Update app.py

app.py

@@ -4,9 +4,9 @@ from bs4 import BeautifulSoup
 import json
 from typing import List, Dict, Any, Optional
 import re
-from urllib.parse import urljoin, urlparse
+from urllib.parse import urljoin
 import time
-from functools import lru_cache
+import functools
 import logging
 from datetime import datetime, timedelta
 
@@ -14,7 +14,8 @@ from datetime import datetime, timedelta
 logging.basicConfig(level=logging.INFO)
 logger = logging.getLogger(__name__)
 
-class HF_API: # Renamed class for brevity
+# Renamed class for brevity to avoid long tool names
+class HF_API:
     def __init__(self):
         self.base_url = "https://huggingface.co"
         self.docs_url = "https://huggingface.co/docs"
@@ -160,28 +161,325 @@ class HF_API: # Renamed class for brevity
         return content
 
     def search_documentation(self, query: str, max_results: int = 3) -> str:
-        # ... (implementation remains the same)
-        return f"Documentation for {query} with {max_results} results."
+        """
+        Searches the official Hugging Face documentation for a specific topic and returns a summary.
+        This tool is useful for finding how-to guides, explanations of concepts like 'pipeline' or 'tokenizer', and usage examples.
+        Args:
+            query (str): The topic or keyword to search for in the documentation (e.g., 'fine-tuning', 'peft', 'datasets').
+            max_results (int): The maximum number of documentation pages to retrieve and summarize. Defaults to 3.
+        """
+        try:
+            max_results = int(max_results) if isinstance(max_results, str) else max_results
+            max_results = min(max_results, 5)
+            query_lower = query.lower().strip()
+            if not query_lower:
+                return "Please provide a search query."
+            doc_sections = {
+                'transformers': {'base_url': 'https://huggingface.co/docs/transformers', 'topics': {'pipeline': '/main_classes/pipelines', 'tokenizer': '/main_classes/tokenizer', 'trainer': '/main_classes/trainer', 'model': '/main_classes/model', 'quicktour': '/quicktour', 'installation': '/installation', 'fine-tuning': '/training', 'training': '/training', 'inference': '/main_classes/pipelines', 'preprocessing': '/preprocessing', 'tutorial': '/tutorials', 'configuration': '/main_classes/configuration', 'peft': '/peft', 'lora': '/peft', 'quantization': '/main_classes/quantization', 'generation': '/main_classes/text_generation', 'optimization': '/perf_train_gpu_one', 'deployment': '/deployment', 'custom': '/custom_models'}},
+                'datasets': {'base_url': 'https://huggingface.co/docs/datasets', 'topics': {'loading': '/load_hub', 'load': '/load_hub', 'processing': '/process', 'streaming': '/stream', 'audio': '/audio_process', 'image': '/image_process', 'text': '/nlp_process', 'arrow': '/about_arrow', 'cache': '/cache', 'upload': '/upload_dataset', 'custom': '/dataset_script'}},
+                'diffusers': {'base_url': 'https://huggingface.co/docs/diffusers', 'topics': {'pipeline': '/using-diffusers/loading', 'stable diffusion': '/using-diffusers/stable_diffusion', 'controlnet': '/using-diffusers/controlnet', 'inpainting': '/using-diffusers/inpaint', 'training': '/training/overview', 'optimization': '/optimization/fp16', 'schedulers': '/using-diffusers/schedulers'}},
+                'hub': {'base_url': 'https://huggingface.co/docs/hub', 'topics': {'repositories': '/repositories', 'git': '/repositories-getting-started', 'spaces': '/spaces', 'models': '/models', 'datasets': '/datasets'}}
+            }
+            relevant_urls = []
+            for section_name, section_data in doc_sections.items():
+                base_url = section_data['base_url']
+                topics = section_data['topics']
+                for topic, path in topics.items():
+                    relevance = 0
+                    if query_lower == topic.lower(): relevance = 1.0
+                    elif query_lower in topic.lower(): relevance = 0.9
+                    elif any(word in topic.lower() for word in query_lower.split()): relevance = 0.7
+                    elif any(word in query_lower for word in topic.lower().split()): relevance = 0.6
+                    if relevance > 0:
+                        full_url = base_url + path
+                        relevant_urls.append({'url': full_url, 'topic': topic, 'section': section_name, 'relevance': relevance})
+            relevant_urls.sort(key=lambda x: x['relevance'], reverse=True)
+            relevant_urls = relevant_urls[:max_results]
+            if not relevant_urls:
+                return f"❌ No documentation found for '{query}'. Try: pipeline, tokenizer, trainer, model, fine-tuning, datasets, diffusers, or peft."
+            result = f"# 📚 Hugging Face Documentation: {query}\n\n"
+            for i, url_info in enumerate(relevant_urls, 1):
+                section_emoji = {'transformers': '🤖', 'datasets': '📊', 'diffusers': '🎨', 'hub': '🌐'}.get(url_info['section'], '📄')
+                result += f"## {i}. {section_emoji} {url_info['topic'].title()} ({url_info['section'].title()})\n\n"
+                content = self._fetch_with_retry(url_info['url'])
+                if content:
+                    soup = BeautifulSoup(content, 'html.parser')
+                    practical_content = self._extract_practical_content(soup, url_info['topic'])
+                    if practical_content['overview']: result += f"**📖 Overview:**\n{practical_content['overview']}\n\n"
+                    if practical_content['installation']: result += f"**⚙️ Installation:**\n{practical_content['installation']}\n\n"
+                    if practical_content['code_examples']:
+                        result += "**💻 Code Examples:**\n\n"
+                        for j, code_block in enumerate(practical_content['code_examples'][:3], 1):
+                            lang = code_block.get('language', 'python')
+                            code_type = code_block.get('type', 'example')
+                            result += f"*{code_type.title()} {j}:*\n```{lang}\n{code_block['code']}\n```\n\n"
+                    if practical_content['usage_instructions']:
+                        result += "**🛠️ Usage Instructions:**\n"
+                        for idx, instruction in enumerate(practical_content['usage_instructions'][:4], 1):
+                            result += f"{idx}. {instruction}\n"
+                        result += "\n"
+                    if practical_content['parameters']:
+                        result += "**⚙️ Parameters:**\n"
+                        for param in practical_content['parameters'][:6]:
+                            param_type = f" (`{param['type']}`)" if param.get('type') else ""
+                            default_val = f" *Default: {param['default']}*" if param.get('default') else ""
+                            result += f"• **{param['name']}**{param_type}: {param['description']}{default_val}\n"
+                        result += "\n"
+                    result += f"**🔗 Full Documentation:** {url_info['url']}\n\n"
+                else:
+                    result += f"⚠️ Could not fetch content. Visit directly: {url_info['url']}\n\n"
+                result += "---\n\n"
+            return result
+        except Exception as e:
+            logger.error(f"Error in search_documentation: {e}")
+            return f"❌ Error searching documentation: {str(e)}\n\nTry a simpler search term or check your internet connection."
 
