[Bug Fix] When loading model from memory, don't need CheckModelFormat

[felixhjh]

PR types(PR类型)

Bug Fix

Description

• When loading model from memory, don't need CheckModelFormat

FastDeploy load model from memory usages(For now temporarily only support Paddle Inference backend):

import fastdeploy as fd
runtime_option = fd.RuntimeOption()
# runtime_option read model form memory
runtime_option.set_model_buffer(model_buffer.read(), model_size, params_buffer.read(), params_size)
runtime_option.use_paddle_backend()
# Initialize model without model path and params path
model = fd.vision.classification.PaddleClasModel( "", "", config_file, runtime_option=runtime_option)
model.predict(im)