license: mit
datasets:
- liuhaotian/LLaVA-Pretrain
- liuhaotian/LLaVA-Instruct-150K
language:
- en
- zh
library_name: transformers
WORK IN PROGRESS
We present TinyLLaVA, a small vision-language chatbot (1.4B) that reaches comparable performances with contemporary vision language models on common benchmarks, using less parameters. TinyLLaVA was trained by finetuning TinyLlama on the LLaVA-1.5 dataset, following the training recipe of LLaVA-1.5. For more details, please refer to the LLaVA-1.5 paper.
Model Performance
We have evaluated TinyLLaVA on GQA, VizWiz, VQAv2, TextVQA and SQA.
| Model | VQAv2 | GQA | SQA | TextVQA | VizWiz |
|---|---|---|---|---|---|
| TinyLLaVA-v1-1.4B | 73.41 | 57.54 | 59.40 | 46.37 | 49.56 |
| BLIP-2 | 41.00 | 41.00 | 61.00 | 42.50 | 19.60 |
| LLaVA-v1.5-7B | 78.50 | 62.00 | 66.80 | 61.3 | 50 |
| LLaVA-v1.5-13B | 80.00 | 63.30 | 71.60 | 61.3 | 53.6 |
| Qwen-VL-7B | 78.80 | 59.30 | 67.10 | 63.8 | 35.2 |
| Qwen-VL-13B | 78.20 | 57.50 | 68.20 | 61.5 | 38.9 |
More evaluations are ongoing.
Model Preparations
- Transformers Version
Make sure to have transformers >= 4.35.3.
- Prompt Template
The model supports multi-image and multi-prompt generation. When using the model, make sure to follow the correct prompt template (USER: <image>xxx\nASSISTANT:), where <image> token is a place-holding special token for image embeddings.
Model Inference from pipeline and transformers
- Using pipeline:
Below we used "bczhou/tiny-llava-v1-hf" checkpoint.
from transformers import pipeline
from PIL import Image
import requests
model_id = "bczhou/tiny-llava-v1-hf"
pipe = pipeline("image-to-text", model=model_id)
url = "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/tasks/ai2d-demo.jpg"
image = Image.open(requests.get(url, stream=True).raw)
prompt = "USER: <image>\nWhat does the label 15 represent? (1) lava (2) core (3) tunnel (4) ash cloud\nASSISTANT:"
outputs = pipe(image, prompt=prompt, generate_kwargs={"max_new_tokens": 200})
print(outputs[0])
>>> {"generated_text': 'USER: \nWhat does the label 15 represent? (1) lava (2) core (3) tunnel (4) ash cloud\nASSISTANT: The label 15 represents lava, which is a type of volcanic rock."}
- Using pure transformers:
Below is an example script to run generation in float16 precision on a GPU device:
import requests
from PIL import Image
import torch
from transformers import AutoProcessor, LlavaForConditionalGeneration
model_id = "bczhou/tiny-llava-v1-hf"
prompt = "USER: <image>\nWhat are these?\nASSISTANT:"
image_file = "http://images.cocodataset.org/val2017/000000039769.jpg"
model = LlavaForConditionalGeneration.from_pretrained(
model_id,
torch_dtype=torch.float16,
low_cpu_mem_usage=True,
).to(0)
processor = AutoProcessor.from_pretrained(model_id)
raw_image = Image.open(requests.get(image_file, stream=True).raw)
inputs = processor(prompt, raw_image, return_tensors='pt').to(0, torch.float16)
output = model.generate(**inputs, max_new_tokens=200, do_sample=False)
print(processor.decode(output[0][2:], skip_special_tokens=True))
Contact
This model was trained by Baichuan Zhou, from Beihang Univerisity, under the supervision of Prof. Lei Huang.