basic trials

autoencoders/eval.py

@@ -23,7 +23,7 @@ def evaluate_linear(
     x_test = encoder.encode(ds.data[train_length:].unsqueeze(1) / 255).numpy()
     y_test = ds.targets[train_length:]
 
-    lr = LogisticRegression(max_iter=300)
+    lr = LogisticRegression(max_iter=1000)
     lr.fit(x_train, y_train)
     labels = lr.predict(x_test)
     labels_ohe = F.one_hot(torch.tensor(labels)).float()

autoencoders/train.py

@@ -28,7 +28,7 @@ def main(cfg):
 
     results = autoencoders.eval.evaluate_linear(module=model, trainer=trainer)
     autoencoders.eval.to_json(
-        results={cfg.model.name: results}, filepath=constants.OUTPUTS.joinpath("results.json")
+        results={cfg.model.name: results}, filepath=constants.OUTPUTS.joinpath("results.txt")
     )
 
 

outputs/.gitignore

@@ -0,0 +1,2 @@
+**/*.ckpt
+**/wandb/

outputs/DAE/train/2023-08-29/21-19-54/.hydra/config.yaml

@@ -0,0 +1,90 @@
+data:
+  batch_size: 256
+  n_workers: 10
+  name: mnist
+  train:
+    _target_: torch.utils.data.DataLoader
+    dataset:
+      _target_: autoencoders.data.AutoEncoderDataset
+      dataset:
+        _target_: autoencoders.data.get_mnist_dataset
+        train: true
+    batch_size: ${data.batch_size}
+    shuffle: true
+    num_workers: ${data.n_workers}
+  valid:
+    _target_: torch.utils.data.DataLoader
+    dataset:
+      _target_: autoencoders.data.AutoEncoderDataset
+      dataset:
+        _target_: autoencoders.data.get_mnist_dataset
+        train: false
+    batch_size: ${data.batch_size}
+    shuffle: false
+    num_workers: ${data.n_workers}
+model:
+  optimizer:
+    _target_: torch.optim.Adam
+    _partial_: true
+    lr: 0.001
+    betas:
+    - 0.9
+    - 0.999
+    weight_decay: 0
+  scheduler:
+    _target_: torch.optim.lr_scheduler.CyclicLR
+    _partial_: true
+    base_lr: 0.001
+    max_lr: 3.0
+    cycle_momentum: false
+  name: DAE
+  nn:
+    _target_: autoencoders.models.base_dae.DenoisingAutoEncoder
+    layers:
+    - 128
+    - 64
+    - 64
+    input_shape:
+    - 28
+    - 28
+    loss_func:
+      _target_: torch.nn.MSELoss
+trainer:
+  _target_: pytorch_lightning.Trainer
+  max_epochs: 100
+  accelerator: mps
+  devices: 1
+  logger:
+    _target_: pytorch_lightning.loggers.WandbLogger
+    project: autoencoders
+    name: null
+    id: null
+    group: null
+    job_type: null
+    save_dir: ${hydra:runtime.output_dir}
+    log_model: true
+    tags: ${tags}
+callbacks:
+  model_summary:
+    _target_: pytorch_lightning.callbacks.RichModelSummary
+  progress_bar:
+    _target_: pytorch_lightning.callbacks.RichProgressBar
+    refresh_rate: 5
+    leave: true
+  early_stopping:
+    _target_: pytorch_lightning.callbacks.EarlyStopping
+    monitor: train-loss
+    min_delta: 0.001
+    patience: 5
+    check_on_train_epoch_end: true
+  model_checkpoint:
+    _target_: pytorch_lightning.callbacks.ModelCheckpoint
+    dirpath: ${hydra:runtime.output_dir}/checkpoints
+    monitor: train-loss
+    save_top_k: 1
+    save_on_train_epoch_end: true
+  log_images:
+    _target_: autoencoders.callbacks.LogReconstructedImagesCallback
+tags:
+- ${data.name}
+- ${model.name}

