VictorMylle/Thesis
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Autoregressive Quantile Training with Policy evaluation

Browse Source
This commit is contained in:
Victor Mylle
2024-02-21 18:11:38 +01:00
parent 2b22b6935e
commit f8823f7efa
6 changed files with 208 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

29
src/training_scripts/autoregressive_quantiles.py
View File

@@ -1,3 +1,5 @@
from src.policies.PolicyEvaluator import PolicyEvaluator
from src.policies.simple_baseline import BaselinePolicy, Battery
from src.models.lstm_model import GRUModel
from src.data import DataProcessor, DataConfig
from src.trainers.quantile_trainer import AutoRegressiveQuantileTrainer
@@ -68,12 +70,17 @@ model_parameters = task.connect(model_parameters, name="model_parameters")
time_embedding = TimeEmbedding(data_processor.get_time_feature_size(), model_parameters["time_feature_embedding"])
# lstm_model = GRUModel(time_embedding.output_dim(inputDim), len(quantiles), hidden_size=model_parameters["hidden_size"], num_layers=model_parameters["num_layers"], dropout=model_parameters["dropout"])
# non_linear_model = NonLinearRegression(time_embedding.output_dim(inputDim), len(quantiles), hiddenSize=model_parameters["hidden_size"], numLayers=model_parameters["num_layers"], dropout=model_parameters["dropout"])
linear_model = LinearRegression(time_embedding.output_dim(inputDim), len(quantiles))
non_linear_model = NonLinearRegression(time_embedding.output_dim(inputDim), len(quantiles), hiddenSize=model_parameters["hidden_size"], numLayers=model_parameters["num_layers"], dropout=model_parameters["dropout"])
# linear_model = LinearRegression(time_embedding.output_dim(inputDim), len(quantiles))
model = nn.Sequential(time_embedding, linear_model)
model = nn.Sequential(time_embedding, non_linear_model)
optimizer = torch.optim.Adam(model.parameters(), lr=model_parameters["learning_rate"])
### Policy Evaluator ###
battery = Battery(2, 1)
baseline_policy = BaselinePolicy(battery, data_path="")
policy_evaluator = PolicyEvaluator(baseline_policy, task)
#### Trainer ####
trainer = AutoRegressiveQuantileTrainer(
model,
@@ -82,12 +89,24 @@ trainer = AutoRegressiveQuantileTrainer(
data_processor,
quantiles,
"cuda",
policy_evaluator=policy_evaluator,
debug=False,
)
trainer.add_metrics_to_track(
[PinballLoss(quantiles), MSELoss(), L1Loss(), CRPSLoss(quantiles)]
)
trainer.early_stopping(patience=30)
trainer.early_stopping(patience=10)
trainer.plot_every(5)
trainer.train(task=task, epochs=epochs, remotely=True)
trainer.train(task=task, epochs=epochs, remotely=False)
### Policy Evaluation ###
idx_samples = trainer.test_set_samples
_, test_loader = trainer.data_processor.get_dataloaders(
predict_sequence_length=trainer.model.output_size)
policy_evaluator.evaluate_test_set(idx_samples, test_loader)
policy_evaluator.plot_profits_table()
policy_evaluator.plot_thresholds_per_day()
task.close()