add application files

.gitattributes

@@ -1,35 +1,36 @@
-*.7z filter=lfs diff=lfs merge=lfs -text
-*.arrow filter=lfs diff=lfs merge=lfs -text
-*.bin filter=lfs diff=lfs merge=lfs -text
-*.bz2 filter=lfs diff=lfs merge=lfs -text
-*.ckpt filter=lfs diff=lfs merge=lfs -text
-*.ftz filter=lfs diff=lfs merge=lfs -text
-*.gz filter=lfs diff=lfs merge=lfs -text
-*.h5 filter=lfs diff=lfs merge=lfs -text
-*.joblib filter=lfs diff=lfs merge=lfs -text
-*.lfs.* filter=lfs diff=lfs merge=lfs -text
-*.mlmodel filter=lfs diff=lfs merge=lfs -text
-*.model filter=lfs diff=lfs merge=lfs -text
-*.msgpack filter=lfs diff=lfs merge=lfs -text
-*.npy filter=lfs diff=lfs merge=lfs -text
-*.npz filter=lfs diff=lfs merge=lfs -text
-*.onnx filter=lfs diff=lfs merge=lfs -text
-*.ot filter=lfs diff=lfs merge=lfs -text
-*.parquet filter=lfs diff=lfs merge=lfs -text
-*.pb filter=lfs diff=lfs merge=lfs -text
-*.pickle filter=lfs diff=lfs merge=lfs -text
-*.pkl filter=lfs diff=lfs merge=lfs -text
-*.pt filter=lfs diff=lfs merge=lfs -text
-*.pth filter=lfs diff=lfs merge=lfs -text
-*.rar filter=lfs diff=lfs merge=lfs -text
-*.safetensors filter=lfs diff=lfs merge=lfs -text
-saved_model/**/* filter=lfs diff=lfs merge=lfs -text
-*.tar.* filter=lfs diff=lfs merge=lfs -text
-*.tar filter=lfs diff=lfs merge=lfs -text
-*.tflite filter=lfs diff=lfs merge=lfs -text
-*.tgz filter=lfs diff=lfs merge=lfs -text
-*.wasm filter=lfs diff=lfs merge=lfs -text
-*.xz filter=lfs diff=lfs merge=lfs -text
-*.zip filter=lfs diff=lfs merge=lfs -text
-*.zst filter=lfs diff=lfs merge=lfs -text
-*tfevents* filter=lfs diff=lfs merge=lfs -text
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text
+models/** filter=lfs diff=lfs merge=lfs -text

.gitignore

@@ -0,0 +1,158 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+lab/
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+database.db
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+#poetry.lock
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#pdm.lock
+#   pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
+#   in version control.
+#   https://pdm.fming.dev/latest/usage/project/#working-with-version-control
+.pdm.toml
+.pdm-python
+.pdm-build/
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/

LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2024 Alireza
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

README.md

@@ -1,13 +1,83 @@
----
-title: Gpt Tools
-emoji: 👁
-colorFrom: gray
-colorTo: indigo
-sdk: gradio
-sdk_version: 5.12.0
-app_file: app.py
-pinned: false
-license: mit
----
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# GPT Tools
+
+![GUI)](https://github.com/user-attachments/assets/6a845c99-6ecc-422f-b662-8069cb5c2324)
+
+---
+این پروژه یک رابط کاربری گرافیکی زیبا و کاربردی برای تولید متن، کد، داستان‌های تعاملی و ارزیابی مدل‌های مختلف مانند GPT-2 و CodeGen ارائه می‌دهد. با استفاده از این ابزار می‌توانید به راحتی مدل‌های زبان طبیعی را مدیریت و از خروجی‌های آن بهره‌برداری کنید
+
+---
+
+## **🚨 Requirements**
+
+این پروژه برای اجرا نیاز به **پایتون نسخه 3.8.6** دارد. لطفاً اطمینان حاصل کنید که نسخه صحیح پایتون روی سیستم شما نصب است.  
+برای بررسی نسخه پایتون، دستور زیر را در خط فرمان اجرا کنید:
+```bash
+python --version
+```
+
+---
+
+## **💫 Main features**
+
+- تولید متن: تولید متن‌های خلاقانه با استفاده از مدل‌های مختلف GPT-2.
+- تولید کد: تولید کدهای برنامه‌نویسی با مدل CodeGen از طریق ورودی‌های توصیفی.
+- داستان‌های تعاملی: ایجاد داستان‌های سفارشی و خلاقانه با همکاری مدل.
+- مدیریت مدل‌ها: دانلود و ذخیره مدل‌ها در مسیرهای سفارشی.
+- آموزش مدل‌ها: آموزش مجدد مدل‌ها با داده‌های دلخواه و ذخیره‌سازی تغییرات.
+---
+## **📁 Project Structure**
+```bash
+.
+├── app.py                 # رابط کاربری گرافیکی (Gradio)
+├── model.py               # مدیریت و بارگذاری مدل‌ها
+├── generate.py            # منطق تولید متن و کد
+├── train.py               # آموزش مجدد مدل‌ها
+├── database.py            # مدیریت پایگاه داده برای ذخیره ورودی‌ها
+├── models/          # مسیر پیش‌فرض برای ذخیره مدل‌ها
+└── lib.txt       # لیست کتابخانه‌های موردنیاز
+```
+---
+## **🚀 Installation and setup**
+
+### **نصب پایتون 3.8.6**
+اگر پایتون نسخه 3.8.6 روی سیستم شما نصب نیست، از صفحه دانلود  آن را نصب کنید.
+در سیستم‌های لینوکسی می‌توانید از دستورات زیر استفاده کنید
+```bash
+sudo apt update
+sudo apt install python3.8
+```
+---
+### **کلون کردن**
+ابتدا مخزن پروژه رو کلون کنید
+```bash
+git clone https://github.com/ali0discord/gpt-text-generator.git
+cd gpt-text-generator
+```
+---
+### **نصب کتابخانه ها**
+با این دستور کتابخانه ها رو نصب کنید
+```bash
+pip install -r lib.txt
+```
+---
+### **دانلود مدل ها**
+با اجرای این فایل مدل های مورد نیاز به صورت خودکار دانلود و در دایرکتوری مخصوص ذخیره میشوند
+```bash
+python download.py
+```
+---
+### **اجرای کدها**
+با دستور زیر کد ها رو اجرا کنید
+```bash
+python app.py
+```
+پس از اجرای کد ها به صورت کامل وارد این آدرش شوید
+```bash
+127.0.0.1:7860
+```
+---
+## **گزارش باگ ها**
+از طریق بخش Issues گیت هاب با ما در ارتباط باشید
+
+---
+### **جان گرفته از علیرضا**

