CatalogWantedMirrorsLedgerOperating StatementDocsStudio statusCard system
alt.freeweights.abliterated / gemma-4-31B-it-SDFT-Heretic-RP
From: Ilya626 <huggingface.co/Ilya626>Newsgroups: alt.freeweights.abliteratedSubject:

gemma-4-31B-it-SDFT-Heretic-RP

Date: 2026-11-15Commit: 6d1a9f8Signed: 2DC28CC7B26A283B ✓

Model with 30.7B parameters. Gguf and safetensors weights, 62.1 GB in all. Published by Ilya626 under Apache-2.0. Changed from the base model.

30.7B params62.1GBgemma-4-31B-it-SDFT-Heretic-RPgguf safetensorspermissive
FactsParams30.7BSize62.1 GBFormatsgguf, safetensorsTasknot taggedLicenseApache-2.0From sourcepermissive
$ fw get Ilya626/gemma-4-31B-it-SDFT-Heretic-RP
# manifest sha256 95d87140150b… · 62.1 GB · 28 files
$ fw verify ./gemma-4-31B-it-SDFT-Heretic-RP Ilya626/gemma-4-31B-it-SDFT-Heretic-RP
From sourceHosted (signed)Ilya626/gemma-4-31B-it-SDFT-Heretic-RPLast check: 2026-11-15 05:16 UTC
Run at homeGet help

Get help: custom fine-tune or private deploy. We review your setup before we quote.

Quantized of gemma-4-31B-it.

Works with

Run18Fine-tune7Convert9Scan2
OverviewModel cardFilesUse itLicenseProvenanceModel tree

Overview

Size
62.1 GB
Weights
214 MB (GGUF)
Parameters
30.7B
Formats
gguf, safetensors
Task
not tagged
License
Apache-2.0 (permissive)
Files
28, each with a sha256
Publisher
Ilya626
Added
2026-11-15
Section
alt.freeweights.abliterated
Lineage

Changed from the base model. The uploader's repo is named abliterated: refusal behavior was removed from google/gemma-4-31B-it by the uploader. Read the model card at the source for the method and the layers touched.

Base model: google/gemma-4-31B-it

Where these bytes come from
Publisher at the pinned commit→You

Sources are tried in order. A file counts only when its size and sha256 match the manifest. Sources are checked when you ask for them, not on a timer.

Sources
BadgeHostLast checkResult
From sourcehuggingface.co2026-11-15 05:16 UTCpassed

Model card

From the publisher at commit 6d1a9f8. Shown as plain text with links; images open at the publisher.

Gemma 4 31B it SDFT Heretic RP#

This repository contains merged Hugging Face weights for an RP-enhanced version of Gemma 4 31B it.

The model is based on the uncensored Heretic SDFT line, then further tuned for roleplay behavior: stronger scene continuation, better character and world continuity, more active narrative pressure, and reduced generic LLM phrasing.

Important Disclaimer#

This model is provided for research and evaluation purposes only.

The underlying Heretic SDFT base intentionally has weakened or removed guardrails compared with standard aligned chat models. Users are responsible for how they run, prompt, deploy, and distribute outputs from this model.

Do not use this model to generate harmful, illegal, abusive, exploitative, or otherwise unsafe content.

What This Is#

  • Merged weights, not a LoRA adapter.
  • Built from google/gemma-4-31B-it.
  • Further RP-tuned from the Heretic SDFT base.
  • Intended for local inference, RP evaluation, narrative testing, and research.
  • Quantized GGUF files may be published separately.

RP Improvements#

Compared with the previous Heretic SDFT RP checkpoints, this version showed statistically meaningful improvements in roleplay-oriented evaluation:

  • Better continuation of ongoing scenes.
  • Stronger use of established character traits and recent actions.
  • More active world and NPC momentum.
  • Better handling of multi-turn RP context.
  • Less generic filler and slightly reduced slop phrasing.
  • More consistent atmospheric detail without as much mechanical repetition.

This model is intended for:

  • Dark fantasy, cyberpunk, and grounded RPG narration.
  • Game Master style scene continuation.
  • Character dialogue and NPC interaction.
  • Multi-turn RP experiments.
  • Research into SDFT, self-training, and RP behavior shaping.

Notes#

This is an experimental release. Outputs may still be unstable, repetitive, unsafe, or stylistically uneven depending on sampling settings and prompt structure.

Recommended starting inference settings:

temperature: 0.8-0.95
top_p: 0.95
top_k: 64

Acknowledgements#

Special thanks to the people who helped make this release possible:

  • kurgan1138 — lead tester, responsible for extensive RP evaluation and feedback.
  • dysfunctionalidi0t — tester, contributed additional evaluation and prompt testing.
  • Greg — sponsor, generously provided the training server resources that made this model possible.

