new runs added Merge remote-tracking branch 'origin/runs'

codes/models/InceptionTime.py

@@ -0,0 +1,101 @@
+from tsai.all import *
+import fastai
+from tsai.models.InceptionTime import InceptionTime
+import numpy as np
+
+class InceptionTime_classifier:
+    """
+    InceptionTime classifier for time series classification
+
+    Parameters  
+    ----------
+    c_in: int
+        Number of input channels
+    c_out: int
+        Number of classes
+    seq_len: int
+        Length of the input sequence
+    nf: int
+        Number of filters
+    nb_filters: int
+        Number of filters in the inception module   
+    ks: int
+        Kernel size 
+    bottleneck: bool
+        Whether to use bottleneck
+
+    Attributes
+    ----------
+    model: nn.Module
+        InceptionTime model
+    """
+
+    def __init__(
+        self,
+        c_in,
+        c_out,
+        seq_len,
+        nf=32,
+        nb_filters=32,
+        ks=40,
+        bottleneck=True,
+        **kwargs
+    ):
+        self.model = InceptionTime(
+            c_in=c_in,
+            c_out=c_out,
+            seq_len=seq_len,
+            nf=nf,
+            nb_filters=nb_filters,
+            ks=ks,
+            bottleneck=bottleneck,
+        )
+
+    def fit(
+            self, 
+            X: np.ndarray,
+            y: np.ndarray,
+            splits: list,
+            test_data: dict,
+            epochs: int,
+            lr: float
+    ):
+        """
+        Fit the model on the training data
+
+        Parameters
+        ----------
+        X: np.ndarray
+            Input data
+        y: np.ndarray
+            Target data
+        splits: list
+            List of indices for training and validation splits
+        test_data: dict
+            Dictionary of test data
+            {'X': X_test, 'y': y_test}
+        epochs: int
+            Number of epochs to train
+        lr: float
+            Learning rate
+        """
+        tfms = [None, [Categorize()]]
+        dsets = TSDatasets(X, y, tfms=tfms, splits=splits, inplace=True)
+        dls = TSDataLoaders.from_dsets(dsets.train, dsets.valid, bs=[64, 128], batch_tfms=[TSStandardize()], num_workers=0)
+        learn = Learner(dls, self.model, metrics=fastai.metrics.accuracy)
+        learn.fit_one_cycle(epochs, lr)
+
+        #now add the test data to the dls
+        test_ds = dls.valid.dataset.add_test(**test_data)
+        test_dl = dls.valid.new(test_ds)
+
+        #get the predictions
+        _, test_targets, test_preds = learn.get_preds(dl=test_dl, with_decoded=True)
+        accuracy = skm.accuracy_score(test_targets, test_preds)
+        cl_report = skm.classification_report(test_targets, test_preds, output_dict=True)
+        return {"accuracy": accuracy, "cl_report": cl_report, "target": test_targets, "pred": test_preds}
+
+
+if __name__ == "__main__":
+    # instantiate the model
+    model = InceptionTime_classifier(c_in=1, c_out=2, seq_len=100)

configs/main_config.yaml

@@ -1,7 +1,7 @@
 defaults:
   - _self_
   - datasets: dataset_params
-  - models: LSTM_FCN
+  - models: InceptionTime
   - [email protected]: ${models}
   - override hydra/sweeper: optuna
 dataset_name:

configs/models/InceptionTime.yaml

@@ -0,0 +1,6 @@
+model:
+  _target_: codes.models.InceptionTime.InceptionTime_classifier
+  nf: 32
+  nb_filters: 32
+  ks: 40
+  bottleneck: True

configs/search_spaces/InceptionTime.yaml

@@ -0,0 +1,4 @@
+models.model.nf: choice(8, 16, 32, 64, 128)
+models.model.nb_filters: choice(8, 16, 32, 64, 128)
+models.model.ks: choice(10, 20, 40, 80, 160)
+models.model.bottleneck: bool (True, False)

mlruns/952712754232464717/00714c6614ac4a79b5bd18121c2c5e2a/artifacts/classification_report.json

