[!NOTE] Includes Unsloth chat template fixes!
Forllama.cpp, use--jinja
| Model Name | Description | Download |
|---|---|---|
| JanusCoder-8B | 8B text model based on Qwen3-8B. | 🤗 Model |
| 👉 JanusCoder-14B | 14B text model based on Qwen3-14B. | 🤗 Model |
| JanusCoderV-7B | 7B multimodal model based on Qwen2.5-VL-7B. | 🤗 Model |
| JanusCoderV-8B | 8B multimodal model based on InternVL3.5-8B. | 🤗 Model |
| Model | JanusCoder-14B | Qwen3-14B | Qwen2.5-Coder-32B-Instruct | LLaMA3-8B-Instruct | GPT-4o |
|---|---|---|---|---|---|
| PandasPlotBench (Task) | 86 | 78 | 82 | 69 | 85 |
| ArtifactsBench | 41.1 | 36.5 | 35.5 | 36.5 | 37.9 |
| DTVBench (Manim) | 8.41 | 6.63 | 9.61 | 4.92 | 10.60 |
| DTVBench (Wolfram) | 5.97 | 5.08 | 4.98 | 3.15 | 5.97 |
Please use transformers >= 4.55.0 to ensure the model works normally.
1from transformers import AutoTokenizer, AutoModelForCausalLM
2import torch
3
4model_name = "internlm/JanusCoder-14B"
5tokenizer = AutoTokenizer.from_pretrained(model_name)
6model = AutoModelForCausalLM.from_pretrained(model_name, device_map="auto", torch_dtype="auto")
7
8messages = [
9 {
10 "role": "user",
11 "content": [
12 {"type": "text", "text": "Create a line plot that illustrates function y=x."},
13 ],
14 }
15]
16
17inputs = tokenizer.apply_chat_template(messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt").to(model.device, dtype=torch.bfloat16)
18
19generate_ids = model.generate(**inputs, max_new_tokens=32768)
20decoded_output = processor.decode(generate_ids[0, inputs["input_ids"].shape[1] :], skip_special_tokens=True)
21print(decoded_output)1@article{sun2025januscoder,
2 title={JanusCoder: Towards a Foundational Visual-Programmatic Interface for Code Intelligence},
3 author={Sun, Qiushi and Gong, Jingyang and Liu, Yang and Chen, Qiaosheng and Li, Lei and Chen, Kai and Guo, Qipeng and Kao, Ben and Yuan, Fei},
4 journal={arXiv preprint arXiv:2510.23538},
5 year={2025}
6}
7
8@article{sun2024survey,
9 title={A survey of neural code intelligence: Paradigms, advances and beyond},
10 author={Sun, Qiushi and Chen, Zhirui and Xu, Fangzhi and Cheng, Kanzhi and Ma, Chang and Yin, Zhangyue and Wang, Jianing and Han, Chengcheng and Zhu, Renyu and Yuan, Shuai and others},
11 journal={arXiv preprint arXiv:2403.14734},
12 year={2024}
13}
14
15@article{chen2025interactscience,
16 title={InteractScience: Programmatic and Visually-Grounded Evaluation of Interactive Scientific Demonstration Code Generation},
17 author={Chen, Qiaosheng and Liu, Yang and Li, Lei and Chen, Kai and Guo, Qipeng and Cheng, Gong and Yuan, Fei},
18 journal={arXiv preprint arXiv:2510.09724},
19 year={2025}
20}
21
22@article{sun2025codeevo,
23 title={CodeEvo: Interaction-Driven Synthesis of Code-centric Data through Hybrid and Iterative Feedback},
24 author={Sun, Qiushi and Gong, Jinyang and Li, Lei and Guo, Qipeng and Yuan, Fei},
25 journal={arXiv preprint arXiv:2507.22080},
26 year={2025}
27}