llm-jp · First MLX build 5 Feb 2025
llm-jp-3-7.2b-instruct3
MLX
Text apache-2.0
A 7.3B parameter text generation model from llm-jp, with 2 MLX builds at 4-bit.
- Parameters
- 7.3B From Hugging Face
- Quantisation
- 4-bit llm-jp-3-7.2b-instruct3-4bit
- Memory needed
- 7 GB Estimate, unified memory
- Download
- 4.1 GB Weights, estimated
- Downloads
- 118 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 7 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 · mlx-lm$ pip install -U mlx-lm
$ mlx_lm.generate --model mlx-community/llm-jp-3-7.2b-instruct3-4bit --prompt "Hello"MLX builds
2 on mlx-community| llm-jp-3-7.2b-instruct3-4bit | 4-bit | 4.1 GB | 7 GB | 118 |
|---|---|---|---|---|
| llm-jp-3-7.2b-instruct3 | Unstated | 14.6 GB | 19 GB | 103 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.