Upload folder using huggingface_hub

README.md

@@ -14,7 +14,7 @@ The code used to train this model is given in `train.py`.
 
 ```bash
 conda env create -n mmgp_tensile2d -f https://huggingface.co/fabiencasenave/mmgp_tensile2d/resolve/main/environment.yml
-conda mmgp_tensile2d
+conda activate mmgp_tensile2d
 pip install git+https://huggingface.co/fabiencasenave/mmgp_tensile2d
 ```
 
@@ -28,9 +28,9 @@ import mmgp_tensile2d
 model = mmgp_tensile2d.load()
 
 hf_dataset = load_dataset("PLAID-datasets/Tensile2d", split="all_samples")
-ids_test = hf_dataset.description["split"]['test']
+ids_test = hf_dataset.description["split"]['test'][:5]
 
-dataset_test, _ = huggingface_dataset_to_plaid(hf_dataset, ids = ids_test, processes_number = 6, verbose = True)
+dataset_test, _ = huggingface_dataset_to_plaid(hf_dataset, ids = ids_test, processes_number = 5, verbose = True)
 
 print("Check the 'U1' field is not present: dataset_test[0].get_field('U1') =", dataset_test[0].get_field('U1'))
 

pipeline.joblib

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b9950a0b4b472fd87721f54fe271aec7da67cec3afef8045f01e319a6d24cbe3
-size 1996169
+oid sha256:39bf5857f390d114fede4a147e156f41c95d860b7870186f2e6cea83bc1ed258
+size 71519577

test.py

@@ -5,9 +5,9 @@ import mmgp_tensile2d
 model = mmgp_tensile2d.load()
 
 hf_dataset = load_dataset("PLAID-datasets/Tensile2d", split="all_samples")
-ids_test = hf_dataset.description["split"]['test']
+ids_test = hf_dataset.description["split"]['test'][:5]
 
-dataset_test, _ = huggingface_dataset_to_plaid(hf_dataset, ids = ids_test, processes_number = 6, verbose = True)
+dataset_test, _ = huggingface_dataset_to_plaid(hf_dataset, ids = ids_test, processes_number = 5, verbose = True)
 
 print("Check the 'U1' field is not present: dataset_test[0].get_field('U1') =", dataset_test[0].get_field('U1'))
 

train.py

@@ -30,12 +30,12 @@ from mmgp.pipelines.mmgp_blocks import MMGPPreparer, MMGPTransformer
 from mmgp_tensile2d.utils import length_scale_init, morphing
 
 
-n_processes = min(max(1, os.cpu_count()), 8)
+n_processes = min(max(1, os.cpu_count()), 24)
 
 
 # load dataset
 hf_dataset = load_dataset("PLAID-datasets/Tensile2d", split="all_samples")
-ids_train = hf_dataset.description["split"]['train_500'][:8]
+ids_train = hf_dataset.description["split"]['train_500']
 
 dataset_train, _ = huggingface_dataset_to_plaid(hf_dataset, ids = ids_train, processes_number = n_processes, verbose = True)
 
@@ -58,7 +58,7 @@ input_scalar_scaler = WrappedPlaidSklearnTransformer(MinMaxScaler(), **config['i
 nodes_preprocessor = Pipeline(
     steps=[
         ("mmgp_nodes_transf", MMGPTransformer(**config['mmgp_nodes_transf'])),
-        ('pca_nodes', WrappedPlaidSklearnTransformer(PCA(n_components=4), **config['pca_nodes'])),
+        ('pca_nodes', WrappedPlaidSklearnTransformer(PCA(), **config['pca_nodes'])),
     ]
 )
 
@@ -82,7 +82,7 @@ kernel = Matern(length_scale_bounds=(1e-8, 1e8), nu = 2.5)
 gpr = GaussianProcessRegressor(
     kernel=kernel,
     optimizer='fmin_l_bfgs_b',
-    n_restarts_optimizer=1,
+    n_restarts_optimizer=2,
     random_state=42)
 
 reg = MultiOutputRegressor(gpr)
@@ -95,7 +95,7 @@ regressor = WrappedPlaidSklearnRegressor(reg, **config['regressor_mach'], dynami
 postprocessor = Pipeline(
     steps=[
         ("mmgp_u1_transf", MMGPTransformer(**config['mmgp_u1_transf'])),
-        ('pca_u1', WrappedPlaidSklearnTransformer(PCA(n_components=4), **config['pca_u1'])),
+        ('pca_u1', WrappedPlaidSklearnTransformer(PCA(), **config['pca_u1'])),
     ]
 )
 
@@ -103,7 +103,6 @@ postprocessor = Pipeline(
 target_regressor = PlaidTransformedTargetRegressor(
     regressor=regressor,
     transformer=postprocessor,
-    # out_features_identifiers = config['pca_u1']['in_features_identifiers']
 )
 
 pipeline = Pipeline(
@@ -117,8 +116,8 @@ pipeline = Pipeline(
 
 # Set hyperameter that have been optimized by cross-valdiation on the training set
 optimized_pipeline = clone(pipeline).set_params(
-    preprocessor__column_preprocessor__nodes_preprocessor__pca_nodes__sklearn_block__n_components = 4,
-    regressor__transformer__pca_u1__sklearn_block__n_components = 8
+    preprocessor__column_preprocessor__nodes_preprocessor__pca_nodes__sklearn_block__n_components = 16,
+    regressor__transformer__pca_u1__sklearn_block__n_components = 32
 )
 
 # Train the model