     def get_model_info(self, model_name: str) -> str:
-        # ... (implementation remains the same)
-        return f"Info for model {model_name}."
+        """
+        Fetches comprehensive information about a specific model from the Hugging Face Hub.
+        Provides statistics like downloads and likes, a description, usage examples, and a quick-start code snippet.
+        Args:
+            model_name (str): The full identifier of the model on the Hub, such as 'bert-base-uncased' or 'meta-llama/Llama-2-7b-hf'.
+        """
+        try:
+            model_name = model_name.strip()
+            if not model_name: return "Please provide a model name."
+            api_url = f"{self.api_url}/models/{model_name}"
+            response = self.session.get(api_url, timeout=15)
+            if response.status_code == 404: return f"❌ Model '{model_name}' not found. Please check the model name."
+            elif response.status_code != 200: return f"❌ Error fetching model info (Status: {response.status_code})"
+            model_data = response.json()
+            result = f"# 🤖 Model: {model_name}\n\n"
+            downloads = model_data.get('downloads', 0)
+            likes = model_data.get('likes', 0)
+            task = model_data.get('pipeline_tag', 'N/A')
+            library = model_data.get('library_name', 'N/A')
+            result += f"**📊 Statistics:**\n• **Downloads:** {downloads:,}\n• **Likes:** {likes:,}\n• **Task:** {task}\n• **Library:** {library}\n• **Created:** {model_data.get('createdAt', 'N/A')[:10]}\n• **Updated:** {model_data.get('lastModified', 'N/A')[:10]}\n\n"
+            if 'tags' in model_data and model_data['tags']: result += f"**🏷️ Tags:** {', '.join(model_data['tags'][:10])}\n\n"
+            model_url = f"{self.base_url}/{model_name}"
+            page_content = self._fetch_with_retry(model_url)
+            if page_content:
+                soup = BeautifulSoup(page_content, 'html.parser')
+                readme_content = soup.find('div', class_=re.compile(r'prose|readme|model-card'))
+                if readme_content:
+                    paragraphs = readme_content.find_all('p')[:3]
+                    description_parts = []
+                    for p in paragraphs:
+                        text = p.get_text(strip=True)
+                        if len(text) > 30 and not any(skip in text.lower() for skip in ['table of contents', 'toc']):
+                            description_parts.append(text)
+                    if description_parts:
+                        description = ' '.join(description_parts)
+                        result += f"**📝 Description:**\n{description[:800]}{'...' if len(description) > 800 else ''}\n\n"
+                code_examples = self._extract_code_examples(soup)
+                if code_examples:
+                    result += "**💻 Usage Examples:**\n\n"
+                    for i, code_block in enumerate(code_examples[:3], 1):
+                        lang = code_block.get('language', 'python')
+                        result += f"*Example {i}:*\n```{lang}\n{code_block['code']}\n```\n\n"
+            if task and task != 'N/A':
+                result += f"**🚀 Quick Start Template:**\n"
+                if library == 'transformers':
+                    result += f"```python\nfrom transformers import pipeline\n\n# Load the model\nmodel = pipeline('{task}', model='{model_name}')\n\n# Use the model\n# result = model(your_input_here)\n# print(result)\n```\n\n"
+                else:
+                    result += f"```python\n# Load and use {model_name}\n# Refer to the documentation for specific usage\n```\n\n"
+            if 'siblings' in model_data:
+                files = [f['rfilename'] for f in model_data['siblings'][:10]]
+                if files:
+                    result += f"**📁 Model Files:** {', '.join(files)}\n\n"
+            result += f"**🔗 Model Page:** {model_url}\n"
+            return result
+        except requests.exceptions.RequestException as e: return f"❌ Network error: {str(e)}"
+        except Exception as e:
+            logger.error(f"Error in get_model_info: {e}")
+            return f"❌ Error fetching model info: {str(e)}"
 
