OpenELM 450M

OpenELM 450M has 457M parameters. Estimated VRAM is 0.8 GB at Q4, 1.1 GB at Q8 and 1.6 GB at FP16. All figures are estimates, covering weights plus about 15% and 0.5 GB for a short context.

Yes. OpenELM 450M fits on a 24 GB, 16 GB, 12 GB or 8 GB GPU at FP16, with an estimated 1.6 GB.

Yes. Estimated VRAM at Q4 is 0.8 GB, and assuming macOS lets the GPU use about 75% of unified memory, the smallest common Mac memory size that fits Q4 is 8 GB (estimate).

Yes. OpenELM 450M has open weights, so you can self-host it for full control of your data.

Run OpenELM 450M locally

No. OpenELM 450M accepts text only, so it cannot read images, audio, video or PDFs.

The closest alternatives, matched on model kind and release date with no comparable price, are OLMo 1.7 7B by Ai2, Hermes 2 Pro Llama 3 8B by Nous and Mistral 7B Instruct v0.3 by Mistral AI.

OLMo 1.7 7BOpenELM 450M vs OLMo 1.7 7BHermes 2 Pro Llama 3 8BOpenELM 450M vs Hermes 2 Pro Llama 3 8BMistral 7B Instruct v0.3OpenELM 450M vs Mistral 7B Instruct v0.3

OpenELM 450M is open weights, not open source, under the apple-amlr license.

The OpenELM 450M weights can be downloaded from the Hugging Face repository apple/OpenELM-450M.

Run OpenELM 450M locally

OpenELM 450M was made by Apple and released on April 23, 2024.

More from Apple

Yes. OpenELM 450M has open weights.