--- base_model: PokeeAI/pokee_research_7b base_model_relation: quantized quantized_by: ArtusDev license: apache-2.0 language: - en tags: - agent - deepresearch - llm - rl - reinforcementlearning - exl3 datasets: - miromind-ai/MiroRL-GenQA ---
EXL3 quants of PokeeAI/pokee_research_7b using exllamav3 for quantization.
| Quant | BPW | Head Bits | Size (GB) |
|---|---|---|---|
| 2.5_H6 | 2.5 | 6 | 3.56 |
| 3.0_H6 | 3.0 | 6 | 3.97 |
| 3.5_H6 | 3.5 | 6 | 4.38 |
| 4.0_H6 | 4.0 | 6 | 4.79 |
| 4.5_H6 | 4.5 | 6 | 5.19 |
| 5.0_H6 | 5.0 | 6 | 5.60 |
| 6.0_H6 | 6.0 | 6 | 6.42 |
| 8.0_H8 | 8.0 | 8 | 8.19 |
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/PokeeAI_pokee_research_7b-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