Dolphin-2.9.3-Mistral-Nemo-12b-4bit
mlx-community/Dolphin-2.9.3-Mistral-Nemo-12b-4bit
Uncategorised · mlx
4-bit Open on Hugging Face
Parameters
12B
Weights (est.)
≈ 6.9 GB
Mac memory (est.)
16 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.
Details
- Base model
- mistralai/Mistral-Nemo-Base-2407 (finetune)
- Vendor
- Mistral
- Task
- Uncategorised
- Quantisation
- 4-bit
- Licence
- apache-2.0
- Downloads
- 384 in the last 30 days · 2.7K all time · 0 likes
- Published
- 2024-07-24 · updated 2024-07-24 (2y ago)
- Tags
- mistral, generated_from_trainer, axolotl
Other MLX builds of mistralai/Mistral-Nemo-Base-2407
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| Dolphin-2.9.3-Mistral-Nemo-12b-4bit (this one) | 4-bit | ≈ 6.9 GB | 16 GB+ | 384 |
| Dolphin-2.9.3-Mistral-Nemo-12b-8bit | 8-bit | ≈ 13.0 GB | 24 GB+ | 129 |
| Dolphin-2.9.3-Mistral-Nemo-12b-bf16 | bf16 | ≈ 24.5 GB | 32 GB+ | 172 |