Instructions to use Minibase/Detoxify-Language-Small with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Minibase/Detoxify-Language-Small with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="Minibase/Detoxify-Language-Small",
	filename="detoxify-small-q8_0.gguf",
)

output = llm(
	"Once upon a time,",
	max_tokens=512,
	echo=True
)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use Minibase/Detoxify-Language-Small with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Minibase/Detoxify-Language-Small:Q8_0
# Run inference directly in the terminal:
llama-cli -hf Minibase/Detoxify-Language-Small:Q8_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Minibase/Detoxify-Language-Small:Q8_0
# Run inference directly in the terminal:
llama-cli -hf Minibase/Detoxify-Language-Small:Q8_0

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 Minibase/Detoxify-Language-Small:Q8_0
# Run inference directly in the terminal:
./llama-cli -hf Minibase/Detoxify-Language-Small:Q8_0

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 Minibase/Detoxify-Language-Small:Q8_0
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Minibase/Detoxify-Language-Small:Q8_0

Use Docker

docker model run hf.co/Minibase/Detoxify-Language-Small:Q8_0

LM Studio
Jan
Ollama
How to use Minibase/Detoxify-Language-Small with Ollama:
```
ollama run hf.co/Minibase/Detoxify-Language-Small:Q8_0
```

Unsloth Studio

How to use Minibase/Detoxify-Language-Small with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for Minibase/Detoxify-Language-Small to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for Minibase/Detoxify-Language-Small to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Minibase/Detoxify-Language-Small to start chatting

Docker Model Runner
How to use Minibase/Detoxify-Language-Small with Docker Model Runner:
```
docker model run hf.co/Minibase/Detoxify-Language-Small:Q8_0
```

Lemonade

How to use Minibase/Detoxify-Language-Small with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Minibase/Detoxify-Language-Small:Q8_0

Run and chat with the model

lemonade run user.Detoxify-Language-Small-Q8_0

List all available models

lemonade list

Minibase commited on Sep 18, 2025

Commit

403101e

verified ·

1 Parent(s): db55146

Update README.md

Browse files

Files changed (1) hide show

README.md +4 -7

README.md CHANGED Viewed

@@ -128,7 +128,10 @@ print(clean_text)  # "This product is really amazing, no kidding!"
 | Metric | Score | Description |
 |--------|-------|-------------|
-| **Toxicity Reduction** | 0.032 (3.2%) | Average reduction in toxicity scores |
 | **Semantic to Expected** | 0.471 (47.1%) | Similarity to human expert rewrites |
 | **Semantic to Original** | 0.625 (62.5%) | How much original meaning is preserved |
 | **Fluency** | 0.919 (91.9%) | Quality of generated text structure |
@@ -138,15 +141,9 @@ print(clean_text)  # "This product is really amazing, no kidding!"
 ### Dataset Breakdown
 #### General Toxic Content (1,000 samples)
-- **Toxicity Reduction**: 3.1%
 - **Semantic Preservation**: 62.7%
 - **Fluency**: 91.9%
-#### High-Toxicity Content (8 samples)
-- **Toxicity Reduction**: 25.0% ⭐ *Strong performance*
-- **Semantic Preservation**: 36.6%
-- **Fluency**: 96.3%
 ### Comparison with Baselines
 | Model | Semantic Similarity | Toxicity Reduction | Fluency |

 | Metric | Score | Description |
 |--------|-------|-------------|
+• Original Toxicity:            0.051 (5.1%)
+• Final Toxicity:               0.020 (2.0%)
+| **Toxicity Reduction** | 0.051 (ParaDetox) --> 0.020 | Reduced toxicity scores by more than 50% |
 | **Semantic to Expected** | 0.471 (47.1%) | Similarity to human expert rewrites |
 | **Semantic to Original** | 0.625 (62.5%) | How much original meaning is preserved |
 | **Fluency** | 0.919 (91.9%) | Quality of generated text structure |
 ### Dataset Breakdown
 #### General Toxic Content (1,000 samples)
 - **Semantic Preservation**: 62.7%
 - **Fluency**: 91.9%
 ### Comparison with Baselines
 | Model | Semantic Similarity | Toxicity Reduction | Fluency |