LLM EXPLORER 62,696 MODELS INDEXED

K EXAONE 152B A23B REAP 80E BF16 V2 by Baekpica

By Baekpica · 20 downloads
HF Repo ↗ Copy URL

K EXAONE 152B A23B REAP 80E BF16 V2 is an open-source language model by Baekpica. Features: 152b LLM, VRAM: 302.9GB, Context: 256K, License: other, LLM Explorer Score: 0.24.

Base model:baekpica/k-exaone-1... Base model:finetune:baekpica/k...   Conversational   De   En   Endpoints compatible   Es   Exaone   Exaone moe   Ja   K-exaone   Ko   Lg-ai   Moe   Post-training   Pruning   Reap   Region:us   Safetensors   Sharded   Tensorflow   Vi
Model Card on HF 🤗: https://huggingface.co/Baekpica/K-EXAONE-152B-A23B-REAP-80E-BF16-v2 

K EXAONE 152B A23B REAP 80E BF16 V2 Parameters and Internals

LLM NameK EXAONE 152B A23B REAP 80E BF16 V2
Repository 🤗https://huggingface.co/Baekpica/K-EXAONE-152B-A23B-REAP-80E-BF16-v2 
Base Model(s)  Baekpica/K-EXAONE-152B-A23B-REAP-80E-BF16   Baekpica/K-EXAONE-152B-A23B-REAP-80E-BF16
Model Size152b
Required VRAM302.9 GB
Updated2026-10-01
MaintainerBaekpica
Model Typeexaone_moe
Model Files  50.0 GB: 1-of-7   50.0 GB: 2-of-7   50.0 GB: 3-of-7   50.0 GB: 4-of-7   50.0 GB: 5-of-7   50.0 GB: 6-of-7   2.9 GB: 7-of-7   1.1 GB
Supported Languagesen ko es de ja vi
Model ArchitectureExaoneMoEForCausalLM
Licenseother
Context Length262144
Model Max Length262144
Transformers Version5.13.0
Tokenizer ClassGPT2Tokenizer
Padding Token[PAD]
Vocabulary Size153600
Errorsreplace