Amkyw-Myanmar-LLMV1

A fine-tuned large language model for Myanmar (Burmese) language processing and generation.

📋 Overview

Amkyw-Myanmar-LLMV1 is a state-of-the-art language model fine-tuned for Myanmar language tasks, including:

• Text generation in Myanmar script
• Conversational AI for Myanmar speakers
• Myanmar language understanding and processing

This model is fine-tuned from a base LLM (Llama, Gemma, etc.) using the Myanmar LLM Dataset from GitHub.

🏗️ Architecture

• Base Model: [To be specified - e.g., meta-llama/Llama-2-7b-hf]
• Fine-tuning Method: LoRA (Low-Rank Adaptation)
• Tokenization: Custom Myanmar tokenizer

📦 Model Files

model/
├── config.json               # Model architecture configuration
├── tokenizer.json            # Custom tokenizer
├── tokenizer_config.json    # Tokenizer settings
├── vocab.json               # Vocabulary file
├── model.safetensors        # Model weights (after training)
├── adapter_config.json      # LoRA adapter configuration
└── adapter_model.safetensors # LoRA adapter weights (after training)

🚀 Quick Start

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "amkyawdev/Amkyw-Myanmar-LLMV1"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)

prompt = "မြန်မာစာနဲ့ စာရေးပါစိုးပါနော်"
inputs = tokenizer(prompt, return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=100)
print(tokenizer.decode(outputs[0]))

🧪 Training

Dataset

This model was fine-tuned on the Myanmar LLM Dataset:

from datasets import load_dataset

dataset = load_dataset(
    "json",
    data_files={
        "train": "https://raw.githubusercontent.com/amkyawdev/myanmar-llm-dataset/main/data/processed/train.jsonl",
        "validation": "https://raw.githubusercontent.com/amkyawdev/myanmar-llm-dataset/main/data/processed/validation.jsonl",
        "test": "https://raw.githubusercontent.com/amkyawdev/myanmar-llm-dataset/main/data/processed/test.jsonl"
    }
)

📁 Project Structure

Amkyw-Myanmar-LLMV1/
├── README.md
├── LICENSE
├── .gitignore
├── .gitattributes
├── model/
├── training/
├── inference/
├── tests/
├── scripts/
└── .github/workflows/

📝 License

MIT License

---

This model was created by amkyawdev