Skip to content
ModelListedExternal
Download

TinyLLaVA Qwen2.5 0.5B Instruct Dinov2 Small

Pretrain stage only, 4630 epochs

Sourcehuggingface.co/sbrzz/TinyLLaVA-Qwen2.5-0.5B-Instruct-dinov2-small/tree/3843430d96f08bbfe5eb34c4321d57b8ef5a8412
KindText modelBase modelQwen/Qwen2.5-0.5B-InstructVersionFP16Cgrade

Model source

Kind
Text model
File size
987 MB
Base model
Qwen/Qwen2.5-0.5B-Instruct
Version
vFP16
Updated
Oct 30, 2026
Downloads
14
Source
Hugging Face

Source description

Pretrain stage only, 4630 epochs

Introduction

We use the powerful TinyLLaVA Factory to create a super small image-text-to-text model.

The goal is to make it possible to run LLaVA models on edge devices (with few gigabytes of memory).

For LLM and vision tower, we choose OpenELM-270M-Instruct and facebook/dinov2-small, respectively.

POPE:

Category# SamplesTPFPTNFNAccuracyPrecisionRecallF1 ScoreYes Ratio
Adversarial3000131212502501880.5210.5120.8750.6460.854
Popular3000131212362641880.5250.5150.8750.6480.849
Random2910131211852251880.5280.5250.8750.6560.858

TEXTVQA

Samples 5000, Accuracy 0% (:-|)

SCIENCEQA

Samples 4241, Correct: -, Accuracy: -%, IMG-Accuracy: -%

MMMU

Category# SamplesAccuracy
Overall9000.280
Overall-Art and Design1200.208
Art300.167
Art Theory300.200
Design300.367
Music300.100
Overall-Business1500.213
Accounting300.100
Economics300.367
Finance300.200
Management300.233
Marketing300.167
Overall-Science1500.300
Biology300.300
Chemistry300.133
Geography300.300
Math300.333
Physics300.433
Overall-Health and Medicine1500.340
Basic Medical Science300.300
Clinical Medicine300.133
Diagnostics and Laboratory Med.300.333
Pharmacy300.400
Public Health300.533
Overall-Humanities and Soc. Sci.1200.342
History300.300
Literature300.567
Sociology300.233
Psychology300.267
Overall-Tech and Engineering2100.276
Agriculture300.300
Architecture and Engineering300.200
Computer Science300.367
Electronics300.200
Energy and Power300.367
Materials300.233
Mechanical Engineering300.267

Sources

1 source
Hugging FacePrimaryVersion 3843430d96f08bbfe5eb34c4321d57b8ef5a8412
huggingface.co/sbrzz/TinyLLaVA-Qwen2.5-0.5B-Instruct-dinov2-small/tree/3843430d96f08bbfe5eb34c4321d57b8ef5a8412

Verified Oct 30

Download files

Repository files

File inventory only; completeness is unverified.

Model artifacts

1 artifact

model.safetensors

safetensors · FP16 (tensor header) · 987 MB · SHA-256 ab9d8464c910…bef1 · Hugging Face

Download file

Source excerpts

3 excerpts
ReadmeAug 14huggingface.co/sbrzz/TinyLLaVA-Qwen2.5-0.5B-Instruct-dinov2-small

--- license: apache-2.0 language: - zho - eng - fra - spa - por - deu - ita - rus - jpn - kor - vie - tha - ara base_model: - Qwen/Qwen2.5-0.5B-Instruct - facebook/dinov2-small pipeline_tag: visual-question-answering --- Pretrain stage only, 4630 epochs # Introduction We use the powerful [TinyLLaVA Factory](https://github.com/TinyLLaVA/TinyLLaVA_Factory) to create a super small image-text-to-text model. The goal is to make it possible to run LLaVA models on edge devices (with few gigabytes of memory). For LLM and vision tower, we choose [OpenELM-270M-Instruct](apple/OpenELM-270M-Instruct) and [facebook/dinov2-small](facebook/dinov2-small), respectively. [POPE](https://tinyllava-factory.readthedocs.io/en/latest/Evaluation.html#pope): | Category | # Samples | TP | FP | TN | FN | Accuracy | Precision | Recall | F1 Score | Yes Ratio | |-----------------|---------------|--------|--------|--------|--------|--------------|---------------|------------|--------------|---------------| | Adversarial | 3000 | 1312 | 1250 | 250 | 188 | 0.521 | 0.512 | 0.875 | 0.646 | 0.854 | | Popular | 3000 | 1312 | 1236 | 264 | 188 | 0.525 | 0.515 | 0.875 | 0.648 | 0.849 | | Random | 2910 | 1312 | 1185 | 225 | 188 | 0.528 | 0.525 | 0.875 | 0.656 | 0.858 | [TEXTVQA](https://tinyllava-factory.readthedocs.io/en/latest/Evaluation.html#textvqa) Samples 5000, Accuracy 0% (:-|) [SCIENCEQA](https://tinyllava-factory.readthedocs.io/en/latest/Evaluation.html#scienceqa) Samples 4241, Correct: -, Accuracy: -%, IMG-Accuracy: -% [MMMU](https://tinyllava-factory.readthedocs.io/en/latest/Evaluation.html#mmmu) | Category | # Samples | Accuracy | |---------------------------------|-----------|----------| | Overall | 900 | 0.280 | | Overall-Art and Design | 120 | 0.208 | | Art | 30 | 0.167 | | Art Theory | 30 | 0.200 | | Design | 30 | 0.367 | | Music | 30 | 0.100 | | Overall-Business | 150 | 0.213 | | Accounting | 30 | 0.100 | | Economics | 30 | 0.367 | | Finance | 30 | 0.200 | | Management | 30 | 0.233 | | Marketing | 30 | 0.167 | | Overall-Science | 150 | 0.300 | | Biology | 30 | 0.300 | | Chemistry | 30 | 0.133 | | Geography | 30 | 0.300 | | Math | 30 | 0.333 | | Physics | 30 | 0.433 | | Overall-Health and Medicine | 150 | 0.340 | | Basic Medical Science | 30 | 0.300 | | Clinical Medicine | 30 | 0.133 | | Diagnostics and Laboratory Med. |...

NoteOct 30huggingface.co/sbrzz/TinyLLaVA-Qwen2.5-0.5B-Instruct-dinov2-small

Source context: 14 downloads · 0 likes · Pipeline visual-question-answering · Repo sbrzz/TinyLLaVA-Qwen2.5-0.5B-Instruct-dinov2-small

NoteAug 14huggingface.co/sbrzz/TinyLLaVA-Qwen2.5-0.5B-Instruct-dinov2-small

Source context: 11 downloads · 0 likes · Pipeline visual-question-answering · Repo sbrzz/TinyLLaVA-Qwen2.5-0.5B-Instruct-dinov2-small