Granite 3 3B A800M Instruct

Granite 3 3B A800M Instruct is a lightweight 3B-parameter IBM open foundation language model for developers and self-hosters who need local multilingual text…

  • AI
  • Tools & Utilities
Granite 3 3B A800M Instruct is a lightweight 3B-parameter IBM open foundation language model for developers and self-hosters who need local multilingual text generation, summarisation, question answering, and reasoning without sending data to external services.

Platform availability

Available on 0 of 10 platforms

  • Unraid (not listed)
  • TrueNAS (not listed)
  • Umbrel (not listed)
  • ZimaOS (not listed)
  • Proxmox (not listed)
  • Helm (not listed)
  • Mac (not listed)
  • Windows (not listed)
  • Linux (not listed)
  • Any server (not listed)

Health score?A 0-100 score computed nightly from four factors: maintained (40%: recent commits and steady activity), popular (25%: GitHub stars, log scale), easy to install (20%: platforms, install notes, setup guides, Docker image) and light to run (15%: minimum RAM, ARM support). When the source code has been checked, a fifth factor, security check (20%: automated scan for committed secrets, unsafe desktop-app settings, known-vulnerable dependencies and project hygiene; any serious finding caps it at 20), is added. Factors without data are left out and the rest are rescaled.

20/100 Low

Maintained
0GitHub repository is archived
Popular
49273 GitHub stars
Easy to install
250 install platforms, install notes, Docker image
Light to run
205.24023 GB minimum RAM, runs on ARM
GitHub stars
273
Open issues
4
Last commit
2024-12-04
License
Apache-2.0
Activity
stale

Checked 3 days ago - source: GitHub (ibm-granite/granite-3.0-language-models)

Resources & Compatibility

Minimum requirements: 5.24023 GB RAM. ARM (e.g. Raspberry Pi) is supported.

Checked 54 days ago - source: Docker Hub image tags and published docs

First-install notes

  • Install PyTorch and vLLM before starting Use pip (the Python package installer) to install PyTorch and the fixed vLLM version from the docs. Without these, the model has no working runner and will fail to start. source
  • Install a recent transformers package Install transformers version 4.49 or newer, a Python package that helps the model read and prepare text and image data. If it is too old, the model may refuse to load or give import errors. source
  • Fix a missing accelerate error If startup says accelerate is missing, install that extra Python package too. It helps set up the model’s compute resources, and without it the first run can crash. source
  • Upstream status The upstream GitHub repository is archived; use the Red Hat Docker image or Hugging Face collection for model artifacts. source
  • Upstream archived The upstream ibm-granite/granite-3.0-language-models repository is now archived, so this Granite 3.0 model will not receive further updates; newer Granite 3.1 models are listed in the catalog. source
  • Upstream repo archived, model still installable The upstream Granite 3.0 repository was archived, so no further commits or releases are expected from it; this is normal for a released model family and does not affect installation. Pull the model from the listed Docker Hub image as usual. source

Fetched 4 days ago - each note links to its source

Alternatives

No alternatives collected yet.

Common questions

Is there a newer version of Granite-3.0-3B-A800M-Instruct?

Yes, a newer version is listed as granite-3.1-3b-a800m-instruct.

Can I use Granite 3.0 3B A800M Instruct commercially?

Yes, the Granite 3.0 language models are released under Apache 2.0 for research and commercial use.

How many parameters are activated in the 3B MoE model?

The 3B MoE model activates 800M parameters.

Which base checkpoint was the Instruct model fine-tuned from?

It was fine-tuned from Granite-3.0-3B-A800M-Base-4K.

Answers sourced from github.com, huggingface.co, huggingface.co

Community

Collected 48 days ago - public community threads