upgrades pytorch lightning

README.md

@@ -85,12 +85,16 @@ python -m autoregressive.scripts.wavenet_mnist classify --config "models/mnist_q
 ## Train
 To train / reproduce a model
 ```bash
-python -m autoregressive.scripts.train --config "models/mnist_q2/config.yaml"
+python -m autoregressive.scripts.train fit --config "models/mnist_q2/config.yaml"
 ```
 Progress is logged to Tensorboard
 ```
 tensorboard --logdir lightning_logs
 ```
+To generate a training configuration file for a specific dataset use
+```
+python -m autoregressive.scripts.train fit --data autoregressive.datasets.FSeriesDataModule --print_config > fseries_config.yaml
+```
 
 ## Test
 To run the tests

models/fseries_q127/config.yaml

@@ -1,11 +1,12 @@
 seed_everything: null
 trainer:
   logger: true
-  checkpoint_callback: true
+  checkpoint_callback: null
+  enable_checkpointing: true
   callbacks: null
   default_root_dir: null
-  gradient_clip_val: 0.0
-  gradient_clip_algorithm: norm
+  gradient_clip_val: null
+  gradient_clip_algorithm: null
   process_position: 0
   num_nodes: 1
   num_processes: 1
@@ -16,30 +17,32 @@ trainer:
   ipus: null
   log_gpu_memory: null
   progress_bar_refresh_rate: null
+  enable_progress_bar: true
   overfit_batches: 0.0
   track_grad_norm: -1
   check_val_every_n_epoch: 1
   fast_dev_run: false
-  accumulate_grad_batches: 1
+  accumulate_grad_batches: null
   max_epochs: 30
   min_epochs: null
-  max_steps: null
+  max_steps: -1
   min_steps: null
   max_time: null
   limit_train_batches: 1.0
   limit_val_batches: 1.0
   limit_test_batches: 1.0
   limit_predict_batches: 1.0
   val_check_interval: 1.0
-  flush_logs_every_n_steps: 100
+  flush_logs_every_n_steps: null
   log_every_n_steps: 25
   accelerator: null
+  strategy: null
   sync_batchnorm: false
   precision: 32
+  enable_model_summary: true
   weights_summary: top
   weights_save_path: null
   num_sanity_val_steps: 2
-  truncated_bptt_steps: null
   resume_from_checkpoint: null
   profiler: null
   benchmark: false
@@ -48,16 +51,16 @@ trainer:
   reload_dataloaders_every_epoch: false
   auto_lr_find: false
   replace_sampler_ddp: true
-  terminate_on_nan: false
+  detect_anomaly: false
   auto_scale_batch_size: false
-  prepare_data_per_node: true
+  prepare_data_per_node: null
   plugins: null
   amp_backend: native
-  amp_level: O2
-  distributed_backend: null
+  amp_level: null
   move_metrics_to_cpu: false
   multiple_trainloader_mode: max_size_cycle
   stochastic_weight_avg: false
+  terminate_on_nan: null
 model:
   wave_dilations:
   - 1

