VlmTableOCR: got multiple values for keyword argument 'resized_shape' error

[artschekoff]

Error:

Error recognizing table: Vlm(model_name=openai/gpt-40-mini) got multiple values for keyword argument 'resized_shape'

Code example:

p2t = Pix2Text(table_ocr=vlm_table_ocr)

total_config = {
    'layout': None,
    "table": {
        "model_type": "VlmTableOCR",  # 指定类名
        "resized_shape": 798,
        "model_name": "openai/gpt-4o-mini",
        "api_key": "key"
    },
}

p2t = Pix2Text.from_config(total_configs=total_config)
result = p2t.recognize_page(img, resized_shape=798)

P.S.: seems like non of those resized_shape have affect