simplescaling · First MLX build 12 Feb 2025
s1.1-32B
MLX
Text apache-2.0
A 33B parameter text generation model from simplescaling, with 6 MLX builds, from 4-bit to fp32.
- Parameters
- 33B From Hugging Face
- Quantisation
- fp32 simplescaling-s1.1-32B-fp32
- Memory needed
- 149 GB Estimate, unified memory
- Download
- 131 GB Weights, estimated
- Downloads
- 43 This build, last 30 days
- Task
- Text generation Text
- MLX builds
- 6 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 149 GB at fp32
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/simplescaling-s1.1-32B-fp32 --prompt "Hello"MLX builds
6 on mlx-community| s1.1-32B-4bit | 4-bit | 1000 MB | 4 GB | 20 |
|---|---|---|---|---|
| simplescaling-s1.1-32B-4bit | 4-bit | 18.4 GB | 23 GB | 29 |
| simplescaling-s1.1-32B-6bit | 6-bit | 26.6 GB | 32 GB | 24 |
| simplescaling-s1.1-32B-8bit | 8-bit | 34.8 GB | 41 GB | 35 |
| simplescaling-s1.1-32B-fp16 | fp16 | 65.5 GB | 76 GB | 23 |
| simplescaling-s1.1-32B-fp32 | fp32 | 131 GB | 149 GB | 43 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.