What's the most elegant && expected way to infere with my trained checkpoint and modified config?

[ryougishikifor214]

Target

inference within my trained checkpoint and modified config, which is a SISR task.

What Had I done

Once I used mmsegmentation providing mmseg.apis.init_model and mmseg.apis.inference_model, where inference_model is passed through img and model created by init_model.
However, mmagic doesn't provide inference_model and there are only init_model and MMagicInferencer.

I tried to use MMagicInferencer instead but occurred some errors. It seems I didn't pass enough parameters for this interface to initialize a model? or If I set model name in the list like "swinir", the mmagic will try to load its checkpoint from openmmlab?

For init without model_name:

import torch
from mmagic.apis import MMagicInferencer

model = MMagicInferencer(
    #model_name = 'swinir',
    model_config = config_file,
    model_ckpt = checkpoint_file,
    device = torch.device('cuda:0'),
)

It cause the following error:

File ~/work/AI6126project2/mmagic/mmagic/apis/inferencers/__init__.py:99, in Inferencers.__init__(self, task, config, ckpt, device, extra_parameters, seed)
     [96](https://vscode-remote+ssh-002dremote-002b7b22686f73744e616d65223a22466561747572697a6531227d.vscode-resource.vscode-cdn.net/home/featurize/work/AI6126project2/src/tmp/~/work/AI6126project2/mmagic/mmagic/apis/inferencers/__init__.py:96)     self.inferencer = DiffusersPipelineInferencer(
     [97](https://vscode-remote+ssh-002dremote-002b7b22686f73744e616d65223a22466561747572697a6531227d.vscode-resource.vscode-cdn.net/home/featurize/work/AI6126project2/src/tmp/~/work/AI6126project2/mmagic/mmagic/apis/inferencers/__init__.py:97)         config, ckpt, device, extra_parameters, seed=seed)
     [98](https://vscode-remote+ssh-002dremote-002b7b22686f73744e616d65223a22466561747572697a6531227d.vscode-resource.vscode-cdn.net/home/featurize/work/AI6126project2/src/tmp/~/work/AI6126project2/mmagic/mmagic/apis/inferencers/__init__.py:98) else:
---> [99](https://vscode-remote+ssh-002dremote-002b7b22686f73744e616d65223a22466561747572697a6531227d.vscode-resource.vscode-cdn.net/home/featurize/work/AI6126project2/src/tmp/~/work/AI6126project2/mmagic/mmagic/apis/inferencers/__init__.py:99)     raise ValueError(f'Unknown inferencer task: {self.task}')

ValueError: Unknown inferencer task: None

For init withmodel_name = swinir

File ~/work/AI6126project2/mmagic/mmagic/apis/inferencers/__init__.py:70, in Inferencers.__init__(self, task, config, ckpt, device, extra_parameters, seed)
     [67](https://vscode-remote+ssh-002dremote-002b7b22686f73744e616d65223a22466561747572697a6531227d.vscode-resource.vscode-cdn.net/home/featurize/work/AI6126project2/src/tmp/~/work/AI6126project2/mmagic/mmagic/apis/inferencers/__init__.py:67)     self.inferencer = TranslationInferencer(
     [68](https://vscode-remote+ssh-002dremote-002b7b22686f73744e616d65223a22466561747572697a6531227d.vscode-resource.vscode-cdn.net/home/featurize/work/AI6126project2/src/tmp/~/work/AI6126project2/mmagic/mmagic/apis/inferencers/__init__.py:68)         config, ckpt, device, extra_parameters, seed=seed)
     [69](https://vscode-remote+ssh-002dremote-002b7b22686f73744e616d65223a22466561747572697a6531227d.vscode-resource.vscode-cdn.net/home/featurize/work/AI6126project2/src/tmp/~/work/AI6126project2/mmagic/mmagic/apis/inferencers/__init__.py:69) elif self.task in ['Image super-resolution', 'Image Super-Resolution']:
...
---> [76](https://vscode-remote+ssh-002dremote-002b7b22686f73744e616d65223a22466561747572697a6531227d.vscode-resource.vscode-cdn.net/home/featurize/work/AI6126project2/src/tmp/~/work/AI6126project2/env/lib/python3.8/posixpath.py:76)     a = os.fspath(a)
     [77](https://vscode-remote+ssh-002dremote-002b7b22686f73744e616d65223a22466561747572697a6531227d.vscode-resource.vscode-cdn.net/home/featurize/work/AI6126project2/src/tmp/~/work/AI6126project2/env/lib/python3.8/posixpath.py:77)     sep = _get_sep(a)
     [78](https://vscode-remote+ssh-002dremote-002b7b22686f73744e616d65223a22466561747572697a6531227d.vscode-resource.vscode-cdn.net/home/featurize/work/AI6126project2/src/tmp/~/work/AI6126project2/env/lib/python3.8/posixpath.py:78)     path = a

TypeError: expected str, bytes or os.PathLike object, not NoneType

I tried to find our tutorials but it only showed the example for pretrained model.

import mmcv
import matplotlib.pyplot as plt
from mmagic.apis import MMagicInferencer

# Create a MMagicInferencer instance and infer
img = './resources/input/restoration/0901x2.png'
result_out_dir = './resources/output/restoration/tutorial_restoration_esrgan_res.png'
editor = MMagicInferencer('esrgan')
results = editor.infer(img=img, result_out_dir=result_out_dir)

problem

1. how to correctly use MMagicInference
2. Can I use init_model to construct a model, and use in way like "results = model(inputs)"