Google · First MLX build 22 Oct 2024
gemma-2-27b
MLX
Text gemma
A 27B parameter text generation model from Google, with 2 MLX builds, from 4-bit to 8-bit.
- Parameters
- 27B From Hugging Face
- Quantisation
- 4-bit gemma-2-27b-bf16-4bit
- Memory needed
- 20 GB Estimate, unified memory
- Download
- 15.3 GB Weights, estimated
- Downloads
- 94 This build, last 30 days
- Task
- Text generation Text
- MLX builds
- 2 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 20 GB at 4-bit
8 GB
16 GB
24 GB
32 GB
64 GB
128 GB
Fits with headroomTight, close other appsWon’t fit
Estimated from the size of the weights, plus a fifth for context and working memory and 2 GB for macOS and your apps. Long contexts need more.
Run it
Terminal · mlx-lm$ pip install -U mlx-lm
$ mlx_lm.generate --model mlx-community/gemma-2-27b-bf16-4bit --prompt "Hello"MLX builds
2 on mlx-community| gemma-2-27b-bf16-4bit | 4-bit | 15.3 GB | 20 GB | 94 |
|---|---|---|---|---|
| gemma-2-27b-bf16-8bit | 8-bit | 28.9 GB | 35 GB | 33 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.