     def get_dataset_info(self, dataset_name: str) -> str:
-        # ... (implementation remains the same)
-        return f"Info for dataset {dataset_name}."
+        """
+        Retrieves detailed information about a specific dataset from the Hugging Face Hub.
+        Includes statistics, a description, and a quick-start code snippet showing how to load the dataset.
+        Args:
+            dataset_name (str): The full identifier of the dataset on the Hub, for example 'squad' or 'imdb'.
+        """
+        try:
+            dataset_name = dataset_name.strip()
+            if not dataset_name: return "Please provide a dataset name."
+            api_url = f"{self.api_url}/datasets/{dataset_name}"
+            response = self.session.get(api_url, timeout=15)
+            if response.status_code == 404: return f"❌ Dataset '{dataset_name}' not found. Please check the dataset name."
+            elif response.status_code != 200: return f"❌ Error fetching dataset info (Status: {response.status_code})"
+            dataset_data = response.json()
+            result = f"# 📊 Dataset: {dataset_name}\n\n"
+            downloads = dataset_data.get('downloads', 0)
+            likes = dataset_data.get('likes', 0)
+            result += f"**📈 Statistics:**\n• **Downloads:** {downloads:,}\n• **Likes:** {likes:,}\n• **Created:** {dataset_data.get('createdAt', 'N/A')[:10]}\n• **Updated:** {dataset_data.get('lastModified', 'N/A')[:10]}\n\n"
+            if 'tags' in dataset_data and dataset_data['tags']: result += f"**🏷️ Tags:** {', '.join(dataset_data['tags'][:10])}\n\n"
+            dataset_url = f"{self.base_url}/datasets/{dataset_name}"
+            page_content = self._fetch_with_retry(dataset_url)
+            if page_content:
+                soup = BeautifulSoup(page_content, 'html.parser')
+                readme_content = soup.find('div', class_=re.compile(r'prose|readme|dataset-card'))
+                if readme_content:
+                    paragraphs = readme_content.find_all('p')[:3]
+                    description_parts = []
+                    for p in paragraphs:
+                        text = p.get_text(strip=True)
+                        if len(text) > 30: description_parts.append(text)
+                    if description_parts:
+                        description = ' '.join(description_parts)
+                        result += f"**📝 Description:**\n{description[:800]}{'...' if len(description) > 800 else ''}\n\n"
+                code_examples = self._extract_code_examples(soup)
+                if code_examples:
+                    result += "**💻 Usage Examples:**\n\n"
+                    for i, code_block in enumerate(code_examples[:3], 1):
+                        lang = code_block.get('language', 'python')
+                        result += f"*Example {i}:*\n```{lang}\n{code_block['code']}\n```\n\n"
+            result += f"**🚀 Quick Start Template:**\n"
+            result += f"```python\nfrom datasets import load_dataset\n\n# Load the dataset\ndataset = load_dataset('{dataset_name}')\n\n# Explore the dataset\n# print(dataset)\n# print(f\"Dataset keys: {{list(dataset.keys())}}\")\n\n# Access first example\n# if 'train' in dataset:\n#     print(\"First example:\")\n#     print(dataset['train'][0])\n```\n\n"
+            result += f"**🔗 Dataset Page:** {dataset_url}\n"
+            return result
+        except requests.exceptions.RequestException as e: return f"❌ Network error: {str(e)}"
+        except Exception as e:
+            logger.error(f"Error in get_dataset_info: {e}")
+            return f"❌ Error fetching dataset info: {str(e)}"
 
