GemMoE Beta 1 by Crystalcareai

 ยป  All LLMs  ยป  Crystalcareai  ยป  GemMoE Beta 1   URL Share it on

  Autotrain compatible   Custom code   Gemmoe   Pytorch   Region:us   Sharded

GemMoE Beta 1 Benchmarks

nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").

GemMoE Beta 1 Parameters and Internals

Model Type 
Mixture of Experts (MoE)
Additional Notes 
Repo deprecated. New models are available at Hugging Face.
Training Details 
Methodology:
Custom MoE architecture tailored specifically for Gemma.
Model Architecture:
8 separately fine-tuned Gemma models, 2 experts per token.
LLM NameGemMoE Beta 1
Repository ๐Ÿค—https://huggingface.co/Crystalcareai/GemMoE-Beta-1 
Required VRAM66.3 GB
Updated2024-11-21
MaintainerCrystalcareai
Model Typegemmoe
Model Files  4.9 GB: 9-of-22   4.9 GB: 10-of-22   4.9 GB: 11-of-22   4.9 GB: 12-of-22   4.9 GB: 13-of-22   4.9 GB: 14-of-22   4.9 GB: 15-of-22   4.9 GB: 16-of-22   4.9 GB: 17-of-22   4.9 GB: 18-of-22   4.9 GB: 19-of-22   5.0 GB: 20-of-22   5.0 GB: 21-of-22   2.4 GB: 22-of-22
Model ArchitectureGemmoeForCausalLM
Licenseother
Context Length8192
Model Max Length8192
Transformers Version4.38.2
Tokenizer ClassGemmaTokenizer
Padding Token<bos>
Vocabulary Size256000
Torch Data Typebfloat16
GemMoE Beta 1 (Crystalcareai/GemMoE-Beta-1)

Rank the GemMoE Beta 1 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? 38199 in total.

Our Social Media →  
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241110