MLX Community · First MLX build 9 Dec 2024
EXAONE-3.5-32B-Instruct
MLX
A 32B parameter model, with 5 MLX builds, from 3-bit to bf16.
- Parameters
- 32B From Hugging Face
- Quantisation
- 4-bit EXAONE-3.5-32B-Instruct-4bit
- Memory needed
- 7 GB Estimate, unified memory
- Download
- 4.3 GB Weights, estimated
- Downloads
- 30 This build, last 30 days
- Task
- —
- MLX builds
- 5 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 7 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 · Hugging Face CLI$ pip install -U huggingface_hub
$ hf download mlx-community/EXAONE-3.5-32B-Instruct-4bitMLX builds
5 on mlx-community| EXAONE-3.5-32B-Instruct-3bit | 3-bit | 3.3 GB | 6 GB | 28 |
|---|---|---|---|---|
| EXAONE-3.5-32B-Instruct-4bit | 4-bit | 4.3 GB | 7 GB | 30 |
| EXAONE-3.5-32B-Instruct-6bit | 6-bit | 6.9 GB | 10 GB | 30 |
| EXAONE-3.5-32B-Instruct-8bit | 8-bit | 10.5 GB | 14 GB | 30 |
| EXAONE-3.5-32B-Instruct-bf16 | bf16 | 64.0 GB | 74 GB | 30 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.