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

Libraries

How to use jarradh/GLM-4.7-heretic with Transformers:

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

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

# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("jarradh/GLM-4.7-heretic")
model = AutoModelForMultimodalLM.from_pretrained("jarradh/GLM-4.7-heretic")
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 jarradh/GLM-4.7-heretic with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "jarradh/GLM-4.7-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": "jarradh/GLM-4.7-heretic",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/jarradh/GLM-4.7-heretic

SGLang

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

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

GLM-4.7 - Heretic (Abliterated)

An abliterated version of Z.ai's GLM-4.7 created using Heretic (spikymoth:implement-mpoa). This model has reduced refusals while maintaining model quality.

Model Details

Base Model: zai-org/GLM-4.7
Abliteration Method: Heretic v1.1.0
Trial Selected: Trial 319
Refusals: 4/100
KL Divergence: 0.0181

Parameters

orthogonalize_direction = True

Parameter	Value
direction_index	36.64
attn.o_proj.max_weight	3.87
attn.o_proj.max_weight_position	56.90
attn.o_proj.min_weight	1.83
attn.o_proj.min_weight_distance	49.40
mlp.down_proj.max_weight	2.29
mlp.down_proj.max_weight_position	55.31
mlp.down_proj.min_weight	0.66
mlp.down_proj.min_weight_distance	31.20

Downloads last month: 54

Safetensors

Model size

353B params

Tensor type

F32

BF16

Model tree for jarradh/GLM-4.7-heretic

Base model

zai-org/GLM-4.7

Finetuned

(22)

this model