Skip to content

Commit

Permalink
Added pretrained models and verified functionality.
Browse files Browse the repository at this point in the history
  • Loading branch information
dwromero committed Feb 11, 2021
1 parent e5a36da commit cf6960d
Show file tree
Hide file tree
Showing 14 changed files with 49 additions and 3 deletions.
2 changes: 1 addition & 1 deletion .gitignore
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
.idea
wandb
ckernel_fitting/wandb
data
data/*
4 changes: 2 additions & 2 deletions run_experiment.py
Original file line number Diff line number Diff line change
Expand Up @@ -41,8 +41,8 @@ def main(_):
np.random.seed(config.seed)

# initialize weight and bias
os.environ["WANDB_API_KEY"] = "3fe624d6a1979f80f1277200966d17bed042ec31" ## Place here your API key.
if not config.train:
os.environ["WANDB_API_KEY"] = "" ## Place here your API key.
if not config.train or True:
os.environ["WANDB_MODE"] = "dryrun"
tags = [
config.model,
Expand Down
46 changes: 46 additions & 0 deletions runs/runs.md
Original file line number Diff line number Diff line change
Expand Up @@ -62,3 +62,49 @@ One can change the frequency of the test set by varying the key `--config.sr_tes

###### Data drop percentage = 30%

`run_experiment.py --config.batch_size=32 --config.clip=0 --config.dataset=CharTrajectories --config.device=cuda --config.drop_rate=30 --config.dropout=0.2 --config.dropout_in=0 --config.epochs=200 --config.kernelnet_activation_function=Sine --config.kernelnet_no_hidden=32 --config.kernelnet_norm_type=LayerNorm --config.kernelnet_omega_0=17.243982444546685 --config.lr=0.001 --config.mfcc=True --config.model=CKCNN --config.no_blocks=2 --config.no_hidden=30 --config.optimizer=Adam --config.permuted=False --config.sched_decay_factor=5 --config.sched_decay_steps=(75,) --config.sched_patience=20 --config.scheduler=plateau --config.weight_decay=0`

###### Data drop percentage = 50%

`run_experiment.py --config.batch_size=32 --config.clip=0 --config.dataset=CharTrajectories --config.device=cuda --config.drop_rate=50 --config.dropout=0 --config.dropout_in=0 --config.epochs=200 --config.kernelnet_activation_function=Sine --config.kernelnet_no_hidden=32 --config.kernelnet_norm_type=LayerNorm --config.kernelnet_omega_0=11.997102427300288 --config.lr=0.001 --config.mfcc=True --config.model=CKCNN --config.no_blocks=2 --config.no_hidden=30 --config.optimizer=Adam --config.permuted=False --config.sched_decay_factor=5 --config.sched_decay_steps=(75,) --config.sched_patience=20 --config.scheduler=plateau --config.weight_decay=0.0001`

###### Data drop percentage = 70%

`run_experiment.py --config.batch_size=32 --config.clip=0 --config.dataset=CharTrajectories --config.device=cuda --config.drop_rate=70 --config.dropout=0.1 --config.dropout_in=0 --config.epochs=200 --config.kernelnet_activation_function=Sine --config.kernelnet_no_hidden=32 --config.kernelnet_norm_type=LayerNorm --config.kernelnet_omega_0=4.237433943113986 --config.lr=0.001 --config.mfcc=True --config.model=CKCNN --config.no_blocks=2 --config.no_hidden=30 --config.optimizer=Adam --config.permuted=False --config.sched_decay_factor=5 --config.sched_decay_steps=(75,) --config.sched_patience=20 --config.scheduler=plateau --config.weight_decay=0`

### SpeechCommands (Preprocessed)

`run_experiment.py --config.batch_size=64 --config.clip=0 --config.dataset=SpeechCommands --config.device=cuda --config.dropout=0.2 --config.dropout_in=0 --config.epochs=200 --config.kernelnet_activation_function=Sine --config.kernelnet_no_hidden=32 --config.kernelnet_norm_type=LayerNorm --config.kernelnet_omega_0=30.90257553169801 --config.lr=0.001 --config.mfcc=True --config.model=CKCNN --config.no_blocks=2 --config.no_hidden=30 --config.optimizer=Adam --config.permuted=False --config.sched_decay_factor=5 --config.sched_decay_steps=(75,) --config.sched_patience=15 --config.scheduler=plateau --config.weight_decay=0`

### SpeechCommands (Raw)

###### Sampling rate test
One can change the frequency of the test set by varying the key `--config.sr_test`.

###### Sampling rate train = 1

`run_experiment.py --config.batch_size=32 --config.clip=0 --config.dataset=SpeechCommands --config.device=cuda --config.dropout=0 --config.dropout_in=0 --config.epochs=300 --config.kernelnet_activation_function=Sine --config.kernelnet_no_hidden=32 --config.kernelnet_norm_type=LayerNorm --config.kernelnet_omega_0=39.450608806633795 --config.lr=0.001 --config.mfcc=False --config.model=CKCNN --config.no_blocks=2 --config.no_hidden=30 --config.optimizer=Adam --config.permuted=False --config.sched_decay_factor=5 --config.sched_decay_steps=(75,) --config.sched_patience=20 --config.scheduler=plateau --config.sr_train=1 --config.weight_decay=0.0001 --config.weight_dropout=0`

