Experience the next evolution of dataset creation with Caption Creator, now featuring a complete UI overhaul and powerful new capabilities. This fast, fully portable GUI tool is designed to generate exceptional image captions and tags with unparalleled ease. It's the ultimate assistant for creating high-quality datasets for AI models, perfect for both LoRA training and advanced image prompting.The application runs entirely on your local machine, ensuring privacy and uncensored output.Official GitHub - <a target="_blank" rel="ugc" href="https://github.com/Merserk/Caption-Creator">https://github.com/Merserk/Caption-Creator</a>Support on Patreon - <a target="_blank" rel="ugc" href="https://www.patreon.com/MM744">https://www.patreon.com/MM744</a><h3 id="features:">Features:</h3><ul><li>Dual Generation Modes: Seamlessly switch between generating detailed Captions or concise, comma-separated Tags.</li><li>Intelligent Tag Formatting: Automatically cleans AI output for tags into a perfect, single-line, comma-separated list, removing notes and unwanted formatting.</li><li>Powerful Batch Processing: Process entire folders of images in a single run with a clear, gallery-style progress view.</li><li>Portable &amp; Self-Contained: No installation needed. Runs from a single folder with its own embedded Python, ensuring it just works.</li><li>Uncensored Local AI: Utilizes locally run models for full creative freedom without content filters.</li><li>Complete UI Overhaul: A sleek, modern, and responsive dark-theme interface designed for a professional workflow.</li><li>LM Studio Integration: Power-users can now connect directly to a running LM Studio instance to use any compatible model.</li><li>Direct Image Pasting: Instantly process an image by simply pasting it from your clipboard (Ctrl+V).</li><li>Interactive Model Management: Download, delete, and manage models directly from within the application's intuitive modal interface.</li><li>Built-in ZIP Archiving: Save your entire generation run (images and text files) into a single ZIP archive with one click.</li><li>Prompt Enrichment: Add extra context or instructions to the AI on the fly to guide its output without editing config files.</li><li>Intuitive Controls: Replaced basic inputs with custom sliders and switches for a more tactile and efficient user experience.</li><li>VRAM Optimization: Choose from models tailored for different GPU VRAM capacities (5GB, 8GB, 10GB, 20GB).</li><li>Low-VRAM Mode: A dedicated checkbox to further reduce VRAM usage on memory-constrained systems.</li><li>Keep Model Loaded: An option to keep the AI model in VRAM after a task, dramatically speeding up subsequent generations.</li><li>Automated Shutdown: Automatically shut down your PC after a long batch process completes.</li><li>Flexible Formatting: Use Trigger Words, define a Max Word count, and format captions as a single paragraph.</li><li>Easy Access: Instantly copy generated text to the clipboard or open the output folder directly from the UI.</li></ul><h3 id="how-to-use:">How to Use:</h3><ol><li>Download &amp; Unpack: Download the program and unpack the .zip archive into a folder.</li><li>Run the Application: Double-click Caption Creator.exe to launch the program.</li><li>Manage Your Model:<ul><li>Click the "Model / VRAM Configuration" button to open the model selection panel.</li><li>To use a built-in model: If a model is not marked "Available," click the download icon (📥) next to it. The app will download and install it for you.</li><li>To use LM Studio: Select the "Custom (LM Studio)" option and click "Connect" to link with your running LM Studio server.</li><li>Select your desired model from the list to make it active.</li></ul></li><li>Load Your Image(s):<ul><li>Single Mode: Drag &amp; drop an image, click to browse, or paste an image from your clipboard.</li><li>Batch Mode: Drag &amp; drop multiple images or click to select a batch of files.</li></ul></li><li>Configure &amp; Generate:<ul><li>Choose your generation type (Captions or Tags).</li><li>Adjust settings like Max Words, Trigger Words, or enable options like Low-VRAM mode.</li><li>Click Generate.</li></ul></li><li>Review Output: Watch the live progress in the status window. Your generated text and images will appear in the right-hand panel and are automatically saved to the output folder, neatly organized by run.</li></ol><h3></h3><h3 id="output-example:">Output Example:</h3><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/3fe82953-e2c8-47a9-b645-7d2cc1798463/width=525/3fe82953-e2c8-47a9-b645-7d2cc1798463.jpeg" />Captions (Format as Single Paragraph enabled):The image is a digital illustration of a female character from the video game "Street Fighter II." She has blonde hair styled in two braids, each tied with red ribbons. Her skin tone is fair, and she has blue eyes that are focused intently forward. She wears a red beret hat with a white button on the front center, a green sleeveless tank top, and red fingerless gloves. Her right arm is extended forward, her fist clenched as if preparing for a punch or throwing a punch. Her left arm is slightly behind her body, also extending forward but less prominently positioned. The background is a gradient from dark gray at the top to black at the bottom, providing contrast to the character's bright clothing colors. The character's expression is one of determination and focus, with her mouth slightly open showing small teeth. Her muscular build is evident through the defined lines of her arms and shoulders. The overall style of the illustration is highly detailed and dynamic, typical of the "Street Fighter" series' art design. The image is framed by gray borders on both sides and top/bottom, creating a rectangular composition. This framing effect adds depth and focus to the central character. The entire image conveys a sense of strength and readiness for combat.Tags:digital art, female character, muscular build, green tank top, red beret with white button, red fingerless gloves, blonde hair in braid, intense expression, right arm extended forward, clenched teeth, dark blue gradient background, vibrant colors, anime style, strong pose, upper body, dynamic lighting, high contrast, Illustrious quality, fighting game character, Camilla (Street Fighter), serious demeanor, confident stance, athletic physique, determined look, bold outlines, realistic shading, vivid details, medium close-up shot, action pose, character design, video game aesthetics, strong facial features, dynamic composition, energetic pose, fierce attitude, expressive eyes, powerful stance, combat-ready appearance<h3></h3><h3 id="tags:">Tags:</h3>#caption-creator #dataset #tagging #portable #uncensored #batch-processing #memory-optimized

Files with SHA256: c476032202dc6fbd18ba0304edcf0fbfa8a2ed848ebc6a82894ffa566bbdee15