超分训练模型转推理失败

[aiyodiulehuner]

python3 tools/export_model.py -c checkpoints/sr_telescope_train/config.yml -o Global.pretrained_model=checkpoints/sr_telescope_train/best_accuracy Global.save_inference_dir=./inference/sr_out
grep: warning: GREP_OPTIONS is deprecated; please use an alias or script
W0928 07:17:32.174465 23625 gpu_context.cc:278] Please NOTE: device: 0, GPU Compute Capability: 7.5, Driver API Version: 10.2, Runtime API Version: 10.2
W0928 07:17:32.179884 23625 gpu_context.cc:306] device: 0, cuDNN Version: 7.6.
[2023/09/28 07:17:40] ppocr INFO: load pretrain successful from checkpoints/sr_telescope_train/best_accuracy
Traceback (most recent call last):
File "tools/export_model.py", line 269, in
main()
File "tools/export_model.py", line 265, in main
model, arch_config, save_path, logger, input_shape=input_shape)
File "tools/export_model.py", line 185, in export_single_model
paddle.jit.save(model, save_path)
File "/usr/local/python3.7.0/lib/python3.7/site-packages/paddle/fluid/dygraph/jit.py", line 629, in wrapper
func(layer, path, input_spec, **configs)
File "/usr/local/python3.7.0/lib/python3.7/site-packages/decorator.py", line 232, in fun
return caller(func, *(extras + args), **kw)
File "/usr/local/python3.7.0/lib/python3.7/site-packages/paddle/fluid/wrapped_decorator.py", line 25, in impl
return wrapped_func(*args, **kwargs)
File "/usr/local/python3.7.0/lib/python3.7/site-packages/paddle/fluid/dygraph/base.py", line 51, in impl
return func(*args, **kwargs)
File "/usr/local/python3.7.0/lib/python3.7/site-packages/paddle/fluid/dygraph/jit.py", line 857, in save
inner_input_spec, with_hook=with_hook)
File "/usr/local/python3.7.0/lib/python3.7/site-packages/paddle/fluid/dygraph/dygraph_to_static/program_translator.py", line 528, in concrete_program_specify_input_spec
*desired_input_spec, with_hook=with_hook)
File "/usr/local/python3.7.0/lib/python3.7/site-packages/paddle/fluid/dygraph/dygraph_to_static/program_translator.py", line 436, in get_concrete_program
concrete_program, partial_program_layer = self._program_cache[cache_key]
File "/usr/local/python3.7.0/lib/python3.7/site-packages/paddle/fluid/dygraph/dygraph_to_static/program_translator.py", line 801, in getitem
self._caches[item_id] = self._build_once(item)
File "/usr/local/python3.7.0/lib/python3.7/site-packages/paddle/fluid/dygraph/dygraph_to_static/program_translator.py", line 790, in build_once
**cache_key.kwargs)
File "/usr/local/python3.7.0/lib/python3.7/site-packages/decorator.py", line 232, in fun
return caller(func, *(extras + args), **kw)
File "/usr/local/python3.7.0/lib/python3.7/site-packages/paddle/fluid/wrapped_decorator.py", line 25, in impl
return wrapped_func(*args, **kwargs)
File "/usr/local/python3.7.0/lib/python3.7/site-packages/paddle/fluid/dygraph/base.py", line 51, in impl
return func(*args, **kwargs)
File "/usr/local/python3.7.0/lib/python3.7/site-packages/paddle/fluid/dygraph/dygraph_to_static/program_translator.py", line 740, in from_func_spec
error_data.raise_new_exception()
File "/usr/local/python3.7.0/lib/python3.7/site-packages/paddle/fluid/dygraph/dygraph_to_static/error.py", line 336, in raise_new_exception
six.exec("raise new_exception from None")
File "", line 1, in
AssertionError: In transformed code:

File "/paddle/PaddleOCR-release-2.6/ppocr/modeling/architectures/base_model.py", line 84, in forward
    x = self.transform(x)
File "/paddle/PaddleOCR-release-2.6/ppocr/modeling/transforms/tbsrn.py", line 208, in forward
    'block%d' % (i + 2))(block[str(i + 1)])
File "/paddle/PaddleOCR-release-2.6/ppocr/modeling/transforms/tbsrn.py", line 261, in forward
    size = residual.shape
    residual = residual.reshape([size[0], size[1], -1])
    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ <--- HERE
    residual = self.feature_enhancer(residual)
    residual = residual.reshape([size[0], size[1], size[2], size[3]])

File "/usr/local/python3.7.0/lib/python3.7/site-packages/paddle/tensor/manipulation.py", line 2139, in reshape
    return paddle.fluid.layers.reshape(x=x, shape=shape, name=name)
File "/usr/local/python3.7.0/lib/python3.7/site-packages/paddle/fluid/layers/nn.py", line 6435, in reshape
    attrs["shape"] = get_attr_shape(shape)
File "/usr/local/python3.7.0/lib/python3.7/site-packages/paddle/fluid/layers/nn.py", line 6411, in get_attr_shape
    % dim_idx)

AssertionError: Only one dimension value of 'shape' in reshape can be -1. But received shape[2] is also -1.

    # N = x.shape()[2]              # N is an int. (NOT recommend under @to_static)
    N = paddle.shape(x)[2]          # N is a Tensor. (Recommend)
    z = paddle.reshape([N, -1, 4])  # z.shape is [-1, -1, 4]

If your target shape in Reshape represents dynamic shape, please turn it into a Tensor under @to_static. See above example for details.