Skywork 13B Base by Skywork

 ยป  All LLMs  ยป  Skywork  ยป  Skywork 13B Base   URL Share it on

  Arxiv:2310.16713   Arxiv:2310.19341   Autotrain compatible   Custom code   Pytorch   Region:us   Sharded   Skywork

Skywork 13B Base Benchmarks

nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Skywork 13B Base (Skywork/Skywork-13B-base)
๐ŸŒŸ Advertise your project ๐Ÿš€

Skywork 13B Base Parameters and Internals

Model Type 
text generation
Additional Notes 
Performance on multilingual tasks including Chinese and code data. Achievements on benchmarks are highlighted.
Supported Languages 
Chinese (High), English (High)
Training Details 
Data Sources:
Multilingual data, Code data
Data Volume:
3.2 trillion tokens
Methodology:
Data cleaning and filtering process
Context Length:
4096
Model Architecture:
Thinner and deeper than Llama-2-13B with 52 layers
Input Output 
Input Format:
Tokenized text
Accepted Modalities:
Text
Output Format:
Generated text
LLM NameSkywork 13B Base
Repository ๐Ÿค—https://huggingface.co/Skywork/Skywork-13B-base 
Model Size13b
Required VRAM20.5 GB
Updated2025-10-10
MaintainerSkywork
Model Typeskywork
Model Files  0.5 GB: 1-of-53   0.5 GB: 2-of-53   0.5 GB: 3-of-53   0.5 GB: 4-of-53   0.5 GB: 5-of-53   0.5 GB: 6-of-53   0.5 GB: 7-of-53   0.5 GB: 8-of-53   0.5 GB: 9-of-53   0.5 GB: 10-of-53   0.5 GB: 11-of-53   0.5 GB: 12-of-53   0.5 GB: 13-of-53   0.5 GB: 14-of-53   0.5 GB: 15-of-53   0.5 GB: 16-of-53   0.5 GB: 17-of-53   0.5 GB: 18-of-53   0.5 GB: 19-of-53   0.5 GB: 20-of-53   0.5 GB: 21-of-53   0.5 GB: 22-of-53   0.5 GB: 23-of-53   0.5 GB: 24-of-53   0.5 GB: 25-of-53   0.5 GB: 26-of-53   0.5 GB: 27-of-53   0.5 GB: 28-of-53   0.5 GB: 29-of-53   0.5 GB: 30-of-53   0.5 GB: 31-of-53   0.5 GB: 32-of-53   0.5 GB: 33-of-53   0.5 GB: 34-of-53   0.5 GB: 35-of-53   0.5 GB: 36-of-53   0.5 GB: 37-of-53   0.5 GB: 38-of-53   0.5 GB: 39-of-53   0.5 GB: 40-of-53   0.5 GB: 41-of-53
Model ArchitectureSkyworkForCausalLM
Licenseother
Context Length4096
Model Max Length4096
Transformers Version4.33.1
Vocabulary Size65519
Torch Data Typebfloat16

Quantized Models of the Skywork 13B Base

Model
Likes
Downloads
VRAM
Skywork 13B Base GGUF63955 GB
Skywork 13B Base GPTQ488 GB

Best Alternatives to Skywork 13B Base

Best Alternatives
Context / RAM
Downloads
Likes
Skywork 13B Math128K / 21.5 GB2410
Skywork 13B Base 8bits4K / 14.6 GB557
Skywork 13B Math 8bits4K / 14.6 GB244
... Spicyboros 3.1 3.0bpw H6 EXL2128K / 5.8 GB51
... Spicyboros 3.1 4.0bpw H6 EXL2128K / 7.5 GB51
... Spicyboros 3.1 6.0bpw H6 EXL2128K / 10.8 GB51
...kywork 13B Base 4.0bpw H6 EXL2128K / 7.5 GB51
... Airoboros Test 4.0bpw H6 EXL2128K / 7.5 GB51
... Airoboros Test 5.0bpw H6 EXL2128K / 9.1 GB51
Skywork 13B Base GPTQ128K / 8.1 GB84

Rank the Skywork 13B Base Capabilities

๐Ÿ†˜ 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! ๐ŸŒŸ

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? 51553 in total.

Our Social Media →  
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241124