What are the hardware requirements for Reglu 100B?

Reglu 100B requires approximately 2.6 GB of VRAM and supports a context window of 2K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Reglu 100B and how large is it?

Reglu 100B is developed by SparseLLM, a model with 100b 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 Reglu 100B?

Reglu 100B is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Reglu 100B by SparseLLM — VRAM 2.6GB, 2K context

Name: Reglu 100B
Author: SparseLLM

Reglu 100B is an open-source language model by SparseLLM. Features: 100b LLM, VRAM: 2.6GB, Context: 2K, License: llama2, LLM Explorer Score: 0.1.

Arxiv:2402.03804 En Endpoints compatible Llama Pytorch Region:us Safetensors

Model Card on HF 🤗: https://huggingface.co/SparseLLM/reglu-100B

Reglu 100B Parameters and Internals

Additional Notes

The study focused on identifying efficient activation functions for sparse computation in large language models through experimentation with ReLU, SwiGLU, ReGLU, and Squared ReLU.

Training Details

Data Sources:

Refinedweb, SlimPajama

Data Volume:

100 billion tokens

Hardware Used:

64xA100(80G)

LLM Name	Reglu 100B
Repository 🤗	https://huggingface.co/SparseLLM/reglu-100B
Model Size	100b
Required VRAM	2.6 GB
Updated	2026-08-01
Maintainer	SparseLLM
Model Type	llama
Model Files	2.6 GB 2.6 GB
Supported Languages	en
Model Architecture	LlamaForCausalLM
License	llama2
Context Length	2048
Model Max Length	2048
Transformers Version	4.36.2
Tokenizer Class	LlamaTokenizer
Vocabulary Size	32000
Torch Data Type	bfloat16

Rank the Reglu 100B 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 Reglu 100B

Best Alternatives	Context / RAM	Downloads	Likes
SmollerLM2 100M Instruct Sft	8K / 0.3 GB	9	1
SmollerLM2 100M	8K / 0.2 GB	5	0
Stockmark 2 100B Instruct Beta	4K / 192.9 GB	13	12
Stockmark 100B	4K / 191.9 GB	9	34
Saily 100b	4K / 235.5 GB	60	7
Plankton 100M	4K / 0.4 GB	8	0
Webcoder 100M Html Css Js	2K / 0.4 GB	487	0
...Climbmix Dolmino Midtrain 100B	2K / 1.5 GB	450	0
...lisLM 100M Layer Hidden Pruned	2K / 0.2 GB	820	0
Mossez 100M Coder Base	1K / 0.4 GB	113	0

Note: green Score (e.g. "73.2") means that the model is better than SparseLLM/reglu-100B.