pitkusharma/supreme-intellect-gemma4-E2B-it

VISIONConcurrent Unit Cost:1Model Size:5.1BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Apr 23, 2026Architecture:Transformer Featherless Exclusive Cold

The pitkusharma/supreme-intellect-gemma4-E2B-it is a 5.1 billion parameter Gemma-4-E2B-it model, fine-tuned and converted to GGUF format. This model supports a 32768 token context length and is notable for its vision capabilities, making it suitable for multimodal applications. It was fine-tuned using Unsloth, emphasizing efficient training.

Loading preview...

Model Overview

The pitkusharma/supreme-intellect-gemma4-E2B-it is a 5.1 billion parameter model based on the Gemma-4-E2B-it architecture. It has been fine-tuned and converted into the GGUF format, making it compatible with various inference engines like llama-cpp. A key aspect of its development is the use of Unsloth for efficient fine-tuning, which reportedly enabled 2x faster training.

Key Capabilities

  • Multimodal Support: This model is designed with vision capabilities, indicated by the availability of F16-mmproj.gguf files, allowing it to process both text and image inputs.
  • GGUF Format: Provided in GGUF format, ensuring broad compatibility and ease of use with llama-cli and similar tools.
  • Efficient Training: Leverages Unsloth for optimized and accelerated fine-tuning processes.

Usage and Compatibility

Users can interact with this model using llama-cli for text-only tasks or llama-mtmd-cli for multimodal applications, both supporting Jinja templating. For Ollama users, a specific process is outlined to create a unified bf16 model due to current limitations in handling separate mmproj files for vision models. This involves placing a Modelfile alongside the merged bf16 model and using the ollama create command.