@@ -0,0 +1,27 @@
+{
+  "0": {
+    "precision": 0.8205128205128205,
+    "recall": 0.8888888888888888,
+    "f1-score": 0.8533333333333333,
+    "support": 36
+  },
+  "1": {
+    "precision": 0.9344262295081968,
+    "recall": 0.890625,
+    "f1-score": 0.9120000000000001,
+    "support": 64
+  },
+  "accuracy": 0.89,
+  "macro avg": {
+    "precision": 0.8774695250105087,
+    "recall": 0.8897569444444444,
+    "f1-score": 0.8826666666666667,
+    "support": 100
+  },
+  "weighted avg": {
+    "precision": 0.8934174022698613,
+    "recall": 0.89,
+    "f1-score": 0.8908800000000001,
+    "support": 100
+  }
+}

mlruns/952712754232464717/00714c6614ac4a79b5bd18121c2c5e2a/artifacts/model_params.json

@@ -0,0 +1,6 @@
+{
+  "nf": 128,
+  "nb_filters": 16,
+  "ks": 80,
+  "bottleneck": false
+}

mlruns/952712754232464717/00714c6614ac4a79b5bd18121c2c5e2a/meta.yaml

@@ -0,0 +1,15 @@
+artifact_uri: file:///home/pops/myProjects/ts_classification_mlflow_hydra_optuna/mlruns/952712754232464717/00714c6614ac4a79b5bd18121c2c5e2a/artifacts
+end_time: 1710766329484
+entry_point_name: ''
+experiment_id: '952712754232464717'
+lifecycle_stage: active
+run_id: 00714c6614ac4a79b5bd18121c2c5e2a
+run_name: ECG200
+run_uuid: 00714c6614ac4a79b5bd18121c2c5e2a
+source_name: ''
+source_type: 4
+source_version: ''
+start_time: 1710766329473
+status: 3
+tags: []
+user_id: pops

mlruns/952712754232464717/00714c6614ac4a79b5bd18121c2c5e2a/metrics/accuracy

@@ -0,0 +1 @@
+1710766329478 0.89 0

mlruns/952712754232464717/00714c6614ac4a79b5bd18121c2c5e2a/metrics/f1

@@ -0,0 +1 @@
+1710766329479 0.8908800000000001 0

mlruns/952712754232464717/00714c6614ac4a79b5bd18121c2c5e2a/metrics/macro_f1_score

@@ -0,0 +1 @@
+1710766329480 0.8826666666666667 0

mlruns/952712754232464717/00714c6614ac4a79b5bd18121c2c5e2a/metrics/precision

@@ -0,0 +1 @@
+1710766329479 0.8774695250105087 0

mlruns/952712754232464717/00714c6614ac4a79b5bd18121c2c5e2a/metrics/recall

@@ -0,0 +1 @@
+1710766329480 0.8897569444444444 0

mlruns/952712754232464717/00714c6614ac4a79b5bd18121c2c5e2a/metrics/support

@@ -0,0 +1 @@
+1710766329480 100.0 0

mlruns/952712754232464717/00714c6614ac4a79b5bd18121c2c5e2a/params/epochs

@@ -0,0 +1 @@
+500

mlruns/952712754232464717/00714c6614ac4a79b5bd18121c2c5e2a/params/lr

@@ -0,0 +1 @@
+0.001

mlruns/952712754232464717/00714c6614ac4a79b5bd18121c2c5e2a/params/model_name

@@ -0,0 +1 @@
+InceptionTime

mlruns/952712754232464717/00714c6614ac4a79b5bd18121c2c5e2a/tags/mlflow.runName

@@ -0,0 +1 @@
+ECG200

mlruns/952712754232464717/00714c6614ac4a79b5bd18121c2c5e2a/tags/mlflow.source.git.commit

@@ -0,0 +1 @@
+fe8042a2f33aa638b2c8945a75a751b3357866df

mlruns/952712754232464717/00714c6614ac4a79b5bd18121c2c5e2a/tags/mlflow.source.name

@@ -0,0 +1 @@
+main.py

mlruns/952712754232464717/00714c6614ac4a79b5bd18121c2c5e2a/tags/mlflow.source.type

@@ -0,0 +1 @@
+LOCAL

mlruns/952712754232464717/00714c6614ac4a79b5bd18121c2c5e2a/tags/mlflow.user

@@ -0,0 +1 @@
+pops

mlruns/952712754232464717/038fa8dac5a84d95842602f3f02dd92e/artifacts/classification_report.json

