Instructions to use concedo/Huihui-GLM-4.5-Air-abliterated-lossytensors with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use concedo/Huihui-GLM-4.5-Air-abliterated-lossytensors with Transformers:

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

pipe = pipeline("text-generation", model="concedo/Huihui-GLM-4.5-Air-abliterated-lossytensors")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("concedo/Huihui-GLM-4.5-Air-abliterated-lossytensors")
model = AutoModelForMultimodalLM.from_pretrained("concedo/Huihui-GLM-4.5-Air-abliterated-lossytensors")
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 concedo/Huihui-GLM-4.5-Air-abliterated-lossytensors with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "concedo/Huihui-GLM-4.5-Air-abliterated-lossytensors"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "concedo/Huihui-GLM-4.5-Air-abliterated-lossytensors",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/concedo/Huihui-GLM-4.5-Air-abliterated-lossytensors

SGLang

How to use concedo/Huihui-GLM-4.5-Air-abliterated-lossytensors 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 "concedo/Huihui-GLM-4.5-Air-abliterated-lossytensors" \
    --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": "concedo/Huihui-GLM-4.5-Air-abliterated-lossytensors",
		"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 "concedo/Huihui-GLM-4.5-Air-abliterated-lossytensors" \
        --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": "concedo/Huihui-GLM-4.5-Air-abliterated-lossytensors",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use concedo/Huihui-GLM-4.5-Air-abliterated-lossytensors with Docker Model Runner:
```
docker model run hf.co/concedo/Huihui-GLM-4.5-Air-abliterated-lossytensors
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

huihui-ai/Huihui-GLM-4.5-Air-abliterated-lossytensors

This is a lossy safetensors version of Huihui-GLM-4.5-Air-abliterated-GGUF that can be run with huggingface transformers, since the original release did not include safetensors files.

It was re-converted back into .safetensors manually from the Q4_K_M GGUF file.

As such, although the weights are now in BF16, it is considered a "lossy" version of inferior quality to the full precision model.

However, you should now be able to use it with backends that cannot use GGUF and require safetensors (e.g. MLX, VLLM).

Avoid requantizing it to formats above Q4_K_M - You will NOT gain any additional quality.

If you require the max precision version, you'll have to buy it from huihui.

Downloads last month: 4

Safetensors

Model size

110B params

Tensor type

BF16

Model tree for concedo/Huihui-GLM-4.5-Air-abliterated-lossytensors

Quantizations

3 models