Frameworks are the central part of Jarvislabs. With each framework, we carefully curate the required dependencies and provide you with a seamless experience to explore, train, and deploy your AI models.

We are continuously adding different frameworks, aimed at making lives easier for all who want to explore the world of AI. Currently, we offer vanilla frameworks like PyTorch, and TensorFlow and customized frameworks like automatic 1111, vLLM, Ollama, etc.

Vanilla Frameworks

These frameworks allow you the most flexibility to explore and train your models. You can install any library or package you want.

Each instance comes with

  • VSCode
  • JupyterLab
  • API Endpoint
  • SSH Access

Specialised frameworks

As AI evolves, a lot of libraries are coming up to solve specific problems focussed on specific audiences like creators, musicians, and developers. We constantly curate such libraries and make them easier to use. We also add popular models to each of the frameworks from time to time, so that you can get started with them in a few clicks.

  • Hugging Face Autotrain - No code solution for training state of art AI models.
  • Cog - Bring in any of the popular COG templates and deploy them in a few clicks.
  • Automatic1111 - A place to unleash your creativity.
  • Text Generation Web UI - Explore different state of art LLM Models
  • vLLM - Deploy state of art LLM models in production.
  • Apple Ferret MLLM - Refer and Ground Anything Anywhere at Any Granularity.
  • Apple ml-mgie - Edit your images using Apple mgie through natural language.
  • Axolotl - Streamline finetuning of AI models.
  • Ollama - Run famous LLMs like Llama 2, Mistral, Gemma using ollama with your favorite UI.
  • Fooocus - High-quality text-to-image without needing much prompt engineering or parameter tuning.
  • Kohya - Kohya GUI allows you to set the training parameters and generate and run the required CLI commands to train the model.
  • ComfyUI - A Stable diffusion UI with a graph/nodes interface.

Comfyui workflows

We also have ready-to-use Comfyui workflows built just for you.

  • SDXL - Generate images using natural language in ComfyUI.
  • Animate-Diff - Generate cool animation videos using natural language in ComfyUI.
  • IP-Adapter FaceID - Create beautiful images using natural language in ComfyUI keeping the facial features intact.
  • Instant ID - Create stunning digital avatars using a single face image in ComfyUI.
  • Stable Cascade - Generate images with 2x super resolution and generation speed than base SDXL.
  • Stable Cascade Fooocus - Create images using stable cascade with different styles using Fooocus.