statisticize in tensor symbolic shapes #404

lixinqi · 2025-12-01T12:35:07Z

PR Category

Feature Enhancement

Description

statisticize in tensor symbolic shapes

example outputs:

      1 [(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(1024,S1),(S0,80,S3),(S0,S2,1024),(S0,),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,1024),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,)]
      1 [(1,128,S1),(1,S0,1024)]
      1 [(1,2,3,S0,S0)]
      1 [(1,S0),(1,S0),(1,S0)]
      1 [(1,S0,512)]
      1 [(1,S0,S0)]
      1 [(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(256,S1),(S0,80,S3),(S0,S2,256),(S0,),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,256),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,),(S1,)]
      1 [(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(512,S2),(S0,),(S1,80,S4),(S1,S3,512),(S1,S3,512),(S1,),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,512),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,),(S2,)]
      1 [(S0,100),(S0,100)]
      1 [(S0,1025,1024),(S0,1025,1024),(S0,1025,1024),(S0,1025,1024)]
      1 [(S0,1152)]
      1 [(S0,11),(S0,11,1024),(S0,11,1024)]
      1 [(S0,11),(S0,11,768),(S0,11,768)]
      1 [(S0,128,S2),(S0,S1,1024),(S0,)]
      1 [(S0,12),(S0,12,512),(S0,12,512)]
      1 [(S0,1370,1024),(S0,1370,1024),(S0,1370,1024),(S0,1370,1024)]
      1 [(S0,13),(S0,13)]
      1 [(S0,17),(S0,17,1024),(S0,17,1024)]
      1 [(S0,192,32,32),(S0,384,16,16),(S0,768,8,8),(S0,96,64,64)]
      1 [(S0,192,64,64),(S0,384,32,32),(S0,768,16,16),(S0,96,128,128)]
      1 [(S0,19),(S0,19)]
      1 [(S0,1),(S0,64),(S0,64)]
      1 [(S0,27),(S0,27)]
      1 [(S0,2917,1024),(S0,2917,1024),(S0,2917,1024),(S0,2917,1024)]
      1 [(S0,32)]
      1 [(S0,3,384,384)]
      1 [(S0,34,1)]
      1 [(S0,3,640,640)]
      1 [(S0,3,S1,S2,S2)]
      1 [(S0,4)]
      1 [(S0,500)]
      1 [(S0,512),(S0,512)]
      1 [(S0,768)]
      1 [(S0,S1,192)]
      1 [(S0,S1,1),(S0,S1,1024)]
      1 [(S0,S1,1),(S0,S1,768)]
      1 [(S0,S1,2304)]
      1 [(S0,S1,2560)]
      1 [(S0,S1),(S0,S1,512),(S0,S1,512)]
      1 [(S0,S1,S1,384)]
      1 [(S0,S1,S2)]
      1 [(S0,S1),(S2,128)]
      2 [(1,1,S1),(1,S0,512),(1,S1,512)]
      2 [(1,8,3,S0,S0)]
      2 [(1,S0,900,256),(1,S0,900,4)]
      2 [(S0,)]
      2 [(S0,100)]
      2 [(S0,128)]
      2 [(S0,1370,768),(S0,1370,768),(S0,1370,768),(S0,1370,768)]
      2 [(S0,3,768,768)]
      2 [(S0,5),(S0,6)]
      2 [(S0,8)]
      2 [(S0,S1,256),(S0,S1,256)]
      2 [(S0,S1,521),(S0,)]
      2 [(S0,S1,S2,S2)]
      3 [(1,3,S0,S1)]
      3 [(1,S0,S1)]
      3 [(S0,S1)]
      3 [(S0,S1,768)]
      5 [(S0,12),(S0,12),(S0,12)]
      5 [(S0,3,512,1024)]
      5 [(S0,80,S1),(S0,)]
      5 [(S0,),(S1,80,S2),(S1,)]
      7 [(1,S0,3,S1,S1)]
     10 [(2,S0),(2,S0)]
     10 [(S0,3,256,256)]
     17 [(S0,S1),(S0,S1)]
     21 None
     24 [(S0,3,256,192)]
     26 [(1,S0)]
     33 [(S0,3,S1,S2)]
     43 [(S0,3,224,224)]
     53 [(S0,S1),(S0,S1),(S0,S1)]
     55 [(S0,3,512,512)]
    346 [(1,3,S0,S0)]
    537 [(S0,3,S1,S1)]
    589 []
    606 [(S0,1),(S0,S1),(S0,S1)]

…nput_tensor_constraints

…or.reshape

…nput_tensor_constraints

… bugs in generalization for torch.Tensor.view and torch.Tensor.reshape

…nput_tensor_constraints

…m_axes_pairs, enabling targeted configuration for specific use cases

…nput_tensor_constraints

…ation process.