     def search_models(self, task: str, limit: str = "5") -> str:
-        # ... (implementation remains the same)
-        return f"Models for task {task} with limit {limit}."
+        """
+        Searches the Hugging Face Hub for models based on a specified task or keyword and returns a list of top models.
+        Each result includes statistics and a quick usage example.
+        Args:
+            task (str): The task to search for, such as 'text-classification', 'image-generation', or 'question-answering'.
+            limit (str): The maximum number of models to return. Defaults to '5'.
+        """
+        try:
+            task = task.strip()
+            if not task: return "Please provide a search task or keyword."
+            limit = int(limit) if isinstance(limit, str) and limit.isdigit() else 5
+            limit = min(max(limit, 1), 10)
+            params = {'search': task, 'limit': limit * 3, 'sort': 'downloads', 'direction': -1}
+            response = self.session.get(f"{self.api_url}/models", params=params, timeout=20)
+            response.raise_for_status()
+            models = response.json()
+            if not models: return f"❌ No models found for task: '{task}'. Try different keywords."
+            filtered_models = []
+            for model in models:
+                if (model.get('downloads', 0) > 0 or model.get('likes', 0) > 0 or 'pipeline_tag' in model):
+                    filtered_models.append(model)
+                    if len(filtered_models) >= limit: break
+            if not filtered_models: filtered_models = models[:limit]
+            result = f"# 🔍 Top {len(filtered_models)} Models for '{task}'\n\n"
+            for i, model in enumerate(filtered_models, 1):
+                model_id = model.get('id', 'Unknown')
+                downloads = model.get('downloads', 0)
+                likes = model.get('likes', 0)
+                task_type = model.get('pipeline_tag', 'N/A')
+                library = model.get('library_name', 'N/A')
+                quality_score = ""
+                if downloads > 10000: quality_score = "⭐ Popular"
+                elif downloads > 1000: quality_score = "🔥 Active"
+                elif likes > 10: quality_score = "👍 Liked"
+                result += f"## {i}. {model_id} {quality_score}\n\n"
+                result += f"**📊 Stats:**\n• **Downloads:** {downloads:,}\n• **Likes:** {likes}\n• **Task:** {task_type}\n• **Library:** {library}\n\n"
+                if task_type and task_type != 'N/A':
+                    result += f"**🚀 Quick Usage:**\n"
+                    if library == 'transformers':
+                        result += f"```python\nfrom transformers import pipeline\n\n# Load model\nmodel = pipeline('{task_type}', model='{model_id}')\n\n# Use model\n# result = model(\"Your input here\")\n# print(result)\n```\n\n"
+                    else:
+                        result += f"```python\n# Load and use {model_id}\n# Check model page for specific usage instructions\n```\n\n"
+                result += f"**🔗 Model Page:** {self.base_url}/{model_id}\n\n---\n\n"
+            return result
+        except requests.exceptions.RequestException as e: return f"❌ Network error: {str(e)}"
+        except Exception as e:
+            logger.error(f"Error in search_models: {e}")
+            return f"❌ Error searching models: {str(e)}"
 
