Is it possible to use auto_scale_batch_size and auto_lr_find in lightning-flash?

[IncubatorShokuhou]

I am trying to use auto_scale_batch_size and auto_lr_find in lightning-flash (like in PytorchLightning) in an ObjectDetection task, but it seems that these two arguments can not work in lightning-flash. How can I use these two arguments properly (with lr and lr_scheduler defined in model) ?

Here is my script based on the official ObjectDetection example in the doc:

import flash
from flash.core.data.utils import download_data
from flash.image import ObjectDetectionData, ObjectDetector
from functools import partial
from torch.optim.lr_scheduler import CyclicLR, ReduceLROnPlateau, CosineAnnealingLR
# 1. Create the DataModule
# Dataset Credit: https://www.kaggle.com/ultralytics/coco128
# download_data("https://github.com/zhiqwang/yolov5-rt-stack/releases/download/v0.3.0/coco128.zip", "data/")

datamodule = ObjectDetectionData.from_coco(
    train_folder="data/coco128/images/train2017/",
    train_ann_file="data/coco128/annotations/instances_train2017.json",
    val_split=0.1,
    transform_kwargs={"image_size": 512},
    batch_size=4,
)

# 2. Build the task
model = ObjectDetector(head="efficientdet", backbone="d0",
                       num_classes=datamodule.num_classes, 
                       image_size=512,
                       learning_rate=1e-3,  # <---  specify  lr and lr_scheduler here
                       lr_scheduler=partial(CosineAnnealingLR,T_max=20)
)

# 3. Create the trainer and finetune the model
trainer = flash.Trainer(max_epochs=1,auto_scale_batch_size="binsearch",auto_lr_find=True)  # <---- add auto_scale_batch_size and auto_lr_find here
trainer.finetune(model, datamodule=datamodule, strategy="freeze")

# 4. Detect objects in a few images!
datamodule = ObjectDetectionData.from_files(
    predict_files=[
        "data/coco128/images/train2017/000000000625.jpg",
        "data/coco128/images/train2017/000000000626.jpg",
        "data/coco128/images/train2017/000000000629.jpg",
    ],
    transform_kwargs={"image_size": 512},
    batch_size=4,
)
predictions = trainer.predict(model, datamodule=datamodule)
print(predictions)

# 5. Save the model!
trainer.save_checkpoint("object_detection_model.pt")

The code is runable, but it seems that both auto_scale_batch_size and auto_lr_find did not work.