Instructions to use saidutta69/gemma-3-1b-it-heretic with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use saidutta69/gemma-3-1b-it-heretic with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf saidutta69/gemma-3-1b-it-heretic:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf saidutta69/gemma-3-1b-it-heretic:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf saidutta69/gemma-3-1b-it-heretic:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf saidutta69/gemma-3-1b-it-heretic:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf saidutta69/gemma-3-1b-it-heretic:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf saidutta69/gemma-3-1b-it-heretic:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf saidutta69/gemma-3-1b-it-heretic:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf saidutta69/gemma-3-1b-it-heretic:Q4_K_M

Use Docker

docker model run hf.co/saidutta69/gemma-3-1b-it-heretic:Q4_K_M

LM Studio
Jan

vLLM

How to use saidutta69/gemma-3-1b-it-heretic with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "saidutta69/gemma-3-1b-it-heretic"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "saidutta69/gemma-3-1b-it-heretic",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/saidutta69/gemma-3-1b-it-heretic:Q4_K_M

Ollama
How to use saidutta69/gemma-3-1b-it-heretic with Ollama:
```
ollama run hf.co/saidutta69/gemma-3-1b-it-heretic:Q4_K_M
```
Unsloth Desktop
Docker Model Runner
How to use saidutta69/gemma-3-1b-it-heretic with Docker Model Runner:
```
docker model run hf.co/saidutta69/gemma-3-1b-it-heretic:Q4_K_M
```

Lemonade

How to use saidutta69/gemma-3-1b-it-heretic with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull saidutta69/gemma-3-1b-it-heretic:Q4_K_M

Run and chat with the model

lemonade run user.gemma-3-1b-it-heretic-Q4_K_M

List all available models

lemonade list

Atomic Chat

gemma-3-1b-it-heretic

A decensored variant of google/gemma-3-1b-it, produced with Heretic v1.4.0 (directional ablation / "abliteration"). Refusal behavior is suppressed via targeted weight edits to the attention output and MLP down-projections rather than fine-tuning, so the base model's knowledge and instruction-following are left largely intact.

Who this is for: developers who want Google's smallest Gemma 3 instruct model without refusals - runs comfortably on CPU, Raspberry Pi class hardware, and tiny edge devices. Not a capability upgrade over base gemma-3-1b-it - same model, refusal guardrails removed.

Why abliteration instead of fine-tuning

Fine-tuning a "helpful" persona on top of RLHF'd refusals fights the base model's training and tends to degrade coherence. Abliteration instead finds and edits the specific weight directions responsible for refusal, leaving the rest of the network (and its capabilities) untouched. See the Heretic repo and the original abliteration writeup for the mechanism.

Files

Safetensors (BF16)

The full-precision weights are in model.safetensors (see the repo file listing for exact sizes).

GGUF quantizations

GGUF quantizations are published for this model (Q4_K_M, Q5_K_M, Q6_K, Q8_0). Pull a specific quant with llama.cpp / ollama.

File	Format	Size
`gemma-3-1b-it-heretic-Q4_K_M.gguf`	GGUF Q4_K_M	(see repo files)
`gemma-3-1b-it-heretic-Q5_K_M.gguf`	GGUF Q5_K_M	(see repo files)
`gemma-3-1b-it-heretic-Q6_K.gguf`	GGUF Q6_K	(see repo files)
`gemma-3-1b-it-heretic-Q8_0.gguf`	GGUF Q8_0	(see repo files)

Quickstart

# llama.cpp - defaults to the Q4_K_M quant
llama serve -hf saidutta69/gemma-3-1b-it-heretic:Q4_K_M

# transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "saidutta69/gemma-3-1b-it-heretic"
model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype="auto", device_map="auto")
tokenizer = AutoTokenizer.from_pretrained(model_name)
# ... inference code

Also runnable via Ollama, LM Studio, Jan, vLLM, SGLang.

Responsible use

Refusal suppression is deliberate and works as intended: this model will comply with requests the base model would refuse, including some it shouldn't. There is no safety filtering layered on top. You are responsible for how you deploy it.