Skip to content
ModelListedExternal
Browse files

Firered Ocr Crispembed GGUF

GGUF conversion of FireRedTeam/FireRed-OCR for CrispEmbed.

Sourcehuggingface.co/cstr/firered-ocr-crispembed-GGUF/tree/6b24bd0aeca967ba886df8ce9557e9ceabffc7fa
KindOtherBase modelFireRedTeam/FireRed-OCRVersionRevision 6b24bd0Cgrade

Model source

Kind
Other
File size
17.4 MB–3.96 GB
Base model
FireRedTeam/FireRed-OCR
Version
vRevision 6b24bd0
Updated
Oct 23, 2026
Downloads
91
Source
Hugging Face

Source description

FireRed-OCR — CrispEmbed GGUF

GGUF conversion of FireRedTeam/FireRed-OCR for CrispEmbed.

FireRed-OCR is a Qwen3-VL 2B fine-tune with Format-Constrained GRPO training for tables, formulas, and structured document OCR. 92.94% on OmniDocBench v1.5.

Architecture

  • Vision: Qwen3-VL ViT (24 layers, 1024d, patch 16, deepstack [5,11,17])
  • LLM: Qwen3-2B (28 layers, 2048d, GQA 16/8, QK norms, mRoPE interleaved)
  • Training: Format-Constrained GRPO with rewards for formula syntax, table integrity, hierarchical closure

Models

FileQuantSizeNotes
firered-ocr-f16.ggufF164.0 GBFull precision
firered-ocr-q8_0.ggufQ8_02.2 GBBest quality/size
firered-ocr-q4_k.ggufQ4_K1.6 GBSmallest
firered-ocr-ref.ggufF3217 MBReference activations for parity testing

Usage

crispembed --ocr firered-ocr-q8_0.gguf document.png

Runs on the qwen2vl_ocr engine (auto-detected from GGUF architecture qwen3vl).

License

Apache-2.0 (FireRedTeam/FireRed-OCR)

Provenance and EU AI Act Art. 53 note

  • Upstream model: FireRedTeam/FireRed-OCR — published by FireRedTeam.
  • Upstream licence: apache-2.0. This repository redistributes under the same terms; it grants no rights the upstream licence does not.
  • What was done here: format conversion and/or quantisation only (GGUF). No training, no fine-tuning, no merging, no distillation, no change to architecture, vocabulary or capability. Only the numeric representation of the upstream weights differs.
  • Training data: documented — where it is documented at all — by the upstream provider; see the upstream model card. No training data was used, added or selected by this repository. No training-content summary was found on the upstream model card at the time of writing; that documentation gap is upstream's and is not filled here.
  • Provider status: under Regulation (EU) 2024/1689 the upstream authors remain the provider of this model. Converting the serialisation format does not make this repository the provider of a new general-purpose AI model, and no such claim is made. Questions about training content, copyright policy or model capability belong upstream.

Sources

1 source
Hugging FacePrimaryVersion 6b24bd0aeca967ba886df8ce9557e9ceabffc7fa
huggingface.co/cstr/firered-ocr-crispembed-GGUF/tree/6b24bd0aeca967ba886df8ce9557e9ceabffc7fa

Verified Oct 23

Download files

Model artifacts

4 artifacts

firered-ocr-f16.gguf

gguf · 3.96 GB · SHA-256 19b74dc8a077…5113 · Hugging Face

Download file

firered-ocr-q4_k.gguf

gguf · 1.54 GB · SHA-256 d85072b974fb…04cc · Hugging Face

Download file

firered-ocr-q8_0.gguf

gguf · 2.20 GB · SHA-256 0c5fc359576f…cf6b · Hugging Face

Download file

firered-ocr-ref.gguf

gguf · 17.4 MB · SHA-256 5b22ebbad120…9d40 · Hugging Face

Download file

Source excerpts

3 excerpts
ReadmeAug 6huggingface.co/cstr/firered-ocr-crispembed-GGUF

--- license: apache-2.0 tags: [gguf, ocr, vlm, qwen3-vl, document-understanding, crispembed] base_model: FireRedTeam/FireRed-OCR library_name: crispembed pipeline_tag: image-to-text --- # FireRed-OCR — CrispEmbed GGUF GGUF conversion of [FireRedTeam/FireRed-OCR](https://huggingface.co/FireRedTeam/FireRed-OCR) for [CrispEmbed](https://github.com/CrispStrobe/CrispEmbed). FireRed-OCR is a Qwen3-VL 2B fine-tune with Format-Constrained GRPO training for tables, formulas, and structured document OCR. 92.94% on OmniDocBench v1.5. ## Architecture - **Vision**: Qwen3-VL ViT (24 layers, 1024d, patch 16, deepstack [5,11,17]) - **LLM**: Qwen3-2B (28 layers, 2048d, GQA 16/8, QK norms, mRoPE interleaved) - **Training**: Format-Constrained GRPO with rewards for formula syntax, table integrity, hierarchical closure ## Models | File | Quant | Size | Notes | |------|-------|------|-------| | `firered-ocr-f16.gguf` | F16 | 4.0 GB | Full precision | | `firered-ocr-q8_0.gguf` | Q8_0 | 2.2 GB | Best quality/size | | `firered-ocr-q4_k.gguf` | Q4_K | 1.6 GB | Smallest | | `firered-ocr-ref.gguf` | F32 | 17 MB | Reference activations for parity testing | ## Usage ```bash crispembed --ocr firered-ocr-q8_0.gguf document.png ``` Runs on the `qwen2vl_ocr` engine (auto-detected from GGUF architecture `qwen3vl`). ## License Apache-2.0 ([FireRedTeam/FireRed-OCR](https://huggingface.co/FireRedTeam/FireRed-OCR)) ## Provenance and EU AI Act Art. 53 note - **Upstream model:** [FireRedTeam/FireRed-OCR](https://huggingface.co/FireRedTeam/FireRed-OCR) — published by `FireRedTeam`. - **Upstream licence:** `apache-2.0`. This repository redistributes under the same terms; it grants no rights the upstream licence does not. - **What was done here:** format conversion and/or quantisation only (GGUF). No training, no fine-tuning, no merging, no distillation, no change to architecture, vocabulary or capability. Only the numeric representation of the upstream weights differs. - **Training data:** documented — where it is documented at all — by the upstream provider; see the upstream model card. No training data was used, added or selected by this repository. No training-content summary was found on the upstream model card at the time of writing; that documentation gap is upstream's and is not filled here. - **Provider status:** under Regulation (EU) 2024/1689 the upstream authors remain th...

NoteOct 23huggingface.co/cstr/firered-ocr-crispembed-GGUF

Source context: 91 downloads · 0 likes · Pipeline image-to-text · Library crispembed · Repo cstr/firered-ocr-crispembed-GGUF

NoteAug 6huggingface.co/cstr/firered-ocr-crispembed-GGUF

Source context: 56 downloads · 0 likes · Pipeline image-to-text · Library crispembed · Repo cstr/firered-ocr-crispembed-GGUF