What are the hardware requirements for Japanese Mistral 300M Base?

Japanese Mistral 300M Base requires approximately 2.8 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 Japanese Mistral 300M Base and how large is it?

Japanese Mistral 300M Base is developed by ce-lery, a model with 300m 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 Japanese Mistral 300M Base?

Japanese Mistral 300M Base is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Japanese Mistral 300M Base by ce-lery — VRAM 2.8GB, 4K context

Name: Japanese Mistral 300M Base
Author: ce-lery

Japanese Mistral 300M Base is an open-source language model by ce-lery. Features: 300m LLM, VRAM: 2.8GB, Context: 4K, Quantized, LLM Explorer Score: 0.1.

Endpoints compatible F32 Generated from trainer Ggml Gguf Mistral Quantized Region:us Safetensors Tensorboard

Model Card on HF 🤗: https://huggingface.co/ce-lery/japanese-mistral-300m-base

Japanese Mistral 300M Base Parameters and Internals

Model Type

text generation

Additional Notes

Suppression of unknown word generation by using byte fallback in SentencePiece tokenizer and conversion to huggingface Tokenizers format.

Supported Languages

Japanese (fluent)

Training Details

Data Sources:

wikipedia, cc100

Methodology:

Pretraining with flash attention2 and torch.compile and DeepSpeed, Fine-tuning with databricks-dolly-15k-ja

Input Output

Input Format:

tokenized input

Accepted Modalities:

text

Output Format:

tokenized output

LLM Name	Japanese Mistral 300M Base
Repository 🤗	https://huggingface.co/ce-lery/japanese-mistral-300m-base
Base Model(s)	None /None
Model Size	300m
Required VRAM	2.8 GB
Updated	2026-07-07
Maintainer	ce-lery
Model Type	mistral
Model Files	1.4 GB 1.4 GB 0.0 GB
GGML Quantization	Yes
GGUF Quantization	Yes
Quantization Type	ggml\|gguf
Model Architecture	MistralForCausalLM
Context Length	4096
Model Max Length	4096
Transformers Version	4.35.2
Tokenizer Class	T5Tokenizer
Padding Token	[PAD]
Vocabulary Size	50257
Torch Data Type	float32

Rank the Japanese Mistral 300M Base 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 Japanese Mistral 300M Base

Best Alternatives	Context / RAM	Downloads	Likes
Mistral 300M Base	8K / 0.6 GB	28	1
Lite Oute 1 300M Instruct	4K / 1.2 GB	46	9
Lite Oute 1 300M	4K / 1.2 GB	21	9
Mistral 300M	4K / 0 GB	7	2
...anese Mistral 300M Instruction	4K / 1.4 GB	3	3

Note: green Score (e.g. "73.2") means that the model is better than ce-lery/japanese-mistral-300m-base.