     def get_transformers_docs(self, topic: str) -> str:
-        # ... (implementation remains the same)
-        return f"Transformer docs for {topic}."
+        """
+        Fetches detailed documentation specifically for the Hugging Face Transformers library on a given topic.
+        This provides in-depth explanations, code examples, and parameter descriptions for core library components.
+        Args:
+            topic (str): The Transformers library topic to look up, such as 'pipeline', 'tokenizer', 'trainer', or 'generation'.
+        """
+        try:
+            topic = topic.strip().lower()
+            if not topic: return "Please provide a topic to search for."
+            docs_url = "https://huggingface.co/docs/transformers"
+            topic_map = {'pipeline': f"{docs_url}/main_classes/pipelines", 'pipelines': f"{docs_url}/main_classes/pipelines", 'tokenizer': f"{docs_url}/main_classes/tokenizer", 'tokenizers': f"{docs_url}/main_classes/tokenizer", 'trainer': f"{docs_url}/main_classes/trainer", 'training': f"{docs_url}/training", 'model': f"{docs_url}/main_classes/model", 'models': f"{docs_url}/main_classes/model", 'configuration': f"{docs_url}/main_classes/configuration", 'config': f"{docs_url}/main_classes/configuration", 'quicktour': f"{docs_url}/quicktour", 'quick': f"{docs_url}/quicktour", 'installation': f"{docs_url}/installation", 'install': f"{docs_url}/installation", 'tutorial': f"{docs_url}/tutorials", 'tutorials': f"{docs_url}/tutorials", 'generation': f"{docs_url}/main_classes/text_generation", 'text_generation': f"{docs_url}/main_classes/text_generation", 'preprocessing': f"{docs_url}/preprocessing", 'preprocess': f"{docs_url}/preprocessing", 'peft': f"{docs_url}/peft", 'lora': f"{docs_url}/peft", 'quantization': f"{docs_url}/main_classes/quantization", 'optimization': f"{docs_url}/perf_train_gpu_one", 'performance': f"{docs_url}/perf_train_gpu_one", 'deployment': f"{docs_url}/deployment", 'custom': f"{docs_url}/custom_models", 'fine-tuning': f"{docs_url}/training", 'finetuning': f"{docs_url}/training"}
+            url = topic_map.get(topic)
+            if not url:
+                for key, value in topic_map.items():
+                    if topic in key or key in topic:
+                        url = value
+                        topic = key
+                        break
+            if not url:
+                url = f"{docs_url}/quicktour"
+                topic = "quicktour"
+            content = self._fetch_with_retry(url)
+            if not content: return f"❌ Could not fetch documentation for '{topic}'. Please try again or visit: {url}"
+            soup = BeautifulSoup(content, 'html.parser')
+            practical_content = self._extract_practical_content(soup, topic)
+            result = f"# 📚 Transformers Documentation: {topic.replace('_', ' ').title()}\n\n"
+            if practical_content['overview']: result += f"**📖 Overview:**\n{practical_content['overview']}\n\n"
+            if practical_content['installation']: result += f"**⚙️ Installation:**\n{practical_content['installation']}\n\n"
+            if practical_content['code_examples']:
+                result += "**💻 Code Examples:**\n\n"
+                for i, code_block in enumerate(practical_content['code_examples'][:4], 1):
+                    lang = code_block.get('language', 'python')
+                    code_type = code_block.get('type', 'example')
+                    result += f"### {code_type.title()} {i}:\n```{lang}\n{code_block['code']}\n```\n\n"
+            if practical_content['usage_instructions']:
+                result += "**🛠️ Step-by-Step Usage:**\n"
+                for i, instruction in enumerate(practical_content['usage_instructions'][:6], 1):
+                    result += f"{i}. {instruction}\n"
+                result += "\n"
+            if practical_content['parameters']:
+                result += "**⚙️ Key Parameters:**\n"
+                for param in practical_content['parameters'][:10]:
+                    param_type = f" (`{param['type']}`)" if param.get('type') else ""
+                    default_val = f" *Default: `{param['default']}`*" if param.get('default') else ""
+                    result += f"• **`{param['name']}`**{param_type}: {param['description']}{default_val}\n"
+                result += "\n"
+            related_topics = [k for k in topic_map.keys() if k != topic][:5]
+            if related_topics: result += f"**🔗 Related Topics:** {', '.join(related_topics)}\n\n"
+            result += f"**📄 Full Documentation:** {url}\n"
+            return result
+        except Exception as e:
+            logger.error(f"Error in get_transformers_docs: {e}")
+            return f"❌ Error fetching Transformers documentation: {str(e)}"
 