app.py

@@ -0,0 +1,206 @@
+import gradio as gr
+from database import create_db
+from functions import *
+from functions import _generate_code
+
+# Supported models
+models_options_general = ['GPT2', 'GPT2-medium', 'GPT2-large', 'GPT2-persian', 'GPT-Neo-125M']
+models_options_codegen = ['codegen']
+models_options_chatbot = ['dialoGPT', 'dialoGPT-medium', 'dialoGPT-large']
+
+# Create database
+create_db()
+
+# Interface setup
+with gr.Blocks() as interface:
+    gr.Markdown(
+        "# **GPT Tools**\n\n"
+        "Generate something using GPT models. Select the model and adjust the parameters for optimal results."
+    )
+    with gr.Tabs():
+        with gr.Tab("Text Generator"):
+            with gr.Row():
+                with gr.Column(scale=1, min_width=350):
+                    input_text = gr.Textbox(label="Input Text", placeholder="Enter your text here...", lines=4, max_lines=6)
+                    selected_model = gr.Radio(choices=models_options_general, value="GPT2", label="Select Model", type="value")
+                    with gr.Row():
+                        max_tokens = gr.Slider(10, 100, value=50, step=1, label="Max New Tokens", interactive=True)
+                with gr.Column(scale=1, min_width=350):
+                    output_text = gr.Textbox(label="Generated Text", interactive=False, lines=8, max_lines=12)
+                    generate_button = gr.Button("Generate Text", variant="primary")
+
+            generate_button.click(
+                generate,
+                inputs=[input_text, selected_model, max_tokens],
+                outputs=output_text,
+            )
+
+
+        with gr.Tab("Multiverse Story Generator"):
+            with gr.Row():
+                with gr.Column(scale=1, min_width=350):
+                    input_text = gr.Textbox(label="Enter your story idea", placeholder="e.g. A scientist discovers a parallel universe...", lines=4, max_lines=6)
+                    selected_model = gr.Radio(choices=models_options_general, value="GPT2", label="Select Model for Story Generation", type="value")
+                    max_length = gr.Slider(50, 300, value=150, step=1, label="Max Length", interactive=True)
+
+                with gr.Column(scale=1, min_width=350):
+                    output_text = gr.Textbox(label="Generated Worlds", interactive=False, lines=12, max_lines=20)
+                    generate_button = gr.Button("Generate Parallel Worlds", variant="primary")
+
+            generate_button.click(
+                generate_multiverse,
+                inputs=[input_text, selected_model, max_length],
+                outputs=output_text,
+            )
+
+        with gr.Tab("Interactive Story Writing"):
+            with gr.Row():
+                with gr.Column(scale=1, min_width=350):
+                    story_input = gr.Textbox(label="Add to Story", placeholder="Enter your part of the story...", lines=4, max_lines=6)
+                    story_model = gr.Radio(choices=models_options_general, value="GPT2", label="Select Model", type="value")
+                    story_max_length = gr.Slider(50, 300, value=50, step=1, label="Max Length", interactive=True)
+                with gr.Column(scale=1, min_width=350):
+                    story_text = gr.Textbox(label="Story So Far", interactive=False, lines=12, max_lines=20)
+                    story_button = gr.Button("Generate Next Part", variant="primary")
+                    reset_button = gr.Button("Reset Story", variant="secondary")
+
+            story_button.click(
+                interactive_story,
+                inputs=[story_input, story_model, story_max_length],
+                outputs=story_text,
+            )
+            reset_button.click(
+                reset_story,
+                inputs=[],
+                outputs=story_text,
+            )
+
+        with gr.Tab("Training"):
+            gr.Markdown("# **Train Model**\n\n")
+            with gr.Column(scale=1, min_width=250):
+                train_model_selector = gr.Radio(choices=models_options_general, value="GPT2", label="Select Model for Training", type="value")
+                train_method = gr.Radio(
+                    choices=["Custom Text", "Database", "Dataset File", "Hugging Face Dataset"],
+                    value="Custom Text",
+                    label="Training Method",
+                    type="value"
+                )
+                dataset_name = gr.Textbox(label="Hugging Face Dataset Name", placeholder="Enter dataset name (e.g., ag_news)")
+                split_name = gr.Textbox(label="Dataset Split", placeholder="e.g., train, test, validation")
+                epochs = gr.Slider(1, 100, value=10, step=1, label="Epochs", interactive=True)
+                batch_size = gr.Slider(1, 100, value=8, step=1, label="Batch Size", interactive=True)
+                password = gr.Textbox(label="Enter Training Password", placeholder="Enter password", type="password")
+                custom_text = gr.Textbox(label="Custom Text (optional)", placeholder="Enter custom text for training...")
+                dataset_file = gr.File(label="Upload Dataset", type="filepath", file_types=[".parquet", ".csv", ".json", ".txt"])
+                train_button = gr.Button("Train Model", variant="primary")
+                train_status = gr.Textbox(label="Training Status", interactive=False)
+            
+            train_button.click(
+                verify_and_train_combined,