Files

28 files · 62.1 GBWeights 214 MB (GGUF). Sizes and sha256 from the signed manifest.
Show all 28 files with their sha256
FileSizesha256
.gitattributes1.66 KB6228a56a3a4dc751bf271e17e3727d5439b876961d0375ac5f81a1143e26bc65
README.md2.61 KBf7a2b5e902eece7b7788279ed5a4fbe385d34b98779ba82b2aeccf77ea088656
chat_template.jinja12.0 KB2dfbfc7d538912f4ea11d29d85b4e25d7bc26386e53f57529f1d707c28b5828c
config.json3.56 KBe225a34c0480b5a785643bd48d344ae20f909b6c93840c9ed0318d3a1a1e54ed
generation_config.json208 Bd4226bbe3117d2d253ba4609720ba82c6c4ce4627a9a6ae05387c78983ac03de
lora/adapter_config.json16.2 KBef4e89f192c6d78ebf310275d36b9372d81be3dafa24f6d2edb47d1854ff9cff
lora/adapter_merge_metadata.json1.08 KBe326597d6aeb8d7675610989792001f258041d90302c6ce60961e79e36ab6df9
lora/adapter_model.safetensors429 MBed6191638879131f7b72dc87461a150875b6a1281afbea4c991db7418fc78138
lora/chat_template.jinja12.0 KB2dfbfc7d538912f4ea11d29d85b4e25d7bc26386e53f57529f1d707c28b5828c
lora/gemma4-31b-rp-guardrails-polish-bds-cat-r14.gguf214 MB7b17bb0b51c1d1e9ceb630853378889f369c7dcf4329286e1e6f2fb88def79aa
lora/tokenizer.json32.2 MBcc8d3a0ce36466ccc1278bf987df5f71db1719b9ca6b4118264f45cb627bfe0f
lora/tokenizer_config.json2.69 KBb9686f02a0bb2d5de25e5bed7949d5858109e62cbb1111ce4cea731885e779e9
model-00001-of-00013.safetensors4.97 GBb743e86d74ab2333d2ab003c951c0b9f5f51191af6f3a7ac2bad86d0d0051d31
model-00002-of-00013.safetensors4.90 GB37838e55d3ab503ac9e8880cc88dcff0e85eea8be529c6ce4b8b77c0325a7926
model-00003-of-00013.safetensors4.90 GBcd68060a01077026dcfa256c212de9cacd2bf0e4ce0f09485768407d0774fc14
model-00004-of-00013.safetensors4.79 GB777ce14ca3d9af041d0c1b451ad71f87832e35c580b1498978a6b9d7f29c2fc7
model-00005-of-00013.safetensors4.90 GB5902f6cffdc13d3f8a19d529a1829553a48244099593016c7e8a0c98f0005024
model-00006-of-00013.safetensors4.90 GB52dd0d7d2a2dedf3207a0eba70f96c8a352e7a9fbb08aa4378e5911702e4e0f5
model-00007-of-00013.safetensors4.90 GBde7054ec582847bef8a9e9d114692c8310703446c7d8dcf351d63411702f581f
model-00008-of-00013.safetensors4.90 GB06f03222dec3284a972b0029df15742679434d48dc78854c38ef92ddb5c2ee7e
model-00009-of-00013.safetensors4.90 GB17f99557ba447fc384d58feb5e2cbdff0bb4e79c84614efed5e92fb878d21bac
model-00010-of-00013.safetensors4.79 GB85300dd0ab7bc41b797df534af9b68ad5ce97fe0b761b07b5d6f95db78e4c261
model-00011-of-00013.safetensors4.90 GB9120b16decec0cbfed6922c438465d9c8938416bee87e4429c36b5f08d09d4c0
model-00012-of-00013.safetensors4.90 GB3886de93fa990ba0db6a69cb13402c8c3297e171ce2ead91bc1467f01c9f1128
model-00013-of-00013.safetensors2.75 GB3683d011bab44510a7d43ad9ad295284e406ab5bdf5503ed8bb8f7bb5c319b8f
model.safetensors.index.json81.7 KB34ab68245faec3961a9bdc97da186289b8e55530735874be8f8688a295732e21
tokenizer.json32.2 MBcc8d3a0ce36466ccc1278bf987df5f71db1719b9ca6b4118264f45cb627bfe0f
tokenizer_config.json2.11 KB8cadfd27f378c9648d98cf47f119b231a4c2b40a413663c2759a3c55da896183
VersionsEach time the pinned upstream commit or file list changed in this repo. Newest first.
Upstream commitDateContents
6d1a9f8139 pinned now2026-11-15 05:16 UTC28 files · 62.1 GB

