible without hovering.
- Pulled fork from: SignalFlagZ's Fork v1.1.0
For Tasks:
Click tags to check more tools for each tasksFor Jobs:
Alternative AI tools for sd-civitai-browser-plus
Similar Open Source Tools
sd-civitai-browser-plus
sd-civitai-browser-plus is an extension designed for Automatic1111's Stable Difussion Web UI, providing features to browse models from CivitAI, check for updates, download specific model versions hassle-free, assign tags to models, access model info quickly, and download models with high-speed using Aria2. The extension offers a sleek and intuitive user interface, actively maintained with feature requests welcome. It also addresses known issues like frozen downloads with possible solutions. The tool is actively developed with regular updates and bug fixes, ensuring a smooth user experience.
Ollama-Colab-Integration
Ollama Colab Integration V4 is a tool designed to enhance the interaction and management of large language models. It allows users to quantize models within their notebook environment, access a variety of models through a user-friendly interface, and manage public endpoints efficiently. The tool also provides features like LiteLLM proxy control, model insights, and customizable model file templating. Users can troubleshoot model loading issues, CPU fallback strategies, and manage VRAM and RAM effectively. Additionally, the tool offers functionalities for downloading model files from Hugging Face, model conversion with high precision, model quantization using Q and Kquants, and securely uploading converted models to Hugging Face.
extensionOS
Extension | OS is an open-source browser extension that brings AI directly to users' web browsers, allowing them to access powerful models like LLMs seamlessly. Users can create prompts, fix grammar, and access intelligent assistance without switching tabs. The extension aims to revolutionize online information interaction by integrating AI into everyday browsing experiences. It offers features like Prompt Factory for tailored prompts, seamless LLM model access, secure API key storage, and a Mixture of Agents feature. The extension was developed to empower users to unleash their creativity with custom prompts and enhance their browsing experience with intelligent assistance.
ai-driven-dev-community
AI Driven Dev Community is a repository aimed at helping developers become more efficient by utilizing AI tools in their daily coding tasks. It provides a collection of tools, prompts, snippets, and agents for developers to integrate AI into their workflow. The repository is regularly updated with new resources and focuses on best practices for using AI in development work. Users can find tools like Espanso, ChatGPT, GitHub Copilot, and VSCode recommended for enhancing their coding experience. Additionally, the repository offers guidance on customizing AI for developers, installing AI toolbox for software engineers, and contributing to the community through easy steps.
polyfire-js
Polyfire is an all-in-one managed backend for AI apps that allows users to build AI applications directly from the frontend, eliminating the need for a separate backend. It simplifies the process by providing most backend services in just a few lines of code. With Polyfire, users can easily create chatbots, transcribe audio files, generate simple text, manage long-term memory, and generate images. The tool also offers starter guides and tutorials to help users get started quickly and efficiently.
JamAIBase
JamAI Base is an open-source platform integrating SQLite and LanceDB databases with managed memory and RAG capabilities. It offers built-in LLM, vector embeddings, and reranker orchestration accessible through a spreadsheet-like UI and REST API. Users can transform static tables into dynamic entities, facilitate real-time interactions, manage structured data, and simplify chatbot development. The tool focuses on ease of use, scalability, flexibility, declarative paradigm, and innovative RAG techniques, making complex data operations accessible to users with varying technical expertise.
SurfSense
SurfSense is a tool designed to help users save and organize content from the internet into a personal Knowledge Graph. It allows users to capture web browsing sessions and webpage content using a Chrome extension, enabling easy retrieval and recall of saved information. SurfSense offers features like powerful search capabilities, natural language interaction with saved content, self-hosting options, and integration with GraphRAG for meaningful content relations. The tool eliminates the need for web scraping by directly reading data from the DOM, making it a convenient solution for managing online information.
llm-answer-engine
This repository contains the code and instructions needed to build a sophisticated answer engine that leverages the capabilities of Groq, Mistral AI's Mixtral, Langchain.JS, Brave Search, Serper API, and OpenAI. Designed to efficiently return sources, answers, images, videos, and follow-up questions based on user queries, this project is an ideal starting point for developers interested in natural language processing and search technologies.
Simplifine
Simplifine is an open-source library designed for easy LLM finetuning, enabling users to perform tasks such as supervised fine tuning, question-answer finetuning, contrastive loss for embedding tasks, multi-label classification finetuning, and more. It provides features like WandB logging, in-built evaluation tools, automated finetuning parameters, and state-of-the-art optimization techniques. The library offers bug fixes, new features, and documentation updates in its latest version. Users can install Simplifine via pip or directly from GitHub. The project welcomes contributors and provides comprehensive documentation and support for users.
