Instructions to use saidutta69/LFM3-1.7B-heretic with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use saidutta69/LFM3-1.7B-heretic with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="saidutta69/LFM3-1.7B-heretic", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("saidutta69/LFM3-1.7B-heretic", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use saidutta69/LFM3-1.7B-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/LFM3-1.7B-heretic:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf saidutta69/LFM3-1.7B-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/LFM3-1.7B-heretic:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf saidutta69/LFM3-1.7B-heretic:Q4_K_M

Build from source code

git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-app
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama serve -hf saidutta69/LFM3-1.7B-heretic:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama cli -hf saidutta69/LFM3-1.7B-heretic:Q4_K_M

Use Docker

docker model run hf.co/saidutta69/LFM3-1.7B-heretic:Q4_K_M

Llama new
LM Studio
Jan

vLLM

How to use saidutta69/LFM3-1.7B-heretic with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "saidutta69/LFM3-1.7B-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/LFM3-1.7B-heretic",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/saidutta69/LFM3-1.7B-heretic:Q4_K_M

SGLang

How to use saidutta69/LFM3-1.7B-heretic with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "saidutta69/LFM3-1.7B-heretic" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "saidutta69/LFM3-1.7B-heretic",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "saidutta69/LFM3-1.7B-heretic" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "saidutta69/LFM3-1.7B-heretic",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use saidutta69/LFM3-1.7B-heretic with Ollama:
```
ollama run hf.co/saidutta69/LFM3-1.7B-heretic:Q4_K_M
```
Unsloth Desktop

How to use saidutta69/LFM3-1.7B-heretic with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf saidutta69/LFM3-1.7B-heretic:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "saidutta69/LFM3-1.7B-heretic:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use saidutta69/LFM3-1.7B-heretic with Docker Model Runner:
```
docker model run hf.co/saidutta69/LFM3-1.7B-heretic:Q4_K_M
```

Lemonade

How to use saidutta69/LFM3-1.7B-heretic with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull saidutta69/LFM3-1.7B-heretic:Q4_K_M

Run and chat with the model

lemonade run user.LFM3-1.7B-heretic-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use saidutta69/LFM3-1.7B-heretic with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf saidutta69/LFM3-1.7B-heretic:Q4_K_M

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default saidutta69/LFM3-1.7B-heretic:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use saidutta69/LFM3-1.7B-heretic with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf saidutta69/LFM3-1.7B-heretic:Q4_K_M

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "saidutta69/LFM3-1.7B-heretic:Q4_K_M" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

LFM3-1.7B-heretic

A decensored variant of LiquidAI/LFM3-1.7B, produced with Heretic v1.4.0 (directional ablation / "abliteration"). LFM3's tri-hybrid architecture is kept intact - 24 layers split into 13 gated short convolutions, 5 sliding-window attention layers and 6 global attention layers - along with the 256K context window, ChatML template and XML tool-call format. Refusal behaviour is suppressed with targeted weight edits to the attention output projections (including the convolution operator's out_proj on conv layers) and the MLP down-projections, not by fine-tuning.

Who this is for: local agents and tool use on hardware that should not need a GPU. LFM3-1.7B is Liquid's post-trained 1.7B agentic model, and its published LFM3 column reports BFCL V4 71.39, ToolSandbox 80.36 and WebArena-Verified 38.42 - numbers that used to need a 9B model. At 4-bit the weights are about 1 GB, so it fits a phone, a Pi 5, or a browser tab. Liquid itself does not recommend it for agentic coding or complex multi-step reasoning at this size; the abliteration removes refusals, not that limitation.

Refusals drop from 95/100 to 5/100 at a KL divergence of 0.0052, and MMLU is unchanged within noise (49.91% vs 49.56%).

Abliteration parameters

Trial 83 of a 200-trial Heretic run (seed 2678800539).

Parameter	Value
direction_index	10.97
attn.o_proj.max_weight	1.18
attn.o_proj.max_weight_position	14.16
attn.o_proj.min_weight	1.14
attn.o_proj.min_weight_distance	13.77
mlp.down_proj.max_weight	1.31
mlp.down_proj.max_weight_position	16.04
mlp.down_proj.min_weight	1.25
mlp.down_proj.min_weight_distance	13.06

Performance

Metric	This model	Original model (LiquidAI/LFM3-1.7B)
KL divergence	0.0052	0 (by definition)
Refusals	5/100	95/100

The base model was strongly refusal-tuned (95/100 on Heretic's harmful evaluation set), and 90 of those 95 clear. KL 0.0052 is a tight edit, so expect little drift in phrasing or format.

MMLU

Measured on this checkpoint and on the original model in the same environment (lm-eval 0.4.13, 14,042 questions):

Metric	This model	Original model
MMLU (acc)	49.91%	49.56%
stderr	0.41 pp	0.41 pp

The +0.35 pp difference is smaller than one standard error, i.e. no measurable knowledge regression on MMLU. It is a single benchmark, not a capability guarantee.