     def get_trending_models(self, limit: str = "10") -> str:
-        # ... (implementation remains the same)
-        return f"Trending models with limit {limit}."
+        """
+        Fetches a list of the most downloaded models currently trending on the Hugging Face Hub.
+        This is useful for discovering popular and widely-used models.
+        Args:
+            limit (str): The number of trending models to return. Defaults to '10'.
+        """
+        try:
+            limit = int(limit) if isinstance(limit, str) and limit.isdigit() else 10
+            limit = min(max(limit, 1), 20)
+            params = {'sort': 'downloads', 'direction': -1, 'limit': limit}
+            response = self.session.get(f"{self.api_url}/models", params=params, timeout=20)
+            response.raise_for_status()
+            models = response.json()
+            if not models: return "❌ Could not fetch trending models."
+            result = f"# 🔥 Trending Models (Top {len(models)})\n\n"
+            for i, model in enumerate(models, 1):
+                model_id = model.get('id', 'Unknown')
+                downloads = model.get('downloads', 0)
+                likes = model.get('likes', 0)
+                task = model.get('pipeline_tag', 'N/A')
+                if downloads > 1000000: trend = "🚀 Mega Popular"
+                elif downloads > 100000: trend = "🔥 Very Popular"
+                elif downloads > 10000: trend = "⭐ Popular"
+                else: trend = "📈 Trending"
+                result += f"## {i}. {model_id} {trend}\n"
+                result += f"• **Downloads:** {downloads:,} | **Likes:** {likes} | **Task:** {task}\n"
+                result += f"• **Link:** {self.base_url}/{model_id}\n\n"
+            return result
+        except Exception as e:
+            logger.error(f"Error in get_trending_models: {e}")
+            return f"❌ Error fetching trending models: {str(e)}"
 
 # Initialize the API server
 hf_api = HF_API()
@@ -189,43 +487,32 @@ hf_api = HF_API()
 # --- Named Functions for Gradio UI ---
 
 def clear_output():
-    """Clears the Gradio output component."""
+    """Clears a Gradio output component."""
     return ""
 
+def set_textbox_value(text):
+    """Sets a Gradio Textbox to a specific value."""
+    return text
+
 # --- Doc Search Tab Functions ---
 def run_doc_search(query, max_results):
     return hf_api.search_documentation(query, int(max_results) if str(max_results).isdigit() else 2)
 
-def set_doc_query(text):
-    return text
-
 # --- Model Info Tab Functions ---
 def run_model_info(model_name):
     return hf_api.get_model_info(model_name)
 
-def set_model_name(text):
-    return text
-
 # --- Dataset Info Tab Functions ---
 def run_dataset_info(dataset_name):
     return hf_api.get_dataset_info(dataset_name)
 