###### Sampling rate train = 1/2

`run_experiment.py --config.batch_size=32 --config.clip=0 --config.dataset=SpeechCommands --config.device=cuda --config.dropout=0 --config.dropout_in=0 --config.epochs=300 --config.kernelnet_activation_function=Sine --config.kernelnet_no_hidden=32 --config.kernelnet_norm_type=LayerNorm --config.kernelnet_omega_0=39.450608806633795 --config.lr=0.001 --config.mfcc=False --config.model=CKCNN --config.no_blocks=2 --config.no_hidden=30 --config.optimizer=Adam --config.permuted=False --config.sched_decay_factor=5 --config.sched_decay_steps=(75,) --config.sched_patience=20 --config.scheduler=plateau --config.sr_train=2 --config.weight_decay=0.0001 --config.weight_dropout=0`

###### Sampling rate train = 1/4

`run_experiment.py --config.batch_size=32 --config.clip=0 --config.dataset=SpeechCommands --config.device=cuda --config.dropout=0 --config.dropout_in=0 --config.epochs=300 --config.kernelnet_activation_function=Sine --config.kernelnet_no_hidden=32 --config.kernelnet_norm_type=LayerNorm --config.kernelnet_omega_0=39.450608806633795 --config.lr=0.001 --config.mfcc=False --config.model=CKCNN --config.no_blocks=2 --config.no_hidden=30 --config.optimizer=Adam --config.permuted=False --config.sched_decay_factor=5 --config.sched_decay_steps=(75,) --config.sched_patience=20 --config.scheduler=plateau --config.sr_train=4 --config.weight_decay=0.0001 --config.weight_dropout=0`

###### Sampling rate train = 1/8

`run_experiment.py --config.batch_size=32 --config.clip=0 --config.dataset=SpeechCommands --config.device=cuda --config.dropout=0 --config.dropout_in=0 --config.epochs=300 --config.kernelnet_activation_function=Sine --config.kernelnet_no_hidden=32 --config.kernelnet_norm_type=LayerNorm --config.kernelnet_omega_0=39.450608806633795 --config.lr=0.001 --config.mfcc=False --config.model=CKCNN --config.no_blocks=2 --config.no_hidden=30 --config.optimizer=Adam --config.permuted=False --config.sched_decay_factor=5 --config.sched_decay_steps=(75,) --config.sched_patience=20 --config.scheduler=plateau --config.sr_train=8 --config.weight_decay=0.0001 --config.weight_dropout=0`

###### Data drop percentage = 30%

`run_experiment.py --config.batch_size=32 --config.clip=0 --config.dataset=SpeechCommands --config.device=cuda --config.drop_rate=30 --config.dropout=0.1 --config.dropout_in=0 --config.epochs=300 --config.kernelnet_activation_function=Sine --config.kernelnet_no_hidden=32 --config.kernelnet_norm_type=LayerNorm --config.kernelnet_omega_0=35.655229877794035 --config.lr=0.001 --config.mfcc=False --config.model=CKCNN --config.no_blocks=2 --config.no_hidden=30 --config.optimizer=Adam --config.permuted=False --config.sched_decay_factor=5 --config.sched_decay_steps=(75,) --config.sched_patience=20 --config.scheduler=plateau --config.sr_train=1 --config.weight_decay=0.0001 --config.weight_dropout=0`

###### Data drop percentage = 50%

`run_experiment.py --config.batch_size=32 --config.clip=0 --config.dataset=SpeechCommands --config.device=cuda --config.drop_rate=50 --config.dropout=0 --config.dropout_in=0 --config.epochs=300 --config.kernelnet_activation_function=Sine --config.kernelnet_no_hidden=32 --config.kernelnet_norm_type=LayerNorm --config.kernelnet_omega_0=31.700745675276966 --config.lr=0.001 --config.mfcc=False --config.model=CKCNN --config.no_blocks=2 --config.no_hidden=30 --config.optimizer=Adam --config.permuted=False --config.sched_decay_factor=5 --config.sched_decay_steps=(75,) --config.sched_patience=20 --config.scheduler=plateau --config.sr_train=1 --config.weight_decay=0.0001 --config.weight_dropout=0`

###### Data drop percentage = 70%

`run_experiment.py --config.batch_size=32 --config.clip=0 --config.dataset=SpeechCommands --config.device=cuda --config.drop_rate=70 --config.dropout=0.1 --config.dropout_in=0.1 --config.epochs=300 --config.kernelnet_activation_function=Sine --config.kernelnet_no_hidden=32 --config.kernelnet_norm_type=LayerNorm --config.kernelnet_omega_0=23.29255834358289 --config.lr=0.001 --config.mfcc=False --config.model=CKCNN --config.no_blocks=2 --config.no_hidden=30 --config.optimizer=Adam --config.permuted=False --config.sched_decay_factor=5 --config.sched_decay_steps=(75,) --config.sched_patience=20 --config.scheduler=plateau --config.sr_train=1 --config.weight_decay=0 --config.weight_dropout=0.1`
Binary file not shown.
Binary file not shown.
Binary file not shown.
Binary file not shown.
Binary file not shown.
Binary file not shown.
Binary file not shown.
Binary file not shown.
Binary file not shown.
Binary file not shown.
Binary file not shown.

0 comments on commit cf6960d

Please sign in to comment.