Use it

$ fw get Ilya626/gemma-4-31B-it-SDFT-Heretic-RP
# manifest sha256 95d87140150b… · 62.1 GB · 28 files
$ fw verify ./gemma-4-31B-it-SDFT-Heretic-RP Ilya626/gemma-4-31B-it-SDFT-Heretic-RP
fw.pyChecks the signature and every sha256, tries each source in order, resumes.
$ python3 fw.py get https://freeweights.allternit.com/manifests/Ilya626/gemma-4-31B-it-SDFT-Heretic-RP.json --dir ./gemma-4-31B-it-SDFT-Heretic-RP $ python3 fw.py verify ./gemma-4-31B-it-SDFT-Heretic-RP ./Ilya626__gemma-4-31B-it-SDFT-Heretic-RP.json
hfThe Hugging Face CLI with one setting. Same repo id.
$ HF_ENDPOINT=https://freeweights.allternit.com hf download Ilya626/gemma-4-31B-it-SDFT-Heretic-RP --revision 6d1a9f813930d4acc300e2656a59f75f83b61f9e --local-dir ./gemma-4-31B-it-SDFT-Heretic-RP
llama.cppDownload one GGUF variant (lora, 214 MB), then serve it.
$ HF_ENDPOINT=https://freeweights.allternit.com hf download Ilya626/gemma-4-31B-it-SDFT-Heretic-RP --include "lora/*" --local-dir ./gemma-4-31B-it-SDFT-Heretic-RP $ llama-server -m ./gemma-4-31B-it-SDFT-Heretic-RP/lora/gemma4-31b-rp-guardrails-polish-bds-cat-r14.gguf
OllamaMake an Ollama model from the downloaded file.
$ printf 'FROM ./gemma-4-31B-it-SDFT-Heretic-RP/lora/gemma4-31b-rp-guardrails-polish-bds-cat-r14.gguf\n' > Modelfile $ ollama create gemma-4-31b-it-sdft-heretic-rp -f Modelfile
Run at home

License

Name
Apache License 2.0
Id
Apache-2.0
Class
Permissive
Redistribution
yes
Listing
Hosted (signed)
Decided
License as declared by the publisher. No person read it. 2026-11-15
Text in
README.md
Text sha256
f7a2b5e902eece7b7788279ed5a4fbe385d34b98779ba82b2aeccf77ea088656
Terms gate
No extra terms beyond the license.

Read the license text

Licenses are checked against the license table. A custom license passes only after a person reads it.

Provenance

From: Ilya626 <huggingface.co/Ilya626>Newsgroups: alt.freeweights.abliteratedSubject: gemma-4-31B-it-SDFT-Heretic-RPDate: 2026-11-15Commit: 6d1a9f8Signed: 2DC28CC7B26A283B ✓
Upstream
Ilya626/gemma-4-31B-it-SDFT-Heretic-RP
Pinned commit
6d1a9f813930d4acc300e2656a59f75f83b61f9e
Manifest
gemma-4-31B-it-SDFT-Heretic-RP.json · .minisig
Manifest sha256
95d87140150bb7a769b3119a977423e532cee696f73298104238fdac4bbbb0d4
Signature
minisign, Ed25519, key 2DC28CC7B26A283B
Created
2026-11-15 05:16 UTC
Generator
freeweights-manifest/0.1.0
Takedowns
None.

How to verify

$ curl -O https://freeweights.allternit.com/manifests/Ilya626/gemma-4-31B-it-SDFT-Heretic-RP.json $ curl -O https://freeweights.allternit.com/manifests/Ilya626/gemma-4-31B-it-SDFT-Heretic-RP.json.minisig $ minisign -Vm gemma-4-31B-it-SDFT-Heretic-RP.json -P RWQ7KGqyx4zCLXtUFOqvfv8L3g+8fpKJ1FsuxPyOYMF+4KufbUxKq1Oq $ shasum -a 256 <file> # compare with the Files tab

A manifest whose signature fails must be rejected. A file whose size or sha256 does not match must be discarded. Verify a download

Model tree

Parent

google/gemma-4-31B-it · this model is a quantization of it

1 model in the catalog name this one as their parent.

Quantizations
1
Quantizations 1
Get it
Download
$ HF_ENDPOINT=https://freeweights.allternit.com hf download Ilya626/gemma-4-31B-it-SDFT-Heretic-RP
Open in Freeweights Studio

Pull, verify and run it in the Allternit app on Desktop, web or phone.

Signed manifest
minisign, key 2DC28CC7B26A283B. gemma-4-31B-it-SDFT-Heretic-RP.json. Provenance. Takedown history: none.
Source
huggingface.co/Ilya626/gemma-4-31B-it-SDFT-Heretic-RP

Variants

1 quantized or adapter build of gemma-4-31B-it-SDFT-Heretic-RP. Each has its own license check and page.

SubjectSizeParamsLicenseDate
Ilya626_gemma-4-31B-it-SDFT-Heretic-RP-GGUFHosted (signed) From: bartowski in alt.freeweights.abliterated525 GB—Permissive2026-11-15

Related models

All of alt.freeweights.abliterated →
From: xtools-at <huggingface.co/xtools-at>Subject: DeepSeek-V4-Pro-Qwen3.5-9B-UNCENSORED-heretic
From: mlx-community <huggingface.co/mlx-community>Subject: DeepSeek-V4-Pro-Qwen3.5-9B-UNCENSORED-heretic-OptiQ-4bit
From: mlx-community <huggingface.co/mlx-community>Subject: gemma-4-26B-A4B-it-heretic-msq-2.6bit
From: coder3101 <huggingface.co/coder3101>Subject: gemma-4-31B-it-heretic
From: llmfan46 <huggingface.co/llmfan46>Subject: gemma-4-31B-it-uncensored-heretic
From: mlx-community <huggingface.co/mlx-community>Subject: gemma-4-31B-it-uncensored-heretic-4bit