-def set_dataset_name(text):
-    return text
-
 # --- Model Search Tab Functions ---
 def run_model_search(task, limit):
     return hf_api.search_models(task, int(limit) if str(limit).isdigit() else 5)
 
-def set_search_task(text):
-    return text
-
 # --- Transformers Docs Tab Functions ---
 def run_transformers_docs(topic):
     return hf_api.get_transformers_docs(topic)
-
-def set_transformer_topic(text):
-    return text
     
 # --- Trending Models Tab Functions ---
 def run_trending_models(limit):
@@ -238,17 +525,8 @@ with gr.Blocks(
     title="🤗 Hugging Face Information Server",
     theme=gr.themes.Soft(),
     css="""
-    .gradio-container {
-        font-family: 'Inter', sans-serif;
-    }
-    .main-header {
-        text-align: center;
-        padding: 20px;
-        background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
-        color: white;
-        border-radius: 10px;
-        margin-bottom: 20px;
-    }
+    .gradio-container { font-family: 'Inter', sans-serif; }
+    .main-header { text-align: center; padding: 20px; background: linear-gradient(135deg, #667eea 0%, #764ba2 100%); color: white; border-radius: 10px; margin-bottom: 20px; }
     """) as demo:
     # Header
     with gr.Row():
@@ -272,10 +550,10 @@ with gr.Blocks(
             doc_clear = gr.Button("🗑️ Clear", variant="secondary")
         gr.Markdown("**Quick Examples:**")
         with gr.Row():
-            gr.Button("Pipeline", size="sm").click(lambda: set_doc_query("pipeline"), outputs=doc_query)
-            gr.Button("Tokenizer", size="sm").click(lambda: set_doc_query("tokenizer"), outputs=doc_query)
-            gr.Button("Fine-tuning", size="sm").click(lambda: set_doc_query("fine-tuning"), outputs=doc_query)
-            gr.Button("PEFT", size="sm").click(lambda: set_doc_query("peft"), outputs=doc_query)
+            gr.Button("Pipeline", size="sm").click(functools.partial(set_textbox_value, "pipeline"), outputs=doc_query)
+            gr.Button("Tokenizer", size="sm").click(functools.partial(set_textbox_value, "tokenizer"), outputs=doc_query)
+            gr.Button("Fine-tuning", size="sm").click(functools.partial(set_textbox_value, "fine-tuning"), outputs=doc_query)
+            gr.Button("PEFT", size="sm").click(functools.partial(set_textbox_value, "peft"), outputs=doc_query)
         
         doc_btn.click(run_doc_search, inputs=[doc_query, doc_max_results], outputs=doc_output)
         doc_clear.click(clear_output, outputs=doc_output)
@@ -289,10 +567,10 @@ with gr.Blocks(
             model_clear = gr.Button("🗑️ Clear", variant="secondary")
         gr.Markdown("**Popular Models:**")
         with gr.Row():
-            gr.Button("BERT", size="sm").click(lambda: set_model_name("bert-base-uncased"), outputs=model_name)
-            gr.Button("GPT-2", size="sm").click(lambda: set_model_name("gpt2"), outputs=model_name)
-            gr.Button("T5", size="sm").click(lambda: set_model_name("t5-small"), outputs=model_name)
-            gr.Button("DistilBERT", size="sm").click(lambda: set_model_name("distilbert-base-uncased"), outputs=model_name)
+            gr.Button("BERT", size="sm").click(functools.partial(set_textbox_value, "bert-base-uncased"), outputs=model_name)
+            gr.Button("GPT-2", size="sm").click(functools.partial(set_textbox_value, "gpt2"), outputs=model_name)
+            gr.Button("T5", size="sm").click(functools.partial(set_textbox_value, "t5-small"), outputs=model_name)
+            gr.Button("DistilBERT", size="sm").click(functools.partial(set_textbox_value, "distilbert-base-uncased"), outputs=model_name)
         
