Instructions to use qspqww/GLM-4.7-Flash-heretic-oQ8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use qspqww/GLM-4.7-Flash-heretic-oQ8 with Transformers:

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

pipe = pipeline("text-generation", model="qspqww/GLM-4.7-Flash-heretic-oQ8")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("qspqww/GLM-4.7-Flash-heretic-oQ8")
model = AutoModelForCausalLM.from_pretrained("qspqww/GLM-4.7-Flash-heretic-oQ8")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use qspqww/GLM-4.7-Flash-heretic-oQ8 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "qspqww/GLM-4.7-Flash-heretic-oQ8"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "qspqww/GLM-4.7-Flash-heretic-oQ8",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/qspqww/GLM-4.7-Flash-heretic-oQ8

SGLang

How to use qspqww/GLM-4.7-Flash-heretic-oQ8 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 "qspqww/GLM-4.7-Flash-heretic-oQ8" \
    --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": "qspqww/GLM-4.7-Flash-heretic-oQ8",
		"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 "qspqww/GLM-4.7-Flash-heretic-oQ8" \
        --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": "qspqww/GLM-4.7-Flash-heretic-oQ8",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use qspqww/GLM-4.7-Flash-heretic-oQ8 with Docker Model Runner:
```
docker model run hf.co/qspqww/GLM-4.7-Flash-heretic-oQ8
```

GLM-4.7-Flash-heretic

WARNING: This model is UNCENSORED, which means it may and likely will generate ANY HARMFUL content if you want. Use this model at your own discreetion. DO NOT use this model in any unlawful way!

Base model: GLM-4.7-Flash

Abliteration tests:

KL divergence: 0.0106

Refusals: 2/100

Speed: no degration compared with base model

Downloads last month: 85

Safetensors

Model size

8B params

Tensor type

BF16

U32

F32