ThomasBaruzier/Qwen2.5-Coder-7B-Instruct-GGUF

Llama.cpp imatrix quantizations of Qwen/Qwen2.5-Coder-7B-Instruct

Using llama.cpp commit 3ad5451 for quantization.

All quants were made using the imatrix option and Bartowski's calibration file.

---

Perplexity table (the lower the better)

Quant	Size (MB)	PPL	Size (%)	Accuracy (%)	PPL error rate
IQ1_S	1815	18.3139	12.49	50.82	0.29
IQ1_M	1947	14.7056	13.40	63.29	0.24
IQ2_XXS	2167	12.3126	14.91	75.59	0.20
IQ2_XS	2354	11.0302	16.20	84.38	0.17
IQ2_S	2475	10.7236	17.03	86.79	0.17
IQ2_M	2651	10.3509	18.24	89.92	0.16
Q2_K_S	2702	10.6991	18.59	86.99	0.17
Q2_K	2876	10.3696	19.79	89.75	0.16
IQ3_XXS	2970	9.8351	20.44	94.63	0.15
IQ3_XS	3191	9.5737	21.96	97.21	0.15
Q3_K_S	3330	9.7610	22.92	95.35	0.15
IQ3_S	3337	9.5520	22.96	97.44	0.15
IQ3_M	3408	9.5248	23.45	97.71	0.15
Q3_K_M	3631	9.6151	24.99	96.80	0.15
Q3_K_L	3899	9.5675	26.83	97.28	0.15
IQ4_XS	4023	9.3942	27.68	99.07	0.15
IQ4_NL	4232	9.3991	29.12	99.02	0.15
Q4_0	4238	9.4483	29.17	98.50	0.15
Q4_K_S	4251	9.3869	29.25	99.15	0.15
Q4_K_M	4466	9.3699	30.73	99.33	0.15
Q4_1	4647	9.3797	31.98	99.22	0.15
Q5_K_S	5068	9.3144	34.88	99.92	0.14
Q5_0	5081	9.3419	34.97	99.63	0.14
Q5_K_M	5192	9.3152	35.73	99.91	0.14
Q5_1	5490	9.3111	37.78	99.96	0.14
Q6_K	5964	9.3168	41.04	99.89	0.14
Q8_0	7723	9.3061	53.15	100.01	0.14
F16	14531	9.3071	100	100	0.14

---

Qwen2.5-Coder-7B-Instruct

Introduction

Qwen2.5-Coder is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen). As of now, Qwen2.5-Coder has covered six mainstream model sizes, 0.5, 1.5, 3, 7, 14, 32 billion parameters, to meet the needs of different developers. Qwen2.5-Coder brings the following improvements upon CodeQwen1.5:

• Significantly improvements in code generation, code reasoning and code fixing. Base on the strong Qwen2.5, we scale up the training tokens into 5.5 trillion including source code, text-code grounding, Synthetic data, etc. Qwen2.5-Coder-32B has become the current state-of-the-art open-source codeLLM, with its coding abilities matching those of GPT-4o.
• A more comprehensive foundation for real-world applications such as Code Agents. Not only enhancing coding capabilities but also maintaining its strengths in mathematics and general competencies.
• Long-context Support up to 128K tokens.

This repo contains the instruction-tuned 7B Qwen2.5-Coder model, which has the following features:

• 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
  • Please refer to this section for detailed instructions on how to deploy Qwen2.5 for handling long texts.

For more details, please refer to our blog, GitHub, Documentation, Arxiv.

Requirements

The code of Qwen2.5-Coder 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

Here provides a code snippet with apply_chat_template to show you how to load the tokenizer and model and how to generate contents.

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "Qwen/Qwen2.5-Coder-7B-Instruct"

model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype="auto",
    device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained(model_name)

prompt = "write a quick sort algorithm."
messages = [
    {"role": "system", "content": "You are Qwen, created by Alibaba Cloud. You are a helpful assistant."},
    {"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=512
)
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]

Processing Long Texts

The current config.json is set for context length up to 32,768 tokens. To handle extensive inputs exceeding 32,768 tokens, we utilize YaRN, a technique for enhancing model length extrapolation, ensuring optimal performance on lengthy texts.

For supported frameworks, you could add the following to config.json to enable YaRN:

{
  ...,
  "rope_scaling": {
    "factor": 4.0,
    "original_max_position_embeddings": 32768,
    "type": "yarn"
  }
}

For deployment, we recommend using vLLM. Please refer to our Documentation for usage if you are not familar with vLLM. Presently, vLLM only supports static YARN, which means the scaling factor remains constant regardless of input length, potentially impacting performance on shorter texts. We advise adding the rope_scaling configuration only when processing long contexts is required.

Evaluation & Performance

Detailed evaluation results are reported in this 📑 blog.

For requirements on GPU memory and the respective throughput, see results here.

Citation

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

@article{hui2024qwen2,
      title={Qwen2. 5-Coder Technical Report},
      author={Hui, Binyuan and Yang, Jian and Cui, Zeyu and Yang, Jiaxi and Liu, Dayiheng and Zhang, Lei and Liu, Tianyu and Zhang, Jiajun and Yu, Bowen and Dang, Kai and others},
      journal={arXiv preprint arXiv:2409.12186},
      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}
}