Gemma 4 31B lineage
Tags identify base_model google/gemma-4-31B-it with finetune lineage.
Open Source Model Profile · DavidAU
gemma-4-31B-it-The-DECKARD-HERETIC-UNCENSORED-Thinking
gemma-4-31B-it-The-DECKARD-HERETIC-UNCENSORED-Thinking is a 31.27B-parameter image-text-to-text fine-tune from DavidAU. According to the model card, it was de-censored then fine-tuned with Unsloth on the DECKARD collection.
- Publisher
- DavidAU
- Task
- image-text-to-text
- Model type
- gemma4
- License
- apache-2.0
- Library
- Unknown
- Publication status
- Accepted · not indexed
Model overview
gemma-4-31B-it-The-DECKARD-HERETIC-UNCENSORED-Thinking is published by DavidAU as an image-text-to-text model. The captured configuration identifies Gemma4ForConditionalGeneration with gemma4 type, and Safetensors metadata reports 31,273,086,512 parameters. Tags identify base_model google/gemma-4-31B-it with finetune lineage.
Recorded capabilities
Deckard Unsloth note
According to the model card, the base was Heretic de-censored then fine-tuned via Unsloth on THE DECKARD 5 dataset collection.
Instruct and thinking templates
According to the model card, Jinja templates are provided for instruct mode and hard-coded-on thinking mode.
31.27B multimodal config
Safetensors metadata reports 31,273,086,512 parameters with Gemma4ForConditionalGeneration architecture and Apache-2.0 licensing.
Use cases in the source record
- Interleaved image-and-text generation following the publisher-described multimodal input behavior.
- Creative writing, roleplay, and fiction experiments aligned with the publisher-listed storytelling tags.
- Instruct- or thinking-mode trials using the publisher-provided Jinja templates.
Limitations and unknowns
- Publisher benchmark and smarter-model claims come from the model card and were not independently verified.
- No Ethen-measured evaluation results were extracted from this record.
- No context-window value was extracted from this record.
- Provider state is historical snapshot data, not independently refreshed current availability.
Source and provenance
Source: DavidAU/gemma-4-31B-it-The-DECKARD-HERETIC-UNCENSORED-Thinking
Captured: Unknown. Processed: 2026-09-07T19:35:38.929118+00:00.
Smarter [beats the base model on 6 of 7 benchmarks ] and uncensored too. gemma-4-31B-it-The-DECKARD-HERETIC-UNCENSORED-Thinking Gemma 4 31B first "Heretic'ed" (de-censored), then fine tuned (Via Unsloth) on "THE DECKARD" 5 dataset collection to improve performance from top to bottom - character, intelligence, depth, observation, and ah... point of view. This model is fully uncensored [ no nanny, no oversight ], and retains all its original power. I have provided Jinja templates for both "instruct mode" and "thinking mode" (hard coded on) so you can choose which one you want to use. The Deckard dataset collection was manually assembl…
F001F002F003F004F005F006F007F008F009F010F011F013F014F016F018