Update model card with detailed documentation

README.md

@@ -6,61 +6,148 @@ tags:
 - llm
 - code-generation
 - fine-tuned
+- lora
+- phi-2
+datasets:
+- amkyawdev/myanmar-llm-data
 ---
 
 # mm-llm-coder-lite-v1
 
-Myanmar LLM model for code generation and conversational tasks, fine-tuned from microsoft/phi-2.
+Myanmar Lightweight LLM for Code Generation and Conversation
 
-## Model Description
+## 📌 Overview
 
-This model is fine-tuned on the myanmar-llm-data dataset for Burmese (Myanmar) language understanding and code generation tasks.
+This is a fine-tuned lightweight LLM model for Myanmar (Burmese) language understanding, code generation, and conversational tasks. The model is based on [microsoft/phi-2](https://huggingface.co/microsoft/phi-2) and fine-tuned using LoRA technique.
+
+## 🏗️ Architecture
 
 - **Base Model**: microsoft/phi-2
+- **Fine-tuning Method**: LoRA (Low-Rank Adaptation)
+- **Training Framework**: Hugging Face Transformers + PEFT + TRL
 - **Language**: Burmese (Myanmar)
-- **Training Data**: amkyawdev/myanmar-llm-data
-- **License**: MIT
 
-## Training Details
+## 📊 Training Details
+
+| Parameter | Value |
+|-----------|-------|
+| Base Model | microsoft/phi-2 |
+| Training Epochs | 3 |
+| Learning Rate | 2e-4 |
+| LoRA Rank (r) | 16 |
+| LoRA Alpha | 32 |
+| LoRA Dropout | 0.05 |
+| Max Length | 512 |
+| Batch Size | 4 |
+| Gradient Accumulation | 4 |
+
+## 📁 Dataset
+
+Trained on [amkyawdev/myanmar-llm-data](https://huggingface.co/datasets/amkyawdev/myanmar-llm-data):
 
-- **Framework**: Transformers + PEFT (LoRA)
-- **Training Epochs**: 3
-- **Learning Rate**: 2e-4
-- **LoRA R**: 16
-- **LoRA Alpha**: 32
+| Tag | Description |
+|-----|-------------|
+| coding | Programming conversations (90%) |
+| translation | English-Myanmar translation (1%) |
+| general | General knowledge Q&A (1%) |
+| greeting | Burmese greetings (1%) |
 
-## Dataset
+### Dataset Statistics
+- Train: ~20,327 samples
+- Test: ~17,155 samples  
+- Validation: ~17,071 samples
 
-The model was trained on [amkyawdev/myanmar-llm-data](https://huggingface.co/datasets/amkyawdev/myanmar-llm-data) dataset which contains:
-- Coding (90%)
-- Translation (1%)
-- General (1%)
-- Greeting (1%)
+## 🚀 Quick Start
 
-## Usage
+### Installation
+
+```bash
+pip install -r requirements.txt
+```
+
+### Fine-tuning
+
+```bash
+python finetune_mm_llm.py
+```
+
+### Inference
 
 ```python
 from transformers import AutoModelForCausalLM, AutoTokenizer
 
+# Load model
 model_name = "amkyawdev/mm-llm-coder-lite-v1"
 tokenizer = AutoTokenizer.from_pretrained(model_name)
 model = AutoModelForCausalLM.from_pretrained(model_name)
 
 # Generate response
-input_text = "ပရိုဂရမ်ရေးရတာကို နှစ်သက်ပါတယ်"
+input_text = "System: သင်သည် မြန်မာစာကျွမ်းကျင်သော AI အကူအညီပေးသူဖြစ်သည်။\n\nUser: မင်္ဂလာပါ။"
+
 inputs = tokenizer(input_text, return_tensors="pt")
 outputs = model.generate(**inputs, max_new_tokens=100)
-print(tokenizer.decode(outputs[0]))
+response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+print(response)
 ```
 
-## Requirements
+### Using with Transformers Pipeline
+
+```python
+from transformers import pipeline
+
+pipe = pipeline(
+    "text-generation",
+    model="amkyawdev/mm-llm-coder-lite-v1",
+    tokenizer="amkyawdev/mm-llm-coder-lite-v1"
+)
+
+result = pipe("User: ဟိုင်း၊ နေကောင်းလား။")
+print(result[0]['generated_text'])
+```
+
+## 📋 Requirements
 
 ```
 torch>=2.0.0
 transformers>=4.35.0
 peft>=0.7.0
+trl>=0.7.0
+accelerate>=0.25.0
+datasets>=2.14.0
 ```
 
----
+## 🔧 Configuration
+
+Edit `Config` class in `finetune_mm_llm.py` to customize:
+
+```python
+class Config:
+    MODEL_NAME = "microsoft/phi-2"  # Change base model
+    OUTPUT_DIR = "./mm-llm-coder-lite-v1"
+    TRAIN_PATH = "/workspace/train.jsonl"
+    # ... more settings
+```
+
+## 📝 Output
+
+After training, the model will be saved to `./mm-llm-coder-lite-v1/` with:
+- `adapter_config.json` - LoRA config
+- `adapter_model.safetensors` - LoRA weights
+- `tokenizer.json` - Tokenizer
+- `tokenizer_config.json` - Tokenizer config
+- `training_config.json` - Training config
+
+## 🏷️ Tags
+
+- `burmese` - Myanmar language
+- `llm` - Large Language Model
+- `code-generation` - Code generation
+- `fine-tuned` - Fine-tuned model
+
+## 📜 License
+
+MIT License
+
+## 👤 Author
 
-Built by [amkyawdev](https://huggingface.co/amkyawdev)
+[amkyawdev](https://huggingface.co/amkyawdev)