outputs/DAE/train/2023-08-29/21-19-54/.hydra/hydra.yaml

@@ -0,0 +1,173 @@
+hydra:
+  run:
+    dir: outputs/${model.name}/${hydra.job.name}/${now:%Y-%m-%d}/${now:%H-%M-%S}
+  sweep:
+    dir: outputs/${model.name}/${hydra.job.name}/multirun
+    subdir: ${hydra.job.override_dirname}/${now:%Y-%m-%d}/${now:%H-%M-%S}
+  launcher:
+    _target_: hydra_plugins.hydra_joblib_launcher.joblib_launcher.JoblibLauncher
+    n_jobs: -1
+    backend: null
+    prefer: processes
+    require: null
+    verbose: 0
+    timeout: null
+    pre_dispatch: 2*n_jobs
+    batch_size: auto
+    temp_folder: null
+    max_nbytes: null
+    mmap_mode: r
+  sweeper:
+    _target_: hydra._internal.core_plugins.basic_sweeper.BasicSweeper
+    max_batch_size: null
+    params: null
+  help:
+    app_name: ${hydra.job.name}
+    header: '${hydra.help.app_name} is powered by Hydra.
+
+      '
+    footer: 'Powered by Hydra (https://hydra.cc)
+
+      Use --hydra-help to view Hydra specific help
+
+      '
+    template: '${hydra.help.header}
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (group=option)
+
+
+      $APP_CONFIG_GROUPS
+
+
+      == Config ==
+
+      Override anything in the config (foo.bar=value)
+
+
+      $CONFIG
+
+
+      ${hydra.help.footer}
+
+      '
+  hydra_help:
+    template: 'Hydra (${hydra.runtime.version})
+
+      See https://hydra.cc for more info.
+
+
+      == Flags ==
+
+      $FLAGS_HELP
+
+
+      == Configuration groups ==
+
+      Compose your configuration from those groups (For example, append hydra/job_logging=disabled
+      to command line)
+
+
+      $HYDRA_CONFIG_GROUPS
+
+
+      Use ''--cfg hydra'' to Show the Hydra config.
+
+      '
+    hydra_help: ???
+  hydra_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][HYDRA] %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: simple
+        stream: ext://sys.stdout
+    root:
+      level: INFO
+      handlers:
+      - console
+    loggers:
+      logging_example:
+        level: DEBUG
+    disable_existing_loggers: false
+  job_logging:
+    version: 1
+    formatters:
+      simple:
+        format: '[%(asctime)s][%(name)s][%(levelname)s] - %(message)s'
+    handlers:
+      console:
+        class: logging.StreamHandler
+        formatter: simple
+        stream: ext://sys.stdout
+      file:
+        class: logging.FileHandler
+        formatter: simple
+        filename: ${hydra.runtime.output_dir}/${hydra.job.name}.log
+    root:
+      level: INFO
+      handlers:
+      - console
+      - file
+    disable_existing_loggers: false
+  env: {}
+  mode: RUN
+  searchpath: []
+  callbacks: {}
+  output_subdir: .hydra
+  overrides:
+    hydra:
+    - hydra.mode=RUN
+    task:
+    - model=base_dae
+  job:
+    name: train
+    chdir: null
+    override_dirname: model=base_dae
+    id: ???
+    num: ???
+    config_name: config
+    env_set: {}
+    env_copy: []
+    config:
+      override_dirname:
+        kv_sep: '='
+        item_sep: ','
+        exclude_keys: []
+  runtime:
+    version: 1.3.2
+    version_base: '1.3'
+    cwd: /Users/chrissantiago/Dropbox/GitHub/autoencoders
+    config_sources:
+    - path: hydra.conf
+      schema: pkg
+      provider: hydra
+    - path: /Users/chrissantiago/Dropbox/GitHub/autoencoders/autoencoders/conf
+      schema: file
+      provider: main
+    - path: ''
+      schema: structured
+      provider: schema
+    output_dir: /Users/chrissantiago/Dropbox/GitHub/autoencoders/outputs/DAE/train/2023-08-29/21-19-54
+    choices:
+      experiment: null
+      callbacks: encoder
+      trainer: default
+      model: base_dae
+      scheduler@model.scheduler: cyclic
+      optimizer@model.optimizer: adam
+      data: mnist
+      hydra/env: default
+      hydra/callbacks: null
+      hydra/job_logging: default
+      hydra/hydra_logging: default
+      hydra/hydra_help: default
+      hydra/help: default
+      hydra/sweeper: basic
+      hydra/launcher: joblib
+      hydra/output: default
+  verbose: false