+                inputs=[train_model_selector, train_method, epochs, batch_size, password, custom_text, dataset_file, dataset_name, split_name],
+                outputs=train_status,
+            )
+            train_button.click(
+                verify_and_train_combined,
+                inputs=[train_model_selector, train_method, epochs, batch_size, password, custom_text, dataset_file, dataset_name, split_name],
+                outputs=train_status,
+            )
+
+        with gr.Tab("Code Generator"):
+            gr.Markdown("### Generate Code from Descriptions")
+            with gr.Row():
+                with gr.Column(scale=1, min_width=350):
+                    code_prompt = gr.Textbox(label="Code Prompt", placeholder="Describe your coding task, e.g., 'Write a Python function to calculate Fibonacci numbers.'")
+                    code_max_tokens = gr.Slider(10, 500, value=150, step=10, label="Max Tokens")
+                with gr.Column(scale=1, min_width=350):
+                    generated_code = gr.Textbox(label="Generated Code", interactive=False, lines=10, max_lines=20)
+                    generate_code_button = gr.Button("Generate Code")
+
+            generate_code_button.click(
+                _generate_code,
+                inputs=[code_prompt, code_max_tokens],
+                outputs=generated_code,
+            )
+
+        # Add AI-Powered Story World Builder Tab
+        with gr.Tab("Story World Builder"):
+            with gr.Row():
+                with gr.Column(scale=1, min_width=350):
+                    world_name = gr.Textbox(label="World Name", placeholder="Enter your world name...")
+                    locations = gr.Textbox(label="Locations", placeholder="Enter locations separated by commas...")
+                    characters = gr.Textbox(label="Characters", placeholder="Enter characters separated by commas...")
+                    create_button = gr.Button("Create World", variant='primary')
+                    generate_story_button = gr.Button("Generate Story")
+                with gr.Column(scale=1, min_width=350):
+                    world_status = gr.Textbox(label="World Status", interactive=False)
+                    generated_story = gr.Textbox(label="Generated Story", interactive=False, lines=12, max_lines=20)
+
+
+            create_button.click(
+                define_world,
+                inputs=[world_name, locations, characters],
+                outputs=world_status,
+            )
+            
+            gr.Markdown("### Generate a Story in Your World")
+            with gr.Row():
+                with gr.Column(scale=1, min_width=350):
+                    story_world = gr.Textbox(label="Enter World Name", placeholder="World name...")
+                    event = gr.Textbox(label="Event", placeholder="Describe an event in the world...")
+                    selected_model = gr.Radio(choices=models_options_general, value="GPT2", label="Select Model", type="value")
+                    max_length = gr.Slider(50, 300, value=150, step=1, label="Max Length")
+
+        with gr.Tab("Chatbot"):
+            gr.Markdown("### **Chat With AI Models**")
+            with gr.Row():
+                with gr.Column(scale=1, min_width=250):
+                    username = gr.Textbox(label="Username", placeholder="Enter your username", lines=1)
+                    chat_id = gr.Textbox(label="Chat ID (optional)", placeholder="Enter chat ID or leave blank for a new chat", lines=1)
+                    selected_model = gr.Radio(models_options_chatbot, label="Select Model", value="dialoGPT")
+                    send_button = gr.Button("Send", variant="primary")
+                    reset_button = gr.Button("Reset Chat", variant="secondary")
+                with gr.Column(scale=1, min_width=250):
+                    input_text = gr.Textbox(label="Your Message", placeholder="Type your message here...", lines=2)
+                    emotion_output = gr.Textbox(label="Detected Emotion", interactive=False)
+                    chat_output = gr.Textbox(label="Chat History", lines=10, interactive=False)
+
+            send_button.click(
+                chatbot_response_with_emotion,
+                inputs=[username, input_text, selected_model, chat_id],
+                outputs=[chat_output, chat_id, emotion_output]
+            )
+
+            reset_button.click(
+                reset_chat,
+                inputs=[username],
+                outputs=[chat_output]
+            )
+            gr.Markdown("---")
+            gr.Markdown("### **Fetch Chat IDs**")
+            with gr.Row():
+                with gr.Column(scale=1, min_width=250):
+                    username = gr.Textbox(label="Username", placeholder="Enter your username", lines=1)
+                    fetch_btn = gr.Button("Fetch", variant="primary")
+                with gr.Column(scale=1, min_width=250):
+                    fetch_output = gr.Textbox(label="Chat IDs", lines=3, interactive=False)
+            fetch_btn.click(
+                chat_ids,
+                inputs=[username],
+                outputs=[fetch_output],
+            )
+    
+    generate_story_button.click(
+        generate_story,
+        inputs=[selected_model, story_world, max_length, event],
+        outputs=generated_story,
+    )
+
+    gr.Markdown("Made by **AliMc2021** with ❤️")
+
+# Launch the interface
+interface.queue().launch(
+    server_port=7860, 
+    show_error=True, 
+    inline=False,
+    #share=True,
+)

