Finalized training parameters

neural_models.py

@@ -44,21 +44,22 @@ def compute_metrics(eval_pred):
 
 
 # load dataset
-food = load_dataset("food101", split="train[:5000]")
+fashion = load_dataset("fashion_mnist", split="train[:4000]")
 # Split into train/test sets
-food = food.train_test_split(test_size=0.2)
+fashion = fashion.train_test_split(test_size=0.2)
 # an example
-print(food["train"][0])
+print(fashion["train"][0])
 
 # Map label names to an integer and vice-versa
-labels = food["train"].features["label"].names
+labels = fashion["train"].features["label"].names
 label2id, id2label = dict(), dict()
 for i, label in enumerate(labels):
     label2id[label] = str(i)
     id2label[str(i)] = label
 
 # Should convert label id into a name
-print(id2label[str(79)])
+# print(label2id)
+# print(id2label)
 
 # Pre-processing with ViT
 # Load image processor to process image into tensor
@@ -88,8 +89,8 @@ val_data_augmentation = keras.Sequential(
     name="val_data_augmentation",
 )
 
-food["train"].set_transform(preprocess_train)
-food["test"].set_transform(preprocess_val)
+fashion["train"].set_transform(preprocess_train)
+fashion["test"].set_transform(preprocess_val)
 
 data_collator = DefaultDataCollator(return_tensors="tf")
 
@@ -97,8 +98,8 @@ accuracy = evaluate.load("accuracy")
 
 # Set hyperparameters
 batch_size = 16
-num_epochs = 5
-num_train_steps = len(food["train"]) * num_epochs
+num_epochs = 4
+num_train_steps = len(fashion["train"]) * num_epochs
 learning_rate = 3e-5
 weight_decay_rate = 0.01
 
@@ -118,11 +119,11 @@ model = TFAutoModelForImageClassification.from_pretrained(
 )
 
 # converting datasets to tf.data.Dataset
-tf_train_dataset = food["train"].to_tf_dataset(
+tf_train_dataset = fashion["train"].to_tf_dataset(
     columns="pixel_values", label_cols="label", shuffle=True, batch_size=batch_size, collate_fn=data_collator
 )
 
-tf_eval_dataset = food["test"].to_tf_dataset(
+tf_eval_dataset = fashion["test"].to_tf_dataset(
     columns="pixel_values", label_cols="label", shuffle=True, batch_size=batch_size, collate_fn=data_collator
 )
 
@@ -133,17 +134,19 @@ model.compile(optimizer=optimizer, loss=loss)
 
 metric_callback = KerasMetricCallback(metric_fn=compute_metrics, eval_dataset=tf_eval_dataset)
 push_to_hub_callback = PushToHubCallback(
-    output_dir="../food_classifier",
+    output_dir="../fashion_classifier",
     tokenizer=image_processor,
     save_strategy="no",
 )
 
 callbacks = [metric_callback, push_to_hub_callback]
 
-model.fit(tf_train_dataset, validation_data=tf_eval_dataset, epochs=num_epochs) #, callback=callbacks)
+model.fit(tf_train_dataset, validation_data=tf_eval_dataset, epochs=num_epochs, callbacks=callbacks)
 
-ds = load_dataset("food101", split="validation[:10]")
-image = ds["image"][0]
-classifier = pipeline("image-classification", model="my_awesome_food_model")
-print(classifier(image))
+# model.push_to_hub()
+
+# ds = load_dataset("fashion_mnist", split="test[:10]")
+# image = ds["image"][0]
+# classifier = pipeline("image-classification", model="my_awesome_fashion_model")
+# print(classifier(image))