What are the hardware requirements for Gemma 2 Baku 2B?

Gemma 2 Baku 2B requires approximately 10.5 GB of VRAM and supports a context window of 8K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Gemma 2 Baku 2B and how large is it?

Gemma 2 Baku 2B is developed by rinna, a model with 2b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Gemma 2 Baku 2B?

Gemma 2 Baku 2B is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Gemma 2 Baku 2B by rinna — VRAM 10.5GB, 8K context

Name: Gemma 2 Baku 2B
Author: rinna

Gemma 2 Baku 2B is an open-source language model by rinna. Features: 2b LLM, VRAM: 10.5GB, Context: 8K, License: gemma, LLM Explorer Score: 0.14.

Arxiv:2404.01657 Base model:finetune:google/gem... Base model:google/gemma-2-2b Dataset:cc100 Dataset:eleutherai/pile Dataset:mc4 Dataset:oscar-corpus/colossal-... Dataset:wikipedia En Gemma2 Ja Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/rinna/gemma-2-baku-2b

Gemma 2 Baku 2B Parameters and Internals

Model Type

text-generation

Supported Languages

ja (advanced), en (intermediate)

Training Details

Data Sources:

Japanese CC-100, Japanese C4, Japanese OSCAR, The Pile, Wikipedia, rinna curated Japanese dataset

Data Volume:

80B tokens

Model Architecture:

A 26-layer, 2304-hidden-size transformer-based language model

Input Output

Accepted Modalities:

text

Output Format:

generated text

Performance Tips:

Recommended to use eager attention when conducting batch inference under bfloat16 precision. Gemma 2 yields NaN values with padding sequences when default attention mechanism is employed with bfloat16.

LLM Name	Gemma 2 Baku 2B
Repository 🤗	https://huggingface.co/rinna/gemma-2-baku-2b
Base Model(s)	Gemma 2 2B google/gemma-2-2b
Model Size	2b
Required VRAM	10.5 GB
Updated	2026-07-21
Maintainer	rinna
Model Type	gemma2
Model Files	5.0 GB: 1-of-3 5.0 GB: 2-of-3 0.5 GB: 3-of-3
Supported Languages	ja en
Model Architecture	AutoModelForCausalLM
License	gemma
Context Length	8192
Model Max Length	8192
Transformers Version	4.42.4
Tokenizer Class	GemmaTokenizer
Padding Token	<pad>
Vocabulary Size	256000
Torch Data Type	float32

Rank the Gemma 2 Baku 2B 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

Best Alternatives to Gemma 2 Baku 2B

Best Alternatives	Context / RAM	Downloads	Likes
Gemma 2B It	0K / 5.1 GB	7	0
Caca Tinny 2B V2	0K / 5.1 GB	5	0
Gemma 2 2B It Chat Lessons	0K / 5.2 GB	6	0
... 2B It Vdm Sq4 Car Motion Beta	0K / 5.2 GB	15	0
Gemma 2 2B It Chat Doctor	0K / 5.2 GB	9	0
Gemma 2 2B It Chat Doctor	0K / 5.2 GB	16	0
Gemma 2 2B It Heisenberg	0K / 5.2 GB	8	0
... It Finetuned Mental Health Qa	0K / 5.1 GB	21	1
...ma 2 2B It Game Recommendation	0K / 5.1 GB	14	0
...emma 2B Lora Lbox Ljp Modified	0K / 10 GB	17	0

Note: green Score (e.g. "73.2") means that the model is better than rinna/gemma-2-baku-2b.