What are the hardware requirements for MoE LLaVA StableLM 1.6B 4e?

MoE LLaVA StableLM 1.6B 4e requires approximately 6.4 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 MoE LLaVA StableLM 1.6B 4e and how large is it?

MoE LLaVA StableLM 1.6B 4e is developed by LanguageBind, a model with 3.2b 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 MoE LLaVA StableLM 1.6B 4e?

MoE LLaVA StableLM 1.6B 4e is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

MoE LLaVA StableLM 1.6B 4e by LanguageBind — VRAM 6.4GB, 4K context

Name: MoE LLaVA StableLM 1.6B 4e
Author: LanguageBind

MoE LLaVA StableLM 1.6B 4e is an open-source language model by LanguageBind. Features: 3.2b LLM, VRAM: 6.4GB, Context: 4K, License: apache-2.0, MoE, LLM Explorer Score: 0.1.

Arxiv:2311.10122 Arxiv:2401.15947 Custom code Endpoints compatible Moe Moe llava stablelm Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/LanguageBind/MoE-LLaVA-StableLM-1.6B-4e

MoE LLaVA StableLM 1.6B 4e Parameters and Internals

Model Type

vision-language, multimodal

Additional Notes

MoE-LLaVA shows excellent performance in multi-modal learning with fewer parameters.

Training Details

Methodology:

A simple MoE tuning stage was added.

Training Time:

2 days

Hardware Used:

8 V100 GPUs

LLM Name	MoE LLaVA StableLM 1.6B 4e
Repository 🤗	https://huggingface.co/LanguageBind/MoE-LLaVA-StableLM-1.6B-4e
Model Size	3.2b
Required VRAM	6.4 GB
Updated	2026-08-07
Maintainer	LanguageBind
Model Type	moe_llava_stablelm
Model Files	5.0 GB: 1-of-2 1.4 GB: 2-of-2
Model Architecture	MoELLaVAStablelmForCausalLM
License	apache-2.0
Context Length	4096
Model Max Length	4096
Transformers Version	4.36.2
Tokenizer Class	Arcade100kTokenizer
Padding Token	<\|reg0\|>
Vocabulary Size	100352
Torch Data Type	bfloat16
Errors	replace

Rank the MoE LLaVA StableLM 1.6B 4e 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 MoE LLaVA StableLM 1.6B 4e

Best Alternatives	Context / RAM	Downloads	Likes
DynMoE StableLM 1.6B	4K / 6.4 GB	5	2

Note: green Score (e.g. "73.2") means that the model is better than LanguageBind/MoE-LLaVA-StableLM-1.6B-4e.