database.py

@@ -0,0 +1,132 @@
+import sqlite3
+
+# مسیر پایگاه داده
+DATABASE_PATH = 'database.db'
+
+# ایجاد یا بازنشانی جداول پایگاه داده
+def create_db():
+    conn = sqlite3.connect(DATABASE_PATH)
+    c = conn.cursor()
+    c.execute("""
+        CREATE TABLE IF NOT EXISTS inputs (
+            id INTEGER PRIMARY KEY,
+            input_text TEXT,
+            selected_model TEXT
+        )
+    """)
+    c.execute("""
+        CREATE TABLE IF NOT EXISTS chats (
+            id INTEGER PRIMARY KEY AUTOINCREMENT,
+            chat_id TEXT NOT NULL,
+            username TEXT NOT NULL,
+            user_message TEXT NOT NULL,
+            ai_response TEXT NOT NULL
+        )
+    """)
+    conn.commit()
+    conn.close()
+
+# درج چت در جدول chats
+def insert_chat(chat_id, username, user_message, ai_response):
+    try:
+        conn = sqlite3.connect(DATABASE_PATH)
+        cursor = conn.cursor()
+        cursor.execute("""
+            INSERT INTO chats (chat_id, username, user_message, ai_response)
+            VALUES (?, ?, ?, ?)
+        """, (str(chat_id), str(username), str(user_message), str(ai_response)))
+        conn.commit()
+    except sqlite3.Error as e:
+        print(f"Error inserting chat: {e}")
+    finally:
+        conn.close()
+
+# درج داده در جدول inputs
+def insert_into_db(input_text, selected_model):
+    try:
+        conn = sqlite3.connect(DATABASE_PATH)
+        c = conn.cursor()
+        c.execute("""
+            INSERT INTO inputs (input_text, selected_model)
+            VALUES (?, ?)
+        """, (str(input_text), str(selected_model)))
+        conn.commit()
+    except sqlite3.Error as e:
+        print(f"Error inserting into inputs: {e}")
+    finally:
+        conn.close()
+
+# پاک کردن داده‌های جدول inputs
+def clear_database():
+    try:
+        conn = sqlite3.connect(DATABASE_PATH)
+        c = conn.cursor()
+        c.execute("DELETE FROM inputs")
+        conn.commit()
+    except sqlite3.Error as e:
+        print(f"Error clearing database: {e}")
+    finally:
+        conn.close()
+
+# بازیابی تمام ورودی‌ها از جدول inputs
+def fetch_all_inputs():
+    try:
+        conn = sqlite3.connect(DATABASE_PATH)
+        c = conn.cursor()
+        c.execute("SELECT input_text, selected_model FROM inputs")
+        results = c.fetchall()
+        return results
+    except sqlite3.Error as e:
+        print(f"Error fetching inputs from database: {e}")
+        return []
+    finally:
+        conn.close()
+
+# بازیابی پیام‌ها و پاسخ‌های مرتبط با یک chat_id
+def fetch_chats_by_id(chat_id):
+    try:
+        conn = sqlite3.connect(DATABASE_PATH)
+        cursor = conn.cursor()
+        cursor.execute("""
+            SELECT user_message, ai_response FROM chats
+            WHERE chat_id = ?
+        """, (str(chat_id),))
+        rows = cursor.fetchall()
+        return rows
+    except sqlite3.Error as e:
+        print(f"Error fetching chats by ID: {e}")
+        return []
+    finally:
+        conn.close()
+
+# بازیابی chat_id ها برای یک کاربر خاص
+def fetch_ids_by_user(username):
+    try:
+        conn = sqlite3.connect(DATABASE_PATH)
+        cursor = conn.cursor()
+        cursor.execute("""
+            SELECT chat_id FROM chats
+            WHERE username = ?
+        """, (str(username),))
+        rows = cursor.fetchall()
+        return rows
+    except sqlite3.Error as e:
+        print(f"Error fetching chat IDs by username: {e}")
+        return []
+    finally:
+        conn.close()
+
+# حذف چت‌های مرتبط با یک کاربر خاص
+def clear_chats_by_username(username):
+    try:
+        conn = sqlite3.connect(DATABASE_PATH)
+        cursor = conn.cursor()
+        cursor.execute("""
+            DELETE FROM chats
+            WHERE username = ?
+        """, (str(username),))
+        conn.commit()
+    except sqlite3.Error as e:
+        print(f"Error clearing chats by username: {e}")
+    finally:
+        conn.close()

download.py

@@ -0,0 +1,40 @@
+import os
+from transformers import AutoModelForCausalLM, AutoTokenizer
+
+# لیست مدل‌ها با مسیر ذخیره مشخص‌شده
+MODEL_LIST = {
+    "gpt2": {"path": "openai-community/gpt2", "save_dir": "./models/gpt2"},
+    "gpt2-medium": {"path": "openai-community/gpt2-medium", "save_dir": "./models/gpt2-medium"},
+    "gpt2-persian": {"path": "flax-community/gpt2-medium-persian", "save_dir": "./models/gpt2-medium-persian"},
+    "gpt2-large": {"path": "openai-community/gpt2-large", "save_dir": "./models/gpt2-large"},
+    "codegen": {"path": "Salesforce/codegen-350M-mono", "save_dir": "./models/codegen"},
+    "dialogpt": {"path": "microsoft/DialoGPT-small", "save_dir": "./models/dialogpt"},
+    "dialogpt-medium": {"path": "microsoft/DialoGPT-medium", "save_dir": "./models/dialogpt-medium"},
+    "dialogpt-large": {"path": "microsoft/DialoGPT-large", "save_dir": "./models/dialogpt-large"}
+}
+
+def download_and_save_models():
+    """
+    دانلود و ذخیره تمام مدل‌ها در مسیرهای مشخص‌شده.
+    """
+    for model_name, model_info in MODEL_LIST.items():
+        model_path = model_info["path"]  # مسیر مدل در Hugging Face
+        save_dir = model_info["save_dir"]  # مسیر ذخیره مدل
+        
+        print(f"Downloading and saving model: {model_name} to folder: {save_dir}")
+        
+        if not os.path.exists(save_dir):  # بررسی اینکه آیا فولدر ذخیره وجود دارد یا نه
+            os.makedirs(save_dir, exist_ok=True)
+            
+            # دانلود و ذخیره مدل
+            model = AutoModelForCausalLM.from_pretrained(model_path)
+            tokenizer = AutoTokenizer.from_pretrained(model_path)
+            model.save_pretrained(save_dir)
+            tokenizer.save_pretrained(save_dir)
+            
+            print(f"Model {model_name} saved to {save_dir}")
+        else:
+            print(f"Model {model_name} already exists in {save_dir}")
+
+if __name__ == "__main__":
+    download_and_save_models()

