open-neo_-_Kyro-n1.1-7B-4bits

Quantization made by Richard Erkhov.

Github

Discord

Request more models

Kyro-n1.1-7B - bnb 4bits

• Model creator: https://huggingface.co/open-neo/
• Original model: https://huggingface.co/open-neo/Kyro-n1.1-7B/

Original model description:

license: other license_name: kyro license_link: LICENSE.md base_model:

• Qwen/Qwen2.5-7B-Instruct tags:
• reasoning
• kyro
• open-neo
• open-source
• deepseek-r1 language:
• en
• zh
• fr
• es
• pt
• de
• it
• ru
• ja
• ko
• vi
• th
• ar
• fa
• he
• tr
• cs
• pl
• hi
• bn
• ur
• id
• ms
• lo
• my
• ceb
• km
• tl
• nl library_name: transformers

---

Kyro-n1.1: Smarter, Sharper, and More Capable

Kyro-n1.1 is an enhanced iteration of Kyro-n1, designed to deliver superior reasoning, improved comprehension, and higher response accuracy. Built upon Qwen2.5-7B-Instruct, this model leverages advanced fine-tuning techniques to refine its ability to analyse complex queries, provide well-structured responses, and engage in more nuanced conversations.

Key Improvements Over Kyro-n1

• Enhanced Reasoning Capabilities: Kyro-n1.1 demonstrates stronger logical thinking, making it more reliable for tasks requiring deep analysis.
• More Accurate Responses: Refined dataset curation and improved fine-tuning methods ensure better factual consistency.
• Broader Context Understanding: With improved context retention, Kyro-n1.1 handles multi-turn conversations with greater coherence.
• Optimised for Open-Source Collaboration: As part of the Open-Neo initiative, Kyro-n1.1 is designed to be a transparent, accessible, and community-driven model.

Why Choose Kyro-n1.1?

• Ideal for Research & Development: Whether you're exploring AI reasoning benchmarks or enhancing your own projects, Kyro-n1.1 is built for performance.
• Balanced for Various Use Cases: From general Q&A to coding assistance and creative writing, the model adapts well across different applications.
• Efficient & Scalable: Designed to be computationally efficient, Kyro-n1.1 delivers strong performance while maintaining manageable resource requirements.
• Fully Open-Source: As part of the Open-Neo ecosystem, Kyro-n1.1 is freely available for modification and integration into various workflows.

Technical Specifications

• Developed by: Spestly (Open-Neo) & Kazex (Open-Neo) & Adversing (Open-Neo)
• Type: Causal Language Models
• Training Stage: Pretraining & Post-training
• Architecture: transformers with RoPE, SwiGLU, RMSNorm, and Attention QKV bias
• Number of Parameters: 7.61B
• Number of Paramaters (Non-Embedding): 6.53B
• Number of Layers: 28
• Number of Attention Heads (GQA): 28 for Q and 4 for KV
• Context Length: Full 131,072 tokens and generation 8192 tokens

Usage

The code of Kyro-n1.1 (Qwen2.5) has been in the latest Hugging face transformers and we advise you to use the latest version of transformers.

With transformers<4.37.0, you will encounter the following error:

KeyError: 'qwen2'

Quickstart

from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "open-neo/Kyro-n1.1-7B"
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype="auto",
    device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained(model_name)
prompt = "What do you think about CRISPR and its effect on the future of humanity?"
messages = [
    {"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
generated_ids = model.generate(
    **model_inputs,
    max_new_tokens=2048
)
generated_ids = [
    output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]

Citation

If you find our work helpful, feel free to give us a cite.

@misc{qwen2.5,
    title = {Qwen2.5: A Party of Foundation Models},
    url = {https://qwenlm.github.io/blog/qwen2.5/},
    author = {Qwen Team},
    month = {September},
    year = {2024}
}
@article{qwen2,
      title={Qwen2 Technical Report}, 
      author={An Yang and Baosong Yang and Binyuan Hui and Bo Zheng and Bowen Yu and Chang Zhou and Chengpeng Li and Chengyuan Li and Dayiheng Liu and Fei Huang and Guanting Dong and Haoran Wei and Huan Lin and Jialong Tang and Jialin Wang and Jian Yang and Jianhong Tu and Jianwei Zhang and Jianxin Ma and Jin Xu and Jingren Zhou and Jinze Bai and Jinzheng He and Junyang Lin and Kai Dang and Keming Lu and Keqin Chen and Kexin Yang and Mei Li and Mingfeng Xue and Na Ni and Pei Zhang and Peng Wang and Ru Peng and Rui Men and Ruize Gao and Runji Lin and Shijie Wang and Shuai Bai and Sinan Tan and Tianhang Zhu and Tianhao Li and Tianyu Liu and Wenbin Ge and Xiaodong Deng and Xiaohuan Zhou and Xingzhang Ren and Xinyu Zhang and Xipin Wei and Xuancheng Ren and Yang Fan and Yang Yao and Yichang Zhang and Yu Wan and Yunfei Chu and Yuqiong Liu and Zeyu Cui and Zhenru Zhang and Zhihao Fan},
      journal={arXiv preprint arXiv:2407.10671},
      year={2024}
}
@misc{kyro-n1.1,
      title={Kyro-n1: Smarter, Sharper, and More Capable }, 
      author={Open-Neo},
      howpublished={https://huggingface.co/collections/open-neo/kyro-n1-67ab2e7bbc76a9aab3030c21},
      year={2025}
}

Get Involved

Kyro-n1.1 is a community-driven effort, and contributions are welcome! Whether it's fine-tuning, testing, or providing feedback, your input helps shape the model's future. Join the Open-Neo community to collaborate and improve Kyro-n1.1 together!