DeepSeek-R1-Distill-Qwen-32B-Japanese-4bit
mlx-community/DeepSeek-R1-Distill-Qwen-32B-Japanese-4bit
Text generation · mlx
4-bit Open on Hugging Face
Parameters
33B
Weights (est.)
≈ 18.4 GB
Mac memory (est.)
24 GB+
Size is estimated from the parameter count and quantisation. Memory allows a fifth extra for context plus 2 GB for the system; long contexts need more.
Run it
With mlx-lm on an Apple silicon Mac:
pip install mlx-lmmlx_lm.generate --model mlx-community/DeepSeek-R1-Distill-Qwen-32B-Japanese-4bit --prompt "Hello"
Details
- Base model
- cyberagent/DeepSeek-R1-Distill-Qwen-32B-Japanese (quantized)
- Vendor
- cyberagent
- Task
- Text generation
- Quantisation
- 4-bit
- Licence
- mit
- Downloads
- 160 in the last 30 days · 1.4K all time · 2 likes
- Published
- 2025-01-27 · updated 2025-01-27 (1y ago)
- Languages
- ja
- Tags
- qwen2, japanese
Other MLX builds of cyberagent/DeepSeek-R1-Distill-Qwen-32B-Japanese
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| DeepSeek-R1-Distill-Qwen-32B-Japanese-4bit (this one) | 4-bit | ≈ 18.4 GB | 24 GB+ | 160 |
| DeepSeek-R1-Distill-Qwen-32B-Japanese-8bit | 8-bit | ≈ 34.8 GB | 48 GB+ | 214 |
| DeepSeek-R1-Distill-Qwen-32B-Japanese | — | ≈ 65.5 GB | 96 GB+ | 151 |