This repository contains the DeepSeek-V3 language model project and related source files.
Open source
24 alternativesmbzuai-oryx/MobiLlama
MobiLlama is a small language model designed for edge devices, according to its repository description.
About mbzuai-oryx/MobiLlama
MobiLlama is an open-source small language model (SLM) developed by Mohamed Bin Zayed University of Artificial Intelligence (MBZUAI) and Linköping University, accepted as a spotlight paper at the ICLR 2025 SLLM Workshop, designed for deployment on resource-constrained edge devices. It uses a parameter sharing scheme adapted from a LLaMA-7B-style architecture to lower pre-training and deployment costs while maintaining performance.
The model is available in 0.5B, 0.8B, and 1B parameter sizes, with dedicated chat variants for each size, and is released under the Apache 2.0 license. Pre-trained models, training code, evaluation scripts, and intermediate training checkpoints are publicly available via its GitHub repository and Hugging Face.
Key features
- Built on LLaMA-7B architecture with parameter sharing for efficiency
- Available in 0.5B, 0.8B, and 1B parameter sizes
- Includes chat-specific versions for each model size
- Released under the open-source Apache 2.0 license
- Public pre-trained models and training code available
- Supports loading intermediate training checkpoints
Best for: Researchers and developers building on-device AI for resource-constrained hardware, such as mobile or IoT devices, benefit most from MobiLlama.
Good for
- Fully transparent open-source design with public code and model access
- Optimized for low resource usage for edge device deployment
- Multiple model sizes to fit varying edge hardware constraints
Watch out for
- Full pre-training requires 20 A100 80GB GPU nodes
- Preprocessed training dataset totals approximately 8TB
- Currently only supports English language tasks
mbzuai-oryx/MobiLlama at a glance
- Category
- AI & Machine Learning › Models & ML Research
- Type
- Open-source project
- Pricing
- Not confirmed. Check the site
- Open source
- Yes · Source code
- Language
- English
- Last checked
- Oct 5, 2026 · Online when checked
- Listed
- Oct 5, 2026
mbzuai-oryx/MobiLlama alternatives
Tools and sites you could use instead, ranked by how closely they match.
This repository contains Google Research code for the Flan-T5 language model project.
Open source
24 alternatives
DeepSeek Coder is a language model project for code generation and programming tasks.
Open source
24 alternatives
OpenVLA provides an open-source vision-language-action model for robotic manipulation tasks.
Open source
24 alternatives
ChatGLM-6B is an open bilingual dialogue language model described in English and Chinese.
Open source
24 alternatives
xTuring supports personalizing open source language models through preprocessing and fine tuning.
Open source
24 alternativesPopular in Models & ML Research
Well-known listings from the same topic.
Falcon LLMfalconllm.tii.aeERNIEernie.baidu.comTokenVersetoken-verse.github.iowTAObittensor.comAlpacacrfm.stanford.edu/2023/03/13/alpaca.htmlSmolVLAhuggingface.co/blog/smolvla
mbzuai-oryx/MobiLlama FAQ
What parameter sizes is MobiLlama available in?
MobiLlama is available in 0.5B, 0.8B, and 1B parameter sizes, with chat variants for each size.
What license governs MobiLlama use?
MobiLlama is released under the Apache 2.0 license.
What hardware is required to pre-train MobiLlama?
Full pre-training requires 20 nodes of A100 80GB GPUs, per the provided pre-training script.
Is something wrong or out of date?
Tell us and an editor will check it.
Own this website?
Claim this listing
Claim the listing and suggest updates yourself. An editor checks every claim first.