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

Nucleus 22B Token 500B by NucleusAI

 »  All LLMs  »  NucleusAI  »  Nucleus 22B Token 500B   URL Share it on

  783   

Nucleus 22B Token 500B is an open-source language model by NucleusAI. Features: 22b LLM, VRAM: 87.5GB, Context: 2K, License: mit, HF Score: 41.3, LLM Explorer Score: 0.15, Arc: 40.7, HellaSwag: 69.4, MMLU: 30.1, TruthfulQA: 39.2, WinoGrande: 67.6, GSM8K: 1.

  En   Endpoints compatible   Llama   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/NucleusAI/nucleus-22B-token-500B 

Nucleus 22B Token 500B Benchmarks

MMLU Pro: 1.8
MUSR: 3.55
BBH: 1.89
IFEval: 2.57 vs 88 (so35)-97.1%
ARC: 40.7 vs 96.7 (so35)-57.9%
HellaSwag: 69.39 vs 95.3 (gpt4)-27.2%
MMLU: 30.11 vs 88.3 (so35)-65.9%
TruthfulQA: 39.16 vs 59 (gpt4)-33.6%
WinoGrande: 67.64 vs 87.5 (gpt4)-22.7%
GSM8K: 0.99 vs 96.4 (so35)-99%
LLME Score: 0.14798
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Nucleus 22B Token 500B (NucleusAI/nucleus-22B-token-500B)
🌟 Advertise your project 🚀

Nucleus 22B Token 500B Parameters and Internals

Model Type 
Causal decoder-only
Use Cases 
Primary Use Cases:
Research on large language models, Foundation for further specialization and finetuning for specific usecases like summarization, text generation, chatbot
Considerations:
Model should be finetuned for specific tasks of interest. Precautions are needed for any production use.
Additional Notes 
This is a raw, pretrained model, which should be further finetuned for most use cases. Nucleus-22B-token-500B is trained on English data and won't generalize to other languages. It may carry online stereotypes and biases.
Supported Languages 
English (NLP)
Training Details 
Data Sources:
RefinedWeb-English, Books, Code, Technical, Math
Data Volume:
500B tokens
Methodology:
Tokenized with the tokenizer similar to Llama-7B
Context Length:
2048
Training Time:
Two weeks in early August 2023
Hardware Used:
256 A100 80GB GPUs
LLM NameNucleus 22B Token 500B
Repository 🤗https://huggingface.co/NucleusAI/nucleus-22B-token-500B 
Model Size22b
Required VRAM87.5 GB
Updated2026-03-29
MaintainerNucleusAI
Model Typellama
Model Files  9.9 GB: 1-of-9   9.7 GB: 2-of-9   9.8 GB: 3-of-9   10.0 GB: 4-of-9   9.8 GB: 5-of-9   9.9 GB: 6-of-9   9.9 GB: 7-of-9   9.8 GB: 8-of-9   8.7 GB: 9-of-9
Supported Languagesen
Model ArchitectureLlamaForCausalLM
Licensemit
Context Length2048
Model Max Length2048
Transformers Version4.30.2
Tokenizer ClassLlamaTokenizer
Beginning of Sentence Token<s>
End of Sentence Token</s>
Unk Token<unk>
Vocabulary Size32000
Torch Data Typefloat32

Quantized Models of the Nucleus 22B Token 500B

Model
Likes
Downloads
VRAM
Nucleus 22B Token 500B GGUF31439 GB
Nucleus 22B Token 500B GPTQ2611 GB
Nucleus 22B Token 500B AWQ1112 GB

Best Alternatives to Nucleus 22B Token 500B

Best Alternatives
Context / RAM
Downloads
Likes
EuroLLM 22B Instruct 251232K / 45.5 GB396962
Calm3 22B Chat16K / 44.9 GB47683
Calm3 22B RP V216K / 44.9 GB3120
Calm3 22B RP V0.116K / 44.9 GB50
Yousei 22B4K / 44.5 GB6872
Llama2 22B Daydreamer V34K / 43.7 GB79211
Llama2 22B4K / 43.7 GB110146
Platypus 2 22B Relora4K / 43.7 GB6101
Llama2 22B Blocktriangular4K / 43.7 GB8344
Llama2 22B Daydreamer V24K / 43.7 GB42
Note: green Score (e.g. "73.2") means that the model is better than NucleusAI/nucleus-22B-token-500B.

Rank the Nucleus 22B Token 500B 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? 52721 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.
Check out Ag3ntum — our secure, self-hosted AI agent for server management.
Release v20260328a
  Support LLM Explorer