Keras RS example fixes (#2094)

* Keras RS example fixes

* Lower LR for Adam

* Doc-string change

* Format

Author: abheesht17

examples/keras_rs/dcn.py

@@ -122,10 +122,6 @@
     "batch_size": 1024,
 }
 
-LOOKUP_LAYERS = {
-    "int": keras.layers.IntegerLookup,
-    "str": keras.layers.StringLookup,
-}
 
 """
 Here, we define a helper function for visualising weights of the cross layer in
@@ -186,9 +182,7 @@ def print_stats(rmse_list, num_params, model_name):
     if num_trials == 1:
         print(f"{model_name}: RMSE = {avg_rmse}; #params = {num_params}")
     else:
-        print(
-            f"{model_name}: RMSE = {avg_rmse} ± {std_rmse}; " "#params = {num_params}"
-        )
+        print(f"{model_name}: RMSE = {avg_rmse} ± {std_rmse}; #params = {num_params}")
 
 
 """
@@ -275,6 +269,7 @@ def get_mixer_data(data_size=100_000):
         keras.layers.Dense(512, activation="relu"),
         keras.layers.Dense(256, activation="relu"),
         keras.layers.Dense(128, activation="relu"),
+        keras.layers.Dense(1),
     ]
 )
 

examples/keras_rs/deep_recommender.py

@@ -668,7 +668,7 @@ def __init__(
         self.update_candidates()  # Provide an initial set of candidates
         self.loss_fn = keras.losses.MeanSquaredError()
         self.top_k_metric = keras.metrics.SparseTopKCategoricalAccuracy(
-            k=100, from_sorted_ids=True
+            k=retrieval_k, from_sorted_ids=True
         )
 
     def update_candidates(self):

examples/keras_rs/listwise_ranking.py

@@ -182,16 +182,19 @@ def get_examples(
     }
     labels = []
     for user_id, user_list in sequences.items():
-        sampled_list = sample_sublist_from_list(
-            user_list,
-            num_examples_per_list,
-        )
-
-        inputs["user_id"].append(user_id)
-        inputs["movie_id"].append(
-            tf.convert_to_tensor([f["movie_id"] for f in sampled_list])
-        )
-        labels.append(tf.convert_to_tensor([f["user_rating"] for f in sampled_list]))
+        for _ in range(num_list_per_user):
+            sampled_list = sample_sublist_from_list(
+                user_list,
+                num_examples_per_list,
+            )
+
+            inputs["user_id"].append(user_id)
+            inputs["movie_id"].append(
+                tf.convert_to_tensor([f["movie_id"] for f in sampled_list])
+            )
+            labels.append(
+                tf.convert_to_tensor([f["user_rating"] for f in sampled_list])
+            )
 
     return (
         {"user_id": inputs["user_id"], "movie_id": inputs["movie_id"]},

examples/keras_rs/multi_task.py

@@ -143,7 +143,7 @@ def __init__(
         )
 
         # Rating model.
-        self.rating_model = tf.keras.Sequential(
+        self.rating_model = keras.Sequential(
             [
                 keras.layers.Dense(layer_size, activation="relu")
                 for layer_size in layer_sizes
@@ -162,7 +162,7 @@ def __init__(
 
         # Top-k accuracy for retrieval
         self.top_k_metric = keras.metrics.SparseTopKCategoricalAccuracy(
-            k=100, from_sorted_ids=True
+            k=10, from_sorted_ids=True
         )
         # RMSE for ranking
         self.rmse_metric = keras.metrics.RootMeanSquaredError()
@@ -306,7 +306,7 @@ def compute_metrics(self, x, y, y_pred, sample_weight=None):
     ranking_loss_wt=1.0,
     retrieval_loss_wt=0.0,
 )
-model.compile(optimizer=tf.keras.optimizers.Adagrad(0.1))
+model.compile(optimizer=keras.optimizers.Adagrad(0.1))
 model.fit(train_ratings, epochs=5)
 
 model.evaluate(test_ratings)
@@ -318,7 +318,7 @@ def compute_metrics(self, x, y, y_pred, sample_weight=None):
     ranking_loss_wt=0.0,
     retrieval_loss_wt=1.0,
 )
-model.compile(optimizer=tf.keras.optimizers.Adagrad(0.1))
+model.compile(optimizer=keras.optimizers.Adagrad(0.1))
 model.fit(train_ratings, epochs=5)
 
 model.evaluate(test_ratings)
@@ -330,7 +330,7 @@ def compute_metrics(self, x, y, y_pred, sample_weight=None):
     ranking_loss_wt=1.0,
     retrieval_loss_wt=1.0,
 )
-model.compile(optimizer=tf.keras.optimizers.Adagrad(0.1))
+model.compile(optimizer=keras.optimizers.Adagrad(0.1))
 model.fit(train_ratings, epochs=5)
 
 model.evaluate(test_ratings)

examples/keras_rs/sequential_retrieval.py

@@ -62,7 +62,7 @@
 TEST_BATCH_SIZE = 2048
 EMBEDDING_DIM = 32
 NUM_EPOCHS = 5
-LEARNING_RATE = 0.05
+LEARNING_RATE = 0.005
 
 """
 ## Dataset
@@ -368,7 +368,7 @@ def compute_loss(self, x, y, y_pred, sample_weight, training=True):
 """
 
 model = SequentialRetrievalModel(
-    movies_count=movies_count + 1, embedding_dimension=EMBEDDING_DIM
+    movies_count=movies_count, embedding_dimension=EMBEDDING_DIM
 )
 
 # Compile.