LLM EXPLORER 59,420 MODELS INDEXED

L3 MoE 16x1B RP Abliterated Test Bad Quality by Disya

By Disya · 13 downloads
HF Repo ↗ Copy URL

L3 MoE 16x1B RP Abliterated Test Bad Quality is an open-source language model by Disya. Features: 13.6b LLM, VRAM: 27.6GB, Context: 128K, License: llama3.2, MoE, LLM Explorer Score: 0.17.

  Conversational   En   Llama   Mergekit   Mixtral   Mixture of experts   Moe   Not-for-all-audiences   Region:us   Roleplay   Rp   Safetensors   Sharded   Story   Tensorflow   Uncensored   Writer
Model Card on HF 🤗: https://huggingface.co/Disya/L3-MOE-16x1B-RP-Abliterated-Test-bad-quality 

L3 MoE 16x1B RP Abliterated Test Bad Quality Parameters and Internals

LLM NameL3 MoE 16x1B RP Abliterated Test Bad Quality
Repository 🤗https://huggingface.co/Disya/L3-MOE-16x1B-RP-Abliterated-Test-bad-quality 
Model Size13.6b
Required VRAM27.6 GB
Updated2025-04-03
MaintainerDisya
Model Typemixtral
Model Files  1.0 GB: 1-of-28   1.0 GB: 2-of-28   1.0 GB: 3-of-28   1.0 GB: 4-of-28   1.0 GB: 5-of-28   1.0 GB: 6-of-28   1.0 GB: 7-of-28   1.0 GB: 8-of-28   1.0 GB: 9-of-28   1.0 GB: 10-of-28   1.0 GB: 11-of-28   1.0 GB: 12-of-28   1.0 GB: 13-of-28   1.0 GB: 14-of-28   1.0 GB: 15-of-28   1.0 GB: 16-of-28   1.0 GB: 17-of-28   1.0 GB: 18-of-28   1.0 GB: 19-of-28   1.0 GB: 20-of-28   1.0 GB: 21-of-28   1.0 GB: 22-of-28   1.0 GB: 23-of-28   1.0 GB: 24-of-28   1.0 GB: 25-of-28   1.0 GB: 26-of-28   1.0 GB: 27-of-28   0.6 GB: 28-of-28
Supported Languagesen
Model ArchitectureMixtralForCausalLM
Licensellama3.2
Context Length131072
Model Max Length131072
Transformers Version4.49.0
Tokenizer ClassPreTrainedTokenizer
Padding Token<|begin_of_text|>
Vocabulary Size128256
Torch Data Typebfloat16