Fixed main.py

src/main.py

@@ -1,22 +1,31 @@
+import os
 import datasets
-import glob
 
-# List all .txt files in the current directory
-txt_files = glob.glob("*.txt")
+script_path = os.path.abspath(__file__)
+src_path = os.path.dirname(script_path)
+root_path = os.path.dirname(src_path)
+res_path = root_path+"/res"
 
-print(txt_files)
-
-# Read each file and store its content in a list
 texts = []
-for file_path in txt_files:
-    with open(file_path, 'r') as file:
-        texts.append(file.read())
+for name in os.listdir(res_path):
+    if name.endswith(".txt"):
+        with open(res_path+"/"+name, "r") as file:
+            text = file.read()
+            texts.append(text)
 
-# Create a Hugging Face dataset from the list of texts
 dataset = datasets.Dataset.from_dict({"text": texts})
+dataset = dataset.train_test_split(train_size=None, test_size=4, shuffle=False)
+
+train_val_dataset = dataset["train"]
+train_val_dataset = train_val_dataset.train_test_split(train_size=None, test_size=2, shuffle=False)
+
+train_dataset = train_val_dataset["train"]
+val_dataset = train_val_dataset["test"]
+test_dataset = dataset["test"]
+dataset = datasets.DatasetDict({
+    "train": train_dataset,
+    "val": val_dataset,
+    "test": test_dataset
+})
 
-dataset_splits = datasets.DatasetDict({
-    'train': train_split,
-    'validation': validation_split,
-    'test': test_split
-})
+dataset.push_to_hub("shakespearefirstfolio")