| --- |
| license: other |
| license_name: lfm1.0 |
| base_model: LiquidAI/LFM2.5-8B-A1B |
| base_model_relation: quantized |
| quantized_by: turboderp |
| tags: |
| - exl3 |
| --- |
| |
| EXL3 quants of [LFM2.5-8B-A1B](https://huggingface.co/LiquidAI/LFM2.5-8B-A1B) |
|
|
| ⚠️ Requires ExLlamaV3 v0.0.38 (or v0.0.37 `dev` branch) |
|
|
| [2.10 bits per weight](https://huggingface.co/turboderp/LFM2.5-8B-A1B-exl3/tree/2.10bpw) |
| [2.53 bits per weight](https://huggingface.co/turboderp/LFM2.5-8B-A1B-exl3/tree/2.53bpw) |
| [3.10 bits per weight](https://huggingface.co/turboderp/LFM2.5-8B-A1B-exl3/tree/3.10bpw) |
| [3.53 bits per weight](https://huggingface.co/turboderp/LFM2.5-8B-A1B-exl3/tree/3.53bpw) |
| [4.10 bits per weight](https://huggingface.co/turboderp/LFM2.5-8B-A1B-exl3/tree/4.10bpw) |
| [5.10 bits per weight](https://huggingface.co/turboderp/LFM2.5-8B-A1B-exl3/tree/5.10bpw) |
| [6.10 bits per weight](https://huggingface.co/turboderp/LFM2.5-8B-A1B-exl3/tree/6.10bpw) |
| [8.00 bits per weight / H8](https://huggingface.co/turboderp/LFM2.5-8B-A1B-exl3/tree/8.00bpw) |
|
|
|  |
|
|