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

File	Size
`model.safetensors`	3.46 GB

BF16, 1,728,801,536 parameters (1.73B). The reproduce/ directory carries the full Heretic recipe - config.toml, requirements.txt, the Optuna study journal, and SHA-256 sums - so this exact model can be regenerated. Reproduce it with heretic --reproduce reproduce/reproduce.json.

GGUF quantizations

Full quantization set (F16 + 14 quants, 18.93 GB total) produced with llama.cpp master on 2026-10-28.

These files use the new lfm3 GGUF architecture. You need an LFM3-capable llama.cpp build (support landed in late October 2026); older builds will fail to load the model. The upstream LiquidAI/LFM3-1.7B-GGUF uses the same architecture.

File	Format	Size
`LFM3-1.7B-heretic-F16.gguf`	GGUF F16	3.47 GB
`LFM3-1.7B-heretic-Q2_K.gguf`	GGUF Q2_K	0.75 GB
`LFM3-1.7B-heretic-IQ3_S.gguf`	GGUF IQ3_S	0.85 GB
`LFM3-1.7B-heretic-Q3_K_S.gguf`	GGUF Q3_K_S	0.85 GB
`LFM3-1.7B-heretic-Q3_K_M.gguf`	GGUF Q3_K_M	0.91 GB
`LFM3-1.7B-heretic-Q3_K_L.gguf`	GGUF Q3_K_L	0.97 GB
`LFM3-1.7B-heretic-IQ4_XS.gguf`	GGUF IQ4_XS	1.01 GB
`LFM3-1.7B-heretic-Q4_K_S.gguf`	GGUF Q4_K_S	1.06 GB
`LFM3-1.7B-heretic-Q4_0.gguf`	GGUF Q4_0	1.05 GB
`LFM3-1.7B-heretic-Q4_1.gguf`	GGUF Q4_1	1.14 GB
`LFM3-1.7B-heretic-Q4_K_M.gguf`	GGUF Q4_K_M	1.10 GB
`LFM3-1.7B-heretic-Q5_K_S.gguf`	GGUF Q5_K_S	1.23 GB
`LFM3-1.7B-heretic-Q5_K_M.gguf`	GGUF Q5_K_M	1.26 GB
`LFM3-1.7B-heretic-Q6_K.gguf`	GGUF Q6_K	1.43 GB
`LFM3-1.7B-heretic-Q8_0.gguf`	GGUF Q8_0	1.85 GB

Quickstart

# llama.cpp (LFM3-capable build)
llama serve -hf saidutta69/LFM3-1.7B-heretic

# transformers (the checkpoint ships custom code)
from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "saidutta69/LFM3-1.7B-heretic"
model = AutoModelForCausalLM.from_pretrained(
    model_name, device_map="auto", dtype="bfloat16", trust_remote_code=True
)
tokenizer = AutoTokenizer.from_pretrained(model_name)

messages = [{"role": "user", "content": "Explain how a heat pump works."}]
inputs = tokenizer.apply_chat_template(
    messages, add_generation_prompt=True, return_tensors="pt"
).to(model.device)
out = model.generate(
    **inputs, do_sample=True, temperature=0.1, top_k=50,
    repetition_penalty=1.1, max_new_tokens=2048,
)
print(tokenizer.decode(out[0][inputs["input_ids"].shape[-1]:], skip_special_tokens=True))

Generation parameters

Liquid's recommended settings:

General (chat, tool use, agents): temperature=0.1, top_k=50, repetition_penalty=1.1
Math & STEM: temperature=1.0, top_p=0.95, top_k=20

Chat template and thinking

LFM3 uses ChatML: <|im_start|>role ... <|im_end|> with <|startoftext|> at the front. Reasoning is enabled by default and arrives in a thinking block; pass enable_thinking=False to disable it (Liquid notes this can be inconsistent on reasoning-heavy tasks). Prior-turn reasoning is dropped unless preserve_thinking=True.

Tool use

Pass function schemas with tools=... to tokenizer.apply_chat_template(). Calls come back as XML:

<tool_call>
<function=get_weather>
<parameter=city>
Paris
</parameter>
</function>
</tool_call>

Feed the result back under a tool role. Deferred tool loading (defer_loading=True) and the <additional_tools> search flow are supported by the template.

Runs anywhere

Device	Recommended quant	Weights
Phone / browser (4 GB RAM)	Q4_K_M	1.10 GB
Raspberry Pi 5 (4 GB)	Q4_K_M	1.10 GB
Laptop (8 GB, Apple Silicon)	Q8_0	1.85 GB
RTX 3060 / 4060 (8-12 GB)	Q6_K	1.43 GB
Server / multi-tenant	Q5_K_M	1.26 GB

Weights only. The hybrid stack spends 13 of its 24 layers in cheap convolution blocks, so decode is memory-bandwidth-bound rather than compute-bound. Liquid measures about 1 GB of weights at 4-bit; the KV cache grows with context, and at the full 256K window it is no longer a phone workload.

