Repository navigation

All pull requests

Labels (8)Milestones (0)New pull request

Open

(34)

Closed

(24)

Add support for nested/OrderedDict outputs and dynamic layer-aligned formatting for FPN-style backbones
#212vlebe opened Nov 25, 2025
add LTSM Layer support
#207intexcor opened Dec 12, 2024
feat: allow nested list/tuple input/output and None type
#200cijinsama opened Jan 28, 2024
4
Fix None Type error while using MultiHeadAttention
#191GCS-ZHN opened Oct 26, 2022
feat: allow more input shapes and weight locations
#189ClashLuke opened Aug 5, 2022
Add license to setup.py
#187normanrz opened Jun 9, 2022
Added OrderedDict as an output datatype
#184SoumyadeepB opened Mar 19, 2022
Return model size info
#182mhamdanx opened Mar 14, 2022
support for list as output item, not only tensors
#167leora-betesh opened Mar 11, 2021
Fix handling input_size with multi-input
#166ahmedhshahin opened Mar 11, 2021
3
Fix the multi-output, dict-input, parameter counting and calculation overflow problem.
#165cainmagi opened Feb 27, 2021
Fix parameter count
#142luisherrmann opened Aug 14, 2020
Minor tweaks and quick fix
#119leriomaggio opened Mar 9, 2020
1
feat: Default device is set to model device
#117frgfm opened Mar 5, 2020
3
fix: Fixed hook for DenseNet
#116frgfm opened Mar 5, 2020
11
Use GPU only when available else default to CPU
#114Gokkulnath opened Feb 26, 2020
add checks for atomic values in output
#111wyc1997 opened Feb 22, 2020
Added two new parameters to summary: verbose and csv. Verbose shows m…
#108eplatero97 opened Jan 10, 2020
Allow user to specify dtype
#104tberckmann opened Dec 23, 2019
Add dictionary support
#101Morpheus3000 opened Nov 20, 2019
1
Multi-gpu compatibility: torch-summary for 'cuda:1' and so on
#93adithyavis opened Sep 21, 2019
1
torchsummary() extensions by input_initializer, dtype
#89DrStoop opened Aug 12, 2019
count running_mean and running_var params for BN
#88mostafaelhoushi opened Aug 9, 2019
2
Add a function that can control the depth of output and display the structure of the model
#83sangyx opened Jul 19, 2019
New functionality: use torchsummary to build pytorch model with scalable input shape
#80ncble opened Jul 4, 2019

Navigation Menu

Repository navigation

All pull requests

Add support for nested/OrderedDict outputs and dynamic layer-aligned formatting for FPN-style backbones

add LTSM Layer support

feat: allow nested list/tuple input/output and None type

Fix None Type error while using MultiHeadAttention

feat: allow more input shapes and weight locations

Add license to setup.py

Added OrderedDict as an output datatype

Return model size info

support for list as output item, not only tensors

Fix handling input_size with multi-input

Fix the multi-output, dict-input, parameter counting and calculation overflow problem.

Fix parameter count

Minor tweaks and quick fix

feat: Default device is set to model device

fix: Fixed hook for DenseNet

Use GPU only when available else default to CPU

add checks for atomic values in output

Added two new parameters to summary: verbose and csv. Verbose shows m…

Allow user to specify dtype

Add dictionary support

Multi-gpu compatibility: torch-summary for 'cuda:1' and so on

`torchsummary()` extensions by `input_initializer`, `dtype`

count running_mean and running_var params for BN

Add a function that can control the depth of output and display the structure of the model

New functionality: use torchsummary to build pytorch model with scalable input shape

Pull requests