LLM EXPLORER 65,548 MODELS INDEXED

EvoLLM JP V1 10B by SakanaAI

By SakanaAI · 74 downloads
HF Repo ↗ Copy URL

EvoLLM JP V1 10B is an open-source language model by SakanaAI. Features: 10b LLM, VRAM: 19.7GB, Context: 32K, License: other, LLM Explorer Score: 0.11.

  Arxiv:2403.13187   Custom code   Evomistral   Ja   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/SakanaAI/EvoLLM-JP-v1-10B 

EvoLLM JP V1 10B Parameters and Internals

Model Type 
Autoregressive Language Model
Additional Notes 
EvoLLM-JP-v1-10B is an experimental general-purpose Japanese LLM created using the Evolutionary Model Merge method by merging models Shisa Gamma 7B v1, WizardMath 7B V1.1, and Abel 7B 002.
Supported Languages 
Japanese (native)
Training Details 
Methodology:
Evolutionary Model Merge
Input Output 
Input Format:
Messages structured in roles and content
Accepted Modalities:
text
Output Format:
Text
LLM NameEvoLLM JP V1 10B
Repository 🤗https://huggingface.co/SakanaAI/EvoLLM-JP-v1-10B 
Model Size10b
Required VRAM19.7 GB
Updated2026-07-31
MaintainerSakanaAI
Model Typeevomistral
Model Files  4.9 GB: 1-of-4   5.0 GB: 2-of-4   4.9 GB: 3-of-4   4.9 GB: 4-of-4
Supported Languagesja
Model ArchitectureEvoMistralForCausalLM
Licenseother
Context Length32768
Model Max Length32768
Transformers Version4.38.2
Tokenizer ClassLlamaTokenizer
Vocabulary Size32000
Torch Data Typebfloat16