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

Starcoder2 3B by bigcode

 »  All LLMs  »  bigcode  »  Starcoder2 3B   URL Share it on

  210769   
  Arxiv:2004.05150   Arxiv:2205.14135   Arxiv:2207.14255   Arxiv:2305.13245   Arxiv:2402.19173   Autotrain compatible   Code Dataset:bigcode/the-stack-v2-t...   Endpoints compatible   Model-index   Region:us   Safetensors   Starcoder2
Model Card on HF 🤗: https://huggingface.co/bigcode/starcoder2-3b 

Starcoder2 3B Benchmarks

MMLU Pro: 7.07
MUSR: 1.43
BBH: 8.91
IFEval: 20.37 vs 88 (so35)-76.9%
ARC: 34.56 vs 96.7 (so35)-64.3%
HellaSwag: 47.62 vs 95.3 (gpt4)-50%
MMLU: 38.65 vs 88.3 (so35)-56.2%
TruthfulQA: 40.49 vs 59 (gpt4)-31.4%
WinoGrande: 54.54 vs 87.5 (gpt4)-37.7%
GSM8K: 19.64 vs 96.4 (so35)-79.6%
HumanEvalA: 23.43
Python Eva: 31.44 vs 92 (so35)-65.8%
Java Eval: 27.41
Javascript: 35.37
CPP Eval: 27.24
PHP Eval: 27.61
Julia Eval: 19.87
D Eval: 12.56
Lua Eval: 28.01
R Eval: 14.22
Racket Eva: 7.8
Rust Eval: 24.52
Swift Eval: 25.09
MATH Lvl 5: 1.51
LLME Score: 0.28877
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Starcoder2 3B (bigcode/starcoder2-3b)

Starcoder2 3B Parameters and Internals

Model Type 
text-generation
Use Cases 
Areas:
research, programming
Primary Use Cases:
code generation
Limitations:
Generated code may contain bugs or exploits., Inefficiency and unintended outputs
Supported Languages 
programming_languages (17 programming languages,)
Training Details 
Data Sources:
GitHub code, Arxiv, Wikipedia
Data Volume:
3+ trillion tokens
Methodology:
Grouped Query Attention, Fill-in-the-Middle objective
Context Length:
16384
Hardware Used:
160 A100 GPUs
Model Architecture:
Transformer decoder with grouped-query and sliding window attention
Input Output 
Accepted Modalities:
text
LLM NameStarcoder2 3B
Repository 🤗https://huggingface.co/bigcode/starcoder2-3b 
Model Size3b
Required VRAM12.1 GB
Updated2025-05-31
Maintainerbigcode
Model Typestarcoder2
Model Files  12.1 GB
Model ArchitectureStarcoder2ForCausalLM
Licensebigcode-openrail-m
Context Length16384
Model Max Length16384
Transformers Version4.37.0.dev0
Tokenizer ClassGPT2Tokenizer
Vocabulary Size49152

Quantized Models of the Starcoder2 3B

Model
Likes
Downloads
VRAM
Starcoder2 3B AWQ02782 GB
StarCoder2 3B GGUF912681 GB

Best Alternatives to Starcoder2 3B

Best Alternatives
Context / RAM
Downloads
Likes
Starcoder2 3b AutoRedteam16K / 12.7 GB60
Starcoder Proto Code16K / 6.1 GB100
Mojo Starcoder216K / 6.4 GB110
NEAR StructTunedStarcoder216K / 6.1 GB190
NEARCoder 3B16K / 6.1 GB100
NEAR PreTrainedStarCoder216K / 6.1 GB130
Bigcode Starcoder2 3B 8bits16K / 3.2 GB130
Starcoder2 3B Instruct16K / 6.1 GB2254
Opencsg Starcoder2 3B V0.116K / 6.4 GB131
OpenCodeInterpreter SC2 3B16K / 6.4 GB137

Rank the Starcoder2 3B 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? 47753 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