01.AI · First MLX build 23 Jun 2024
Yi-1.5-34B-32k
MLX
apache-2.0
A 9.7B parameter model from 01.AI, with 3 MLX builds, from 2-bit to 8-bit.
- Parameters
- 9.7B From Hugging Face
- Quantisation
- 8-bit dolphin-2.9.3-Yi-1.5-34B-32k-8bit
- Memory needed
- 15 GB Estimate, unified memory
- Download
- 11.3 GB Weights, estimated
- Downloads
- 15 This build, last 30 days
- Task
- —
- MLX builds
- 3 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 15 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/dolphin-2.9.3-Yi-1.5-34B-32k-8bitMLX builds
3 on mlx-community| dolphin-2.9.3-Yi-1.5-34B-32k-2bit | 2-bit | 2.8 GB | 6 GB | 9 |
|---|---|---|---|---|
| dolphin-2.9.3-Yi-1.5-34B-32k-4bit | 4-bit | 4.6 GB | 8 GB | 14 |
| dolphin-2.9.3-Yi-1.5-34B-32k-8bit | 8-bit | 11.3 GB | 15 GB | 15 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.