Skip to content

ModelListedExternal

LightOnOCR 2 1B Bbox Soup GGUF

This are the quantizations of the model LightOnOCR-2-1B-bbox-soup

Sourcehuggingface.co/noctrex/LightOnOCR-2-1B-bbox-soup-GGUF/tree/a620a66c4d57c8f2f253ab44c9ba0e48486c92bc

KindOtherBase modellightonai/LightOnOCR-2-1B-bbox-soupVersionRevision a620a66

Model source

Kind: Other
File size: 1.12 MB–1.53 GB
Base model: lightonai/LightOnOCR-2-1B-bbox-soup
Version: vRevision a620a66
Updated: Sep 26, 2026
Downloads: 682
Tasks: Captioning
Source: Hugging Face

Source description

This are the quantizations of the model LightOnOCR-2-1B-bbox-soup

Try to use the best quality you can run.

Quality order: BF16 > F16 > Q8_0 > Q6_K > Q5_K_M > IQ4_NL > IQ4_XS > Q4_K_M

For the mmproj, try to use the F32 version as it will produce the best results.
Quality order: F32 > BF16 > F16

Sources

1 source

Hugging FacePrimaryVersion a620a66c4d57c8f2f253ab44c9ba0e48486c92bc

huggingface.co/noctrex/LightOnOCR-2-1B-bbox-soup-GGUF/tree/a620a66c4d57c8f2f253ab44c9ba0e48486c92bc

Verified Sep 10

Model artifacts

12 artifacts

imatrix.gguf

gguf · 1.12 MB · SHA-256 824e59913f7c…f4cd · Hugging Face

LightOnOCR-2-1B-bbox-soup-BF16.gguf

gguf · 1.12 GB · SHA-256 c2e68132e7bf…922f · Hugging Face

LightOnOCR-2-1B-bbox-soup-F16.gguf

gguf · 1.12 GB · SHA-256 f19a660298ed…eea0 · Hugging Face

LightOnOCR-2-1B-bbox-soup-IQ4_NL.gguf

gguf · 364 MB · SHA-256 0c727f8ddab7…f5b9 · Hugging Face

LightOnOCR-2-1B-bbox-soup-IQ4_XS.gguf

gguf · 351 MB · SHA-256 dcbe8878e01e…dba5 · Hugging Face

LightOnOCR-2-1B-bbox-soup-Q4_K_M.gguf

gguf · 378 MB · SHA-256 366f769999a5…456f · Hugging Face

LightOnOCR-2-1B-bbox-soup-Q5_K_M.gguf

gguf · 424 MB · SHA-256 2ef08199a002…8ad3 · Hugging Face

LightOnOCR-2-1B-bbox-soup-Q6_K.gguf

gguf · 472 MB · SHA-256 ce38f1b5f071…afc8 · Hugging Face

LightOnOCR-2-1B-bbox-soup-Q8_0.gguf

gguf · 610 MB · SHA-256 232d287ed22e…7e07 · Hugging Face

mmproj-BF16.gguf

gguf · 791 MB · SHA-256 499bb7544e72…ae93 · Hugging Face

mmproj-F16.gguf

gguf · 781 MB · SHA-256 c118c7d389fd…baad · Hugging Face

mmproj-F32.gguf

gguf · 1.53 GB · SHA-256 b79b4e3f5d67…4136 · Hugging Face

Source excerpts

2 excerpts

ReadmeSep 10huggingface.co/noctrex/LightOnOCR-2-1B-bbox-soup-GGUF

--- pipeline_tag: image-to-text base_model: lightonai/LightOnOCR-2-1B-bbox-soup --- This are the quantizations of the model [LightOnOCR-2-1B-bbox-soup](https://huggingface.co/lightonai/LightOnOCR-2-1B-bbox-soup) Try to use the best quality you can run. Quality order: BF16 > F16 > Q8_0 > Q6_K > Q5_K_M > IQ4_NL > IQ4_XS > Q4_K_M For the mmproj, try to use the F32 version as it will produce the best results. Quality order: F32 > BF16 > F16

NoteSep 10huggingface.co/noctrex/LightOnOCR-2-1B-bbox-soup-GGUF

Source context: 682 downloads · 2 likes · Pipeline image-to-text · Repo noctrex/LightOnOCR-2-1B-bbox-soup-GGUF