Jan-v3-4B-base-instruct: a 4B baseline model for fine-tuning

Overview

Jan-v3-4B-base-instruct is a 4B-parameter model obtained via post-training distillation from a larger teacher, transferring capabilities while preserving general-purpose performance on standard benchmarks. The result is a compact, ownable base that is straightforward to fine-tune, broadly applicable and minimizing the usual capacity–capability trade-offs.

Building on this base, Jan-Code, a code-tuned variant, will be released soon.

Model Overview

This repo contains the BF16 version of Jan-v3-4B-base-instruct, which has the following features:

• Type: Causal Language Models
• Training Stage: Pretraining & Post-training
• Number of Parameters: 4B in total
• Number of Layers: 36
• Number of Attention Heads (GQA): 32 for Q and 8 for KV
• Context Length: 262,144 natively.

Intended Use

• A better small base for downstream work: improved instruction following out of the box, strong starting point for fine-tuning, and effective lightweight coding assistance.

Performance

Quick Start

Integration with Jan Apps

Jan-v3 demo is hosted on Jan Browser at chat.jan.ai. It is also optimized for direct integration with Jan Desktop, select the model in the app to start using it.

Local Deployment

Using vLLM:

vllm serve janhq/Jan-v3-4B-base-instruct \
    --host 0.0.0.0 \
    --port 1234 \
    --enable-auto-tool-choice \
    --tool-call-parser hermes 
    

Using llama.cpp:

llama-server --model Jan-v3-4B-base-instruct-Q8_0.gguf \
    --host 0.0.0.0 \
    --port 1234 \
    --jinja \
    --no-context-shift

Recommended Parameters

For optimal performance in agentic and general tasks, we recommend the following inference parameters:

temperature: 0.7
top_p: 0.8
top_k: 20

🤝 Community & Support

• Discussions: Hugging Face Community
• Jan App: Learn more about the Jan App at jan.ai

📄 Citation

Updated Soon