MLX Community · First MLX build 17 Sept 2024
Mistral-Small-Instruct-2409
MLX
mrl
A 22B parameter model, with 3 MLX builds, from 4-bit to bf16.
- Parameters
- 22B From Hugging Face
- Quantisation
- 4-bit Mistral-Small-Instruct-2409-4bit
- Memory needed
- 16 GB Estimate, unified memory
- Download
- 12.5 GB Weights, estimated
- Downloads
- 53 This build, last 30 days
- Task
- —
- MLX builds
- 3 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 16 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/Mistral-Small-Instruct-2409-4bitMLX builds
3 on mlx-community| Mistral-Small-Instruct-2409-4bit | 4-bit | 12.5 GB | 16 GB | 53 |
|---|---|---|---|---|
| Mistral-Small-Instruct-2409-8bit | 8-bit | 23.6 GB | 29 GB | 24 |
| Mistral-Small-Instruct-2409-bf16 | bf16 | 44.5 GB | 52 GB | 20 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.