LLM EXPLORER 59,358 MODELS INDEXED

Defne Llama3 2x8B by Eurdem

By Eurdem · 15 downloads
HF Repo ↗ Copy URL

Defne Llama3 2x8B is an open-source language model by Eurdem. Features: 13.7b LLM, VRAM: 27.4GB, Context: 8K, License: llama3, MoE, Instruction-Based, LLM Explorer Score: 0.13, Arc: 70.2, HellaSwag: 86.1, MMLU: 68.2, GSM8K: 67.7.

  Conversational   Deploy:azure   En   Endpoints compatible   Instruct   Llama-3   Merge   Mixtral   Moe   Region:us   Safetensors   Sharded   Tensorflow   Tr
Model Card on HF 🤗: https://huggingface.co/Eurdem/Defne_llama3_2x8B 

Defne Llama3 2x8B Parameters and Internals

Model Type 
text-generation, Mixure of Experts (MoE)
Supported Languages 
en (proficient), tr (proficient)
Input Output 
Accepted Modalities:
text
Output Format:
text
LLM NameDefne Llama3 2x8B
Repository 🤗https://huggingface.co/Eurdem/Defne_llama3_2x8B 
Model Size13.7b
Required VRAM27.4 GB
Updated2026-05-18
MaintainerEurdem
Model Typemixtral
Instruction-BasedYes
Model Files  10.0 GB: 1-of-3   10.0 GB: 2-of-3   7.4 GB: 3-of-3
Supported Languagesen tr
Model ArchitectureMixtralForCausalLM
Licensellama3
Context Length8192
Model Max Length8192
Transformers Version4.40.2
Tokenizer ClassPreTrainedTokenizerFast
Padding Token<|begin_of_text|>
Vocabulary Size128256
Torch Data Typebfloat16

Best Alternatives to Defne Llama3 2x8B

Best Alternatives
Context / RAM
Downloads
Likes
L3.1 Celestial Stone 2x8B128K / 27.3 GB3123
...ma 3 2x8B Instruct MoE 64K Ctx64K / 27.3 GB124
Inixion 2x8B V28K / 27.4 GB62
MoE Llama3 8bx2 Rag8K / 27.3 GB70
Inixion 2x8B8K / 27.5 GB91
Llama 3 Chatty 2x8B8K / 27.3 GB611
FinalFintetuning XVIII 2x8B8K / 27.5 GB114
Llama 3 Teal Instruct 2x8B MoE8K / 27.3 GB80
Llama 3 8Bx2 MoE DPO8K / 27.4 GB41
...lama3 2x8b MoE 41K Experiment18K / 27.3 GB52
Note: green Score (e.g. "73.2") means that the model is better than Eurdem/Defne_llama3_2x8B.