What are the hardware requirements for ELYZA Japanese Llama 2 13B Fast Instruct?

ELYZA Japanese Llama 2 13B Fast Instruct requires approximately 26.4 GB of VRAM and supports a context window of 4K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed ELYZA Japanese Llama 2 13B Fast Instruct and how large is it?

ELYZA Japanese Llama 2 13B Fast Instruct is developed by elyza, a model with 13b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does ELYZA Japanese Llama 2 13B Fast Instruct perform on standard benchmarks?

ELYZA Japanese Llama 2 13B Fast Instruct has the following published scores: MMLU 54.52. Compare against reference models on this page or on the LLM Explorer leaderboards.

ELYZA Japanese Llama 2 13B Fast Instruct by elyza — VRAM 26.4GB, 4K context

Name: ELYZA Japanese Llama 2 13B Fast Instruct
Author: elyza

ELYZA Japanese Llama 2 13B Fast Instruct is an open-source language model by elyza. Features: 13b LLM, VRAM: 26.4GB, Context: 4K, License: llama2, Instruction-Based, LLM Explorer Score: 0.14, Arc: 57.5, HellaSwag: 81.8, MMLU: 54.5, GSM8K: 2.7.

Arxiv:2307.09288 Deploy:azure En Endpoints compatible Instruct Ja Llama Pytorch Region:us Sharded

Model Card on HF 🤗: https://huggingface.co/elyza/ELYZA-japanese-Llama-2-13b-fast-instruct

ELYZA Japanese Llama 2 13B Fast Instruct Parameters and Internals

Model Type

text generation, language model

Supported Languages

en (proficient), ja (proficient)

LLM Name	ELYZA Japanese Llama 2 13B Fast Instruct
Repository 🤗	https://huggingface.co/elyza/ELYZA-japanese-Llama-2-13b-fast-instruct
Model Size	13b
Required VRAM	26.4 GB
Updated	2026-05-07
Maintainer	elyza
Model Type	llama
Instruction-Based	Yes
Model Files	10.0 GB: 1-of-3 9.9 GB: 2-of-3 6.5 GB: 3-of-3
Supported Languages	ja en
Model Architecture	LlamaForCausalLM
License	llama2
Context Length	4096
Model Max Length	4096
Transformers Version	4.34.1
Tokenizer Class	LlamaTokenizer
Padding Token	</s>
Vocabulary Size	44581
Torch Data Type	bfloat16

Rank the ELYZA Japanese Llama 2 13B Fast Instruct capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Quantized Models of the ELYZA Japanese Llama 2 13B Fast Instruct

Model	Likes	Downloads	VRAM
...Llama 2 13B Fast Instruct GGUF	1	816	5 GB

Best Alternatives to ELYZA Japanese Llama 2 13B Fast Instruct

Best Alternatives	Context / RAM	Downloads	Likes
NexusRaven V2 13B	16K / 26 GB	105	472
CodeLlama 13B Instruct Hf	16K / 26 GB	2469	160
CodeLlama 13B MORepair	16K / 26 GB	9	2
CodeLlama 13B Instruct Hf	16K / 26 GB	369	29
TableLLM 13B	16K / 26 GB	98	34
NexusRaven 13B	16K / 26 GB	37	106
Panda Coder 13B	16K / 26 GB	11	14
... Llama 2 13B Instruct Text2sql	16K / 26 GB	8	28
Gen Sim	16K / 0.3 GB	4	3
Llama 3 13B Instruct Ft	8K / 26.1 GB	10	2

Note: green Score (e.g. "73.2") means that the model is better than elyza/ELYZA-japanese-Llama-2-13b-fast-instruct.