functions.py

@@ -0,0 +1,270 @@
+import torch
+from model import load_model_lazy, unload_model
+from generate import generate_code, generate_text
+from database import *
+import train
+import uuid
+
+train_pass = '6818'
+
+# AI-Powered Story World Builder Functions
+world_data = {}
+
+def _generate_code(code_prompt, max_tokens, selected_model='codegen'):
+    """
+    Generate code based on the code prompt and selected model.
+    """
+    # Load the model lazily
+    model_data = load_model_lazy(selected_model)
+
+    # Generate code
+    generated_code = generate_code(model_data, code_prompt, max_tokens)
+
+    # Unload the model after use
+    unload_model(selected_model)
+
+    return generated_code
+
+def generate(input_text, selected_model, max_new_token):
+    """
+    Generate text based on the selected model and input text.
+    """
+    # Load the model lazily
+    model_data = load_model_lazy(selected_model)
+
+    # Generate text
+    generated_text = generate_text(model_data, input_text, max_new_token)
+    insert_into_db(input_text, selected_model)
+
+    # Unload the model after use
+    unload_model(selected_model)
+
+    return generated_text
+
+def define_world(world_name, locations, characters):
+    """
+    Define a new story world with locations and characters.
+    """
+    world_data["world_name"] = world_name
+    world_data["locations"] = locations.split(", ")
+    world_data["characters"] = characters.split(", ")
+    return f"World '{world_name}' created with locations: {locations} and characters: {characters}"
+
+def generate_story(model, world_name, event, max_length):
+    """
+    Generate a story based on the defined world and an event.
+    """
+    if not world_name or not world_data.get("world_name"):
+        return "Error: Please define a world first."
+
+    if world_name != world_data["world_name"]:
+        return f"Error: World '{world_name}' not found. Define it first."
+
+    prompt = f"In the world of {world_name}, {event}. Locations: {', '.join(world_data['locations'])}. Characters: {', '.join(world_data['characters'])}."
+
+    generated_story = generate(prompt, model, max_length)
+    return generated_story
+
+
+# Story Mode
+story = []
+
+# Main Function For Story Generating
+def interactive_story(input_text, selected_model, max_length):
+    global story
+    if input_text.strip():
+        story.append(input_text)  # Add user input to story
+    current_text = " ".join(story)  # Build cumulative story
+
+    generated_text = generate(current_text, selected_model, max_length)
+    story.append(generated_text)  # Add generated text to story
+
+    return current_text + "\n\n" + generated_text
+
+
+def reset_story():
+    global story
+    story = []  # Reset story
+    return ""
+
+def generate_multiverse(input_text, selected_model, max_new_tokens, num_worlds=3):
+    """
+    Generate multiple parallel worlds from a single input text.
+    """
+    worlds = []
+
+    for i in range(num_worlds):
+        world_intro = f"World {i + 1}: "
+        # Custom logic for different parallel worlds
+        if i == 0:
+            world_intro += f"{input_text} This world leads to a parallel universe!"
+        elif i == 1:
+            world_intro += f"{input_text} In this world, time splits into different periods!"
+        elif i == 2:
+            world_intro += f"{input_text} This world faces a strange physical anomaly that changes everything!"
+
+        # Generate the story for this world
+        generated_text = generate(world_intro, selected_model, max_new_tokens)
+
+        worlds.append(generated_text)
+
+    return "\n\n".join(worlds)
+
+
+
+# Function to verify password, train the model, and clear the database
+def verify_and_train_combined(selected_model, train_method, epochs, batch_size, password, custom_text, dataset_file, dataset_name, split_name):
+    if password != train_pass:
+        return "Error: Incorrect password. Training not started."
+
+    if train_method == "Custom Text" and custom_text.strip():
+        train.train_model_with_text(selected_model, custom_text, epochs, batch_size)
+        return f"Training completed for model: {selected_model} using custom text."
+
+    elif train_method == "Database":
+        train.train_model_with_database(selected_model, epochs, batch_size)
+        clear_database()
+        return f"Training completed for model: {selected_model} using database. Database cleared."
+
+    elif train_method == "Dataset File" and dataset_file is not None:
+        try:
+            dataset_path = dataset_file.name
+            train.train_model_with_dataset(selected_model, epochs, batch_size, dataset_path)
+            return f"Training completed for model: {selected_model} using uploaded dataset."
+        except Exception as e:
+            return f"Error during training with dataset: {str(e)}"
+
+    elif train_method == "Hugging Face Dataset" and dataset_name.strip():
+        try:
