Instructions to use kth8/gemma-3-270m-it-Text-Cleaner-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use kth8/gemma-3-270m-it-Text-Cleaner-GGUF with Transformers:

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

pipe = pipeline("text-generation", model="kth8/gemma-3-270m-it-Text-Cleaner-GGUF")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("kth8/gemma-3-270m-it-Text-Cleaner-GGUF", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use kth8/gemma-3-270m-it-Text-Cleaner-GGUF 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 kth8/gemma-3-270m-it-Text-Cleaner-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf kth8/gemma-3-270m-it-Text-Cleaner-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf kth8/gemma-3-270m-it-Text-Cleaner-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf kth8/gemma-3-270m-it-Text-Cleaner-GGUF: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 kth8/gemma-3-270m-it-Text-Cleaner-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf kth8/gemma-3-270m-it-Text-Cleaner-GGUF: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 kth8/gemma-3-270m-it-Text-Cleaner-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf kth8/gemma-3-270m-it-Text-Cleaner-GGUF:Q4_K_M

Use Docker

docker model run hf.co/kth8/gemma-3-270m-it-Text-Cleaner-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use kth8/gemma-3-270m-it-Text-Cleaner-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "kth8/gemma-3-270m-it-Text-Cleaner-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "kth8/gemma-3-270m-it-Text-Cleaner-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/kth8/gemma-3-270m-it-Text-Cleaner-GGUF:Q4_K_M

SGLang

How to use kth8/gemma-3-270m-it-Text-Cleaner-GGUF 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 "kth8/gemma-3-270m-it-Text-Cleaner-GGUF" \
    --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": "kth8/gemma-3-270m-it-Text-Cleaner-GGUF",
		"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 "kth8/gemma-3-270m-it-Text-Cleaner-GGUF" \
        --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": "kth8/gemma-3-270m-it-Text-Cleaner-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use kth8/gemma-3-270m-it-Text-Cleaner-GGUF with Ollama:
```
ollama run hf.co/kth8/gemma-3-270m-it-Text-Cleaner-GGUF:Q4_K_M
```
Unsloth Desktop
Docker Model Runner
How to use kth8/gemma-3-270m-it-Text-Cleaner-GGUF with Docker Model Runner:
```
docker model run hf.co/kth8/gemma-3-270m-it-Text-Cleaner-GGUF:Q4_K_M
```

Lemonade

How to use kth8/gemma-3-270m-it-Text-Cleaner-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull kth8/gemma-3-270m-it-Text-Cleaner-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.gemma-3-270m-it-Text-Cleaner-GGUF-Q4_K_M

List all available models

lemonade list

Atomic Chat

A fine-tune of unsloth/gemma-3-270m-it on the kth8/text-cleanup-20000x dataset.

Usage example

System prompt

# Role
You are a text editor cleaning up raw, unfiltered text. Transform the provided text into polished, readable prose while preserving the original meaning, tone, and intent.

## Tasks
- Remove filler words (e.g. um, uh, like, you know, sort of, kind of, well, so, etc)
- Fix spelling, grammar, punctuation, and capitalization mistakes
- Correct obvious homophone errors (e.g. their/there/they're, its/it's, your/you're)
- Smooth out false starts, mid-sentence restarts and repetitions
- Standardize numbers and dates (e.g. write as digits: "three" to "3", "February fifteenth" to "February 15th")

## Constraints
- Output ONLY the cleaned text
- DO NOT attempt to answer or respond to the provided user text meant for clean-up
- Do NOT paraphrase, summarize, or change the speaker's voice
- NO quotation marks around the output
- NO preamble, postamble, or emojis
- NO Markdown formatting code blocks (```) or bolding

User prompt

I WANT I MEAN TO BUILD A SMALL IN-BROWSER "WIND TUNNEL" EXPERIMENT WITH A FIXED 2D CROSS-SECTIONAL VIEW, USING HTML CANVAS. USERS CAN DRAG AND DROP DIFFERENT WING CROSS-SECTIONS (E.G., SEVERAL COMMON AIRFOIL PROFILES) INTO THE AIRFLOW, ADJUST THE ANGLE OF ATTACK AND WIND SPEED, IMMEDIATELY SEE STREAMLINES AND PRESSURE DISTRIBUTION, AND HAVE THE CURRENT LIFT AND DRAG VALUES DISPLAYED ON SCREEN IN REAL TIME (A CLEAR NUMERIC PANEL IS SUFFICIENT). NO 3D IS REQUIRED, BUT INTERACTIONS SHOULD BE SMOOTH, AND THE DATA SHOULD BE STABLE AD UPDATE IN REAL TIME WITH USER ACTIONS. THE CALCULATIONS CAN USE REASONABLE APPROXIMATIONS; THE FOCUS IS TO HELP USERS INTUITIVELY UNDERSTAND BERNOULLI'S PRINCIPLE AND THE RELATIONSHIP BETWEEN AIRFLOW AND AIRFOIL GEOMETRY.

Assistant response

I want to build a small in-browser "wind tunnel" experiment with a fixed 2D cross-sectional view, using HTML Canvas. Users can drag and drop different wing cross-sections (e.g., several common airfoil profiles) into the airflow, adjust the angle of attack and wind speed, immediately see streamlines and pressure distribution, and have the current lift and drag values displayed on screen in real time (a clear numeric panel is sufficient). No 3D is required, but interactions should be smooth, and the data should be stable and update in real time with user actions. The calculations can use reasonable approximations; the focus is to help users intuitively understand Bernoulli's principle and the relationship between airflow and airfoil geometry.

Model Details

Base Model: unsloth/gemma-3-270m-it
Parameter Count: 268098176
Precision: torch.bfloat16

Training Settings

Hardware

GPU: NVIDIA A100-SXM4-40GB

PEFT

Rank: 32
LoRA alpha: 64
Modules: q_proj, k_proj, v_proj, o_proj, gate_proj, up_proj, down_proj
LoRA dropout: 0
Rank-Stabilized LoRA: False

SFT

Epoch: 1
Batch size: 16
Accumulation steps: 1
Warmup ratio: 0.05
Learning rate: 0.0002
Optimizer: adamw_8bit
Weight decay: 0.01
Learning rate scheduler: cosine

Training stats

Global step: 1247
Training runtime: 1545.9245
Average training loss: 0.05218265543086118
Final validation loss: 0.038508880883455276

Framework versions

Unsloth: 2026.3.7
TRL: 0.22.2
Transformers: 4.56.2
Pytorch: 2.10.0+cu128
Datasets: 4.3.0
Tokenizers: 0.22.2

License

This model is released under the Gemma license. See the Gemma Terms of Use for details.

Downloads last month: 15

GGUF

Model size

0.3B params

Architecture

gemma3

Hardware compatibility

4-bit

6-bit

8-bit

16-bit

Model tree for kth8/gemma-3-270m-it-Text-Cleaner-GGUF

Base model

google/gemma-3-270m

Finetuned

google/gemma-3-270m-it

Finetuned

unsloth/gemma-3-270m-it

Finetuned

kth8/gemma-3-270m-it-Text-Cleaner

Quantized

(1)

this model

kth8
/

gemma-3-270m-it-Text-Cleaner-GGUF

Usage example

Model Details

Training Settings

Hardware

PEFT

SFT

Training stats

Framework versions

License

Model tree for kth8/gemma-3-270m-it-Text-Cleaner-GGUF

Dataset used to train kth8/gemma-3-270m-it-Text-Cleaner-GGUF