DrRiceIO7/gemma-3-270m-MaxSlop-3e-high

TEXT GENERATIONConcurrent Unit Cost:1Model Size:0.3BQuant:BF16Context Size:32kPublished:Jun 26, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

The DrRiceIO7/gemma-3-270m-MaxSlop-3e-high is a 0.3 billion parameter Gemma-3 model developed by DrRiceIO7, fine-tuned from unsloth/gemma-3-270m-it. This model was trained using Unsloth and Huggingface's TRL library, emphasizing faster training. With a 32768 token context length, it is suitable for applications requiring efficient processing of longer sequences.

Loading preview...

Model Overview

DrRiceIO7/gemma-3-270m-MaxSlop-3e-high is a compact 0.3 billion parameter language model, developed by DrRiceIO7. It is a fine-tuned variant of the unsloth/gemma-3-270m-it base model, leveraging the Unsloth framework and Huggingface's TRL library for optimized training efficiency.

Key Characteristics

  • Architecture: Based on the Gemma-3 family, providing a balance of performance and efficiency.
  • Parameter Count: Features 0.3 billion parameters, making it a lightweight option for deployment.
  • Context Length: Supports a substantial context window of 32768 tokens, enabling it to handle longer inputs and generate coherent, extended outputs.
  • Training Optimization: Benefits from training with Unsloth, which facilitated a 2x faster fine-tuning process.

Use Cases

This model is particularly well-suited for applications where computational resources are a consideration, but a large context window is still required. Its efficient training and compact size make it a strong candidate for:

  • Edge device deployment: Due to its smaller parameter count.
  • Rapid prototyping and experimentation: Leveraging the faster training methodology.
  • Tasks requiring long-context understanding: Such as summarization of lengthy documents or complex conversational agents.