Instructions to use CoNDeNse-AI/GLM-5.1-Qwen3-1.7B-CoNDeNse with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use CoNDeNse-AI/GLM-5.1-Qwen3-1.7B-CoNDeNse with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3-1.7B")
model = PeftModel.from_pretrained(base_model, "CoNDeNse-AI/GLM-5.1-Qwen3-1.7B-CoNDeNse")

Transformers

How to use CoNDeNse-AI/GLM-5.1-Qwen3-1.7B-CoNDeNse with Transformers:

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

pipe = pipeline("text-generation", model="CoNDeNse-AI/GLM-5.1-Qwen3-1.7B-CoNDeNse")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("CoNDeNse-AI/GLM-5.1-Qwen3-1.7B-CoNDeNse")
model = AutoModelForCausalLM.from_pretrained("CoNDeNse-AI/GLM-5.1-Qwen3-1.7B-CoNDeNse", device_map="auto")
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=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use CoNDeNse-AI/GLM-5.1-Qwen3-1.7B-CoNDeNse with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "CoNDeNse-AI/GLM-5.1-Qwen3-1.7B-CoNDeNse"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "CoNDeNse-AI/GLM-5.1-Qwen3-1.7B-CoNDeNse",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/CoNDeNse-AI/GLM-5.1-Qwen3-1.7B-CoNDeNse

SGLang

How to use CoNDeNse-AI/GLM-5.1-Qwen3-1.7B-CoNDeNse 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 "CoNDeNse-AI/GLM-5.1-Qwen3-1.7B-CoNDeNse" \
    --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": "CoNDeNse-AI/GLM-5.1-Qwen3-1.7B-CoNDeNse",
		"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 "CoNDeNse-AI/GLM-5.1-Qwen3-1.7B-CoNDeNse" \
        --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": "CoNDeNse-AI/GLM-5.1-Qwen3-1.7B-CoNDeNse",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Unsloth Desktop
Docker Model Runner
How to use CoNDeNse-AI/GLM-5.1-Qwen3-1.7B-CoNDeNse with Docker Model Runner:
```
docker model run hf.co/CoNDeNse-AI/GLM-5.1-Qwen3-1.7B-CoNDeNse
```

Discussion: Research alignment and potential synergy with Qubik

by Phase-Technologies - opened May 22

Discussion

Phase-Technologies

May 22

Hi Andy,
I’ve been following the work at CoNDeNse-AI with great interest
your focus on reasoning-driven compact models really resonates with the current development of Qubik, a 5B-parameter search-optimized model for edge intelligence.
I'm currently building out our roadmap for nested search and real-time verification and would love to exchange thoughts on the methodologies you've been using. Do you have any availability for a brief technical chat on how our research goals might align?
Looking forward to your thoughts.
Best,
Soham Pal
Founder, Xerv

Andy-ML-And-AI

CoNDeNse-AI org May 22

Hi Soham,
Thanks for reaching out — I appreciate the note. Qubik sounds really interesting, especially the focus on search-optimized reasoning and edge intelligence.
The nested search and verification direction is something I’ve been thinking about a lot as well, particularly around efficient tool use and maintaining reasoning quality in smaller models.
I’d definitely be open to a short technical discussion to learn more about what you’re building and see where our ideas overlap. Feel free to share a few time slots that work for you, along with any papers, demos, or technical notes you’d like me to look at beforehand.
Looking forward to chatting.

Best Regards,
Andy

Phase-Technologies

May 23

Thanks Andy. i really wanna work with you. it will really be interesting. kindly send a mail to xerv.org@gmail.com so we can start 🔥

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment