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

Dzakwan MoE 4x7b Beta by dzakwan

 »  All LLMs  »  dzakwan  »  Dzakwan MoE 4x7b Beta   URL Share it on

  9072   

Dzakwan MoE 4x7b Beta is an open-source language model by dzakwan. Features: 24.2b LLM, VRAM: 48.4GB, Context: 32K, License: apache-2.0, MoE, HF Score: 75.8, LLM Explorer Score: 0.24, Arc: 72.1, HellaSwag: 88.9, MMLU: 64.9, TruthfulQA: 74.5, WinoGrande: 83.5, GSM8K: 70.8.

  Autotrain compatible Base model:beowolx/codeninja-1... Base model:merge:beowolx/coden... Base model:merge:mlabonne/alph... Base model:merge:mlabonne/neur... Base model:merge:sanjiwatsuki/... Base model:mlabonne/alphamonar... Base model:mlabonne/neuraldare... Base model:sanjiwatsuki/kunoic... Beowolx/codeninja-1.0-openchat...   Conversational   Endpoints compatible   Frankenmoe   Lazymergekit   Merge   Mergekit   Mixtral   Mlabonne/alphamonarch-7b   Mlabonne/neuraldaredevil-7b   Moe   Region:us   Safetensors Sanjiwatsuki/kunoichi-dpo-v2-7...   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/dzakwan/dzakwan-MoE-4x7b-Beta 

Dzakwan MoE 4x7b Beta Benchmarks

MMLU Pro: 23.42
GPQA: 4.81
MUSR: 12.11
BBH: 32.07
IFEval: 44.43 vs 88 (so35)-49.5%
ARC: 72.1 vs 96.7 (so35)-25.4%
HellaSwag: 88.87 vs 95.3 (gpt4)-6.7%
MMLU: 64.89 vs 88.3 (so35)-26.5%
TruthfulQA: 74.51 vs 59 (gpt4)26.3%
WinoGrande: 83.5 vs 87.5 (gpt4)-4.6%
GSM8K: 70.81 vs 96.4 (so35)-26.5%
MATH Lvl 5: 7.78
LLME Score: 0.23945
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Dzakwan MoE 4x7b Beta (dzakwan/dzakwan-MoE-4x7b-Beta)
🌟 Advertise your project 🚀

Dzakwan MoE 4x7b Beta Parameters and Internals

Model Type 
text generation
Use Cases 
Areas:
research, commercial applications
Applications:
text generation, chatbots, programming assistance, storywriting
Primary Use Cases:
text-based assistant, code generation, creative writing
Training Details 
Methodology:
Mixture of Experts using LazyMergekit
Model Architecture:
Mixture of Experts composed of multiple models
Input Output 
Input Format:
chat-like prompt with roles
Accepted Modalities:
text
Output Format:
text
LLM NameDzakwan MoE 4x7b Beta
Repository 🤗https://huggingface.co/dzakwan/dzakwan-MoE-4x7b-Beta 
Base Model(s)  AlphaMonarch 7B   beowolx/CodeNinja-1.0-OpenChat-7B   Kunoichi DPO V2 7B   NeuralDaredevil 7B   mlabonne/AlphaMonarch-7B   beowolx/CodeNinja-1.0-OpenChat-7B   SanjiWatsuki/Kunoichi-DPO-v2-7B   mlabonne/NeuralDaredevil-7B
Model Size24.2b
Required VRAM48.4 GB
Updated2025-09-23
Maintainerdzakwan
Model Typemixtral
Model Files  1.9 GB: 1-of-25   2.0 GB: 2-of-25   2.0 GB: 3-of-25   2.0 GB: 4-of-25   2.0 GB: 5-of-25   2.0 GB: 6-of-25   2.0 GB: 7-of-25   2.0 GB: 8-of-25   2.0 GB: 9-of-25   2.0 GB: 10-of-25   2.0 GB: 11-of-25   2.0 GB: 12-of-25   2.0 GB: 13-of-25   2.0 GB: 14-of-25   2.0 GB: 15-of-25   2.0 GB: 16-of-25   2.0 GB: 17-of-25   2.0 GB: 18-of-25   2.0 GB: 19-of-25   2.0 GB: 20-of-25   2.0 GB: 21-of-25   2.0 GB: 22-of-25   2.0 GB: 23-of-25   2.0 GB: 24-of-25   0.5 GB: 25-of-25
Model ArchitectureMixtralForCausalLM
Licenseapache-2.0
Context Length32768
Model Max Length32768
Transformers Version4.41.1
Tokenizer ClassLlamaTokenizer
Padding Token<s>
Vocabulary Size32000
Torch Data Typefloat16

Best Alternatives to Dzakwan MoE 4x7b Beta

Best Alternatives
Context / RAM
Downloads
Likes
Beyonder 4x7B V332K / 48.3 GB785660
Calme 4x7B MoE V0.232K / 48.3 GB81272
Mera Mix 4x7B32K / 48.3 GB815719
Calme 4x7B MoE V0.132K / 48.3 GB80032
MixtureofMerges MoE 4x7b V532K / 48.3 GB81171
MixtureofMerges MoE 4x7b V432K / 48.3 GB90684
CognitiveFusion2 4x7B BF1632K / 48.3 GB81853
Proto Athena 4x7B32K / 48.4 GB50
Proto Athena V0.2 4x7B32K / 48.4 GB60
NeuralStar FusionWriter 4x7b32K / 48.3 GB249

Rank the Dzakwan MoE 4x7b Beta 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