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

Starcoderbase by bigcode

 »  All LLMs  »  bigcode  »  Starcoderbase   URL Share it on

  1695   
  Arxiv:1911.02150   Arxiv:2205.14135   Arxiv:2207.14255   Arxiv:2305.06161   Autotrain compatible   Code   Codegen Dataset:bigcode/the-stack-dedu...   Endpoints compatible   Gpt bigcode   Model-index   Pytorch   Region:us   Sharded
Model Card on HF 🤗: https://huggingface.co/bigcode/starcoderbase 

Starcoderbase Benchmarks

ARC: 30.29 vs 96.7 (so35)-68.7%
HellaSwag: 47.21 vs 95.3 (gpt4)-50.5%
MMLU: 32.12 vs 88.3 (so35)-63.6%
TruthfulQA: 40.02 vs 59 (gpt4)-32.2%
WinoGrande: 55.8 vs 87.5 (gpt4)-36.2%
GSM8K: 7.88 vs 96.4 (so35)-91.8%
HumanEvalA: 22.4
Python Eva: 30.35 vs 92 (so35)-67%
Java Eval: 28.53
Javascript: 31.7
CPP Eval: 30.56
PHP Eval: 26.75
Julia Eval: 21.09
D Eval: 10.01
Lua Eval: 26.61
R Eval: 10.18
Racket Eva: 11.77
Rust Eval: 24.46
Swift Eval: 16.74
LLME Score: 0.17791
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Starcoderbase (bigcode/starcoderbase)

Starcoderbase Parameters and Internals

Model Type 
text generation
Use Cases 
Areas:
code generation and technical assistance
Applications:
code completion, debugging
Limitations:
The generated code may not work as intended, might be inefficient, contain bugs or exploits.
Considerations:
Generated code may require further validation.
Additional Notes 
StarCoderBase models are not instruction models.
Supported Languages 
Programming Languages (80+)
Training Details 
Data Sources:
The Stack (v1.2)
Data Volume:
1 trillion tokens
Methodology:
Fill-in-the-Middle objective
Context Length:
8192
Training Time:
24 days
Hardware Used:
512 Tesla A100 GPUs
Model Architecture:
GPT-2 model with multi-query attention
LLM NameStarcoderbase
Repository 🤗https://huggingface.co/bigcode/starcoderbase 
Required VRAM63.3 GB
Updated2025-02-22
Maintainerbigcode
Model Typegpt_bigcode
Model Files  9.9 GB: 1-of-7   9.9 GB: 2-of-7   9.8 GB: 3-of-7   9.9 GB: 4-of-7   9.8 GB: 5-of-7   9.9 GB: 6-of-7   4.1 GB: 7-of-7
Gated ModelYes
Generates CodeYes
Model ArchitectureGPTBigCodeForCausalLM
Licenseproprietary
Transformers Version4.28.1
Tokenizer ClassGPT2Tokenizer
Vocabulary Size49152
Torch Data Typefloat32
Activation Functiongelu

Best Alternatives to Starcoderbase

Best Alternatives
Context / RAM
Downloads
Likes
Opencoderplus0K / 31.8 GB1964103
...rcoder Gpteacher Code Instruct0K / 63.3 GB186681
StarCoder Alpaca0K / 31 GB19531
Sqlcoder0K / 31.2 GB1122319
Sqlcoder20K / 31.2 GB612111
Minicoder Random0K / 0.1 GB1810
SteloCoder0K / 31.2 GB71
Minicoder Lua0K / 0.1 GB1650
DPO Santacoder1b0K / 2.2 GB640
Tiny Starcoder Py0K /  GB1491

Rank the Starcoderbase 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? 43470 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