Llama-3.3-70B-Instruct-bf16
mlx-community/Llama-3.3-70B-Instruct-bf16
Text generation · transformers
bf16 Open on Hugging Face
Parameters
71B
Weights (est.)
≈ 141 GB
Mac memory (est.)
192 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/Llama-3.3-70B-Instruct-bf16 --prompt "Hello"
Details
- Base model
- meta-llama/Llama-3.3-70B-Instruct (finetune)
- Vendor
- Meta · Llama
- Task
- Text generation
- Quantisation
- bf16
- Licence
- llama3.3
- Downloads
- 172 in the last 30 days · 14K all time · 2 likes
- Published
- 2024-12-06 · updated 2024-12-06 (1y ago)
- Languages
- en, fr, it, pt, hi, es, th, de
- Tags
- llama, facebook, meta, pytorch, llama-3
Other MLX builds of meta-llama/Llama-3.3-70B-Instruct
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| Llama-3.3-70B-Instruct-3bit | 3-bit | ≈ 30.9 GB | 48 GB+ | 1.4K |
| Llama-3.3-70B-Instruct-4bit | 4-bit | ≈ 39.7 GB | 48 GB+ | 6.1K |
| Llama-3.3-70B-Instruct-6bit | 6-bit | ≈ 57.3 GB | 96 GB+ | 678 |
| Llama-3.3-70B-Instruct-8bit | 8-bit | ≈ 75.0 GB | 96 GB+ | 3.7K |
| llama-3.3-70b-instruct-fp16 | fp16 | ≈ 141 GB | 192 GB+ | 1.7K |
| Llama-3.3-70B-Instruct-bf16 (this one) | bf16 | ≈ 141 GB | 192 GB+ | 172 |