alexnt2/Gemma4-Gutenberg-Literotica-12B

TEXT GENERATIONPricing:Input $1.2 / Cached $0.24 / Output $4.8Concurrent Unit Cost:1Model Size:12BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Jul 23, 2026License:gemmaArchitecture:Transformer0.0K Featherless Exclusive Cold

alexnt2/Gemma4-Gutenberg-Literotica-12B is a 12 billion parameter Gemma-4 model, fine-tuned using Supervised Fine-Tuning (SFT) with Unsloth (LoRA r=16) and merged into the base weights in BF16. This model is specifically prepared for inference and GGUF export, offering a specialized variant of the Gemma-4 architecture. Its training process, though interrupted, indicates a focus on adapting the base model for specific textual generation tasks.

Loading preview...

Model Overview

alexnt2/Gemma4-Gutenberg-Literotica-12B is a 12 billion parameter language model built upon the gemma4_unified architecture. It has undergone Supervised Fine-Tuning (SFT) using the Unsloth library with a LoRA adapter (r=16). The LoRA adapter was subsequently merged back into the base model weights in BF16 format, making it ready for direct inference or conversion to GGUF format.

Key Characteristics

  • Base Architecture: Utilizes the Gemma-4 (gemma4_unified) foundation.
  • Parameter Count: Features 12 billion parameters, offering a balance between capability and computational requirements.
  • Fine-Tuning Method: Employed SFT with Unsloth and LoRA (r=16) for efficient adaptation.
  • Precision: The final merged model weights are in BF16 (bfloat16) format.
  • Readiness: Designed for immediate deployment in inference tasks or for export to GGUF for broader compatibility.
  • Training Status: The model was merged from a LoRA adapter at checkpoint-1950, indicating that training was interrupted before full completion (1950 out of 2429 steps).

Intended Use

This model is suitable for developers looking for a specialized Gemma-4 variant that has been fine-tuned for specific text generation tasks. Its BF16 format and readiness for GGUF export make it versatile for various deployment environments.