Name: Kogpt J 350M
Author: heegyu

Kogpt J 350M is an open-source language model by heegyu. Features: 350m LLM, VRAM: 1.4GB, License: mit, LLM Explorer Score: 0.05.

Dataset:heegyu/korean-petition... Dataset:heegyu/kowikitext Dataset:heegyu/namuwiki-extrac... Endpoints compatible Gptj Jax Ko Pytorch Region:us

Model Card on HF 🤗: https://huggingface.co/heegyu/kogpt-j-350m

Kogpt J 350M Benchmarks

LLME Score: 0.05111

^nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").

What is the LLM Explorer Rank (Score)

🌟 Advertise your project 🚀

Kogpt J 350M Parameters and Internals

Model Type

text-generation

Additional Notes

This model's training data may include various discriminatory/offensive data, and no removal process has been conducted. Consequently, the model may generate sentences that contain discriminatory/offensive statements regarding certain individuals, races, genders, or disabilities.

Supported Languages

ko (Fluent)

Training Details

Data Sources:

AIHub SNS 대화, AIHub 구어체, AIHub 도서, AIHub 대규모 웹데이터 기반 한국어 말뭉치, 한국어 위키, 나무위키, 국립국어원 메신저 대화, 국립국어원 일상대화 말뭉치, 국립국어원 문어 말뭉치, 국립국어원 구어 말뭉치, 국립국어원 신문 말뭉치, 청와대 국민청원

Data Volume:

Approx. 7B tokens

Context Length:

1024

Training Time:

2023/1/25 ~ 2023/1/29

Hardware Used:

TPU V2-8

Model Architecture:

20 Layers, 1024 hidden dim, 4096 intermediate, 16 heads, 51200 vocab size

LLM Name	Kogpt J 350M
Repository 🤗	https://huggingface.co/heegyu/kogpt-j-350m
Model Size	350m
Required VRAM	1.4 GB
Updated	2026-03-29
Maintainer	heegyu
Model Type	gptj
Model Files	1.4 GB
Supported Languages	ko
Model Architecture	GPTJForCausalLM
License	mit
Transformers Version	4.25.1
Tokenizer Class	GPT2Tokenizer
Beginning of Sentence Token	<\|endoftext\|>
End of Sentence Token	<\|endoftext\|>
Unk Token	<\|endoftext\|>
Vocabulary Size	51200
Torch Data Type	float32
Activation Function	gelu_new
Errors	replace

Best Alternatives to Kogpt J 350M

Best Alternatives	Context / RAM	Downloads	Likes
Codegen 350M List Manip 5 Len	0K / 1.5 GB	5	1
Codegen 350M Multi Gptj	0K / 0.8 GB	8	4
Codegen 350M Mono Gptj	0K / 0.8 GB	7	2

Rank the Kogpt J 350M Capabilities

🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟

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

What open-source LLMs or SLMs are you in search of? 52473 in total.

Email us: info@extractum.io. Our Privacy Policy | Terms and Conditions | Suggest an improvement.

Our Social Media →

Original data from HuggingFace, OpenCompass and various public git repos.

Check out Ag3ntum — our secure, self-hosted AI agent for server management.

Release v20260328a

Support LLM Explorer

Kogpt J 350M by heegyu

» All LLMs » heegyu » Kogpt J 350M URL Share it on

Kogpt J 350M Benchmarks

Kogpt J 350M Parameters and Internals

Best Alternatives to Kogpt J 350M

Rank the Kogpt J 350M Capabilities

What open-source LLMs or SLMs are you in search of? 52473 in total.