@@ -0,0 +1,27 @@
+{
+  "0": {
+    "precision": 0.7804878048780488,
+    "recall": 0.8888888888888888,
+    "f1-score": 0.8311688311688312,
+    "support": 36
+  },
+  "1": {
+    "precision": 0.9322033898305084,
+    "recall": 0.859375,
+    "f1-score": 0.8943089430894309,
+    "support": 64
+  },
+  "accuracy": 0.87,
+  "macro avg": {
+    "precision": 0.8563455973542786,
+    "recall": 0.8741319444444444,
+    "f1-score": 0.862738887129131,
+    "support": 100
+  },
+  "weighted avg": {
+    "precision": 0.8775857792476229,
+    "recall": 0.87,
+    "f1-score": 0.871578502798015,
+    "support": 100
+  }
+}

mlruns/952712754232464717/038fa8dac5a84d95842602f3f02dd92e/artifacts/model_params.json

@@ -0,0 +1,6 @@
+{
+  "nf": 128,
+  "nb_filters": 16,
+  "ks": 80,
+  "bottleneck": false
+}

mlruns/952712754232464717/038fa8dac5a84d95842602f3f02dd92e/meta.yaml

@@ -0,0 +1,15 @@
+artifact_uri: file:///home/pops/myProjects/ts_classification_mlflow_hydra_optuna/mlruns/952712754232464717/038fa8dac5a84d95842602f3f02dd92e/artifacts
+end_time: 1710766637731
+entry_point_name: ''
+experiment_id: '952712754232464717'
+lifecycle_stage: active
+run_id: 038fa8dac5a84d95842602f3f02dd92e
+run_name: ECG200
+run_uuid: 038fa8dac5a84d95842602f3f02dd92e
+source_name: ''
+source_type: 4
+source_version: ''
+start_time: 1710766637721
+status: 3
+tags: []
+user_id: pops

mlruns/952712754232464717/038fa8dac5a84d95842602f3f02dd92e/metrics/accuracy

@@ -0,0 +1 @@
+1710766637725 0.87 0

mlruns/952712754232464717/038fa8dac5a84d95842602f3f02dd92e/metrics/f1

@@ -0,0 +1 @@
+1710766637726 0.871578502798015 0

mlruns/952712754232464717/038fa8dac5a84d95842602f3f02dd92e/metrics/macro_f1_score

@@ -0,0 +1 @@
+1710766637728 0.862738887129131 0

mlruns/952712754232464717/038fa8dac5a84d95842602f3f02dd92e/metrics/precision

@@ -0,0 +1 @@
+1710766637726 0.8563455973542786 0

mlruns/952712754232464717/038fa8dac5a84d95842602f3f02dd92e/metrics/recall

@@ -0,0 +1 @@
+1710766637727 0.8741319444444444 0

mlruns/952712754232464717/038fa8dac5a84d95842602f3f02dd92e/metrics/support

@@ -0,0 +1 @@
+1710766637727 100.0 0

mlruns/952712754232464717/038fa8dac5a84d95842602f3f02dd92e/params/epochs

@@ -0,0 +1 @@
+500

mlruns/952712754232464717/038fa8dac5a84d95842602f3f02dd92e/params/lr

@@ -0,0 +1 @@
+0.001

mlruns/952712754232464717/038fa8dac5a84d95842602f3f02dd92e/params/model_name

@@ -0,0 +1 @@
+InceptionTime

mlruns/952712754232464717/038fa8dac5a84d95842602f3f02dd92e/tags/mlflow.runName

@@ -0,0 +1 @@
+ECG200

mlruns/952712754232464717/038fa8dac5a84d95842602f3f02dd92e/tags/mlflow.source.git.commit

@@ -0,0 +1 @@
+fe8042a2f33aa638b2c8945a75a751b3357866df

mlruns/952712754232464717/038fa8dac5a84d95842602f3f02dd92e/tags/mlflow.source.name

@@ -0,0 +1 @@
+main.py

mlruns/952712754232464717/038fa8dac5a84d95842602f3f02dd92e/tags/mlflow.source.type

@@ -0,0 +1 @@
+LOCAL

mlruns/952712754232464717/038fa8dac5a84d95842602f3f02dd92e/tags/mlflow.user

@@ -0,0 +1 @@
+pops

mlruns/952712754232464717/05edaa4acc1548f087835c211c9f5f49/artifacts/classification_report.json

