MLX Community · First MLX build 20 Dec 2023
Mixtral-8x7B-Instruct-v0.1
MLX
apache-2.0
A 47B parameter model, with 3 MLX builds, from 4-bit to 8-bit.
- Parameters
- 47B From Hugging Face
- Quantisation
- 4-bit Mixtral-8x7B-Instruct-v0.1-4bit
- Memory needed
- 32 GB Estimate, unified memory
- Download
- 26.3 GB Weights, estimated
- Downloads
- 310 This build, last 30 days
- Task
- —
- MLX builds
- 3 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 32 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/Mixtral-8x7B-Instruct-v0.1-4bitMLX builds
3 on mlx-community| Mixtral-8x7B-Instruct-v0.1-4bit | 4-bit | 26.3 GB | 32 GB | 310 |
|---|---|---|---|---|
| Mixtral-8x7B-Instruct-v0.1-8bit | 8-bit | 49.6 GB | 58 GB | 47 |
| Mixtral-8x7B-Instruct-v0.1 | Unstated | 93.4 GB | 107 GB | 190 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.