diff options
author | Nathan Bailey <nathan.bailey@arm.com> | 2024-06-03 11:57:14 +0100 |
---|---|---|
committer | Nathan Bailey <nathan.bailey@arm.com> | 2024-06-07 16:17:49 +0100 |
commit | 75d9dd877051fe46f885178ede9ba7623e294a63 (patch) | |
tree | 4a1f11cf3d5cb474c589a884aab2e39f8ccdf5dc /README.md | |
parent | 409a32a521f88a27c0f06fd67d21bc4648c86e73 (diff) | |
download | mlia-75d9dd877051fe46f885178ede9ba7623e294a63.tar.gz |
feat: Add kernel size to user parameter list for conv2D rewrites
Adds the kernel size as a user input parameter for conv2D to add flexibility
Signed-off-by: Nathan Bailey <nathan.bailey@arm.com>
Change-Id: If8fa70e79a64dffafd5cf6b1c3e482e89ac6f8de
Diffstat (limited to 'README.md')
-rw-r--r-- | README.md | 14 |
1 files changed, 7 insertions, 7 deletions
@@ -237,13 +237,13 @@ There are a number of predefined profiles for rewrites shown below: | :-----------------------------------: | :--------: | :--: | :-----------: | :---: | :---------: | :-------: | :---------: | :---------: | :--------: | :--------: | | optimization-fully-connected-pruning | 32 | 1e-3 | True | 48000 | "cosine" | 1 | 0 | None | 2 | 4 | -| Name | Batch Size | LR | Show Progress | Steps | LR Schedule | Num Procs | Num Threads | Checkpoints | Num Clusters | Cluster Centroids Init | Activation | -| :-------------------------------------: | :--------: | :--: | :-----------: | :---: | :---------: | :-------: | :---------: | :---------: | :----------: | :--------------------------------: | :--------: | -| optimization-conv2d-clustering | 32 | 1e-3 | True | 48000 | "cosine" | 1 | 0 | None | 16 | "CentroidInitialization.LINEAR" | "relu" | +| Name | Batch Size | LR | Show Progress | Steps | LR Schedule | Num Procs | Num Threads | Checkpoints | Num Clusters | Cluster Centroids Init | Activation | Kernel Size | +| :-------------------------------------: | :--------: | :--: | :-----------: | :---: | :---------: | :-------: | :---------: | :---------: | :----------: | :--------------------------------: | :--------: | :---------: | +| optimization-conv2d-clustering | 32 | 1e-3 | True | 48000 | "cosine" | 1 | 0 | None | 16 | "CentroidInitialization.LINEAR" | "relu" | 3x3 | -| Name | Batch Size | LR | Show Progress | Steps | LR Schedule | Num Procs | Num Threads | Checkpoints | Sparsity M | Sparsity N | Activation | -| :-----------------------------------: | :--------: | :--: | :-----------: | :---: | :---------: | :-------: | :---------: | :---------: | :--------: | :--------: | :--------: | -| optimization-conv2d-pruning | 32 | 1e-3 | True | 48000 | "cosine" | 1 | 0 | None | 2 | 4 | "relu" | +| Name | Batch Size | LR | Show Progress | Steps | LR Schedule | Num Procs | Num Threads | Checkpoints | Sparsity M | Sparsity N | Activation | Kernel Size | +| :-----------------------------------: | :--------: | :--: | :-----------: | :---: | :---------: | :-------: | :---------: | :---------: | :--------: | :--------: | :--------: | :---------: | +| optimization-conv2d-pruning | 32 | 1e-3 | True | 48000 | "cosine" | 1 | 0 | None | 2 | 4 | "relu" | 3x3 | These are summarized below: @@ -267,7 +267,7 @@ The user can also specify custom augmentations as part of the training parameter The augmentations consist of 2 parameters: mixup strength and gaussian strength. -Augmenations can be selected from a number of pre-defined profiles (see the table below) or each individual parameter can be chosen (see optimization_custom_augmentation above for an example): +Augmentations can be selected from a number of pre-defined profiles (see the table below) or each individual parameter can be chosen (see optimization_custom_augmentation above for an example): | Name | MixUp Strength | Gaussian Strength | | :------------------: | :------------: | :---------------: | |