8-bit Open on Hugging Face
Parameters
71B
Weights (est.)
≈ 75.0 GB
Mac memory (est.)
96 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/shisa-v2-llama3.3-70b-8bit --prompt "Hello"
Details
- Base model
- shisa-ai/shisa-v2-llama3.3-70b (quantized)
- Vendor
- shisa-ai
- Task
- Text generation
- Quantisation
- 8-bit
- Licence
- llama3.3
- Downloads
- 23 in the last 30 days · 562 all time · 0 likes
- Published
- 2025-04-15 · updated 2025-04-15 (1y ago)
- Languages
- ja, en
- Tags
- llama
Other MLX builds of shisa-ai/shisa-v2-llama3.3-70b
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| shisa-v2-llama3.3-70b-4bit | 4-bit | ≈ 39.7 GB | 48 GB+ | 60 |
| shisa-v2-llama3.3-70b-6bit | 6-bit | ≈ 57.3 GB | 96 GB+ | 23 |
| shisa-v2-llama3.3-70b-8bit (this one) | 8-bit | ≈ 75.0 GB | 96 GB+ | 23 |
| shisa-v2-llama3.3-70b-fp16 | fp16 | ≈ 141 GB | 192 GB+ | 28 |