CursorLens
Cursor Lens is an open-source tool that acts as a proxy between Cursor and various AI providers, logging interactions and providing detailed analytics to help developers optimize their use of AI in their coding workflow. It supports multiple AI providers, captures and logs all requests, provides visual analytics on AI usage, allows users to set up and switch between different AI configurations, offers real-time monitoring of AI interactions, tracks token usage, estimates costs based on token usage and model pricing. Built with Next.js, React, PostgreSQL, Prisma ORM, Vercel AI SDK, Tailwind CSS, and shadcn/ui components.
easydiffusion
Easy Diffusion 3.0 is a user-friendly tool for installing and using Stable Diffusion on your computer. It offers hassle-free installation, clutter-free UI, task queue, intelligent model detection, live preview, image modifiers, multiple prompts file, saving generated images, UI themes, searchable models dropdown, and supports various image generation tasks like 'Text to Image', 'Image to Image', and 'InPainting'. The tool also provides advanced features such as custom models, merge models, custom VAE models, multi-GPU support, auto-updater, developer console, and more. It is designed for both new users and advanced users looking for powerful AI image generation capabilities.
omniscient
Omniscient is an advanced AI Platform offered as a SaaS, empowering projects with cutting-edge artificial intelligence capabilities. Seamlessly integrating with Next.js 14, React, Typescript, and APIs like OpenAI and Replicate, it provides solutions for code generation, conversation simulation, image creation, music composition, and video generation.
gemini-android
Gemini Android is a repository showcasing Google's Generative AI on Android using Stream Chat SDK for Compose. It demonstrates the Gemini API for Android, implements UI elements with Jetpack Compose, utilizes Android architecture components like Hilt and AppStartup, performs background tasks with Kotlin Coroutines, and integrates chat systems with Stream Chat Compose SDK for real-time event handling. The project also provides technical content, instructions on building the project, tech stack details, architecture overview, modularization strategies, and a contribution guideline. It follows Google's official architecture guidance and offers a real-world example of app architecture implementation.
Magick
Magick is a groundbreaking visual AIDE (Artificial Intelligence Development Environment) for no-code data pipelines and multimodal agents. Magick can connect to other services and comes with nodes and templates well-suited for intelligent agents, chatbots, complex reasoning systems and realistic characters.
J.A.R.V.I.S.-Ai-Assistant-V1-
Jarvis Version 3 is a versatile personal assistant application designed to enhance productivity by automating common tasks. It can interact with websites and applications, perform searches, manage device functions, and control music. Users can give commands to open websites, search on Google or YouTube, scroll pages, manage applications, check time, internet speed, battery percentage, battery alerts, charging status, play music, and synchronize clapping with music. The tool offers features for web navigation, search functionality, scrolling, application management, device management, and music control.
For similar tasks
sd-civitai-browser-plus
sd-civitai-browser-plus is an extension designed for Automatic1111's Stable Difussion Web UI, providing features to browse models from CivitAI, check for updates, download specific model versions hassle-free, assign tags to models, access model info quickly, and download models with high-speed using Aria2. The extension offers a sleek and intuitive user interface, actively maintained with feature requests welcome. It also addresses known issues like frozen downloads with possible solutions. The tool is actively developed with regular updates and bug fixes, ensuring a smooth user experience.
For similar jobs
weave
Weave is a toolkit for developing Generative AI applications, built by Weights & Biases. With Weave, you can log and debug language model inputs, outputs, and traces; build rigorous, apples-to-apples evaluations for language model use cases; and organize all the information generated across the LLM workflow, from experimentation to evaluations to production. Weave aims to bring rigor, best-practices, and composability to the inherently experimental process of developing Generative AI software, without introducing cognitive overhead.
LLMStack
LLMStack is a no-code platform for building generative AI agents, workflows, and chatbots. It allows users to connect their own data, internal tools, and GPT-powered models without any coding experience. LLMStack can be deployed to the cloud or on-premise and can be accessed via HTTP API or triggered from Slack or Discord.
VisionCraft
The VisionCraft API is a free API for using over 100 different AI models. From images to sound.
kaito
Kaito is an operator that automates the AI/ML inference model deployment in a Kubernetes cluster. It manages large model files using container images, avoids tuning deployment parameters to fit GPU hardware by providing preset configurations, auto-provisions GPU nodes based on model requirements, and hosts large model images in the public Microsoft Container Registry (MCR) if the license allows. Using Kaito, the workflow of onboarding large AI inference models in Kubernetes is largely simplified.
PyRIT
PyRIT is an open access automation framework designed to empower security professionals and ML engineers to red team foundation models and their applications. It automates AI Red Teaming tasks to allow operators to focus on more complicated and time-consuming tasks and can also identify security harms such as misuse (e.g., malware generation, jailbreaking), and privacy harms (e.g., identity theft). The goal is to allow researchers to have a baseline of how well their model and entire inference pipeline is doing against different harm categories and to be able to compare that baseline to future iterations of their model. This allows them to have empirical data on how well their model is doing today, and detect any degradation of performance based on future improvements.
tabby
Tabby is a self-hosted AI coding assistant, offering an open-source and on-premises alternative to GitHub Copilot. It boasts several key features: * Self-contained, with no need for a DBMS or cloud service. * OpenAPI interface, easy to integrate with existing infrastructure (e.g Cloud IDE). * Supports consumer-grade GPUs.
spear
SPEAR (Simulator for Photorealistic Embodied AI Research) is a powerful tool for training embodied agents. It features 300 unique virtual indoor environments with 2,566 unique rooms and 17,234 unique objects that can be manipulated individually. Each environment is designed by a professional artist and features detailed geometry, photorealistic materials, and a unique floor plan and object layout. SPEAR is implemented as Unreal Engine assets and provides an OpenAI Gym interface for interacting with the environments via Python.
Magick
Magick is a groundbreaking visual AIDE (Artificial Intelligence Development Environment) for no-code data pipelines and multimodal agents. Magick can connect to other services and comes with nodes and templates well-suited for intelligent agents, chatbots, complex reasoning systems and realistic characters.