@@ -0,0 +1,27 @@
+{
+  "0": {
+    "precision": 0.8378378378378378,
+    "recall": 0.8611111111111112,
+    "f1-score": 0.8493150684931507,
+    "support": 36
+  },
+  "1": {
+    "precision": 0.9206349206349206,
+    "recall": 0.90625,
+    "f1-score": 0.9133858267716536,
+    "support": 64
+  },
+  "accuracy": 0.89,
+  "macro avg": {
+    "precision": 0.8792363792363792,
+    "recall": 0.8836805555555556,
+    "f1-score": 0.8813504476324021,
+    "support": 100
+  },
+  "weighted avg": {
+    "precision": 0.8908279708279707,
+    "recall": 0.89,
+    "f1-score": 0.8903203537913926,
+    "support": 100
+  }
+}

mlruns/952712754232464717/05edaa4acc1548f087835c211c9f5f49/artifacts/model_params.json

@@ -0,0 +1,6 @@
+{
+  "nf": 128,
+  "nb_filters": 128,
+  "ks": 40,
+  "bottleneck": false
+}

mlruns/952712754232464717/05edaa4acc1548f087835c211c9f5f49/meta.yaml

@@ -0,0 +1,15 @@
+artifact_uri: file:///home/pops/myProjects/ts_classification_mlflow_hydra_optuna/mlruns/952712754232464717/05edaa4acc1548f087835c211c9f5f49/artifacts
+end_time: 1710766537806
+entry_point_name: ''
+experiment_id: '952712754232464717'
+lifecycle_stage: active
+run_id: 05edaa4acc1548f087835c211c9f5f49
+run_name: ECG200
+run_uuid: 05edaa4acc1548f087835c211c9f5f49
+source_name: ''
+source_type: 4
+source_version: ''
+start_time: 1710766537795
+status: 3
+tags: []
+user_id: pops

mlruns/952712754232464717/05edaa4acc1548f087835c211c9f5f49/metrics/accuracy

@@ -0,0 +1 @@
+1710766537800 0.89 0

mlruns/952712754232464717/05edaa4acc1548f087835c211c9f5f49/metrics/f1

@@ -0,0 +1 @@
+1710766537800 0.8903203537913926 0

mlruns/952712754232464717/05edaa4acc1548f087835c211c9f5f49/metrics/macro_f1_score

@@ -0,0 +1 @@
+1710766537802 0.8813504476324021 0

mlruns/952712754232464717/05edaa4acc1548f087835c211c9f5f49/metrics/precision

@@ -0,0 +1 @@
+1710766537801 0.8792363792363792 0

mlruns/952712754232464717/05edaa4acc1548f087835c211c9f5f49/metrics/recall

@@ -0,0 +1 @@
+1710766537801 0.8836805555555556 0

mlruns/952712754232464717/05edaa4acc1548f087835c211c9f5f49/metrics/support

@@ -0,0 +1 @@
+1710766537802 100.0 0

mlruns/952712754232464717/05edaa4acc1548f087835c211c9f5f49/params/epochs

@@ -0,0 +1 @@
+500

mlruns/952712754232464717/05edaa4acc1548f087835c211c9f5f49/params/lr

@@ -0,0 +1 @@
+0.001

mlruns/952712754232464717/05edaa4acc1548f087835c211c9f5f49/params/model_name

@@ -0,0 +1 @@
+InceptionTime

mlruns/952712754232464717/05edaa4acc1548f087835c211c9f5f49/tags/mlflow.runName

@@ -0,0 +1 @@
+ECG200

mlruns/952712754232464717/05edaa4acc1548f087835c211c9f5f49/tags/mlflow.source.git.commit

@@ -0,0 +1 @@
+fe8042a2f33aa638b2c8945a75a751b3357866df

mlruns/952712754232464717/05edaa4acc1548f087835c211c9f5f49/tags/mlflow.source.name

@@ -0,0 +1 @@
+main.py

mlruns/952712754232464717/05edaa4acc1548f087835c211c9f5f49/tags/mlflow.source.type

@@ -0,0 +1 @@
+LOCAL

mlruns/952712754232464717/05edaa4acc1548f087835c211c9f5f49/tags/mlflow.user

@@ -0,0 +1 @@
+pops