LLM EXPLORER 59,962 MODELS INDEXED

Elly 768 MoE 4bit by assassindesign

By assassindesign · 84 downloads

Elly 768 MoE 4bit is an open-source language model by assassindesign. Features: 31m LLM, VRAM: 0.1GB, Context: 32K, MoE, Quantized, LLM Explorer Score: 0.3.

  4-bit   4bit   Children   Chinese   Companion   Conversational   En   Mlx   Moe   Quantized   Qwen3   Qwen3 moe   Region:us   Safetensors   Zh

Elly 768 MoE 4bit Parameters and Internals

LLM NameElly 768 MoE 4bit
Repository πŸ€—https://huggingface.co/assassindesign/Elly-768-MoE-4bit 
Base Model(s)  minimind   /minimind
Model Size31m
Required VRAM0.1 GB
Updated2026-08-30
Maintainerassassindesign
Model Typeqwen3_moe
Model Files  0.1 GB
Supported Languageszh en
Quantization Type4bit
Model ArchitectureQwen3MoeForCausalLM
Context Length32768
Model Max Length32768
Transformers Version5.8.1
Tokenizer ClassTokenizersBackend
Padding Token<|endoftext|>
Vocabulary Size6400