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

Magic Dolphin 7B by InferenceIllusionist

 »  All LLMs  »  InferenceIllusionist  »  Magic Dolphin 7B   URL Share it on

  98   

Magic Dolphin 7B is an open-source language model by InferenceIllusionist. Features: 7b LLM, VRAM: 14.4GB, Context: 32K, License: apache-2.0, Merged, HF Score: 67.5, LLM Explorer Score: 0.14, Arc: 65.8, HellaSwag: 85.6, MMLU: 64.6, TruthfulQA: 58, WinoGrande: 79.6, GSM8K: 51.2.

  Merged Model   Arxiv:2203.05482 Base model:dphn/dolphin-2.6-mi... Base model:ibm-research/merlin... Base model:locutusque/hyperion...   Code   Endpoints compatible   Mistral   Model-index   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/InferenceIllusionist/Magic-Dolphin-7b 

Magic Dolphin 7B Benchmarks

ARC: 65.78 vs 96.7 (so35)-32%
HellaSwag: 85.61 vs 95.3 (gpt4)-10.2%
MMLU: 64.64 vs 88.3 (so35)-26.8%
TruthfulQA: 58.01 vs 59 (gpt4)-1.7%
WinoGrande: 79.64 vs 87.5 (gpt4)-9%
GSM8K: 51.18 vs 96.4 (so35)-46.9%
LLME Score: 0.13585
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Magic Dolphin 7B (InferenceIllusionist/Magic-Dolphin-7b)
🌟 Advertise your project 🚀

Magic Dolphin 7B Parameters and Internals

Model Type 
text-generation
Additional Notes 
The model is a linear merge of three base models and is tested for LAB tuning.
LLM NameMagic Dolphin 7B
Repository 🤗https://huggingface.co/InferenceIllusionist/Magic-Dolphin-7b 
Base Model(s)  cognitivecomputations/dolphin-2.6-mistral-7b-dpo-laser   Locutusque/Hyperion-1.5-Mistral-7B   Merlinite 7B   cognitivecomputations/dolphin-2.6-mistral-7b-dpo-laser   Locutusque/Hyperion-1.5-Mistral-7B   ibm/merlinite-7b
Merged ModelYes
Model Size7b
Required VRAM14.4 GB
Updated2026-04-24
MaintainerInferenceIllusionist
Model Typemistral
Model Files  9.9 GB: 1-of-2   4.5 GB: 2-of-2
Model ArchitectureMistralForCausalLM
Licenseapache-2.0
Context Length32768
Model Max Length32768
Transformers Version4.37.1
Tokenizer ClassLlamaTokenizer
Padding Token</s>
Vocabulary Size32000
Torch Data Typefloat16

Best Alternatives to Magic Dolphin 7B

Best Alternatives
Context / RAM
Downloads
Likes
...Nemo Instruct 2407 Abliterated1000K / 24.5 GB25420
MegaBeam Mistral 7B 512K512K / 14.4 GB1144354
SpydazWeb AI HumanAI RP512K / 14.4 GB141
SpydazWeb AI HumanAI 002512K / 14.4 GB181
...daz Web AI ChatML 512K Project512K / 14.5 GB120
MegaBeam Mistral 7B 300K282K / 14.4 GB377916
MegaBeam Mistral 7B 300K282K / 14.4 GB796816
Hebrew Mistral 7B 200K256K / 30 GB131615
Astral 256K 7B V2250K / 14.4 GB60
Astral 256K 7B250K / 14.4 GB50
Note: green Score (e.g. "73.2") means that the model is better than InferenceIllusionist/Magic-Dolphin-7b.

Rank the Magic Dolphin 7B 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? 53185 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