Instructions to use janhq/Jan-v3.5-4B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use janhq/Jan-v3.5-4B with Transformers:

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

pipe = pipeline("text-generation", model="janhq/Jan-v3.5-4B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("janhq/Jan-v3.5-4B")
model = AutoModelForCausalLM.from_pretrained("janhq/Jan-v3.5-4B", 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=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use janhq/Jan-v3.5-4B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "janhq/Jan-v3.5-4B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "janhq/Jan-v3.5-4B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/janhq/Jan-v3.5-4B

SGLang

How to use janhq/Jan-v3.5-4B 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 "janhq/Jan-v3.5-4B" \
    --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": "janhq/Jan-v3.5-4B",
		"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 "janhq/Jan-v3.5-4B" \
        --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": "janhq/Jan-v3.5-4B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use janhq/Jan-v3.5-4B with Docker Model Runner:
```
docker model run hf.co/janhq/Jan-v3.5-4B
```

Training methods and/or data?

by boneylizardwizard - opened 11 days ago

Discussion

boneylizardwizard

11 days ago

The model card claims that your parent company, Menlo Research, finetuned and trained the model, and it provides a link to follow. I misunderstood and assumed I could find information about this training by following the link, which seemed the logical inference at the time, because although Menlo appears to primarily focus on advanced robotics, I went through pretty much the entire site and could not find a single reference to Jan or LLM finetuning.

It would be fascinating to see the methods that Menlo used to organically achieve the remarkable output demonstrated on the model card, and, given that Jan 3.5 itself proclaims that it wants to be "trained by open-source principles and transparency," transparency on how you achieved this conversational result would be helpful. If I somehow missed it on the Menlo website, my bad, and if you could point me specifically to where the training methods and data sets might be outlined, that would also be great.

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