RichardErkhov/aigcode_-_AIGCodeGeek-DS-6.7B-gguf

Quantization made by Richard Erkhov.

Github

Discord

Request more models

AIGCodeGeek-DS-6.7B - GGUF

• Model creator: https://huggingface.co/aigcode/
• Original model: https://huggingface.co/aigcode/AIGCodeGeek-DS-6.7B/

Name	Quant method	Size
AIGCodeGeek-DS-6.7B.Q2_K.gguf	Q2_K	2.36GB
AIGCodeGeek-DS-6.7B.Q3_K_S.gguf	Q3_K_S	2.75GB
AIGCodeGeek-DS-6.7B.Q3_K.gguf	Q3_K	3.07GB
AIGCodeGeek-DS-6.7B.Q3_K_M.gguf	Q3_K_M	3.07GB
AIGCodeGeek-DS-6.7B.Q3_K_L.gguf	Q3_K_L	3.35GB
AIGCodeGeek-DS-6.7B.IQ4_XS.gguf	IQ4_XS	3.4GB
AIGCodeGeek-DS-6.7B.Q4_0.gguf	Q4_0	3.56GB
AIGCodeGeek-DS-6.7B.IQ4_NL.gguf	IQ4_NL	3.59GB
AIGCodeGeek-DS-6.7B.Q4_K_S.gguf	Q4_K_S	3.59GB
AIGCodeGeek-DS-6.7B.Q4_K.gguf	Q4_K	3.8GB
AIGCodeGeek-DS-6.7B.Q4_K_M.gguf	Q4_K_M	3.8GB
AIGCodeGeek-DS-6.7B.Q4_1.gguf	Q4_1	3.95GB
AIGCodeGeek-DS-6.7B.Q5_0.gguf	Q5_0	4.33GB
AIGCodeGeek-DS-6.7B.Q5_K_S.gguf	Q5_K_S	4.33GB
AIGCodeGeek-DS-6.7B.Q5_K.gguf	Q5_K	4.46GB
AIGCodeGeek-DS-6.7B.Q5_K_M.gguf	Q5_K_M	4.46GB
AIGCodeGeek-DS-6.7B.Q5_1.gguf	Q5_1	4.72GB
AIGCodeGeek-DS-6.7B.Q6_K.gguf	Q6_K	5.15GB
AIGCodeGeek-DS-6.7B.Q8_0.gguf	Q8_0	6.67GB

Original model description:

library_name: transformers tags: - code datasets: - Leon-Leee/wizardlm_evol_instruct_v2_196K_backuped - m-a-p/Code-Feedback - openbmb/UltraInteract_sft - ise-uiuc/Magicoder-Evol-Instruct-110K - flytech/python-codes-25k metrics: - code_eval pipeline_tag: text-generation license: other license name: deepseek

AIGCodeGeek-DS-6.7B

Introduction

AIGCodeGeek-DS-6.7B is our first released version of a Code-LLM family with competitive performance on public and private benchmarks.

Model Details

Model Description

• Developed by: Leon Li
• License: DeepSeek
• Fine-tuned from deepseek-ai/deepseek-coder-6.7b-base with full parameters

Training data

A mixture of samples from high-quality open-source (read Acknowledgements) and our private datasets. We have made contamination detection as Magicoder/Bigcode did (https://github.com/ise-uiuc/magicoder/blob/main/src/magicoder/decontamination/find_substrings.py).

Evaluation

results to be added.

Requirements

It should work with the same requirements as DeepSeek-Coder-6.7B or the following packages:

tokenizers>=0.14.0
transformers>=4.35.0
accelerate
sympy>=1.12
pebble 
timeout-decorator 
attrdict

QuickStart

import torch
from transformers import AutoTokenizer, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("aigcode/AIGCodeGeek-DS-6.7B", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("aigcode/AIGCodeGeek-DS-6.7B", trust_remote_code=True, torch_dtype=torch.bfloat16).cuda()
messages=[
    { 'role': 'user', 'content': "write a merge sort algorithm in python."}
]
inputs = tokenizer.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt").to(model.device)
# tokenizer.eos_token_id is the id of <|EOT|> token
outputs = model.generate(inputs, max_new_tokens=512, do_sample=False, top_k=50, top_p=0.95, num_return_sequences=1, eos_token_id=tokenizer.eos_token_id)
print(tokenizer.decode(outputs[0][len(inputs[0]):], skip_special_tokens=True))

Acknowledgements

We gain a lot of knowledge and resources from the open-source community:

• DeepSeekCoder: impressive model series and insightful tech reports
• WizardCoder: Evol Instruct and public datasets
  • We used a (Leon-Leee/wizardlm_evol_instruct_v2_196K_backuped) since this original has been deleted.
• Magicoder: OSS-Instruct, Magicoder-Evol-Instruct-110K from theblackcat102/evol-codealpaca-v1(https://huggingface.co/datasets/theblackcat102/evol-codealpaca-v1)
• Eurus: creative datasets for reasoning, openbmb/UltraInteract_sft
• OpenCoderInterpreter: well-designed system and datasets m-a-p/Code-Feedback
• flytech/python-codes-25k: diversity
• LLaMA-Factory: easily used to finetune base models