+            train.train_model_with_hf_dataset(selected_model, epochs, batch_size, dataset_name, split=split_name.strip())
+            return f"Training completed for model: {selected_model} using Hugging Face dataset {dataset_name}."
+        except Exception as e:
+            return f"Error during training with Hugging Face dataset: {str(e)}"
+
+    else:
+        return "Error: Invalid input for training. Please check your selections."
+
+def limit_chat_history(chat_history, max_turns=3):
+    """
+    محدود کردن تعداد پیام‌های تاریخچه به max_turns.
+    """
+    turns = chat_history.split("\n")
+    if len(turns) > max_turns * 2:  # هر سوال و پاسخ دو خط می‌شود
+        turns = turns[-max_turns * 2:]  # فقط n پیام اخیر را نگه می‌دارد
+    return "\n".join(turns)
+
+def chatbot_response(username, input_text, selected_model, chat_id=None):
+    if not username.strip():
+        return "Error: Please enter a username.", "", str(uuid.uuid4())  # تولید شناسه جدید
+
+    # اگر شناسه چت وارد نشده باشد، یک شناسه جدید تولید می‌شود
+    if not chat_id or chat_id.strip() == "":
+        chat_id = str(uuid.uuid4())  # تولید شناسه جدید
+
+    # Load model lazily
+    model_data = load_model_lazy(selected_model)
+
+    # Retrieve previous chats from database
+    previous_chats = fetch_chats_by_id(chat_id)
+    chat_history = "\n".join([f"User: {msg}\nAI: {resp}" for msg, resp in previous_chats])
+
+    # محدود کردن تاریخچه چت
+    if chat_history:
+        chat_history = limit_chat_history(chat_history, max_turns=3)
+        prompt = f"{chat_history}\nUser: {input_text}\nAI:"
+    else:
+        prompt = f"User: {input_text}\nAI:"
+
+    # Generate response
+    max_new_token = 150  # تعداد توکن‌های جدید
+    full_response = generate_text(model_data, prompt, max_new_token)  # حذف آرگومان‌های اضافی
+
+    # Extract only the new AI response
+    ai_response = full_response.split("AI:")[-1].strip()
+
+    unload_model(selected_model)
+
+    # Save chat to database
+    insert_chat(chat_id, username, input_text, ai_response)
+
+    # Return updated chat history and chat_id
+    updated_history = chat_history + f"\nUser: {input_text}\nAI: {ai_response}"
+    return updated_history, chat_id
+
+def chat_ids(username):
+    return fetch_ids_by_user(username)
+
+def reset_chat(username):
+    clear_chats_by_username(username)  # حذف چت‌های مرتبط با کاربر
+    return f"Chat history cleared for user: {username}", ""
+
+# توابع تحلیل احساسات
+def analyze_emotion(user_input):
+    # بارگذاری مدل احساسات
+    model_data = load_model_lazy("bert-emotion")
+    
+    # اگر مدل از pipeline پشتیبانی می‌کند
+    if "pipeline" in model_data:
+        emotion_pipeline = model_data["pipeline"]
+        result = emotion_pipeline(user_input)
+        emotion = result[0]['label']
+        confidence = result[0]['score']
+    else:
+        # روش قدیمی برای مدل‌هایی که از pipeline پشتیبانی نمی‌کنند
+        emotion_tokenizer = model_data['tokenizer']
+        emotion_model = model_data['model']
+        inputs = emotion_tokenizer(user_input, return_tensors="pt", truncation=True, padding=True)
+        outputs = emotion_model(**inputs)
+        probs = torch.nn.functional.softmax(outputs.logits, dim=-1)
+        emotion = probs.argmax().item()
+        confidence = probs.max().item()
+    
+    unload_model("bert-emotion")
+    return emotion, confidence
+
+def emotion_label(index):
+    emotions = ["anger", "joy", "sadness", "fear", "love", "surprise"]
+    return emotions[index]
+
+def chatbot_response_with_emotion(username, input_text, selected_model, chat_id=None):
+    if not username.strip():
+        return "Error: Please enter a username.", "", str(uuid.uuid4())
+
+    if not chat_id or chat_id.strip() == "":
+        chat_id = str(uuid.uuid4())
+
+    # بارگذاری مدل چت و احساسات
+    model_data = load_model_lazy(selected_model)
+
+    # تحلیل احساسات پیام کاربر
+    emotion, confidence = analyze_emotion(input_text)
+    user_emotion = emotion  # برچسب احساسات
+
+    # بازیابی چت‌های قبلی از پایگاه داده
+    previous_chats = fetch_chats_by_id(chat_id)
+    chat_history = "\n".join([f"User: {msg}\nAI: {resp}" for msg, resp in previous_chats])
+
+    # محدود کردن تاریخچه چت
+    if chat_history:
+        chat_history = limit_chat_history(chat_history, max_turns=3)
+        prompt = f"[Emotion: {user_emotion}]\n{chat_history}\nUser: {input_text}\nAI:"
+    else:
+        prompt = f"[Emotion: {user_emotion}]\nUser: {input_text}\nAI:"
+
+    # تولید پاسخ
+    max_new_token = 150
+    full_response = generate_text(model_data, prompt, max_new_token)
+
+    # استخراج پاسخ AI
+    ai_response = full_response.split("AI:")[-1].strip()
+
+    # آزادسازی مدل‌ها
+    unload_model(selected_model)
+    unload_model("bert-emotion")
+
+    # ذخیره چت در پایگاه داده
+    insert_chat(chat_id, username, input_text, ai_response)
+
+    # بازگرداندن تاریخچه به‌روز شده و شناسه چت
+    updated_history = chat_history + f"\nUser: {input_text}\nAI: {ai_response}"
+    return updated_history, chat_id, user_emotion

