Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct-bf16
mlx-community/Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct-bf16
Text generation · mlx
bf16 Open on Hugging Face
Parameters
8.0B
Weights (est.)
≈ 16.1 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/Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct-bf16 --prompt "Hello"
Details
- Base model
- nvidia/Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct (finetune)
- Vendor
- NVIDIA
- Task
- Text generation
- Quantisation
- bf16
- Licence
- cc-by-nc-4.0
- Downloads
- 31 in the last 30 days · 933 all time · 0 likes
- Published
- 2025-06-30 · updated 2025-06-30 (1y ago)
- Languages
- en
- Tags
- llama
Other MLX builds of nvidia/Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct-4bit | 4-bit | ≈ 4.5 GB | 8 GB+ | 165 |
| Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct-6bit | 6-bit | ≈ 6.5 GB | 16 GB+ | 104 |
| Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct-bf16 (this one) | bf16 | ≈ 16.1 GB | 24 GB+ | 31 |