         model_btn.click(run_model_info, inputs=model_name, outputs=model_output)
         model_clear.click(clear_output, outputs=model_output)
@@ -306,10 +584,10 @@ with gr.Blocks(
             dataset_clear = gr.Button("🗑️ Clear", variant="secondary")
         gr.Markdown("**Popular Datasets:**")
         with gr.Row():
-            gr.Button("SQuAD", size="sm").click(lambda: set_dataset_name("squad"), outputs=dataset_name)
-            gr.Button("IMDB", size="sm").click(lambda: set_dataset_name("imdb"), outputs=dataset_name)
-            gr.Button("GLUE", size="sm").click(lambda: set_dataset_name("glue"), outputs=dataset_name)
-            gr.Button("Common Voice", size="sm").click(lambda: set_dataset_name("common_voice"), outputs=dataset_name)
+            gr.Button("SQuAD", size="sm").click(functools.partial(set_textbox_value, "squad"), outputs=dataset_name)
+            gr.Button("IMDB", size="sm").click(functools.partial(set_textbox_value, "imdb"), outputs=dataset_name)
+            gr.Button("GLUE", size="sm").click(functools.partial(set_textbox_value, "glue"), outputs=dataset_name)
+            gr.Button("Common Voice", size="sm").click(functools.partial(set_textbox_value, "common_voice"), outputs=dataset_name)
         
         dataset_btn.click(run_dataset_info, inputs=dataset_name, outputs=dataset_output)
         dataset_clear.click(clear_output, outputs=dataset_output)
@@ -327,10 +605,10 @@ with gr.Blocks(
             search_clear = gr.Button("🗑️ Clear", variant="secondary")
         gr.Markdown("**Popular Tasks:**")
         with gr.Row():
-            gr.Button("Text Classification", size="sm").click(lambda: set_search_task("text-classification"), outputs=search_task)
-            gr.Button("Question Answering", size="sm").click(lambda: set_search_task("question-answering"), outputs=search_task)
-            gr.Button("Text Generation", size="sm").click(lambda: set_search_task("text-generation"), outputs=search_task)
-            gr.Button("Image Classification", size="sm").click(lambda: set_search_task("image-classification"), outputs=search_task)
+            gr.Button("Text Classification", size="sm").click(functools.partial(set_textbox_value, "text-classification"), outputs=search_task)
+            gr.Button("Question Answering", size="sm").click(functools.partial(set_textbox_value, "question-answering"), outputs=search_task)
+            gr.Button("Text Generation", size="sm").click(functools.partial(set_textbox_value, "text-generation"), outputs=search_task)
+            gr.Button("Image Classification", size="sm").click(functools.partial(set_textbox_value, "image-classification"), outputs=search_task)
             
         search_btn.click(run_model_search, inputs=[search_task, search_limit], outputs=search_output)
         search_clear.click(clear_output, outputs=search_output)
@@ -344,10 +622,10 @@ with gr.Blocks(
             transformers_clear = gr.Button("🗑️ Clear", variant="secondary")
         gr.Markdown("**Core Topics:**")
         with gr.Row():
-            gr.Button("Pipeline", size="sm").click(lambda: set_transformer_topic("pipeline"), outputs=transformers_topic)
-            gr.Button("Tokenizer", size="sm").click(lambda: set_transformer_topic("tokenizer"), outputs=transformers_topic)
-            gr.Button("Trainer", size="sm").click(lambda: set_transformer_topic("trainer"), outputs=transformers_topic)
-            gr.Button("Generation", size="sm").click(lambda: set_transformer_topic("generation"), outputs=transformers_topic)
+            gr.Button("Pipeline", size="sm").click(functools.partial(set_textbox_value, "pipeline"), outputs=transformers_topic)
+            gr.Button("Tokenizer", size="sm").click(functools.partial(set_textbox_value, "tokenizer"), outputs=transformers_topic)
+            gr.Button("Trainer", size="sm").click(functools.partial(set_textbox_value, "trainer"), outputs=transformers_topic)
+            gr.Button("Generation", size="sm").click(functools.partial(set_textbox_value, "generation"), outputs=transformers_topic)
         
         transformers_btn.click(run_transformers_docs, inputs=transformers_topic, outputs=transformers_output)
         transformers_clear.click(clear_output, outputs=transformers_output)