generate.py

@@ -0,0 +1,80 @@
+import torch
+
+seed = 0
+
+def generate_text(model_data, input_text, max_new_token):
+    """
+    Generate text using the given model and tokenizer.
+    """
+    if "pipeline" in model_data:
+        # اگر مدل از pipeline پشتیبانی می‌کند
+        model_pipeline = model_data["pipeline"]
+        generated_text = model_pipeline(
+            input_text,
+            max_length=max_new_token,
+            do_sample=False,  # غیرفعال کردن نمونه‌گیری (حالت حریصانه)
+            truncation=True  # فعال کردن truncation
+        )[0]["generated_text"]
+        return generated_text
+    else:
+        # روش قدیمی برای مدل‌هایی که از pipeline پشتیبانی نمی‌کنند
+        model = model_data["model"]
+        tokenizer = model_data["tokenizer"]
+
+        if tokenizer.pad_token is None:
+            tokenizer.pad_token = tokenizer.eos_token
+
+        torch.manual_seed(seed)
+        torch.cuda.manual_seed_all(seed)
+
+        encodings = tokenizer(
+            input_text,
+            return_tensors="pt",
+            padding=True,
+            truncation=True,  # فعال کردن truncation
+            max_length=512
+        )
+        input_ids = encodings.input_ids
+        attention_mask = encodings.attention_mask
+
+        outputs = model.generate(
+            input_ids=input_ids,
+            attention_mask=attention_mask,
+            max_new_tokens=max_new_token,
+            do_sample=False,  # غیرفعال کردن نمونه‌گیری (حالت حریصانه)
+            pad_token_id=tokenizer.eos_token_id,
+            repetition_penalty=1.2,
+            no_repeat_ngram_size=3,
+        )
+
+        return tokenizer.decode(outputs[0], skip_special_tokens=True)
+
+def generate_code(model_data, prompt, max_new_tokens):
+    """
+    Generate code based on the provided prompt using a code-specific model.
+    """
+    model = model_data["model"]
+    tokenizer = model_data["tokenizer"]
+
+    # تنظیم seed برای خروجی ثابت
+    torch.manual_seed(seed)
+    torch.cuda.manual_seed_all(seed)
+
+    # توکنایز کردن ورودی
+    input_ids = tokenizer.encode(prompt, return_tensors="pt")
+
+    # ایجاد attention mask
+    attention_mask = torch.ones(input_ids.shape, device=input_ids.device)  # ایجاد یک ماسک توجه برای ورودی‌ها
+
+    # تولید کد
+    outputs = model.generate(
+        input_ids=input_ids,
+        attention_mask=attention_mask,  # ارسال attention mask
+        max_new_tokens=max_new_tokens,
+        do_sample=False,
+        pad_token_id=tokenizer.eos_token_id,  # تنظیم شناسه توکن پایان به عنوان پرکننده
+        repetition_penalty=1.2,  # جلوگیری از تکرار
+        no_repeat_ngram_size=3,  # جلوگیری از تکرار n-gram
+    )
+    
+    return tokenizer.decode(outputs[0], skip_special_tokens=True)

lib.txt

@@ -0,0 +1,15 @@
+transformers
+torch
+gradio
+datasets
+numpy
+Pillow
+scikit-learn
+wandb
+pyarrow
+pandas
+chardet
+accelerate
+safetensors
+diffusers
+jax

model.py

@@ -0,0 +1,74 @@
+import torch
+import gc
+from transformers import AutoModelForCausalLM, AutoTokenizer, GPT2LMHeadModel, GPT2Tokenizer, pipeline, AutoModelForSequenceClassification
+
+# Dictionary of models and paths
+model_dict = {
+    "GPT2": {"path": "./models/gpt2", "library": GPT2LMHeadModel, "tokenizer": GPT2Tokenizer, "use_pipeline": False},
+    "GPT2-medium": {"path": "./models/gpt2-medium", "library": GPT2LMHeadModel, "tokenizer": GPT2Tokenizer, "use_pipeline": False},
+    "GPT2-large": {"path": "./models/gpt2-large", "library": AutoModelForCausalLM, "tokenizer": AutoTokenizer, "use_pipeline": False},
+    "GPT2-persian": {"path": "./models/gpt2-medium-persian", "library": GPT2LMHeadModel, "tokenizer": AutoTokenizer, "use_pipeline": False},
+    "codegen": {"path": "./models/codegen", "library": AutoModelForCausalLM, "tokenizer": AutoTokenizer, "use_pipeline": False},
+    "dialoGPT": {"path": "./models/dialogpt", "library": AutoModelForCausalLM, "tokenizer": AutoTokenizer, "use_pipeline": False},
+    "dialoGPT-medium": {"path": "./models/dialogpt-medium", "library": AutoModelForCausalLM, "tokenizer": AutoTokenizer, "use_pipeline": False},
+    "dialoGPT-large": {"path": "./models/dialogpt-large", "library": AutoModelForCausalLM, "tokenizer": AutoTokenizer, "use_pipeline": False},
+    "GPT-Neo-125M": {"path": "./models/GPT-neo-125M", "library": AutoModelForCausalLM, "tokenizer": AutoTokenizer, "use_pipeline": True},  # اضافه کردن مدل جدید
+    "bert-emotion": {"path": "./models/bert-emotion", "library": AutoModelForSequenceClassification, "tokenizer": AutoTokenizer, "use_pipeline": True},
+}
+
+loaded_models = {}
+
+def load_model_lazy(model_name):
+    if not isinstance(model_name, str):
+        raise ValueError(f"Model name must be a string, not {type(model_name)}")
+    if model_name not in model_dict:
+        raise ValueError(f"Model {model_name} not found!")
+
+    model_info = model_dict[model_name]
+    print(f"Loading model: {model_name}")
+
+    # اگر مدل از pipeline پشتیبانی می‌کند
+    if model_info.get("use_pipeline", False):
+        print(f"Using pipeline for model: {model_name}")
+        if model_name == "bert-emotion":
+            # برای مدل bert-emotion از text-classification استفاده کنید
+            model_pipeline = pipeline(
+                "text-classification",  # تغییر وظیفه به text-classification
+                model=model_info["path"],
+                truncation=True
+            )
+        else:
+            # برای سایر مدل‌ها از text-generation استفاده کنید
+            model_pipeline = pipeline(
+                "text-generation",
+                model=model_info["path"],
+                truncation=True,
+                pad_token_id=50256
+            )
+        loaded_models[model_name] = {"pipeline": model_pipeline}
+        return {"pipeline": model_pipeline}
+
+    # در غیر این صورت، مدل و توکنایزر را به روش قدیمی بارگذاری کنید
+    model = model_info["library"].from_pretrained(model_info["path"])
+    tokenizer = model_info["tokenizer"].from_pretrained(model_info["path"])
+
+    # تنظیمات پیش‌فرض
+    if tokenizer.pad_token is None:
+        tokenizer.pad_token = tokenizer.eos_token
+
+    loaded_models[model_name] = {"model": model, "tokenizer": tokenizer}
+    return {"model": model, "tokenizer": tokenizer}
+
+def unload_model(model_name):
+    global loaded_models
+    if model_name in loaded_models:
+        if "pipeline" in loaded_models[model_name]:
+            del loaded_models[model_name]["pipeline"]
+        elif "model" in loaded_models[model_name]:
+            del loaded_models[model_name]["model"]
+            del loaded_models[model_name]["tokenizer"]
+        torch.cuda.empty_cache()
+        gc.collect()
+        print(f"Model {model_name} unloaded and memory cleared.")
+    else:
+        print(f"Model {model_name} was not loaded.")

