LLM EXPLORER 65,695 MODELS INDEXED

DeepSeek V4.1 Flash 1017 EXL3 2.5bpw by bertholomus

By bertholomus · 209 downloads

DeepSeek V4.1 Flash 1017 EXL3 2.5bpw is an open-source language model by bertholomus. Features: 92.3b LLM, VRAM: 183.4GB, License: mit.

Base model:deepseek-ai/deepsee... Base model:quantized:deepseek-...   Deepseek   Deepseek v41   Dgx-spark   Exl3   Exllamav3   Gb10   Image-text-to-text   Region:us   Safetensors   Sharded   Tensorflow   Tensorfold

DeepSeek V4.1 Flash 1017 EXL3 2.5bpw Parameters and Internals

LLM NameDeepSeek V4.1 Flash 1017 EXL3 2.5bpw
Repository 🤗https://huggingface.co/bertholomus/DeepSeek-V4.1-Flash-1017-EXL3-2.5bpw 
Base Model(s)  deepseek-ai/DeepSeek-V4.1-Flash-1017   deepseek-ai/DeepSeek-V4.1-Flash-1017
Model Size92.3b
Required VRAM183.4 GB
Updated2026-10-27
Maintainerbertholomus
Model Typedeepseek_v41
Model Files  5.8 GB: 1-of-41   4.4 GB: 2-of-41   4.4 GB: 3-of-41   4.3 GB: 4-of-41   4.3 GB: 5-of-41   4.3 GB: 6-of-41   4.3 GB: 7-of-41   4.3 GB: 8-of-41   4.4 GB: 9-of-41   4.3 GB: 10-of-41   4.3 GB: 11-of-41   4.3 GB: 12-of-41   4.3 GB: 13-of-41   4.4 GB: 14-of-41   4.4 GB: 15-of-41   4.3 GB: 16-of-41   4.3 GB: 17-of-41   4.3 GB: 18-of-41   4.3 GB: 19-of-41   4.3 GB: 20-of-41   4.4 GB: 21-of-41   4.3 GB: 22-of-41   4.3 GB: 23-of-41   4.3 GB: 24-of-41   4.4 GB: 25-of-41   4.3 GB: 26-of-41   4.3 GB: 27-of-41   4.3 GB: 28-of-41   4.4 GB: 29-of-41   4.3 GB: 30-of-41   4.3 GB: 31-of-41   4.3 GB: 32-of-41   4.4 GB: 33-of-41   4.3 GB: 34-of-41   4.3 GB: 35-of-41   4.3 GB: 36-of-41   4.4 GB: 37-of-41   4.3 GB: 38-of-41   4.3 GB: 39-of-41   7.3 GB: 40-of-41   5.9 GB: 41-of-41
Model ArchitectureDeepseekV41ForCausalLM
Licensemit
Model Max Length1048576
Transformers Version5.6.0
Tokenizer ClassPreTrainedTokenizerFast
Beginning of Sentence Token<|begin▁of▁sentence|>
End of Sentence Token<|end▁of▁sentence|>