LLM EXPLORER 59,358 MODELS INDEXED

DeepSeek V3.0324 MoE Pruner E192 Bf16 by tflsxyy

By tflsxyy · 6 downloads
HF Repo ↗ Copy URL

DeepSeek V3.0324 MoE Pruner E192 Bf16 is an open-source language model by tflsxyy. Features: 521b LLM, VRAM: 285.4GB, Context: 160K, License: mit, MoE, LLM Explorer Score: 0.17.

  Arxiv:2412.19437   Conversational   Custom code   Deepseek v3   Endpoints compatible   Moe   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/tflsxyy/DeepSeek-V3-0324-MoE-Pruner-E192-bf16 

DeepSeek V3.0324 MoE Pruner E192 Bf16 Parameters and Internals

LLM NameDeepSeek V3.0324 MoE Pruner E192 Bf16
Repository 🤗https://huggingface.co/tflsxyy/DeepSeek-V3-0324-MoE-Pruner-E192-bf16 
Model Size521b
Required VRAM285.4 GB
Updated2026-07-13
Maintainertflsxyy
Model Typedeepseek_v3
Model Files  8.6 GB: 1-of-163   7.5 GB: 2-of-163   4.9 GB: 3-of-163   7.7 GB: 4-of-163   7.3 GB: 5-of-163   4.4 GB: 6-of-163   8.3 GB: 7-of-163   5.8 GB: 8-of-163   6.0 GB: 9-of-163   7.0 GB: 10-of-163   5.3 GB: 11-of-163   1.9 GB: 12-of-163   7.3 GB: 13-of-163   6.5 GB: 14-of-163   6.4 GB: 15-of-163   7.5 GB: 16-of-163   5.7 GB: 17-of-163   6.8 GB: 18-of-163   6.3 GB: 19-of-163   5.5 GB: 20-of-163   8.2 GB: 21-of-163   6.2 GB: 22-of-163   5.6 GB: 23-of-163   7.3 GB: 24-of-163   6.0 GB: 25-of-163   6.6 GB: 26-of-163   6.8 GB: 27-of-163   5.5 GB: 28-of-163   7.1 GB: 29-of-163   6.3 GB: 30-of-163   6.0 GB: 31-of-163   7.1 GB: 32-of-163   6.0 GB: 33-of-163   2.4 GB: 34-of-163   8.0 GB: 35-of-163   5.7 GB: 36-of-163   6.1 GB: 37-of-163   7.2 GB: 38-of-163   6.1 GB: 39-of-163   5.8 GB: 40-of-163   6.6 GB: 41-of-163   6.4 GB: 42-of-163   7.1 GB: 43-of-163   6.4 GB: 44-of-163   6.2 GB: 45-of-163
Model ArchitectureDeepseekV3ForCausalLM
Licensemit
Context Length163840
Model Max Length163840
Transformers Version4.46.3
Vocabulary Size129280
Torch Data Typebfloat16