--- base_model: Hcompany/Holo2-8B base_model_relation: quantized quantized_by: ArtusDev license: apache-2.0 language: - en pipeline_tag: image-text-to-text tags: - multimodal - action - agent - pytorch - computer use - gui agents - exl3 ---
EXL3 quants of Hcompany/Holo2-8B using exllamav3 for quantization.
| Quant | BPW | Head Bits | Size (GB) |
|---|---|---|---|
| 2.5_H6 | 2.5 | 6 | 5.06 |
| 3.0_H6 | 3.0 | 6 | 5.49 |
| 3.5_H6 | 3.5 | 6 | 5.93 |
| 4.0_H6 | 4.0 | 6 | 6.36 |
| 4.5_H6 | 4.5 | 6 | 6.79 |
| 5.0_H6 | 5.0 | 6 | 7.23 |
| 6.0_H6 | 6.0 | 6 | 8.10 |
| 8.0_H8 | 8.0 | 8 | 9.99 |
You can download quants by targeting specific size using the Hugging Face CLI.
pip install -U "huggingface_hub[cli]"
2. Download a specific quant:
huggingface-cli download ArtusDev/Hcompany_Holo2-8B-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.
Made possible with cloud compute from lium.io