…nput_tensor_constraints

paddle-bot · 2025-12-01T16:37:47Z

Thanks for your contribution!

lixinqi added 30 commits July 31, 2025 08:23

support checking model redundancy

ea77602

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into develop

c3b3ea9

revert change of vision_model_test

c21cb49

reformat python code.

ca9017f

reformat bert_model_test.py and utils.py

52cc34d

minor fix

d8c6213

fix failed check by comparing directories after os.path.realpath()

6bd1370

merge paddle repo develop

5db0b63

fix bugs in check_validate.sh

165ae4b

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into develop

6059328

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into develop

d2d9e0b

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into develop

2cfa175

set dynamic=False in single_device_runner.py

3a75ddd

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into develop

b394760

reset graph hash

868b686

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into develop

4c473cf

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into develop

5530841

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into develop

0caf96a

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into develop

718cd39

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into develop

d46c810

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into develop

985d3cc

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into develop

782858e

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into develop

a7acb51

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into develop

b49edb4

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into develop

6be9482

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into develop

7b950b9

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into develop

956fb0d

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into develop

41fb776

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into develop

99bbefe

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into develop

16264a7

lixinqi added 28 commits November 23, 2025 09:40

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into gen_i…

e85b357

…nput_tensor_constraints

Introduce input_tensor_constraints.py using shape propagation logic.

2a78c74

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into gen_i…

ebc1dd0

…nput_tensor_constraints

support dimension generalization for torch.Tensor.view and torch.Tens…

a9b0066

…or.reshape

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into gen_i…

4945f85

…nput_tensor_constraints

1) support dimension generalization for torch.Tensor.expand(); 2) fix…

bcf5df6

… bugs in generalization for torch.Tensor.view and torch.Tensor.reshape

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into gen_i…

8ccb9b5

…nput_tensor_constraints

dimension_generalization_passes

dc95020

Refactored DimensionGeneralizationPass.__init__ to accept argument di…

b00fd55

…m_axes_pairs, enabling targeted configuration for specific use cases

save dimension generalization pass names into graph_net.json

0acd80c

Generalize sequence dimension

880248d

merge develop

8173fea

more dimension generalization passes for token dimension

2d35e3b

merge develop

aba28f1

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into gen_i…

bc5a23d

…nput_tensor_constraints

refactor parse_sole_graph_module

8d2874c

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into gen_i…

29dbd62

…nput_tensor_constraints

Enhance the performance of the input_tensor_constraints.py file gener…

32f9105

…ation process.

minor fix

c90ed7d

merge develop

5415f4a

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into gen_i…

f35ee0a

…nput_tensor_constraints

more dimension generalization pass

61e5ba4

more dimension generalization pass

6a7780b

fix some hint bugs

dd9a05c

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into gen_i…

4b4c6ee

…nput_tensor_constraints

generate input_tensor_constraints.py

7dc23d8

Merge branch 'develop' of github.com:PaddlePaddle/GraphNet into gen_i…

261d6e2

…nput_tensor_constraints

statisticize in tensor symbolic shapes

57cda04

lixinqi merged commit dc0eb9a into PaddlePaddle:develop Dec 1, 2025
3 checks passed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

statisticize in tensor symbolic shapes #404

statisticize in tensor symbolic shapes #404

Uh oh!

lixinqi commented Dec 1, 2025 •

edited

Loading

Uh oh!

Uh oh!

paddle-bot bot commented Dec 1, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

statisticize in tensor symbolic shapes #404

statisticize in tensor symbolic shapes #404

Uh oh!

Conversation

lixinqi commented Dec 1, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

PR Category

Description

Uh oh!

Uh oh!

paddle-bot bot commented Dec 1, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

lixinqi commented Dec 1, 2025 •

edited

Loading