Skip to content
ModelListedExternal
Browse files

Gharscan Qwen2vl GGUF

GGUF export of the GharScan building-defect detection model.

Sourcehuggingface.co/ritvik360/gharscan-qwen2vl-gguf/tree/8e289539c93f144a5a2d9b12162cb0626a863921
KindOtherBase modelQwen/Qwen2-VL-2B-InstructVersionRevision 8e28953Bgrade

Model source

Kind
Other
File size
940 MB–2.88 GB
Base model
Qwen/Qwen2-VL-2B-Instruct
Version
vRevision 8e28953
Updated
Oct 23, 2026
Downloads
36
Source
Hugging Face

Source description

GharScan Qwen2-VL GGUF

GGUF export of the GharScan building-defect detection model.

Base Model

Qwen/Qwen2-VL-2B-Instruct

Fine-Tuning

LoRA fine-tuned on Indian residential wall-defect imagery.

LoRA Adapter: https://huggingface.co/ritvik360/gharscan-qwen2vl-lora

Datasets

Defect Dataset: https://huggingface.co/datasets/ritvik360/gharscan-defect-dataset

Agent Traces: https://huggingface.co/datasets/ritvik360/gharscan-agent-traces

Quantizations

  • gharscan-f16.gguf
  • gharscan-q4_k_m.gguf

Use Case

GharScan analyzes residential building defects including:

  • Hairline cracks
  • Structural cracks
  • Settlement cracks
  • Water seepage
  • Efflorescence
  • Concrete spalling
  • Rebar corrosion
  • Plaster delamination

The model is designed for rapid first-pass assessment of residential wall defects in Indian housing contexts.

Conversion

Converted to GGUF and quantized using llama.cpp.

Sources

1 source
Hugging FacePrimaryVersion 8e289539c93f144a5a2d9b12162cb0626a863921
huggingface.co/ritvik360/gharscan-qwen2vl-gguf/tree/8e289539c93f144a5a2d9b12162cb0626a863921

Verified Oct 23

Download files

Model artifacts

2 artifacts

gharscan-f16.gguf

gguf · 2.88 GB · SHA-256 e0f4167592f8…2cf5 · Hugging Face

Download file

gharscan-q4_k_m.gguf

gguf · 940 MB · SHA-256 a69bddc9d6c4…b483 · Hugging Face

Download file

Source excerpts

3 excerpts
ReadmeAug 6huggingface.co/ritvik360/gharscan-qwen2vl-gguf

--- license: apache-2.0 datasets: - ritvik360/gharscan-defect-dataset language: - en base_model: - Qwen/Qwen2-VL-2B-Instruct pipeline_tag: image-to-text tags: - gguf - llama.cpp - image-to-text - qwen2-vl - computer-vision --- # GharScan Qwen2-VL GGUF GGUF export of the GharScan building-defect detection model. ## Base Model Qwen/Qwen2-VL-2B-Instruct ## Fine-Tuning LoRA fine-tuned on Indian residential wall-defect imagery. LoRA Adapter: https://huggingface.co/ritvik360/gharscan-qwen2vl-lora ## Datasets Defect Dataset: https://huggingface.co/datasets/ritvik360/gharscan-defect-dataset Agent Traces: https://huggingface.co/datasets/ritvik360/gharscan-agent-traces ## Quantizations * gharscan-f16.gguf * gharscan-q4_k_m.gguf ## Use Case GharScan analyzes residential building defects including: * Hairline cracks * Structural cracks * Settlement cracks * Water seepage * Efflorescence * Concrete spalling * Rebar corrosion * Plaster delamination The model is designed for rapid first-pass assessment of residential wall defects in Indian housing contexts. ## Conversion Converted to GGUF and quantized using llama.cpp.

NoteOct 23huggingface.co/ritvik360/gharscan-qwen2vl-gguf

Source context: 36 downloads · 0 likes · Pipeline image-to-text · Repo ritvik360/gharscan-qwen2vl-gguf

NoteAug 6huggingface.co/ritvik360/gharscan-qwen2vl-gguf

Source context: 99 downloads · 0 likes · Pipeline image-to-text · Repo ritvik360/gharscan-qwen2vl-gguf