Hyphonical/gemma-4-E4B-it-Uncensored-MAX-opus-4.7

VISIONConcurrency Cost:1Model Size:7.9BQuant:FP8Ctx Length:32kTool Calling:SupportedPublished:Apr 29, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

Hyphonical/gemma-4-E4B-it-Uncensored-MAX-opus-4.7 is a 7.9 billion parameter Gemma-4 based instruction-tuned language model developed by Hyphonical. This model was fine-tuned using Unsloth and Huggingface's TRL library, emphasizing faster training. It is designed for general instruction-following tasks, leveraging its uncensored base for broader response generation. The model's 32768 token context length supports processing extensive inputs for various applications.

Loading preview...

Model Overview

Hyphonical/gemma-4-E4B-it-Uncensored-MAX-opus-4.7 is a 7.9 billion parameter instruction-tuned language model developed by Hyphonical. It is based on the Gemma-4 architecture and was fine-tuned from the prithivMLmods/gemma-4-E4B-it-Uncensored-MAX model. The fine-tuning process utilized Unsloth and Huggingface's TRL library, which enabled a 2x faster training speed.

Key Characteristics

  • Architecture: Gemma-4 based.
  • Parameters: 7.9 billion.
  • Context Length: Supports a substantial 32768 tokens.
  • Training Efficiency: Fine-tuned with Unsloth for accelerated training.
  • Base Model: Derived from an uncensored variant, suggesting a broader range of response capabilities.

Intended Use Cases

This model is suitable for general instruction-following tasks where an uncensored base model is desired. Its efficient training process and substantial context window make it a candidate for applications requiring robust language understanding and generation over longer inputs.