You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

Tree: f45f9657ab

detached

master

resnet

Branches Tags

${ item.name }

Create tag ${ searchTerm }

Create branch ${ searchTerm }

from 'f45f9657ab'

${ noResults }

cimp-impl/cimp/resnet

History

mht f45f9657ab Fix: Remove invalid debug_get_conv3_1t_output and debug_get_conv4_1t_output calls from test_models.cpp - Removed calls to non-existent BBRegressor methods debug_get_conv3_1t_output and debug_get_conv4_1t_output in test/test_models.cpp. - Restored correct BBRegressor processing logic using get_iou_feat, get_modulation, and predict_iou. - This resolves build errors introduced by referencing methods that were never implemented. - Build and test workflow now completes successfully.		3 weeks ago
..
resnet.cpp	Fix ResNet BatchNorm parameter loading and enhance BN1 debugging. Corrected loading of BatchNorm running_mean, running_var, and num_batches_tracked parameters in C++ ResNet BottleneckImpl and ResNetImpl to use direct member assignment instead of named_buffers(). This resolved discrepancies with Python's BatchNorm behavior. Added detailed intermediate output saving for bn1 in both C++ ResNet and Python comparison script to facilitate debugging. Ensured Python comparison script correctly loads and compares these new ResNet intermediate tensors. This series of changes led to numerical equivalence for ResNet conv1, bn1, and subsequently layer1-4 outputs between Python and C++.	2 months ago
resnet.h	Feat: Achieved perfect Conv1, good BN1/ReLU1/MaxPool similarity. README updated.	2 months ago