Soulfate24/LFM2.5-2.6B-DSpark-ASHQ1-Remix-GGUF overview
LFM2.5 2.6B with DSpark ASHQ1 Remix This is a GGUF quantized version of the original model. ๐ Release Benchmarks wiki.test.raw, symmetric FA auto reference | โฆ
Runs locally from ~2.3 MB disk (4 GB VRAM class GPUs with llama.cpp / guIDE).
Repository Files & Downloads
| File | Type | Quantization | Size | Link |
|---|---|---|---|---|
| dspark-ASHQ1-Quality-36pc.gguf | GGUF | GGUF | 234.8 MB | Download |
| imatrix.gguf | GGUF | GGUF | 2.3 MB | Download |
| model-ASHQ1-Compact-33pc.gguf | GGUF | GGUF | 1.67 GB | Download |
| model-ASHQ1-Fidelity-48pc.gguf | GGUF | GGUF | 2.42 GB | Download |
| model-ASHQ1-Mini-30pc.gguf | GGUF | GGUF | 1.52 GB | Download |
| model-ASHQ1-Nano-27pc.gguf | GGUF | GGUF | 1.37 GB | Download |
| model-ASHQ1-Pico-24pc.gguf | GGUF | GGUF | 1.25 GB | Download |
| model-ASHQ1-Precision-42pc.gguf | GGUF | GGUF | 2.15 GB | Download |
| model-ASHQ1-Quality-36pc.gguf | GGUF | GGUF | 1.81 GB | Download |
Model Details
| Model ID | Soulfate24/LFM2.5-2.6B-DSpark-ASHQ1-Remix-GGUF |
|---|---|
| Author | Soulfate24 |
| Pipeline | text-generation |
| License | other |
| Base model | LiquidAI/LFM2.5-2.6B,LiquidAI/LFM2.5-2.6B-DSpark |
| Last modified | 2026-09-10T11:08:39.000Z |
Model README
---
library_name: transformers
license: other
license_name: lfm1.0
license_link: LICENSE
language:
- ar
- zh
- en
- fr
- de
- hi
- id
- it
- ja
- ko
- pl
- pt
- ru
- es
- th
- vi
pipeline_tag: text-generation
base_model:
- LiquidAI/LFM2.5-2.6B
- LiquidAI/LFM2.5-2.6B-DSpark
tags:
- liquid
- lfm2.5
- edge
- speculative-decoding
- dspark
- lfm2
- draft-model
- quantization
- gguf
- ashq1
- imatrix
---
LFM2.5-2.6B with DSpark - ASHQ1-Remix
This is a GGUF quantized version of the original model.
๐ Release Benchmarks (wiki.test.raw, symmetric FA-auto reference)
| Model | Size | PPL | KLD | RMS ฮp | top-p | Speed |
| :--- | ---: | ---: | ---: | ---: | ---: | ---: |
| Q8_0 (stock) | 2742 MiB | 55.0069 | 0.0030 | 1.22% | 97.4% | 3413 t/s |
| Fidelity-48pc | 2481 MiB | 55.6994 | 0.0074 | 1.97% | 96.0% | 2844 t/s |
| Precision-42pc ๐ฅ | 2199 MiB | 55.2931 | 0.0087 | 2.20% | 95.7% | 2695 t/s |
| Q6_K-imx (stock) | 2119 MiB | 55.6407 | 0.0122 | 2.54% | 94.7% | 2844 t/s |
| Quality-36pc โญ | 1850 MiB | 54.9504โ | 0.0347 | 3.95% | 91.1% | 3200 t/s |
| Q5_K_M-imx (stock) | 1850 MiB | 54.9504โ | 0.0347 | 3.95% | 91.1% | 2844 t/s |
| Compact-33pc | 1708 MiB | 53.0954โ | 0.0734 | 6.01% | 87.5% | 2844 t/s |
| Mini-30pc | 1554 MiB | 49.8535โ | 0.1291 | 8.10% | 83.4% | 2844 t/s |
| IQ4_XS-imx (stock) | 1447 MiB | 55.5126 | 0.1486 | 8.59% | 81.8% | 2560 t/s |
| Nano-27pc โ | 1399 MiB | 58.9388 | 0.2130 | 10.16% | 78.3% | 3012 t/s |
| Pico-24pc โ | 1276 MiB | 54.4368โ | 0.3644 | 12.94% | 72.7% | 3200 t/s |
| IQ3_M-imx (stock) | 1225 MiB | 59.3136 | 0.3884 | 13.44% | 72.2% | 3012 t/s |
โน๏ธ About ASHQ1-Remix Suite
Activation-aware GGUF quantization whose every ratio, floor, and cap traces to a measured experiment. Plain-BF16-native first; AutoRound lineage supported with explicit saturation bounds. Full seven-tier ladder validated across six model families.
๐ Link: https://huggingface.co/Soulfate24/AutoRound-ASHQ1-Remix_Double-Quantization_Suite
Run Soulfate24/LFM2.5-2.6B-DSpark-ASHQ1-Remix-GGUF with guIDE
Download guIDE โ the AI-native code editor with local LLM inference and 69 built-in tools.
Source: Hugging Face ยท Compare models