https://github.com/kratugautam99/logiclink-project

LogicLink is a conversational AI chatbot developed by Kratu Gautam (AIML Engineer). Powered by the TinyLlama-1.1B-Chat-v1.0 model, it provides an interactive interface for engaging conversations, query resolution, and task assistance. Version 5 features streaming responses, conversation management, and a sleek GUI.
https://github.com/kratugautam99/logiclink-project

antd-design chatbot-application conversational-ai cuda gradio graphical-user-interface huggingface-spaces huggingface-transformers jupyter-notebooks keras large-language-models mlops model-service-controller modelscope-studio natural-language-generation natural-language-processing pytorch reasoning-agent tensorflow

Last synced: 2 months ago
JSON representation

Host: GitHub
URL: https://github.com/kratugautam99/logiclink-project
Owner: Kratugautam99
Created: 2025-05-02T18:10:12.000Z (about 1 year ago)
Default Branch: main
Last Pushed: 2025-06-27T01:19:17.000Z (12 months ago)
Last Synced: 2025-06-27T02:21:10.065Z (12 months ago)
Topics: antd-design, chatbot-application, conversational-ai, cuda, gradio, graphical-user-interface, huggingface-spaces, huggingface-transformers, jupyter-notebooks, keras, large-language-models, mlops, model-service-controller, modelscope-studio, natural-language-generation, natural-language-processing, pytorch, reasoning-agent, tensorflow
Language: Jupyter Notebook
Homepage: https://huggingface.co/spaces/KraTUZen/LogicLink-Project-Space
Size: 37.9 MB
Stars: 0
Watchers: 1
Forks: 0
Open Issues: 0
Metadata Files:
- Readme: README.md

Awesome Lists containing this project

README

          # 💬 LogicLink: Version 5

**LogicLink** is a conversational AI chatbot developed by **Kratu Gautam** (AIML Engineer). Powered by the **TinyLlama-1.1B-Chat-v1.0** model, it provides an interactive interface for engaging conversations, query resolution, and task assistance. Version 5 features streaming responses, conversation management, and a sleek GUI.



  



---

## 🔍 Topic Index

- [✨ Key Features](#-key-features)

- [📸 GUI Display](#-gui-display)

- [🛠️ Installation](#-installation)

  - [Prerequisites](#prerequisites)

  - [Setup](#setup)

  - [Directory Structure](#directory-structure)

- [💬 Usage](#-usage)

- [⚙️ Technical Architecture](#-technical-architecture)

  - [Model Configuration](#model-configuration)

  - [Key Components](#key-components)

- [🧪 Troubleshooting Guide](#-troubleshooting-guide)

- [🚀 Future Roadmap](#-future-roadmap)

- [📜 License](#-license)

--- 

## ✨ Key Features

| Feature | Description | Benefit |

|---------|-------------|---------|

| **🤖 Conversational AI** | TinyLlama-1.1B-Chat-v1.0 powered responses | Natural, engaging dialogue |

| **⚡ Streaming Responses** | Real-time token generation with `TextIteratorStreamer` | Smooth user experience |

| **🎨 Customizable GUI** | Red/blue/black theme with Gradio & ModelScope Studio | Professional interface |

| **🗂️ Conversation Management** | New chat, clear history, delete conversations | Full control over interactions |

| **⏱️ Single Time Stamp** | Regex-cleaned response timing `*(4.50s)*` | Consistent performance metrics |

| **🚀 CUDA Support** | Automatic GPU detection with CPU fallback | Optimized performance |

| **🛡️ Error Handling** | Graceful failure for memory/input issues | Robust user experience |

---

## 📸 GUI Display

---

### 💬 Full-Fledged Conversation



  



LogicLink engaging in a complete dialogue, handling multiple turns seamlessly.  

This demonstrates its ability to maintain context, respond naturally, and adapt to user intent across an extended session.

---

### 🧑‍💻 Coding Response (Part 1)



  



LogicLink generating a structured coding solution.  

Notice how it explains the reasoning step-by-step, making the output not just correct but also **educational**.

---

### 🧑‍💻 Coding Response (Part 2)



  



A continuation of the coding workflow, where LogicLink refines and expands on its earlier solution.  

This shows its iterative reasoning ability — improving code quality when prompted.

---

### 🔑 Core Response



  



A snapshot of LogicLink delivering a **core logical explanation**.  

This highlights its strength in breaking down abstract queries into clear, actionable insights.

---

### ⚡ While Processing



  



The system mid‑inference, showing its **real-time feedback loop**.  

This reassures users that LogicLink is actively working on their request.

---

### 🔄 With vs Without Latest Output Text Box



  

  



A side‑by‑side comparison of LogicLink’s performance **with** and **without LOTB (Latest Output Text Box)**.  

