Instructions to use jialinyyzz/humanizer with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use jialinyyzz/humanizer with Transformers:

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

pipe = pipeline("text-generation", model="jialinyyzz/humanizer")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("jialinyyzz/humanizer")
model = AutoModelForMultimodalLM.from_pretrained("jialinyyzz/humanizer", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use jialinyyzz/humanizer 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 jialinyyzz/humanizer:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf jialinyyzz/humanizer:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf jialinyyzz/humanizer:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf jialinyyzz/humanizer: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 jialinyyzz/humanizer:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf jialinyyzz/humanizer: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 jialinyyzz/humanizer:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf jialinyyzz/humanizer:Q4_K_M

Use Docker

docker model run hf.co/jialinyyzz/humanizer:Q4_K_M

LM Studio
Jan

vLLM

How to use jialinyyzz/humanizer with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/jialinyyzz/humanizer:Q4_K_M

SGLang

How to use jialinyyzz/humanizer 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 "jialinyyzz/humanizer" \
    --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": "jialinyyzz/humanizer",
		"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 "jialinyyzz/humanizer" \
        --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": "jialinyyzz/humanizer",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use jialinyyzz/humanizer with Ollama:
```
ollama run hf.co/jialinyyzz/humanizer:Q4_K_M
```
Unsloth Desktop
Docker Model Runner
How to use jialinyyzz/humanizer with Docker Model Runner:
```
docker model run hf.co/jialinyyzz/humanizer:Q4_K_M
```

Lemonade

How to use jialinyyzz/humanizer with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull jialinyyzz/humanizer:Q4_K_M

Run and chat with the model

lemonade run user.humanizer-Q4_K_M

List all available models

lemonade list

Atomic Chat

jialinyyzz commited on 3 days ago

Commit

8a23577

verified ·

1 Parent(s): 15530db

Figures: light paper cards (evaluation card, before/after, training, banner, quantization card); README uses the evaluation card

Browse files

Files changed (17) hide show

.gitattributes +2 -0
README.md +3 -7
assets/app-en.png +2 -2
assets/app-showcase-en.png +2 -2
assets/app-showcase-zh.png +2 -2
assets/app-zh.png +2 -2
assets/banner-en.png +2 -2
assets/banner-zh.png +2 -2
assets/compare-en-email.png +2 -2
assets/compare-en-forum.png +2 -2
assets/compare-zh-email.png +2 -2
assets/compare-zh-zhihu.png +2 -2
assets/eval-en.png +3 -0
assets/eval-zh.png +3 -0
assets/quant-top1-en.png +2 -2
assets/training-en.png +2 -2
assets/training-zh.png +2 -2

.gitattributes CHANGED Viewed

@@ -66,3 +66,5 @@ humanizer-12b-bf16.gguf filter=lfs diff=lfs merge=lfs -text
 humanizer-12b-IQ2_XS-QAT.gguf filter=lfs diff=lfs merge=lfs -text
 humanizer-12b-Q3-QAT.gguf filter=lfs diff=lfs merge=lfs -text
 assets/quant-top1-en.png filter=lfs diff=lfs merge=lfs -text

 humanizer-12b-IQ2_XS-QAT.gguf filter=lfs diff=lfs merge=lfs -text
 humanizer-12b-Q3-QAT.gguf filter=lfs diff=lfs merge=lfs -text
 assets/quant-top1-en.png filter=lfs diff=lfs merge=lfs -text
+assets/eval-en.png filter=lfs diff=lfs merge=lfs -text
+assets/eval-zh.png filter=lfs diff=lfs merge=lfs -text

README.md CHANGED Viewed

@@ -42,13 +42,9 @@ More examples (a forum answer, a Zhihu answer) are in the [GitHub README](https:
 ## Results
-<img src="assets/results-detector-en.png" alt="Originality.ai: 95% of rewrites judged human" width="100%">
-**95% judged human by Originality.ai** at its strictest setting (210 English drafts, bf16 weights, 2026-10-02): 11 of 210 rewrites were flagged as AI, against 26 for the previous release, and no AI detector was used anywhere in training. On the same 60 drafts, the most popular de-AI skill on GitHub, applied by Claude Sonnet, had all 60 flagged; this release had 4.
-<img src="assets/results-fidelity-en.png" alt="Fact fidelity compared with the previous releases" width="100%">
-**376 of 420 English rewrites came back with no factual problem** from a strict LLM judge, measured on the `humanizer-12b-Q8_0.gguf` file you download; where it did find one, more than 9 in 10 fixes are a single word or phrase. Method, every genre and the Chinese results: [Evaluation details](#evaluation-details).
 ## Quantization: smaller files, closer to the full model
@@ -290,7 +286,7 @@ If a chat reply greets you, repeats the instruction or doesn't stop, the file ha
 ## Evaluation details
-Evaluation set: 312 drafts (210 English, 102 Chinese), 18 genres, written from scratch by GLM-5.3, GPT-5.6 luna and Claude Sonnet (about a third each), never used in training. Two samples per draft.
 **AI detection (external check only): 95% judged human.** Originality.ai, API v3, AI Allowance 0% (strictest), 2026-10-02, 210 English drafts, first sample each, bf16 weights: 11 of 210 rewrites flagged as AI.

 ## Results
+<img src="assets/eval-en.png" alt="Evaluation card: 95% of English rewrites judged human by Originality.ai at its strictest setting (11 of 210 flagged; previous release 26); 376 of 420 English rewrites with no factual problem (previous release 369)" width="100%">
+**95% judged human by Originality.ai** at its strictest setting (210 English drafts, bf16 weights, 2026-10-02): 11 of 210 rewrites were flagged as AI, against 26 for the previous release, and no AI detector was used anywhere in training. **376 of 420 English rewrites came back with no factual problem** from a strict LLM judge, measured on the `humanizer-12b-Q8_0.gguf` file you download; where it did find one, more than 9 in 10 fixes are a single word or phrase. Method, every genre and the Chinese results: [Evaluation details](#evaluation-details).
 ## Quantization: smaller files, closer to the full model
 ## Evaluation details
+Evaluation set: 312 drafts (210 English, 102 Chinese), 18 genres, written from scratch by GLM-5.3, GPT-5.6 Luna and Claude Sonnet (about a third each), never used in training. Two samples per draft.
 **AI detection (external check only): 95% judged human.** Originality.ai, API v3, AI Allowance 0% (strictest), 2026-10-02, 210 English drafts, first sample each, bf16 weights: 11 of 210 rewrites flagged as AI.