models/mnist_q2/config.yaml

@@ -1,11 +1,12 @@
 seed_everything: null
 trainer:
   logger: true
-  checkpoint_callback: true
+  checkpoint_callback: null
+  enable_checkpointing: true
   callbacks: null
   default_root_dir: null
-  gradient_clip_val: 0.0
-  gradient_clip_algorithm: norm
+  gradient_clip_val: null
+  gradient_clip_algorithm: null
   process_position: 0
   num_nodes: 1
   num_processes: 1
@@ -16,30 +17,32 @@ trainer:
   ipus: null
   log_gpu_memory: null
   progress_bar_refresh_rate: null
+  enable_progress_bar: true
   overfit_batches: 0.0
   track_grad_norm: -1
   check_val_every_n_epoch: 1
   fast_dev_run: false
-  accumulate_grad_batches: 1
-  max_epochs: 3000
+  accumulate_grad_batches: null
+  max_epochs: 30
   min_epochs: null
-  max_steps: null
+  max_steps: -1
   min_steps: null
   max_time: null
   limit_train_batches: 1.0
-  limit_val_batches: 0.2
+  limit_val_batches: 1.0
   limit_test_batches: 1.0
   limit_predict_batches: 1.0
   val_check_interval: 1.0
-  flush_logs_every_n_steps: 100
+  flush_logs_every_n_steps: null
   log_every_n_steps: 25
   accelerator: null
+  strategy: null
   sync_batchnorm: false
   precision: 32
+  enable_model_summary: true
   weights_summary: top
   weights_save_path: null
   num_sanity_val_steps: 2
-  truncated_bptt_steps: null
   resume_from_checkpoint: null
   profiler: null
   benchmark: false
@@ -48,16 +51,16 @@ trainer:
   reload_dataloaders_every_epoch: false
   auto_lr_find: false
   replace_sampler_ddp: true
-  terminate_on_nan: false
+  detect_anomaly: false
   auto_scale_batch_size: false
-  prepare_data_per_node: true
+  prepare_data_per_node: null
   plugins: null
   amp_backend: native
-  amp_level: O2
-  distributed_backend: null
+  amp_level: null
   move_metrics_to_cpu: false
   multiple_trainloader_mode: max_size_cycle
   stochastic_weight_avg: false
+  terminate_on_nan: null
 model:
   wave_dilations:
   - 1

models/mnist_q256/config.yaml

@@ -1,11 +1,12 @@
 seed_everything: null
 trainer:
   logger: true
-  checkpoint_callback: true
+  checkpoint_callback: null
+  enable_checkpointing: true
   callbacks: null
   default_root_dir: null
-  gradient_clip_val: 0.0
-  gradient_clip_algorithm: norm
+  gradient_clip_val: null
+  gradient_clip_algorithm: null
   process_position: 0
   num_nodes: 1
   num_processes: 1
@@ -16,30 +17,32 @@ trainer:
   ipus: null
   log_gpu_memory: null
   progress_bar_refresh_rate: null
+  enable_progress_bar: true
   overfit_batches: 0.0
   track_grad_norm: -1
   check_val_every_n_epoch: 1
   fast_dev_run: false
-  accumulate_grad_batches: 1
-  max_epochs: 3000
+  accumulate_grad_batches: null
+  max_epochs: 30
   min_epochs: null
-  max_steps: null
+  max_steps: -1
   min_steps: null
   max_time: null
   limit_train_batches: 1.0
-  limit_val_batches: 0.2
+  limit_val_batches: 1.0
   limit_test_batches: 1.0
   limit_predict_batches: 1.0
   val_check_interval: 1.0
-  flush_logs_every_n_steps: 100
+  flush_logs_every_n_steps: null
   log_every_n_steps: 25
   accelerator: null
+  strategy: null
   sync_batchnorm: false
   precision: 32
+  enable_model_summary: true
   weights_summary: top
   weights_save_path: null
   num_sanity_val_steps: 2
-  truncated_bptt_steps: null
   resume_from_checkpoint: null
   profiler: null
   benchmark: false
@@ -48,16 +51,16 @@ trainer:
   reload_dataloaders_every_epoch: false
   auto_lr_find: false
   replace_sampler_ddp: true
-  terminate_on_nan: false
+  detect_anomaly: false
   auto_scale_batch_size: false
-  prepare_data_per_node: true
+  prepare_data_per_node: null
   plugins: null
   amp_backend: native
-  amp_level: O2
-  distributed_backend: null
+  amp_level: null
   move_metrics_to_cpu: false
   multiple_trainloader_mode: max_size_cycle
   stochastic_weight_avg: false
+  terminate_on_nan: null
 model:
   wave_dilations:
   - 1

requirements/common.txt

@@ -1,6 +1,6 @@
-torch
-torchvision
-pytorch_lightning
+torch>=1.9.0
+torchvision>=0.10.0
+pytorch_lightning>=1.5.5
 jsonargparse[signatures]
 numpy
 matplotlib

requirements/dev.txt

@@ -1 +1,2 @@
-scipy
+scipy
+pytest