The difference illustrates how LOTB enhances reasoning depth and response clarity.

---

### 📊 Bottom Section



  



The footer view of the interface, where conversation summaries and quick actions are displayed.  

This ties the user experience together, making LogicLink feel like a polished, end‑to‑end assistant.

---



## 🛠️ Installation

### Prerequisites

- Python 3.8+

- CUDA-enabled GPU (recommended)

- Dependencies:

  ```bash

  pip install gradio torch transformers modelscope-studio

  ```

  

---

### Setup

1. Clone repository:

   ```bash

   git clone https://github.com/Kratugautam99/LogicLink-Project.git

   cd LogicLink-Project

   ```

2. Install dependencies:

   ```bash

   pip install -r requirements.txt

   ```

3. Run application:

   ```bash

   python app.py

   ```

---

### Directory Structure

```

LogicLink-Project/

├── LogicLinkVersion5.ipynb

├── README.md

├── app.py

├── config.py

├── .gitattributes

├── requirements.txt

├── assets/

├── Documents/

├── Screenshots/

├── ui_components/

└── Different Versions of LogicLink/  (not expanded)

```

---

## 💬 Usage

```python

# Sample interaction flow

user >> "Who are you?"

LogicLink >> "I'm LogicLink V5, created by Kratu Gautam. How can I assist you today? *(4.50s)*"

```

1. **Interface Controls**:

   - 💬 Input field: Type queries

   - ➕ New Chat: Start fresh conversation

   - 🧹 Clear History: Reset current chat

   - 🗑️ Delete: Remove conversations from sidebar

2. **Performance Metrics**:

   - ⏱️ Response time: 3-5s (GPU), 5-8s (CPU)

   - 💾 RAM usage: 2-3GB (CPU), ~1.5GB (GPU)

---





  

    

  

      Technical Architecture



### Model Configuration

```python

# Core model parameters

model = AutoModelForCausalLM.from_pretrained(

    "TinyLlama/TinyLlama-1.1B-Chat-v1.0",

    torch_dtype=torch.float16 if cuda else torch.float32

)

# Generation settings

generation_kwargs = {

    "max_new_tokens": 1024,

    "temperature": 0.7,

    "top_k": 50,

    "top_p": 0.95,

    "num_beams": 1

}

```

---

### Key Components

1. **Prompt Engineering**:

   ```

   <|system|>You are LogicLink V5 created by Kratu Gautam

   <|user|>{user_input}

   <|assistant|>

   ```

   

2. **Streaming Pipeline**:

   ```mermaid

   graph LR

   A[User Input] --> B(Tokenizer)

   B --> C{TextIteratorStreamer}

   C --> D[Model Generation]

   D --> E[Real-time Output]

   E --> F[Regex Cleaner]

   F --> G[Timestamp Append]

   ```

3. **GUI Components**:

   - `pro.Chatbot`: Conversation display

   - `antdx.Sender`: Input field

   - `antdx.Conversations`: Sidebar manager

   - `antd.Button`: Action controls

---

## 🧪 Troubleshooting Guide

| Issue | Solution |

|-------|----------|

| Double timestamps | Verify regex: `re.sub(r'\*\(\d+\.\d+s\)\*', '', response)` |

| Slow responses | Enable CUDA, reduce `max_new_tokens` to 512 |

| GUI rendering issues | Update packages: `pip install --upgrade gradio modelscope-studio` |

| Delete button failure | Check `menu_click` event binding in JS |

| Model loading errors | Validate RAM ≥3GB, test with minimal example |

**Minimal Test Script**:

```python

from transformers import AutoModelForCausalLM, AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained("TinyLlama/TinyLlama-1.1B-Chat-v1.0")

model = AutoModelForCausalLM.from_pretrained("TinyLlama/TinyLlama-1.1B-Chat-v1.0")

inputs = tokenizer(["Test input"], return_tensors="pt")

outputs = model.generate(**inputs, max_new_tokens=10)

print(tokenizer.decode(outputs[0]))

```

---

## 🚀 Future Roadmap

- **Persistent Storage**: SQLite conversation history

- **Multimodal Support**: Image/text inputs

- **Enhanced Prompting**: Context-aware responses

- **Deployment Options**: Docker containerization

- **Performance**: Quantization for CPU optimization

---

## 📜 License

MIT License - See [LICENSE](https://github.com/Kratugautam99/LogicLink-Project/LICENSE)

---



  Developed with 🧠 by Kratu Gautam | AIML Engineer


  GitHub | 

  HFT Space | 

  UI Framework

ecosyste.ms

Data

Tools

Indexes

Applications

Experiments

Awesome

https://github.com/kratugautam99/logiclink-project

Awesome Lists containing this project

README

Technical Architecture