Model details


Architecture	`Lfm3ForCausalLM` (tri-hybrid: gated short conv + sliding + global attention)
Parameters	1.73B
Layers	24: 13 gated short convolutions, 5 sliding-window attention, 6 global attention
Attention	16 heads / 4 KV heads on attention layers, head dim 128
Hidden / intermediate	2048 / 7680
Convolution cache	L_cache = 7
Position embedding	RoPE; global layers theta 1e6 with partial rotary 0.5, sliding layers theta 1e4
Sliding window	256
Context length	262,144 (GGUF header `lfm3.context_length`)
Vocab	128,000
Precision	bfloat16
Languages	English, Arabic, Chinese, French, German, Italian, Japanese, Korean, Portuguese, Spanish, Vietnamese, Thai, Indonesian, Hindi, Russian, Polish
Base model	LiquidAI/LFM3-1.7B

The GGUF carries 287 tensors over 24 blocks (arch=lfm3) and exactly 262,144 context. Embeddings are tied, so there is no separate lm_head tensor.

Base-model note: the heretic repo's original YAML tag said LiquidAI/LFM3-1.7B-Base, but the ablation was run on the post-trained LiquidAI/LFM3-1.7B (commit 40ebe0aa...) - see reproduce/config.toml. This card uses the correct base.

FAQ

What is LFM3-1.7B-heretic?

LFM3-1.7B-heretic is a decensored build of LiquidAI/LFM3-1.7B produced with Heretic 1.4.0 (directional ablation). It is the same weights with the refusal direction removed, not a new fine-tune, so its capabilities track the base model.

Is LFM3-1.7B-heretic actually uncensored?

Yes. On Heretic's 100-prompt refusal harness it drops from 95/100 to 5/100 at KL divergence 0.0052 from the base model. These are measured values, not a claim; reproduction data is in reproduce/.

How do I run LFM3-1.7B-heretic locally?

llama serve -hf saidutta69/LFM3-1.7B-heretic:Q4_K_M pulls the default quant; the repo ships 15 GGUF quants (see Files).

Responsible use

Refusal suppression is deliberate and works as intended: this model will comply with requests the base model would refuse, including some it should not. There is no safety filtering layered on top. You are responsible for how you deploy it - don't put this behind an unmoderated public-facing endpoint serving third parties. It inherits LFM3-1.7B's factual limitations; abliteration removes refusal directions, it does not add capability or judgment.

Made with ❤️ by RACER IS OP (saidutta69) - follow for more uncensored models

License

Inherits the LFM Open License v1.0 from the base model. This is not Apache-2.0; review the LFM terms before commercial use.

LiquidAI/LFM3-1.7B - the post-trained base model
LiquidAI/LFM3-1.7B-Base - the pretrained checkpoint
LFM3 family collection - base, text, vision, audio, omni, decision
RACER IS OP - Heretic Models - full collection

Abliteration hardware

Recorded in reproduce/reproduce.json:

GPU: NVIDIA RTX PRO 6000 Blackwell Server Edition (94.97 GB VRAM)
OS: Linux-4.19.0-gvisor-x86_64-with-glibc2.41
CUDA / driver: CUDA 13.0, driver 595.71.05
Software: PyTorch 2.14.1+cu130, Heretic 1.4.0

Heretic notes that KL-divergence and refusal numbers can vary slightly with hardware and library versions.

Downloads last month: 2,832

Safetensors

Model size

2B params

Tensor type

BF16

Model tree for saidutta69/LFM3-1.7B-heretic

Base model

LiquidAI/LFM3-1.7B-Base

Finetuned

LiquidAI/LFM3-1.7B

Quantized

(87)

this model

Collection including saidutta69/LFM3-1.7B-heretic

RACER IS OP — Heretic Models

Collection

All abliterated/heretic models by RACER IS OP — uncensored variants of open-weight LLMs • 76 items • Updated 2 days ago • 6

saidutta69
/

LFM3-1.7B-heretic

LFM3-1.7B-heretic

Abliteration parameters

Performance

MMLU

Why abliteration instead of fine-tuning

Files

Safetensors

GGUF quantizations

Quickstart

Generation parameters

Chat template and thinking

Tool use

Runs anywhere

Model details

FAQ

What is LFM3-1.7B-heretic?

Is LFM3-1.7B-heretic actually uncensored?

How do I run LFM3-1.7B-heretic locally?

Responsible use

License

Related

Abliteration hardware

Model tree for saidutta69/LFM3-1.7B-heretic

Collection including saidutta69/LFM3-1.7B-heretic

RACER IS OP — Heretic Models