ausboss · First MLX build 18 Oct 2024
llama-30b-supercot
MLX
A 33B parameter model from ausboss, with one MLX build at 8-bit.
- Parameters
- 33B From Hugging Face
- Quantisation
- 8-bit llama-30b-supercot
- Memory needed
- 41 GB Estimate, unified memory
- Download
- 34.6 GB Weights, estimated
- Downloads
- 21 This build, last 30 days
- Task
- —
- MLX builds
- 1 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 41 GB at 8-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/llama-30b-supercotMLX builds
1 on mlx-community| llama-30b-supercot | 8-bit | 34.6 GB | 41 GB | 21 |
|---|
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.