outputs/DAE/train/2023-08-29/21-19-54/.hydra/overrides.yaml

@@ -0,0 +1 @@
+- model=base_dae

outputs/DAE/train/2023-08-29/21-19-54/checkpoints/best_k_models.yaml

@@ -0,0 +1 @@
+/Users/chrissantiago/Dropbox/GitHub/autoencoders/outputs/DAE/train/2023-08-29/21-19-54/checkpoints/epoch=10-step=2585.ckpt: 0.03252919763326645

outputs/autoencoder/train/2023-08-29/21-13-51/.hydra/config.yaml

@@ -0,0 +1,90 @@
+data:
+  batch_size: 256
+  n_workers: 10
+  name: mnist
+  train:
+    _target_: torch.utils.data.DataLoader
+    dataset:
+      _target_: autoencoders.data.AutoEncoderDataset
+      dataset:
+        _target_: autoencoders.data.get_mnist_dataset
+        train: true
+    batch_size: ${data.batch_size}
+    shuffle: true
+    num_workers: ${data.n_workers}
+  valid:
+    _target_: torch.utils.data.DataLoader
+    dataset:
+      _target_: autoencoders.data.AutoEncoderDataset
+      dataset:
+        _target_: autoencoders.data.get_mnist_dataset
+        train: false
+    batch_size: ${data.batch_size}
+    shuffle: false
+    num_workers: ${data.n_workers}
+model:
+  optimizer:
+    _target_: torch.optim.Adam
+    _partial_: true
+    lr: 0.001
+    betas:
+    - 0.9
+    - 0.999
+    weight_decay: 0
+  scheduler:
+    _target_: torch.optim.lr_scheduler.CyclicLR
+    _partial_: true
+    base_lr: 0.001
+    max_lr: 3.0
+    cycle_momentum: false
+  name: autoencoder
+  nn:
+    _target_: autoencoders.models.base.AutoEncoder
+    layers:
+    - 128
+    - 64
+    - 16
+    input_shape:
+    - 28
+    - 28
+    loss_func:
+      _target_: torch.nn.MSELoss
+trainer:
+  _target_: pytorch_lightning.Trainer
+  max_epochs: 100
+  accelerator: mps
+  devices: 1
+  logger:
+    _target_: pytorch_lightning.loggers.WandbLogger
+    project: autoencoders
+    name: null
+    id: null
+    group: null
+    job_type: null
+    save_dir: ${hydra:runtime.output_dir}
+    log_model: true
+    tags: ${tags}
+callbacks:
+  model_summary:
+    _target_: pytorch_lightning.callbacks.RichModelSummary
+  progress_bar:
+    _target_: pytorch_lightning.callbacks.RichProgressBar
+    refresh_rate: 5
+    leave: true
+  early_stopping:
+    _target_: pytorch_lightning.callbacks.EarlyStopping
+    monitor: train-loss
+    min_delta: 0.001
+    patience: 5
+    check_on_train_epoch_end: true
+  model_checkpoint:
+    _target_: pytorch_lightning.callbacks.ModelCheckpoint
+    dirpath: ${hydra:runtime.output_dir}/checkpoints
+    monitor: train-loss
+    save_top_k: 1
+    save_on_train_epoch_end: true
+  log_images:
+    _target_: autoencoders.callbacks.LogReconstructedImagesCallback
+tags:
+- ${data.name}
+- ${model.name}