metadata
base_model: DMindAI/DMind-2-107B
base_model_relation: quantized
quantized_by: ArtusDev
license: apache-2.0
language:
- en
- zh
tags:
- exl3
ArtusDev/DMindAI_DMind-2-107B-EXL3
EXL3 quants of DMindAI/DMind-2-107B using exllamav3 for quantization.
Quants
| Quant | BPW | Head Bits | Size (GB) |
|---|---|---|---|
| 2.5_H6 | 2.5 | 6 | 34.98 |
| 2.91_H6 (optimized) | 2.91 | 6 | 40.15 |
| 3.0_H6 | 3.0 | 6 | 41.60 |
| 3.38_H6 (optimized) | 3.38 | 6 | 46.33 |
| 3.5_H6 | 3.5 | 6 | 48.18 |
| 3.85_H6 (optimized) | 3.85 | 6 | 52.62 |
| 4.0_H6 | 4.0 | 6 | 54.80 |
| 4.25_H6 | 4.25 | 6 | 58.10 |
| 5.0_H6 | 5.0 | 6 | 68.00 |
| 6.0_H6 | 6.0 | 6 | 81.20 |
| 8.0_H8 | 8.0 | 8 | 107.75 |
How to Download and Use Quants
You can download quants by targeting specific size using the Hugging Face CLI.
Click for download commands
1. Install huggingface-cli:
pip install -U "huggingface_hub[cli]"huggingface-cli download ArtusDev/DMindAI_DMind-2-107B-EXL3 --revision "5.0bpw_H6" --local-dir ./EXL3 quants can be run with any inference client that supports EXL3, such as TabbyAPI. Refer to documentation for set up instructions.
Acknowledgements
Made possible with cloud compute from lium.io