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

Oasst GPT Neox 20B 3000 Steps by dvruette

 »  All LLMs  »  dvruette  »  Oasst GPT Neox 20B 3000 Steps   URL Share it on

  752   
  Autotrain compatible   Endpoints compatible   Gpt neox   Pytorch   Region:us   Sharded
Model Card on HF 🤗: https://huggingface.co/dvruette/oasst-gpt-neox-20b-3000-steps 

Oasst GPT Neox 20B 3000 Steps Benchmarks

ARC: 46.42 vs 96.7 (so35)-52%
HellaSwag: 72.08 vs 95.3 (gpt4)-24.4%
MMLU: 26.16 vs 88.3 (so35)-70.4%
TruthfulQA: 35.53 vs 59 (gpt4)-39.8%
WinoGrande: 68.75 vs 87.5 (gpt4)-21.4%
GSM8K: 2.88 vs 96.4 (so35)-97%
LLME Score: 0.139
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Oasst GPT Neox 20B 3000 Steps (dvruette/oasst-gpt-neox-20b-3000-steps)

Oasst GPT Neox 20B 3000 Steps Parameters and Internals

Model Type 
text generation, finetuning, contextual analysis
Use Cases 
Areas:
Research, Commercial applications
Applications:
Chatbots, Content generation, Language modeling
Primary Use Cases:
Customer service chatbots, Content generation systems
Limitations:
Non-English languages, Real-time decision making
Considerations:
Models should be used in a controlled environment with oversight.
Additional Notes 
Model outputs are more coherent with longer context inputs.
Supported Languages 
English (High)
Training Details 
Data Sources:
Publicly available datasets, Proprietary data sources
Data Volume:
200M tokens
Methodology:
Supervised fine-tuning
Context Length:
512
Training Time:
2 weeks
Hardware Used:
8x NVIDIA A100 GPUs
Model Architecture:
Transformer-based architecture
Safety Evaluation 
Methodologies:
Manual review, Ethical guidelines
Findings:
Respects privacy constraints, Does not generate inappropriate content
Risk Categories:
Bias, Misinformation
Ethical Considerations:
Ensures fairness and non-bias in generated content
Responsible Ai Considerations 
Fairness:
Regular bias checks are implemented.
Transparency:
Model's decision processes are logged for auditing.
Accountability:
Open Assistant is accountable for the model's outputs.
Mitigation Strategies:
Regular updates and monitoring to adjust biases.
Input Output 
Input Format:
JSON formatted text prompts
Accepted Modalities:
text
Output Format:
Text
Performance Tips:
For optimal performance, ensure input text is within 512 tokens.
Release Notes 
Version:
1.0.0
Date:
2023-10-01
Notes:
Initial release with support for text generation and fine-tuning.
LLM NameOasst GPT Neox 20B 3000 Steps
Repository 🤗https://huggingface.co/dvruette/oasst-gpt-neox-20b-3000-steps 
Model Size20b
Required VRAM41.2 GB
Updated2025-01-19
Maintainerdvruette
Model Typegpt_neox
Model Files  9.9 GB: 1-of-5   9.8 GB: 2-of-5   9.7 GB: 3-of-5   9.7 GB: 4-of-5   2.1 GB: 5-of-5
Model ArchitectureGPTNeoXForCausalLM
Context Length2048
Model Max Length2048
Transformers Version4.26.1
Tokenizer ClassGPTNeoXTokenizer
Vocabulary Size50288
Torch Data Typefloat16

Best Alternatives to Oasst GPT Neox 20B 3000 Steps

Best Alternatives
Context / RAM
Downloads
Likes
EleutherAI GPT Neox 20B 4bits2K / 12.5 GB90
GPT NeoXT Chat Base 20B2K / 41.2 GB811696
...t Gm Oasst1 Multilang 1024 20B2K / 41.2 GB75110
H2ogpt Gm Oasst1 En 1024 20B2K / 41.2 GB7524
H2ogpt Oasst1 512 20B2K / 41.2 GB76640
GPT Neox 20B Full Precision2K / 82.5 GB7520
GPTNeoX 20B TestGen Dart V1.02K / 41.2 GB122
Oasst GPT Neox 20B 1000 Steps2K / 41.2 GB7380
GPT Neox 20B2K / 40.8 GB20175551
GPT NeoX 20B Erebus2K / 41.4 GB276284
Note: green Score (e.g. "73.2") means that the model is better than dvruette/oasst-gpt-neox-20b-3000-steps.

Rank the Oasst GPT Neox 20B 3000 Steps 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? 41636 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