train.py

@@ -0,0 +1,136 @@
+import torch
+from torch.utils.data import Dataset, DataLoader
+from transformers import AdamW
+from model import load_model_lazy, unload_model
+from database import fetch_all_inputs, clear_database  # مدیریت دیتابیس
+from datasets import load_dataset
+
+class TextDataset(Dataset):
+    def __init__(self, texts, tokenizer, max_length=512):
+        self.texts = texts
+        self.tokenizer = tokenizer
+        self.max_length = max_length
+
+    def __len__(self):
+        return len(self.texts)
+
+    def __getitem__(self, idx):
+        text = self.texts[idx]
+        encodings = self.tokenizer(
+            text,
+            truncation=True,
+            padding="max_length",  # پُر کردن توکن‌ها تا طول مشخص
+            max_length=self.max_length,
+            return_tensors="pt"
+        )
+        attention_mask = encodings.attention_mask.squeeze(0)
+        return encodings.input_ids.squeeze(0), attention_mask
+
+def train_model_with_text(selected_model, custom_text, epochs, batch_size):
+    """
+    آموزش مدل با متن سفارشی.
+    """
+    model, tokenizer = load_model_lazy(selected_model)
+    dataset = TextDataset([custom_text], tokenizer)
+    dataloader = DataLoader(dataset, batch_size=min(batch_size, len(dataset)), shuffle=True)
+
+    _train_model(model, tokenizer, dataloader, epochs, selected_model, "custom_text")
+    unload_model(selected_model)
+
+def train_model_with_database(selected_model, epochs, batch_size):
+    """
+    آموزش مدل با داده‌های موجود در دیتابیس.
+    """
+    model, tokenizer = load_model_lazy(selected_model)
+    inputs_data = fetch_all_inputs()
+    texts = [input_text for input_text, model_name in inputs_data if model_name == selected_model]
+
+    if not texts:
+        print("Error: No data found in the database for the selected model.")
+        return
+
+    dataset = TextDataset(texts, tokenizer)
+    dataloader = DataLoader(dataset, batch_size=batch_size, shuffle=True)
+
+    _train_model(model, tokenizer, dataloader, epochs, selected_model, "database")
+    clear_database()
+    unload_model(selected_model)
+
+def train_model_with_dataset(selected_model, epochs, batch_size, dataset_path):
+    """
+    آموزش مدل با فایل دیتاست آپلود‌شده.
+    """
+    model, tokenizer = load_model_lazy(selected_model)
+
+    # خواندن دیتاست
+    with open(dataset_path, "r", encoding="utf-8") as f:
+        texts = f.readlines()
+
+    if not texts:
+        print("Error: Dataset is empty.")
+        return
+
+    dataset = TextDataset(texts, tokenizer)
+    dataloader = DataLoader(dataset, batch_size=batch_size, shuffle=True)
+
+    _train_model(model, tokenizer, dataloader, epochs, selected_model, "dataset")
+    unload_model(selected_model)
+
+def _train_model(model, tokenizer, dataloader, epochs, model_name, method):
+    """
+    منطق مشترک آموزش مدل.
+    """
+    optimizer = AdamW(model.parameters(), lr=5e-5)
+
+    # انتقال مدل به GPU در صورت وجود
+    device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+    model.to(device)
+
+    model.train()
+    for epoch in range(epochs):
+        total_loss = 0
+        for step, (input_ids, attention_mask) in enumerate(dataloader):
+            optimizer.zero_grad()
+            input_ids = input_ids.to(device)
+            attention_mask = attention_mask.to(device)
+
+            # محاسبه خروجی و خطا
+            outputs = model(input_ids, attention_mask=attention_mask, labels=input_ids)
+            loss = outputs.loss
+            loss.backward()
+            optimizer.step()
+            total_loss += loss.item()
+
+        print(f"Epoch {epoch + 1}/{epochs}, Loss: {total_loss / len(dataloader)}")
+
+    # ذخیره مدل
+    save_path = f"trained_{model_name}_{method}"
+    model.save_pretrained(save_path)
+    tokenizer.save_pretrained(save_path)
+    print(f"Model {model_name} trained with {method} and saved to {save_path}.")
+
+def train_model_with_hf_dataset(selected_model, epochs, batch_size, dataset_name, split="train"):
+    """
+    آموزش مدل با استفاده از دیتاست‌های Hugging Face.
+    
+    Args:
+        selected_model (str): نام مدل برای آموزش.
+        epochs (int): تعداد epochs.
+        batch_size (int): اندازه batch.
+        dataset_name (str): نام دیتاست در Hugging Face.
+        split (str): بخش دیتاست برای بارگذاری (train, test, validation).
+    """
+    model, tokenizer = load_model_lazy(selected_model)
+
+    # بارگذاری داده‌ها از Hugging Face
+    texts = load_dataset(dataset_name, split)
+    
+    if not texts:
+        print(f"Error: Dataset {dataset_name} ({split} split) is empty or invalid.")
+        return
+
+    dataset = TextDataset(texts, tokenizer)
+    dataloader = DataLoader(dataset, batch_size=batch_size, shuffle=True)
+
+    _train_model(model, tokenizer, dataloader, epochs, selected_model, f"huggingface_{dataset_name}")
+    unload_model(selected_model)