GraySoft
Projects Models Compare Cloud benchmarks FAQ Download guIDE โ†’
Model Intelligence Sheet

Soulfate24/OvisOCR2-ASHQ1-Remix-GGUF overview

OvisOCR2 ASHQ1 Remix This is a GGUF quantized version of the original model. ๐Ÿ“ˆ Release Benchmarks wiki.test.raw, symmetric FA auto reference | Model | Size | โ€ฆ

transformersggufocrdocument-parsingmultimodalmarkdowntablesformulasvllmqwen3_5quantizationashq1imatriximage-text-to-textbase_model:ATH-MaaS/OvisOCR2base_model:quantized:ATH-MaaS/OvisOCR2license:apache-2.0endpoints_compatibleregion:usconversational

Runs locally from ~1.1 MB disk (4 GB VRAM class GPUs with llama.cpp / guIDE).

Downloads
1,416
Likes
0
Pipeline
image-text-to-text

Repository Files & Downloads

7 GGUF files detected
Direct downloads for local inference
FileTypeQuantizationSizeLink
imatrix.ggufGGUFGGUF1.1 MBDownload
mmproj-ASHQ1-Balanced-54pc.ggufGGUFGGUF121.8 MBDownload
model-ASHQ1-Fidelity-48pc.ggufGGUFGGUF704.5 MBDownload
model-ASHQ1-Nano-27pc.ggufGGUFGGUF457.0 MBDownload
model-ASHQ1-Pico-24pc.ggufGGUFGGUF441.8 MBDownload
model-ASHQ1-Precision-42pc.ggufGGUFGGUF669.6 MBDownload
model-ASHQ1-Quality-36pc.ggufGGUFGGUF556.2 MBDownload

Model Details

Model IDSoulfate24/OvisOCR2-ASHQ1-Remix-GGUF
AuthorSoulfate24
Pipelineimage-text-to-text
Licenseapache-2.0
Base modelATH-MaaS/OvisOCR2
Last modified2026-09-10T10:39:42.000Z

Model README

---

license: apache-2.0

library_name: transformers

pipeline_tag: image-text-to-text

base_model:

- ATH-MaaS/OvisOCR2

tags:

- ocr

- document-parsing

- multimodal

- markdown

- tables

- formulas

- vllm

- qwen3_5

- quantization

- gguf

- ashq1

- imatrix

---

OvisOCR2 - ASHQ1-Remix

This is a GGUF quantized version of the original model.

๐Ÿ“ˆ Release Benchmarks (wiki.test.raw, symmetric FA-auto reference)

| Model | Size | PPL | KLD | RMS ฮ”p | top-p | Speed |

| :--- | ---: | ---: | ---: | ---: | ---: | ---: |

| Q8_0 (stock) | 774 MiB | 28.7134 | 0.0009 | 0.61% | 98.3% | 3413 t/s |

| Fidelity-48pc | 704 MiB | 28.7754 | 0.0019 | 0.96% | 97.3% | 2560 t/s |

| Precision-42pc ๐Ÿฅˆ | 670 MiB | 28.8111 | 0.0023 | 1.04% | 97.2% | 3200 t/s |

| Q6_K-imx (stock) | 601 MiB | 28.8894 | 0.0033 | 1.25% | 96.5% | 3413 t/s |

| Quality-36pc โญ | 556 MiB | 29.0026 | 0.0077 | 1.93% | 94.7% | 3200 t/s |

| Q5_K_M-imx (stock) | 551 MiB | 29.0602 | 0.0088 | 2.02% | 94.2% | 3200 t/s |

| IQ4_XS-imx (stock) | 481 MiB | 30.5363 | 0.0305 | 3.94% | 90.2% | 3413 t/s |

| Nano-27pc | 457 MiB | 30.5109 | 0.0598 | 5.82% | 87.2% | 3200 t/s |

| Pico-24pc | 442 MiB | 31.9258 | 0.0699 | 6.16% | 85.6% | 3200 t/s |

| IQ3_M-imx (stock) | 433 MiB | 32.0551 | 0.0895 | 7.35% | 84.2% | 3200 t/s |

โ„น๏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.

๐Ÿ”— Link: https://huggingface.co/Soulfate24/AutoRound-ASHQ1-Remix_Double-Quantization_Suite

Run Soulfate24/OvisOCR2-ASHQ1-Remix-GGUF with guIDE

Download guIDE โ€” the AI-native code editor with local LLM inference and 69 built-in tools.

Download guIDE โ†’ ยท Browse 524k+ models ยท Compare models

Source: Hugging Face ยท Compare models