diff --git a/pytorch/main.py b/pytorch/main.py
index e69de29..134a217 100644
--- a/pytorch/main.py
+++ b/pytorch/main.py
@@ -0,0 +1,52 @@
+import requests
+import json
+import base64
+
+class GitHubDatasetGenerator:
+    def __init__(self, username, token):
+        self.username = username
+        self.token = token
+        self.base_url = "https://api.github.com"
+        self.headers = {'Authorization': f'token {self.token}'}
+
+    def get_repos(self):
+        url = f"{self.base_url}/users/{self.username}/repos"
+        response = requests.get(url, headers=self.headers)
+        return json.loads(response.text)
+
+    def get_files(self, repo_name):
+        url = f"{self.base_url}/repos/{self.username}/{repo_name}/contents"
+        response = requests.get(url, headers=self.headers)
+        return json.loads(response.text)
+
+    def get_file_content(self, download_url):
+        response = requests.get(download_url)
+        return response.text
+
+    def generate_dataset(self):
+        dataset = []
+        repos = self.get_repos()
+        for repo in repos:
+            repo_name = repo['name']
+            files = self.get_files(repo_name)
+            for file in files:
+                if file['name'].endswith('.py') or file['name'] == 'README.md':
+                    file_content = self.get_file_content(file['download_url'])
+                    dataset.append({
+                        'repo': repo_name,
+                        'file': file['name'],
+                        'url': file['html_url'],
+                        'content': file_content
+                    })
+        return dataset
+
+    def save_dataset(self, dataset, filename):
+        with open(filename, 'w') as f:
+            for data in dataset:
+                f.write(json.dumps(data))
+                f.write('\n')
+
+
+# generator = GitHubDatasetGenerator('username', 'token')
+# dataset = generator.generate_dataset()
+# generator.save_dataset(dataset, 'dataset.jsonl')
\ No newline at end of file