Skip to content
Local AI Wizard
Your machine
Theme

Aquila-VL-2B-llava-qwen

BAAI · BAAI/Aquila-VL-2B-llava-qwen

Weights license: Apache 2.0 · code and components may use other terms

Reported by the publisher25 resultsBAAI's model card · read 2026-10-25 · not ranked · see them
Memory needed~2 GBQ4 weights + working room, before the machine's reserve
Parameters2.2 BHub safetensors total
Repository updated2026-07-13A commit is not a release date

Quality

No comparable score

We track no public, comparable benchmark that covers this model yet. No score is inferred.

Reported by the publisher

25 results · BAAI's model card · read 2026-10-25
MMBench-EN test
78.8VLMEvalKit
MMBench-CN test
76.4VLMEvalKit
MMBench_V1.1 test
75.2VLMEvalKit
MMT-Bench test
58.2VLMEvalKit
RealWorldQA
63.9VLMEvalKit
HallusionBench
43.0VLMEvalKit
SEEDBench2 plus
63.0VLMEvalKit
LLaVABench
68.4VLMEvalKit
Every reported result25
MMStar
54.9VLMEvalKit
POPE
83.6VLMEvalKit
MMVet
44.3VLMEvalKit
MMMU val
47.4VLMEvalKit
ScienceQA test
95.2VLMEvalKit
AI2D test
75.0VLMEvalKit
MathVista testmini
59.0VLMEvalKit
MathVerse testmini
26.2VLMEvalKit
MathVision
18.4VLMEvalKit
DocVQA test
85.0VLMEvalKit
InfoVQA test
58.3VLMEvalKit
ChartQA test
76.5VLMEvalKit
TextVQA val
76.4VLMEvalKit
OCRVQA testcore
64.0VLMEvalKit
VCR en easy
70.0VLMEvalKit
OCRBench
772VLMEvalKit
Average
64.1VLMEvalKit

Values as printed on the card, with the publisher's own settings (card section: Evaluation); column “Aquila-VL-2B”. Not comparable across publishers and not used in any ranking. Model card

Runs on

Fast memory holds it 119

The planning estimate fits the memory the model can use quickly: graphics card VRAM, or unified memory within the macOS GPU share.

Show all 119

Hardware profiles grouped by the Q4 planning estimate on each machine (dedicated GPUs reserve 2 GB; unified and system memory 25%, 2 to 8 GB): the fast memory holds it, it is within memory but outside the fast part (layers offloaded to PC RAM, or above the macOS GPU share), or the machine is one memory tier short. Not a measured fit or speed.

Consider it for

BAAI released Aquila-VL-2B-llava-qwen on 2024-10-25: a visual question answering model, with 2.2B parameters on the Hugging Face count and the Apache 2.0 license. Languages the publisher declares: en, zh.

Check first

License: commercial use allowed. At Q4 the weights load about 1.7 GB before each machine's reserve; image resolution, video frames and context add memory, and the vision encoder needs support in the runtime you pick.

Measured speed

No published measurement for this checkpoint

Tokens per second are shown only when a source publishes them with hardware, build and setup. They are never estimated.

How to run

Runtime compatibility and tokens per second have not been measured for this entry. Published metadata and a memory estimate do not prove a successful local run.

Library in source: transformers

Documented ways to run it

Copied as printed in the linked source (read 2026-10-18); not run by this site.

LLaVA · model card · How to use
# pip install git+https://github.com/LLaVA-VL/LLaVA-NeXT.git
from llava.model.builder import load_pretrained_model
from llava.mm_utils import process_images, tokenizer_image_token
from llava.constants import IMAGE_TOKEN_INDEX, DEFAULT_IMAGE_TOKEN
from llava.conversation import conv_templates
from PIL import Image
import requests
import copy
import torch
import warnings

warnings.filterwarnings("ignore")

pretrained = "BAAI/Aquila-VL-2B-llava-qwen"

model_name = "llava_qwen"
device = "cuda"
device_map = "auto"
tokenizer, model, image_processor, max_length = load_pretrained_model(pretrained, None, model_name, device_map=device_map)  # Add any other thing you want to pass in llava_model_args

model.eval()

# load image from url
url = "https://github.com/haotian-liu/LLaVA/blob/1a91fc274d7c35a9b50b3cb29c4247ae5837ce39/images/llava_v1_5_radar.jpg?raw=true"
image = Image.open(requests.get(url, stream=True).raw)

# load image from local environment
# url = "./local_image.jpg"
# image = Image.open(url)

image_tensor = process_images([image], image_processor, model.config)
image_tensor = [_image.to(dtype=torch.float16, device=device) for _image in image_tensor]

conv_template = "qwen_1_5"  # Make sure you use correct chat template for different models
question = DEFAULT_IMAGE_TOKEN + "\nWhat is shown in this image?"
conv = copy.deepcopy(conv_templates[conv_template])
conv.append_message(conv.roles[0], question)
conv.append_message(conv.roles[1], None)
prompt_question = conv.get_prompt()

input_ids = tokenizer_image_token(prompt_question, tokenizer, IMAGE_TOKEN_INDEX, return_tensors="pt").unsqueeze(0).to(device)
image_sizes = [image.size]

cont = model.generate(
    input_ids,
    images=image_tensor,
    image_sizes=image_sizes,
    do_sample=False,
    temperature=0,
    max_new_tokens=4096,
)

text_outputs = tokenizer.batch_decode(cont, skip_special_tokens=True)

print(text_outputs)

Choose my computer

Technical details

What the source and the catalog store for this checkpoint. Missing fields are left out, never zero.

Checkpoint / source
BAAI/Aquila-VL-2B-llava-qwen
Task
visual-question-answering
Architecture
LlavaQwenForCausalLM
Library in source
transformers
Weight license
Apache 2.0BAAI/Aquila-VL-2B-llava-qwen model card (Hugging Face) · read 2026-10-18source
Commercial use
Allowed Apache License 2.0 text (SPDX) · read 2026-10-18license text
Languages
en, zhModel card metadata · read 2026-10-25source
Released
Aquila-VL-2B model card News (Hugging Face, BAAI) · read 2026-10-18source
Every stored field12 fields
Revision checked
32a63cf22d73a5f253d994987fa96e0bf5515e15
Architecture
LlavaQwenForCausalLM
Library
transformers
Hub task tag
visual-question-answering
Declared base models
Qwen/Qwen2.5-1.5B-Instruct · google/siglip-so400m-patch14-384
Formats
safetensors
Access
Open download
Publisher kind
Official
Repository created
2024-10-17creation is not the model's release
Repository updated
2026-07-13a commit is not a release date
Hub downloads (30 days) · likes
256 · 62this repository only · read 2026-10-25
Catalog state
Hub config or weights found; no inventory benchmark link

Sources

Identity and published specifications come from the checkpoint page. Benchmark scores belong to the named release and do not measure speed on your computer.

Curated facts checked: 2026-10-18. The license and release date cite the publisher's sources; parameters, files and task come from the Hugging Face record, and memory is a labelled estimate from its parameter count. A value no source states stays empty; a repository commit is not a model release date.