LLM EXPLORER 59,420 MODELS INDEXED

RakutenAI 7B Instruct by Rakuten

By Rakuten · 251 downloads
HF Repo ↗ Copy URL

RakutenAI 7B Instruct is an open-source language model by Rakuten. Features: 7b LLM, VRAM: 14.8GB, Context: 32K, License: apache-2.0, Instruction-Based, LLM Explorer Score: 0.12.

  Arxiv:2403.15484 Base model:finetune:rakuten/ra... Base model:rakuten/rakutenai-7...   Deploy:azure   En   Endpoints compatible   Instruct   Ja   Mistral   Pytorch   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/Rakuten/RakutenAI-7B-instruct 

RakutenAI 7B Instruct Parameters and Internals

Model Type 
Instruction-based
Use Cases 
Areas:
Research, Commercial applications
Primary Use Cases:
Japanese language understanding, English language tasks
Limitations:
Can produce biased, inaccurate, or unsafe outputs
Considerations:
Exercise caution and judgment while interacting with the model
Additional Notes 
The model's vocabulary was increased to improve Japanese token representation.
Supported Languages 
Japanese (Excellent), English (Good)
Training Details 
Data Sources:
JSNLI, RTE, KUCI, BELEBELE, JCS, JNLI, Dolly-15K, OpenAssistant1
Methodology:
Fine-tuned using open source and internally hand-crafted datasets
Model Architecture:
Usage of Mistral model architecture with extended vocabulary from 32k to 48k
LLM NameRakutenAI 7B Instruct
Repository 🤗https://huggingface.co/Rakuten/RakutenAI-7B-instruct 
Base Model(s)  RakutenAI 7B   Rakuten/RakutenAI-7B
Model Size7b
Required VRAM14.8 GB
Updated2026-07-20
MaintainerRakuten
Model Typemistral
Instruction-BasedYes
Model Files  10.0 GB: 1-of-2   4.8 GB: 2-of-2   10.0 GB: 1-of-2   4.8 GB: 2-of-2
Supported Languagesen ja
Model ArchitectureMistralForCausalLM
Licenseapache-2.0
Context Length32768
Model Max Length32768
Transformers Version4.34.0
Tokenizer ClassLlamaTokenizer
Padding Token</s>
Vocabulary Size48000
Torch Data Typebfloat16

Best Alternatives to RakutenAI 7B Instruct

Best Alternatives
Context / RAM
Downloads
Likes
...Nemo Instruct 2407 Abliterated1000K / 24.5 GB22020
SpydazWeb AI HumanAI RP512K / 14.4 GB161
SpydazWeb AI HumanAI 002512K / 14.4 GB181
...daz Web AI ChatML 512K Project512K / 14.5 GB120
... Summarize 64K QLoRANET Merged128K / 4.1 GB100
...1 Summarize 64K LoRANET Merged128K / 14.4 GB130
Mistral 7B Instruct V0.232K / 14.4 GB11153373184
Mistral 7B Instruct V0.132K / 14.4 GB2536291846
Mistral 7B Instruct V0.232K / 14.4 GB13380
Lexi Coder V5.132K / 14.5 GB1970
Note: green Score (e.g. "73.2") means that the model is better than Rakuten/RakutenAI-7B-instruct.