MLX Community · First MLX build 9 Jan 2024
Magicoder-S-DS-6.7B
MLX
Text deepseek
A 6.7B parameter text generation model, with 2 MLX builds at 4-bit.
- Parameters
- 6.7B From Hugging Face
- Quantisation
- Unstated Magicoder-S-DS-6.7B-MLX
- Memory needed
- 18 GB Estimate, unified memory
- Download
- 13.5 GB Weights, estimated
- Downloads
- 40 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 18 GB at Unstated
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/Magicoder-S-DS-6.7B-MLX --prompt "Hello"MLX builds
2 on mlx-community| Magicoder-S-DS-6.7B-MLX | Unstated | 13.5 GB | 18 GB | 40 |
|---|---|---|---|---|
| Magicoder-S-DS-6.7B-4bit-mlx | 4-bit | — | — | 39 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.