LLM EXPLORER
 LLM Explorer: A Curated Large Language Model Directory and Analytics
Dark Theme
  LLM  List    LLM Hosting    LLM Leaderboards    Blog    Newsfeed    Advertise

Deepseek Coder 6.7B Instruct by deepseek-ai

 »  All LLMs  »  deepseek-ai  »  Deepseek Coder 6.7B Instruct   URL Share it on

  88134   
  Autotrain compatible   Codegen   Conversational   Endpoints compatible   Instruct   Llama   Pytorch   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/deepseek-ai/deepseek-coder-6.7b-instruct 

Deepseek Coder 6.7B Instruct Benchmarks

ARC: 38.14 vs 96.7 (so35)-60.6%
HellaSwag: 55.09 vs 95.3 (gpt4)-42.2%
MMLU: 39.02 vs 88.3 (so35)-55.8%
TruthfulQA: 45.56 vs 59 (gpt4)-22.8%
WinoGrande: 56.83 vs 87.5 (gpt4)-35.1%
GSM8K: 26.76 vs 96.4 (so35)-72.2%
HumanEvalA: 48.17
Python Eva: 80.22 vs 92 (so35)-12.8%
Java Eval: 53.34
Javascript: 65.8
CPP Eval: 59.66
PHP Eval: 59.4
Julia Eval: 38.84
D Eval: 21.59
Lua Eval: 47.78
R Eval: 38.56
Racket Eva: 20.87
Rust Eval: 47.73
Swift Eval: 44.22
LLME Score: 0.26996
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Deepseek Coder 6.7B Instruct (deepseek-ai/deepseek-coder-6.7b-instruct)

Deepseek Coder 6.7B Instruct Parameters and Internals

Model Type 
code generation
Use Cases 
Areas:
coding assistant, software development
Applications:
code completion, project-level code corpus analysis
Primary Use Cases:
code completion, code infilling
Supported Languages 
English (fluent), Chinese (fluent)
Training Details 
Data Sources:
2T tokens consisting of 87% code and 13% natural language in both English and Chinese
Data Volume:
2T tokens
Context Length:
16000
Input Output 
Accepted Modalities:
text
LLM NameDeepseek Coder 6.7B Instruct
Repository 🤗https://huggingface.co/deepseek-ai/deepseek-coder-6.7b-instruct 
Model Size7b
Required VRAM13.5 GB
Updated2025-03-14
Maintainerdeepseek-ai
Model Typellama
Instruction-BasedYes
Model Files  10.0 GB: 1-of-2   3.5 GB: 2-of-2   10.0 GB: 1-of-2   3.5 GB: 2-of-2
Generates CodeYes
Model ArchitectureLlamaForCausalLM
Licenseother
Context Length16384
Model Max Length16384
Transformers Version4.34.1
Tokenizer ClassLlamaTokenizerFast
Beginning of Sentence Token<｜begin▁of▁sentence｜>
End of Sentence Token<|EOT|>
Vocabulary Size32256
Torch Data Typebfloat16

Quantized Models of the Deepseek Coder 6.7B Instruct

Model
Likes
Downloads
VRAM
...pseek Coder 6.7B Instruct GGUF195257952 GB
...epseek Coder 6.7B Instruct AWQ1615063 GB
...pseek Coder 6.7B Instruct GGUF67532 GB
...pseek Coder 6.7B Instruct GPTQ244933 GB

Best Alternatives to Deepseek Coder 6.7B Instruct

Best Alternatives
Context / RAM
Downloads
Likes
CodeLlama 7B Instruct Hf16K / 13.5 GB41864225
CodeLlama 7B Instruct Hf16K / 13.5 GB2815444
GetCode Slerp16K / 13.6 GB17371
CodeV CL 7B16K / 13.5 GB921
Cdlm 7 Ko Nl2sql V1.016K / 13.5 GB20154
Stack Codellama 7B Inst16K / 13.5 GB100
Glaive Coder 7B16K / 27 GB216954
...ama 7B Instruct Hf Dequantized16K / 13.5 GB1970
...deLlama 7B Instruct Hf 8bits Q16K / 4.2 GB211
Llama 2 7B Chat Finetune516K / 13.8 GB111

Rank the Deepseek Coder 6.7B Instruct Capabilities

🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟

Instruction Following and Task Automation  
Factuality and Completeness of Knowledge  
Censorship and Alignment  
Data Analysis and Insight Generation  
Text Generation  
Text Summarization and Feature Extraction  
Code Generation  
Multi-Language Support and Translation  

What open-source LLMs or SLMs are you in search of? 45095 in total.

Email us: info@extractum.io. Our Privacy Policy  | Terms and Conditions  | Suggest an improvement.
Our Social Media →  
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241227
  Support LLM Explorer